The full text on this page is automatically extracted from the file linked above and may contain errors and inconsistencies.
Volume 2 REAPPRAISAL OF THE FEDERAL RESERVE DISCOUNT MECHANISM Board of Governors of the Federal Reserve System PU BLISH ED IN D EC EM B ER 1971 Library of Congress Catalogue Card Number 70-609110 Copies of this report m ay be obtained from Publica tions Services, Division of Adm inistrative Services, Board of G overnors of the Federal Reserve System, W ashington, D.C., 20551. The price is $3.00 per copy; in quantities of 10 or m ore sent to one address, $2.50 each. Remittances should be m ade payable to the Board of G overnors of the Federal Reserve System in a form collectible at par in U.S. currency. (Stamps and coupons not accepted.) Volume 2 REAPPRAISAL OF THE FEDERAL RESERVE DISCOUNT MECHANISM CONTENTS THE LEGITIMACY OF CENTRAL BANKS 1 Kenneth E. Boulding SELECTIVE CREDIT CONTROL 15 Lester V. Chandler A REVIEW OF RECENT ACADEMIC LITERATURE ON THE DISCOUNT MECHANISM 23 David M. Jones SUMMARY OF ISSUES RAISED AT THE ACADEMIC SEMINAR ON DISCOUNTING 47 Priscilla Ormsby SOME PROPOSALS FOR A REFORM OF THE DISCOUNT WINDOW Franco Modigliani 59 CONTENTS cont. AN EVALUATION OF SOME DETERMINANTS OF MEMBER BANK BORROWING 77 Leslie M. Alperstein TOWARD A SEASONAL BORROWING PRIVILEGE: A STUDY OF INTRAYEAR FUND FLOWS AT COMMERCIAL BANKS 93 Emanuel Melichar CAPITAL AND CREDIT REQUIREMENTS OF AGRICULTURE, AND PROPOSALS TO INCREASE AVAILABILITY OF BANK CREDIT Emanuel Melichar and Raymond J. Doll 107 THE LEGITIMACY OF CENTRAL BANKS Kenneth E. Boulding University of Colorado THE LEGITIMACY OF CENTRAL BANKS therefore, depend on the willingness of other role occupants to give outputs, and they will not do this continuously unless there is legitimacy. Where people feel that certain outputs are illegitimate they will eventually find ways of stopping them. The corresponding inputs will likewise stop. To use a rather crude illustration, a bandit can take your money once, but anyone who wants to take it every week either has to be a landlord or a tax collector, or perhaps even a bank. There are a considerable number of sources of legitimacy, and the functions that relate the determinants of legitimacy to its amount are extremely complex. Such func tions are certainly nonlinear and they ex hibit discontinuities that are, to say the least, disconcerting. Sometimes an institution, the legitimacy of which seems to be absolutely unquestioned, collapses overnight. All of a sudden we reach some kind of a “cliff” in the legitimacy function and the institution suddenly becomes illegitimate. The same thing perhaps can even happen the other way, in which institutions quite suddenly become legitimate after having been illegit imate. A good example of the former is the collapse of the monarchy, beginning in the 17th century. The legitimacy of monarchy survived the Cromwellian war in England, largely because an ancient legitimacy is like a capital stock—it takes a great deal of The problem of legitimacy is one of the most neglected aspects of the study of so cial systems. There may be good reasons for this because legitimacy is inevitably a hot subject. One can hardly discuss the legitimacy of anything without seeming to threaten it, for a great deal of legitimacy depends on things being taken for granted and not being talked about at all. The more one looks at the dynamics of social systems, however, the more it becomes clear that the dynamics of legitimacy is one of the most important elements in the total long-run dynamics of society. It certainly ranks with such things as population and demographic movements, and even with technological change with which it is closely intertwined. Its importance can be seen in the remark that a person or institution that loses legiti macy loses everything and can no longer maintain itself in the social system. No amount of wealth—that is, exchange capa bility—or power—that is, threat capability —can keep an institution alive if there is a widespread denial of the legitimacy of its role in society. This is because the perform ance of any continuous and repeated role requires an acceptance of its legitimacy on the part of those role occupants whose roles are related to it. A role in the social system is a focal point or node of inputs and out puts of many different kinds, the output of one role being the input of another. Inputs, 3 4 spending before it can be exhausted. At the time of Louis XIV in the following cen tury one might have thought the legitimacy of monarchy was absolutely unquestioned and secure. In the 19th century, however, monarchies collapsed nearly everywhere and the only monarchs who survived were those who abandoned their power and became symbols of legitimacy, like the British, Dutch, and Scandinavian monarchs. On the other side, abortion has been an institution that has been regarded as highly illegit imate and now in the face of the popula tion problem seems to be acquiring a sud den legitimacy. We may distinguish at least six classes of sources of legitimacy; that is, variables in society that are functionally related to legitimacy. The first consists of the payoffs of the institution in question. If an institu tion provides good terms of trade with those who are related to it, this contributes to its legitimacy up to a point, especially in the long run. The case is clearer on the negative side. An institution that has very poor payoffs—that is, demands a great deal of input from other people and gives very little output to them—is likely to have its legitimacy eventually eroded on this ac count. The relationship, however, is cer tainly nonlinear and quite complex, and at times may even be negative. Just because an institution is useful and pays off well is not sufficient to give it legitimacy. Paradoxically enough, it is not merely good payoffs that give legitimacy but also bad payoffs; that is, sacrifices—the second source. A sacrifice or “grant” may be de fined as a one-way transfer from one decision unit to another, in contrast with exchange, which is a two-way transfer, from A to B and also from B to A. The structure of one-way transfers of commodi ties and exchangeables, I call the “grants” economy, and it is a good first-approxima tion measure of the extent and structure of the integrative system in general. If A makes a grant to B, the implication is that A identifies with B, A and B are in a com munity together, and A clearly regards B as legitimate. The dynamics of the grants system is very complex because to some extent grants are self-justifying. If A makes sacrifices for B, it is very hard for A to admit to himself that these sacrifices have been in vain. This would be a threat to his identity, which is the greatest threat that any person can feel. There is, there fore, a strong tendency to “throw good money after bad” and to continue making sacrifices for some institution, even after some possibly expected long-run payoffs have failed to materialize. This is what I call the “sacrifice trap.” We see this in the family, for instance, where the devotion of one spouse to an unsatisfactory partner often continues for a long time in spite of very unsatisfactory internal terms of trade. A spouse who gives a lot to a marriage and gets very little out of it may continue to do so because of the threat to the personal identity should the process ever stop. There may come a point, of course, at which the terms of trade become too bad altogether and a break-up ensues. This is the “cliff” phenomenon in the legitimacy function. The same thing evidently happened to the mon archy, and it can happen to religion, like the religion of the Aztecs. It could even happen to the national state. The third source of legitimacy is age. Institutions build up legitimacy just by sticking around, as long as there is an ex cess of production over consumption. Even this function, however, may be nonlinear. Up to a point increase in age increases legit imacy; beyond a certain point, however, the senator becomes senile and the good old things become old-fashioned. One can LE G IT IM A C Y OF C EN T R A L B A N K S detect, perhaps, three phases of the func tion. When things are new, they have the special legitimacy of babies, young people, or the new fashion. At a certain point they become middle-aged or old-fashioned and legitimacy declines sharply. Then as time goes on, they become antiques and legiti macy increases once again. In the case of a creative person, for instance, one often finds a phase of rising legitimacy with age and then a declining phase as he gets out of date, and then an increasing phase as he acquires a posthumous reputation, which is presum ably the personal equivalent of being an antique. The fourth source of legitimacy is mys tery. Something that is not understood but that is dimly perceived as obscurely grand and magnificent acquires an aura of legiti macy simply because it is secret and we do not understand it. The temples and impres sive ceremonies of religion, the state of kings, the mystique of the brass hat and the military leader, the sanctity of priest hoods of all kinds, and even the mystery of science and the laboratory are all related to this aspect of legitimacy. It depends, of course, on a class structure, on a distinc tion between the initiates and the common people. Historically, it has been a very powerful source of the willingness of the common people to make sacrifices for the benefit of the initiates and to afford them a great deal of legitimacy, often in the absence of much in the way of tangible returns. Closely related to this aspect of legiti mation is ritual or artificial order. Man has always feared the randomness of his en vironment, the uncertainty of the weather, the crops, accidental injury or death, dis ease, his whole future state. One of his responses to this has been to create little islands of artificial order, regularly repeated 5 rituals, liturgies, and human law. The role that law plays in legitimation is closely re lated to this aspect of it as ritual. To say that law and ritual are artificial orders is not in any sense to deny them validity, nor does it mean that these artificial orders are arbi trary. Where they are successful it is pre cisely because they reflect an order in the real world, whatever that is. Nonetheless, they are artificial in the sense that they create an island, as it were, of life and ex perience that is separated from the rest of the world. A monastery is a good example of such artificial order; so is a law court. Insofar as the need for legitimation is closely related to the need for regularity and for law in the broad sense of regularity and nonrandomness, we can easily see why the development of these artificial orders of liturgy and legal procedure, of due proccess, and of repeatable and predictable be haviors and decisions are an important as pect in the legitimation process. Here, too, however, we may run into nonlinear rela tionships. Beyond a certain point an arti ficial order becomes too artificial, and if protest arises against it, the legitimacy of the institution that is based on it may sud denly collapse. The Reformation, perhaps, may be interpreted as a protest against too artificial an order in the Roman Catholic Church. The fact that law does not always maintain legitimacy, as the experience of prohibition indicated, also suggests that law too may be “a hass” in the memorable words of some unmemorable character in Dickens, and when it is perceived to be such, the legitimacy on which it is based easily collapses. There are many countries today, indeed, in which law is much less legitimate than it is in the United States, and the legitimacy of law itself is a prob lem to which we have given far too little attention. The sixth source of legitimacy consists of the alliance of an institution with other legitimacies. This is what might be called the legitimacy syndrome. If there are in stitutions that already possess a great deal of legitimacy, it is possible sometimes for new and nonlegitimate institutions to ac quire legitimacy by identifying themselves with the legitimate ones. It is easy to cite examples of this. The United States built Washington in the classical tradition of ancient times. The United States, being a new and therefore rather illegitimate Re public, sought to establish its legitimacy by means of a “tie-in” with Corinthian col umns and handsome domes. The legitimacy of a religion often permits highly radical and otherwise illegitimate movements to spring up within it, like the Franciscans, or, in our day, the movement for racial equal ity or even the peace movement. Here again we may run into nonlinearities in the rela tionship. The nouveau riche person who builds a very fancy house may thereby diminish rather than enhance his legitimacy in the eyes of those he most wishes to im press. A country that wastes its scarce re sources on building a vast presidential pal ace or a grand new capital may not acquire much legitimacy thereby, but only the sub tle sneers reserved for unwise decision makers. One interesting phenomenon here is that the more legitimacy an institution has, the less it has to worry about these alliances. In the early days of a university, for instance, it often builds elaborate Gothic or classical buildings to tie in with the legitimacy of the past and to pretend that it has the legitimacy of spurious age. As it acquires genuine legitimacy, how ever, perhaps in the process of providing payoffs, its buildings become skimpier and more austere and it puts less and less into ritual and into elaborate architecture until finally it ends up by abandoning gowns, Gothic buildings, ivy, and even grass as it lays down its campus to enormous parking lots. Let us now apply this analysis as far as we can to the problem of the legitimacy of the banking system, and of the central banks in particular. The existence of social ist states shows that this is not an idle prob lem. Socialism indeed can be interpreted largely as an attack on the legitimacy of certain institutions of exchange, and in the socialist states we see the very interesting phenomenon of the gradual re-establish ment of many of these same institutions with a different framework of legitimation. In the Western World and especially in the United States, the legitimacy of the bank ing system is almost completely taken for granted. It must not be assumed, however, that the banking system or any other insti tution necessarily creates its own legit imacy, and it must not be assumed that this legitimacy could never disappear, even though it might seem at the moment to be quite unshakable. The possessors of un shakable legitimacy should always remem ber Louis XVI at least once a day, even though the Federal Reserve is not the sort of place where heads are likely to roll. It will at least be an interesting exercise, there fore, to apply the six major sources of legit imacy to the banking system, and see if any dynamic patterns emerge. The payoffs of the banking system to the rest of society are fairly positive and also are fairly visible. Most people outside the banking system have contact with it either through using a checking account, which is clearly a great convenience and for which the payment does not seem ex orbitant, or through borrowing money, which again we would not do unless we thought that the returns were likely to be LE G IT IM A C Y OF C E N T R A L B A N K S greater than the costs. The banking system is perhaps the purest example of an ex change institution. It lives almost entirely by exchange, it does very little physical transformation, and the utilities that it creates out of which payoffs to the various parties come are essentially exchange utili ties, such as the creation of convenient forms of exchangeables like checking ac counts, or the separation of ownership from control and the placing of asset complexes in the control of those who presumably know how to manage them best. The legit imacy of banking, therefore, falls or rises with the legitimacy of exchange itself. Even though the payoffs to the banking system for those who deal with it are clearly positive—for it is an essential characteristic of exchange systems that continued ex change would not take place unless there are positive payoffs to all parties—this in itself is not sufficient to give legitimacy, al though it helps. The somewhat loose rela tionship between payoffs and legitimacy may happen for two reasons. The first, which applies to all exchange institutions, is that an exchange, perhaps because it in volves so little in the way of sacrifice, does not generate strong integrative sentiments and feelings. My own bank once advertised as “the bank that puts people first.” Every body knows, however, that this is a ritualis tic remark designed solely to create favor able sentiments. If, indeed, I thought it true, I probably would not bank there, for what we really want in a bank is that it puts money first. In other words, we want ex treme probity in accounting, with not a cent out of place, and if this involves some sacrifice of a charming but careless accoun tant or a benevolent embezzler, I doubt very much if we would fight for putting people first. There have been a number of cases, indeed, of benevolent bank officers 7 who embezzled in order to do good, and this act is usually frowned upon quite severely. I, at least, want banks to be honest, im peccable, and full of rectitude. I do not necessarily want them to be lovable, in spite of some of their advertising. Nevertheless, this absence of lovability in exchange insti tutions not only seems to worry banks, but it may occasionally lead to their over throw. Schumpeter, we may recall, argued that capitalism would be overthrown by its very success and because the rationalistic attitude that it generated would destroy the integrative institutions in, say, the family or the church, or even the state, which en able exchange to be legitimated. Exchange and exchange institutions, in other words, simply pay off too well. They do not de mand any sacrifice. Thus an institution that bases its legitimacy on its payoffs may be challenged by another institution that claims to have even better payoffs. This is one reason, perhaps, why legitimacy that is based merely on payoffs is a little insecure, whereas a legitimacy that is based on sacrifice is remarkably stable. It is at least an amusing fantasy to sup pose that we might do a cost-benefit analy sis of the financial system and, indeed, of competing financial systems. The costs are fairly easy to identify. We could, for in stance, do a comparative study of, say, Austria and Hungary, two countries at about the same level of development, one of which has a predominantly marketbased financial system and the other being a socialist state. We could find out fairly easily the costs of the two systems in terms of resources absorbed into them based on the economy in general. We could find out, for instance, what proportion of the gross na tional product in each case was absorbed by the financial system. The benefits, of course, would be much harder to assess. 8 Indeed, I would almost despair of ever making a quantitative assessment of them. It is on judgments of this kind, however, that the long-run competition between so cialism and capitalism may ultimately be determined. Merely asking a question of this kind, however, may seem somewhat threatening to the legitimacy of either kind of institu tion. The legitimacy of the institutions of capitalism could depend a good deal on their simple age; that is, just on the fact that they are not questioned and that we have gotten along with them for a long time with reasonable success. It is one of the curious problems of the dynamics of legitimacy, indeed, that a threat to legiti macy is very hard to counter where the legit imacy itself is a function of age and ritual, for even an attempt to defend a legitimacy of this kind may destroy it. This perhaps is one reason why the Marxist threat to the legitimacy of capitalism was so much more dangerous than would be the case if the legitimacy depended merely on payoffs. The payoffs to capitalism are actually quite high. A good deal of its legitimacy, however, depends on institutions like pri vate property, the legitimacy of which had never really been questioned, and rests not on the perception of long-run payoffs at all, but simply on age, long use, and the ritual of law. The legitimacy of socialist institu tions likewise depends in good measure on the enormous sacrifices that have been made to create them. The socialist state stresses much more fiercely than the late President Kennedy: “Ask not what your country can do for you, ask only what you can do for your country.” Because it has demanded enormous sacrifices of its peo ple, in the interests of an ideal, its people do not like to admit that the ideal might not have much in the way of payoffs. Hence, the suggestion that the relative merits of the systems should be tested by cost-benefit analysis would probably be even more threatening to the socialist than it is to the capitalist. Let us now take a brief look at some of the other sources of legitimacy and see how they apply to the banking system. We have already noticed that banks are not institu tions that demand sacrifice, except perhaps sacrifice of temptations to dishonesty and extravagance. Banks, therefore, are not “heroic” institutions, and they cannot hope to inspire the kind of love and loyalty that such institutions as the church and the national state inspire. The banking system is, relatively speak ing, a fairly modem institution. It cannot perhaps draw a great deal of legitimacy from its age, although we do find banks and institutions of all kinds advertising the date of their foundation—when that is suit ably distant in time—as evidence of their integrity, respectability, and legitimacy. The Bank of England’s affectionate title as “The Old Lady of Threadneedle Street” indicates that age is perhaps not a negli gible factor. The sense of mystery and charisma is also far from a negligible factor in establishing the legitimacy of banks. The bank may not be a heroic institution, but it is certainly mysterious to the ordinary person. Most people who use banks, and indeed a good many people who operate them, really do not understand the operations of the bank ing system as a whole. There is, further more, a lingering sacred quality about money itself. There is something a little mysterious about the fact that mere green pieces of paper or, even more remarkable, a signature on a check is sufficient to buy tangible objects. In the past, at least, banks have contributed to the sense of mystery by LE G IT IM A C Y OF C E N T R A L B A N K S their very architecture, which has often tended to be quasi-religious. Even if banks shied away from the more subtle mysteries of the Gothic, they have frequently en shrined themselves in pagan temples and Corinthian columns, lofty ceilings, marble floors, and a general air of hushed magnifi cence that hopefully induces in the cus tomer the frame of mind of proper respect and reverence. Ritual, likewise, plays a not insignificant role in establishing the legitimacy of banks. Regular hours, standardized procedures, and a highly formalized accounting system contribute to a sense of regularity and or der. The banking system, furthermore, is strongly hedged about by legal safeguards and the ritualistic language of contracts. Alliances with other legitimacies are seen not only in the architecture but also in the institution of boards of directors—the mem bers of which are drawn from other respect able institutions in the community—and also in the institution of the charter granted by the state or by the nation, which brings along with it a certain apparatus of in spection and oversight. We could even re gard national deposit insurance, quite apart from its strictly economic aspects, as an alliance with the enormous legitimacy of the national state, for then behind even the most private of banks stands the majesty and legitimacy of government. We now come rather belatedly to what is supposed to be the main object of this paper, which is the problem of the legit imacy of central banks. Central banking is a rather late development in the banking system. Even in Great Britain the Bank of England did not begin to act as a central bank until well into the 19th century. The United States got along for the most part without any central bank until 1913, al though before that it had something that 9 might almost be called an informal central banking system. Until the establishment of the Federal Reserve System, the neces sity of central banking was still a matter of debate. The Japanese, for instance, when they began to introduce Western institu tions started with something like the Amer ican national banking system, and devel oped a central bank only after a number of financial crises. Today, however, the legit imacy, indeed almost the necessity, of cen tral banking seems unquestioned. Every new country sets up a central bank almost as soon as it is established. It is part of what every well-dressed country will wear. If we look down our six sources of legit imacy, we will see that almost everything that can be said of the banking system in general applies also to central banks. Here they have unquestionably risen in response to a felt need. There must, there fore, be some kind of a payoff to the or ganization. These, however, may be of two kinds: market payoffs and political pay offs. The fact that even under a free bank ing system some strategically located banks tended to perform the functions of a central bank—in that part of their deposits were owned by other banks and regarded as reserves—suggests that the function of cen tral banking is something that will develop even in a pure market system, simply be cause there are payoffs for this kind of organization; that is, central banking can provide adequate terms of trade for all those with whom it exchanges. There are clearly great conveniences, for instance, in the clearing function and in commercial banks holding their reserves in the form of deposits in some central banks—whether this clearing be public or private; the sheer dynamics of a free financial market would almost certainly throw up the institution of central banking in one form or another. 10 Without any exception, as far as I know, however, societies have not permitted cen tral banking to grow simply as a result of market forces, but have always intervened in the matter politically. At some point in the development of the system, those who are in control of the legislative process of society perceive certain payoffs in the de velopment of a government central bank that can then be used to control the pri vate banking system. In its political aspects the government central bank can then be seen as a partial movement toward the socialization of the banking system; such socialization leaves the ownership of most of the institutions of the system in private hands, but uses the government central bank as an instrument of control. This may be regarded for the most part as a problem in the legitimation of power. Because of the very structure of the system, a central bank, whether public or private, will have a great deal of power; that is, the decisions of its responsible decision-makers will have repercussions extending through the whole system of the society. Power, however, as we have seen, to be exercised continuously must be legitimated, and governmental in stitutions are the principal agency of legit imation in modern society. Private power will only be tolerated if it is small. This, indeed, is the theory behind the encourage ment of competition as a regulating factor, for in a competitive society the power ex ercised by any particular private decision maker is relatively small and is constantly checked by his competitors. Central bank ing, however, as in electric power or tele phones, has the great advantage of monop oly, which means a concentration of power, and if this concentration is to be legitimated, it must be regulated in some way through governmental organization. Hence, it is not surprising to find a strong tendency for government to take over the central banks, even though, as in the case of the Bank of England, nationalization may make practically no difference to its day-to-day operation or even its general policy. In this picture the Federal Reserve Sys tem presents some rather curious anomalies, which may, however, in the American con text be more apparent than real. The Fed eral Reserve System, like the Bank of Eng land before its nationalization, is theoret ically privately owned and is a series of interlocking corporations, theoretically owned and controlled in large measure by the member banks themselves. In reality, of course, the Federal Reserve Banks are public institutions, exercising the great power that they have, not to make profit for themselves, but to advance what they con ceive to be the public interest. Public rep resentatives sit on their boards of directors and the members of the Board of Governors of the Federal Reserve System are ap pointed by the President of the United States and confirmed by the Senate. The structure is thus less socialized than that of the Post Office and more socialized than the American Telephone and Telegraph Com pany, although there are certain parallels between the Board of Governors of the Federal Reserve System and a regulatory commission for public utilities. In the American system of legitimacy these apparent anomalies actually make a good deal of sense, for the American peo ple have a curious ambivalence towards government. On the one hand government is a strong source of legitimacy; on the other hand it is also regarded as something that is always potentially illegitimate and can get out of hand. Hence, government has to be hedged around with all sorts of constitutional safeguards. The American LE G IT IM A C Y OF C EN T R A L B A N K S Constitution can be interpreted in consider able measure as a kind of treaty between a people and its own government regarded as a potential enemy! Consequently, in the United States government does not have any monopoly of the legitimating process, and private institutions—simply because they are private—have a certain legitimacy of their own. It is not surprising, therefore, to find in the United States this curious mix of the public and the private that we find in the Federal Reserve System, and it can certainly be regarded, for its time, an op timum solution for the maximization of legitimacy. Today, certainly, there seems to be no major threat within the American system to the legitimacy of the Federal Re serve System, although there have been frequent and perhaps justified criticisms of its policies. As far as I know, there are no serious proposals either to nationalize the Federal Reserve Banks, or to put them under the U.S. Treasury, or to dissolve them and go back to a system of free bank ing. The principle of separation of powers is still very strong and the notion of the Treasury and the Federal Reserve System as two separate fiefs within a broad struc ture of governmental legitimation does not seem to be seriously threatened. Most of the other aspects of legitimacy that we noticed as being characteristic of the banking system also apply to the Fed eral Reserve System. Like the rest of the banking system, Federal Reserve Banks are not heroic institutions, although their association with the national state hangs over them a certain cloak of sacrificelegitimation, especially insofar as they may have to sacrifice their own ideals of finan cial probity in times of war. Bankers of all sorts tend to be deflationary- rather than inflationary-minded and it must hurt their souls a little to be accomplices in the infla 11 tionary financial policy that invariably ac companies a war. This sacrifice of finan cial honor, however, is small compared with the sacrifices of the soldier, although it may not be insignificant in contribut ing to the legitimacy of the institution. Certainly, if central banks were to oppose a war effort on the grounds that it offended their financial principles, their unwilling ness to sacrifice their principles would not be taken kindly and would contribute rap idly toward the loss of their legitimacy. Central banking is now old enough to have acquired a little of the sanctity of age, and it is certainly shrouded in a great deal of mystery and acquires a certain legit imacy from this fact as well. Where or dinary men and ordinary brokers have at least some familiarity with the operations of the member banks, they may have no famili arity at all with the operations of the cen tral bank. I must confess myself that I was an economist for 30 years, although not a specialist in money and banking, before I personally set foot within a central bank of any kind, and my knowledge of such banks and their operations, are derived wholly from books and talk. Even in the mind of a professional economist, therefore, the central banks appear as abstractions and cannot be visualized as flesh and blood realities. Whether the central banks should try to enlighten the public and to dispel the mystery is a nice point. It may well be that their own legitimacy is best fostered by pre serving a certain air of charismatic obscur ity about their operations. Their officers might even take to wearing gowns and robes and their public pronouncements might be couched in even more mysterious and im pressive language than they now use. The concept of a central bank as a creator of artificial order and financial ritual has some interpretive power and should not 12 be dismissed lightly. One of the real prob lems of central banking policy is that at the heart of it there is a certain arbitrariness. The movements of the bank rate, the deci sion to change the asset structures, the changes in legal reserve ratios, and other in struments of central bank control have a certain Delphic quality about them. They emerge as the result of arguments that are not disclosed, and yet they have very power ful effects on the total system. Furthermore, the effects of these decisions are not always easy to trace, and the feedbacks of informa tion are not easy to relate to particular decisions. Under these circumstances, the ritualizing of these decisions may be a very important aspect in their legitimation. One might even speculate on the value of ritual izing them more than is now the case. The decisions of a board, for instance, might be entrusted to a dramatically attired rider who would deliver them to the White House with the pounding of hooves and the flour ish of trumpets! We might conclude with a brief look at the possible threats to the legitimacy of the System. The fact that the System survived the Great Depression is a tribute to the re markable stock of legitimacy that it pos sesses. The extent to which the Federal Reserve System contributed to the Great Depression is still somewhat a matter of controversy. It certainly cannot be blamed for the whole episode; nonetheless, a strong argument can be made that in this period the payoffs of the System for the society as a whole were strongly negative and that disastrous mistakes in policy were made. In the short run, however, as we have noticed, the payoffs of the System are only loosely related to its legitimacy and the other sources of legitimacy for the Federal Re serve System are quite strong—strong enough, indeed, to enable it to survive a considerable decline in its payoffs to society. The only source of loss of legitimacy that seems even remotely on the horizon arises out of the sixth factor; that is, the alliances with other legitimacies. The Federal Re serve System is not allied at all with the legitimacies that derive from religion, from the family, from the arts, and from the more poetic, heroic, and evocative aspects of life. It is essentially and almost wholly an institution of exchange. Its inputs and outputs are exchangeables, and in itself exchange is too rational an institution to create much loyalty and affection and the kind of legitimacy that proceeds from these sources. I would argue indeed that an ex change institution should not try to derive legitimacy from these other sources, for if it does so it makes itself ridiculous. The Federal Reserve should certainly not try to become patron of the arts, an inspirer of heroism, or a producer of poetry. To at tempt to do so would be like tying peacock feathers on a work horse, and the ridiculous incongruities that would result would lessen rather than enhance the legitimacy of the institution. Insofar as the legitimacy of the central banks is enhanced by alliances, it is with the national state, and the national state alone. In these days the national state is so fantas tically legitimate an institution that to sup pose that its legitimacy might decline or even collapse seems almost absurd. Never theless, stranger things have happened. Par ticularly if the international system deterio rates much beyond its present deplorable condition, the payoffs of the international system for the human race will be so nega tive that the legitimacy of the national state as the essential and primary institution of the international system will itself be af fected. It may be, indeed, that before many decades are up, if we live that long, the LE G IT IM A C Y OF C EN T R A L B A N K S national state itself will have to be “desacralized.” This, indeed, is what general and complete disarmament and stable peace would involve. To put the matter brutally, some time in the future it may seem as absurd to die for one’s country as it would be today to die for the Federal Reserve. In the long run, therefore, we may see something very peculiar. The very common place and nonheroic aspects of the national state may save it, and the strong alliance that exists between central banks and gov ernments may turn out to be a two-way street. At the moment, indeed, it is govern ment that confers legitimacy on central banks to a considerable extent. It is not wholly inconceivable that in the future it will be the fact that the central bank is pri marily an agency for human welfare and not for human destruction that will confer legitimacy on the government, as we make the subtle transition from the warfare state, which threatens to engulf us all in a com mon destruction, to the desacralized com monplace, unheroic welfare state, which 13 works simply for human betterment. In the long run I have a good deal of confidence that payoffs in terms of human welfare are the only ultimate and self-sustaining sources of legitimacy. Sacrifice, age, mystery, and ritual can fool some of the people some of the time. If, however, they are not associ ated with real payoffs, they will be found out. This, of course, does not answer the question that we raised earlier as to whether there is not some other form of social or ganization that has still higher payoffs and lower costs than the existing banking struc ture. It would be rash indeed to argue that we have exhausted the potential of social invention in this regard. I am fairly cer tain, however, that whatever mutation may supplant the existing system has not yet been made, but if the legitimacy of the sys tem rests firmly on its payoffs then the so cial invention that will supplant it, if any, should be welcomed with joy rather than fear. It is only what I do not now mind calling the fraudulent legitimacies that fear competition. SELECTIVE CREDIT CONTROL Lester V. Chandler Princeton University Contents Federal Reserve Discount Policy a s an Instrum ent for Selective C redit C ontrol_____________ 17 O ther B ases for Selective Credit C ontrols_______________________________________________ 19 Som e Further O bservations on Discounting_____________________________________________ 20 15 SELECTIVE CREDIT CONTROL FEDERAL RESERVE DISCOUNT POLICY AS AN INSTRUMENT FOR SELECTIVE CREDIT CONTROL One strand of thought in the Federal Re serve Act and in the statements and actions of Federal Reserve officials through the years is that discount policy should be used, at least on occasion, to influence the uses to which credit will be put—that it should be an important instrument, if not the principal instrument, for selective credit control. This entire idea requires rethinking. For purposes of argument, I shall concede that on occasion the Federal Reserve may wish to exercise some degree of selective control over the uses of credit, by types of bor rower and/or by types of use to which the credit is put. Selective control may be at tempted over (1) credit from all sources, (2) credit from all commercial banks, (3) credit from all member banks, or (4) credit from banks that are currently borrowing from the Federal Reserve. My thesis is that primary reliance on discount policy to achieve such results is likely to prove in effective, or to have undesirable side ef fects, or both. I believe Federal Reserve history bears me out. If the Federal Re serve wishes to exercise selective credit con trols, it should rely primarily on other more comprehensive controls. To analyze this, consider some methods that have been used. Attempt to encourage certain types of paper and to hold down interest costs on it by giving it privileged access to the Fed eral Reserve Banks. This can, of course, be successful if the Reserve Banks stand pas sively ready to buy all of the paper pre sented at the posted buying rate with no onus on the sellers. For example, this worked in the late 1920’s when the Federal Reserve was the willing residual buyer of acceptances. It would also work if the Fed eral Reserve stood ready to discount or lend on the paper with no quantitative limit and no onus on the borrower. The rate on the paper could be held low relative to other market rates. Note, however, that (1) the Federal Reserve loses control over the vol ume of its holdings of the paper, and it has to buy all of the supply that others are un willing to hold at the Federal Reserve rate, and (2) it has no control over the types of credit created on the basis of the newly issued reserves. Now alter the situation by imposing some sort of quantitative control over the volume of borrowing by individual banks. This might be a “tradition against continuous or excessive borrowing” or anything else that would make it impossible or disadvanta geous to borrow fully on the basis of the “favored” type of paper. In this case it no longer follows that the favored type of paper will enjoy a lower market yield rela tive to other yields or that there will neces sarily be a net increase in total demand for 17 18 the paper. To attract the marginal holder necessary to make demand equal to supply, the rate may have to be fully competitive with other market rates. And it still remains true that the Federal Reserve cannot con trol the types of credit created on the basis of the new reserves. Make “undesirable” types of paper in eligible as a basis for borrowing at the Fed eral Reserve. This might indeed reduce the banks’ demand for this type of paper if they did not hold “eligible” paper of other types sufficient to cover likely needs for borrowing at the Federal Reserve or to cover the quantities the Federal Reserve would be willing to lend, whichever is smaller. But if banks have plenty of eligible paper, their willingness to acquire ineligible paper will be little reduced by its ineligibility. Deny the discounting privilege to banks that hold “undesirable types of paper” or too much of it. The classic case oc curred in 1929, when the Federal Re serve wished simultaneously to curb “spec ulative security loans” and to maintain rea sonable rates for “legitimate business.” The technique attempted was to deny, or at least limit, Federal Reserve loans to banks with speculative securities loans. Note that this did not affect at all the broad classes of lenders: all nonbank lenders, nonmem ber banks, and member banks that were not in debt to the Federal Reserve and ex pected that they would not need to borrow. This narrow coverage was enough to doom the experiment. It probably did restrict somewhat such loans by member banks that were borrowing at the Federal Reserve or who feared they might have to do so. But the restrictive effects on such banks were not nearly so selective as the Federal Re serve had hoped. Such banks had several ways of getting out of debt to the Federal Reserve or of avoiding borrowing there, while maintaining speculative loans on se curities. (1) They bid the Federal funds rate considerably above the discount rate. (2) They sold acceptances, Government se curities, and other open market assets ex tensively. (3) They sold mortgages or re frained from buying them. (4) They even went so far as to limit loans to business customers. In short, the whole attempt was a failure. Loans on securities continued to rise up to the eve of the crash and the restrictive ef fects were not selective; credit of all kinds to all kinds of users was restricted. Gov ernor Harrison, of the New York Reserve Bank, later claimed that this whole “moral suasion” effort aimed only at borrowing members made banks less willing to bor row at the Federal Reserve in the early 1930’s. Whether or not this is true, it is plausible. Not only this case but also other evidence and a priori reasoning lead me to the con clusion that it is unwise to rely primarily, or even heavily, on discounting policy as an instrument for selective credit control. When this is done, the offense is not that of making “undesirable” types of loans; it is that of making or holding such types of loans by banks in debt to the Federal Re serve. Banks and others not in debt to the Federal Reserve can make such loans with out restriction or onus. Moral: stay out of debt to the Federal Reserve and thus main tain freedom of action. I believe that the ef fects of such policies, if resorted to fre quently, would be to: 1. Inhibit use of the Federal Reserve discount window and militate against the development of discounting. Reluctance to borrow would be augmented by “reluctance to become subject to Federal Reserve selec tive controls.” Banks, especially the larger ones, would develop even further their capacity to “stay out of the Federal Re serve” through Federal funds, CD’s, re 19 S E L E C T IV E C R E D IT C O N T R O L purchase agreements, and other financial arrangements. 2. Penalize the wrong thing; that is, bor rowing at the Federal Reserve rather than making “undesirable loans.” Presumably the prime purpose of selective controls is to regulate the making of undesirable loans. To penalize borrowing at the Federal Reserve is a clumsy, ineffective, and inequitable way of trying to inhibit banks from making un desirable loans. 3. Lead to an inefficient allocation of credit. Consider, for example, an attempt to limit business loans by controlling access to the discount window. Is there any reason to believe that the most efficient allocation of credit would require the smallest expan sion by those banks that, for one reason or another—such as deposit drains or inability to attract CD money—were borrowing at the Federal Reserve, or feared they would have to? OTHER BASES FOR SELECTIVE CREDIT CONTROLS As indicated earlier, I believe that selec tive credit controls, if they are to be used, should be based upon Federal Reserve powers broader than the power to discount and applied more widely than to member banks who are in debt to the Federal Re serve or fear that they soon may be. Ideally, such controls should have at least the fol lowing characteristics: They should apply to all lenders, or at least to all potentially important lenders, in the market involved. They should not discriminate against banks borrowing at the Federal Reserve, or member banks, or commercial banks. In some cases it may be enough for the regulations to cover only commercial banks; in other cases they should apply more widely. They should be based upon the social desirability of controlling selectively the type of credit involved and justified on the basis of Federal Reserve responsibility to exercise such controls, rather than on its power to discount. They should be implemented with meas ures appropriate to the selective ends being sought. I do not pretend to know what these measures should be. However, some possibilities may be suggested, at least some of which would require permissive legislation. (Some readers may be shocked by the degree of selective intervention im plied. But can selective controls be ex pected to work otherwise?) (1) Margin or downpayment require ments. (Implies knowledge of true values.) (2) Maximum periods of repayment. (3) Differential reserve requirements against various types of assets, or differential marginal reserve re quirements against increases in various types of assets above some base. (4) Quantitative limitations on in creases of selected types of assets above some base date; that is, not more than 5 per cent above the level at the end of 1966. (5) Limitations on selected types of assets as a percentage of total as sets, or total deposits, or net worth, or some other base. (6) Limitations of changes of selected types of assets as a percentage of changes in total assets, or total deposits, and so forth. (7) Methods of encouraging banks to hold or even to increase their holdings of selected assets: 20 (a) Secondary reserve require ments in the form of the fa vored types of assets equal to at least a stated percentage of deposits or of assets other than cash. (b) Marginal secondary reserve requirements calling for in crease of favored assets equal to at least some percentage of increase of other earning as sets. This, and variations of it, offer interesting possibili ties and problems. (c) Permit banks, in computing required reserves, to deduct from their deposits (demand or time) all or a fraction of their holdings of the favored assets. (This percentage need not be the same as the per centage reserve requirement for the bank.) Those who are more ingenious can think of other possibilities. The types of measures suggested above, and modifications of them, could be used in various combinations. Just one imaginary example. Consider the case in 1966 when the Federal Reserve wished to discourage both the expansion of busi ness loans and bank liquidation of certain favored assets. It might (given the legal power) have proclaimed the following: “Until further notice, the required reserves of any bank will be equal to its regular required reserves against deposits plus an amount equal to 10 per cent of (the change of its business loans over a specified base date minus the change of its holdings of favored assets over the same specified base date).” Consider three cases in which a bank increases its business loans by $100. 1. It increases its holdings of favored assets by the same amount. It has no re quired reserves against assets. 2. It holds constant its holdings of fa vored assets. It has required reserves against assets of $10. 3. It decreases by $100 its holdings of favored assets. It has required reserves against assets of $20. Of course, this would encourage banks to try to borrow rather than sell favored assets. So, if you wish, you can impose a reserve requirement on all bank liabilities, includ ing outstanding repurchase agreements. This is one principle; you work out the details! SOME FURTHER OBSERVATIONS ON DISCOUNTING Here are a few brief comments on lessons from the 1920’s and early 1930’s. As I have indicated earlier, I believe bank willingness to borrow from the Federal Reserve and to remain in debt to it, as well as bank demand for excess reserves, fluc tuate in a procyclical manner, even when cycles are mild. One reason is the wide fluc tuations of customer demands for loans. Another is that banks share the euphoria of boom and the hesitancy of recession. A third is that banks want to retire debt and build up liquid assets in depression because they fear less liquidity will be provided by net flows of funds to them. The moral of this, as many have pointed out, is that out standing discounts should be liquidated through open market purchases at the on set of recession. This should be obvious but was not to most Federal Reserve officials in 1930 and 1931. Discounting does not appear to be a very SE L E C T IV E C R E D IT C O N T R O L effective device for supplying additional funds to credit-scarce areas over a pro longed period. In the 1920’s this was tried; country banks in some cases were encour aged to borrow, agricultural paper of longer maturity was made eligible for re discount, and the Federal intermediate credit system was created. But the results seem to have been rather limited. The rea sons for this were probably numerous, but important were the facts that banks had to 21 endorse the paper and bear the risk, that they had in many cases too little capital to make this a sound practice, and that many banks lacked the inclination to extend them selves. As I read the lesson, success in remedy ing credit scarcity over a prolonged period requires credit institutions that can bypass the banks and put credit in the hands of ultimate users. Admittedly, however, “suc cess” is a relative term. A REVIEW OF RECENT ACADEMIC LITERATURE ON THE DISCOUNT MECHANISM David M. Jones Federal Reserve Bank of New York Contents Introduction_________________________________________________________________________ 25 Major Issues and Related Findings_____________________________________________________ 25 Borrowing and monetary restraint Determinants of member bank borrowing Nonprice rationing Announcement effects Proposals for change Concluding observations Discounting and M onetary C ontrol_____________________________________________________ 29 Borrowing and monetary restraint Determinants of member bank borrowing Nonprice rationing Announcement effects Proposed C hanges in th e Discount M echanism__________________________________________ 37 Abolition of discount mechanism Nondiscretionary approach Discretionary approach Tobin's proposals Bibliography_________________________________________________________________________ 4 2 23 A REVIEW OF RECENT ACADEMIC LITERATURE ON THE DISCOUNT MECHANISM INTRODUCTION in interest rates is of particular concern in this regard. An effort will also be made to cover in some detail the wide range of pro posed changes in the current discounting ar rangement. The primary intent of this paper is to present the post-accord literature on dis counting in such a way as to highlight the major points of emphasis in recent analysis. Hopefully, information of this type can serve as important background material for a reconsideration of the role of the Federal Reserve discount mechanism. The paper does not attempt to make an assessment of the affirmative and negative sides of the many technical issues that are raised in the academic literature. After approximately two decades of disuse, the Treasury-Federal Reserve accord of 1951 prompted renewed interest in the nature and effectiveness of the discount mechanism. ‘Analysis since the accord has been devoted in large part to the unre solved controversy over the nature of the relationship between discounting and mone tary control. This paper discusses only the post-accord academic literature that bears directly on the implications of discounting for mone tary control. Special emphasis will be placed on the determinants of member bank borrowing, including a review of the major issues and related empirical findings. The responsiveness of borrowing to movements MAJOR ISSUES AND RELATED FINDINGS borrowing; (3) the significance of nonprice rationing; and (4) the announcement ef fects of changes in the discount rate. The fundamental issue raised by post accord literature dealing with the Federal Reserve discount mechanism is whether this mechanism operates to subvert or to supplement over-all monetary control. Crit ics have argued that the discount function as it currently operates is fundamentally antagonistic to monetary management. Re lated to this position, issues have developed around a number of topics, namely: (1) the effects of borrowing during periods of restraint; (2) the factors that determine Borrowing and m onetary restrain t On one hand, the discount mechanism may be viewed as a sort of “safety valve” that cushions but does not offset the usually uneven impact on individual banks of re strictive shifts in monetary policy. Tempo rary reserves are allocated through the dis count window directly to those banks com25 26 ing under greatest stress, and thus the Sys tem is free to act more decisively than other wise would be the case. The case favoring the present discount ing arrangement turns on the contention that reserves supplied through the discount window are by nature more restrictive in terms of credit and deposit expansion than reserves supplied through other means. Bor rowing from the Federal Reserve is looked upon as only a temporary source of funds for the individual bank, usually requiring some form of asset adjustment in order to effect prompt repayment. Thus, the larger the over-all volume of borrowing relative to other sources of reserves, the greater the restrictive impact on credit growth. The academic critics of the existing dis count mechanism have not sought to refute directly the points raised above. Their posi tion is founded instead upon the following three general considerations: 1. The initiative in using the discount mechanism rests with the borrowing banks themselves rather than with those charged with the responsibility for monetary control. 2. Member bank borrowing from the Federal Reserve adds to total reserves, whereas sales of Treasury bills or other means of reserve adjustment available to the banks do not. 3. Member bank borrowing tends to rise during periods of monetary restraint and fall during periods of monetary ease. In essence, the critics hold that over-all monetary control is weakened to the extent that discounting counters the impact of Federal Reserve open market operations on the reserve base. Working in the con text of models linking bank reserves to the money supply, and the money supply to economic activity, some economists have argued that borrowing accentuates cyclical swings. D eterm inants of m em ber bank borrowing Inasmuch as discounting is at the banks’ own initiative and, therefore, difficult to predict, post-accord inquiry has focused on the determinants of member bank demand for borrowed reserves. To what extent are banks’ decisions to borrow influenced by profitability considerations? And how strong is the so-called “tradition against borrow ing”? These questions are remnants of the old need versus profitability issue, which was debated at length in the 1920’s and 1930’s.1 The “need” concept has never been clearly defined by its advocates, but according to common interpretation banks that borrow out of “need” do so only to meet temporary, unexpected reserve deficiencies. At the same time, the needy banks supposedly make every effort to repay these debts as soon as possible. This view of borrowing behavior presumes a strong traditional re luctance on the part of banks to be in debt to the Federal Reserve. On the other hand, the strict version of the “profitability” thesis posits that banks will borrow whenever additional funds can be invested in assets that earn yields higher than the discount rate. In short, banks bor row out of a calculated effort to profit from rate differentials, rather than simply in re sponse to the unpredictable swings in mar ket factors that produce temporary reserve deficits. Expressed in these terms, “need” and “profitability” appear to be conflicting mo 1 See, for example, W. Randolph Burgess, The R e serve B anks and the M on ey M a rk e t ; Lauchlin Currie, The Supply and C ontrol o f M on ey in the U n ited States; Charles O. H ardy, C redit P olicies o f the F ed eral R eserve S ystem ; Seymour E. H arris, T w en ty Y ears o f F ederal R eserve P olicy; Winfield W. Riefler, M on ey R ates and M on ey M arkets in the U nited States; and R obert C. T urner, M em b er Bank B o r rowing, . F o r an excellent discussion of the points raised in these earlier writings, see A. Jam es Meigs, Free R eserves and the M on ey S u pply , pp. 6-31. 27 A C A D E M IC LIT E R A T U R E ON D IS C O U N T M E C H A N IS M tives. In effect, the borrowing-out-of-“need” proponents postulated that such borrowing was insensitive to levels of interest rates, while the “profitability” school visualized that borrowing was affected by rate levels. One of the few important contributions of the post-accord literature on discounting has been the theoretical resolution of the need versus profitability issue. But even this accomplishment rests in large part on a modified concept of profitability that dates back to Turner’s work in the 1930’s. The argument runs roughly as follows: Given a reserve deficiency or the need to borrow —whether the cause is an unexpected surge in required reserves, or a sudden cash drain, or some other reserve-absorbing factor— the extent to which a bank makes use of the discount window for its reserve adjust ments depends upon the relative costs of borrowing and of other means of replenish ing reserves. For example, the higher the Treasury bill rate—that is, the larger the loss of revenue from reducing the bill port folio—relative to the discount rate, the less the relative cost of borrowing (or the greater the profitability) to meet a given re serve deficit. Thus, a reluctant bank that borrows only to meet its immediate needs can, at the same time, be sensitive to the rate differentials between its alternative sources of short-term funds. By using this modified concept of profitability, it has been demonstrated with some rigor that it is possible to integrate, into a consistent theory, bank reluctance to be in debt to the Federal Reserve and the profit incentive for such borrowing. During periods of monetary restraint, the discount rate tends to lag behind rising market rates on alternative sources of funds, and borrowings rise. Conversely, the dis count rate remains above falling market rates on the same sources of funds during periods of monetary ease, and borrowings fall. This fact represents one basis of the contention that borrowings tend to accen tuate cyclical swings. Nonprice rationing The attitude of banks toward the nonprice terms applied at the discount window has an important bearing on their decisions to borrow. Yet it appears that the Reserve Banks find it quite difficult to admin ister these terms. A wide variation in nonprice terms, among the various Federal Reserve districts and/or over time, can serve to diminish significantly the predic tability of borrowings. It is difficult, if not impossible, to separate—and measure in relative terms—the effects of nonprice ra tioning from the effects of bank reluctance to borrow. It has been argued that these two factors have a mutually reinforcing ef fect on bank borrowing. But there has been very little in the literature on this subject. In general, there seems to be a dissatisfaction with nonprice rationing, explicitly on the grounds that the price mechanism would operate more effectively. A nnouncem ent effects A major source of contention in the litera ture has been the question of whether dis cretionary changes in the discount rate have undesirable effects on expectations. On the one hand, it is argued that one must make inconsistent assumptions about the behavior of lenders and borrowers in order for the announcement feature of discount rate changes to have desired effects. It has also been argued that, at best, the announce ment effects will be unpredictable. There are, however, those who see some merit in announcement effects. They argue that discretionary changes in the discount rate have two basic advantages. First, the 28 changes are widely publicized and espe cially useful as a universal means of signal ing the intent, for example, to stem a bal ance of payments drain. Second, discount rate adjustments, the only m ajor m onetary instrument that has no direct effect on re serves, can play a unique and often helpful role as an index of the course of policy. P roposals for change Proposals for changing the discount m ech anism have run the gam ut from, abolishing the mechanism altogether to allegedly m ak ing it the most powerful tool of m onetary policy. Elim ination of the discretionary aspect of discount-window administration is the object of nearly all of the proposed modifications. A plan often suggested would eliminate discretionary discount rate changes by tying the discount rate to the m arket rate on some alternative source of ready funds. This type of arrangem ent usually involves setting the discount rate high enough above the anchor rate to m ake it a “penalty” rate. M ost ad vocates of such a device would rely on the price mechanism alone to allocate Federal Reserve credit and to keep borrowing in check; they would, in effect, discard the present borrowing “privilege” with its non price connotations in favor of granting banks the “right” to borrow. There has been controversy, however, on the appro priate m arket rate to which the discount rate would be anchored. A somewhat m ore radical plan calls for the payment of interest at the discount rate on m em ber banks’ excess reserves. Through adjustments in the discount rate, the F ed eral Reserve would then have direct con trol over the opportunity cost of bank lend ing. U nder such an arrangem ent, banks would be tem pted to increase their excess reserves and reduce their holdings of short term Government securities. The discount rate would then take on sharply increased importance among the m ajor instruments of m onetary policy. There are, in addition, those who would abolish the discount mechanism. Two rea sons for such a move have been advanced. First, by doing away with borrow ing at the banks’ initiative, the Federal Reserve would greatly improve its control over total reserves. Second, it has been argued that the discounting function is no longer neces sary in view of the substantial postw ar growth in banks’ holdings of short-term Government securities, which can be used to m ake the necessary adjustments in re serve positions. Needless to say, the latter argum ent has little relevance under circum stances in which bank holdings of short term Government securities are minimal. It has also been proposed, however, that the discounting terms should be fully dis cretionary. The basic contention is that the discretionary approach not only entails the power to control total borrowing but also makes possible the selective control of bank lending practices. Concluding observations Although most of the m ajor issues raised in the academic dialogue on discounting re m ain unresolved, it is possible to draw some general conclusions. Discounting does not, for example, appear to weaken m onetary control to any significant extent during periods of m onetary restraint. Indeed, the discount mechanism is, for the most part, a useful complement to open m arket opera tions. Those favoring the current arrange ment argue, in particular, that shifts in m onetary policy are cushioned by the pro vision of tem porary reserves through the discount window to those banks that suffer the greatest stress. A t the same time, bor A C A D E M IC LITERAT U RE ON D ISC O U N T M E C H A N IS M rowed reserves have less expansive implica tions for credit and deposit growth than a corresponding am ount of reserves supplied through other means. On the other hand, regardless of how limiting the effect of borrowed reserves on credit growth may be, the fact remains that m onetary control is rendered less precise under conditions in which banks borrow at their own initiative. Hopefully, the predict ability of borrow ing can be improved by reliable quantitative measurements of the relative effects of interest rates and other factors that influence banks’ decisions to borrow. W ith regard to adm inistration of the dis count window and general supply consid erations, there is almost unanimous agree ment among economists on the desirability of complete reliance on the price m ech anism to control borrowing. But regardless of how appealing the “tied” rate plans may be, there has been no agreement on the m arket rate to which the discount rate 29 should be linked nor on the appropriate spread to be m aintained. A lthough experi ence suggests that there should be some substantial revisions in the present nonprice discounting guidelines, it seems that both price and nonprice terms will continue to be necessary to insure effective m onetary control. Finally, the predom inant view in the literature is that under present circum stances the announcem ent effects of changes in the discount rate will be am biguous at best. A t the same time, those who fear that changes in the rate will have adverse effects on expectations may have overrated their case a bit. In particular, it is not likely that discount rate changes alone, whatever their effects on expecta tions may be, dom inate the behavior of borrowers and lenders. Indeed, these rate adjustments are only one of m any factors that influence expectations about the course of m onetary policy and future economic conditions. DISCOUNTING AND MONETARY CONTROL Borrowing and m onetary restraint As noted earlier, those favoring the current discount procedures often assume that bor rowed reserves are less expansive in terms of credit growth than a corresponding am ount of reserves provided through open m arket operations.2 It is argued that banks will seek to extinguish their borrow ed re serves promptly, usually through some form of asset adjustment. In Roosa’s w ords:3 2 See, for example, B oard of G overnors of the Federal Reserve System and the U.S. Treasury, The Federal R eserve and the Treasury: A n sw ers to Q ues tions from the C om m ission on M on ey and Credit, p. 118. 3 R obert V. Roosa, “Credit Policy at the Discount Window: Com m ent,” p. 334. In the A m e ric an settin g th e fa c t th a t b an k s b o r ro w o nly as a privilege m e a n s th a t even th o u g h an y in d iv id u al b a n k ca n te m p o ra rily , in effect, cau se th e crea tio n o f reserv es b y •b o rro w in g a t th e d isco u n t w indow , th a t sam e b a n k sim u ltan e ously takes on an o b lig atio n to find w ays o f ex tin g u ish in g th o se reserv es— th e m o re p ro m p tly th e b etter, in o rd e r to p reserv e its p rivilege fo r use ag ain w h en u n e x p e c te d reserv e d rain s o ccu r. T h u s, as a gen eral ru le, th e la rg e r th e aggregate vo lu m e o f b a n k b o rro w in g fro m th e F e d e ra l R eserve, the g re ate r w ill b e th e effo rt going on, th ro u g h th e b a n k in g system , to lim it cred its an d b rin g reserves in to b a lan c e w ith th e re q u ire m e n ts against deposits. The fact that Roosa casts his discussion in terms of the actions of an individual bank is not to deny that a high or rising volume of borrowings for the banking sys 30 tem as a whole m ay persist for long periods — as for example, when an increasing num ber of banks turn to the discount window for tem porary reserve relief. But the key point is that aggregate borrowed reserves have a restrictive im pact on credit expansion; and the higher the level of such borrowing, the greater the restriction involved. A part from the special nature of bor rowed reserves, Samuelson has argued that the tendency for borrowings to offset in part the reserve effects of open m arket opera tions actually strengthens m onetary policy. He observes th a t:4 W hile it is tru e th a t d isco u n tin g o fte n acts c o u n te r to o p e n -m a rk e t o p eratio n s, th e re is n o evidence th a t a u n it ch an g e in o p e n -m a rk e t o p eratio n s in d u ces an o p p o sin g ch an g e in d isc o u n t ing larg e en o u g h to rev erse o r su b stan tially w ip e o u t th e orig in al effect. So it is n o t really diffi c u lt fo r th e p la n n e rs o f o p e n -m a rk e t o p e ra tio n s to tak e all th is in to ac co u n t; an d precisely b e cau se th e y k n o w th a t th e d isc o u n t w in d o w p ro vides an escap e valve, th ey can be m o re c o u ra geous in th e use o f o p e n -m a rk e t o p eratio n s. Am ong the critics of the present dis counting arrangem ent, M ilton Friedm an looks upon borrowing with somewhat more alarm. He contends that since the banks can discount at their own initiative, the System is unable to exert direct control over m onetary expansion.5 W arren Smith, another academic critic of the current discount mechanism, asserts that those who emphasize the restrictive nature of borrow ed reserves overlook the all-im portant fact that member bank bor rowing adds to total reserves. “Therefore . . . borrow ing constitutes an offset to the restraint that brought it about to the extent that the supply of reserves is thereby in creased.”6 Finally, the procyclical fluctuations in borrowings have been criticized by Aschheim 7 and Brunner and M eltzer8 among others. In this regard, Aschheim observes that “. . . however strong the commercial bank tradition and however potent the Fed eral Reserve policy, they have not stood in the way of cyclical fluctuations in the vol ume of rediscounting.”9 B runner and M elt zer go into somewhat m ore detail on this m atter: 10 T h e a d m in istra tio n o f th e d isc o u n t w in d o w co n trib u te d b o th in th e tw en ties an d th e fifties to th e cyclical v a ria b ility o f th e m o n e y supply. T h e d isco u n t ra te ty p ically lags b e h in d th e m o v e m en ts o f th e m a rk e t rates. A cy clical u p sw ing, g en e ra ted o r re in fo rc e d b y n o n m o n e ta ry fa c to rs, p u sh es m a rk e t rate s ah e a d o f th e d isco u n t rate, a n d in d u ces b an k s to e x p a n d th e ir b o rro w ing. T h e risin g v o lu m e o f d isco u n ts a n d ad v an ces increases th e [reserve] base a n d co n seq u e n tly in creases th e m o n ey su p p ly. A reverse o p e ra tio n o c cu rs in a dow nsw ing. T h e cy clical v ariab ility o f th e m o n e y su p p ly is th u s am plified b y th e o p e ra tio n o f th e d isc o u n t w in d o w . D eterm inants of m em ber bank borrowing M ost of the post-accord dialogue on the factors that influence banks in their deci sions to borrow has been conditioned by the need versus profitability issue that was de bated extensively in the 1920’s and 1930’s. Recent attempts have been m ade to isolate and quantify the im pact of interest rates on borrowing, and general comments on the 6 W arren L. Smith, “The Discount Rate as a Credit Control W eapon,” p. 172. 7 Joseph Aschheim, Techniques o f M on etary C on trol. 8 U.S. House of Representatives, Subcomm ittee on Domestic Finance, A n A ltern ative A pproach to the 4 Paul A. Samuelson, “Reflections on M onetary Policy,” p. 266. 5 M ilton Friedm an, A P rogram fo r M on etary S tability , p. 38. M on etary M echanism . 9 Aschheim, op. cit., p. 91. 10 U.S. House of Representatives, Subcommittee on Domestic Finance, op. cit., p. 35. A C A D E M IC LITERA T U RE ON D ISC O U N T M E C H A N IS M sensitivity of borrowing to rate movements are abundant in the literature. Somewhat less attention has been devoted to the ques tion of bank reluctance to borrow. One of the more interesting contributions in the post-accord literature is a theoretical recon ciliation of these two motives. Interest rates and borrowing. M any of those who feel that the present discount m echan ism weakens m onetary control are alarmed by evidence suggesting that borrowings are sensitive to interest rates and that they therefore work systematically against open m arket operations. Although the extent to which borrowings respond to rate move ments is clearly an empirical question, the evidence is scanty. Typical of the casual observation in this area is the following: “No doubt it is true that banks are reluc tant to borrow, but like m any ordinary per sons, bankers allow their reluctance to be overcome by more attractive alternatives.”11 Aschheim theorizes in a similar vein:12 T h e F e d e ra l R eserv e p refe rs to state th a t in tim e o f m o n e ta ry tig h tn ess th e re is a g re a t “n e e d ” on th e p a rt o f m e m b e r b an k s fo r red isco u n tin g . E co n o m ically , th e m o re in fo rm a tiv e fo rm u la tio n , h ow ever, is th a t in tim es o f m o n e ta ry tightness it is m o re p ro fitab le fo r b a n k s to b o rro w fro m th e F e d e ra l R eserve th a n in o th e r p erio d s. W arren Smith is somewhat m ore specific about the way in which he feels that in terest rates influence borrowing decisions, but he too stays prim arily in the realm of supposition in observing that while bank dem and for readily available funds to sat isfy the kind of urgent needs that commonly induce banks to borrow at the discount window is probably quite interest-insensi tive, the extent to which banks actually turn to the Federal Reserve to satisfy these needs H E a rl Rolph, “Discussion,” pp. 413 and 414. 12 Aschheim, op. cit., p. 91. 31 rather than relying on other sources m ay be significantly affected by rate m ovem ent:13 In m o st cases, b a n k s h a v e a c h o ice o f o b ta in in g a d d itio n al fu n d s b y b o rro w in g a t th e F e d e ra l R eserve o r by liq u id a tin g seco n d a ry reserv es o r o th e r in v estm en t securities. S urely, th e m a jo r fa c to r influ en cing th e c h o ice w ill be th e re le v a n t cost o f fu n d s o b tain ed b y th e v a rio u s m eth o d s, a n d th is d ep en d s chiefly o n th e re la tio n b etw een th e d isco u n t rate an d th e ex p ected yield o n assets th a t th e b a n k m ay c o n sid e r liq u id a tin g . Meigs, who actually focuses on bank demand for free reserves (excess reserves less borrow ing), concludes that “aggregate m ember bank borrowing is indeed influ enced by the net yields obtainable on bo r rowed funds, within a considerable p art of the range of interest rates and other condi tions observed.”14 In this connection Meigs makes the point that the hypothesis that member bank borrowing is not responsive to changes in m arket interest rates cannot be confirmed solely by demonstrating that banks are reluctant to borrow. R ather, the characteristics of the dem and schedule m ust be determined by direct empirical observa tion of borrowing and interest rates. M ore recently, de Leeuw has concluded from empirical estimates of bank dem and for borrowed reserves (based on quarterly data for the 1954-62 period) that the re sponse of borrowings to the differential be tween the discount rate and the yield on 3-month Treasury bills is “m oderate,” with implied long-run elasticities with respect to the discount rate and the yield on Treasury bills of —0.7 and + 0 .5 , respectively.15 de Leeuw uses a stock-adjustment form ula tion of the borrowings dem and function in deriving these results. A ccording to the Smith, op. cit., p. 172. 14 Meigs, op. cit., p. 89. 15 F rank de Leeuw, “A M odel of F inancial Be havior,” pp. 512 and 513. 32 stock-adjustment principle, changes in bank borrowings in any given period are a func tion of the discrepancy between the desired level of borrowings in that period and the actual level of borrowings in the preceding period, de Leeuw posits that desired amounts of borrowing are dependent, in turn, upon (1 ) the differential between the Treasury bill rate and the discount rate, (2 ) the Treasury bill rate level, and (3 ) the net inflow of bank funds (that is, changes in private dem and deposits plus Federal Governm ent dem and deposits plus private time deposits less m em ber bank re quired reserves less holdings of loans and other private securities). In an empirical study patterned closely after de Leeuw’s work, Stephen Goldfeld has estimated borrowing dem and functions for city and country banks, separately.16 He found the short-run elasticity of changes in borrowings with respect to the discount rate to be —0.875 for country banks and —0.979 for city banks. Com parable elas ticities with respect to the Treasury bill rate were + 0 .7 8 5 and + 0 .8 7 7 for country and city banks, respectively. Goldfeld’s long-run elasticity estimates for these vari ables were substantially higher than de Leeuw’s and, surprisingly, were higher for country banks than for city banks. Specifi cally, the estimates of long-run elasticity of borrowings with respect to the discount rate were —2.926 and —2.382 for country and Stephen M. Goldfeld, C om m ercial B ank B e havior and E conom ic A c tiv ity . G oldfeld’s short-run elasticities were calculated by — . _L_ where B dr B represents bank borrowing and r is the relevant in terest rate. N ote th at the relevant m ean used was that of the level of borrowing, B. The m ean of the flow variable cannot be used because it could well be zero in some cases. The long-run elasticities were obtained by setting the borrowings flow, AB, equal to zero, solving fo r the steady-state B, and differen tiating as above. city banks, respectively. The Treasury bill rate elasticities were + 2 .6 2 5 for country banks and + 2 .1 3 4 for city banks. In yet another empirical study, Goldfeld and Kane have gone still further by deriv ing estimates of dem and for borrowings for four separate classes of member banks.17 A nother distinguishing feature of this study is that the empirical dem and estimates are based on w eekly data for borrowings. From a demand function that relates borrowings to the Treasury bill-discount rate differen tial, lagged borrowings, and changes in non borrowed reserves, Goldfeld and Kane cal culated implicit short-run elasticities with respect to the bill rate of 0.56 for New Y ork City banks, 0.08 for Chicago banks, 0.15 for other reserve city banks, 0.21 for country banks, and similarly, 0.21 for total m ember banks. Goldfeld and Kane note that the long-run elasticity of borrowings with respect to the Treasury bill rate ranged from 2.8 to 3.9 for the various groups of member banks and that such figures are thus generally consistent with Goldfeld’s quarterly results. (C om parable elasticity estimates for the discount rate were not presented in this article.) The Federal Reserve System has not always been completely clear on the im por tance it attributes to interest rate considera tions in the decisions of banks to borrow. The following is among its pronouncem ents on the subject:18 B anks are g en erally re lu c ta n t to b eco m e in d eb ted to th e F e d e ra l R eserve ex c e p t fo r v ery sh o rt p erio ds, a n d w h en in d e b t feel co n stra in e d to liq u id ate assets. T h e d e te rre n ts to b o rro w in g 17 Stephen M. G oldfeld and Edward J. Kane, “The D eterm inants of M em ber Bank Borrowing: A n Econom etric Study.” 18 U.S. Congress, Joint Economic Committee, E m ploym en t G row th and Price L evels, Hearings, p. 755. A C A D E M IC LITERA T U RE ON D ISC O U N T M E C H A N IS M are greatly w e ak en ed if m a rk e t yields o n sec u ri ties ow n ed b eco m e an d re m a in su b stan tially h ig h e r th a n th e d isc o u n t rate. Going into greater detail on the relation ship among borrowings, m arket rates, and the discount rate under conditions of m one tary restraint, the System has commented th a t:19 . . . it is o f p rim e im p o rta n c e th a t th e g eneral re lu ctan ce o f b a n k s to b o rro w a t th e F e d e ra l R eserve be re in fo rc e d b y a d isc o u n t ra te w ith real d e te rre n t p o w e r a t tim es w h en a te m p e rin g o f b a n k cred it g ro w th is in th e p u b lic in terest. In o th e r w ords, in o rd e r to m a k e th e d isco u n t m ech a n ism an effective su p p le m e n t to o p en m a r k e t o p eratio n s th e F e d e ra l R eserve is obliged to m a in ta in d isco u n t rate s n o t m a rk e d ly lo w er th a n m a rk e t yields o n th e m o st re ad ily av ailable a lte r n ativ e so u rce o f b a n k reserves, T re a su ry bills. If th e F e d e ra l R eserve in these circu m sta n ce s did n o t ad ju st its d isc o u n t rates to k eep th em “in to u c h ” w ith m a rk e t rate s, th e ta sk o f a d m in iste r ing th e d isco u n t w in d o w to p re v e n t excessive cre d it expansio n w o u ld b eco m e v ery difficult. On the other hand, the System has more recently concluded that a comparison of the costs of alternative sources of ready funds with changing amounts of borrowed funds “does not suggest that there is a pow erful borrowing response to changing cost considerations.”20 Reluctance to borrow. Attem pts to discern the nature of the tradition against borrow ing date back to the need versus profitabil ity discussions of the 1920’s. Bank reluc tance to borrow is commonly associated with the notion that since banks are al ready “in debt” to their depositors, with repaym ent due in m any cases on demand, it is im prudent for them to incur additional debt that is of a prior-claim. nature.21 C on tinued borrowing has been viewed as a con19 Ibid., p. 756. Board of G overnors of the Federal Reserve System and the U.S. Treasury, op. cit., p. 134. 21 Ibid., p. 129. 20 33 fession either of weakened condition or of poor m anagem ent.22 There is general agree ment that the reluctance to borrow varies m arkedly in intensity among banks. N ever theless, it has been argued that “in most cases” bank reluctance to borrow is “a deterrent sufficiently strong to prevent ex cessive use of discounting.”23 A t first glance, one might readily inter pret any prem ium in excess of the discount rate that banks pay for Federal funds as a manifestation of bank reluctance to bor row from the Federal Reserve. In fact, however, large banks, which are the ones prim arily responsible for bidding up the Federal funds rate, are almost certainly not insensitive to rates in a way that the tradi tional meaning of reluctance would imply. R ather, these banks m ay be viewed as add ing an implicit cost factor to the discount rate in order to take account of scrutiny by the discount authorities. U nder such cir cumstances, the effective cost of borrowing to these large banks will exceed the pub lished discount rate, and the Federal funds prem ium may be largely illusory. Theoretical reconciliation. Polakoff has demonstrated that it is possible to integrate into a consistent theory bank reluctance to be in debt to the Federal Reserve and the profit incentive for such borrowing.24 The key assumption in Polakoff’s theory is that member banks display a “reluctance elas ticity” when borrowing from the Federal Reserve; in other words, it is true not only that there is a reluctance to borrow at all times but also that this reluctance increases 22 Charles R. W hittlesey, “C redit Policy at the D is count W indow,” p. 213. 23 Board of G overnors of the Federal Reserve System and the U.S. Treasury, op. cit., p. 130. 24 M urray E. Polakoff, “Reluctance Elasticity, Least Cost, and M em ber Bank Borrowing: A Sug gested Integration.” 34 as the volume of discounting grows. View ing member bank decisions to borrow in the context of a “preference” system, Polakoff reasons that as borrowings rise in response to an increasing differential between the yield on Treasury bills and the discount rate, the disutility of borrowing relative to the utility of profit will eventually become so great that member banks will no longer borrow. He argues, in effect, that the banks’ marginal propensity to borrow declines as the spread between the bill and the discount rates widens. To test his hypothesis, Polakoff relates (in scatter diagrams) both weekly and monthly data on member bank borrowings to specific spreads between the bill and dis count rates over the July 1953 to Decem ber 1958 period. He concludes that “the expansion paths of borrowings suggested by the various scatter diagrams are all con sistent with the theoretical results deduced from the integration hypothesis.”25 How ever, these empirical findings are not sup ported by Goldfeld’s results from quarterly data for the somewhat longer period, 1950III to 1962-11. Taking account of the im pact of loan demand and reserve avail ability on borrowing behavior (something Polakoff failed to do) Goldfeld tests spe cifically for the relationship between bor rowings and the rate spread postulated by Polakoff. He finds that while borrowings are in general interest-sensitive, there is no tendency for the marginal propensity to borrow to fall as the rate differential widens.26 25 Ibid., p. 18. In a more recent article, Polakoff has fitted a quadratic function to his empirical data and offered this as further proof of his theoretical scheme. See Murray E. Polakoff, “Federal Reserve Discount Policy and Its Critics,” pp. 205-07. 26 Goldfeld, op. cit., pp. 150 and 151. In their more recent test using weekly data covering the July 1953 to December 1963 period, Goldfeld and Kane Nonprice rationing The guiding principles of Regulation A (as amended in 1955) have been interpreted and applied only with considerable diffi culty. The appropriateness of borrowing under these nonprice terms turns on the intent of the borrower. A bank is not, for example, to borrow willfully in order to make a profit from, rate differentials. But this is basically a subjective determination, and it is difficult, if not impossible, to pin point the uses to which borrowed reserves are put. Distinctions between appropriate and in appropriate borrowing may be quite fine, as evidenced by the following case cited by a former Federal Reserve discount officer:27 . . . if a bank borrowed temporarily to meet a commitment to make a loan to a business concern at 4 per cent, with reasonable expectations of having funds at hand shortly to pay out, the bank would not be borrowing to earn a rate differen tial even though it was borrowing at the lower rate (in one market) and re-lending at a higher rate (in another m arket). With regard to the stability of discount ing terms over time, Professor Whittlesey has set out to correct what he terms a “common misconception” that nonprice discount window standards are adjusted to changing business conditions. According to came up with what they consider to be “limited support” for the relationship between borrowings and rates hypothesized by Polakoff. See Goldfeld and Kane, op. cit., p. 513. The evidence offered by Goldfeld and Kane in support of the Polakoff hy pothesis has recently been brought into question by Polakoff and Silber in “Reluctance and MemberBank Borrowing: Additional Evidence.” Polakoff and Silber argue that high collinearity in Goldfeld and Kane’s observations bearing on Polakoff’s hy pothesis “sheds serious doubt on the validity of these results.” In place of Goldfeld and Kane’s analysis, Polakoff and Silber present their own evidence, which is interpreted as verifying the operation of the “reluctance/surveillance” motive in periods of “tight” money. 27 George W. McKinney, Jr., The Federal Reserve Discount Window, pp. 106 and 107. 35 ACADEM IC LITERATURE ON DISCOUNT MECHANISM Whittlesey, “the fact is that neither the way in which the discount window is adminis tered nor the standards by which member bank borrowing is judged are modified to conform to over-all monetary policy.”28 Roosa is of a similar opinion:29 Insofar as human frailties permit, it is always the same [discount] window, open in the same way at all times for borrowers of the same circum stances. What makes the impact of these con tinuous standards seem to vary is that the cir cumstances of the banks themselves change. The relative importance of discount-win dow administration and the tradition against borrowing in borrowing decisions has been a point of contention. Professor Whittlesey argues, for example, that the ad ministration of the discount window is not a significant feature of over-all credit con trol but that it merely acts in an indirect and admonitory manner “. . . to keep alive and reinforce the tradition against borrow ing, without which discount policy as pres ently conducted could quickly break down.”30 According to Whittlesey, “. . . the privilege of borrowing, despite conventional statements to the contrary, is, in practice, tantamount to a right.”31 Roosa, for one, does not appear to be convinced that discount-window adminis tration does in fact play such an unimpor tant role in borrowing decisions. Without attempting to determine precisely where the influence of discount-window surveillance begins and the influence of the traditional reluctance to borrow runs out, he contends that “both are certainly present; and when ever the check imposed by tradition might begin to falter, the limits imposed by sur veillance would begin to take hold.”32 28 Whittlesey, op. cit., p. 209 29 Roosa, op. cit., p. 334. 30 Whittlesey, op. cit., p. 216. si Ibid., pp. 214 and 215. 32 Roosa, op. cit., p. 336. Finally, the difficulties in administering the provisions of Regulation A may have contributed to variations in nonprice terms among Federal Reserve districts. This is con tended in a recent study of the relationship between borrowed reserves and total re serves in the various Federal Reserve dis tricts. The evidence provided, however, cannot be considered conclusive.83 The dominant view in the literature is that there should be greater reliance on the price mechanism and less on nonprice ra tioning in the allocation of Federal Reserve credit through the discount window. As will be seen in a subsequent section of this pa per, proposals by Aschheim, Brunner and Meltzer, and Tobin all call explicitly for an “open” discount window— that is, one at which banks may borrow all they wish at the existing discount rate. Announcement effects There has recently been a growing concern with the impact of discount rate policies on expectations. Some do not necessarily agree with C. E. Walker’s observation that changes in the discount rate are “a simple and easily understandable technique for in forming the market of monetary authori ties’ views on the economic and tredit sit uation.”34 According to Kareken, some asymmet rical assumptions about the behavior of lenders and borrowers are necessary in or der to argue that the “announcement ef fects” of discount rate adjustments are necessarily stabilizing. In particular, lenders must be expected to interpret an increase in 33 See David T. Lapkin and Ralph W. Pfouts, “The Administration of the Discount Function”; and Jimmie R. Monhollon and James Parthemos, “Administration of the Discount Function: A Com ment.” 34 C. E. Walker, “Discount Policy in the Light of Recent Experience,” p. 229. 36 the discount rate as a sign that tighter credit conditions lie ahead and to react with a more conservative lending policy; borrow ers, on the other hand, must view the in crease in the discount rate as a signal of the end of good times and cut back their spend ing plans and loan demands accordingly.35 Samuelson is not so sure that the borrow ers in fact react in such a manner. He rea sons that:36 Today, financial men know that the Federal Re serve “leans against the breeze,” tightening money when it thinks the forces of expansion are strong and easing money when deflation seems a threat. Therefore it is rational for an investor to say, “Aha! the ‘Fed’ is raising interest rates; they must know that the current outlook is very bull ish, and if that is going to be so, I’d better ex pand my operations.” Conclusion: Announce ment effects are often ambiguous. Taking a position similar to Samuelson’s, Warren Smith concludes that “the effects of discount rate increases on business expec tations are likely to be destabilizing or, at best, neutral,” but he hastens to add that he believes such effects to be “rarely of major importance” because the discount rate is only one of many kinds of informa tion that go into the formulation of busi ness expectations.37 According to Smith, changes in the dis count rate also induce shifts in expectations about monetary policy and bring on related “unsteadiness” in market rates. For ex 35 John H. Kareken, “Federal Reserve System Dis count Policy: An Appraisal,” p. 109. In a more recent article, Warren Smith has ex pressed these conditions under which announcement effects can be assumed to be stabilizing in Hicksian terms. That is, lenders must have elastic expectations about future interest rate movements while borrowers act on inelastic expectations. See Warren L. Smith, “The Instruments of General Monetary Control,” pp. 61-63. 36 Samuelson, “Recent American Monetary Con troversy,” p. 10. 37 Smith, “The Discount Rate,” p. 174. A similar argument is advanced by Smith in “The Instruments,” pp. 63 and 64. ample, failure to increase the discount rate when the Treasury bill rate rises to or above the level of the discount rate may trigger a decline in interest rates, especially if cur rent business indicators happen to be point ing downward even the slightest bit. In attempting to smooth such a swing, the System might bring about tighter monetary conditions than would otherwise be de sirable. Monetary control may also be undermined, Smith argues, when a technical increase in the discount rate, to bring it in line with market rates, is interpreted as a sign of tighter monetary policy ahead and causes a sharp rise in those rates. Appro priate action by the monetary authorities in this case might result in a relaxation of restrictive policies, before such a move were deemed appropriate on general grounds. Still another expression of concern with announcement effects is offered by Culbert son, who observes in particular that the November 1957 reduction in the discount rate “precipitated the most extraordinary bull market in bonds, a development that would have been most untimely had reces sion not been in the offing. The [Novem ber 1957] discount rate reduction seems to have served waiting debt speculators in the capacity of a starter’s gun, and thus, to have contributed unduly to the speculative flavor of the bond market.”38 On the other hand, the System has ob served that discretionary changes in the discount rate are a useful complement to the other major tools of credit policy be cause they are probably the most widely publicized step that a central bank can take— and yet they have no direct effect on the available supply of bank reserves.39 38 John M. Culbertson, “Timing Changes in Mone tary Policy,” pp. 157 and 158. 39 Board of Governors of the Federal Reserve System and the U.S. Treasury, op. cit., p. 146. 37 ACADEM IC LITERATURE ON DISCOUNT MECHANISM PROPOSED CHANGES IN THE DISCOUNT MECHANISM The critics of the present discounting ar rangement have offered alternative pro posals that range from abolishing the practice to making it the most powerful tool in the central banker’s kit. Abolition of discount mechanism Perhaps the most adamant advocate of abolishing discounting is Milton Friedman, who argues that since member banks dis count at their own initiative, the Federal Reserve System cannot determine the amount of money it creates either through the discount window or through a combina tion of discounting and open market opera tions.40 Regarding discount rate policy in particular, Friedman is highly critical of those who have looked to the level of the discount rate rather than its position rela tive to other rates as an indication of the tone of monetary policy. Under a discre tionary discount rate policy, an unchanged rate is accompanied, according to Fried man, by unintended shifts between mone tary tightness and ease as market rates change relative to the discount rate. More over, the occasional but usually substantial changes in the discount rate are viewed as a source of general instability. Friedman sums up his feelings as follows:41 . . . rediscounting should be eliminated. The Fed eral Reserve would then no longer have to an nounce a discount rate or to change it; it would then have direct control over the amount of high-powered money it created; it would not be a source of instability alike by its occasional changes in the discount rate and by the un intended changes in the “tightness” or “ease” of policy associated with an unchanged rate, nor would it be misled by these unintended changes; and it would be less subject to being diverted from its main task by the attention devoted to the “credit” effects of its policy. 40 Friedman, op. cit., p. 38. 41 Ibid., p. 44. However, Friedman adds one vital quali fication to his argument for total abolish ment. He reasons that since required re serves are calculated after the fact, some discrepancies between required and actual reserves are unavoidable. As an alternative to the current charge of the discount rate plus 2 percentage points on realized reserve deficits, Friedman offers a fixed rate of “fine” that “should be large enough to make it well above likely market rates of interest. The fine would then become the equivalent of a truly ‘penalty’ discount rate . . . [but] no collateral, or eligibility requirements, or the like would be involved.”42 It seems that Friedman was not aware of how much this one qualification weakens his solution. As Ahearn has pointed out, this qualification would replace the discount mechanism with an “overdraft system” un der which everything would depend on the height of the penalty rate. If market rates of interest moved up, the penalty rate might have to be adjusted upward to keep it a penalty, which means in essence that the discount mechanism would have crept back under another name.43 Professor Kareken also views the aboli tion of discounting as a possible alternative to the present system.44 He reasons that in view of the growth in the public debt— and especially, of the expansion in the stock of Treasury bills— during and after World War II, there is no longer any need for discounting in order to make reserve adjust ments. With the closing down of discount facilities, banks short of reserves would, ac cording to Kareken, be forced to sell short term Government securities. But those 42 Ibid., p. 45. 43 Daniel S. Aheam, Federal Reserve Policy R e appraised, 1951-1959, p. 140. 44 Kareken, op. cit., pp. I l l and 112. 38 banks with reserve excesses would have a strong incentive to retain their Treasury obligations, and perhaps to acquire more. Aheam contends that this analysis is faulty because Kareken assumes that Gov ernment securities sold by reserve-deficient banks will be bought up by other banks, and this assumption ignores the fact that broad swings in reserve positions affect nearly all banks in roughly the same way at about the same time. If bank reserve posi tions were tightening, Ahearn asserts, it would actually be rational for banks with excess reserves to husband their reserves and, indeed, to sell Government securities in anticipation, before reserve positions tightened further and depressed prices of securities lower.45 In the light of more re cent developments, there is, of course, the additional argument that bank holdings of short-term Government securities may ac tually drop so low as to limit sales of such securities as a means of reserve adjustment. Nondiscretionary approach A general dissatisfaction with the discre tionary features of discount policy is re flected in nearly all of the suggested modi fications in this mechanism. The proposals along this line rest on the assumption that considerations of “profitability” do, in fact, bear heavily on borrowing decisions. The central feature of the proposed nondiscre tionary discounting arrangements is a dis count rate that is “tied” to the Treasury bill rate or some other money market rate that is relevant to borrowing decisions. Such an arrangement appears to be motivated in large part by the desire to: (1) stabilize the rate differentials that influence borrowing decisions, thus hopefully stabilizing the aggregate amount of borrowing, and (2) eliminate the threat of adverse announce ment effects stemming from discretionary ■15 Ahearn, op. cit., p. 140 (n. 51). changes in the discount rate. When coupled with a penalty-rate concept, this system establishes a basis for relying entirely on the price mechanism for the allocation of credit at the discount window. The practical problem of how high to set the penalty rate is an important one. If the rate is set too high, borrowing from the Federal Reserve Banks may cease to be a practical alternative for banks unexpectedly in need of reserves. Many regard this lender-of-last-resort function as an impor tant central bank responsibility, however, and the adverse effect on the attractiveness of membership in the System is also a con sideration. On the other hand, if the penalty rate is set too low, the volume of borrow ing may become “excessive.” Another prob lem— perhaps even more thorny— is created by the fact that market interest rates do not move in perfect tandem with each other. Thus if the discount rate were tied to some particular rate, movements of other market rates relative to the chosen rate could result in continued interest-rateinduced instability in the aggregate volume of borrowing. The choice of the market rate to which the discount rate would be tied and of the size of the differential to be used hinges in significant part on the question of whether banks balance borrowings against rates on other sources of readily available funds or whether borrowings are related to the rate that banks can earn on loans. A penalty discount rate that is effective under condi tions in which borrowings are balanced against rates on marginal assets (that is, Treasury bills) may not inhibit borrowing decisions that are related to the higher re turns on other types of earning assets. Moreover, even if the discount-window authorities effectively preclude borrowing to lend at a profit under the terms of Regu lation A, a given penalty rate may become ACADEM IC LITERATURE ON DISCOUNT MECHANISM ineffective as banks shift from one short term source of funds to another. For exam ple, if the discount rate is set at some spec ified margin above the Treasury bill rate and a substantial number of banks turn to other sources of short-term funds such as certificates of deposit (CD’s), the discount rate may lose its initial penalty properties. Warren Smith has observed that, on prac tical grounds, the discount rate should ex ceed the Treasury bill rate by a margin that is sufficient to discourage unnecessary bor rowing without imposing too heavy a pen alty on banks that are forced to borrow because they lack salable securities. On this basis, he determines that the discount rate should be set a full 1 percentage point or more above the Treasury bill rate.46 Smith has been careful to distinguish be tween his penalty-rate system, in which the use of the discount window is penalized in cost terms relative to other sources of short term funds, and the British plan, in which the penalty rate is related to the return on earning assets— which happens in the case of the British discount houses to be almost exclusively Treasury bills. The British pen alty-rate concept is held to be impracticable in the United States “because there are sev eral thousand member banks able to bor row directly from the Federal Reserve and invest their funds in a broad range of assets carrying widely varying interest rates.”47 46 Smith, “The Discount Rate,” p. 176. More recently, however, Smith has voiced reservations about using the Treasury bill rate as an anchor rate. At a Federal Reserve seminar on the discount mechanism in May 1966 he noted that in the last few years many banks have come to use CD’s rather than Treasury bills in their reserve adjust ments. At the same time, Smith indicated that he has become less certain of the appropriate penaltyrate spread: “There is a fuzziness about what a penalty rate is here. Does it have to be sort of higher than any rate that any bank can earn on an asset, at one extreme, or does it have just to be a little bit above the lowest rate [at which] any bank can turn out any asset, at the other extreme? It’s probably somewhere in between.” 47ibid., p. 171 (n. 3). 39 In another “tied” discount rate plan, Ahearn proposes that the discount rate be anchored to the bill rate but that the Fed eral Reserve be allowed to vary the differen tial in accordance with monetary policy aims. “This would retain needed flexibility in the relation of the discount rate to other money market rates but also minimize the possibility of market misinterpretation of the meaning of discount rate changes.”48 Brunner and Meltzer also call for an ar rangement in which the discount rate would always exceed the bill rate, but not neces sarily by a fixed margin. They envision a market-determined discount rate and sug gest that the discount window should be kept “open” at the penalty rate.49 Precisely how the penalty rate would be determined is not spelled out, however. Aschheim presents a plan in which the discount rate would be tied to the rate on Federal funds instead of the Treasury bill rate because Federal funds are considered to be the closest substitute for reserve ac commodation from the Federal Reserve. As did Brunner and Meltzer, Aschheim also envisions (but fails to spell out) a penaltyrate scheme in which “the ‘principles of prudent discounting’ that are currently ap plicable to the System’s rediscount facility could be dispensed with.”50 He concludes:51 Where . . . open market operations are feasible, nonpenal rediscounting is— in effect— an escape mechanism for commercial banks seeking to over come the constraint o f restrictive open-market policy. Last-resort reserve accommodation via a penalty rate eliminates this escape mechanism while retaining the safety valve of central-bank lending to member banks at the latter’s initiative. Thus, in monetary systems possessing the insti tutional setting for open-market operations, pen alty-rate rediscounting enhances the effectiveness of central bank control. 48 Ahearn, op. cit., p. 144. 49 U.S. House, Subcommittee on Domestic Fi nance, op. cit., pp. 89 and 90. 50 Aschheim, op. cit., p. 94. 51 Ibid., p. 98. 40 Some technical difficulties appear to exist, however, in attempting to tie the dis count rate to the Federal funds rate in instances where resort to the “window” is unlimited. The predetermined and fixed penalty spread would have to be added to some past value of the funds rate to deter mine the current discount rate, say, the average effective funds rate for the pre ceding week. The balance of supply and demand in the funds market is very shiftable, however, and the rate tends to be quite unstable. As long as the current funds rate remained below the current week’s dis count rate, borrowings would probably be very low. If the current funds rate should rise to the discount rate, however, banks would be indifferent between the funds market and the “window” as a source of reserves, the funds rate would rise no fur ther, and borrowings could rise indefinitely until the demand for reserves was satisfied at the existing discount rate. Thus, it would appear that considerable instability in the volume of borrowings would be reintro duced. In summary, those advocating a nondiscretionary discount mechanism would at tempt to minimize the variability in borrow ings by fixing the differential between rates pertinent to the borrowing decision and to hold down the average level of borrowings by setting the discount rate at a penalty level.52 Discretionary approach As an alternative to his proposal for abolish ing discounting, Kareken suggests that the discretionary features of discounting be strengthened.53 In his view, there is no basis for thinking that nonprice rationing is in principle any less effective than price rationing in curbing unwanted expansions of Federal Reserve credit. Indeed, the dis cretionary approach entails the power to control total indebtedness and also to con trol bank lending practices selectively. The selective control of bank lending is considered to be a means of influencing two factors of “special significance” in the con temporary inflationary process— namely, inventory speculation and money wage pressures. To the extent that funds needed to finance an inventory build-up or an in crease in corporate transactions balances (in order to make larger wage payments) must be limited by member banks that make use of the discount window, the in ventory and wage sources of inflationary pressures would be blunted. Kareken notes that his plan would re quire the establishment of appropriate non price eligibility conditions such as a maxi mum figure for the ratio of loans to total loans and investments. In addition, and in marked contrast to most of the proposed modifications in discounting, it would be necessary to keep the discount rate below the penalty level. “If banks are to avail 52 Some interesting variants of the “tied” rate plan themselves of the System’s discount facili were offered at the Federal Reserve seminar re ties and thereby to submit to the regulation ferred to in footnote 46. It was proposed, for ex ample, that the discount rate should be linked to of their activities it must in some sense be the Federal funds rate but that the spread should profitable for them to do so.”54 Aschheim increase with both the size and duration of an in dividual bank’s borrowing from the Federal Reserve. indicates general disapproval of this scheme Another plan called for a given bank to pay a bor by asking the obvious:55 rowing rate that is fixed in relation to its return per dollar of loans and invc s<n:cv,ts on the grounds that since the most efficient bankers con "itute the hard core of borrowers, a single penalty discount rate for the system as a whole might have perverse effects by penalizing least those that tend to borrow the most. For an excellent summary of the dialogue and proposals at the seminar on discounting in May 1966, see: Priscilla Ormsby, “Summary of Issues Raised at the Academic Seminar on Discounting,” starting on p. 47. 53 Kareken, op. cit., pp. 119 and 120. 54 Ibid., p. 121. 55 Aschheim, op. cit., pp. 96 and 97. ACADEM IC LITERATURE ON DISCOUNT MECHANISM If the purpose is selective lending control, why confine it to those banks that choose to subject themselves to it? If many banks choose to shun the discount window to avoid central-bank reg ulation of their lending practices, how far down shall the discount rate go or how watered down shall the selective lending control be in the effort to lure more banks to the discount window? Tobin’s proposals Professor Tobin advocates a radical de parture from the current discounting ar rangement that would make the discount rate “the most powerful tool in the central bankers’ kit.”56 He makes two basic pro posals:57 (1) The Federal Reserve Banks should pay in terest at the discount rate on member bank re serve balances in excess of requirements; (2 ) Banks should be released from the prohibi tion of interest payments on demand deposits and from the ceilings on interest rates on time and savings deposits. According to Tobin, the purpose of the first proposal is to tighten the control of the Federal Reserve over the opportunity cost of bank lending. By raising the discount rate, the Federal Reserve would “clearly, directly, and quickly” make lending less attractive to all banks, regardless of whether they are in debt to the Federal Reserve or not. The discount rate would become a floor to the rate on Treasury bills and simi lar short-term paper that banks might hold as secondary reserves. The purpose of the second proposal is to tighten the Federal Reserve’s control over the opportunity cost that bank deposi tors charge against any alternative invest ment of funds. “The rate that banks pay depositors will be closely geared to the dis count rate since a bank will always be able 56 James Tobin, “Towards Improving the Effi ciency of the Monetary Mechanism,” p. 279. 57 ibid., pp. 277 and 278. 41 to earn a fraction of the discount rate (one minus the required reserve ratio) on a new deposit.” Among the advantages claimed by Tobin for the second proposal are the elim ination of the “unproductive efforts” de voted to economizing on cash in periods of high interest rates, and the replacement of the existing “wasteful and imperfect” non price competition with price competition. “Better to pay depositors interest than to seek their patronage by organ music, free silverware, and plush surroundings.”58 Another important feature of Tobin’s plan is that the Federal Reserve would make a perfect Federal funds market at the dis count rate. Among the implications fore seen by Tobin for his proposals are that much of the short-term Government debt would be transferred to the Federal Re serve from banks and corporations, leaving them to hold excess reserves and bank de posits, respectively. Also, Tobin suspects that monetary control under his system might require much wider fluctuations in discount rates and connected short-term interest rates “than we have yet had the courage to try.”59 By way of criticism of the Tobin scheme, Ahearn points out that the potential for inflationary enlargement of the reserve base would be enormous; yet the only adminis trative defense against member bank bor rowing would be the power to raise the discount rate.60 The problem would be com pounded by the difficulties in carrying out offsetting open market operations under conditions of dried-up public short-term Government security holdings.61 58 Ibid., p. 278. 59 Ibid., p. 279. 60 Ahearn, op. cit., p. 133. 61 This point is made by Jonathan Levin in “Pro fessor Tobin on the Monetary Mechanism,” an internal memorandum of the Federal Reserve Bank of New York, Sept. 8, 1960, p. 5. 42 BIBLIOGRAPHY Books Ahearn, Daniel S. Federal Reserve Policy Reappraised, 19511959. New York: Columbia University Press, 1963. Aschheim, Joseph. Techniques of Monetary Control. Baltimore: Johns Hopkins University Press, 1961. Board of Governors of the Federal Reserve System and the U.S. Treasury. The Federal Reserve and the Treasury: Answers to Questions from the Commission on Money and Credit. Englewood Cliffs, N.J.: Prentice-Hall, Inc., 1963. Burgess, W. Randolph. The Reserve Banks and the Money Mar ket, rev. ed. New York: Harper and Bros., 1936. Currie, Lauchlin. The Supply and Control of Money in the United States, rev. ed. Cambridge: Harvard University Press, 1935. Friedman, Milton. A Program for Monetary Stability. New York: Fordham University Press, 1959. Goldfeld, Stephen M. Commercial Bank Behavior and Economic Activity. Amsterdam: North Holland Publishing Co., 1966. Hardy, Charles O. Credit Policies of the Federal Reserve System. Washington: The Brookings Institution, 1932. Harris, Seymour E. Twenty Years of Federal Reserve Policy. Cambridge: Harvard University Press, 1933. McKinney, George W., Jr. The Federal Reserve Discount Win dow. New Brunswick, N.J.: Rutgers University Press, 1960. Meigs, A. James. Free Reserves and the Money Supply. Chicago: University of Chicago Press, 1962. Riefler, Winfield W. Money Rates and Money Markets in the United States. New York: Harper and Bros., 1930. Roosa, Robert V. Federal Reserve Operations in the Money and Government Securities Markets. New York: Federal Reserve Bank of New York, 1956. Turner, Robert C. Member-Bank Borrowing. Columbus: Ohio State University Press, 1938. U.S. Congress, Joint Economic Committee. Employment Growth and Price Levels, Hearings, “Part 4— The Influence on Prices of Changes in the Effective Supply of Money,” 86th Cong., 1st sess. (May 25-28, 1959). U.S. House of Representatives, Subcommittee on Domestic Fi nance. An Alternative Approach to the Monetary Mechan ism by Karl Brunner and Allan H. Meltzer, 88th Cong., 2nd sess. (Aug. 17, 1964). Willis, Parker B. The Federal Funds Market. Boston: Federal Reserve Bank of Boston, 1957. ACADEM IC LITERATURE ON DISCOUNT M ECHANISM Periodicals and Other References Alhadeff, David A. “Monetary Policy and the Treasury Bill Mar ket,” American Economic Review, vol. 42 (June 1952), pp. 326-46. Board of Governors of the Federal Reserve System. Annual Re port. 1952, 1953, 1957. ---------- . “Excess Profits Taxes of Commercial Banks,” Federal Reserve Bulletin, vol. 37 (June 1952), pp. 602-19. ---------- . “Interest Rates and Monetary Policy,” by Stephen H. Axilrod and Ralph A. Young, Federal Reserve Bulletin, vol. 48 (Sept. 1962), pp. 1110-37. ---------- . “Regulation A— Effective February 15, 1955,” Federal Reserve Bulletin, vol. 41 (Jan. 1955), pp. 9-14. Carr, Hobart C. “Federal Funds,” in Money Market Essays. New York: Federal Reserve Bank of New York, 1952, pp. 13-16. Culbertson, John M. “Timing Changes in Monetary Policy,” Journal of Finance, vol. 14 (May 1959), pp. 145-60. de Leeuw, Frank. “A Model of Financial Behavior,” in The Brookings Quarterly Econometric Model of the United States, eds. James S. Duesenberry et al. Chicago: Rand, Mc Nally and Co., 1965, pp. 465-530. Federal Reserve Bank of Atlanta. “The Discount Rate and Bank Lending,” by Thomas R. Atkinson, Monthly Business Re view (Feb. 1953), pp. 9 and 10. Federal Reserve Bank of New York. “Borrowing from the Fed,” Monthly Review, vol. 41 (Sept. 1959), pp. 138-42. ---------- . “The Significance and Limitations of Free Reserves,” Monthly Review, vol. 40 (Nov. 1958), pp. 162-67. ---------- . “Techniques of Member Bank Borrowing at the Fed eral Reserve,” Monthly Review, vol. 46 (Apr. 1964), pp. 66 - 68 . Federal Reserve Bank of Philadelphia. “Discount Rate and Dis count Policy,” Business Review (Jan. 1959), pp. 16-26. Federal Reserve Bank of St. Louis. “The Discount Mechanism and Monetary Policy,” Monthly Review, vol. 42 (Sept. 1960), pp. 5-9. ----------. “Some Misconceptions in Public Understanding of Mone tary Policy,” Monthly Review, vol. 41 (Nov. 1959), pp. 1 2 4 -2 8 . Goldfeld, Stephen M., and Kane, Edward J. “The Determinants of Member-Bank Borrowing: An Econometric Study,” Jour nal of Finance, vol. 21 (Sept. 1966), pp. 499-514. Hodgman, Donald R. “Member-Bank Borrowing: A Comment,” Journal of Finance, vol. 16 (Mar. 1961), pp. 90-97. 44 Kareken, John H. “Federal Reserve System Discount Policy: An Appraisal,” Banca Nazionale del Lavoro Quarterly Review, No. 48 (Mar. 1959), pp. 103-24. Lapkin, David T., and Pfouts, Ralph W. “The Administration of the Discount Function,” National Banking Review, vol. 3 (Dec. 1965), pp. 179-86. Lindow, Wesley. “The Federal Funds Market,” Banker’s Monthly, vol. 77 (Sept. 1960). McKinley, David H. “The Discount Rate and Rediscount Policy,” The Federal Reserve System, ed. Herbert V. Prochnow. New York: Harper and Bros., 1960, pp. 90-112. McWhinney, Madeline. “Member Bank Borrowing from the Fed eral Reserve Banks,” in Money Market Essays. New York: Federal Reserve Bank of New York, 1952, pp. 8-12. Monhollon, Jimmie R., and Parthemos, James. “Administration of the Discount Function: A Comment,” National Banking Review, vol. 4 (Sept. 1966), pp. 89-92. Polakoff, Murray E. “Federal Reserve Discount Policy and Its Critics,” in Banking and Monetary Studies, ed. Deane Car son. Homewood, Illinois: Richard D. Irwin, Inc., 1963, pp. 190-212. ---------- . “Reluctance Elasticity, Least Cost, and Member Bank Borrowing: A Suggested Integration,” Journal of Finance, vol. 15 (Mar. 1960), pp. 1-18. Polakoff, Murray E., and Silber, William L. “Reluctance and Member-Bank Borrowing: Additional Evidence,” Journal of Finance, vol. 22 (Mar. 1967), pp. 88-92. Rolph, Earl. “Discussion,” American Economic Review, Papers and Proceedings, vol. 45 (May 1955), pp. 411-14. Roosa, Robert V. “Credit Policy at the Discount Window: Com ment,” Quarterly Journal of Economics, vol. 73 (May 1959), pp. 333-38. ---------- . “Monetary Policy Again: Comments,” Bulletin of the Oxford University Institute of Statistics, vol. 14 (Aug. 1952), pp. 253-61. Samuelson, Paul A. “Recent American Monetary Controversy,” Three Banks Review, No. 29 (Mar. 1956), pp. 3-21. ---------- . “Reflections on Monetary Policy,” Review of Economics and Statistics, vol. 42 (Aug. 1960), pp. 263-69. Simmons, Edward C. “A Note on the Revival of Federal Reserve Discount Policy,” Journal of Finance, vol. 11 (Dec. 1956), pp. 413-21. ACADEM IC LITERATURE ON DISCOUNT MECHANISM ----------. “Federal Reserve Discount-Rate Policy and MemberBank Borrowing, 1944-50,” Journal of Business of the University of Chicago, vol. 25 (Jan. 1952), pp. 18—29. Smith, Warren L. “The Discount Rate as a Credit-Control Weapon,” Journal of Political Economy, vol. 66 (Apr. 1958), pp. 171-77. ----------. “The Instruments of General Monetary Control,” Na tional Banking Review, vol. 1 (Sept. 1963), pp. 47-76. Tobin, James. “Towards Improving the Efficiency of the Mone tary Mechanism,” Review of Economics and Statistics, vol. 42 (Aug. 1960), pp. 276-79. Walker, C. E. “Discount Policy in the Light of Recent Experi ence,” Journal of Finance, vol. 12 (May 1957), pp. 2 2337. Whittlesey, Charles R. “Credit Policy at the Discount Window,” Quarterly Journal of Economics, vol. 73 (May 1959), pp. 207-16. Young, Ralph A. “Tools and Processes of Monetary Policy,” United States Monetary Policy. New York: The American Assembly, Columbia University, 1958, pp. 13-48. Youngdahl, Richard. “Monetary Policy in Recent Years,” Amer ican Economic Review, Papers and Proceedings, vol. 45 (May 1955), pp. 402-08. SUMMARY OF ISSUES RAISED AT THE ACADEMIC SEMINAR ON DISCOUNTING Priscilla Ormsby Board of Governors of the Federal Reserve System Contents Introduction______________________________________________________________ 49 Role of the Discount Mechanism_______________________________________________ 50 Reallocation of Reserves_____________________________________________________ 50 Regional reserve reallocation Sector reserve reallocation Use of Discount Rate to Control Volume of Borrowing______________________________ 47 53 SUMMARY OF ISSUES RAISED AT THE ACADEMIC SEMINAR ON DISCOUNTING INTRODUCTION On May 11, 1966, an Academic Seminar on Changes in the Discount Mechanism was held at the offices of the Board of Gov ernors of the Federal Reserve System in conjunction with the “fundamental reap praisal of the discount mechanism” under way within the System. This paper repre sents an attempt to organize and sum marize what was by design a far-ranging and unstructured exchange of ideas and opinions at that seminar. The issues dis cussed there have been explored much more extensively in the academic literature by these same professors and by others. How ever, this paper presents the arguments only as they developed during the seminar and does not evaluate them— other than to com ment at times on the course of the seminar discussion, or to trace their origin. The following professors participated in the seminar: Professor Lester V. Chandler, Prince ton University, Chairman Professor G. L. Bach, Carnegie Insti tute of Technology1 Professor Edward E. Edwards, In diana University Professor Hyman Minsky, Washington University Professor Franco Modigliani, Massa chusetts Institute of Technology Professor Paul A. Samuelson, Massa chusetts Institute of Technology Professor Richard T. Selden, Cornell University Professor Warren L. Smith, University of Michigan In addition, a large number of academi cians submitted brief papers on the gen eral topic of the role of the discount mech anism. The ideas presented in those papers appear in this summary only insofar as they were again reflected at the seminar itself. The first section reviews, in a general way, the present and possible future roles of the discount mechanism: whether it is neces sary, and if so, what purposes it should serve. The other sections contain detailed considerations of the two major issues dis cussed at length during the seminar: (1 ) the role of discounting in the reallocation of reserves; and (2 ) the use of the discount rate to control the volume of borrowing. A number of connected issues, which while important in themselves were treated only peripherally in the course of the seminar, are also discussed in these sections. They include the influence of existing banking structure on credit needs and the operation of the discount window, the relationship of discounting to general monetary policy, an nouncement effects of changes in the dis i Now of Stanford University. 49 50 count rate, and nonprice rationing at the window. For the most part, the ideas brought forth were not restricted by the working of the present discount mechanism and there fore they should not be evaluated in terms of laws and regulations that are in existence today. ROLE OF THE DISCOUNT MECHANISM The participants in the seminar were dis satisfied with the discount mechanism as it currently existed; the two chief complaints concerned “nonprice rationing” and “an nouncement effects.” However, unsatisfac tory as the present discount mechanism might be in the estimation of the partici pants, there was little sentiment at the semi nar for its complete elimination. The sug gestion was made by several participants early in the discussion that perhaps, in the interest of tightening up aggregate mone tary controls, the discount mechanism could be dispensed with. However, the suggestion was not pursued and subsequent discussion lent no support to the proposal. One major reason seen for keeping the discount mechanism was uncertainty about the future. It was noted that the banking system is constantly changing and what seems superfluous today may become vital tomorrow; for instance, if the banking system were to run out of the assets em ployed in open market operations, the win dow could become the major source of re serves. The possibility was also cited of the window becoming important in a changing political climate where oral suasion became the order of the day. Although no one felt that, in today’s economy, the discount mechanism should be used as the major tool of monetary policy, a number of possible roles for the discount window were suggested by the par ticipants. Some of these suggestions— with varying degrees of support— were to pro vide a safety valve for correcting mistakes in open market operations, to permit ad justments by the individual banks and re gions and to protect the unit banking sys tem. It was noted in this discussion that the discount window could simultaneously serve a variety of purposes and thus need not be limited to a single, narrow role. REALLOCATION OF RESERVES One of the most basic and widely accepted functions of the discount mechanism is to provide temporary assistance to individual banks and regions in adjusting to changing reserve pressures. Thus, in a sense, shortrun reallocation of reserves is inherent in the window’s operation, and none of the professors questioned this sort of realloca tion. The desirability of permitting longerterm reallocation of reserves through the discount window was regarded as much more controversial, posing problems of both a political and an economic nature. While not totally separable, the realloca tion problem falls into two categories— re gional reallocation and sector reallocation. Although many of the same considerations apply to both, the pattern of the discussion at the seminar seemed to warrant separate treatment in this paper. ISS U E S RAISED AT SEM INAR ON DISCOUNTING Regional reserve reallocation A basic consideration in evaluating the need for regional reserve reallocation is the rela tive freedom of capital flows among differ ent parts of the country. The rapid growth of the California economy in recent decades was pointed to as demonstrating the ade quacy of this flow. It was estimated that, during the period of most rapid growth, about 40 per cent of the money in mort gages came from out of State. It was ques tioned whether one could generalize from this experience, however. The example was cited of an agricultural area in which meth ods of farming were becoming increas ingly capitalistic; however, deposits were not growing at a pace sufficient to finance this potentially profitable trend and no ap parent means, such as Federal insurance of farm loans, existed for drawing in the neces sary funds from other areas. It was agreed, however, that capital flows have become somewhat more flexible and responsive throughout the country since the early 1930’s, due to the development of Govern ment-sponsored protective measures, such as deposit insurance for both commercial banks and thrift institutions and mortgage insurance through the Federal Housing Au thority and also the Veterans Adminis tration. One of the professors severely criticized the one-way flow of capital that he con tended was encouraged by the present dis count mechanism. According to his anal ysis, the New York money market banks “raided” the small country banks, drawing funds away— mainly through the use of certificates of deposit— by paying higher interest rates than the small banks could af ford. He would resolve this inequity by having the Federal Reserve lend liberally to the money market banks, satisfying their 51 demand for funds and keeping them out of the small banks’ markets. To offset the re sulting reserve creation, the Federal Re serve would sell Government securities in the open market. It is conceivable that these securities could be bought directly by cus tomers of the country banks, resulting in the same loss of funds on their part; but in con trast to the case with CD’s, such purchases would not involve a personal commitment of customers to a specific money market bank. This proposal met with very little sup port from the other professors. According to traditional economic logic— and assum ing away any barriers to credit mobility— the fact that the New York banks could pay a higher interest rate indicated that the funds “belonged” in New York. It was also noted that the above analysis was con cerned with a “bank” allocation problem, which might be quite distinct from the “customer” allocation problem. If the de positors sending their money to New York could, with equal ease, obtain loans from the New York banks, the net result for the region might be beneficial. If this were true, then the capital flow would really not be one way; it would be merely bypassing the local bank. In the final analysis, however, a majority of the participants felt that the “small bank problem” probably existed in some degree. A second solution was offered that would create a dichotomy of banks— money mar ket banks and nonmoney market banks. For the money market banks, the Federal Re serve would adopt something similar to the British technique of discounting— denying them access to the window and fostering the development of market operators who would use the window. The nonmoney market banks would retain access to the 52 window at a rate “considerably higher than the money market rate.” This suggestion likewise elicited little support from other participants. It was noted that one of the underlying causes of whatever small bank problem existed was the currently existing banking structure. Those banks that were handi capped were typically unit banks and as such were probably inherently limited in the level of fundraising efficiency they could attain relative to the larger banks. Mixed views were apparent among the participants as to whether the Federal Reserve should work toward the liberalization of branch ing laws, should protect the traditional unit banking system, or should take any action in the area. Sector reserve reallocation The most commonly cited purpose of sector reallocation of reserves was to bypass mar ket forces and to insulate part of the econ omy— the most frequently cited example is probably the homebuilding industry— from general monetary policy. This was generally envisioned as being undertaken to offset im perfections already existing in the markets and to ameliorate what would otherwise be a disproportionately large impact of policy decisions on specific sectors. The Federal Reserve might offer such selective credit assistance through the indirect method of accepting the paper of those sectors for discounting by member banks on a more liberal basis, perhaps at a preferential rate. The professors saw problems with this ac tion, apart from the question of its desir ability. It was pointed out that it was in fact a reincarnation of the commercial loan theory, long since proven ineffective. Unless the window were to accept the specified paper on a massive and perhaps unlimited scale— a policy that could have serious con sequences for monetary management, especially at a time when the over-all pos ture of the System was probably one of tightness— there was very little assurance that the funds thus provided would be used for the desired purpose. Other possible actions of the System in cluded the subsidizing of various agencies that support specific sectors, such as the Federal National Mortgage Association, or the direct purchase of the paper of specific sectors.3 The overwhelming sentiment at the seminar, however, was toward keeping any assistance as indirect as possible. None of the professors felt that the Federal Re serve had a responsibility to support any sector on a long-term basis. Perhaps if an interest rate were obviously out of line the System would be justified in stepping into the relevant market temporarily, but the border between a temporary situation and a fundamental trend is necessarily hazy, and it was noted that direct assistance to one segment of the market, even in an extreme situation, could set a precedent that would result in an increasing number of requests for such assistance. The professors therefore favored con tinuing the present system of establishing separate agencies, not endowed with the power of reserve creation, to foster spe cific sectors. This left open the question of how deeply the Federal Reserve should in volve itself in assisting these agencies; no one doubted that the System would have a responsibility to protect them from com plete failure, but the professors would pre fer to see the Federal Reserve remain, as much as possible, in the role of lender of last resort. 3 This statute. course would require a change in the ISSU ES RAISED AT S E M IN A R ON DISCOUNTING 53 USE OF DISCOUNT RATE TO CONTROL VOLUME OF BORROWING If a single recommendation could be said to have come out of the academic seminar, it would be for the Federal Reserve to make more and better use of the discount rate as a means of rationing credit. The present rate system was almost unanimously criticized, and most of the professors recom mended that the discount rate be tied to some market rate. Recommendations were also made for graduated rates based on the amount of borrowing. Finally, a number of specific models using rate as the principal control device were recommended and dis cussed. The major criticism of the present rate system was of the ambiguous “announce ment effects” of a rate change. Without careful inspection of market rate patterns — and sometimes even with such inspection — it can be difficult or impossible to deter mine whether the Federal Reserve is lead ing the market and opening a new phase in monetary policy or lagging the market and merely adjusting to existing conditions. The value of an announcement effect was not completely rejected, however; it was pointed out that it might be extremely im portant in restoring international confidence in a shaky currency, A number of recommendations were made to permit less ambiguous announce ment effects. The simplest suggestion was for the Federal Reserve to issue a statement saying exactly what it wanted to convey. Such a direct method would seem to offer less chance for misinterpretation, but it would have drawbacks of its own. A state ment agreed to by seven Governors or 12 members of the Federal Open Market Com mittee would, by the very facts of human nature, almost unavoidably be rather bland, and even the most clear-cut statement would probably have less effect than a lasting change in rate. A second proposal was to establish a regular schedule of changes in the discount rate. These changes would be frequent and very small and therefore should be ac cepted and almost unnoticed by the public, but they would allow the discount rate to keep pace with changing market rates. When the time came to announce a change in monetary policy, a relatively large change in the discount rate should accomplish this without confusion. If the sole criticism of the present rate system were the ambiguity of announcement effects, adoption of this proposal would probably be viewed as a major improvement by the academic com munity. The final proposal for improving the announcement effect assumed that the dis count rate would be tied to some market rate and vary with that rate automatically. Any desired announcement effect could be achieved by changing the differential be tween the two rates. The impact of a change in this case might be even stronger than in the previous proposal, since it would be a voted and announced change after a period of automatic and continuous adjustment. Although the ambiguity of announce ment effects was the most frequently men tioned criticism of the present rate system, most of the professors saw other benefits that would result from a tied-rate policy. One was simply an appeal to the principle of parsimony; it is wasted effort to control the money supply through open market op erations and set the discount rate more or less independently when the work could be cut in half by letting the market handle the second task. 54 Other more positive benefits suggested for the tied rate included the insurance of a nationally determined rate automatically and a stabilization in the amount of bor rowing. There was some doubt as to whether the amount of borrowing would be an invariant function of the spread between the controlling market rate and the discount rate, but it was agreed that shifts in this relationship could probably be predicted and offset. Assuming the desirability of a tied dis count rate system, a number of specific problems were identified in the establish ment of that rate. These included: the choice of the market rate to which the dis count rate would be tied; whether the dis count rate should be a penalty rate, and if so, just what constituted a penalty; and whether there should be a freely open win dow at the established rate and what the implications of such an arrangement would be for general monetary policy. The deciding factor in the choice of a controlling market rate was deemed to be the manner in which commercial banks were financing their positions— that is, the most typical source of short-term funds, apart from the discount window, to which they turned to adjust to changing reserve pressures. It was recognized that, in the cur rent financial environment, a variety of such sources were used by the banks and there was therefore no one obvious answer to the question. However, several nominations were made for the role of controlling mar ket rates. Historically, the most logical seemed to be the Treasury bill rate; here was an extremely well-organized market for an almost universally held instrument. However, present trends suggested that the sale of bills was becoming— and in some cases had already become— obsolete as a means of bank reserve adjustment. For many banks, almost the whole of their de clining holdings of Government securities are tied to specific purposes, such as col lateralizing public deposits. A second possibility was to tie the dis count rate to the rate paid on certificates of deposit, which is of increasing importance but is still not a universally important rate for commercial banks. Another problem in this case was seen to arise from the fact that CD’s are normally for maturities signifi cantly longer than the typical adjustment borrowing at the discount window. The Federal funds rate, also suggested, was not regarded as a significant rate for all banks and also was highly volatile in the short run. While not seriously proposed, it was sug gested that the only really relevant rate for some very small country banks might be the rate they were making on their loans. A proposal made earlier, to tie the discount rate to some measure of the individual bank’s profit rate, was rejected almost with out discussion, apparently because this would not provide the single nationally de termined price for reserve credit that they felt to be important. There was also some discussion as to whether the rate chosen really mattered. It was pointed out that, in all but the very short run, all the rates proposed were highly correlated. Thus the choice of a base was somewhat immaterial. However, it was recognized as important that the initial value of the discount rate be appropriately set in relation to a market rate that was significant for all commercial banks. The wrong choice of the initial level was seen as having possibly disastrous results, if with that choice the Federal Reserve relin quished further control of the rate. For in stance, if the discount rate were set 50 basis points above the Treasury bill rate ( 4 Vi per cent) and banks were actually ISSU ES RAISED AT SEMINAR ON DISCOUNTING financing their positions with CD’s at 5 Vi per cent, the discount rate would be V2 of 1 percentage point below the relevant rate. With a freely open window, the money sup ply would increase drastically until banks were out of CD’s and financing their posi tions at the discount window. The second question relating to discount rate policy was whether the rate should be at a “penalty” level relative to market rates. One of the participants, a consistent advo cate of increased discount-window use, recommended that it actually be set below market levels in periods of tight money. The other professors present agreed that it should be a penalty rate, but were hesitant to commit themselves to specific figures, which they regarded as incidental to the concepts they were developing and better worked out in practice. There did seem to be general support for a penalty in the vicinity of V2 of 1 percentage point— some thing that would provide a deterrent to bor rowing, but would “maintain the virtues of a system that permits the individual bank adjustment possibilities.” It was expected that, should the Federal Reserve adopt a tied penalty rate, a period of experimenta tion would be required before the appro priate differential could be judged. It was suggested that, even with a freely open window, discounting should be worth more than other methods of obtaining funds, since it provided increased liquidity to the banking system as a whole. Therefore, perhaps the entire differential of the dis count rate above the market rate should not be regarded as a penalty. Although participants in the seminar were almost unanimously in favor of a tied rate, there was some doubt expressed as to how much such an arrangement would ac tually accomplish. It was pointed out that the Federal Reserve had a major effect, 55 through its open market operations, in de termining the level of market rates, so perhaps it was wrong to speak of an in dependently determined discount rate. In the final analysis, the System played a major role in both money supply policy and in terest rate policy, regardless of how it chose to exercise these roles. The question of an open window with a tied rate was recognized as important for general monetary policy. The problem in the case of a poorly chosen discount rate was discussed above. But even if the dis count rate were set above the appropriate market rate, possible problems were fore seen. The increased cost of credit at the window had to be looked at in conjunction with the increased availability. Basically, no matter how high the rate was set, the central bank gave up direct control of the volume of reserves created and supplied through the discount window. Advocates of an open window pointed out that any undesired reserve creation through the discount window could be off set by open market operations. This was fairly generally accepted as true in princi ple— at least in the absence of a shortage of assets employed in such operations, seen by some as a possibility— but it was felt that it could require continuous offsetting operations and, in the extreme case, on a massive and unheard-of scale. Some noted that in the proper environment an open window could have the advantage of avoid ing great scrambles for funds that some times result in disorderly rate patterns. None of the professors advocated a per manently open discount window at a rate that the Federal Reserve could not control. The System should always have the option of increasing the rate spread if borrowing became obviously excessive. Two more continuous methods of controlling the vol 56 ume of borrowing within a tied-rate system were suggested: (1 ) a graduated penalty rate based in some way on the amount of borrowing; and (2) the kind of administra tive, nonprice rationing employed today. The first possibility, while not actively supported by all those at the seminar, was unanimously viewed as feasible. Advocates of a graduated rate schedule variously sup ported two versions; the more popular would base the schedule on the amount of borrowing by the individual bank. This was seen as consistent with the fact that the Sys tem was trying to influence individual deci sions and as more effective from an alloca tive point of view. The alternative version would base the penalty schedule on the aggregate amount of borrowing by member banks. It was felt by some that such an arrangement would avoid penalizing an in dividual bank that was the innocent victim of an adverse situation rather than a guilty party, or a bank with unusually profitable opportunities which was therefore justified in unusually large borrowing. Nonprice rationing was rejected by the seminar participants more by omission than commission. The general feeling seemed to be that it was too personal and apt to be arbitrary. The only statement offered in its defense was that it might prove useful if, because of circumstances beyond the Sys tem’s control, the U.S. economy found it self in an extreme inflationary spiral. Under anything approaching normal circum stances, however, the professors were unan imously opposed to its use. The remainder of this section describes a number of specific models using rate to control the volume of member bank bor rowing that were proposed by seminar par ticipants. None of these represents a closed and self-sufficient system, but they never theless represent a more structured level of thinking than the earlier discussion. In the first model proposed, the discount rate would be tied to an average for the past 2 weeks of the Federal funds rate. This rate was chosen because the Federal funds market is, like discounting, a short-term source of funds and is, therefore, for the member bank— although not for the Sys tem— the closest alternative to borrowing at the window. The discount rate would be set 100 basis points above this past average of the Federal funds rate. With a freely open discount window, this would limit swings in the Federal funds rate by creating a ceiling on that rate. In the short run, this could create a spiral situation, since there would be a tendency for the Federal funds rate to increase until it was equal to the dis count rate. Thus, in the extreme, the dis count rate would equal a past value of it self plus the differential. However, in prac tice this tendency could be controlled by open market operations conducted in such a way as to reduce the banks’ necessity to stay in debt, and it would have a natural limit since, if the rate became too high, banks would presumably find it desirable to adjust their basic positions by such methods as calling loans and selling off other assets. The second model had as its immediate goal the stabilization of the amount of free reserves that, it was argued, would “give a fairly rigid relation between what the cen tral bank directly controls and the total amount of reserves.” The amount of bor rowing would be controlled by tying the discount rate to a market rate— the rate was not specified in this case. The discount rate would be above this and would in crease with extensive use of the window; the term “extensive” probably encom ISSU ES RAISED AT SEMINAR ON DISCOUNTING passes both duration and amount of borrow ing. The amount of excess reserves would be controlled by paying interest on them. This rate would also vary with a market rate— and here the Federal funds rate was spec ified— but would be below that market rate. Advocates of this model would even go so far as to impose a penalty on the holding of excess reserves if the Federal funds rate fell that low. A somewhat more conventional model proposed would employ a nationwide dis count rate tied to and slightly above some market rate. Proponents of this model chose the Treasury bill rate for this purpose but felt that the choice was somewhat arbitrary. They would decrease over time the amount of nonprice rationing and eventually elimi nate it altogether. The rate could be ad justed as the system was perfected, but mas sive change would be employed only when 57 the Federal Reserve wanted to create an “announcement effect.” The final model would provide for auto matic determination of the discount rate, not by tying it to any one market rate but rather by having the System regularly auc tion a specified amount of borrowed re serves in a manner somewhat analogous to the weekly Treasury bill auction. The Sys tem could then directly control the money supply and would have the added advantage that the discount rate would be auction determined and would not be tied to any market rate that could become obsolete. This arrangement could result in tremendous rate instability, however, in cases where unusual circumstances caused a great scramble for funds. To offset this, a penalty rate could be instituted, above the auction rate, at which reserves were available for the rest of the period, or very frequent— even daily— auctions could be held. SOME PROPOSALS FOR A REFORM OF THE DISCOUNT WINDOW Franco Modigliani Massachusetts Institute of Technology Contents Goals to be Achieved________________________________________________________ 61 Proposed Devices to Improve Federal Reserve Control Over the Money Supply While Permitting Unrestricted Use of the Window_____________________________ 61 Sources of slippage between nonborrowed reserves and the supply of demand deposits, and how to reduce them Outline of proposed reform—basic features Elaboration of the proposal Behavior of free reserves and supply of demand deposits under the proposed system Choice of penalty rate—case for a sliding differential Possible minor improvement: payment of interest on excess reserves Proposal for Special Borrowing Facilities Aimed at Improving the Spatial Allocation of Bank Credit-------------------------------------------------------------------------------------------- 72 Proposal outlined How the term window could contribute to allocative efficiency Operational aspects How large a target volume of borrowing? Some minor complementary suggestions 59 SOME PROPOSALS FOR A REFORM OF THE DISCOUNT WINDOW GOALS TO BE ACHIEVED The purpose of this paper is to outline several proposals for a reform of the Fed eral Reserve discount window. These pro posals are aimed at achieving the follow ing major goals: 1. To eliminate the discretionary and sometimes capricious elements that charac terize the present administration of the win dow by permitting unrestricted use of such borrowing facilities by all creditworthy bor rowers that are willing to pay the discount rate. 2. To reduce the slippage that exists between nonborrowed (bank) reserves and the supply of demand deposits by reducing swings in free reserves, especially those of a procyclical character, and thus improve control of the Federal Reserve over the money supply and interest rates. 3. To make available to smaller banks facilities analogous to those provided by the markets for Federal funds and certif icates of deposit, from which these banks are now partially or totally excluded be cause of the small size of their operations. 4. To contribute through goal 3 and other devices to an improved spatial alloca tion of bank credit. It is believed that the proposed reform would also make possible the achievement of two other goals: 5. To eliminate the announcement ef fects that result from sporadic and hence sizable changes in the discount rate, and 6. To provide a stronger inducement than now exists for banks to become mem bers of the Federal Reserve System, which would contribute to goal 2. PROPOSED DEVICES TO IMPROVE FEDERAL RESERVE CONTROL OVER THE MONEY SUPPLY WHILE PERMITTING UNRESTRICTED USE OF THE WINDOW This section discusses six devices that it is believed would improve Federal Reserve control over the money supply while permit ting unrestricted use of the discount win dow. serves, which the Federal Reserve controls largely through open market operations, and the supply of demand deposits can be traced primarily to variations in free re serves. Our primary concern in this section is with methods of reducing such variations while at the same time keeping the discount window open to all creditworthy borrowers willing to pay the price. Other major sources of slippage— such Sources of slippage between nonborrowed reserves and the supply of demand deposits, and how to reduce them As is well known, the slippage that exists between the volume of nonborrowed re 61 62 as variations in the reserve ratio as a result of shifts of deposits between banks with different reserve requirements and between member and nonmember banks, changes in time deposits, and currency drain— would not be affected by the proposed re forms except indirectly through goal 6 to be achieved (see description on the pre ceding page). The operation of the discount window obviously affects free reserves through bor rowing. It is reasonable to suppose that the volume of bank borrowing is influenced in part by the profitability of borrowing, as measured by the spread between the dis count rate and short-term market yields; this supposition is supported by the em pirical evidence. It is also generally agreed that a rise in aggregate demand and eco nomic activity tends initially to be accom panied by a rise in short-term market yields unless it is accommodated by a commen surate expansion of the money supply.1 Under these conditions, as long as the dis count rate is kept unchanged, a rise in de mand tends to increase the profitability of borrowing; and since the rise in market rates also tends to reduce the demand for excess reserves, the result is likely to be a reduction of free reserves and thus a pro cyclical movement in the money supply, relative to nonborrowed reserves. Under the present system this tendency is, of course, moderated by various limita tions on the use of the discount window through regulations, frowns, and suasion. Such administrative limitations in turn seem unavoidable so long as the discount rate is changed only infrequently, permitting siz able fluctuations in the spread between it and short-term market yields and corre sponding variations in the incentive to bor row. Furthermore, since changes in the dis count rate have tended to occur infre 1 See the discussion on pp. 64-69. quently, and often only after some debate within the Federal Reserve System, they have come to acquire a symbolic meaning (even if often a rather obscure one) apt to generate wide repercussions. And this very feature in turn has contributed to the practice of avoiding frequent changes in the rate. Clearly the source of slippage between nonborrowed reserves and the money sup ply described above would be reduced if the discount window could be redesigned so as to minimize fluctuations in the incen tive to borrow. The simplest way to achieve this result, of course, would be to shut the window altogether. But this solution is clearly inconsistent with preserving the es sential role of the central bank as a lender of last resort. Individual banks must have an outlet to which they can turn in case of “justified need” and, similarly, some methods must be provided by which the banking system as a whole can manage to satisfy reserve requirements in a way that is not unreasonably painful. However, there is no reason in principle why borrowing from the discount window should not involve some significant penalty. Accordingly, one possible device to limit substantially the use of the window would be to set the discount rate at some level substantially above short-term market rates. There would then be an incentive for banks to avoid the risk of having to borrow and to repay promptly any borrowing that might have been incurred due to miscalculations of or unanticipated contractions in non borrowed reserves. Yet banks could be al lowed unrestricted use of the window sub ject only to normal and prudent standards of creditworthiness, for use of the window would be limited by the cost of borrowing, without any need for fiat or frowns. However, this approach has two major drawbacks: (1) Since the “penalty” would PROPOSALS FOR REFORM OF DISCOUNT WINDOW depend on the relation between the dis count rate and market rates, it would still be necessary, in order to keep the penalty reasonably uniform over time, to change the discount rate from time to time. And that would perpetuate the announcement effects. (2) The method would in effect discriminate against small banks, which cannot make effective use of the Federal funds market as a source of funds. Indeed, if the banking system as a whole were out of debt, which presumably would be the normal circumstance under a penalty-bor rowing rate, the Federal funds rate would tend to hover below the discount rate and around short-term market yields, say the rate on 3-month Treasury bills (hereinafter referred to as 3-month bills, or in some in stances, bills). Thus, individual banks hav ing access to that market could make up their deficiencies at that cost. Yet the smal ler banks would have to pay the signifi cantly higher penalty rate. It is suggested that these shortcomings could be eliminated, while retaining the basic idea of a wide-open window at a pen alty rate, by reorganizing the operation of the window along the lines set forth in the remainder of this section. Outline of proposed reform— basic features 1. The window would be open to all borrowers willing to pay the discount rate as long as they met some appropriate tests of creditworthiness. To avoid uncertainties, each bank would be informed about the maximum amount of accommodation that it could expect to receive. The ceiling would be reviewed at stated intervals, ex cept under special circumstances requiring a reappraisal of the bank’s credit standing. 2. The borrowing rate would be tied to a short-term market rate, say for the mo ment, the 3-month bill rate. This device would eliminate sizable, discontinuous 63 changes in the discount rate and associated announcement effects. 3. To maintain the penalty character of the window, the borrowing rate would be fixed at, say, last week’s bill rate plus a fixed number of basis points, or plus a fixed per centage. Considerations relevant in setting the size of the penalty are set forth later. 4. Borrowing at the window would be for very short terms— usually for a single day— although automatically renewable at the option of the borrower. 5. To avoid discrimination against smaller banks, the Federal Reserve would provide, for such banks, accommodations similar to those obtainable through the Fed eral funds (hereinafter abbreviated FF) market. Specifically, those entitled to the special accommodation would be allowed to borrow at the window at a daily rate equal to that day’s average FF rate plus a com mission, consisting of a fixed but moderate number of basis points or a moderate per centage charge, as noted above. This facil ity could be provided for banks not exceed ing a certain size, or for those at particular locations, or perhaps more equitably, for loans not exceeding a stated modest size. If the last device were adopted, one might expect that this facility would, in fact, be used only by the smaller banks with in adequate access to the FF market. Elaboration of the proposal It should be noted that the reform outlined above is only part of a broader plan. In deed, what has been proposed so far would be of no help in achieving goal 4 — im proved spatial allocation of bank credit. To that end there is a separate proposal, de scribed in the last section, to provide facil ities for longer-term borrowing. Accord ingly, the rest of this section is concerned only with the operation of the “1-day win dow.” 64 The first questions that need to be con sidered are: To what rate should the dis count rate be tied? How large should the premium be? These two questions are closely interrelated. Clearly, it would be desirable to anchor the discount rate to the yield of some market instrument of major importance— one that has a broad, wellorganized market. This would insure that the chosen rate would be “representative” and relatively free of erratic movements. From this point of view, the 3-month bill rate would seem to be an obvious choice, at least under present arrangements. There are, however, two related prob lems to be considered. First, any specific instrument may, at times, reflect special in fluences. Second, there are some delicate issues involved in tying a 1-day rate to a 3-month rate, if at the same time borrow ing is unrestricted. In particular, “term structure effects” (for example, expectations of a fall in the 3-month rate) might make it profitable to borrow short at a rate negligi bly higher than the 3-month rate. To avoid these problems, it would seem desirable to peg the discount rate substan tially above the 3-month bill rate. One relevant guide in deciding on the size of the premium is provided by the consideration that, with a truly open discount window, the discount rate, by and large, would set the ceiling for the FF rate. In other words, as the FF rate approached the discount rate, the demand for funds would become highly elastic as would-be borrowers turned to the window. This consideration suggests that the premium should be sufficiently large to allow the FF rate to deviate from the bill rate as much as might be justified by term structure and other special circumstances affecting the bill rate, without making it “profitable” to borrow at the window. While it is impossible to set an absolute limit, the above considerations suggest a premium on the order of 100 basis points, a margin somewhat larger than the largest amount by which the weekly average FF rate has exceeded the 3-month bill rate in recent years. (Unfortunately, this experi ence is a very limited one since, as is well known, until early 1965 habits and con vention prevented the FF rate from being bid above the discount rate; this, of course, also tended to distort the relation of the FF rate to other rates.) With such a differential, one could ac commodate substantial variability in the proper relation between the FF rate and the short-term rate to which the discount rate was tied, without creating incentives to borrow at the window and hence without causing undesirable flurries in the volume of borrowing. An alternative, and probably more effective, device to guard against this source of difficulty is discussed below. It would be to rely on a floating differential. But first, it is well to examine more closely both the long-run and short-run be havior of the proposed system,— assuming the premium over the bill rate to be sub stantial but of fixed size. We propose to show that, under this system, free reserves would tend to fluctuate rather narrowly around a substantially constant “equilib rium” level. In other words, while fluctua tions of free reserves would not be— and should not be— altogether eliminated, de viations of free reserves from the constant equilibrium level would set up strong forces tending to move these reserves back toward equilibrium. Behavior of free reserves and supply of demand deposits under the proposed system The analysis that follows is based largely on some very definite views as to the major forces that shape banks’ portfolio manage ment and their use of the discount window, as well as the behavior of short-term mar PROPOSALS FOR REFORM OF DISCOUNT WINDOW 65 to their equilibrium level tends to occur gradually over time— somewhat along the lines of the textbook description of the process of expansion of deposits in response to an initial disequilibrium. In addition, the process of adjustment gets disturbed by the changes under (a). Thus free reserves tend to be high when there is an unforeseen in crease in nonborrowed reserves or an un foreseen slackening in the demand for com mercial loans, and to be low when the un foreseen changes are in the opposite direc tion. The evidence referred to above also sup ports the view that changes in short-term market yields (say, the 3-month bill rate or the commercial paper rate) are accounted for largely by the interaction of the supply of demand deposits, controlled by the forces outlined above, and the demand for demand deposits, which is basically controlled by current and past levels of income and short term market yields. It further suggests that in the short run (say, a quarter or less) the level of income is largely unaffected by variations in the money supply or short term interest rates (at least as long as these remain within realistic limits). It therefore follows that, in the short run, the level of short-term market yields is controlled, in the last analysis, by the behavior of the out standing stock of demand deposits. In light of the above interpretation of bank behavior, we can examine first what would happen to the “equilibrium” level of free reserves under the proposed system. To this end, it is convenient initially to ignore operations at the special discount window provided for small borrowings. It should be evident that by floating the 2 See especially Modigliani, Rasche, and Cooper, discount rate sufficiently above short-term “Central Bank Policy, the Money Supply, and the market yields the desired or equilibrium Short-Term Rate of Interest.” Journal of Money, Credit, and Banking (May 1970). Other results are level of borrowing can be brought essen contained in yet unpublished memoranda of the proj tially to zero, and this would be true regard ect; it is expected that these results will be published in the near future. less of the level of the short-term rate. (By ket yields. These views in turn appear to receive strong support from the empirical analysis of recent experience undertaken in the course of the MIT-Federal Reserve econometric research on the working of stabilization tools.2 This evidence supports the view that the volume of free reserves outstanding at any given time reflects two basic sets of forces: 1. An “equilibrium” component, to wit, the desired or equilibrium level of free reserves. This equilibrium level itself is the difference between (a) desired excess re serves, which depend on short-term market yields and tend to decrease as these yields increase, and (b) the optimum volume of borrowing at the window, which is basically controlled by the spread between short term market yields— such as the FF rate or the 3-month bill rate— and the discount rate. (Note, however, that the FF rate is an adequate measure of short-term yields only for the very recent period when that rate was not conventionally kept at, or below, the discount rate.) 2. A disequilibrium component reflect ing the inability and/or undesirability of banks to adjust instantaneously to un foreseen (or transient) changes in their de posits or in the demand for commercial loans. The unforeseen changes in demand deposits in turn reflect (a) unforeseen changes in nonborrowed bank reserves due to Federal Reserve operations and changes in currency holdings (and time deposits), and (b) the unforeseen effect on demand deposits of expansion and contraction in bank credit itself. Component (b) implies that, even in the absence of the changes under (a), the adjustment of free reserves 66 contrast, under the present system a rise in market yields increases the equilibrium level of borrowings unless, and until, coun teracted by a rise in the discount rate.) It then follows that the equilibrium level of free reserves would itself tend to be con stant except for the effect of market yields on excess reserves. However, this effect ap pears to be fairly moderate, except possibly for extremely low levels of market yields where “liquidity-trap phenomena” could become significant.3 Furthermore, even this effect could be eliminated by the device of paying interest on excess reserves, as is dis cussed in the last part of this section. But even without this device we are led to the conclusion that the equilibrium level of free reserves would tend to be stable (and pre vailingly positive) under normal condi tions, except for some tendency to decline mildly with the prevailing level of market yields. Finally, we may note that if the system were in a position of equilibrium, with bor rowings near zero and excess reserves in equilibrium, then one could expect the FF rate to hover around the bill rate. More precisely, in the absence of term-structure effects (that is, if short-term rates were anticipated to stay unchanged in the near future), it should tend to be quite close to the bill rate, although term-structure effects could cause it to deviate, within limits, on either side of the bill rate. This proposition seems fairly obvious and can be supported by a more rigorous analysis, which we need not spell out here.4 The above consideras Analysis of recent years suggests that an in crease of 100 basis points in short-term yields—in, say, the 3-month bill rate—tends to reduce excess reserves by somewhat less than $50 million within 1 month and by somewhat less than $100 million within one to two quarters. 4 The above conclusion rests on the assumption that bills will continue to represent an important com ponent of secondary reserves and sources of short term liquidity. Should this premise lose its validity and bills cease to be held by banks in significant tions have the following implication, which is important for an understanding of the workings of the proposed reform: provided the discount rate were set significantly above the 3-month bill rate, and if the sys tem were in a position of equilibrium, the FF rate could be expected to be signifi cantly below the discount rate. We can now reintroduce the special “Federal funds window” for small operators and show that this does not significantly change the above conclusion. We need ob serve only that those who are eligible to use the special window and who find it economical to do so would be borrowing at a rate differing only by a small commission from the rate available to any would-be borrower; namely, the FF rate. Since the relation between the FF rate and short-term yields was just shown to be such as to in duce the banking system as a whole to hold positive-free reserves, regardless of the level of short-term yields, we can infer that this relationship would provide the same incen tive to special borrowers as a whole. Hence, these borrowers would also tend to hold a positive and relatively stable amount of free reserves; these reserves would prob ably exhibit some tendency to move in versely, but moderately, with the level of market yields. Note, however, that the amount of borrowing at the special, in con trast to the regular, discount window would not be zero since some operators would on the average be borrowing there, just as many other banks would, on the average, be borrowing from the FF market. All we are asserting is that the net free reserve position of the group as a whole would tend to be stable.5 quantity—except for the purpose of satisfying col lateral requirements—then the bill rate would no longer provide a reliable yardstick of short-term yields and hence would no longer be a suitable rate on which to anchor the discount rate. 5 This note appears on opposite page. PROPOSALS FOR REFORM OF DISCOUNT WINDOW 67 (whether at the window or in the FF mar ket) relative to other short-term market yields would put pressure on banks to re duce their asset portfolios, thereby shrink ing the supply of demand deposits and re quired reserves, until the borrowing had been eliminated and free reserves had moved back to equilibrium. In the process short-term market yields would, of course, tend to move up, which is presumably what the Federal Reserve intended. But note that this rise would not per se reduce the pres sure for the banking system to get out of debt. Indeed, with the discount rate floating above the bill rate, a rise in the latter would not reduce the incentive for individual banks to avoid a net borrowed position. 2. Suppose instead that the Federal Reserve wished to expand the money sup ply and lower short-term market rates, and accordingly brought about an unanticipated expansion of nonborrowed reserves. Here initially free reserves would exceed the planned amount, causing an increase in sup ply and a fall in demand in the FF market. This would lower the FF rate relative to the bill rate, encouraging an expansion of banks’ portfolios and the money supply and leading to an increase in required reserves. The incentive to expansion would persist until free reserves had moved back to equi librium, and thus the FF rate had re-estab lished its equilibrium relation to the bill rate. Here again the bill rate would pre 5 At a more refined level of analysis, one should sumably fall in response to bank expan recognize that the incentive structure would be a little different for those operating at the special win sion, as intended. But this fall would not dow. In the first place, if banks were short of reserves, reduce the incentive to expand the money they would be paying somewhat more than the larger operator using the FF market. In addition, they supply until the additional reserves had been would probably earn less if they were long on re absorbed by higher required reserves; for as serves. In fact, small operators would probably tend to hold any excess funds in the form of excess re long as free reserves remained above equi serves yielding nothing, instead of lending them in librium, the FF rate would tend to remain the FF market where they would yield the FF rate less transactions costs. (However, these qualifica below the rate on bills and other short-term tions do not require modifying our conclusions that instruments. their “equilibrium” level of net free reserves would tend to be stable though probably somewhat higher 3. By relying on the reasoning devel than for the larger operators, and probably also some oped earlier, one can also readily establish what less responsive to variations in market yields.) We can now examine the short-run, dy namic behavior of the model in response to developments pushing it out of equilib rium. For analytical purposes we can distin guish between disturbances originating in the economy and those originating from Federal Reserve actions, although of course in general both types of disturbances may occur simultaneously and reinforce or offset each other. 1. Consider first the response of the banking system to a situation in which the Federal Reserve wished to hold down the money supply and raise short-term market rates. To this end the Federal Reserve would force an (unanticipated) contrac tion in nonborrowed reserves (relative to the normal seasonal and secular pattern). As a result, free reserves would initially fall short of the planned level. This implies an increase in the demand for, and/or a decrease in the supply of, funds in the FF market, which would immediately raise the FF rate. If the Federal Reserve action were sufficiently strong, the shortage of reserves would be such as to push the FF rate to the ceiling provided by the discount rate. At this point some banks would be induced to borrow at the window, thus acquiring the additional reserves needed to satisfy reserve requirements (plus the demand for excess reserves, probably somewhat reduced). But now the higher cost of borrowing 68 that the pressures and responses described in (1) and (2 ) apply equally to the subset of “small” operators having access to the special “Federal funds window.” In summary, an (unanticipated) expan sion or contraction of nonborrowed reserves would, initially, be reflected largely in op posite movements of free reserves and of the FF rate, relative to the bill rate. But this would be only a temporary and (appro priate) cushioning reaction. For the move ment of the FF rate in turn would generate incentives to actions that would tend to bring free reserves back to the initial equi librium (except for the small effect of the change in short-term market yields on ex cess reserves). With free reserves moving back to the original position, the supply of demand deposits would tend to move commensurately with the change in nonbor rowed reserves. The final change in short term market yields (that is, the bill rate) would then depend on the size of the change in nonborrowed reserves and the (shortrun) elasticity of demand deposits with respect to short-term yields. 4. Consider next the effect of an in crease in the demand for money— an up ward shift in the demand schedule relating money demand to short-term yields. This would tend to raise short-term market yields, unless the money supply increased. But there could be no significant increase in the money supply so long as the Federal Reserve kept the level of nonborrowed re serves unchanged. Indeed, under these con ditions, the money supply could expand significantly only through an increase in borrowing. But since the rise in the bill rate would be accompanied by a commensurate rise in the discount rate, there could be no incentive for banks to expand their borrow ings. With borrowing unchanged, free re serves would also be unchanged, except again for a moderate decrease in response to the higher market yields, and hence the money supply would be basically un changed, as stated above. Needless to say, if the Federal Reserve wished to prevent the bill rate from rising, it could do so by sup plying additional reserves, in amounts suf ficient to increase the supply of demand de posits pari passu with the increased demand. The reasoning can be repeated mutatis mu tandis, in the presence of a decrease in the demand for money and falling interest rates. 5. A different and very important type of disturbance originating from, the econ omy would be an (unanticipated) surge of demand for commercial loans. Banks, as suggested earlier, would initially tend to ac commodate the increase without a commen surate reduction in the rest of their port folio. Hence, the supply of demand deposits and required reserves would in the first in stance rise. But with nonborrowed reserves unchanged, the system would be short of reserves, and hence the FF rate would tend to be pushed to the discount rate ceiling, opening up the window to an amount of borrowing needed to satisfy reserve require ments. But again the increase in the cost of borrowing (whether at the window or in the FF market) relative to short-term mar ket yields would generate an inducement for banks to reduce their portfolios and the sup ply of demand deposits until the borrowing had been eliminated and free reserves had moved back to equilibrium. Of course, the reshuffling of bank portfolios would likely involve some net liquidation of bills and other market instruments to accommodate the expansion of loans, which would result in some increase in short-term market yields. But once more this would not modify the incentive for the banking system to get out of debt to the Federal Reserve since the discount rate would be moving pari passu with the bill rate. Similar conclusions hold mutatis mutan PROPOSALS FOR REFORM OF DISCOUNT WINDOW dis, if an unanticipated decline occurred in the demand for loans. The above analysis has one implication that is worth noting. It should be apparent that under the proposed reform a level of free reserves in excess of the constant equi librium level would tend to be accompanied by an expansion of the money supply. Fur thermore, the larger the excess, the larger the rate of expansion of the money supply would tend to be. Conversely, free reserves below that constant level would tend to be accompanied by a contraction of the money supply at a rate commensurate with the negative gap. The proposed reform would thus tend to validate a view of long stand ing that there is a direct, reliable association between the volume of free reserves and the rate of change of bank credit and the money supply. Yet, paradoxically, this view is not warranted under the existing set-up in which the equilibrium level of free reserves is not stable over time because of variations in the spread between the discount rate and market yields. It is not inconceivable that reliance on that unwarranted view may have been responsible for certain past failures in monetary management. Choice of penalty rate— case for a sliding differential We are now in a position to set forth the main considerations that would seem rele vant in setting the size of the differential be tween the discount rate and the bill rate, or other short-term rate, to which it was tied. It follows from the analysis of the pre ceding section that the essential implication of a large differential is that banks would tend to find it undesirable to stay sub stantially in debt for extended periods. But this means, in turn, that the volume of de mand deposits could be kept under close control by the Federal Reserve through its 69 control over nonborrowed reserves (and re serve requirements). At the same time, since a temporary shortage of reserves could push the FF rate as high as the discount rate, a large differential would imply the possi bility of substantial short-run variability of the FF rate and related very short-term market rates. By the same token, a small differential would imply more limited vari ability of the FF rate but at the cost of tolerating a larger and longer-lasting de parture of the money supply from the level determined by nonborrowed reserves— that is, in essence, a looser coupling between nonborrowed reserves and the money sup ply. The above considerations suggest that the choice of the differential would be de pendent in large part on one’s view concern ing the nature of the monetary mechanism. Those holding that the cutting edge of monetary policy rests on the effects of such policy on interest rates and related financial yields would presumably be led to favor a set-up that minimized unintended move ments of interest rates and hence to prefer a relatively small differential. On the other hand, those leaning toward the view that the money supply affects economic activity di rectly might well be led to favor a system that minimized unintended movements in the money supply, even at the cost of larger short-run fluctuations in interest rates. In my view, however, a reasonable choice of differential does not really require set tling the thorny issue about the nature of monetary linkages. For whatever one’s view on that issue, presumably it is generally agreed that departures from the intended course, whether of the money supply or of very short-term market rates, can have a noticeable effect on the economy if they persist— but not if they are ephemeral. And this is particularly true once the rules of the 70 game are well understood and stable and the participants have had a chance to adjust to them. Hence, insofar as purely transient dis turbances are concerned, the choice of the differential is unlikely to be of real conse quence. On the other hand, in the case of marked and/or persistent departures, the Federal Reserve would soon have to reach a decision as to whether the most suitable response involved modifying the interest rate target or the money supply target, or some combination thereof. In such circum stances, the choice of the differential would therefore control only the character of the short-run, semiautomatic response of the model, while the Federal Reserve made up its mind as to the appropriate eventual re sponse. In any event, the dilemma of choosing between a high or a low differential could be avoided by adopting a “compromise” system, which should prove largely agree able to both points of view. The compro mise would consist of tying the discount rate to the bill rate with a variable peg. Un der this scheme the differential would re main fixed at some base level as long as ag gregate borrowing at the discount window remained below some stated amount. But if borrowing were to exceed this amount, then the differential would rise with the volume of aggregate borrowing, according to a preestablished schedule. By making the base level of the differen tial relatively modest, moderate and tran sient variations in the demand for money could be absorbed by an elastic money sup ply, with minor effects on market yields. This feature appears especially desirable in light of the difficulty in determining reliably the demand-for-money schedule, and hence the supply of deposits appropriate to a cer tain level of short-term rates. Yet, larger and more persistent disturbances, while still initially accommodated at the window, would be accompanied by a larger increase in the differential cost of borrowing, which would put pressure on banks to eliminate rapidly at least a portion of their borrowing. Such disturbances would thus tend to be communicated promptly to interest rates, unless of course the central bank decided to accommodate the larger demand by increas ing the supply of reserves, thus reducing the effect on interest rates. Finally, it should be noted that while the schedule of penalty rates would influence the response to a tight money situation, such a schedule would have little influence in shaping the response to a loose situation, characterized by a rise in free reserves. That response would be controlled by the fall in the FF rate below market yields and by the speed with which banks would respond to this situation by expanding their portfolios of earning assets. Possible minor improvement: payment of interest on excess reserves I have noted, in setting forth the anticipated behavior of the banking system under the proposed reform, that some variation in free reserves would continue to be present because of the negative association between equilibrium excess reserves and short-term market yields. Even this source of variation in equilibrium free reserves could be largely, if not totally, eliminated by the simple de vice of paying interest on excess reserves at a rate tied to short-term market yields. It is suggested that the most effective ar rangement would be to peg the interest on excess reserves a certain number of basis points (or a fixed percentage) below the FF rate. (The differential could be thought of as something in the nature of a commission paid to the Federal Reserve Bank for in vesting the free reserves of the banks own ing them.) PROPOSALS FOR REFORM OF DISCOUNT WINDOW This arrangement could be expected to have the following major effects: (1) If the differential were made sufficiently large— say on the order of 50 basis points, or even somewhat larger— there would still be an incentive, at least for larger banks, to invest unneeded reserves directly in the FF market to avoid the differential. Thus, one would largely preserve a well-working FF market. (2) Smaller banks not having ready access to the FF market would be able to derive an income from their reserve surpluses com mensurate with that earned by the larger banks, except for a reasonable commission. (3) Because those now relying on the FF market as an outlet for their surplus funds would gain less from this activity than under the old system— to be precise, they would earn the differential instead of the full FF rate— one would expect that the equilibrium demand for excess reserves would increase. (4) But under the new system the amount gained by investing surplus funds in the FF market instead of keeping them as excess reserves would become a constant— the dif ferential— that would be independent of the level of the FF rate. Thus, while the equilib rium level of excess reserves would presum ably tend to be generally larger than under the present system, it would become in dependent of fluctuations in short-term market yields. Even greater stability in the level of ex cess reserves could be achieved by use of a sliding differential similar to that proposed above for the discount rate. That is, the differential would be kept constant as long as excess reserves were below some stated amount; but if excess reserves grew larger, the differential could be increased— thereby encouraging investment in the FF market, which would lead to a lower FF rate, and thus finally, through portfolio expansion, to a reduction in excess reserves. One important caveat must be entered 71 at this point. The stabilization of excess reserves results from making the opportu nity cost of holding excess reserves in dependent of the level of market rates. How ever, a difficulty would arise if the FF rate became so depressed as to be lower than the posted differential. Under such circum stances, because the interest paid on excess reserves cannot be less than zero, the dif ferential itself would have to be reduced— which would lead to an increase in the de sired level of excess reserves. What this means, of course, is that the payment of in terest on excess reserves is an effective sta bilizer of excess reserves only so long as the banking system does not encounter liquid ity traps; but it affords no protection against a liquidity trap.6 6 It would be possible, in principle, to design the proposal so that it could afford some protection even against situations of very low returns from invest ments and associated very low market yields. But this would require the radical step of applying the differential even when the FF rate were so low as to imply a negative interest—or in other words, the levying of a penalty—on excess reserves. The gen eral effect of such a penalty, of course, would tend to be that of making it possible for market yields to become extremely low—indeed in principle even negative. This, in turn, would clearly be a useful stabilizing mechanism. I must hasten to add, how ever, that it is hard to say just how effective this mechanism would in fact prove to be. For one thing, faced with a penalty on excess reserves, banks not finding any adequately yielding market instrument might endeavor to turn depositors away. It is unlikely that they would refuse deposits outright—because of long-run considerations. They might instead have recourse to service charges aiming at the same re sults. But this would still be a stabilizing influence for it would imply a negative return for holding money—a storage charge—which again would fa cilitate bringing market rates to very low or even negative levels and would encourage investments in real assets. A great danger is the possibility that banks might artificially increase their deposits, in order to absorb excess reserves, by making fictitious loans to cus tomers. It is impossible to predict just how wide spread such a practice might become, what its effects might be, and how it could be prevented or limited. However, it hardly seems worthwhile to dwell on the issue of a penalty on excess reserves because, for the moment at least, the likelihood of short-term market yields being so low as to create real problems seems rather remote. 72 PROPOSAL FOR SPECIAL BORROWING FACILITIES AIMED AT IMPROVING THE SPATIAL ALLOCATION OF BANK CREDIT This section describes in brief fashion a proposal for creation of special borrowing facilities, the purpose of which would be to improve the allocation of bank credit— that is, goal 4. Proposal outlined The reform sketched in the previous sec tion would go a long way toward achieving the first three goals set forth in the first section. But it would do little toward goal 4 — improvement of the spatial allocation of bank credit— except possibly insofar as it would tend to make more uniform across the banking system the cost of very short term borrowing and the rate of return from very short-term lending. In order to achieve goal 4 and as a further contribution to goal 3, I would like to advance a second proposal, the adop tion of which, incidentally, would be largely independent of the implementation of the reform set forth in the previous section. The essence of the proposal is to set up a second discount window (hereinafter referred to as the “term window”) — one that would grant credit for an essentially fixed term, say 3 months, not repayable until maturity (ex cept under special circumstances and/or with some appropriate penalty). The term window, too, would be open to any bank willing to pay the price, up to some limit determined by a creditworthiness standard. At the same time, lending conditions would again be structured so as to stabilize the amount of borrowing, making it independ ent of the level of short-term market yields or of other indicators of monetary string ency. But in contrast to the fiui window, which would be designed for minimum usage, the term window would be de signed to function as a substitute for an interbank loan market; to perform this func tion on an adequate scale, the volume of borrowing outstanding might have to be substantial. A market for interbank lending seems to have developed only to a very limited ex tent, except for overnight lending in the FF market and through the correspondent banking system— a rather surprising phe nomenon considering the large number of banks that make up the U.S. banking system. To be sure, a satisfactory spatial allocation of funds could be achieved even in the absence of interbank lending, if there were adequate devices by which banks could attract funds from “surplus” areas to “short” areas. But until rather recently such a possibility has been very much limited by the levels of ceiling rates on time deposits and the prohibition of interest payments on demand deposits. More recently, increases in ceilings on time deposit rates and the development of a market for CD’s have pre sumably led to some improvements. But there is reason to believe that even these developments fall short of adequacy since the CD market is, in practice, accessible only to large, prime banks. The proposed term window could also be regarded as a device to extend to smaller banks facilities that are analogous to those provided by the CD market. How the term window could contribute to allocative efficiency Before inquiring how the terms of borrow ing could be set so as to reconcile the goals of an open and extensively used window with that of a stable volume of loans out standing, it would be well to ascertain in PROPOSALS FOR REFORM OF DISCOUNT WINDOW what sense the existence of the term window could be expected to improve the spatial allocation of bank credit. Basically, the answer lies in the consid eration that a window open to all on the same terms would tend to equalize the op portunity cost of funds among banks, and hence presumably also the terms on which credit would be available to would-be bor rowers. It might be argued that this uni formity already tends to prevail, in the sense that under the present system all banks have the opportunity to invest in a common set of market instruments, and— what is more im portant— they all do invest, by and large, in certain instruments such as Treasury bills. It would therefore appear that the Treasury bill rate represents the common opportunity cost for all banks. But this, in fact, is not a valid inference because such bills are held not merely for their cash in come but also, in part at least, to satisfy liquidity requirements (as well as certain other requirements). One rather striking piece of evidence in support of this proposi tion is provided by the observation that many banks holding some bills in their port folios have been willing to issue CD’s at a significant premium over the rate on bills. We must conclude then that, even though the cash yield is the same for all holders, the “total” yield, including the “liquidity” component, need not be the same. It fol lows that the opportunity cost of funds in vested in other assets need not be the same for all banks, even if they all hold bills. In particular, one would expect that when banks are compared on the basis of the relation of their supply of funds for lending relative to their lending opportunities, the opportunity cost would be higher for banks with lower supply-to-demand ratios than for banks with higher ratios. Under these conditions we might expect 73 that if banks that are relatively short of funds were enabled to borrow from banks with excess funds at a rate somewhat above the current bill rate they would, within some limits, tend to take advantage of this oppor tunity. The borrowing banks would then use the funds for expanding their loan port folios (and possibly even their portfolios of short-term market instruments). For the lending bank the investment in the loan would presumably displace other assets, in cluding some loans. And the redistribution of loans would presumably increase alloca tive efficiency. The proposed term window would ac complish the same general result, although by a somewhat different route. Suppose the rate at the term window had been set some how and that at this rate banks would bor row a certain volume of funds with which to expand their loans. In order to accom modate this demand, while keeping total reserves unchanged, the Federal Reserve would have to liquidate some of its port folio of market instruments. This would raise market yields, thereby encouraging some banks— presumably those better sup plied with funds— to acquire market instru ments at the expense of their other invest ments, including loans. Thus, the final effect would be a redistribution of loans from more amply provided to less well-provided areas through a somewhat circuitous route. In other words, the surplus bank would choose as a substitute for direct loans to its regular customers not loans to the less wellsupplied customers of other banks but rather market instruments such as Treasury bills; such purchases by the bank with sur plus funds would enable the Federal Re serve to exchange securities for cash, which it would lend to the “short” bank, which in turn would use those funds to expand its loans. 74 It might be noted from the above that one implication of the proposed reform might well be an increase in the yield on market instruments, especially short-term ones such as bills. As is well known, this is an effect that typically tends to accompany any restructuring of the financial system, the result of which is more reliance on pure price rationing and less on other forms of rationing. It also follows from this analysis that the improvement in allocative efficiency one might expect from the proposed reform would depend on the views one had about the effectiveness of present arrangements for the spatial allocation of funds. Operational aspects Having thus laid out the basic argument in favor of a term window, we can take up the problem of how to achieve simultaneously an open window and a substantial and yet relatively stable volume of borrowing. Abstracting for a moment from “practi cality,” one could readily suggest a device that would accomplish the desired aim. Spe cifically, one could auction off on a regular schedule, say every week, a block of funds equal to the volume of loans that would come due in that week, somewhat along the lines of the present bills auction. This approach probably deserves con sideration in the light of the experience gained with the bills auction. Major draw backs might be (1) administrative com plexity and (2) the fact that the auction might again give an edge to the larger banks, which are better equipped to partici pate in it. It is hard to say without further careful study how serious these shortcom ings might be. As an alternative, it may be possible to “simulate” closely an auction by a device similar to that suggested for the 1-day win dow: reset the borrowing rate at frequent intervals— say, once a week— and tie that rate to a short-term market rate— say, the 3-month bill rate or the CD rate— with a flexible differential, one that increases as the volume of borrowing increases. With this arrangement one could not al together avoid some variations in the vol ume of borrowing, but the variations could be kept within moderate bounds. One im portant feature that would tend to insure this result is that the shrinkage in the vol ume outstanding in any given week could not exceed the amount reaching maturity. Assuming a 3-month maturity, this amount would be approximately 1/13 of the out standing volume. Furthermore, since this window would not be designed as a device for meeting short-run reserve requirements (which would be handled through the 1-day window), it would be quite appropriate to require that applications for loans to be taken down in a given week be filed some time in advance. Under these conditions the Federal Reserve would know in advance how much variation in the volume of bor rowing at the term window would occur in each week and could, if it wished, offset such variations by open market operations. One could readily conceive of slightly more complex designs. For instance, the window could announce, say, 2 weeks in advance two or more possible rates and ask for preliminary applications at each of the indicated rates. On the basis of this information it could set a final rate 1 week in advance and could accept as final all ap plications received at that rate. In short, it should not prove too difficult to design a system that would minimize fluctuations in the volume of borrowing outstanding and that furthermore could offset any remain ing fluctuations through open market policy. It should be noted in this connection that there is little reason to be concerned with PROPOSALS FOR REFORM OF DISCOUNT WINDOW the danger that, in slack periods, the volume of borrowings would shrink beyond con trol. In fact, with a variable differential one could always go so far as to push the rate to a level below the bill rate, at which point the volume of borrowings would obviously become highly elastic because any bank could make a “hedged” profit by borrowing and using the proceeds to buy bills. (It is an open question whether under such con ditions it would be preferable to let the borrowing shrink and to let the central bank purchase bills through open market opera tions.) How large a target volume of borrowing? It should be apparent from the preceding discussion that the smaller the target volume of borrowing, the easier the task of mini mizing fluctuations in reserves caused by fluctuations in borrowings at the proposed term window. But it should be equally ap parent that keeping the volume of borrow ing low would reduce the effectiveness of the proposal in achieving a better spatial al location of bank credit. To see how far these goals might be reconciled we might first ask this question: If one neglects the problem of stabilizing borrowing, how large a volume of borrow ing at the term window might, in the long run, be optimal? An answer to this question might be obtained by pursuing the idea that for smaller banks the term window should provide an alternative to the CD market. This criterion suggests the following an swer: The volume of borrowing should be such that the rate necessary to induce it would be somewhat above prevailing CD rates for a maturity comparable to that of fered at the window. To understand the rationale for the sug gested criterion, we may first note that a term-window rate close to the CD rate 75 would tend to equalize roughly the oppor tunity cost of funds for all banks that were issuing CD’s and/or using the window. To be sure, for banks not using either device the opportunity cost could be lower, pre sumably as low as the bill rate. However, since CD rates have tended, at least so far, to stay reasonably close to the bill rate, the difference in opportunity costs would re main within modest limits. At the same time it should be recognized that a lower rate at the term window would hardly be feasible, unless the window were somehow closed to banks issuing CD’s— which would seem quite undesirable and even inconsis tent with the spirit of the proposal. The reason is that, with a completely open win dow, the borrowing rate would tend to set a ceiling on the CD rate. Or to look at this from a different angle, the demand for bor rowing at the term window might be ex pected to become very elastic as the rate approached the CD rate. These considerations suggest an opera tional and pragmatic approach toward the development of the term window. Suppose the Federal Reserve started out with a fairly modest target, say around $2 billion to $3 billion, which would imply a weekly turn over of $150 million to $200 million. If it then turned out that the rate needed to clear that volume of borrowing were on the average substantially above the CD rate, one could make two inferences: (1) that the window seemed to be contributing sig nificantly to an improved allocation, filling a function not performed by present institu tions (this inference could of course be further tested by examining the distribution of borrowing among banks and the ap parent use made of the marginal funds ac quired); and (2) that there was a primafacie case for moving in the direction of increasing the target. It would then be pos 76 sible to plan to have such an expansion oc cur gradually over time as experience was gained with operation of the window and with problems that might conceivably arise. If on the other hand, even with a modest target, one should find that the rate tended to hover close to the CD rate and that the window was being used by banks that could have issued CD’s, then one could infer that the target should be reduced, or even that the reform was contributing so little to the improvement of the system as to justify abandoning it. Some minor complementary suggestions 1. It would seem appropriate to make the term window available only to banks that are members of the Federal Reserve System. This limited-availability feature, when coupled with the payment of interest on excess reserves (and a graduated system of reserve requirements), could go some distance toward providing an incentive for nonmember banks to become members, contributing to a better control of the money supply and short-term market rates. 2. One could relax the requirement that all borrowing at the window be for a single fixed term and allow some choice of terms, say between 2 and 6 months. The rate for such loans could be tied to rates for the corresponding maturities on the chosen market instruments, be it CD’s or bills, through a single differential applied to all maturities. However, this greater flexi bility would complicate the task of stabiliz ing the volume of loans maturing in any given week. It is not clear that this refine ment is worth the cost since banks could presumably manage, through other transac tions in short-term markets, to reconcile a fixed-term borrowing with their require ments. If the volume of borrowing at the window were sufficiently large, one might, as an alternative, conceive of developing some sort of secondary market, with or without the participation of the Federal Reserve. AN EVALUATION OF SOME DETERMINANTS OF MEMBER BANK BORROWING Leslie M. Alperstein Board of Governors of the Federal Reserve System Contents Summary of Findings and Conclusions__________________________________________ 80 Design of the Study_________________________________________________________ 82 Data Description of explanatory variables Results___________________________________________________________________ 83 Cross-sectional analysis Time-series analysis 77 AN EVALUATION OF SOME DETERMINANTS OF MEMBER BANK BORROWING borrow, sensitivity to interest rates has tra ditionally been a popular determinant. The literature from time to time has reflected discussions, both theoretical and empirical, on the influences of interest rate spreads on banks’ sensitivity to borrowing. Interest rates are examined closely for their influence upon borrowing per se, as well as for their degree of interaction with other determi nants of borrowing. A discussion dating back only a few years centered around the effects on member bank borrowing of possible differences in the ad ministration of the discount window among the various Federal Reserve districts. In order to facilitate a closer look at this prob lem, an attempt is made to distinguish be tween supply and demand factors in deter mining borrowing among several Federal Reserve districts. The variables tested for their significance in explaining borrowing from the Federal Reserve and borrowing from other sources were: (1) a liquidity ratio, designed to re flect the banks’ ability to meet loan de mands and unexpected deposit withdrawals out of internally generated short-term as sets; (2) an interest rate differential be tween the discount rate and the 3-month Treasury bill rate, to reflect the impact of banks’ response to least-cost considerations; (3) the size of the bank, to indicate dif ferences in the likelihood that some banks, The major focus of this paper is to identify the determinants of member bank borrow ing and to measure the relationships that exist among these determinants. Regression analysis is used to contrast various forms of borrowing from the Federal Reserve with borrowing from sources other than the Fed eral Reserve. The methodology was selected so that it would be possible to assign meas urable weights to the casual factors felt to be most responsible for borrowing and then to compare for differences in the ability of these factors to explain variations in bor rowing from alternative sources. It is becoming increasingly clear that the distribution of financial assets throughout the banking system can materially affect the speed and effectiveness of a given mone tary policy. Banks of various sizes may be expected to respond differently over the business cycle to changes in financial vari ables such as liquidity and interest rates. To the extent that banks react to fluctuating credit conditions by adjusting their liquidity positions, they can affect to a significant degree the amount of borrowing and credit expansion that takes place. Consequently, both bank size and liquidity are examined in this paper and tested for their signifi cance in explaining different types of bor rowing. Insofar as the structural characteristics of banks that borrow from the Federal Re serve influence the degree to which they 79 80 for example those associated with financial centers, might be less reluctant than others to borrow from the Federal Reserve; and (4) Federal Reserve district, to shed light on the problem of alleged differences in the administration of the discount window. To lessen the degree to which the statis tical results would be affected by problems associated with aggregated data, the initial part of the study examined the borrowing behavior of individual banks. Tests were made on weekly reporting member banks for six Federal Reserve districts during the period July 1959 to October 1961. The tests were divided into two parts: cross-sectional analyses and time-series analyses. The cross sections estimated the relationship between the likelihood of bor rowing by a particular bank and factors as sociated with its indebtedness. Additional cross sections estimated the relationship be tween the frequency of borrowing from the Federal Reserve and the postulated deter minants of borrowing. The time-series regressions estimated the relationship between borrowing and the in dependent variables— liquidity, size, and district— and then with the addition of the temporal variable, the interest rate differen tial. SUMMARY OF FINDINGS AND CONCLUSIONS The results of the cross sections and the time-series regressions suggest that the liquidity condition of a bank’s short-term asset portfolio, as well as the interest dif ferential between the discount rate and the bill rate, contributed significantly toward ex plaining variations in borrowing. Of the two factors, liquidity had a greater impact on borrowing in all of the periods studied; but in those periods when the discount rate was less than the bill rate, the importance of liquidity as an explanatory factor dimin ished somewhat in favor of the interestdifferential factor. The behavior of the liquidity and interestdifferential variables supports the least-cost hypothesis; that is, that banks, in general, are sensitive to interest rate differentials to the extent that they will borrow from the least expensive source even when that source is the Federal Reserve. In the period when the bill rate exceeded the discount rate, the ability of the bank liquidity variable to ex plain the likelihood of indebtedness to the Federal Reserve was only about half as great as when the discount rate exceeded the bill rate. This suggests that banks are less reluctant to borrow from the Federal Reserve when it is the least expensive alter native source of funds. During the period studied, banks bor rowed more often when it was profitable to do so. The frequency of borrowing from the Federal Reserve was negatively asso ciated with liquidity, corroborating the re sults of the conditional probability esti mates. Banks with relatively higher levels of liquidity were more likely to borrow, and borrow more often, from the Federal Reserve when the bill rate exceeded the discount rate. The cross sections demon strated that in the period when the bill rate exceeded the discount rate (rb > rd) banks were less willing to sell Treasury bills as a secondary source of reserves and shifted in stead to the Federal Reserve. The time-series regressions supported the inferences made from the cross-sectional analysis. When the interest rate differential was explicitly included in the regressions, banks were found to be sensitive to varia tions in the interest rate spread as well as to variations in liquidity levels. Inasmuch as aggregate data can sometimes mask the ef DETERMINANTS OF M EMBER BANK BORROWING fects of microrelations, it is not immediately clear whether the variations in relative amounts of borrowing were being caused by the same number of banks borrowing greater amounts, the same number of banks bor rowing the same amounts but more often, or fewer banks borrowing greater amounts. To distinguish between these effects, an estimate was made of the proportion of banks in each Federal Reserve district that were borrowing over time. The results in dicated that the number of banks per district as well as the relative amounts of borrow ing from the Federal Reserve increased when the discount rate was the least-cost alternative. On the basis of the regression results, it was concluded that for the period studied the incentive of banks to borrow stemmed from the liquidity condition of their short term assets as well as, and to a lesser ex tent, the profitability of borrowing. It can not be concluded from these results that banks borrow from the Federal Reserve to reinvest in short-term Government securi ties. It was demonstrated that Treasury bills were liquidated in periods when the bill rate exceeded the discount rate. The re sults would have been the reverse if banks had been engaged in “profiteering.” Among the other determinants, bank size was shown to have a significant, al though uncertain, effect on borrowing from the Federal Reserve. Although considerable differences were found to exist among banks of varying sizes with respect to their likeli hood, as well as to their frequency, of bor rowing, no discernible pattern emerged among size groups. Borrowing from other sources, however, was found to increase with size, a result which was not unexpected. There were variations among Federal Re serve districts in relative amounts of bor rowing from the Federal Reserve, in the 81 proportion of banks borrowing, and in the frequency of borrowing per bank. These re sults suggest the existence of other causal factors not explicitly considered, such as those with characteristics of demand or supply. If nonuniformities in the administration of the discount function were responsible for the disparate pattern of borrowing among districts, this pattern would be expected to differ substantially from a market in which nonprice rationing was nonexistent. Banks are not precluded from borrowing else where, for example the Federal funds mar ket, on grounds other than price constraints or smallness of transaction. Therefore, the patterns of ex post borrowing among dis tricts between a price-determining market and the discount window would differ to the extent that funds were more accessible at some discount windows by virtue of easier lending policies. Although the six districts differed substantially with respect to borrowing from the Federal Reserve after liquidity and cost were taken into account, there was a similarity in the bor rowing for each district by type of borrow ing. The cross-sectional findings indicated that roughly the same patterns of borrow ing existed among Federal Reserve districts for borrowing from the Federal Reserve and for borrowing from other sources. The timeseries analyses indicated that the patterns of borrowing were precisely the same after taking into account the liquidity condition of the district as a whole and the interest rate differential. This does not prove that supply factors are of no importance in the determination of differences in borrowing among districts, but rather that for the period of time covered and the districts involved, differences in demand explained to some degree the im portance of the Federal Reserve district as a determinant of borrowing. 82 DESIGN OF THE STUDY Data The sample consisted of nearly all weekly reporting member banks in the six Federal Reserve districts for which adequate data were available: Boston, Richmond, St. Louis, Minneapolis, Dallas, and San Fran cisco. Banks for which complete records were not available for one reason or another were dropped from the sample. The Bank of America was dropped because it was believed that this bank was not of the same nature as the other banks in the sample and, because of its size, might bias the re sults. Each bank was checked for changes in structure. Banks that merged during the period July 1959 to October 1961 were deleted. When the sample of banks was finally completed, the raw data from the weekly balance sheets were averaged for 2-week periods, giving 35 biweekly observa tions for each of the 143 banks. Information on interest rates was ob tained from the Federal Reserve Bulletin and from Section 12 of the Supplement to Banking and Monetary Statistics, 1966. Special calculations required for weekly averages of daily figures on 3-month Treas ury bill rates were made available by the Government Finance Section of the Board’s Division of Research and Statistics. coin + Balances in banks in the United States — Borrowing from the Federal Re serve — Lagged borrowing from, others) to (Demand deposits adjusted + Time de posits — Required reserves). The approach taken in this study is to regard the maintenance of good bankercustomer relationships as the deciding fac tor in assessing the short-run “needs” or liquidity of the bank. Accordingly, bankers will alter their optimum asset portfolio in order to accommodate customers who they feel have long-run profit potentials out weighing current considerations. P, measure of the cost of borrowing from the Federal Reserve in contrast to borrow ing from other sources where P — (rd — rb), rd = the discount rate, and rb = Treasury bill rate. When (rd — rb) is negative, there is a negative cost associated with borrowing from the Federal Reserve. Li, measure of the bank’s demand deposit size where Li = under $25 million; L2 = $25 million to $50 million; L3 = $50 million to $100 million; L4 = $100 million to $300 million; L 5 = over $300 million. In the regressions, the Li’s are represented by dummy variables. Description of explanatory variables C, dummy variable indicating the bank’s reserve classification The variables used to explain the various forms in which borrowing is presented in clude the following: where reserve city banks were assigned a value of 1 for variable C, country banks were given a value of 0. LQ, measure of a bank’s liquidity Di, dummy variables representing the six reserve districts where LQ = the ratio of: (Loans to domes tic commercial banks + Loans to brokers and dealers for purchasing or carrying other securities + Treasury bills + Currency and where D i = Boston; D 5 = Richmond; D 8 = St. Louis; D 9 = Minneapolis; D u = Dal las; D i 2= San Francisco. 83 DETERMINANTS OF MEMBER BANK BORROWING RESULTS Cross-sectional analysis August 19, 1959 Ordinary least-square regressions were used to ascertain if, and to what extent, a rela tionship existed between a bank’s borrow ing from the Federal Reserve and such characteristics as size of bank, Federal Re serve district, and the liquidity position of the asset portfolio. The conditional prob ability estimates predicted the likelihood of nonzero borrowing. In addition, they served to indicate the nature of the relationship between the independent variables and the likelihood of borrowing; that is, would banks in one particular district be more likely to borrow than those in another. Likelihood of indebtedness. Four cross sec tions were taken: two for dates when the Treasury bill rate was greater than the dis count rate— August 19, 1959, and Decem ber 23, 1959— and two for dates when the bill rate was less than the discount rate— July 22, 1959, and March 16, 1960. Esti mates were made for the likelihood of bor rowing from the Federal Reserve (Bf) and from other sources (B0). The final equa tions for the conditional probability esti mates were:1 (2) Bf = July 22, 1959 (1) B , = .195 - .110(Di) (.370) (.119) - .088(Ds) + .208(Do) + .017(Dn) (.118) (.136) (.116) - .084(Di2) + .224(L 2 )* + .126(L3) (.121) (.090) (.100) + . 1 5 3 ( 1 , 4 ) - . 0 7 4 ( 1 * ) - .001 (L Q )* (.093) (.145) (.000) i?2 = .152, F = 2.360* December 23,1959 (3) Bf — .313 - .105(Di) (.402) (.128) - .031 (D8) - .001 (Do) - .066( Du) (.129) (.144) (.125) - .129(Di2) + .07 3(1,2 )+. 030(L3) (.130) (.097) (.111) + .147 (Li) — .154(1,5) — .001 (LQ)* (.400) (.149) (.000) R 2 = . 079, F = 1.126 March 16, 1960 (4) Bf = .624 - .236(Di) — .154(D S) (.465) (.149) (.150) - .058(Ds) — .038(D n) — ,184(Di2) (.170) (.147) (.154) + ,052(L 2) + .0 8 7 ( L 3) - .065(1,4) (.110) (.130) (.117) + . 1 2 2 ( 1 * ) - .002( LQ) * (.206) (. 0 0 0 ) R 2 = .113, F = 1.674 .549 - .353(Di)* (.439) (.141) - .203(Ds) + ,039(Dg) — .115(D u) (.140) (.160) (.137) - .224(Di2) + .256(1,2)* + .108(L3) (.141) (.107) (.121) + .194(L 4) + .015 (Ls) — .0 0 1 (L g )* (.116) (.161) (.000) R 2 = .203, F = 3.352t + 1 In all of the equations * indicates significant at .05 level of confidence and f indicates significant at .01 level of confidence. Standard errors are in paren theses. R 2 = .317, F = 6.121+ July 22, 1959 (5) B0 = .399 - . 0 5 4 ( D i ) - .099(D 8) (.422) (.135) (.135) + .180(D9) - .024(D u) — .235(D 12) (.155) (.132) (.137) .22 6(1,2 )* + .251 (L3)* + .493(L O t (.1 0 2 ) (.116) (.1 1 2 ) + .533(L5) * - .002( LQ) * (.155) (.000) 84 August 19, 1959 (6) B o = .388 + . 0 3 3 ( D i ) - . 0 2 2 ( D s ) (.392) (.126) (.125) + .201 (De) + ,024(D u) — .175(Di2) (.143) (.123) (.128) + .273(La) * + .279(L3)* + .595(L4) f (.095) (.106) (.099) + .609 (L5)* — .002 (L<2) * (.154) (.000) R 2 = .404, F = 8.934f December 23, 1959 (7) B0 = .144 - . 0 2 2 ( D i ) + .122(D8) (.384) (.123) (.123) + .236(D») + .020(D U) + .077 (Du) (.137) (.120) (.126) + ,374(L 2) f + 189(L3) + .680(L4) t (.093) (.103) (.094) + .491 (L b) * - .001 (LQ )* (.153) (.000) R2 = .407, F = 9.071t March 16, 1960 (8) B0 = .365 - ,1 06 (D i)— .017(D8) (.379) (.121) (.121) + .004(A)) — .008( D u ) - .098(D 12) (.138) (.121) (.123) + .260(L2) f + ,409(L3) t + .572(L4) t (.091) (.104) (.095) + .706(1*) t - .002(L<2)* (.157) (.000) R 2 = .442, F = 10.484+ Of the eight regressions, six revealed the presence of debt as being significantly related to the three characteristics: size, district, and liquidity. On each date borrowing from other sources was more fully explained by the independent variables than was borrow ing from the Federal Reserve. This was to be expected, however, as certain unobserv able variables would affect borrowing from the Federal Reserve in a different way than borrowing from other sources. For instance, an ostensibly important factor, the profit spread, has not been considered explicitly. Differences in the availability of funds from the discount window in contrast to other sources would also affect the ability of the independent variables to explain borrowing. It is inferred, therefore, that the relatively lower R 2’s for borrowing from the Federal Reserve are explained at least in part by factors that constrain borrowing from the Federal Reserve but not borrowing from other sources; to wit, reluctance to borrow and availability of supply. In the two equations that showed a sig nificant relationship between borrowing from the Federal Reserve and the factors determining the likelihood of borrowing, each of the variables was tested to determine its net contribution in explaining the total variation in indebtedness. The partial rela tionships for borrowing from the Federal Reserve and borrowing from other sources are given in Table 1. These partial relation ships indicate the amount of explanation contributed by the addition of the factor considered. All three variables— liquidity, district, and size— were found to contribute to the explanation of borrowing. Worth noting are the differences in impact of the independent factors on borrowing from the Federal Re serve in contrast to borrowing from other sources. In terms of the size of the partial coeffi cient of determination, the Federal Reserve district made the largest contribution to the explained variation in borrowing from the Federal Reserve. By comparison, the district played a much smaller role in borrowing from other sources, while size explained most of the variation. On August 19, 1959, when the bill rate was greater than the dis count rate indicating that banks could bor row more cheaply at the discount window, the partial correlation coefficient for the DETERMINANTS OF M EM B E R BANK BORROWING TABLE 1 PARTIAL RELATIONSHIPS OF CONDITIONAL PROBABILITY ESTIMATES OF BORROWING FROM THE FEDERAL RESERVE AND FROM OTHER SO U RCES1 Variables in regression Added variable July 22, 1959 Di Li Li Di Di LQ L{ LQ d vlq Li l vlq Di D i Li LQ Partial coefficient of determination Federal Reserve Other sources .065 (2.32)* .010 (2.66) f .083 (12.29) f .082 (12.18) f .052 (1.82) .080 (2.28) * .070 (9.93) f .214 (9.04) f .092 (2.69) t .163 (26.38) f .061 (2.16) * .079 (2.28) * .029 (4.06)f .041 (1.57) .062 (2.19) * .068 (1.93) * .030 (6.08) f .272 (12.40) f .095 (2.79) f .186 (31.02) f .168 (27.70) t .232 (9.96) f .066 (1.86) .014 (21.70) | .011 (16.55) f .146 (5.60) f .071 (2.00) * .089 (12.94) f Aug. 19, 1959 Di Li Li Di Di LQ Li LQ d vlq Li L VLQ Di D i>L i LQ 1 F-ratios are in parentheses. * indicates significant at 0.05 level of confidence, f indicates significant at 0.01 level of confidence. liquidity variable dropped more than 50 per cent of the value that had prevailed when the bill rate was less than the discount rate. The finding that less importance is attrib uted to liquidity considerations when bor rowing from the Federal Reserve is profitable lends support to the least-cost hypothesis. Borrowing from other sources was in fluenced very little by liquidity when the bill rate was greater than the discount rate, indicating that banks are sensitive to changes in interest rate differentials. During this period, the bill rate was also greater than the rate on Federal funds, and so it is likely that banks tended to absorb Treasury bills by meeting liquidity considerations through the Federal funds market. This hypothesis is supported by the behavior of the size variable. During the period when 85 the bill rate was low in relation to other rates, the partial coefficient for size was 0.146 and significant at the 1 per cent level. When the bill rate was relatively high, the larger banks were less reluctant to borrow, as indicated by the increase in the partial correlation coefficient to 0.232, significant at the 1 per cent level.2 Comparisons by reserve districts between borrowing from the Federal Reserve and from others. The most interesting finding with respect to the determinants of borrowing was the behavior of the variables for Fed eral Reserve districts. It has been argued that there are differences in borrowing among districts and that these differences reflect nonuniformities in the administration of the discount window. If the different bor rowing patterns that emerge among districts could be attributed to nonuniform adminis tration of the discount window, the patterns of borrowing from the Federal Reserve would differ among districts from borrow ing from other sources. On the other hand, if differences in demand as distinct from differences in supply, were responsible for the borrowing patterns that emerged, the patterns of borrowing among districts would be similar for both types of borrowing. In Table 2, the districts have been ranked in order of the likelihood of borrowing from the Federal Reserve and from other sources. The order was obtained from the regres sions and derived from the coefficient at tached to the district variables. A negative district coefficient in the regressions places the district below D5 (Richmond was the base district in the regression) in Table 2. Banks located in the district with the largest negative value on a given date show the least probability of borrowing. Banks in the 2 The assertion that borrowing from other sources is functionally related to size is explained more fully later. 86 TABLE 2 DISTRICTS RANKED BY LIKELIHOOD OF BANKS BORROWING FROM THE FEDERAL RESERVE AND FROM OTHER SOURCES July 22, 1959 Order Bf 1 2 D 12 3 4 5 Ds P it D* D9 6 N ote. Bo d 8 jD1 D m d5 D9 Aug. 19, 1959 Bt Dec. 23, 1959 Mar. 16, 1960 Bo Bf Bo Bf d *>12 Dx Ds Dt 12 D S Ds IX K D, Dx D0 D, D UL £>8 d . Bo Ds *>11 D 12 Ds D .i d 5 — Order is from least likely to most likely to borrow. of which is the possibility that the liquidity variable does not accurately reflect the de mands for credit in the various Federal Re serve districts. district with the largest positive value show the greatest probability of being in debt. The comparative likelihoods of borrow ing from the Federal Reserve (Bf) and of borrowing from other sources ( B0) are roughly similar. For July 22, 1959, only Di differed in its ranking among the six districts borrowing from the Federal Reserve and borrowing from other sources. On August 19, 1959, Di and D i2 changed position while on De cember 23, 1959, D 5 and D 12 changed their order of rank. On March 15, 1960, only D 9 changed its rank between Bf and B 0. Had the patterns of borrowing among districts differed by type of borrowing, one might conclude that some discount windows are more accessible than others. However, since the patterns are roughly the same be tween borrowing from the Federal Reserve and from other sources, it remains only to explain differences in the demand borrow ing among districts. There are, no doubt, a number of conceivable explanations, one Relationship between size of bank and borrow ing. A comparison by size of bank was made for borrowing from the Federal Re serve and borrowing from other sources. The order of comparison is shown in Table 3 where the banks by size are listed from least likely to most likely to be in debt. There appears to be a strong functional relationship between size of bank and bor rowing from other sources. For each date except December 23, 1959, the likelihood of a bank’s indebtedness was an increasing function of size. These results were not un expected for as credit tightens, smaller banks draw down balances with their larger correspondent banks— shifting the burden of liquidity to them. The larger, more ag gressive banks usually carry smaller relative quantities of excess reserves and therefore would be expected to borrow more. In contrast to borrowing from other TABLE 3 BANK SIZE RANKED BY LIKELIHOOD OF BORROWING FROM THE FEDERAL RESERVE AND FROM OTHER SOURCES July 22, 1959 Order 1 2 3 4 5 N ote. Aug. 19, 1959 Dec. 23, 1959 Bo Bo Bf Bo Bf Lx Li l 5 Ls l 5 Ls l 2 Lx Lx L2 Lx L3 Ls L, h Ls K K L2 Bt L* L . —•Order is l 2 l 2 from least likely to most likely to borrow. Mar. 16, 1960 Bf Bo L, L, Lx L2 Lz L2 DETERMINANTS OF MEMBER BANK BORROWING 87 sources, borrowing from the Federal Re serve showed considerably less association with size of bank. On August 19, 1959, and December 23, 1959, the largest banks were less likely to borrow than the smallest banks, as reflected in Table 3. Borrowing from the Federal Reserve, unlike borrowing from other sources, is neither anticipated nor orderly. It is more spontaneous, and conse quently would be expected to show a less discernible pattern. (9) Ff = 14.575 - 1.398(Di) (8.882) (2.950) - 1.767(D8) + 5.194 (Z>9) (2.872) (3.288) - 2.337( D u ) - 6.611 (D u )* (2.817) (2.950) - 4.151(C)* + 5.399(L2)* (1.805) (2.144) Frequency of borrowing at the discount win dow. Cross-sectional regressions were used to estimate the relationship between the frequency of borrowing from the Federal Reserve and the variables that had been used to estimate the likelihood of indebted ness, with the addition of a reserve classi fication variable (C). The period from July 8, 1959, to No vember 2, 1960, was divided into three subperiods. The subperiods were chosen to emphasize patterns of borrowing when the relationship between the bill rate and the discount rate differed. In particular, all indi vidual bank data were aggregated to pro vide totals for the subperiod when the dis count rate was above the bill rate. A second aggregation covered the subperiod when the bill rate was greater than the discount rate; the third aggregation covered the en tire period. District, size, class, and liquidity variables were used to explain variations in the fre quency of borrowing by banks in each of the designated subperiods. The frequency of borrowing from the Federal Reserve (Ff) represented the number of weeks that a given bank was indebted during the sub period considered. A weekly average was made of the data within each subperiod. For the entire period, July 8, 1959, to November 2, 1960, frequency of borrowing from the Federal Reserve was estimated by: + 4.050(1*) * + 2.073 (L4) (2.498) (2.398) + 2.362(1,5) — M 0 ( L Q ) * (3.741) (.012) R 2 = .209, F = 3.144t In order to examine the effects of holding short-term Treasury bills, the variable T was substituted for LQ in equation 10. (10) Ff = 9.768 .221 (Di) (8.874) (2.984) - 1.640(D8) + 7.104(D<))* (3.209) (2.784) - 2.890(D n) — 5.138(Dia)* (2.893) (2.893) - 4 . 1 1 0 ( C ) * + 5.488(L2)* (1.801) (2.145) + 5.157(L3) * + 5.440(L4)* (2.506) (2.384) - 1 0 . 3 7 1 ( 1 , 5 ) * - .055(T )* (4.351) (. 0 2 2 ) R 2 = .210, F = 3.169* From August 19, 1959, to March 2, 1960, when the bill rate was above the dis count rate, the estimating equation was able to explain to a lesser extent the variation in Ff. In particular: (11) Ff = 5.649 + .444(D i) (5.009) (1.671) + .904(D 8) + 3.059 (D»)* (1.616) (1.834) + .352(D n ) — 2.157(D i2) (1.583) (1.636) - 1.440(C ) + 2.852 (L2)* (1.015) (1.223) 88 + 2.326( L3) * + 1.335 (Li) (1.364) (1.334) + 1.498(Lb) - .0 16(L g)* (2.149) (.006) R 2 = .153, F = 2.149* (12) Ff = 2.897 + .753(Di) (5.071) (1.723) + . 7 6 9 (D s ) + 3.954(Do)* (1.635) (1.830) .147 (D u) - 1.492(Di2) — 1.201(C) (1.584) (1.639) (1.019) + 2.995 (L2) * + 2.735 (L3)* (1.240) (1.385) + 2.889(L4)* + 4.984(L5)* (1.337) (2.433) - .019(7")* (.009) R 2 = .132, F = 1.805 In the subperiod when the discount rate (rd) was greater than the bill rate (rb), the estimating equations using LQ and T were: (13) Ff = 10.257 - 1.798(Di) (5.046) (1.679) - 2.382(D8) + 1.809(D9) - 2.270(Du) (1.630) (1.823) (1.599) - 4.909(D 1 2 ) * — 2.843(C)* (1.680) (1.022) + 2.363(L2)* + 1.971(1*) + .220 (L*) (1.219) (1.420) (1.362) + .734 (Lb) .022 (L<2)* (2.129) (.007) R 2 = .241, F = 3.792f (14) Ff = 6.764 - 1.135(D i) (5.126) (1.709) - 2.365(D « )+ 3.219(Do) — 2.741 ( Du) (1.657) (1.853) (1.609) - 3.744(D 12) * - 2.795(C )* (1.671) (1.041) + 2.309(L 2)* + 2.618(L 3) + 2.338(L 4) (1.240) (1.444) (1.374) + 5.394(L S) * - .031 (T )* (2.483) (.014) R 2 = .217, F = 3.306* The cross-sectional regressions suggest that the frequency with which a bank bor rows from the Federal Reserve is related to its size, reserve district, portfolio of liquid assets, and perhaps its reserve classification. A compact arrangement of the variable co efficients is presented in Table 4. In all three subperiods there was a signifi cant association between liquidity and fre quency of borrowing from the Federal Re serve. When the bill rate was greater than the discount rate (rb > rd), liquidity be came less important as a determinant of frequency. In contrast, when rd was greater than rb, a smaller drop in the level of liquidity prompted an increase in the fre quency of borrowing. The movements in Treasury bills showed much the same pattern as liquidity. Al though movements in bills were inversely related to frequency in every subperiod, TABLE 4 FREQUENCY OF BORROWING FROM FEDERAL RESERVE By district, size, reserve classification, liquidity, and Treasury bills held Equation Period Entire When rb > rd ( ) 1V 9 10 ( 11 \ When rd > rb ( 1I 12 13 14 Constant 14.575 (8.882) 9.768 (8.874) 5.649 (5.009) 2.897 (5.071) 10.257* (5.046) 6.764 (5.126) D8 Do Du £>12 —1.767 (2.873) —1.641 (2.873) .940 (1.616) .769 (1.635) -2.382 (1.630) -2.365 (1.657) 5.194 (3.288) 7.104* (3.209) 3.059 (1.834) 3.954* (1.830) 1.809 (1.873) 3.218 (1.853) -2.337 (2.817) —2.890 (2.784) .352 (1.583) - .147 (1.584) —2.270 (1.599) -2.741 (1.609) -6.611* (2.950) —5.138* (2.894) -2.157 (1.636) —1.492 (1.639) —4.909f (1.680) —3.744* (1.672) Di. -1.398 (2.950) — .221 (2.984) .444 (1.671) .753 (1.723) —1.798 (1.679) —1.135 (1.709) * indicates significant at 0.05 level of confidence, f indicates significant at 0.01 level of confidence. DETERMINANTS OF MEMBER BANK BORROWING 89 banks were less willing to reduce holdings of bills when their yield exceeded the cost of funds at the discount window. This find ing supports the hypothesis that banks ad just reserves in accordance with the leastcost alternative; that is, banks with low levels of liquidity chose to borrow more often at the Federal Reserve when it was the least expensive source of funds. When the discount rate exceeded the bill rate (rd > rb), the frequency of borrowing was related to larger swings in holdings of bills, and banks displayed a greater willingness to liquidate bills rather than borrow from the more costly discount window. The regressions explained less of the variation in the frequency of borrowing when the yield on Treasury bills exceeded the discount rate. Two reasons for this may be cited. First, the exclusion of the price variable reflects more importance when the bill rate is greater than the discount rate. Least-cost considerations demonstrate their impact during this time at the expense of liquidity considerations. It appears, then, that banks are more inclined to borrow from the Federal Reserve when it is least costly. Consequently, the reduction in li quidity, which has been shown to be a sig nificant factor in determining borrowing, reduces the explanatory power of the equa tion when the bill rate is greater than the discount rate. The second reason for the lower explana tory values concerns the element of control over borrowing at the discount window. As the bill rate rises relative to the discount rate, discount officers must remain alert to the potential for banks to take advantage of the interest rate differential. As administrative factors and, therefore, unspecified supply factors increase in importance, the demand factors are less able to explain the variations that occur in the frequency of borrowing from the Federal Reserve. L2 £3 l4 l 5.399* (2.144) 5.488* (2.145) 2.852* (1.223) 2.995* (1.241) 2.363 (1.220) 2.309 (1.240) 4.050 (2.498) 5.157* (2.506) 2.326 (1.364) 2.735 (1.385) 1.971 (1.420) 2.618 (1.444) 2.073 (2.398) 5.441* (2.384) 1.335 (1.334) 2.889* (1.337) .220 (1.362) 2.338 (1.374) 2.362 (3.741) —10.371* (4.351) 1.498 (2.140) 4.984* (2.434) .734 (2.128) 5.394* (2.483) 5 C LQ -4.151* (1.805) -4.119* (1.801) -1.440 (1.015) —1.201 (.019) —2.843f (1.023) —2.795* (1.041) -.030* (.012) Time-series analysis The inclusion of the price variable P = (rd — rb) in a time-series analysis substan tiated a large part of what was suggested by the cross-sectional analysis. The results of the test indicate that both liquidity and rela tive prices play significant roles in the amounts of borrowing that banks, on bal ance, will wish to undertake and that these relative amounts differ among Federal Re serve districts. The objective of the time-series was to observe over time the effect of liquidity, dis trict, and cost on the patterns of borrowing. In order to do this, the cross-sectional vari able for the reserve district had to be pooled with the temporal variable P. The procedure was to aggregate all banks within each dis trict for each date; this provided six district observations for each of the 35 dates, or a T —.055* (.022) -.016* (.006) —.019 (.010) —.022f (.007) —.031* (.014) R2 .209f (3.144) .210f (3.169) .153* (2.149) .132 (1.805) .241f (3.772) .217f (3.306) N o t e . — Standard errors are in parentheses under variables. F-ratios are in paren theses under R 2. 90 total of 210 observations. The districts were represented by dummy variables. Two forms of borrowing were estimated. The first was designed to show the relation ship between relative amounts of borrowing and the explanatory variables. This relation ship is represented by the ratio of borrowing from the Federal Reserve to demand de posits adjusted (Bf / DDa) and of borrow ings from other sources to demand deposits adjusted (B„/DDa). The second form of borrowing was designed to show how the proportion of banks that borrowed in each district was affected by changes in liquidity conditions or interest rate differentials. The proportion of banks borrowing from the Federal Reserve was represented by (Bn) and the proportion borrowing from other sources by (B„f). The amount of borrowing from the Fed eral Reserve and from other sources is esti mated in equations 15 and 16, respectively. (15) Bf = .0 3 9 (P )f — .150(L<2)f DDa (.008) (.019) + 18.86(D 1) f + 2 5 .1 5 ( D ,,) f (2.120) (2.478) + 30.43 (D s) f + 31.33 (Da) f (2.983) (1.600) + 3 3 .9 6 (D n ) t + 11.02(Dio)f (3.494) (1.539) R* = .74 , F = 79.28 (16) Bo/DDa ~ .019(Pi)* (.009) - .348(L <2)f-f 5 0 .3 1 2 (D i)f (.023) (2.673) + 55.827 (D s)t + 78.997 (D s)f (3.125) (3.762) + 76.675(D 9) f + 99.761 (D n )f (2.018) (4.407) + 37.895(D 12) f (1.940) ^2 = .95, F — 514.876 In both equations 15 and 16, borrowing was explained by the interest rate differen tial, liquidity, and district variables.3 It should be recalled that borrowing in the cross sections was related to Federal Re serve district. The variation in borrowing among districts gave rise to the question of causal factors not explicitly considered in the equations. This problem was handled by demonstrating that borrowing from other sources also varied significantly among dis tricts and the patterns that resulted from ranking the districts in order of their likeli hood of borrowing were roughly the same for Bf and B0. It was concluded, therefore, on the basis of the cross sections, that the variations among districts reflected mainly demand factors. The relative amounts of borrowing from the Federal Reserve differed among districts in the time series as well as the cross sec tions. However, when the districts were ranked in order of the amounts borrowed from the Federal Reserve and from other sources, the rankings were found to be precisely the same (Table 5). Once again the differences in borrowing among districts TABLE 5 DISTRICTS RANKED BY RELATIVE AMOUNTS OF BORROWING FROM THE FEDERAL RESERVE AND FROM OTHER SOURCES Order B f/ D D a 1 2 3 4 5 6 D5 D8 N o te . — B 0/ D D a d 12 5 Db d *>n Order is from least likely to most likely to borrow. 3 The coefficients of the district variable reflect deviations from the district average. For instance, if this average were 25.0, membership in would mean a less-than-average level of borrowing from the Federal Reserve in equation 15, given the liquidity condition and rate differential. Similarly, membership in D n would imply a greater-than-average level of borrowing. DETERMINANTS OF M EMBER BANK BORROWING 91 would seem to be associated with demand factors. Equations 17 and 18 are used to estimate the proportion of banks in each district that borrowed from the Federal Reserve (Bff) and other sources ( B0f), respectively. proportion of banks borrowing from the Federal Reserve is roughly the same as the relative amounts borrowed. The rankings by district for relative amounts and propor tions are combined in Table 6. (17) Bf f = - .779(Fi)f — 1.915( L 0 ) f (.115) (.264) + 38.562(Di) + 541.160(Ds)t (30.142) (35.231) + 488.728 (Ds) f + 465.152(D9) f (42.417) (22.752) + 513.350 (Du) f + 280.008 (Di2) t (49.684) (21.884) R 2 = .871, F = 181.300 .048(F) — 1.746(L<2)f (.068) (.197) + 568.572(Di)f + 714.378(D5) t (22.454) (26.246) + 518.473 (D8) f + 480.353 (Do) t (31.598) (16.949) TABLE 6 DISTRICTS RANKED BY RELATIVE AMOUNTS ( B f/DDa AND B 0/DDa ) AND PROPORTIONS OF BANK BORROWING (B fr AND B of) Order 1 2 3 4 5 6 B f /D D a D 12 D'i D, Da Do On B 0/ D D a D 12 B or D, Ds Do Ds Dt Dn Dn d 5 *>13 £>5 *>1 5 Ds d *>n v* N o t e .— Order is from least to most amount and proportion of borrowing. (18) B0f = - + 5 7 1.2 5 9 (D n )f+ 5 9 0 .2 3 3 (D i2) f (37.012) (16.303) R 2 = .971, F = 961.70 Equation 17 demonstrates that a strong relationship existed between the proportion of banks in each district that borrowed from the Federal Reserve and the explanatory variables. The proportion of banks that borrowed from the Federal Reserve varied inversely with the profit spread and liquidity position. This indicates that at least part of the increased amounts of borrowing that took place when the bill rate exceeded the discount rate resulted from an increase in the number of borrowing banks. This find ing supports the view that the effects of tight money are passed from one bank to another, affecting greater proportions as alternative sources of liquidity dry up. As in previous equations, there were dif ferences in borrowing among districts. The Equation 18, which focuses on the pro portion of banks borrowing from other sources, is rather difficult to explain. In the first place, the sign of the cost variable P is negative. This suggests that as the cost of discounting becomes greater than the rate on borrowing from other sources, the pro portion of banks borrowing from other sources falls. This seems unlikely. The vari able P itself, however, was not statistically significant. As with borrowing from the Federal Re serve, the proportion of banks borrowing from other sources varied noticeably among districts. However, in the former this varia tion was attributed to demand factors on the basis that the same patterns of borrowing were prevalent for borrowing from the Fed eral Reserve as for borrowing from other sources. The proportion of banks borrow ing from other sources varied among dis tricts, but the rankings were not similar to the rankings of borrowing from the Federal Reserve. The reason for the unusual pattern of borrowing from other sources is not im mediately clear. An appropriate explana tion would require further investigation. TOWARD A SEASONAL BORROWING PRIVILEGE: A STUDY OF INTRAYEAR FUND FLOWS AT COMMERCIAL BANKS Emanuel Melichar Board of Governors of the Federal Reserve System Contents Objectives and Procedures----------------------------------------------------------------------------------- 95 Fund Flows at Individual Banks: Definition_______________________________________ 96 Calculation of trend-adjusted intrayear changes Special terminology for assets, liabilities, and flows Fund Flows at Individual Banks: Two Examples___________________________________ 98 How the examples were chosen Computation of trend-adjusted intrayear flows Relative changes and flows Portfolio adjustments in response to fund flows Preliminary lessons from the examples Origin of Intrayear Fund Flows_________________________________________________101 Changes in I PC deposits Changes in nonfinancial loans Independence of loan and deposit changes Coincidence of loan and deposit drains Fund Outflows: Summary of Aggregate Data______________________________________102 Relative Outflows at Individual Banks___________________________________________103 A Seasonal Borrowing Privilege________________________________________________104 Design of a seasonal discount program Potential borrowings by period and by size of bank Potential borrowings under alternative deductible levels 93 Preface In contrast, the seasonal borrowing privi lege later proposed by the Steering Commit tee specifies somewhat different definitions of loans and deposits to be used in the calculation of seasonal needs of individual banks, suggests a monthly or 4-week moving-average basis, and presumably utilizes more rigorous methods to separate seasonal from trend, cyclical, and irregular move ments in loans and deposits. In formulating its proposal, the Committee was also guided by other studies that tested its effect while using data and methods more closely at tuned to its specifications. Therefore, the reader is warned that the estimates presented herein cannot be inter preted as representing, even roughly, flows and borrowings to be expected under the seasonal borrowing privilege as now pro posed. But hopefully, the relationships found in this study continue to be broadly indicative of the need for and design of a seasonal borrowing privilege. An earlier version of this paper that presents more detailed and disaggregated data is available from the author upon re quest. Emanuel Melichar December 1970 When the Federal Reserve System initiated a reappraisal of the discount mechanism, one avenue of investigation was concerned with facilitating the provision of discount credit for seasonal purposes. Such study re quired evidence on seasonal fund flows at individual commercial banks throughout the Nation. To help meet this need for data, the project reported herein was developed to estimate intrayear flows of funds between call dates on a uniform basis for all banks. These data helped in assessing the need for seasonal discount credit and in designing a mechanism through which such credit could be provided on a routinized basis. This study, undertaken during 1966-67, measures intrayear fund flows at individual banks and provides estimates of potential borrowings under assumed alternative sea sonal discount arrangements. These esti mates are based on certain definitions of loan and deposit flows, and flows so defined were calculated solely for standard (calen dar) quarterly and semiannual periods. It is unlikely that even these intrayear flows were truly ascertained, as only crude adjustments could be made for trends in loans and de posits. 94 TOWARD A SEASONAL BORROWING PRIVILEGE: A STUDY OF INTRAYEAR FUND FLOWS AT COMMERCIAL BANKS OBJECTIVES AND PROCEDURES To develop this statistical base for a sea sonal borrowing proposal, the following se quence is followed in this paper. First, a specific definition of intrayear fund flows is adopted and the fund flows at individual banks are calculated for selected periods. The flows at individual banks are then summed to show their aggregate origin, di rection, and magnitude. Next, attention is focused on banks at which seasonal flows of funds are large in relation to the size of the banks. Within the considerable limitations imposed by the data used, it is shown that a significant pro portion of banks do have large relative seasonal outflows of funds; that these banks tend to be small, presumably with limited access to financial markets that larger banks could use to meet such pressures; and that, because these banks are generally small, their borrowings under a seasonal discount program designed to serve them would be compatible with continuation of a limited over-all role for the discount mechanism. The report concludes with estimates of total borrowings under alternative discount programs that would allow these individual banks to borrow a portion of the funds that they need to meet their typical seasonal outflows. Among the aspects of discount operations being reviewed during the fundamental re appraisal of the discount mechanism is the attitude of the Federal Reserve System to ward member bank use of the discount win dow for seasonal needs. Present policy con templates that each member bank will main tain sufficient liquidity to meet seasonal swings that it might normally expect, with assistance at the window confined to deal ing with variations of unusual amplitude. However, because the liquidity of many banks has been reduced since this policy was formulated, it is pertinent to examine whether banks should now be permitted to meet a larger portion of their seasonal needs through discounting. Basic data required for such study are the flows of funds at individual banks, rather than the published summaries of banking data that reveal only the net flow totals for large groups of banks. The mag nitude and duration of individual seasonal flows, as well as their distribution among different types and sizes of banks, could af fect the advisability of liberalizing borrow ing at the discount window for seasonal pur poses. These factors would also be impor tant in formulating rules under which more liberal borrowing might be implemented. 95 96 FUND FLOWS AT INDIVIDUAL BANKS: DEFINITION Fund flows at each bank were calculated from data obtained on reports of condition (call reports) during two 12-month periods. The year from July 1962 to June 1963 was processed on a quarterly basis and was used because it is the latest period for which spring and autumn call data are available in machine-language form. To provide more recent data as well as a second year for comparison, the period from July 1965 to June 1966 was also processed. These were the latest data available when the analysis was performed. Reports of condition were available for every member and insured nonmember bank, but the only banks included in the study were those for which comparable re ports could be constructed for each call date in the period examined. Thus, banks that had been newly created or liquidated during the period were excluded, but banks in volved in a merger were retained by summing their separate reports prior to the merger to create data comparable to that reported by the merged bank on subsequent call dates. In this manner, the study of the 1962-63 period was able to cover 98 per cent of all banks in existence on June 30, 1963. The 1965-66 study covered 99 per cent of the banks in existence on June 30, 1966. At the time of the study, call report data constituted the only readily available series on assets and liabilities of individual banks of all sizes and in all regions. Such compre hensive coverage was greatly desired, given the purpose of the work. The principal dis advantage of these series, however, was that not enough measurements were provided during the year to ascertain the peaks and troughs of the seasonal swings at each bank or to measure accurately the duration of the flows. Also, in common with all data cover ing past bank performances, the call statis tics could not provide a measure of the ex tent to which banks may have curtailed lending seasonally or held back on seasonal loan expansion because of unavailability of funds. Different seasonal loan patterns might emerge at some banks if a seasonal discount program were adopted. Calculation of trend-adjusted intrayear changes Because each set of data covered only 12 months, it was impossible to calculate sea sonal components of pertinent bank asset and liability items with anywhere near the degree of sophistication commonly em ployed in seasonal adjustment of economic time series. The trend could be estimated only in a crude fashion, and there was no basis for separation of irregular movements from seasonal changes. In recognition of these large departures from the usual mean ing of “seasonal” in economic studies, the term “intrayear” is applied to the trendadjusted quarterly or semiannual changes computed in this study. Allowance for trend in a bank asset or liability item was achieved by first calcu lating the June-to-June change in the item. Then, one-fourth of this value was sub tracted from each observed quarterly change in the item, and one-half of the value was subtracted from each observed semiannual change. To illustrate this procedure, suppose that a bank experienced an increase of $100 million in deposits between the June 1962 and June 1963 call dates. The quarterly trend is calculated to be one-fourth of this IN T R A Y E A R FUND FLO W S AT C O M M E R C IA L B A N K S value, or an increase of $25 million. F ur ther, suppose that deposits at this bank ac tually increased by $10 million between the June 1962 and September 1962 call dates. The trend-adjusted change in deposits dur ing this quarter is therefore a decrease of $15 million, calculated by subtracting the quarterly trend increase of $25 million from the observed increase of $10 million. Examples for individual banks given in the next section will further demonstrate this statistical adjustment. Then, in the re mainder of the paper, all references to changes in bank asset and liability items will be to the trend-adjusted changes. To simplify the exposition, the qualifying term “trend-adjusted” will be omitted in those sections of the paper. Special term inology for asse ts, liabilities, and flows Several bank asset and liability items used in this study represent special combinations or adjustments of call report items. These are defined as follows: Net deposits are total deposits less cash items in process of collection and unposted debits drawn on the bank. IPC deposits are demand, time, and sav ings deposits of individuals, partnerships, and corporations, including time deposits accumulated for payment of personal loans. Nonfinancial loans are total gross loans less loans to financial institutions and loans for purchasing or carrying securities. The bulk of the study is concerned with examination of the net result of simultane ous changes in IPC deposits and nonfinan 97 cial loans, which is termed the fund flow. The study concentrated on changes in these particular assets and liabilities because they were thought to best approximate the sea sonal impact originating within the area served by the bank. Also, these items are relatively free of “window dressing” on call dates. The value of the fund flow for any given period was obtained by subtracting the change in nonfinancial loans from the change in IPC deposits, after both had been adjusted for trend. If the result is positive, it is called a fund inflow; if negative, a fund outflow. The calculation of fund flows in each period was performed separately for each bank. Aggregate fund flows for groups of banks were later computed by summing the flows at the individual banks. It was there fore possible to compute aggregate gross outflow or gross inflow by summing individ ual flows only at banks with outflow or in flow, respectively. The aggregate net fund flow for a group of banks was obtained by summing their individual fund flows irre spective of their direction. The result is either a net outflow or a net inflow, depend ing on its sign. To facilitate comparison of fund flows among banks of different size, relative fund flow for a given bank in a given period was calculated as the percentage that fund flow in that period was of the trend value of net deposits at the beginning of the period. In some analyses, the relative change in par ticular bank asset and liability items was also of interest, and in each case it was com puted in the same manner. 98 FUND FLOWS AT INDIVIDUAL BANKS: TWO EXAMPLES The concept of a fund flow that is the net result of intrayear changes in both deposits and loans is not a common one, and so an examination of how fund flows originated at individual banks may assist the reader in becoming comfortable with these data. The two examples presented here each represent the average experience of three banks that were similar in deposit size and in the direc tion and magnitude of their fund flows. In addition to showing how changes in deposits and in loans entered into the calculation of fund flow, the examples show how the rela tive magnitude of the flow is related to intra year changes in the loan/deposit ratio, and how certain other asset items fluctuated in response to the varying fund inflows and outflows. How the exam ples were chosen These examples are being given prior to the presentation and analysis of data for all banks. But the examples were in reality constructed after the analysis was com pleted, and therefore could be chosen to illustrate some of the principal findings of the analysis. One such finding was that while the major part of national intrayear fund flows originates at large banks, the flows at such banks are usually moderate in size when compared to the assets of the banks. To il lustrate this typical case, Bank A was con structed by averaging actual data for three fairly large Eastern banks, each of which had total deposits in the neighborhood of $500 million. The changes in deposits and loans at these banks were representative of the upper range of the moderate changes found at other large banks, and thus pro vide an example of how the principal (in terms of dollar amount) national flows originated. A second key finding was that numerous smaller banks, usually serving rural areas, are subject to rather severe intrayear fluc tuations in deposits and loans. The dollar amounts involved in these cases are small when compared to total national flows, but large relative to the assets of the smaller banks at which they tend to occur. These cases are illustrated by Bank B, which was constructed by averaging data for three banks in Nebraska, each of which had total deposits of about $3 million. Com putation of trend-adjusted intrayear flows The presentation that follows uses data from the 1965-66 period, in which the computa tions and analyses were made on a semi annual basis. Data for Bank A and Bank B are shown in adjacent columns to facilitate comparison. Changes in IPC deposits. Outstanding IPC deposits (in dollars) on each of the three call dates used during 1965-66 were: D ate Bank A June 30, 1965 ........... 488,409,000 D ecem ber 31, 1965 ..589,233,000 June 30, 1966 ............. 562,361,000 Bank B 2,684,000 3,549,000 2,952,000 The actual deposit changes during each half-year period were therefore: Period Bank A July-D ecem ber ...+ 1 0 0 ,8 2 4 ,0 0 0 Jan u ary -Ju n e .........— 26,872,000 Bank B +865,000 —597,000 To find the trend-adjusted deposit changes, the first step is to calculate the average semiannual change that is due to the deposit trend experienced at both banks. One-half of the change in deposits between June 1965 and June 1966 is regarded as the change due to trend in each semiannual period, as follows: Period Bank A July-D ecem ber ....+ 3 6 ,9 7 6 ,0 0 0 Jan u ary -Ju n e ........... +36,976,000 Bank B + 134,000 +134,000 99 IN T R A Y E A R FUND FLO W S AT C O M M E R C IA L B A N K S The trend-adjusted change in each period is then computed by subtracting the change due to trend from the actual change, which gives the following trend-adjusted deposit changes: Period Bank A J u ly -D e c e m b e r.........+63,848,000 January-June ........... —63,848,000 Bank B + 731,000 —731,000 Note that, with only two “seasons,” the trend-adjusted change for one period is necessarily the mirror image of that for the other period, as the sum of the trend-ad justed intrayear changes within any given annual period must be zero. Changes in nonfinancial loans. Similar com putations yield the trend-adjusted changes in nonfinancial loans. The outstanding amounts (in dollars) on the call dates were: D ate Bank A June 30, 1965 ........... 367,640,000 Decem ber 31, 1965 . 409,705,000 June 30, 1966 ......... 439,497,000 Bank B 2,414,000 1,664,000 2,485,000 Thus the actual semiannual changes were: Period Bank A July-D ecem ber . . . . +42,065,000 January-June ........... +29,792,000 Bank B —750,000 +821,000 And the semiannual changes ascribed to trend were: Period Bank A J u ly -D e c e m b e r........... +35,928,500 January-Jun e ............. +35,928,500 Bank B +35,500 + 35,500 Therefore, the trend-adjusted changes in nonfinancial loans were as follows: Period Bank A July-D ecem ber .........+6,136,500 January-Jun e ............. —6,136,500 Bank B -7 8 5 ,5 0 0 +785,500 Fund flows. The fund flow in each semi annual period is obtained by subtracting the trend-adjusted change in nonfinancial loans from the trend-adjusted change in IPC deposits, which yields: Period Bank A July-D ecem ber ...+ 5 7 ,7 1 1 ,5 0 0 January-Jun e . ...- 5 7 ,7 1 1 ,5 0 0 Bank B +1,516,500 -1 ,5 1 6 ,5 0 0 Relative changes and flows If the deposit and loan changes at each bank are related to a common base, such as assets or deposits of the bank, their relative contributions to the fund flow be come readily apparent. The base employed for all such comparisons in this study is the trend value of total net deposits at the bank at an appropriate date. Expressed as per centages of this deposit figure at each bank, the trend-adjusted changes and the fund flow are shown in Table 1. TABLE 1 RELATIVE TREND-ADJUSTED CHANGES AND FUND FLOWS In per cent Item Change in deposits Change in l o a n s ............ Fund flow ...................... July-December January-June Bank A Bank B Bank A Bank B +10.5 + 1.0 + 9.5 +23.3 -25.1 +48.4 -1 0 .5 — 1.0 — 9.5 -2 3 .3 +25.1 —48.4 These relative percentages provide desired comparisons among banks and will be much used in later sections. In this example, semi annual relative fund flow at Bank A, the larger Eastern bank, was 9.5 per cent of net deposits, whereas that at Bank B, the smaller rural bank, amounted to 48.4 per cent of net deposits, or about five times the magnitude at Bank A. At Bank A, the main semiannual swing occurred in deposits. To the extent that loans did change, they rose and fell with deposit volume and thus offset part of the deposit flow. But at Bank B, both deposits and loans exhibited relatively large semiannual swings. Furthermore, they moved in opposite directions— loan volume fell while deposits increased, and rose while deposits decreased— and thus accentuated the intrayear fund flow with which manage ment had to cope. The intrayear fund flow at each bank obviously caused seasonal movements in loan/deposit ratios, and it is interesting to compare the changes in this familiar statis 100 tic— (net lo a n s)/(n e t deposits), expressed as a percentage: D ate June 30, 1965 ............................... December 31, 1965 .................... June 30, 1966 ............................. Bank A Bank B 75.8 72.7 77.6 77.8 40.4 74.9 The severe intrayear fund flow at Bank B, and the correspondingly sharp seasonal swing in its liquidity position, reflected the seasonal demands of the agricultural econ omy on which this bank was largely depen dent. More than three-fourths of its loan volume was in loans to farmers, both in December and at the June peaks. To meet crop production expenses in the spring, farmer-customers of this bank drew down their deposits and also secured additional loans. The significant event from the bank’s standpoint, however, was that these funds did not remain with its other customers. Ap parently the funds were not spent locally, or they tended to flow out of the community after being spent. (Two of the three banks averaged to form Bank B were in one-bank towns; the third was in a two-bank town.) The high degree of association with agri culture at Bank B was not unique, as 22 per cent of the Nation’s banks had more than half of their loan volume in loans to farmers on June 30, 1966. Portfolio ad justm ents in resp o n se to fund flows When a bank experiences a fund inflow or outflow as defined here, other asset or liabil ity items must together show a net change that exactly compensates for the fund flow. As one part of this study, a look is taken at changes in two asset groups— balances on deposit at other domestic banks and hold ings of U.S. Government securities— in which banks are generally able and likely to make discretionary changes in direct re sponse to fund inflow or outflow. Trend-adjusted semiannual changes in these items were computed in the same man ner as previously illustrated for deposits and loans. Then, to facilitate comparison of these figures with the fund flow or with other changes at the bank, and to permit direct interbank comparisons, the relative trend-adjusted changes in these items are also shown in Table 2 as percentages of the December trend value of net deposits. TABLE 2 INTRAYEAR CHANGES IN TWO ASSET GROUPS In dollars unless otherwise indicated Asset group Balances with other banks: July-December Bank A + 7,800,000 Trendadjusted .. + 6,930,500 Relative +1.1 (per cent) U.S. Govt. securities: A ctual.......... +14,141,000 Trendadjusted. . +14,189,000 Relative +2.3 (per cent) Bank B January-June Bank A Bank B + 231,000 — 6,061,000 — 156,000 + 194,500 — 6,930,500 — 194,500 +6.2 -1 .1 -6 .2 +1,160,000 -14,237,000 —1,236,000 +1,198,000 —14,189,000 -1,198,000 +38.2 —2.3 —38.2 At both banks, semiannual changes in each of these components were in the direc tion expected as a consequence of the fund flows. At Bank A, the two items together accounted for the use of 36 per cent of the semiannual fund inflow (or conversely, sup plied the same proportion of funds to meet the outflow). At Bank B, the major part of the adjustment to the severe fund flow ap peared to occur in these two items, partic ularly in the holdings of U.S. Government securities. Prelim inary lessons from the exam ples The contrast between fund flows at the two composite “banks” is typical. The general ization of these differences is the purpose of this study and is thought to have major im plications for seasonal discount policies of the Federal Reserve System. At the larger and primarily urban bank, 101 IN T R A Y E A R FUND FLO W S AT C O M M E R C IA L B A N K S the fund flow arose mainly through changes in deposits. Deposits increased during the second half of the year and fell during the first half, whereas loan volume displayed little change. But at the smaller rural bank, loan volume underwent a substantial change because of very high dependence of the bank and the community on a single indus try that had a marked seasonal need for funds. For the same reason, deposits at the smaller bank also exhibited a greater rela tive intrayear change. Relative to its deposits, the larger bank found that small changes in U.S. Govern ment securities and in balances with other banks sufficed to cope with its fund flow. In contrast, the smaller bank had to make relatively large changes in these items, and on a relative basis its portfolio adjustment problem loomed much larger than that of the bigger bank. During part of the year, relatively large amounts had to be kept idle or invested in securities that could be readily liquidated to meet the coming seasonal out flow of funds. At the larger bank, the dollar amount of the semiannual fund flow was much greater than at the small bank— $57.7 million com pared to $1.5 million. However, the magni tude relative to the size of the bank— the im portance of the intrayear fluctuation to banking operations— was much smaller at the large bank— 9.5 per cent of net deposits, compared to 48.4 per cent at the small bank. Thus by providing a relatively small amount of funds, the Federal Reserve could materially assist the small bank in meeting its relatively large seasonal pressures. ORIGIN OF INTRAYEAR FUND FLOWS As defined for this study, fund outflows and inflows in a given quarter or half-year period can originate through various com binations of trend-adjusted changes in IPC deposits and in nonfinancial loans. (In the remainder of this paper, all intrayear data cited are trend adjusted.) These changes are examined on a semiannual basis during 1965-66 and on a quarterly basis during 1962-63, the latest period for which quar terly flows could be calculated. (An ex amination of semiannual flows during 1962-63 showed marked resemblance to those of 1965-66.) C hanges in IPC deposits The most prevalent influence on flows was a tendency for deposits to increase during the second half of the year and to decrease during the first half. In 1965-66, 81 per cent of member banks experienced this movement. Deposits rose at 64 per cent of the mem ber banks during the third quarter of 1962 and at 70 per cent over the fourth quarter. They then decreased at 75 per cent of mem ber banks in the first quarter of 1963 and at 60 per cent during the second quarter of that year. C hanges in nonfinancial loans At many banks, loan volume tended to in crease in the spring and decrease in the fall. For instance, loans fell at 57 per cent of the member banks during the second half of 1965 and rose from January through June of 1966. Loans decreased at 65 per cent of the member banks in the third quar ter of 1962 and at 49 and 62 per cent in the fourth quarter of 1962 and first quarter of 1963, respectively. But on the other hand, the volume of loans rose at 73 per cent of the member banks during the second quar ter of 1963. 102 Independence of loan and deposit changes During each period, the direction of the change in loans at an individual bank ap peared largely independent of the direction of the change in deposits. For instance, the proportion of member banks at which loans decreased during a given period was about the same in the group in which deposits were up as in the group in which deposits were down. Among the member banks at which deposits decreased from July to De cember 1965, for example, loans fell at 58 per cent and rose at 42 per cent. Among the member banks at which deposits increased, loans fell at 57 per cent and rose at 43 per cent. Similar independence between loan and deposit changes was observed in all periods studied. Coincidence of loan and deposit drains However, such independence of loan and deposit changes did not preclude loan and deposit drains from coinciding at a large proportion of banks during the spring sea son. Thus, in the first half of 1966 nearly one-half of member banks experienced a reduction in deposits combined with an in crease in loans, resulting in an outflow of $5.5 billion. Another one-third experienced a reduction in both deposits and loans, but the generally greater deposit changes re sulted in a net fund outflow of $3.2 billion. The first quarter of 1963 was charac terized by deposit reductions, which oc curred at three-fourths of the member banks. Loan volume also went down at more than half of these banks, but the total of the deposit changes was much larger and re sulted in a net fund outflow of $3.3 billion. Additional fund outflow of $2.1 billion oc curred at banks that experienced a rise in loan volume while their deposits decreased; such banks constituted 27 per cent of all member banks. Only one-fourth of member banks had deposit gains, for net fund inflow of just $0.7 billion. Net outflow therefore totaled $4.6 billion at all member banks, the largest quarterly net outflow of the year. The deposits-down, loans-up squeeze was most common during the second quarter, when 42 per cent of member banks experi enced a drop in deposits while loan volume rose. Fund outflow arising from this squeeze totaled $2.7 billion. At another 18 per cent of the banks, at which both loan and deposit volumes were reduced, there was an addi tional $0.6 billion of net fund outflow. Off setting net fund inflow of $0.8 billion oc curred at the 40 per cent of banks at which deposits rose during this quarter. Total net fund outflow at all member banks was there fore $2.5 billion, or only slightly more than one-half of the net outflow of the first quar ter. As noted, however, more banks ex perienced the deposits-down, loans-up squeeze than in the first quarter. FUND OUTFLOWS: SUMMARY OF AGGREGATE DATA In this section the primary focus settles on banks with fund outflow during a given period, in the belief that these banks con stitute the prime candidates for use of sea sonal discount credit during that period. In examining fund outflow at a given group of banks, two statistics seem very relevant: the proportion of banks that ex perienced outflow, and the dollar amount of that outflow. On the semiannual basis, 22 per cent of member banks had outflow during the sec ond half of 1965. This outflow totaled $0.9 billion and amounted to 1.7 per cent of net deposits at these banks. In the first half of 1966, 78 per cent of 103 IN T R A Y E A R FUND FLO W S AT C O M M E R C IA L B A N K S member banks had outflow that totaled $9.1 billion and amounted to 4.7 per cent of net deposits at such banks. Semiannual data for 1962-63 exhibited approximately the same relationships. Quar terly member bank data for this period were as shown in the accompanying table. At the banks with outflow, the amount ranged from 2.2 per cent of their net deposits in the third quarter to 3.8 per cent in the first quarter. The proportion of all member and in sured nonmember banks with outflow in each period was virtually identical to the proportion of member banks alone. But in Q uarter Percentage of banks w ith outflow Outflow (billions of dollars) 3 4 1 2 31 37 68 69 2.1 .9 5.7 4.1 each period the outflow at those nonmember banks with outflow was greater relative to their net deposits. In the remainder of this report, data will be limited to member banks. These banks constitute the group that would be immedi ately eligible to take advantage of discount regulations designed to provide more sea sonal credit. RELATIVE OUTFLOWS AT INDIVIDUAL BANKS The relative size of the fund outflow for a given period at each bank was measured by comparing the outflow to the trend value of total net deposits at the start of the period, as described and illustrated earlier. These percentages provide a basis for com paring outflows among banks of different size that more nearly reflects the magnitude of the portfolio adjustment and other prob lems posed by the outflow than does the dol lar amount of the outflow. In each period studied, outflows at most banks were limited to less than 10 per cent of deposits. In fact, during each semiannual period, about one-half of the banks with outflow experienced outflows amounting to less than 5 per cent of their net deposits, and during each quarter over three-fifths of the banks with outflow were within this figure. The bulk of the total outflow oc curred at these banks with small or mod erate individual outflows. But in each period some banks had relatively large outflow. There were few large outflows on a rela tive basis during the second half of both 1965 and 1962, and the total outflow at these banks was small (Table 3 ). TABLE 3 FUND OUTFLOWS AT INDIVIDUAL BANKS Relative outflow (percentage of net deposits) Percentage of all member banks with specified outflow July-December 1965 ........ Under 5.0 ...................... 5.0-9.9 ............................ 10.0 and over .............. July-December 1962 ........ Under 5.0 ...................... 5.0-9.9 ............................ 10.0 and over .............. 22 January-June 1966 .......... Under 5.0 ...................... 5.0-9.9 ............................ 10.0 and o v e r ................ January-June 1963 .......... Under 5.0 ...................... 5.0-9.9 ............................ 10.0 and over ................ 78 39 25 14 78 37 25 16 17 4 2 22 17 3 1 Total outflow (billions of dollars) .9 .6 .2 .1 .6 .4 .1 .1 9.1 3.8 3.9 1.5 7.7 2.6 3.6 1.6 D ata for the fourth quarter of 1962, however, reveal a greater frequency of the larger relative outflows— a finding that is submerged in the semiannual data because these banks had either fund inflow or only small fund outflow during the third quarter. During the fourth quarter, in which the banking system as a whole was experiencing its greatest net fund inflow, 12 per cent of member banks had outflows amounting to at least 5 per cent of their deposits, and at 4 per cent of banks the outflow equaled 10 per cent or more of deposits. 104 Large relative outflows were common in the first half of both 1966 and 1963. About one in every seven member banks experi enced outflow equal to at least 10 per cent of its net deposits, and at some banks the ratio of outflow to deposits was consider ably above this level (Table 3). However, the banks with large relative outflows were evidently the smaller banks. For instance, in 1966 the 14 per cent of banks at which outflow was at least one-tenth of deposits accounted for only 16 per cent of the total outflow of $9.1 billion. About two-thirds of member banks had outflows in each of the quarters in the first half of 1963. In each quarter, one-fourth of the member banks had outflow that equaled or exceeded 5 per cent of their deposits. About 8 per cent of the banks had outflows exceeding one-tenth of deposits, and these banks accounted for only a minor propor tion of the total quarterly outflows. A SEASONAL BORROWING PRIVILEGE Design of a seasonal discount program It has been shown that only a minor part of total fund outflow occurs at banks with large relative outflow. Data presented in this sec tion further indicate that large relative out flows occur much more frequently among small banks than among the larger insti tutions. Considered jointly, these findings have several implications for the design of a discount program that would permit more seasonal borrowing by member banks. To serve only banks with large relative outflows. First, if the seasonal discount program were limited to banks with the larger relative outflows, a significant number of small banks would be assisted in making portfolio adjustments but the total amount of funds supplied would constitute a small propor tion of total reserves in the banking system. Second, the small banks that would com prise the majority of banks eligible for the program are likely to be operating at a disadvantage in the present financial mar kets that are commonly employed for port folio adjustment purposes. The discount route for seasonal funds should therefore be a relatively attractive one for such banks. Third, many of the small banks with large relative seasonal flows are probably involved heavily in financing agriculture, a sector that in recent decades has been generating credit demands in excess of its contribution to the growth of rural banking resources. The sea sonal discount program would therefore provide a net addition to the lending re sources of such banks that currently find it difficult to meet the aggregate local de mands for farm credit. To require banks to fund part of outflow. An additional practical consideration en ters into the design of a discount program to serve banks with large relative seasonal fund flows. Only that part of the seasonal outflow exceeding a specified relative level ought to be funded through discounting; otherwise, banks would unwisely be given an incentive to achieve seasonal outflows. But if each bank were required to meet, through portfolio adjustment, all seasonal outflow up to a specified proportion of its resources, the incentive to undertake opera tions that deliberately create or accentuate seasonal flows would be largely removed. At the same time, a bank experiencing large relative seasonal credit demands that it be lieves should be met would be encouraged to do so, and by discounting would be able to obtain funds for this purpose to meet 105 IN T RA YEA R FUND FLO W S AT C O M M E R C IA L B A N K S demands that it might otherwise be unable to fulfill. The level of the “deductible” quantity— the amount of its seasonal outflow that a bank would be required to meet from sources other than borrowing at the discount window— could be set at a given percentage of average deposits. The level at which this percentage is set affects both total poten tial borrowings and the distribution of the potential borrowings among large and small banks. As the level is lowered, more of the larger banks that experience moderate in trayear outflows qualify for seasonal bor rowing, and potential total borrowings in crease rapidly. As the deductible is raised, potential total borrowings diminish, but so does the value of the program to those banks with large relative outflows. On the basis of the intrayear outflow data that have been presented, deductible levels of 5 and 10 per cent of net deposits may represent approximate lower and upper limits, respectively, of the deductible for a seasonal borrowing program that might pro vide significant assistance to many banks, yet keep potential borrowings within the scope permitted to the discount mechanism in recent years. Under the 5 per cent de ductible, potential borrowings in the spring are estimated at $2 billion, with just over one-half of the sum going to banks with deposits of $100 million and over. Potential springtime borrowings under the 10 per cent deductible plan are estimated at $400 million, with perhaps two-fifths of the total being borrowed by the large banks. As a representation of real potential bor rowings, these estimates are subject to the same basic weakness as the fund outflows considered throughout the study— the in herent disadvantage of being based on quar terly or semiannual observations that are unlikely to have measured the true seasonal peaks and troughs, and the likelihood that with discounting providing a source of addi tional seasonal funds, some banks would make additional seasonal loans that they were unable to make during the past periods that have been examined. Potential borrowings by period and by size of bank In both 1965 and 1962, relatively few member banks had large or even moderate relative outflows in the second half as a whole. The amounts exceeding the deduct ible were small even under the 5 per cent rule (Table 4 ). Most of the seasonal bor rowing that might occur in this period would evidently be at small banks. This would be particularly true in the fourth quarter of the year, in which a fair proportion of small banks, but no large banks, had large rela tive outflow. On a quarterly basis, poten tial borrowing demands appear larger in the fourth quarter than in the second half as a whole but are still relatively small sums. TABLE 4 POTENTIAL BORROWINGS BY SIZE OF BANK Net deposits at bank (millions of dollars) July-December 1965 . . Under 1 0 .................. 10-99 ........................ 100 and over .......... Tuly-December 1962 .. Under 1 0 .................. 10-99 ........................ 100 and over .......... January-June 1966 . . . Under 1 0 .................. 10-99 ........................ 100 and over .......... January-June 1963 . . . Under 10 .................. 10-99 ........................ 100 and over .......... Percentage of specified banks with outflow over— 5% of 10% of deposits deposits 5 7 3 1 6 3 1 5 39 43 34 32 41 44 35 36 2 2 1 1 2 1 14 19 7 5 16 20 8 6 Total outflow (millions of dollars) over— 5% of deposits 140 48 75 17 91 37 43 11 1,989 385 565 1,039 2,087 420 514 1,153 10% of deposits 54 19 36 25 15 11 409 150 100 159 441 184 114 144 Major borrowing under a seasonal dis count program would evidently occur in the first half of the year. A rather large proportion of small member banks— about one-fifth of those with deposits under $10 106 million— might be eligible for borrowing even under the 10 per cent deductible plan. Some large banks also would be eligible and would account for a significant portion of the total potential borrowings, on the basis of data for 1966 and 1963 (Table 4 ). At $2 billion, total potential borrowings un der the 5 per cent deductible were almost five times the potential borrowings under the 10 per cent plan. The proportion of potential borrowings at large banks was larger under the 5 per cent deductible. On a quarterly basis, potential borrow ings appeared lower than those just shown for the first half as a whole because many banks had outflow in both quarters of the period. The cumulative data covering both quarters may therefore be more indicative of the level that potential borrowings could reach in this peak outflow period. Potential borrowings under alternative deductible levels There are large differences in banks eligible and in potential borrowings under the 5 per cent and 10 per cent deductible levels. The 10 per cent level appears rather restric tive, unless it is found that many banks have in fact been forced to limit seasonal lending significantly in recent years. But the 5 per cent deductible, under which two-fifths of member banks might be eligible, perhaps violates the intent to limit the program to banks with relative outflows significantly above average. Under other alternative de ductible levels within this range, using data for 1965-66, potential borrowing is as shown in Table 5. TABLE 5 POTENTIAL BORROWINGS BY DEDUCTIBLE LEVELS Outflow during period exceeds net deposits by at least— July-December 1965: 5 per cent ................ 6 ................ ................ 7 ................ 8 9 ................ 10 ................ January-June 1966: 5 per cent ................ 6 ................ 7 .......... 8 ................ 9 ................ 10 ................ Percentage of member banks with specified outflow Total outflow (millions of dollars) exceeding specified percentage of net deposits 5 4 3 3 140 113 90 75 63 54 39 32 26 1,989 1,422 2 2 22 17 14 1,022 724 531 409 CAPITAL AND CREDIT REQUIREMENTS OF AGRICULTURE, AND PROPOSALS TO INCREASE AVAILABILITY OF BANK CREDIT Emanuel Melichar, Board of Governors of the Federal Reserve System Raymond J. Doll, Federal Reserve Bank of Kansas City Contents I. Introduction_______________________________________________________________________ 109 II. Sum m ary of Findings and P roposals_____________________________________________ _ _ 1 0 9 Part 1. CAPITAL AND CREDIT REQUIREMENTS OF AGRICULTURE_____________________ 114 III. Capital R equirem ents, 1 9 5 0 -7 9 ----------------------------------------------------------------------------- 115 Capital stock of agriculture, 1950-80 Capital requirements by asset group, 1950-79 Total capital flows, 1950-79 IV. Credit R equirem ents, 1 9 5 0 -7 9 ____________________________________________________ 129 How have capital requirements been financed? Projected credit requirements, 1970-79 V. Sources of Credit, 1 9 5 0 -7 9 ________________________________________________________ 136 Sources of outstanding credit, 1950-68 Relative role of banks, 1950-68 Projected credit expansion by major lenders, 1970-79 Supply of funds at rural banks VI. Seasonal Production C redit------------------------------------------------------------------------------------145 Seasonal capital requirements Seasonal credit extensions Institutional sources of seasonal credit Part 2. PROPOSALS TO INCREASE AVAILABILITY OF BANK CREDIT____________________150 VII. C orrespondent and Branch Banking----------------------------------------------------------------------- 151 Correspondent banking Branch banking VIII. Federal Reserve C redit---------------------------------------------------------------------------------------- 159 Seasonal discount credit Longer-term credit IX. Unified M arkets to Serve Rural B anks--------------------------------------------------------------------- 167 Organization A secondary market for rural bank loans Other services of unified markets X.--Concluding C om m ents----------------------------------------------------------------------------------------— 1 7 2 107 CAPITAL AND CREDIT REQUIREMENTS OF AGRICULTURE, AND PROPOSALS TO INCREASE AVAILABILITY OF BANK CREDIT I. INTRODUCTION A large proportion of the Nation’s banks are located in rural areas where agriculture is the primary economic base. Deposit trends at these banks— and loan demands made on them— derive mainly from devel opments in the agricultural economy. Thus, many aspects of the well-known “revolu tion” in the structure of agricultural produc tion and related rural business have had major impact on rural banks and promise to continue to exert similar influence for some time. This study gives special atten tion to those problems of rural banks that arise from the peculiar nature of and changes in their agricultural environment. It seeks to determine how the Federal Re serve discount mechanism might be made more helpful to those banks. The examination of past and projected agricultural and rural banking trends, how ever, suggested that maintenance of the present leading role of banks in rural lend ing will likely require institutional changes beyond those that appear feasible in dis count administration and other Federal Re serve policies. Thus after documenting the growing capital requirements of the agri cultural sector and the increasing inability of rural banks to finance their usual share of the resulting credit demands, this report outlines a broad program designed to in crease materially the flow of funds from national capital and money markets into rural areas through the banking system. This paper was first prepared in 1966, and was expanded and updated in October 1969. II. SUMMARY OF FINDINGS AND PROPOSALS economy, which brought higher prices for some purchased inputs and added to de mand for land. And a substantial portion resulted from land price increases to which farm enlargement, land improvement pro grams, and other technologically induced pressures contributed. In addition, as agri culture purchased more production inputs, capital requirements of related rural busi nesses also rose. Several agricultural economists have re cently studied farm capital growth. Each concluded that the value of capital stocks Capital used in agriculture has been in creasing rapidly. Since 1950, for example, the value of farm assets of a primarily pro ductive nature has risen by 131 per cent. This growth is traced mainly to techno logical developments that have prompted enlargement of individual farms and substi tution of purchased inputs for labor and farm-produced inputs. Some of the capital growth occurred as farmers added to physi cal stocks of machinery, livestock, and other assets. Another part can be ascribed to growth and price inflation in the nonfarm 109 110 will rise further, though they differed on the rate of growth and on which assets will lead the advance. By using information mainly from these studies, three alternative capital models are developed in this study. In the lowest of these estimates, the value of farm assets projected for 1980 is 28 per cent above that of 1969, whereas the highest estimate indicates a gain of 74 per cent. From the projected capital stocks, esti mates are made of the implied yearly capi tal flows— the capital requirements that must be financed in some manner. As the capital assets of agriculture increase, larger annual flows of capital are generally re quired to make real additions to stocks, re place equipment as it depreciates, and trans fer assets from one farmer to the next. An nual capital flows for these purposes are estimated to have averaged $7 billion during the 1950’s and to have been fairly stable during the late 1950’s and early 1960’s. By 1965-68, however, the annual flow aver aged $11 billion per year. And under the three alternative capital models formulated herein, capital flows are projected at from $13 to $19 billion in 1975-79. Annual capital flows are financed either internally from cash flow— depreciation al lowances and net income— or externally by expanded use of credit. Upon comparing estimated capital flows, known expansion of credit, and estimated cash flows, it appears that the proportion of cash flow allocated by farmers to capital needs declined during the 1950’s. Consequently, the share of capi tal spending financed by debt rose from 13 per cent in the early 1950’s to 31 per cent in the early 1960’s. Then, the proportion of income allocated to capital apparently sta bilized, but because capital spending rose more sharply than income, the share fi nanced by debt reached 37 per cent during 1965-68. These findings provide a framework within which future farm credit demands may be projected. For the estimates made herein, capital flow requirements and de preciation allowances were taken as pro jected by the three alternative capital models, net farm and nonfarm income was projected on the basis of recent trends, and the proportion of cash flow that farmers would allocate to capital spending was pro jected at the level that prevailed in the 1960’s. Outstanding farm debt, which rose from $10.7 billion in 1950 to $23.6 billion in 1960 and $52.0 billion in 1969, in the lowest projection increases to $91 billion by 1980 and in the highest to $137 billion. The lowest projection implies that debt will in crease by about 5 per cent annually, a sig nificant slowdown from recent growth rates averaging 9 per cent, but a rate that never theless calls for $3 to $4 billion of net addi tions to outstanding debt annually between now and 1980. The highest projection calls for debt to rise by $79 billion during the next decade, which would require annual rates of increase similar to those of the 1960’s. Increased credit to agriculture has been supplied in three important ways. First, more sellers of farms have been taking mort gages or using land contracts. Individuals have been providing about one-fifth of the additions to outstanding farm debt. Second, money and capital market funds have been channeled into agriculture through the lend ing operations of life insurance companies, Federal land banks, production credit as sociations, and national farm supply corpo rations. Such funds have provided about one-half of the growth in farm credit. Third, commercial banks have been supplying about one-fourth of the additional credit. Some of these loans have been made by large money market banks, either directly C A PIT A L AN D C R E D IT R E Q U IR E M E N T S OF A G R IC U LT U R E or through correspondent relationships with rural banks; much of the loan expansion, however, has occurred at smaller country banks. Rural banks have increased loans at a much faster pace than their deposits have grown, a divergence made possible by the low ratio of loans to deposits found at most banks when World War II ended. Through lending supported by the past accumula tion of deposits, bank credit to farmers has almost kept pace with the total expansion of farm credit, even though deposits, being dependent on gains in aggregate farm in comes and savings, rose at a much slower rate. However, expansion of bank lending by a relative shift from security investments to loans obviously could not be sustained for ever. Individual rural banks began to reach a “tight” position during the 1950’s, and a large proportion have now reached the point at which further reductions in liquidity do not appear feasible, given present institu tional arrangements. As these banks in clude most of the larger institutions and those that have been most active in meeting the credit demands of their areas, much of the Nation’s farm loan volume is affected. In the last few years, loan demands would have pressed harder against rural banking resources had not time deposits grown at an extremely rapid pace. Unfortunately, a lower rate of deposit expansion may realis tically be expected over the next decade. When the three alternative farm credit de mand projections are compared with pro jected deposit expansion, two indicate that banks as a whole will find it difficult to sup ply from their own resources the same share of farm credit growth that they have pro vided since 1950. If rural banks are to maintain their relative role in farm lending, this analysis indicates that they must draw 111 increasing proportions of their loan funds from sources other than local deposits. Several existing arrangements permit fund flows between urban and rural areas via banks. In unit-banking States, city bank participations in farm loans channel urban funds into farm lending. A thorough ex amination of this mechanism, however, leads to serious doubts that it can develop sufficiently to fill the credit gap. Its present use is largely restricted to dealing with over lines rather than with general credit deficits at country banks; in fact, since the usual “payment” for the service consists of de posits maintained at the urban correspon dent, the net flow of funds in most cases appears to be to the city rather than the rural bank. For those rural banks that are short of loanable funds, correspondent credit would be more helpful if it could be paid for by fees rather than balances, and development of this practice is advised. However, the generally tight liquidity posi tions of city banks will hardly lead them to favor this change or to increase significantly the supply of correspondent credit if it were adopted. In States with large branch-banking sys tems, funds can flow internally from urban offices to rural branches where loan demand exceeds deposit inflow. Studies of branch systems show that such flows do occur, and that at particular branches the funds so ob tained are often relatively greater than a unit bank would have been likely to obtain through the correspondent-banking system. Thus, in States that have well-developed statewide branch systems and also urban areas sufficiently large either to provide sur plus funds or to support a bank large enough to tap national money markets, the supply of bank funds to farm lending ap pears more likely to remain adequate pro vided that the managers of the branch sys- 112 terns maintain both interest and competence in farm lending. But even if the latter con dition were met, it seems doubtful that ex pansion of branch banking to rural areas of present unit-banking States will provide an adequate near-term, solution to main tenance of banking’s role in farm lending. If laws restricting branching are liberalized at all, initial changes are likely to permit only limited branching arrangements. Fur thermore, in some rural States with limited urban development, even statewide branch banking might not have a sufficient urban base to increase materially the flow of funds into the rural areas. New approaches are therefore recom mended. To maintain farm lending opera tions of commercial banks in a fully viable condition— in fact, to improve them at banks that are already experiencing the diffi culties cited— two broad proposals for chan neling funds to rural banks are made herein. First, greater amounts of Reserve Bank credit should be provided directly to rural banks through changes in the nature and administration of the discount mechanism. Second, new institutional arrangements should be established to permit greatly in creased rural bank participation in national capital and money markets. Small rural member banks have made limited use of System discount facilities in recent decades. The discount window may have been avoided partly because of the manner in which it was administered— the “reluctance to borrow” may have devel oped into a considerably larger deterrent against borrowing by the smaller banks. In addition, temporary fund needs at rural banks are usually for relatively lengthy periods such as a crop production season, and borrowing arrangements at most Re serve Banks have been ill-adapted to han dling such needs. In fact, a strict interpre tation of the regulation held that borrowing for normally expected seasonal outflows of funds was inappropriate. Thus, administration of the discount win dow that removes any previous stigma as sociated with borrowings for small short term adjustments, and that permits borrow ing for lengthy seasonal periods under equally clear guidelines, should encourage use of the discount window by rural banks. Seasonal borrowing privileges, in particular, would benefit the significant number of small rural banks and the communities they serve, because farm customers have a large relative seasonal fund demand. By borrow ing from the Federal Reserve to meet such seasonal outflows, these banks could employ for other community loan needs the funds that now must be set aside for the seasonal demands and that therefore either remain idle, or are temporarily invested outside the community, for up to half the year. A seasonal borrowing privilege appears able to provide prompt and significant as sistance to rural member banks facing rela tively large seasonal demands, but could not be employed by the many rural non member banks and would likely be rela tively insignificant to rural member banks in areas of balanced crop and livestock production, in which farm credit demands occur throughout the year rather than sea sonally. A complementary and more gen eral approach— one that would benefit all rural banks— would aim to reduce the capi tal market imperfections that now largely prevent small and rural banks from using these national markets as a source of funds. To this end, a second set of proposals is set forth under “Unified markets to serve rural banks.” These markets would be de signed to place small and rural banks on a more nearly equal competitive footing with other participants in the national capital and C A PITA L A N D C R E D IT R E Q U IR E M E N T S OF A G R IC U LT U R E money markets by minimizing the disadvan tages that result from the small size and isolated location of these banks. The major objective of unified markets is seen as facili tating sale of a wide variety of bank assets and liabilities, thereby encouraging national money market funds to flow into rural areas through the banking system much as they presently can through the cooperative credit system. Unified markets could provide rural banks with information and arrangements 113 for effective trading in Federal funds, Gov ernment securities, and certificates of de posit issued by these banks, in addition to a secondary market for loans. In each of these endeavors, they would strive to overcome the market imperfections that now place small and rural banks at a relative dis advantage, and would thereby secure more equitable allocation of money market funds among sectors of the economy and regions of the Nation. Part 1. CAPITAL AND CREDIT REQUIREMENTS OF AGRICULTURE Farmers’ use of credit increased almost five times in the aggregate and nine times on a per-farm basis between January 1, 1950, and January 1, 1969. Total debt (exclusive of Commodity Credit Corporation debt) rose from $11 billion to $52 billion; debt per farm increased from $1,900 to $17,000. Several factors combined to bring about this large expansion: new technology spurred upward trends in total farm capital stocks and production expenses; technology also permitted enlargement of individual farms, with associated capital demands; prices of some capital goods— particularly real estate and machinery— advanced con siderably; and finally, farmers financed an increasing proportion of their capital re quirements by borrowing. Farm debt as a percentage of selected production assets rose from 8.8 per cent in 1950 to 18.5 per cent in 1969. Since the major forces responsible for the rapid growth of farm debt from its low point of 1946 continue to prevail, there is widespread expectation of further credit expansion. Few studies, however, have at tempted to quantify these expectations in a reasonably rigorous and comprehensive fashion. One study that did cover all farm debt was generally assumed to have reached a bullish— perhaps even alarming— con clusion by projecting outstanding farm debt of $100 billion in 1980. In fact, however, this projection implied a substantial slow down in the rate of credit expansion, which followed as a consequence of the much re duced rates of future capital spending and land price inflation that were assumed in the study. Other analyses of investment and land prices appear to support much higher expectations, but their authors stopped at projecting the value of capital stocks rather than also examining the implied capital flows and credit demands. This paper therefore attempts first to ascertain and analyze postwar capital flows in agriculture and then to remedy the paucity of projections of such flows. In Section III, the nature and magnitude of past and future capital requirements are explored. Uses of capital are identified, and the flow of capital into each use is esti mated. Projections of capital flows for 1970-79 are then derived for each of three projections of farm capital stocks in 1980 that have been published in studies by other analysts. Section IV then attempts to determine likely future credit demands, given the pro jected capital flows. To provide a basis for such credit projections, financial data for 1950-68 are examined to ascertain trends in the manner that capital flows required in this period were financed— whether intern ally from depreciation allowances and net income or externally through increase in 114 CAPITAL AND CREDIT REQUIREMENTS OF AGRICULTURE debt. Then, with the aid of specific assump tions about future income and financial be havior of farmers, probable additions to debt are projected. (With additional time and resources, development of a model in which capital, income, and savings flows are jointly determined would be a preferable procedure, and perhaps will be inspired by these preliminary efforts.) After projection of total credit demands, attention turns to the various lenders that may supply these funds. Again, although apprehensions have often been expressed about the continued ability of certain farm lenders— particularly commercial banks— to continue rapid expansion of farm credit, no previous study has pitted specific alterna tive projections of credit demands against projections of bank lending resources, in order to determine the situations in which those fears might be justified. This analysis is attempted in Section V. First, sources of additions to farm debt during 1950-68 are examined in order to ascertain the share of credit provided by each lender group. Then, 115 for each of the alternative credit projections derived in the preceding section, estimates are made of the amount by which banks would have to expand their farm lending in order to maintain their relative role in this market. The various required rates of ex pansion in loans are compared with the projected rate of growth in deposits, to de termine the conditions under which banks are likely to experience future difficulty in meeting farm loan demands from their own resources. Credit extensions to meet seasonal capital requirements are treated separately in Sec tion VI. Because neither seasonal expenses nor total seasonal loans are measured di rectly, little quantitative analysis of these flows has been attempted at the national level. However, in Section VI an attempt is made to provide indicators of the trend in seasonal capital needs and in seasonal credit provided by banks and production credit associations. The relative extent to which these two lenders have met the increased seasonal needs is then estimated. III. CAPITAL REQUIREMENTS, 1950-79 Measurement, analysis, and projection of capital used in agriculture have primarily dealt with stocks of assets and with past and expected changes in those stocks. The U.S. Department of Agriculture annually publishes the value of several categories of farm assets such as real estate, machinery, and livestock. Analytical studies have re lated observed changes in these series to changes in various farm and nonfarm fac tors. On the basis of these observed relation ships, together with estimates of future trends in the causal factors, several recent studies have projected values of major farm assets to 1980. This section begins with a brief review of past developments and of three selected projections of capital stocks. • These data alone, however, prove inadequate as indica tors of the actual flow of capital into agri culture, both past and future. The annual capital flows, although related, are not equivalent to changes in the value of stocks. In particular, large amounts of capital are required annually to replace machinery that has worn out or become obsolete and to finance transfers of real estate. Thus, in a given year the value of stocks could remain unchanged because of stable prices and no net real investment, but several billion dol lars of capital would be required by replace ment and transfer transactions. Conversely, 116 although price increases of machinery or land that cause assets to be revalued upward would have the same proportional effect on replacement and transfer transactions, the dollar increase in the latter would be only a small fraction of that in stocks, because only a portion of the stocks is replaced or transferred in any given year. A significant analytical contribution of this section, therefore, is calculation of past annual capital flows and of flows implied by the stocks projected for 1980. Data on most kinds of capital spending were avail able from the USD A, but one very impor tant category— real estate transfers prior to 1965— had to be estimated. Capital spend ing and transfers implied by each projection of stocks were also estimated, with attention to whether an increase in stocks was ex pected to result from price rises or from real additions. Each type of asset is dis cussed separately, to consider the factors that probably caused past changes in the an nual capital flow that it required and hope fully to establish a basis for projection of probable future change. The projected com ponents are then summed to obtain three alternative projections of farm capital flows during the 1970’s. C a p ita l s to c k o f a g ric u ltu re , 1 9 5 0 - 8 0 The stock of various types of farm capital, valued at current market prices, is estimated annually by the USDA. Table 1 shows that selected assets of a primarily produc tive nature totaled $281.1 billion as of Jan uary 1, 1969. These assets— machinery, livestock, stored crops, working capital, and real estate— constitute the capital analyzed in this study. The account includes some nonproductive assets such as dwellings, per sonal cars, and some forms of personal sav ings. It excludes the two other personal as sets included in the USDA’s Balance Sheet of Agriculture— household equipment and investments in cooperatives— as well as other personal assets owned by farmers, such as nonfarm investments and the cash value of life insurance policies, that are not included in the Balance Sheet. As in the Balance Sheet, all farm assets of the selected types are included in the totals, whether owned by farmers, nonfarm landlords, or other persons or institutions. Composition and trends. The selected agri cultural assets increased in value in every postwar year except 1950 and 1954, for a total gain of $159.3 billion since the be ginning of 1950. Annual increases during the 1950’s averaged 4.5 per cent, fell to 3.4 per cent during 1960-64, but then ac celerated to 5.9 per cent in the 1965-68 period. Real estate remains the most important farm asset, and indeed its relative value rose from 62 per cent of total assets in 1950 to 72 per cent in 1969. Of the real estate value, perhaps one-fifth is contributed by TABLE 1 VALUE OF SELECTED ASSETS USED IN AGRICULTURE Billions of dollars A sset Per cent of total 1950 1955 1960 1965 1969 1950 1955 1960 1965 1969 Vehicles, machinery, and e q u ip m e n t................. L iv e sto c k ....................................................................... Stored c r o p s ................................................................. Demand deposits and cu rren cy ............................ Time deposits and savings bonds ...................... Real estate ................................................................... 12.2 12.9 7.6 7.0 6.8 75.3 18.6 11.6 9.6 6.9 7.5 98.2 22.2 15.2 7.7 6.2 7.6 130.2 25.5 14.5 9.2 5.9 7.9 160.9 32.6 20.1 10.5 6.3 9.0 202.6 10 11 6 6 6 62 12 7 6 5 5 65 12 8 4 3 4 69 11 6 4 3 4 72 12 7 4 2 3 72 Total selected assets ........................................... 121.8 152.0 189.1 223.9 281.1 100 100 100 100 100 S o u r c e . — T h e B alan ce S h e et o f A g ricu ltu re, 1968, U SD A , Jan. 1969, pp. 10, and 26 and 27. Data are shown as of January 1 of each year. 117 CAPITAL AND CREDIT REQUIREMENTS OF AGRICULTURE farm dwellings and service buildings and the remaining four-fifths by land and land im provements. In second place among asset groups, the machine stock— vehicles, machinery, and equipment— comprised 12 per cent of assets in 1969 and has roughly maintained this proportion since 1950. Livestock ranked third in 1969, at 7 per cent of the total. Stored crops and financial working balances each represented about 5 per cent and have been declining in relative importance. Changes in asset values over 5-year in tervals since 1950 are shown more explicitly in Table 2 (dollar changes occurring during 1965-68 were multiplied by 1.25 to express them as a 5-year rate comparable to the previous periods). Prominent features in clude the following: (1 ) increases in real estate values accounted for a large propor tion— an average of 80 per cent— of the gain in total assets; (2 ) growth in value of machinery and livestock involved consider able sums in some years, but varied con siderably over the period; and (3 ) asset growth in 1965-68 proceeded at an extra ordinarily rapid rate, as growth in machin ery, livestock, and real estate values each accelerated. Real versus price changes. In contrast to the changes in current value discussed above, the total farm physical plant, often referred to as real assets, has expanded rather slowly since 1950 (Table 2 ). According to USDA estimates, the se lected farm assets, when valued at constant prices, rose by only 15 per cent in 1 9 5 0 68. As the current value of these assets increased by 131 per cent, by implication the total price rise during the period was estimated as 101 per cent. The separation of capital growth into its real and price components is important to analysis and projection of capital flows, TABLE 2 CHANGES IN VALUE OF SELECTED ASSETS USED IN AGRICULTURE Asset 1950-54 1955-591960-64 1965-69* 5-year total (billions of dollars) Vehicles, machinery, and e q u ip m en t............................ 8.9 7.0 1.6 6.4 —1.7 2.0 3.6 4.0 —1.9 3.3 —.7 1.5 —.1 —.7 —.3 .5 Real estate .............................. .7 22.9 .1 32.0 .3 30.7 1.4 52.1 Total selected assets . . . 30.2 37.1 34.8 71.5 Stored crops ............................ Demand deposits and currency .............................. Time deposits and savings Percentage change in current value Vehicles, machinery, and eq u ip m en t............................ Stored c r o p s ............................ Demand deposits and currency .............................. Time deposits and savings 52 —13 26 19 36 -2 0 15 -5 19 —1 -1 0 -5 Real estate .............................. 10 30 Total selected assets . . . 25 35 48 18 8 1 33 4 24 17 32 24 18 32 Percentage change in real assets Vehicles, machinery, and eq u ip m en t............................ 37 11 11 -3 Stored c r o p s ............................ Demand deposits and currency .............................. Time deposits and savings ” 3 4 7 -2 —10 -1 6 -1 2 Real estate ............................... —1 4 -6 2 —3 2 2 2 8 0 2 6 Total selected assets . . . 18 2 40 -3 Average annual percentage change in current value Vehicles, machinery, and e q u ip m en t............................ Stored c r o p s ............................ Demand deposits and currency .............................. Time deposits and savings Real estate .............................. Total selected assets . . . 8.8 -2 .7 4.8 3.6 6.3 -4 .3 2.8 -.9 3.6 -.3 -2 .2 -1 .0 1.6 2.0 5.5 .3 5.8 .8 4.3 3.3 5.9 4.5 4.5 3.4 5.9 6.3 8.5 3.4 Average annual percentage change in real assets Vehicles, machinery, and eq u ip m en t............................ 6.5 2.2 2.1 -.7 ” .6 .8 1.3 —.4 3.4 .5 7.1 -2 .2 -3 .5 -2 .4 —.6 Real estate .............................. -.3 .8 - 1 .3 .4 —.7 .4 .4 .3 Total selected assets . . . 1.4 0 .3 1.3 Stored c r o p s ............................ Demand deposits and currency ............................... Time deposits and savings * Data shown for 1965-69 are actual values for 1965-68 multiplied by 1.25 to facilitate comparison with previous 5-year periods. N o t e . —-Users of the data on real assets are referred to p. 118-19 for a discussion of a probable bias in these estimates. S o u r c e . —Table 1 and additional data from U SD A . simply because these flows over time differ for varying mixes of real and price increases in stock. Efforts to allocate changes in stock values to real and price components are 118 greatly handicapped, however, by the fact that capital goods change over time as technology advances. The tractors and land of today are not the same products as in 1950, and so one cannot be sure how much of the increase in their current price rep resents price inflation and how much is due to gains in quality or productivity of the assets. As such gains often occur in subtle ways that defy measurement, the USDA estimates of real assets may understate the progress that has occurred, and the price increase may therefore be overstated. Nevertheless, it appears that real estate and machinery prices rose rather steadily during 1950-68, with very significant im pact on total asset values. On the other hand, prices of livestock moved in a direc tion opposite to livestock numbers, so that when the real livestock inventory increased, its current value tended to decrease, as in 1950-54 and 1960-64. The rate at which physical additions were made to stocks of machinery, livestock, and crops varied substantially from one period to the next. Machinery stocks were easily the most volatile component, with especially rapid increases in the early 1950’s and again in 1963-67. Projected capita! stocks in 1980. Three widely circulated projections of 1980 stocks constitute the point of departure for esti mation of capital flows in the intervening period. The stocks projected for 1980 in current (1980) dollars are summarized in Table 3. To facilitate comparison with cur rent values, Model NC (no change) shows the value of stocks (and later also of flows) if neither price nor real changes occurred after January 1, 1969. The first set of projected stocks, Model HT, is based primarily on projections for 1960-79 published by Heady and Tweeten in 1963 after extensive econometric anal ysis of the determinants of demand for vari ous farm capital goods.1 The HeadyTweeten projections were made in real terms only, but the machinery, financial assets, and real estate values shown in Table 3 are altered to reflect moderate price ad vances. For real estate, the current-dollar projection employs a Heady-Tweeten price equation that is relatively successful in ex plaining the postwar course of farmland values. The second projection, Model B, is based on current-dollar projections of 1980 stocks published by Brake in 1966, with the real 1 E arl O. Heady and L uther G. Tweeten, R esource D em and and Structure of the A gricultu ral Industry (Ames, Iowa: Iowa State University Press, 1963). TABLE 3 ALTERNATIVE PRO JECTIO NS OF SELEC T ED FARM A S SE T S Asset Model NC Model HT Model B Model HM 32.6 20.1 10.5 15.3 202.6 281.1 Vehicles, machinery, and equipment Livestock ........................................................... Stored crops .................................................... Deposits, currency, and savings bonds . . Real estate ........................................................ Total selected assets ............................ 12 7 4 5 72 100 40.5 21.4 10.0 25.2 392.9 490.1 36.4 23.2 11.4 15.7 272.2 358.9 64.2 21.9 10.0 25.2 288.4 409.7 (per cent of total) 8 4 2 5 80 100 10 6 3 4 76 100 Model HT Model B Model HM Change during 1970’s (billions of dollars) Amount in 1980 (billions of dollars) Vehicles, machinery, and equipment Livestock .......................................................... Stored crops .................................................... Deposits, currency, and savings bonds . . Real estate ........................................................ Total selected assets ............................ M odel NC 7.3 1.2 -.5 9.2 177.8 195.0 3.5 2.8 .8 .4 64.1 71.5 29.6 1.6 —.5 9.2 79.2 119.1 (average annual percentage change) 16 5 2 6 70 100 2.0 .6 —.4 4.6 6.2 5.2 1.0 1.3 .7 .2 2.7 2.2 6.4 .8 —.4 4.6 3.3 3.5 CAPITAL AND CREDIT REQUIREMENTS OF AGRICULTURE estate estimate as updated by Brake in 1968.2 The final projection, Model HM, is based primarily on one of several projections of real stocks of machinery and livestock and of price changes of real estate published by Heady and Mayer in 1967, in a project executed for the National Advisory Com mission on Food and Fiber.3 The estimates used here assumed that land retirement pro grams of the present “feed-grain” type are continued for wheat and feed grains and are also applied to cotton production, and that exports increase in accordance with 195065 trends. As with Model HT, the machin ery and real estate projections were modified to reflect trends in machinery prices and in the general price level, respectively. In ad dition, because Heady and Mayer did not project values of stored crops or of financial assets, these items were projected at the same levels as in Model HT. The three projections agree in one impor tant respect: that the total value of farm assets will increase considerably during the next decade. Beyond this, there are differ ences that appear likely to have considerable impact on capital and credit demands: (1 ) the projected total increase in value varies from $71.5 billion under Model B to $195.0 billion under Model HT— an average dif ference of $12 billion per year over the de cade, and (2 ) growth projected for major asset components differs greatly. Model HT projects a relatively rapid rise in real estate values, but only moderate gains in the ma2 John R. Brake, “Im pact of Structural Changes on Capital and Credit N eeds,” Journal of Farm Economics (Dec. 1966), pp. 1536-45. Also “Dim en sions of the C redit D oor,” unpublished speech at Blacksburg, Va., Aug. 5, 1968. 3 E arl O. Heady and Leo V. Mayer, Food Needs and U.S. Agriculture in 1980, Technical Papers, vol. 1, U.S. N ational Advisory Commission on Food and Fiber (W ashington: Govt. Printing Office, Aug. 1967). 119 chine stock. The reverse is true of Model HM, whereas Model B anticipates relatively moderate growth in all components but with rising real estate values dominant. In the next subsection, the bases for these stock projections are briefly noted, and the capital flow requirement that appears im plied by each model is calculated. The framework for the analysis both here and in the next section draws heavily on the pio neering capital study of Tostlebe, which is also the source of many insights into long term trends.4 A comprehensive and more recent capital and credit study by Johnson was also very useful.5 C a p ita l re q u ire m e n ts b y a sse t g r o u p , 1950-79 Farm capital flows and credit demands arise in three important ways. First, they origi nate from expenditures to maintain or ex pand the capital plant. In this category one finds spending for (1 ) replacements and additions to the stock of vehicles, machin ery, equipment, buildings, and land im provements; (2 ) additions to inventories of livestock and of crops stored for feed and seed; and (3 ) additions to financial working balances. Second, capital flows and credit demands arise when the capital plant— especially real estate— is transferred from one owner to the next by means other than gift or inheritance. Estimates for 1950-68 of the various capital flow requirements of these two types are summarized in Table 4. Third, seasonal credit demands occur when additional working capital is needed to fi nance seasonal production processes for 4 Alvin S. Tostlebe, Capital in Agriculture: Its Formation and Financing Since 1870 (Princeton: Princeton University Press, 1957). 5 D. Gale Johnson, “A gricultural Credit, Capital and C redit Policy in the U nited States,” Federal Credit Programs, Commission on M oney and Credit (Englewood Cliffs, N.J.: Prentice-Hall, Inc., 1963), pp. 355-423. 120 TABLE 4 CAPITAL FLOWS, 1950-69 In billions of dollars 5-year total* Type of flow Gross capital expenditures: Vehicles, machinery, and equipment Buildings and land improvements . . . To increase: Livestock inventory ................................. Stored crop inventory ............................ Demand deposits and currency ......... Time deposits and savings bonds Required by real estate p u rc h a se s............... Total capital flow 1950-54 15.4 7.7 2.4 .4 1955-59 14.0 6.9 .5 .8 1.3 - .2 1.7 11.0 13.5 37.4 35.1 39.4 .1 * Data shown for 1965-69 are estimates for 1965-68 multi plied by 1.25 to facilitate comparison with previous 5-year periods. S o u r c e .— Machinery and building expenditures from F arm In com e S itu a tio n , U SD A , July 1969, p. 60; increase in live- which the level of cash assets normally maintained does not fully provide. These seasonal demands are discussed in Section VI. Vehicles, machinery, and equipment. Im proved vehicles, machinery, and equipment (all grouped under “machinery”) consti tute a readily visible example of the impact of technological change on the capital goods of agriculture. And in addition to all the new equipment purchased for production on farms (with which this study is con cerned), there has been considerable non farm investment in such allied industries as hatcheries and feed mills, which perform work that in earlier years had been done on farms. Expenditures for machinery now consti tute a significant capital requirement, over two-fifths of the total flow. Analytically, these expenditures are of two types: to re place stock that has worn out or has be come obsolete, and to expand the total stock in order to increase output or reduce labor requirements. Expenditures arising from either need are affected by the course of machinery prices. To maintain the machine stock at a given real level requires an annual expenditure equal to about 14 per cent of the value of the stock, according to recent depreciation 16.0 6.4 - .3 .3 16.0 -.1 - .7 1960-64 Annual average 1965-69 1950-54 23.6 6.4 3.1 1.5 .2 .5 1955-59 2.8 1.4 1960-64 1965-69 3.2 1.3 4.7 1.3 —.1 .1 1.4 .5 1.4 20.7 2.2 2.7 3.2 .3 .1 .3 4.2 54.2 7.5 7.0 7.9 1 0 .8 .1 .2 -.1 .1 .1 stock and crop inventories are unpublished data from U SD A (livestock and crop total is published in F arm In c o m e S itu a tion, July 1969, p. 53); increase in financial assets from T he B alan ce S h eet o f A g ricu ltu re, 1968, U SD A , Jan. 1969, p. 10; capital flows required by real estate purchases are estimated by Emanuel Melichar. allowances estimated by the USDA .6 With the stock valued at $32.6 billion in 1969, annual replacement requirements are thus around $4.6 billion. Machinery prices, however, appear likely to increase over time. Prices set by manu facturers are likely to reflect the general upward course of unit costs in the capital goods sector of the nonfarm economy. The implicit price deflator for the total farm machine stock rose at annual rates of 4.4 per cent in 1955-59, 2.0 per cent in 1 96064, and 2.8 per cent in 1965-68. If, in view of this record, one projects annual machin ery price increases averaging 2.5 per cent in 1969-79 and no real growth, the value of the stock would still rise to $42.8 billion by 1980. Annual replacement requirements would by then average $6.1 billion. Any physical additions to the total stock constitute a capital flow requirement super imposed on the replacement expenditures. In this century, periods of rapid real expan sion have alternated with extended periods of little or no growth. A spending boom that nearly tripled the real stock between 1945 and 1954 was succeeded by 10 years of little growth or of small declines. Re6 USDA, Farm Incom e Situation Govt. Printing Office, July 1969), p. 61. (W ashington: 121 CAPITAL AND CREDIT REQUIREMENTS OF AGRICULTURE TABLE 5 PAST A N D PROJECTED RATES OF C H A N G E IN M A C H IN E ST O C K S A N D PRICES, 1946-79 Average annual rate of change, per cent Period 1946-48 ............................. 1949-51 ............................ 1952-55 ............................ 1956-60 ............................ 1961-62 ............................ 1963-64 ............................ 1965-67 ............................ 1968 ................................... 1970-79: Model HT ................. B ...................... HM ............... Total Real Price 23.2 18.2 3.7 2.5 2.1 6.0 6.7 4.5 18.4 14.5 2.7 - 1 .3 -.8 4.0 3.8 1.8 4.0 3.2 1.0 3.9 2.9 1.9 2.8 2.8 2.0 .9 6.2 -.5 .3 3.6 2.5 .6 2.5 N o t e . — Users of the data on prices and real stocks are referred to pp. 118-19 for a discussion of a probable bias in these estimates. S o u r c e . — Past annual rates of change in total stock were computed from data in The B alan ce S h eet o f A g ricu ltu re, 1968, U SD A , Jan. 1969, pp. 26 and 27. Estimates of past real stocks were supplied by the U SD A . Price changes shown are for the implicit price deflator for the total machine stock, as computed from these two series. newed rapid expansion beginning in 1963 lifted real stocks by another 23 per cent be fore 1969. Some projections of machinery require ments emphasize the spur from continued technical innovation, combined with desires and incentives (higher wage rates) to re duce labor requirements. Such projections, as in Model HM, indicate substantial real increases in future machine stocks. Other analysts have been more impressed with the substantial upgrading of stocks that can occur in the course of the large replace ment expenditures. For instance, structural analysis by Heady and Tweeten suggested “a mature agricultural economy in terms of machinery. A large amount of new machinery will continue to be purchased not only to replace worn-out machines but also to substitute for machines that are inade quate for large holdings. This will offer siz able opportunities for machinery to replace labor, despite the rather small increment in machinery assets.”7 This view is represented in Models HT and B. The historical record since World War II taxes analysts seeking to determine the more appropriate view, as the growth rates shown in Table 5 demonstrate. Heady and Mayer analyzed the record of 1949-64 and found a strong upward trend over these years. The large expenditures shown for Model HM in Table 6 are based mainly on assumed con tinuation of this trend. In 1975-79, annual expenditures would average $9.6 billion. But Heady and Tweeten, writing in the early 1960’s, thought the relative stability of 1952-60 to be more representative of the future, and thus projected little real expan sion. Brake, although writing in 1966 after expenditures had again accelerated, also expected relatively slow future growth. Models HT and B both project average annual expenditures of about $5 billion in 1975-79, or little higher than those at the peak of the recent boom. Buildings and land improvements. Construc tion of farm dwellings, service buildings, and various other structures and land im provements such as fences, wells, ponds, terraces, and tile lines comprises a sub stantial continuing capital expenditure, cur rently about 12 per cent of total capital ? Heady and Tweeten, op. cit., p. 492. TABLE 6 ALTERNATIVE PROJECTED EX P E N D IT U R E S FOR M A C H IN E R Y In billions of dollars 5-year total Projection 195054 195559 196064 196569 14.0 16.0 23.6* Actual .................................................. 15.4 Model NC ......................................... HT ........................................................................................................ B ............................................................................................................. HM ........................................................................................................ * Expenditures for 1965-68 multiplied by 1.25. Annual average 197074 197579 195054 195559 196064 196569 3.1 2.8 3.2 4.7 23.1 24.3 24.4 35.6 23. i 26.8 25.5 48.2 ...................................................... ....................................................... ....................................................... 197074 4^6 4.9 4.9 7.1 197579 4.6 5.4 5.1 9.6 122 flow. Tn some regions, construction of items such as irrigation systems and commercial feed lots has been expanding. Nationally, however, expenditures have been declining absolutely as well as relative to other capi tal uses. The downward drift in construction fol lowed large gains in the years immediately after World War II. Expenditures for farm operators’ dwellings reached a peak of $702 million in 1948 but by 1968 were reduced to $493 million. Construction of other buildings and land improvements topped at $949 million in 1952 and was down to $812 million in 1968. One factor reducing new farm construc tion is the rapidly declining number of farm units and families. From 1950 to 1968, the number of farms fell by 46 per cent, or by about 2.6 million units. Each farmstead that was abandoned or became a rural residence for a nonfarm family tended to reduce fu ture farm building needs. In addition, expenditures for new service buildings have been negatively affected by various technological developments. Greater efficiency in livestock production— more milk per cow, faster growth of hogs and broilers— enabled farmers to increase out put without proportional increases in animal housing space. Greater use of purchased mixed feeds and virtual elimination of horses and mules tended to reduce farm feed storage requirements. Less costly types of buildings, such as those employing poletype construction, were increasingly adopted. Projected construction expenditures used in Models HT, B, and HM are based on a recent study by Scott and Heady.8 They 8 John T. Scott, Jr., and E arl O. Heady, A ggregate In vestm ent D em an d fo r Farm Buildings: A N ation al , R egional and State Tim e-Series A nalysis, Research Bulletin 545 (Ames, Iow a: A gricultural and H ome Economics Experim ent Station, Iow a State University, July 1966), pp. 704-36. project an average annual real decrease of 0.9 per cent and assume that prices of build ing materials will continue to rise at the 2 per cent annual average experienced from 1947 to 1963. Thus, yearly current-dollar spending would average $1.4 billion during 1970 to 1974, and $1.5 billion in 1975 to 1979. Livestock inventory. Additions to the quan tity of livestock on farms entail a capital flow equal to the value of the physical quan tities added. There is general agreement that expanding domestic population and rising per capita income will continue to raise aggregate demand for livestock pro ducts, and that the greater output will re quire larger livestock inventories on farms. However, inventories are likely to rise more slowly than output. As Tostlebe noted after his study of 1890-1950, “the most signifi cant technological advances in agriculture . . . have quite consistently been connected with the production of livestock and of livestock products. . . . Improvements in the breeds of livestock and in livestock feed and management have been sufficient to permit animal products to become increas ingly important in the farm-product mix, while the investment in productive live stock per dollar of total farm product de clined greatly.”9 This effect remains impor tant. Excluding horses and mules, the num ber of animal units of breeding livestock on farms in 1967 was the same as in 1919 and somewhat below levels of the 1940’s and 1950’s. However, production per breeding unit was 116 per cent larger than in 1919, 38 per cent above that of 1950, and up 13 per cent since I960.10 The larger numbers of feeder livestock and poultry have since 1950 required capital flows that varied 9 Tostlebe, op cit., p. 126. 10 USDA, Changes in Farm Production and Effi ciency (W ashington: Govt. Printing Office) June 1955, p. 23; June 1968, p. 10. CAPITAL AND CREDIT REQUIREMENTS OF AGRICULTURE greatly from year to year, but averaged only $228 million annually (Table 4 ). As with the machinery projections, anal ysts again apparently differ as to relative future impact on inventories of the divergent influences of greater consumer demand and increased production efficiency. Heady and Tweeten projected an average annual gain of only 0.75 per cent in the real livestock inventory, which would require yearly ex penditures of about $120 million during the next decade. But Heady and Mayer specifically assumed no further improve ment in the inventory/output ratio and thus projected an average real gain of 2.8 per cent annually between 1965 and 1979. Even if livestock prices receded to 1965 levels by 1980, this growth would require expenditures of more than $700 million an nually during the 1970’s. Brake also pro jected similar real growth, with 1980 prices 14 per cent over those of 1965. Annual ex penditures of $600 million would be re quired to achieve this projection. Although the projections vary consider ably, even a relatively faulty livestock fore cast does not introduce a large relative error in projected total capital flows. Pro jected livestock expenditures have the great est relative importance in Model B, but even there they account for only 5 per cent of total capital flows anticipated. Inventory of stored crops. The value of net physical additions to farmers’ holdings of stored crops constitutes a volatile but minor capital flow that averaged $144 million an nually in 1950-68. Diverse influences ap pear to be operating on the long-term trend. Larger livestock production leads to growth in feed inventories, but the rise is moderated by upward trends in the animal output ob tained from a given quantity of feed and in the proportion of total feed purchased from commercial mixers. To the extent that feed inventories are held by feed companies 123 and dealers, the associated capital require ment has been transferred to the nonfarm economy. Each capital model projects a continued small upward trend in real stocks. How ever, because 1969 inventories represent a considerable bulge over the long-term trend — one of several sizable fluctuations ex hibited over the postwar period— these pro jections translate into a small amount of dis investment between 1969 and 1980. Financial assets. Farmers must hold money balances to carry on their business trans actions, primarily involving payment for current operating and family living ex penses. Historically, these balances have risen both in absolute terms and as a pro portion of total assets, reflecting the growth of cash operating expenses as each farm unit has become less self-sufficient and more dependent on purchases from other farms or from the nonfarm sector. During 1950-65, however, growth in money holdings was at least temporarily interrupted as farmers reduced their demand deposits and currency by $1.1 billion, or 16 per cent. The upward trend in operating expenses continued during these years, but offsetting influences on the money stocks — such as the decline in the number of farms and in the farm population— were ap parently more powerful. In addition, an up ward movement in interest rates put an increasing opportunity cost on cash bal ances. Ready availability of seasonal pro duction credit may also have enabled farmers to reduce the relative amount of cash assets held on January 1, the day on which these stocks are estimated for the Bal ance Sheet. In response to higher interest rates paid on time and savings deposits and perhaps also as a result of improved farm financial management, farmers may have been more likely to hold seasonally-idle working capi 124 tal in time and savings rather than demand deposits. Thus the change in these assets, which tended to increase during the post war period, has been included among capi tal requirements. At the same time, farmers have reduced their holdings of U.S. savings bonds, which have also been included among the financial assets here enumerated. Projection of financial balances must con tend with these diverse influences. Heady and Tweeten projected a 23 per cent total real gain in cash for operating expenses be tween 1960 and 1980. To achieve this real growth as prices paid by farmers rise by an assumed 2 per cent a year, farmers would have to add $917 million per year to their holdings of the financial assets listed. This estimate is used in Models HT and HM. But Brake projected a slow rise in current dollars; farmers would have to add only $36 million annually to financial assets to fulfill his projection, which is used in Model B. Real estate purchases. Most farm real estate is owned by individuals and is transferred from one owner to the next by sale rather than inheritance. Of the total number of transfers in the year ending March 1, 1969, for example, only 13 per cent were inheri tance or gift transfers. Voluntary sales by retiring or retired farmers and others and by executors of estates averaged $5.5 bil lion annually over the 4 years ending on March 1, 1969.11 Thus, annual purchases of land are somewhat larger than expenditures for vehicles and machinery. Capital flows required by land transfers are lower than the value of sales, however. The total capital flow required equals the money removed from the agricultural pro duction sector by sellers who are retiring or retired farmers, nonfarmer heirs, or nonn USDA, Farm R eal Estate M arket D evelopm en ts (W ashington: Govt. Printing Office) Aug. 1969, p. 22; M ar. 1969, p. 11; Apr. 1968, p. 14; June 1967, p. 13. farmer investors who are withdrawing from farmland ownership. To calculate the capi tal flow, therefore, the value of sales must be adjusted for the amount of outstanding debt on the property— which is either as sumed by the purchaser or is repaid as a result of the sale— and also for the proceeds of land sales that are used to buy other farm land. There is little data on which to estimate these adjustments and so derive required capital flows from value of sales. One in dication of the amount of outstanding debt is provided by a 1967 survey showing that assumption of outstanding property mort gages accounted for 9 per cent of credit in volved in land transfers, which puts assump tions at about 5 per cent of transfer value. A 1964 survey indicated that about 10 per cent of total voluntary sales were made by farmers who continued in farming after the sale, and who therefore may have bought other tracts with the proceeds.12 No data seem to be available on debt repayments or on the subsequent activities of nonfarmer sellers. For estimates of capital flows, land sales were adjusted downward by 25 per cent to obtain the capital flow required. In 1965— 68, capital flows associated with real estate transfers were therefore estimated to aver age $4.2 billion per year, or 38 per cent of total farm capital flows. For the years prior to 1965, estimates are made still more difficult by lack of data on the value of real estate sales. For these years, only transfer rates and total real estate values are provided by the USDA. Since in 1965-68 the value of sales aver aged 78 per cent of the figure obtained by multiplying the transfer rate by total value, this relationship was used to estimate capi tal flows required in 1950-64 (Table 7 ). 12 ib id ., Dec. 1968, p. 23; Aug. 1965, p. 31. CAPITAL AND CREDIT REQUIREMENTS OF AGRICULTURE 125 TABLE 7 ALTERNATIVE PROJECTED CAPITAL FLOWS R EQ U IR E D BY REAL ESTATE P U R C H A SE S In billions of dollars 5-year total Projection Estimated actual ............................ Model NC ....................................... HT ....................................... B ........................................... HM ....................................... 195054 195559 196064 196569 11.0 13.5 16.0 20.7* Annual average 197074 197579 22.7 28.1 24.9 25.4 22.6 37.3 28.0 29.3 195054 195559 196064 196569 2.2 2.7 3.2 4.2 197074 197579 4.5 5.6 5.0 5.1 4.5 7.5 5.6 5.9 * Estimated flows in 1965-68 multiplied by 1.25. The estimates indicate a steady upward trend that about doubled the required flow between 1950 and 1968, as the effect of higher land prices easily overwhelmed the effect of lower transfer rates. The same relationships were used in pro jecting future capital flows. With a con tinued small decline in the transfer rate, re quired annual capital flows would average about 2.2 per cent of any projected value of the real estate stock. Thus if the value of land and buildings were to stabilize at the 1969 level, as in Model NC, the transfer capital required would be $4.5 billion per year. In the other models, the capital flows depend on the projected course of real estate prices. An econometric study by Twee ten and Nelson that attempted to measure the rela tive strength of pressures on farmland prices in 1950-63 ascribed 52 per cent to farm enlargement (of which an unspecified portion was thought due to Government programs), 20 per cent to demand for nonfarm uses, 17 per cent to the expectation of further capital gains, and most of the remainder to reduction in quantity of land.13 Since the land price index was deflated by the wholesale price index prior to analysis, participation by farmland in a general price 13 Luther G. Tweeten and Ted R. Nelson, Sources and R epercussions of Changing U.S. Farm R eal Estate Values, Technical Bulletin T-120 (Stillwater, Okla.: O klahom a State University A gricultural Ex perim ent Station, Apr. 1966), p. 18. uptrend was also assumed. Because of the many alternative ways in which a land-price model could be specified and estimated, this one study is not definitive. But perhaps it indicates the principal forces bearing on land prices and exerting through them a major influence on capital and credit re quirements. In this view, the basic factor behind increases in land prices is technological change. First, innovations have increased the productivity of land. Higher crop yields resulting from new technology and better management have tended, ceteris paribus, to lower unit production costs and increase net returns. Second, other new technologies — principally larger tractors and machines — have permitted a farmer to operate a larger land area and thereby also to lower unit overhead costs.14 This incentive to en large farm units has created an active de mand for land. Competitive bidding among the more successful farmers— those able to achieve above-average net returns from each added tract— has led to increased prices; in effect, the higher net returns have been capitalized into land prices.15 Also, as this experience prevails over many years, 14 Ibid., pp. 45-47. 15 A lbert A. M ontgom ery and Joseph R. Tarbet, “Land Returns and Real Estate Values,” A gricultu ral Econom ics Research, USDA (W ashington: Govt. Printing Office, Jan. 1968), pp. 5-16. W illiam H. Scofield, “Land Prices and F arm Earnings,” Farm R eal Estate M arket D evelopm en ts, U SD A (W ashing ton: Govt. Printing Office, Oct. 1964), pp. 39^42. 126 the upward course of land prices is prob ably further reinforced as buyers discount expected future advances in technology and therefore in net returns— or, what is equiv alent if less sophisticated, they discount capital gains from an expected future up ward trend in land prices.16 Much of the same new technology that reduced unit costs, however, also tended to increase total farm output.17 Output gains could occur in two ways: through improve ment in inputs and farming practices and as farm consolidation places more of the total resources into the hands of the more effi cient and specialized operators. Government output control programs kept the potential output increase from being fully achieved, but the gain has been sufficiently large rela tive to the slower expansion of demand to exert a depressing influence on output prices. The latter effect tended to offset the favorable impact of unit cost reductions on net returns and would have been more pronounced in the absence of the Govern ment programs.18 In these circumstances the commodity programs, by restricting total production and either maintaining output prices or sup plementing net incomes, have allowed a higher portion of the benefits of cost-reducing technology to accrue to farmers rather than to consumers. To the extent that Gov ernment programs have thus preserved the technologically induced gains in net returns that have in turn been capitalized into land prices, such programs may contribute to the rise in land prices.19 The effect has been particularly obvious in cases where benefits 16 Tweeten and Nelson, op. c i t pp. 19-22. 17 Gene L. Swackhamer, “Agriculture and Tech nology,” M on th ly R eview , Federal Reserve Bank of Kansas City (M ay-June 1967), pp. 5 and 6. is Tweeten and Nelson, op. cit., pp. 23-25. 19 Ibid., pp. 15-18 and 47. of an effective program have been tied to specific parcels of land; for instance, land with a tobacco allotment has been valued at several times the price of similar land that lacked an allotment.20 Insofar as the future course of real estate values depends on technological advances and the extent to which these foster further farm enlargement, their direction in the relatively near future does not seem in doubt. Numerous studies continue to indicate that the optimum sizes of family farms— given known technology— are far above present averages. It is reasonable that price projec tions to 1980, as made in the three models, be based mainly on the upward thrust from this source, but with realization that prices can be materially affected within that time by changes in the nature and extent of Gov ernment programs and in export levels, gen eral price trends, and the degree to which expected land price increases are discounted. Over a longer period, changes in the rate and nature of technological advances— par ticularly in the extent to which they would continue to foster enlargement of the land area of individual farms— become a greater source of uncertainty.21 Changes in popula tion growth and in the nature of urban ap petites for residential and recreational lands also become larger considerations. Of the projected real estate values, that of Model HT represents most closely an extension of the past historical relationship between land prices and farm enlargement. Prices are projected to rise by 6.2 per cent 20 William H. Scofield, “Land Returns and Farm Income,” Farm R eal E state M arket D evelopm en ts, USDA (Washington: Govt. Printing Office, Aug. 1965), p. 51. 21 Bruce B. Johnson, “An Active Land Market in Perspective,” F arm R eal Estate M arket D evelopm en ts, USDA (Washington: Govt. Printing Office, Dec. 1968), pp. 34 and 35. 127 C A PITA L AND C R E D IT R E Q U IR E M E N T S OF A G R IC U LT U R E annually, causing required transfer capital flow to rise rapidly to an annual average of $7.5 billion in 1975-79 (Table 7 ). In Model HM, on the other hand, an average yearly price increase of 3.3 per cent is de rived by assuming that land values will re flect projected increases in the economic rent to cropland as well as general price in flation averaging 2 per cent yearly. Annual capital flows required by this model attain an average level of only $5.9 billion in 1975-79. Model B reflects Brake’s assump tion that land prices will rise by an average of 3 per cent yearly, with implied capital flows therefore similar to those of Model HM. Total capital flows, 1 9 5 0 -7 9 Total capital flows— past, present, and pro jected— are summarized in Table 8. In the 1950’s, total flows averaged $7.3 billion annually. Real estate purchases rose throughout the decade, but in the second half machinery expenditures and additions to livestock inventory slackened enough to stabilize the total. In 1960-64, additions to machinery and livestock holdings were re sumed and together with increasing real estate purchases raised total flows to an average of $7.9 billion per year. Then in 1965-68, a sharp increase in machinery expenditures and a steady rise in land prices combined to raise capital flows to an annual average of $10.8 billion. If the capital stock were to be stabilized at the level existing at the beginning of 1969, both in real terms and in current dol lars as Model NC assumes, future capital flows would average $11.3 billion per year. About two-fifths of this sum would arise from real estate transfers, a similar share from expenditures required to maintain the stock of vehicles and machinery, and the remaining one-fifth from maintenance of the stock of buildings and land improve ments. It is evident, therefore, that any further increases in prices of capital goods and any further additions to the physical plant would raise total capital flows above the present level. Each of the other three models en vision some price and real increases during the next decade and therefore project higher capital requirements. They differ only in the magnitude of the increases in requirements expected. Model B, which projects moderate land price increases and small gains in machinery expenditures, envisions only a moderate gain in the required capital flow. By the second half of the next decade, annual flows would average $12.8 billion. Real estate transfers would rise somewhat in relative importance, from 38 per cent of total flow in the period 1965-68 to 44 per cent a decade later. Model HT projects only moderate gains in machinery expenditures and very small additions to livestock inventories, but strong increases in farmland prices. By 1975-79, TABLE 8 ALTERNATIVE PROJECTED TOTAL CAPITAL FLOWS In billions of dollars 5-year total Projection 195054 195559 196064 196569 Estimated actual ............................... 37.4 39.4 54.2* 35.1 Model NC ......................................... HT ........................................................................................................ B ............................................................................................................. HM ........................................................................................................ * Estimated flows for 1965-68 multiplied by 1.25. Annual average 197074 197579 195054 195559 196064 196569 7.5 7.0 7.9 10.8 56.2 63.7 59.0 75.9 56.8 76.8 64.0 94.4 ....................................................... ....................................................... ...................................................... 197074 197579 11.2 12.7 11.8 15.2 11.4 15.4 12.8 18.9 128 required capital flows would consequently average $15.4 billion per year, with real estate transfers contributing 49 per cent of this total. Model HM, on the other hand, projects moderate increases in prices of land, but very large real additions to machinery stocks. Because of the latter, the capital flows projected are the largest of the three models, averaging $18.9 billion annually during the late 1970’s. Of this total, 52 per cent would consist of machinery expendi tures and only 31 per cent would stem from real estate purchases. The projections differ considerably. But to emphasize the differences, and the un knowns that they reflect, would be to lose the principal message of the estimates. Re call that over the past 10 years annual capi tal flows rose by $4 billion; in relative terms, by 54 per cent. The projections for the next decade show annual requirements rising by $2 to $8 billion; in relative terms, by 19 to 75 per cent. The message is clear: capital demands will rise further from the high level of the last few years; in number of ad ditional dollars, the gain could easily ex ceed that of the last 10 years; relative to the new high level of current requirements, the additional demands may represent a somewhat slower advance, but under some conditions might equal or exceed the recent sharp rise. The unanimous projection of a signifi cant further increase in capital flows ap pears well grounded. The two primary sources of future capital flows— machinery purchases and farm enlargement— have a common root in technological advance. The fund of technological knowledge now avail able but not yet applied and the high like lihood of additional discoveries indicate that growth in total investment and investment per farm will continue for some time.22 The National Advisory Commission on Food and Fiber recently summarized these expec tations as follows:23 T here is little doubt th at farm ing will continue to use m ore capital in the future. First, science and technology are continually advancing n ot only in application to farm ing but throughout the economy. Second, reflecting increased productivity, the relative cost of capital keeps declining. Capital becomes continually cheaper, com pared w ith labor and land, so farm ers will continue to use m ore capital. These changes n ot only m ake it possible for the individual farm er to increase his volum e of opera tions— they m ake it necessary fo r him to do so. He m ust expand his investm ent and then spread costs over m ore units of product to rem ain com petitive. Thus, even though agriculture is already one of the more capital-intensive sectors of the American economy, a further rise in the capital/output ratio in current prices seems certain. The ratio of the value of farm productive assets to the gross national prod uct produced in agriculture has been esti mated as at least 6:1 in the 1950’s com pared with a ratio of about 1.5:1 in the nonfarm economy.24 By 1964-66, the ratio in agriculture averaged 8:1. These data hint that the annual capital demands of farming place a relatively severe and rising strain on the income flows from which they are either initially or ultimately financed. These relationships are examined next— first as they have evolved since 1950, and then as they might develop under each of the altern ative capital projections. 22 Federal Reserve Bank of Kansas City, “F in an cial Requirem ents of A griculture,” M on th ly R eview (Sept.-O ct. 1964), pp. 5-7. 23 U.S. N ational Advisory Commission on F ood and Fiber, F ood and F iber for the Future (W ash ington: Govt. Printing Office, July 1967), p. 240. 24 D. G ale Johnson, op. cit., p. 355. 129 C A PITAL AN D C R E D IT R E Q U IR E M E N T S OF A G R IC U LT U R E IV. CREDIT REQUIREMENTS, 1950-79 Given the prospect of substantial capital flow requirements, this section projects the share that will be financed from cash flow— depreciation allowances and net income— as opposed to the share financed by ex panded use of credit. Thus, one preliminary task is to project depreciation and net in come, and the other is to project the share of these amounts that may be allocated to ward meeting capital needs. After examina tion of the postwar history of these series, such projections of internal financing are made here. They are then compared with the projected capital flows to secure esti mates of future credit demands and farm debt expansion. investment is expected to pay its own way from, farm income and land price apprecia tion. Of total annual cash flow averaging $32.5 billion in 1965-68, net farm income rep resented 50 per cent, capital consumption allowances 17 per cent, and nonfarm in come 33 per cent (Table 9). Although the principal component is still net farm in come, its relative importance has been de clining. Fifteen years earlier it had con tributed 61 per cent, while capital con sumption allowances had represented 13 and nonfarm income only 26 per cent. TABLE 9 FINANCING OF CAPITAL FLOWS, 1950-69 How have capital requirem ents been financed? In billions of dollars unless otherwise indicated Sources of— 1950-54 1955-59 1960-64 1965-69 5-year total* It is analytically useful to view capital flows required by the farm production sec tor— including nonfarm landlords— as being met either (1) from a cash flow consisting of income remaining after operating ex penses are paid or (2) by borrowing.25 Financing from cash flow. Cash flow is esti mated as the sum of net farm income of operators and landlords, plus the capital consumption allowances that were included in estimated production expenses (estimated depreciation of buildings, land improve ments, vehicles, machinery, and equipment, as well as accidental damage to these capital goods), plus nonfarm income of the farm population. Nonfarm income is included in cash flow because most farm families ap parently continue to pool farm and non farm income prior to meeting living and capi tal investment needs. Nonfarm income of farm landlords is not included because such 25 Tostlebe, op. cit., p. 132. Capital financing Increase in debt ........... From cash flow ........... Total capital flow . . . Cash flow Capital consumption allowances ............. Net farm income . . . . Nonfarm income ........... Total cash flow . . . . 4.7 32.7 8.2 27.0 12.4 27.0 20.0 34.3 37.4 35.1 39.4 54.2 16.2 73.9 31.9 19.6 63.3 33.1 22.0 68.3 40.7 27.7 80.9 54.0 116.0 131.0 162.5 122.0 Annual average Capital financing Increase in debt ........... From cash flow ........... .9 6.5 1.6 5.4 2.5 5.4 4.0 6.9 Total capital flow .. Cash flow Capital consumption allowances ............. Net farm income . . . . Nonfarm income ......... 7.5 7.0 7.9 10.8 3.2 14.8 6.4 3.9 12.7 6.6 4.4 13.7 8.1 5.5 16.2 10.8 Total cash flow . . . . 24.4 23.2 26.2 32.5 Per cent Analytical ratios Capital flow /cash flow Proportion of cash flow used for capital Average annual growth rate during period: Selected assets (Table 2) ............... Debt ............................ Debt/assets, end of period . . . . 31 30 30 33 27 23 21 21 4.5 7.6 4.5 8.9 3.4 8.8 5.9 9.6 10.2 12.5 16.1 18.5f * Data shown for 1965-69 are estimates for 1965-68 multi plied by 1.25 to facilitate comparison with previous 5-year periods. f As of Jan. 1, 1968. S o u r c e . —Capital flows from Table 4, debt from Table 13, assets from Table 2, and cash flow components from F arm In com e S itu a tio n , U SD A , July 1969, pp. 48, 52, 57, and 61. 130 Cash flow averaged $24.4 billion annu ally in the first 5 years of the 1950’s, de clined slightly when farm income dropped in the next 5-year period, more than made up this loss during the first half of the 1960’s, and then jumped to an annual rate of $32.5 billion in 1965-68. In this last period, however, the relative gains in cash flow did not keep up with those in required capital flows. Whereas capital flow aver aged 30 per cent of cash flow in the 1950’s and early 1960’s, this ratio increased to 33 per cent in 1965-68 (Table 9 ). Thus the burden posed by capital requirements, viewed in relation to the cash flow from which they might be financed, has increased. Over the 1950’s, when the relative capi tal burden was running at about 30 per cent of cash flow, farmers progressively reduced the proportion of cash flow they devoted to meeting capital requirements. In the first half of the 1950’s, the 27 per cent of cash flow that was used for capital purposes met the bulk of capital requirements. By the early 1960’s, however, only 21 per cent of cash flow was being used for capital pur chases, and the same share was also used in 1965-68 in spite of relatively greater capital spending. Relative reliance on credit, 1950-68. Credit thus became increasingly important as a source of funds for capital expenditures. In the early 1950’s, only 13 per cent of capi tal flows were met by an increase in debt. Ten years later this ratio had risen to 32 per cent, and by 1965-68 it averaged 37 per cent. One must go back 50 years to find a similar degree of reliance on credit. Writing in the 1950’s, Tostlebe noted:26 To a rem arkable degree, farm ers have financed the increase in farm capital w ith their own in comes and savings. A com parison of the volume 26 ibid., p. 19. of new capital th at was financed by loans and book credits w ith th a t w hich was financed w ith funds derived from gross farm incom e and sav ings shows th at in every decade fo r w hich we have inform ation, save the one im m ediately pre ceding 1920, farm ers supplied by far the greater part of the funds th at financed the capital acquisi tions. A few years later, Johnson was still able to state:27 . . . even when it is assumed th at all increases in loans and credit were used to increase agricultural assets, their contribution has generally been less im portant over the past two decades [1940-59] than either depreciation or net income as a source of financing. According to our estimates, however, in 1958 credit became a more important source of capital than net income. In fact, increases in debt have recently rivaled de preciation allowances for the lead in sup plying capital, whereas 10 years earlier they were only one-third as large. Thus, in projecting credit demands it is not enough to cite the capital flows antici pated. It appears equally important to pro ject the cash flow and also the proportion of that flow that farmers will be willing or forced to apply to satisfaction of the pro jected capital needs. In so doing, additional uncertainties are obvious. Is a major change in net income probable? Has the postwar trend toward less internal financing reached its lowest point? How probable is a higher savings rate in the near future? Current factors affecting credit use. In the last period of markedly increased participa tion of creditors in the financing of agricul ture, that of 1900-20, Tostlebe found two primary factors in operation. One was the pressure of financing farm transfers at the newly inflated prices.28 Physical farm en largement was not a major factor, but the average dollar value of assets per farm rose 27 D. G ale Johnson, op. cit., p. 377. 28 Tostlebe, op. cit., pp. 140 and 145-147. CAPITAL AND CREDIT REQUIREMENTS OF AGRICULTURE by 79 per cent between 1900 and 1910 and by 91 per cent between 1910 and 1920.29 Tostlebe also speculates that family living offered stiff competition for available funds:30 First, and m ost im portant, inflated expenditures for fam ily living probably m ade heavy inroads on the incomes of m any farm ers. The rise in prices of that period m ade necessary m uch greater outlays to m aintain the prew ar level of living. But m ore than that, the prosperity of the times encouraged farm ers to spend freely, so that the level of living for m any farm ers was substantially higher during this period than before. These elements appear in the current situation, although they do not seem to dominate it. Typical farms are now so large that an average person seeking to enter farming through purchase necessarily has to borrow a large portion of the funds re quired. Also, family farms approaching op timum size increasingly represent a quan tity of assets that a typical farmer is not expected to save during a lifetime, and so farmers are more likely to remain indebted throughout their career. On the other hand, in these circumstances, alternatives to the use of credit— leasing of land and equip ment from nonfarm investors, vertical inte gration, or corporate ownership— are spreading and lessening the credit demands made directly by farmers, although prob ably increasing credit demands of these other entities of the farm production sector. The desire to raise family living levels may also be a powerful factor in current borrowing. National television networks have exposed farmers more to consumer amenities. Among landowners, income after depreciation and expansion allowances might not permit a significant rise in living conditions, but paper capital gains may have imparted a sense of financial prosperity 29 I b id ., p. 85. 30 I b id ., p. 145. 131 reflected in spending. Several authors have noted that short of selling their land, farm ers can tap these gains only by offering them as collateral for increased debt. This proc ess has probably occurred in subtle ways. Perhaps depreciation allowances, which after all look just like net income, are con sumed in current living, and the tractors that were purchased for cash 10 years ago are today bought on the instalment plan. Or, instead of saving toward a downpayment on the adjacent “80,” a farmer simply plans to use his inflated equity in his present hold ings to effect a completely debt-financed purchase. The attitudes of farmers and lenders to ward the future of farming and toward what constitutes appropriate uses and terms of farm credit are obviously important deter minants of the proportion of capital needs financed by debt. Farmers must be willing to borrow and lenders to lend if outstanding credit is to increase— and both were ob viously willing over the last 20 years. The outlook for product and land prices must be important in the determination of these attitudes. In recent years, lenders that iden tify their interests most closely with those of agriculture— retiring farmers and the co operative credit system— have provided a larger share of credit, and other lenders have employed more agriculturally trained loan officers. Thus, it is not surprising that the outlook and attitudes of borrowers and lenders have apparently tended to coincide. The question of attitudes leads into the dominant feature of the present situation— farm enlargement— that was largely absent in the previous period. According to Tost lebe, average physical assets per farm na tionwide remained almost unchanged be tween 1870 and 1940, although their composition was altered.31 A slow decline in the size of Southern farms concealed a 31 Ibid., p. 85. 132 slow increase in the size of Midwest and Western enterprises, but nowhere did ex pansion match that started in the 1940’s. Since then, the benefits of, or competitive necessity for, enlargement became more ob vious to farm lenders. In fact, numerous educational efforts attempted not only to instruct lenders in these matters, but also to advise them to tell farmers about the need to expand in order to raise income and about how credit could assist this endeavor. Greater appreciation of the leverage that could be attained through credit was in stilled in lender and farmer alike. The importance of these considerations emerged in the 1960 Sample Survey of Agri culture, in which a large national sample of farmers for the first time was asked to enumerate debts owed to various sources. About 58 per cent of all farmers were in debted to varying degrees. When Garlock compared indebted operators to those with out debt, he found that:32 Regardless of w hether the farm ers were clas sified by age, years on the farm , tenure, or type of farm they operated, the indebted farm ers, on the average, conducted larger-scale operations than the debt-free farm ers. The value of the land and buildings they operated was greater, they leased m ore land, and they owned m ore land. Also they sold products of greater value, earned m ore net cash income from farm ing, and had larger offfarm incomes and m ore net incom e from all sources than did the debt-free farm ers. A lthough credit was indispensable to indebted farm ers in building up and operating large farm businesses, it is questionable w hether use of credit was fundam entally responsible for their larger, m ore profitable operations. W hat the data p rob ably m ean is th at the farm ers who used credit were m ore energetic and aggressive, m ore willing to take risks, and less willing to w ork only with the assets they owned outright than were the debtfree farm ers. This is indicated by their m ore extensive use of leased land as well as by their use of credit. 32 Fred L. G arlock, F arm ers and T heir D e b ts . . . The role o f credit in the farm econ om y, A gricultural Economic R eport No. 93, USDA (W ashington: Govt. Printing Office, June 1966), pp. 8 and 9. These expansionist characteristics of the credit users— particularly the heavy credit users— are pointed up m ore sharply w hen farm ers are clas sified according to the extent of their indebtedness. . . . D espite their small equities, the m ost heavily indebted farm ers ow ned farm s of nearly as high value as those owned by the debt-free farm ers. But the m ost significant point is the extent to w hich the indebted farm ers used their equities as a ful crum for developing larger operations than their own financial resources would support. T he m ost heavily indebted farm ers ow ned 3 Vi tim es as m uch land, and operated 6 times as m uch land, as they could have ow ned or operated w ithout b o r rowing and leasing. By using these m ethods of expanding operations, they raised their net cash farm incomes to levels approxim ating those of the other groups whose equities were m uch greater. In the past, these expansionary desires, grounded in the economics cited by G ar lock, might have been financed in large part by saving. But in view of the fact that many farmers have come to regard credit as an appropriate tool for achieving these ends, that lenders encourage this use of credit, and that both young farmers and holders of paper gains are probably disinclined to post pone attainment of family living goals, a continued high or perhaps even increased use of credit relative to required capital flow seems probable as long as the factors forc ing farm enlargement continue operative. These have been found to be rooted in tech nological innovation and seem in no danger of expiring before 1980. They have already been found responsible for higher capital requirements and now are also found re sponsible for greater relative use of credit in meeting these requirements, given the farm income situation since 1950. Barring a drastic rise in net farm income, it seems reasonable to expect a savings rate no higher than that of the 1960’s. On the other hand, if real net income per farm rises at a rea sonable pace, there would be no great pressure to reduce the savings rate. The credit projections that follow incorporate this reasoning. CAPITAL AND CREDIT REQUIREMENTS OF AGRICULTURE Projected credit requirem ents, 1 9 7 0 -7 9 The analytical framework outlined above and exemplified by Table 9 can be used to project credit requirements to 1980 under the various capital models that have been developed. For each past period shown in Table 9, capital flow, capital consumption, net income, and increase in debt were all known, and thus the amount of cash flow that was used to meet capital requirements could be obtained by subtraction. For the projections, however, only the capital flows and the capital consumption allowances that are consistent with these flows are initially given. But if the course of net farm and non farm income is projected, the total cash flow becomes known. Then, if a further projection is made about how much of their cash flow farmers will devote to meeting capital requirements, the amount of increase in borrowing— the credit requirement— can be determined as the residual. Internal financing in the 1970fs. To project internal financing, the course of net farm income, nonfarm income of farmers, and the savings rate must first be estimated. Instinctively, one wants to project total net farm income by projecting gross farm income and production expenses and cal culating the difference. But a different ap proach is taken here, based on the belief that over a period of years (1) advances in real per-farm income will parallel gains in per capita income achieved in the nonfarm economy; (2) technological advances will cause farm numbers to decline indepen dently of the course of farm income; and (3) the general price level will tend to rise. These trends have been in evidence over the postwar period. Between 1958 and 1968, for example, operators’ real net farm income per farm (income adjusted for changes in the index of prices paid by farm ers) rose by 3.3 per cent annually, while 133 national per capita real personal income rose at a yearly rate of 3.4 per cent. How ever, the number of farms decreased by 3.2 per cent annually, and so total real opera tors’ net farm income was unchanged. On the other hand, prices paid by farmers rose at an average annual rate of 1.6 per cent, and so total net farm income in current dollars increased at the same rate. For the 1970’s the National Planning Association projects an annual advance of 3.25 per cent in national per capita real personal income.33 In the long run, the inter play of competitive and political forces will tend to ensure that farmers participate to roughly the same extent in this advance in the national level of living. At the same time, as farm enlargement continues and farm numbers therefore decline, this rate of gain in real income per farm may be achieved by a merely stable total real net income. However, if prices paid by farmers tend to rise by an average of 2 per cent annually, as is projected in Models HT and HM, total net farm income would also have to rise by 2 per cent annually to yield the projected real gains. Note that this projection of net farm income implicitly requires that gross income rise sufficiently not only to provide the in crease in net income, but also to cover any rise in production expenses (including in terest payments on projected increases in debt) and in projected depreciation allow ances. Total nonfarm income of the farm popu lation has been rising rapidly and the trend is expected to continue as nonfarm employ ment and investment opportunities become increasingly available to rural residents. 33 Ahmad Al-Samarrie, Morris Cobern, and Take shi Hari, National Economic Projections to 1978/79 (Washington: National Planning Association, Jan. 1969), p. 117. 134 Between 1958 and 1968, nonfarm income rose at an annual rate of 5.8 per cent in spite of an average yearly drop of 4.8 per cent in the farm population. For the 1970’s, total nonfarm income is projected to in crease by 5 per cent annually, with twofifths of the gain reflecting projected price inflation. To summarize the cash flow projections, net farm income is projected to increase from $16.1 billion in 1968 to an annual average of $17.4 billion in 1970-74 and $19.3 billion in 1975-79. Annual nonfarm income is projected to rise from $11.8 bil lion in 1968 to an average of $14.3 billion in 1970-74 and $18.3 billion in 1975-79. Capital consumption allowances, which vary among the capital models according to the growth of the machine stock foreseen, are projected at annual levels of $7 billion to $8 billion in 1970-74 and $7 billion to $10 billion in 1975-79. Total cash flow, which was $34 billion in 1968, is therefore pro jected to rise to about $39 billion per year in 1970-74, and about $45 billion in 197579, with some variation among models as shown in Table 10. Of this cash flow, 21 per cent is projected to be allocated to meeting capital require ments— the same proportion that was so allocated on average during 1960-68. In ternal financing of capital flow is thus ex pected to average about $8 billion per year in 1970-74 and $9 billion to $10 billion in 1975-79, up from the average of $6.9 bil lion in 1965-68. TABLE 10 PROJECTED FINANCING OF ALTERNATIVE CAPITAL FLOWS In billions of dollars unless otherwise indicated 1970-74 Sources of — Model NC Model HT Model B 1975;-79 Model HM Model NC Model HT Model B M odel HM 5-year total Capital financing From cash flow (21 per cent of cash flow) ............... Increase in debt ................................................................... Total capital flow ...................................................... Cash flow Capital consumption allowances ................................ N et farm income ................................................................. Nonfarm income ................................................................. Total cash flow ........................................................... 40.4 15.8 56.2 40.8 22.9 63.7 40.5 18.5 59.0 41.5 34.4 75.9 46.7 10.0 56.8 47.6 29.3 76.8 47.0 17.0 64.0 49.9 44.5 94.4 33.6 87.1 71.7 192.4 35.5 87.1 71.7 194.3 34.3 87.1 71.7 193.1 38.9 87.1 71.7 197.7 34.5 96.5 91.5 222.5 38.4 96.5 91.5 226.4 35.8 96.5 91.5 223.9 49.6 96.5 91.5 237.6 Annual average Capital financing From cash flow (21 per cent of cash flow) ............... Increase in debt ................................................................. Total capital flow ...................................................... Cash flow Capital consumption allowances ................................... N et farm income ................................................................. Nonfarm in c o m e ................................................................... Total cash flow .......................................................... 8.1 3.2 11.2 8.2 4.6 12.7 8.1 3.7 11.8 8.3 6.9 15.2 9.3 2.0 11.4 9.5 5.9 15.4 9.4 3.4 12.8 10.0 8.9 "1^9 6.7 17.4 14.3 38.5 7.1 17.4 14.3 38.9 6.9 17.4 14.3 38.6 7.8 17.4 14.3 39.5 6.9 19.3 18.3 44.5 7.7 19.3 18.3 45.3 7.2 19.3 18.3 44.8 9.9 19.3 18.3 47.5 490.1 108.1 358.9 91.3 409.7 136.8 End of period Assets and debt Selected assets ....................................................................... Outstanding debt ................................................................. 281.1 71.5 378.6 78.9 320.5 74.3 344.2 92.3 281.1 81.6 Per cent Analytical ratios Capital flow /cash flow ...................................................... Increase in debt/capital flow ......................................... Average annual change during period: Selected assets ........................................................... Outstanding debt ........................................................ Debt/assets, end of p e r io d ................................................ 29 28 33 36 31 31 38 45 26 18 34 38 29 27 40 47 5.1 25.4 5.1 7.1 20.8 2.2 5.9 23.2 3.4 9.8 26.8 2.7 29.0 5.3 6.5 22.1 2.3 4.2 25.4 3.5 8.2 33.4 135 CAPITAL AND CREDIT REQUIREMENTS OF AGRICULTURE Projected debt expansion under alternative capital models. For each capital model, the projected increase in outstanding farm debt consists of the difference between the capital flow and the internal financing projected by that model. These calculations are shown in Table 10, and the resulting debt projec tions are summarized in Table 11. Under Model NC, in which farm assets remain unchanged in value, substantial but decreasing annual additions to debt would still be required, and outstanding debt would reach $81.6 billion in 1980, up from $52.0 billion in 1969. Thus it appears that farm ers would for some time tend to incur siz able amounts of new debt simply in the course of replacement and transfer of to day’s capital plant at today’s prices. In Model B, in which capital flows advance moderately, outstanding farm debt would reach $91.3 billion in 1980. The rate of debt expansion would fall to an annual rate of about 4 per cent by 1980, compared with the actual rate of 9.6 per cent in 1965-68. However, debt would grow more than twice as rapidly as assets and in 1980 would con stitute 25 per cent of assets, compared with 19 per cent in 1968. The relatively greater land price increases projected in Model HT would lead to an outstanding farm debt of $108.1 billion in 1980. Annual gains in debt would average about 7 per cent; in dollars, the annual in crease during 1975-79 would average $5.9 billion. However, because of the large rise in farm real estate values projected by this model, the ratio of debt to assets would rise only slowly, to perhaps 22 per cent in 1980. Model HM represents the greatest in crease in capital flows, resulting mainly from large real additions to farm machine stocks. Outstanding debt would continue to grow almost as fast as in recent years and TABLE 11 ALTERNATIVE PROJECTED CREDIT REQUIREMENTS In billions of dollars unless otherwise indicated Projection 1950-54 1955-59 1960-64 1965-69 1970-74 1975-79 Debt: increase during period Actual Model NC HT 4.7 8.2 12.4 20.0* 15.8 22.9 18.5 34.4 B HM 10.0 29.3 17.0 44.5 Debt: average annual increase Actual Model NC HT .9 1.6 2.5 4.0* B HM 3.2 4.6 3.7 6.9 2.0 5.9 3.4 8.9 Outstanding debt:: end of period Actual Model NC HT 15.4 23.6 36.0 56.0* 71.5 78.9 74.3 92.3 B HM 81.6 108.1 91.3 136.8 Outstanding debt: annual growth rate (per cent) Actual Model NC HT 7.6 8.9 8.8 9.2* 2.7 6.5 4.2 5.1 7.1 5.9 9.8 B HM 8.2 D ebt/asset ratio: end of period (per cent) Actual Model NC HT 10 12 16 B HM 191* 25 21 23 27 29 22 25 33 * Estimate based on data in Table 9. f As of Jan. 1, 1968. would reach $136.8 billion in 1980. In 1975-79, annual additions to debt would average $8.9 billion. The rise in debt would far outpace growth in the value of farm assets, so that the debt/asset ratio would rise to 33 per cent by 1980. In comparison with the experience of recent years, these projected credit demands represent somewhat slower rates of expan sion in debt. However, no model represents continuation of the combination of capital growth that has actually prevailed in 196368— significant real additions to machine stocks plus relatively rapid increases in land values. If this experience were to continue through all the years to 1980, credit de mands would probably prove larger than any of those projected. But the historical perspective on capital flows provided in Sec tion III indicates this to be a somewhat ex treme expectation. 136 Sizable increases in debt and in the debt/ asset ratio are projected by each model. The levels reached in each by 1980 are not so large as to be impossible, but neither can the process continue indefinitely, especially at the rate represented by Model HM. At some point, capital flows may recede and/or farmers’ savings may increase. Or nonfarm ers may supply significantly more of the capital needed. A watch should be main tained for the occurrence of such structural changes on a significant scale, as these events would lead to changes in demands for credit. V. SOURCES OF CREDIT, 1950-79 From 1950 through 1968, $41.3 billion of additional credit was supplied to farmers by a great many individual and institutional lenders. As will be shown, the share sup plied by some lender groups increased over this period, whereas other groups became less important sources. Commercial banks, however, maintained about the same rela tive role over the entire period. In the preceding section, further sizable increases in total farm debt were projected for the 1970’s. To continue to provide their recent historical share of such expansion, banks would have to continue to expand their farm lending substantially. The amount of increase necessary to achieve this target varies among models, as it depends on the projected size of total credit requirements and also on how much of the credit is in curred to support non-real-estate rather than real estate spending. These projections of required bank credit growth are made here for Models HT, B, and HM. Attention then turns to the supply of funds at rural banks— to examine how banks have been able to increase farm lending rapidly since 1950 and whether they will be able to continue the pace. Future deposit growth is projected and then compared with the various projections of future farm loan demands, to provide an indication of the degree to which internal growui of rural banks is or is not likely to be adequate to meet farm credit demands arising in a variety of possible future farm capital situa tions. Sources of outstanding credit, 1 9 5 0 -6 8 Credit to farmers is provided by a large number of individuals, dealers, and institu tions. Estimates of the amount outstanding from each of several classes of lenders are published annually by USDA. For the major institutional lenders, these estimates are based on lender reports submitted at least annually. Commercial banks hold the larg est outstanding farm loan total among these reporting lenders. Other institutions in this group are insurance companies, the Farmers Home Administration, and the agencies (Federal land banks, Federal intermediate credit banks, and production credit associa tions) that comprise the cooperative credit system supervised by the Farm Credit Ad ministration. Many other lending institutions make small amounts of loans to farmers, but in general they do not report their volume of farm loans. In the USDA estimates of farm lending, such loans are grouped with credit provided by individuals. Taken together, individuals, dealers, and these nonreporting institutions are the most important source of farm, credit. Retiring farmers and other sellers of farms, in par ticular, provide large amounts of credit to the purchasers by taking mortgages or land contracts. Merchants, dealers, and individ 137 C A PITAL AN D C R E D IT R E Q U IR E M E N T S OF A G R IC U LT U R E uals such as farm landlords also supply large amounts of non-real-estate credit to finance purchases of production inputs, machinery, and livestock. In this group, national farm supply and machinery cor porations have emerged as a major source of financing. In general, these creditors do not report their loan volume annually, and so USDA estimates of the debt they hold may contain relatively large errors. This is particularly true of recent yearly changes, as census surveys of farm debt made in 1960 and 1965 have permitted improved evaluation of the relative longer-term role of these lenders. Holders of outstanding debt. Outstanding farm debt on January 1, 1969, totaled $52.0 billion, up from $23.6 billion at the start of the decade and $10.7 billion in 1950 (Table 12). Individuals, dealers, and nonreporting institutions held about two-fifths of the out standing debt throughout this period. OUTSTANDING FARM DEBT, 1950-69 By major lender groups excluding CCC T o t a l ......................................... 1950 1955 1960 1965 1969 3.0 4.9 4.1 7.2 6.5 11.5 9.7 17.5 13.6 26.2 1.4 1.2 1.9 2.1 3.8 2.8 6.1 4.3 10.1 5.8 2.3 3.2 4.9 7.1 10.3 2.3 .5 3.4 .7 4.9 .8 7.6 1.3 10.9 1.3 15.4 23.6 36.0 52.0 10.7 Per cent of total Banks ................................................ Money market lenders ............... Cooperative credit system . . . Life insurance companies . . . Dealers and individuals (non-real-estate)* ............... Individuals (real estate) * ......... Farmers Home Administration. T o t a l......................................... 28 46 27 46 27 49 27 49 26 50 13 11 12 13 16 12 17 12 19 11 22 21 21 22 5 21 4 20 21 4 20 22 5 100 100 100 100 100 21 3 Per cent of bank and money market total Banks ................................................ Money market lenders ............... 38 62 37 63 36 64 36 64 34 66 T o t a l ......................................... 100 100 100 100 100 * Includes other nonreporting lenders. S o u r c e . — A gricu ltu ral F in an ce R e v ie w , U SD A , Apr. 1969, pp. 2, and 22 and 23. Th e B alan ce S h e et o f A g ricu ltu re, 1968, U SD A , Jan. 1969, pp. 13 and 15. Data are as of January 1 of each year. AVERAGE ANNUAL GROWTH RATE OF OUTSTANDING FARM DEBT, 1950-68 By major lender groups excluding CCC Lender group Billions of dollars Banks ................................................ Money market lenders ............... Cooperative credit system . . . Life insurance companies . . . Dealers and individuals (non-real-estate)* ............... Individuals (real estate)* ......... Farmers Home Administration. TABLE 13 In per cent TABLE 12 Lender group Among the major lending institutions, com mercial banks ranked first with $13.6 bil lion in 1969, about one-fourth of the total farm debt. Outstanding loans at banks had also increased markedly from $6.5 billion in 1960 and $3.0 billion in 1950. The cooperative credit system held nearly one-fifth of outstanding farm debt in 1969. Its volume of $10.1 billion represented a rapid rise from $3.8 billion in 1960 and $1.4 billion in 1950. Life insurance com panies held $5.8 billion of farm mortgage loans in 1969, representing 11 per cent of total farm debt. Finally, the Farmers Home Administration held 3 per cent of the total debt. Farm debt has been rising by about 9 per cent yearly since the mid-1950’s (Table 13). During 1965-68, total debt rose at an 1950-54 1955-59 1960-64 1965-68 6.8 8.0 Money market lenders . . . . Cooperative credit system 6.4 Life insurance companies 11.8 Dealers and individuals 6.6 (non-real-estate) * . . . . Individuals (real estate) * .. 8.1 Farmers Home Administration .................. 5.5 9.3 9.8 8.4 8.8 8.9 10.6 14.6 6.6 10.0 8.7 13.5 7.7 8.7 7.9 9.8 7.3 9.5 9.4 3.5 8.6 1.0 8.9 8.8 9.6 7.6 * Includes other nonreporting lenders. annual rate of 9.6 per cent, paced by an nual gains of 13.5 per cent in outstanding debt held by the cooperative credit system and assisted by rapid expansion of farm loans at each of the other principal lenders. Expansion at banks averaged 8.9 per cent annually during 1965-68 and 8.4 per cent during 1960-64, in each case only slightly below the growth rate of total farm debt. Sources of additions to debt. Another per spective on farm credit is provided by ex amination of the sources of net additions to outstanding debt, which are shown in 138 TABLE 14 SOURCES OF ADDITIONS TO FARM DEBT, 1950-69 Excluding CCC In billions of dollars unless otherwise indicated Lender group 1950-54 1955-59 1960-64 1965-69* Relative role of banks, 1 9 5 0 -6 8 5-year total Banks ....................................... Money market lenders . . . . Cooperative credit system Life insurance companies. Dealers and individuals (non-real-estate) f . . . . Individuals (real estate)f .. Farmers Home A dm inistration.................... 1.2 2.3 .5 .9 2.3 4.3 3.2 6.0 4.9 10.9 1.9 .8 2.3 1.5 5.0 1.8 .9 1.1 1.7 2.2 4.0 1.4 2.8 4.1 .2 .1 .4 .1 Total ................................. 4.7 8.2 12.4 20.0 Banks ....................................... Money market lenders Cooperative credit system Life insurance companies. Dealers and individuals (non-real-estate) f . . . . Individuals (real estate) f . . Farmers Home A dm inistration.................... .2 .7 .7 Annual average .5 .9 .2 .4 .2 .2 .3 .2 .3 .6 1.2 .5 1.0 2.2 1.0 .4 .3 .5 .6 .8 .8 .1 Total ................................. .9 Banks ....................................... Money market lenders Cooperative credit system Life insurance companies. Dealers and individuals (non-real-estate) f . . . . Individuals (real estate) f .. Farmers Home A dm inistration.................... 25 48 1.6 4.0 28 52 26 48 25 54 11 19 23 9 19 12 25 9 19 20 18 20 24 18 22 21 3 2 3 100 100 2.5 Per cent of total Total ................................. 100 100 Percentage of bank and money market total Banks ....................................... Money market lenders 34 66 35 65 35 65 31 69 Total ................................. 100 100 100 100 * Data shown for 1965-69 are actual values for 1965-68 multiplied by 1.25 to facilitate comparison with previous 5-year periods. f Includes other nonreporting lenders. Table 14 for 5-year intervals since 1950. In each 5-year period, dealers and individuals provided about two-fifths of the increase, while banks provided about one-fourth. The increase at the cooperative credit system was low in 1950-54, when real estate lending by the Federal land banks was restrained by appraisal methods that proved outmoded. More recently, the cooperative credit sys tem has been supplying around 20 to 25 per cent of the additions to farm credit. In 1965-68, banks and the cooperative credit system each supplied 25 per cent of the total increase. After providing 19 per cent of the gain in farm credit in 1950-54, the share of life insurance companies dropped to about 10 per cent in subsequent periods. New insights into the relative role of vari ous lenders, as well as a better basis for projection of future roles, may be secured by further aggregation of lenders into groups with key common characteristics. Thus, it is useful to group three lender classes— the cooperative credit system, life insurance companies, and the nonreporting creditors who supply non-real-estate credit — into one category called “money market lenders” because the supply and the real or opportunity cost of the credit provided by each is influenced by conditions in the na tional money market. The cooperative credit system obtains its funds by selling money market instruments. Such instruments also comprise a major alternative investment for funds of life insurance companies. And, the non-real-estate-lending volume of non reporting creditors is dominated by national corporations that supply production inputs to farmers, and such concerns are likely to have obtained funds for these loans in the money market or by borrowing from money market banks. Local merchant credit re mains in this category as separate data are not available, but the amount of this misclassification is relatively small. After this consolidation, four groups of farm lenders remain: (1) the money market lenders, (2) individuals who hold real estate debt, (3) banks, and (4) the Farmers Home Administration. Individual holders of real estate debt are sellers of farms who took mortgages or sold by land contract for tax reasons, in order to make the sale, to obtain a higher price, or to retain a continuing investment in their farm. Their volume of lending depends on CAPITAL AND CREDIT REQUIREMENTS OF AGRICULTURE the strength of these considerations and on prices and activity in the farm real estate market. Banks are set apart as a third major lender group because their farm lending is significantly affected by factors different from either the sellers of farms or the money market lenders. Loans for real estate pur chases constitute only 15 per cent of banks’ farm lending, and so land market considera tions are less important in determining loan volume than in the case of sellers of farms. Also, the bulk of banks active in farm lend ing find it difficult to participate effectively in national money markets under present conditions and so are dependent on local sources of funds. Thus, whereas money mar ket participants active in farm lending face a very elastic supply function, in that their demands on the money market constitute only a small portion of total national de mands for funds, most rural banks face a relatively inelastic supply. In seeking to en hance the growth of their lending resources, they are limited both by legal ceilings on interest rates they can offer on deposits and by the over-all economic growth being achieved by their community. In the long run, therefore, constraints are thereby placed on growth of farm lending at most rural banks, given present institutional ar rangements that exclude such banks from effective participation in the money market. The Farmers Home Administration com pletes the four lender groups. In its direct farm lending program, this agency makes supervised loans to farmers unable to obtain credit from the other lenders. The outstand ing volume of these loans failed to increase during the 1965-68 period and conse quently declined to 3 per cent of total out standing farm debt. In the projections that follow, it is assumed that the volume of farm lending by the Farmers Home Administra 139 tion will remain unchanged over the next decade. Share of outstanding debt. As outstanding farm debt rose from $10.7 billion in 1950 to $52.0 billion in 1969, the share con sisting of real estate debt held by individ uals fluctuated narrowly between 20 and 22 per cent of the total. The share held by banks declined slowly from 30 per cent in 1952 to 26 per cent in 1969. Conversely, the portion held by money market lenders rose from 46 per cent in 1950 to 50 per cent in 1969 (Table 12). Of the total debt held only by banks and money market lenders, the share held by banks declined from 39 per cent in 1952 to 34 per cent by 1969. In eight of the years from 1950 to 1968, bank credit grew faster than credit from money market lenders, but on average the latter tended to expand more rapidly throughout the period. In the 4 years 1965-68, debt at banks rose by 8.9 per cent annually, whereas debt held by money market lenders grew at a rate of 10.6 percent (Table 13). Share of additions to debt. While the share of outstanding debt held by banks eroded slowly over the entire period since 1950, banks’ share of additions to farm credit showed no downward trend during 195064. Although year-to-year fluctuations were large, banks on average provided slightly over one-fourth of the total increase in farm credit and slightly over one-third of the total gain at banks and money market lend ers. These shares of new credit were slightly below the shares of outstanding credit with which banks entered the period. This differ ence explains the erosion observed in the shares of outstanding credit. In 1965-68, banks’ share of additions to credit dropped to 25 per cent of the grand total and to 31 per cent of the sum provided by banks and money market lenders. The 140 share provided by money market lenders in creased (Table 14). Projected credit expansion by major lenders, 1 9 7 0 -7 9 The preceding section presented three al ternative projections of increases in farm debt to 1980, based on three different capi tal models and a single farm income and savings rate projection. Given these pro jected additions to total farm credit, esti mates are here made of the corresponding increases in farm lending by banks that would be necessary for banks to maintain their recent share— about one-third— of the total credit expansion projected for banks and money market lenders together. Credit from sellers of farms. As a prelim inary step, it is necessary to estimate the amount of additional real estate credit that may be provided by individuals, particularly sellers of farms. Credit from this source is virtually certain to be related to the value of farms transferred by sale; thus more would be expected if land prices rise rapidly, as projected by Model HT, than if they rise more moderately, as projected by Models B and HM. In addition, credit pro vided by sellers has recently been increas ing relative to the value of transfers. It is estimated that such credit may have equaled 18 per cent of the value of real estate sales in 1955, 22 per cent in 1960, 24 per cent in 1965, and 29 per cent in 1968. Several factors have contributed to this increase, chief among them being substantial capital gains tax advantages to sellers who provide credit under a land contract and the ability of sellers to offer lower downpayments than most institutional lenders. Again, the future trend of this ratio seems likely to be posi tively related to the rate of gain in real estate values. With these considerations in mind, past increases in real estate credit provided by individuals were related to estimated capital flows required by real estate transfers. This ratio was estimated at 11 per cent in 1955— 59, 17 per cent in 1960-64, and 20 per cent in 1965-67. For Models B and HM, the ratio was projected to average 22 per cent in 1970-74 and 24 per cent in 1975-79. For Model HT, in view of its more rapid rise in land prices, the ratio was projected at 25 and 29 per cent, respectively. These relationships were applied to the value of real estate capital flows projected by these models (Table 7) to obtain the estimated amounts of additional real estate credit that may be supplied by individuals and other nonreporting lenders. The estimates are shown in Table 15. Projected loan demands on banks. Subtrac tion of the projected seller-supplied credit from the total credit requirements shown in Table 15 (from Table 10) yields projections TABLE 15 ALTERNATIVE PROJECTED FARM LOAN EXPANSION AT MAJOR LENDER GROUPS, 1970-79 In billions of dollars Projection, and period Total Individuals (real estate) Banks and At banks money to maintain market relative share lenders Total 5-year total Model HT 1970-74 ................................... 1975-79 ................................... Model B 1970-74 ................................... 1975-79 ................................... Model HM 1970-74 ................................... 1975-79 ................................... Individuals (real estate) Banks and A t banks money to maintain market lenders relative share Annual average 22.9 29.3 7.0 11.0 15.9 18.3 5.4 6.3 4.6 5.9 1.4 2.2 3.2 3.7 1.1 1.3 18.5 17.0 5.5 6.7 13.0 10.3 4.4 3.5 3.7 3.4 1.1 1.3 2.6 2.1 .9 .7 34.4 44.5 5.6 7.0 28.8 37.5 9.8 12.8 6.9 8.9 1.1 1.4 5.8 7.5 2.0 2.6 141 CAPITAL AND CREDIT REQUIREMENTS OF AGRICULTURE This pace would be significantly below ac tual rates of farm loan expansion at banks since 1955. But with the much greater capital flows represented by Model HM, farm loans at banks have to expand at an annual rate of 10 per cent, or even faster than they have been growing since 1955. of the amounts to be supplied by banks and money market lenders together. If banks are to supply one-third of the latter totals, they would have to increase their farm loans by the amounts shown in Table 15. In Table 16, the required additions to farm lending by banks are shown in a more familiar context, as the sum that outstanding farm loans would reach in 1975 and 1980 and as the average annual rate of increase in outstanding loans in each 5-year interval. Outstanding loan volume required in 1980 ranges from $22.5 billion (Model B) to $37.9 billion (Model H M ). These sums may be compared with estimated outstand ing volume of $14.6 billion at the end of 1969. The projections show that credit de mands on banks would be very large if the high rate of machinery investment projected by Model HM should materialize. Rapidly rising land prices, as represented by Model HT, would have a more moderate effect on credit demands on banks, because the pro jected increase in seller-financing of real estate transfers meets a significant portion of credit demands arising from that source. Under the moderately greater capital flows of Models B and HT, and given the projected internal financing, annual rates of farm loan expansion averaging 6 per cent would suffice to maintain the relative role of banks in farm lending during 1970-74. Supply of funds at rural banks Would it be easy or difficult for banks to expand farm lending at the various rates projected above and thereby to maintain their share of the farm credit market? At present, rural banks depend primarily on growth in their deposits, most of which originate locally, for expansion of their lending resources. Farmers' deposits. To some extent, banks act as financial intermediaries among farmers. In the early 1950’s, the outstanding loans to farmers represented about one-half of farmers’ deposits, and at that time over all farm lending activity at banks could be viewed as being on average sustained com pletely from the deposits of farmers them selves. By 1960, however, the volume of farm loans was almost equal to that of deposits, and in 1968 it was 43 per cent larger (Table 17). It is evident that banks have been making farm loans from funds received from sources other than farmers alone. TABLE 16 ALTERNATIVE PROJECTED FARM LOAN EXPANSION REQUIRED AT BANKS TO MAINTAIN BANKS' RELATIVE ROLE IN FARM LENDING, 1970-79 In billions of dollars unless indicated otherwise Outstanding loans Projection 195054 195559 196064 196569 1970-74 197579 195054 195559 4.1 6.5 Increase during period Actual .............................................. Model HT ..................................... B ......................................... H M ..................................... 1.2 Actual ........................................... Model HT ..................................... B ............................ .. HM ................................... .2 2.3 3.2 4.9* ..................................... ..................................... ..................................... 5.4 4.4 9.8 * Estimate based on data shown in Tables 12 and 14. 9.7 197074 197579 20.0 19.0 25.1 26.3 22.5 37.9 14.6* 6.3 3.5 12.8 Annual growth rate during period (per cent) 6.8 1.1 .9 2.0 196569 End of period Average annual increase .5 .6 1.0* ..................................... ..................................... ..................................... 1960-64 1.3 .7 2.6 9.3 8.4 8.5* 6.5 5.5 10.5 5.6 3.4 8.6 142 TABLE 17 COMMERCIAL BANK FARM LOANS COMPARED WITH FARM ERS’ DEPOSITS, 1950-69 In billions of dollars unless otherwise indicated Year Bank loans to farmers Demand and time deposits of farmers Farmers’ loans as percentage of farmers’ deposits 1950............... 1955............... 1960............... 1965............... 1969............... 3.0 4.1 6.5 9.7 13.6 6.6 7.2 7.2 7.7 9.5 45 57 90 126 143 S o u r c e . — A g ricu ltu ra l F inance R e v ie w , U SD A , Apr. 1969, pp. 2, and 22 and 23. The B alan ce S h eet o f A g ricu ltu re, 1968, U SD A , Jan. 1969, pp. 10, 13, and 15. The discussion of farmers’ cash assets in Section III turned up conflicting views on the extent of the growth that banks can ex pect in farmers’ deposits during the next decade. As farmers purchase more inputs, cash working capital has become more im portant in farm operation, but farmers have also learned how to economize on these balances. If these offsetting trends continue, farmers’ demand deposits may show only moderate growth. It is possible, however, that banks could achieve more significant expansion in time deposits of farmers by attracting current and past savings away from alternative investments. Total deposit growth, 1950-68. Fortunately, total deposits at rural banks increased at a faster pace than farmers’ deposits alone. An indication of this is provided by the USDA’s index of deposits of country banks, which measures changes in deposits at banks in towns with population under 15,000 in 20 agricultural States. These primarily rural banks achieved annual growth in total de posits averaging 3.4 per cent in the 1950’s, 5.8 per cent in 1960-64, and 8.5 per cent in 1965-68 (Table 18). The very significant recent acceleration in the growth of total deposits at these banks can be traced primarily to the expansion of time deposits. Since 1950, demand deposits have increased slowly, with annual expan sion averaging less than 3 per cent. Time deposits, however, rose at an average annual rate of about 10 per cent in the 1950’s and 15 per cent in 1960-68. At first, these rapid rates of expansion did not contribute much to total deposit growth, because time de posits represented only a small fraction— 13 per cent in 1950—of total deposits at these banks. But as the rapid pace continued, time deposits became more important, reaching 44 per cent of total deposits in January 1969. With this sizable component growing at 15 per cent annually, total deposits rose by 8.5 per cent a year in 1965-68 even though annual demand deposit growth aver aged only 4.2 per cent during these years. TABLE 18 DEPOSITS OF SELECTED COUNTRY BANKS, 1950-69* Period Total Demand Index of volume (1947-49 1950.................... 1955.................... 1960.................... 1965.................... 1969.................... 102 124 142 188 261 103 122 127 144 170 Time 100) 103 152 260 502 883 Average annual growth rate (per cent) 1950-54............. 1955-59............. 1960-64............. 1965-68............. 4.0 2.7 5.8 8.5 3.4 .8 2.5 4.2 8.1 11.3 14.1 15.2 * Data are for banks in towns with population under 15,000 in 20 agricultural States, as compiled and published by U SD A . Data for 1950-65 are averages for January of each year. Data for 1969 are as of January 1. S o u r c e . — Agricultural Finance Branch, U SD A . Deposit and farm loan growth compared, 1950-68. Data on farm loan growth at the universe of banks used in compiling the USDA index of country bank deposits are not available. It is likely, however, that they parallel the course of farm loans at all banks and that impressions obtained from a com parison of the deposit index with total farm loan growth will not be misleading. Such comparison shows that farm loans have tended to increase faster than rural bank deposits (Table 19). The gap was espe cially large in 1955-59, was reduced some what by faster deposit growth in 196064, and then almost closed by still faster deposit growth in 1965-68. In 1965-68 CAPITAL AND CREDIT REQUIREMENTS OF AGRICULTURE TABLE 19 COMPARISON OF DEPOSIT AND FARM LOAN GROWTH, 1950-68 Average annual rate of growth, per cent Period Total deposits of selected country banks 1950-54.......................... .....................4.0 1955-59.......................... .....................2.7 1960-64.......................... .................... 5.8 1965-68.......................... .....................8.5 Total farm loans at all banks 6.8 9.2 8.4 8.9 deposits rose by 8.5 per cent annually while farm loans expanded at a rate of 8.9 per cent. Confirmation of these recent relationships between loans and deposits is afforded by data from surveys of banks active in agri cultural lending, made annually since 1962 by The American Bankers Association (ABA). From 1962 to 1965, loan/deposit ratios at these banks tended to move up ward. In 1968, however, the distribution of banks by loan/deposit ratio was still much the same as in 1965 (Table 20). At the naTABLE 20 PERCENTAGE DISTRIBUTION OF ABA AGRICULTURAL BANKS, MIDYEAR 1962-68 By loan/deposit ratio Loans as percentage of deposits Year 1962........................ 1963........................ 1964........................ 1965........................ 1966........................ 1967........................ 1968........................ Under 50 41 38 31 27 27 21 25 50 to 59 34 33 35 32 31 33 33 60 and over 25 29 34 41 42 46 42 N o t e . —Data are for banks designated as “agricultural” banks by The American Bankers Association and thereby covered by the midyear agricultural credit situation survey. Approximately two-thirds of all insured commercial banks qualify as agricul tural banks under the criteria used. See footnote 39. S o u r c e . — T ren ds in A g ricu ltu ra l B an kin g: R e p o r t o f M id ye a r 1968 A g ricu ltu ra l C r e d it S itu ation S u rve y, The American Bank ers A ssociation, N ew York, 1968, p. 8. tional “median” bank covered by the ABA surveys, the 58 per cent gain in farm loans in 1963-68 was virtually identical to the 59 per cent increase in total deposits. At banks in the Plains States, however, deposits had risen by only 26 per cent compared to farm loan growth of 67 per cent.34 34 Trends in A gricultural Banking: R ep o rt of M id yea r 1968 A gricultural C redit Situation Survey (New Y ork: The A m erican Bankers Association, 1968), p. 8. 143 Projected deposit growth, 1970-79. It seems reasonable to believe that demand deposit growth at rural banks in 1970-79 may resemble postwar expansion to date, aver aging perhaps 3 per cent annually. This pro jection reflects the growing money needs of an expanding rural economy. However, the rate of money expansion is somewhat below the anticipated rate of economic growth, as persons and businesses continue to reduce the money balances that they hold in relation to their volume of trans actions. Projection of time deposit growth seems more speculative. To some extent, the large recent gains represent an adjustment by the public to increased attractiveness of time deposits relative to both demand deposits and nonbank investments. First, ceiling in terest rates prescribed by regulatory authori ties on passbook and other time deposits were raised to a level more competitive with those paid by nonbank financial intermedi aries, such as savings and loan associations, and with returns available from U.S. sav ings bonds and marketable securities. Second, in response to their increased loan requests and reduced liquidity, many banks began to offer time certificates of deposit on which they were permitted to pay higher rates of interest than on passbook savings accounts. Small banks, while unable to par ticipate in the sale of large-denomination certificates that have become a popular short-term investment for businesses with surplus cash, have been quite successful in marketing small-denomination certificates and passbook-notice accounts to the public. From one point of view, further realloca tion of personal savings to time deposits could occur, if time deposits continue to be attractive. Nationally, time deposits at banks still constitute a relatively small share of the total financial assets of consumers—in 144 1968, less than 10 per cent of the total. Thus, even relatively small shifts of funds from other assets into time deposits would enable the latter to increase at a very rapid rate, and this process conceivably could continue for many years. However, there is a second and prob ably dominant consideration that militates against such expectations. The rate of time deposit growth is obviously an important influence on the rate at which total bank credit can expand— and expansion of total bank credit will continue to be greatly in fluenced by national economic policies seek ing full employment without price inflation. Thus, policies of the Federal Reserve Sys tem can be expected to result in rates of bank credit expansion consistent with po tential real economic growth, while taking into account trends toward greater relative use of credit. In this environment, total de posit growth at banks will at times be en couraged and at other times restrained, as appropriate in the light of national economic goals and current business conditions. The relative attractiveness of time deposit rates is likely to be regulated accordingly. On average, an annual increase of 6.5 per cent in total deposits may represent a reasonable projection. This reasoning implies that a slowdown from recent rates of time deposit growth must be projected, with the estimated future gains dependent on the growth actually ex perienced in demand deposits. If the latter do increase at an average rate of 3 per cent annually at rural banks covered by the USDA index, the following annual rates of time deposit growth at these banks would yield the projected 6.5 per cent increase in total deposits in the 1970’s: 11 per cent in 1969, 10 per cent between 1970 and 1974, and 9 per cent between 1975 and 1979. Projected deposit and farm loan growth com pared, 1970-79. If farm credit demands on banks were to increase at the same rate as deposits, it may be assumed that on aver age banks would be able to meet these de mands without excessive difficulty; that is, on the average, banks could increase farm loans at this rate without (1) increasing the proportion of farm loans in their loan port folios, or (2) increasing their over-all loan/ deposit ratio in order to make the addi tional farm loans. Farm loan expansion at a TABLE 21 PROJECTED FARM LOAN EXPANSION AT BANKS COMPARED WITH BANKS' INTERNAL RESOURCE GROWTH, 1970-79 In billions of dollars unless otherwise indicated Banks’ percentage share of total farm loan expansion if bank expansion is— Increase in outstanding loans Projection, and period Needed to maintain banks’ relative role in farm lending Supported by annual deposit growth of 6.5 per cent Excess of needed amount over expansion supported by internal growth Needed to maintain banks’ relative role in farm lending Excess of needed amount over expansion supported by internal growth Enough to maintain relative role 6.5 per cent annually Annual average 5-year total Model HT 1 9 7 0 -7 4 ... 1 9 7 5 -7 9 ... Model B 1 9 7 0 -7 4 ... 1 9 7 5 -7 9 ... M odel HM 1 9 7 0 -7 4 ... 1 9 7 5 -7 9 ... Supported by annual deposit growth of 6.5 per cent 5.4 6.3 5.4 7.3 .1 —1.1 1.1 1.3 1.1 1.5 “ .2 24 21 23 25 4.4 3.5 5.4 7.3 -.9 —3.8 .9 .7 1.1 1.5 -.2 —.8 24 21 29 43 9.8 12.8 5.4 7.3 4.5 5.5 2.0 2.6 1.1 1.5 .9 1.1 29 29 16 16 C A PITA L A N D C R E D IT R E Q U IR E M E N T S OF A G R IC U LT U R E yearly rate of 6.5 per cent during the 1970’s would increase outstanding farm loans at banks by about $5.4 billion in 1970-74 and $7.3 billion in 1975-79. Table 21 shows how these amounts compare with ex pansion that would have to take place under the various capital models in order that banks maintain their relative share of farm lending. In only one of the model situations would this rate of increase in farm credit avail ability at banks exceed the credit demand projected. This result occurs in Model B, which combined a stable real stock of farm machinery with a moderate 3 per cent yearly increase in land prices. In the other situations projected, in which one of these capital projections is different— either real machinery stocks in crease (Model HM) or land prices rise faster (Model H T )— expansion of deposits and loans at 6.5 per cent yearly would at best just permit banks to maintain their one-third share of total bank and money market lending (Model HT) or funds would fall substantially short of this goal (Model 145 H M ). As these capital trends may easily continue, the projections on balance point toward probable difficulty for banks as they try to meet the farm loan demands of their present customers. If, as in Model HT, pro jected national supply and demand for bank credit are roughly in balance, a significant proportion of banks can still be expected to be out of balance because of differing local conditions. In the situation projected by Model HM, a majority of rural banks would experience farm lending difficulty. On balance, the analysis and projections indicate a fairly high probability that rural banks may be unable to maintain their usual share of farm lending on the basis of growth in their deposits. True, low farm capital requirements would reduce credit demands to a rate that could be met by probable de posit growth. Or a continued very high rate of increase in time deposits would enable higher credit demands to be met. But more probable events appear likely to result in a shortage of internally originated loanable funds relative to farm credit demands on rural banks. VI. SEASONAL PRODUCTION CREDIT In addition to using credit to maintain, add to, and transfer capital assets, farmers de mand seasonal credit to carry on produc tion processes not financed from their stock of cash and liquid assets. Seasonal credit extensions particularly require recognition here because the preceding analysis em ployed capital stocks and debt measured as of January 1, whereas the national seasonal peak in the demand for farm working capital occurs in the spring and summer. Unfortunately, seasonality of total ex penses and working capital has not been measured. Only annual estimates of farm expenditures and capital are available. Data on seasonal credit extensions are also incomplete. Such credit is provided mainly by three lender groups: merchants and dealers, commercial banks, and produc tion credit associations (PCA ’s). Advances and outstanding debt at PCA’s are reported monthly, but PCA’s hold only about onesixth of total non-real-estate debt. Loans from commercial banks, which represent two-fifths of the total, have been reported only semiannually in recent years. Debt out standing at merchants, dealers, and other individuals is estimated only as of January 1. Although the magnitude of seasonal capi tal and credit requirements is not known, 146 the rate at which such demands have been expanding can be estimated with the help of some plausible assumptions. For instance, the growth rate of operating expenses that clearly have a significant seasonal com ponent can be computed. If the relative seasonality of these expenses is assumed to have remained roughly unchanged, that growth rate becomes an estimate of the rate at which seasonal capital requirements have been rising. On the credit extension side of the puz zle, the January-July variation in loans out standing at PCA ’s and banks, after adjust ment for trend, can serve as an index of seasonal credit extensions by these lenders. The assumption here is that a change in the amount of seasonal lending would change the difference between January and July outstanding loans by about the same propor tion. Thus, the growth rate of that difference becomes an estimate of the growth rate of total seasonal lending by these institutions. The validity of this estimate may be helped by the fact that January is the low month and July the high month in outstanding farm debt nationally, as indicated by data for PCA’s. Seasonal capital requirem ents Operating expenditures with major seasonal elements include purchases of seed, fertil izer, and lime; operation and repair of motor vehicles and machinery; and wages and per quisites paid to hired workers who do not reside on the farm by which they are em ployed. These expenses are tabulated in Table 22. Purchases of feed and livestock are omitted from this table both because additions to these inventories have been in cluded in capital requirements previously considered and because the national sea sonal peak in credit extended for these items likely occurs near the January 1 date on T ABLE 22 SELECTED CURRENT FARM OPERATING EXPENSES In per cent unless otherwise indicated Item and period Expenditures in 1968 (millions of dollars) ......... Total Seed pur chases Operation and Wages to Fertilizer repair non of and lime vehicles resident workers and machinery 8,788 668 2,095 3,916 2,109 Annual rate .. 38.2 3.3 28.3 2.5 74.2 5.7 19.8 1.8 53.3 4.4 Average annual rate of change in specified period: 1950-54 ......... 1955-59 ......... 1960-64 1965-68 ......... 3.2 3.4 2.2 4.6 - .7 —1.4 2.9 4.2 6.2 1.3 5.7 5.3 4.1 3.7 —.3 4.5 .1 6.6 4.0 4.0 Increase from 1956-58 to 1966-68: S o u r c e . — F arm In c o m e S itu a tio n , U SD A , July 1969, pp. 56, 58, and 59. which capital stocks and debt were meas ured for the preceding analysis. Repairs and maintenance of buildings and land im provements are omitted on the conjecture that they did not have a strong seasonal element. For the same reason, wages paid to hired workers who reside on the farm are also excluded. The selected expenses totaled $8.8 billion in 1968 and over the previous decade had risen at an average annual rate of about 3.3 per cent. An exponential least-squares trend for 1950-68 rises by 3.1 per cent yearly. Data for each component, as shown in Table 22, reveal that these longer-term averages are depressed by the relative stabil ity in vehicle and machinery expenses during the early 1960’s. With such expenses ad vancing more rapidly in the last few years, the total selected costs rose by 4.6 per cent annually during 1965-68. If the degree of seasonality in these ex penditures has not changed in recent years, the seasonal capital requirement that under lies demand for seasonal production credit has also been advancing at these rates— less than half as fast as total outstanding farm debt. However, seasonal credit demands CAPITAL AND CREDIT REQUIREMENTS OF AGRICULTURE may also have increased because farmers wanted to finance a higher proportion of their seasonal costs by borrowing. Further evidence is provided by examination of the seasonality in non-real-estate debt. 147 found more useful in showing average an nual growth rates, along with the longerterm changes presented in the last section of Table 24. TABLE 24 Seasonal credit extensions Semiannual variation in outstanding nonreal-estate farm debt held by reporting lend ing institutions was calculated by averaging the amounts outstanding at the beginning and end of each year and subtracting this average from the amount outstanding on July 1, with results as shown in Table 23. The seasonal credit increase thus obtained has an upward trend both in total and at banks and PCA’s separately; however, the year-to-year fluctuations have been so large that the trend is not properly revealed by use of the 5-year periods that this study has employed in analysis of other data. Ex ponential trend curves fitted by the leastsquares technique to data for 1950-68 were TABLE 23 SEMIANNUAL VARIATION IN INSTITUTIONAL NON-REAL-ESTATE DEBT OWED BY FARMERS, 1949-68 Debt on July 1 exceeded average of debt at beginning and end of year by— Year Millions of dollars Per cent Total i Bank PCA Total Bank PCA 1 9 4 9 .... 1 9 5 0 .... 1951___ 1952___ 1 9 5 3 .... 437 269 464 680 634 271 127 247 418 400 146 108 171 211 191 11.6 8.7 12.5 16.4 15.9 13.6 5.6 8.8 13.2 13.4 38.7 25.8 33.8 36.4 33.5 1 9 5 4 .... 1 9 5 5 .... 1 9 5 6 .... 1957___ 1 9 5 8 .... 609 664 640 473 530 354 385 363 222 255 171 185 183 161 178 15.8 15.8 14.4 10.0 9.9 12.4 12.3 11.0 6.4 6.6 30.5 30.3 27.2 20.3 17.8 1 9 5 9 .... I 9 6 0 .. .. 1 9 6 1 .... 1 9 6 2 ... . 1 9 6 3 .... 686 650 626 627 872 336 329 311 278 425 262 241 244 243 313 11.0 9.5 8.6 7.8 9.7 7.5 6.7 6.0 4.9 6.7 21.2 17.0 15.6 14.0 15.9 1 9 6 4 .... 1 9 6 5 .... 1 9 6 6 .... 1 9 6 7 .... 1 9 6 8 .... 956 781 855 940 1,030 500 326 385 433 485 324 299 324 356 393 9.8 7.4 7.3 7.2 7.3 7.3 4.4 4.8 4.9 5.1 14.8 12.3 11.6 10.9 10.7 i Total non-real-estate loans to farmers held by all operating banks, production credit associations, Federal intermediate credit banks, and the Farmers Home Administration, excluding loans guaranteed by the CCC. S o u r c e . — A g ricu ltu ra l F in an ce R e v ie w , U SD A , Oct. 1957, pp. 30 and 31; Apr. 1969, pp. 22 and 23. SELECTED CURRENT EXPENSES AND SEASONAL COMPONENT OF DEBT: AMOUNT AND RELATIVE CHANGE, 1950-68 In uer cent unless otherwise indicated Period Selected current expenses Seasonal component of institu tional non-real-estate debt Total Bank PCA Annual average (millions of dollars) 1950-54 1955-59 1960-64 1965-68 5,332 6,182 6,998 8,328 531 599 746 902 309 312 369 407 170 194 273 343 Change from previous period 1950-54 1955-59 1960-64 1965-68 15.9 13.2 19.0 12.8 12.5 20.9 1.0 18.3 10.3 14.1 40.7 25.6 Change from specified period to 1965-68 1950-54 1955-59 . 1960-64.. .. 56.2 34.7 19.0 69.9 50.6 20.9 31.7 30.4 10.3 101.8 76.8 25.6 The estimated seasonal component of total institutional non-real-estate debt reached $1,030 million in 1968 and ex hibited an average annual growth of 4.4 per cent over 1950-68. The upward trend thus exceeded the annual average growth of 3.1 per cent estimated for seasonal production expenses. The excess may be explained by a seasonal element in intermediate-term credit that has been captured in the cal culated seasonal component of debt— farm ers tend to buy machinery in the spring, and debt incurred for this purpose would be paid down somewhat by year-end. However, farmers may also have been increasing the proportion of their seasonal production ex penses financed by debt. The increase in the seasonal component of outstanding debt fell far short of keeping pace with the rise in total non-real-estate debt at institutional lenders. In the late 1950’s debt in July was about 10 per cent higher than the January-December average, 148 but by the late 1960’s the average difference was reduced to 7 per cent. Given the expectation that farm produc tion will continue to rise at the modest rates that prim arily reflect domestic population growth and higher consumer living levels, seasonal farm operating costs can reason ably be expected to continue to increase at rates approxim ating those of the recent past, in which the same influences were dominant. The historical record indicates that seasonal credit demands can be expected to reflect this increase, and thus to rise by perhaps 3 to 5 per cent annually. A t this rate, the trend-adjusted January-July increase in non-real-estate farm debt might range from $1.3 billion to $1.7 billion around 1980. But if total non-real-estate debt should meanwhile increase in line with projections for total farm debt, this seasonal fluctuation would represent only about 4 to 5 per cent of outstandings. This analysis and projection indicates that provision of quantities of seasonal credit desired by farmers as a whole will not be a m ajor or growing problem. However, this prognosis is not likely to apply to each farm ing region. As will be shown, there is great regional variation in relative seasonal farm credit demands, and in some regions the seasonal factor can only be classified as huge. In these areas, any shortfall in total credit supply is synonymous with a short age of seasonal credit. Also, changes in re gional production patterns— particularly increasing specialization in a seasonal com modity— will continue to place at least tem porary strains on seasonal credit re sources in some areas from time to time. Institutional sources of seasonal credit As estimated seasonal lending at banks and PC A ’s together has recently been rising somewhat faster than estimated seasonal capital requirements, one might reason that these institutional lenders have been respon sive to farm ers’ seasonal demands. However, the increased credit demands have been met more vigorously by PC A ’s than banks. The large year-to-year fluctuations in the cal culated seasonal component, particularly in the bank debt series, precludes explicit judg ments, but it appears that during the last two decades P C A ’s may have provided ad ditional seasonal credit that surpassed the volume supplied by banks, in spite of the P C A ’s lesser role in total non-real-estate lending (Table 2 4 ). The 1950-68 leastsquares trend shows that the seasonal com ponent of non-real-estate loans at banks rose at an average annual rate of only 3.0 per cent, whereas that at PC A ’s rose by 5.5 per cent annually. A t both PC A ’s and banks, seasonal credit extensions have become a substantially smaller proportion of outstanding credit. The semiannual variation at P C A ’s fell from 18 per cent of outstanding loans in 1958 to 11 per cent in 1968, while at banks it de creased from 7 per cent to 5 per cent (T able 2 3 ). As already noted, however, the am ount of seasonal funds provided by both lenders actually increased and the seasonal role of P C A ’s rose in relation to that of banks. The ratios give the wrong impression because total farm lending (1 ) increased greatly at both lenders and (2 ) grew faster at P C A ’s than at banks. PC A credit exhibits greater seasonality than bank non-real-estate credit in all m ajor production areas except the A ppalachian and Southeastern States (Table 2 5 ). The same relationship is found in m any im por tant farm States— in 1968, P C A ’s showred larger relative seasonal variation in 18 of the 29 States in which bank non-real-estate farm loans exceeded $100 million, while in 1966 the proportion was 20 of 28 States. CAPITAL AND CREDIT R E Q U IR E M EN T S OF AG RICU LTURE TABLE 25 SEMIANNUAL VARIATION IN BANK AND PCA NON-REAL-ESTATE DEBT OWED BY FARMERS, 1968 By farm production areas Debt on July 1 exceeded average of debt at beginning and end of year by— Per cent United States ................. Northeast ...................... Lake States ................. Corn Belt ...................... Northern Plains ......... Appalachian ............... Southeast ...................... Delta States . .............. Southern Plains ......... Mountain ...................... Pacific .......................... Millions of dollars Bank PCA Bank PCA 5.1 .8 4.3 1.4 —.3 14.6 9.9 26.3 4.1 6.4 12.8 10.7 —1.5 4.9 7.0 3.3 9.3 6.0 53.8 13.4 13.7 16.3 485 3 37 35 4 66 29 74 40 65 140 393 4 20 51 13 39 20 117 40 50 40 S o u r c e . — A g ricu ltu ra l F in an ce R e v ie w , U SD A , Apr. 1969, pp. 24 and 25. 149 Seasonality in farm loans is greatest in Southern, Plains, and W estern States. In the Mississippi River Delta States, for example, the semiannual variation in 1968 was 26 per cent of outstanding loans at banks and 54 per cent at P C A ’s. These regional data indicate the con tinued great im portance of seasonal credit in some farming areas. From the historical trends previously noted, it appears that PC A ’s have been better able than banks to meet seasonal credit demands in areas where such demands have been large and increasing. Part 2. PROPOSALS TO INCREASE AVAILABILITY OF BANK CREDIT Part 1 outlined two important projections: (1) the agricultural sector is likely to con tinue to seek significantly larger amounts of credit, and (2) the ability of rural banks to provide their historical share of such in creases and adequately finance their com munities from their own resources is likely to be further impaired. To maintain their role as a leading farm lender, commercial banks will therefore in creasingly have to assume the role of inter mediaries who channel nonlocal— urban and money market— capital into agricul tural loans of either a term or a seasonal character. The avenues for securing such funds fall into two general categories: (1) discount or sale of assets— in particular, of loans; and (2) borrowing, such as by pur chase of Federal funds or sale of time cer tificates of deposit. These avenues for obtaining reserves to support additional lending have been partly or totally closed to many banks that are extensively engaged in rural lending. Whereas secondary markets have been de veloped for some bank paper, such as ac ceptances and mortgages, there is virtually no market for many rural loans. Thus, most banks that make a high percentage of their loans to agriculture and for other rural pur poses are unable to obtain any significant volume of reserves through rediscount or sale of notes. Their volume of lending is re duced by this imperfection in financial markets. Borrowing— the second route by which additional reserves can be secured— has been employed in significant proportions by large banks but is much less available to small institutions. Present markets in such instruments as Federal funds, time certifi cates of deposit, and Euro-dollars are largely designed to meet efficiently the needs of large banks. Thus small banks, including most banks engaged primarily in rural lend ing, are in many cases virtually precluded from participation or can participate only as effectively as the interest of their city cor respondent permits. The ability of small or isolated banks to employ these sources of funds is further restricted by lack of man agerial skills in this area, lack or relatively high cost of market information, and the relative lack of geographical and economic diversification of their resources, which out side investors tend to view as prima-facie evidence of higher risk. These imperfections in financial markets prevent an optimum allocation of money market resources, with attendant social cost. Economic sectors that must deal with the disadvantaged banks— industries such as agriculture, with large numbers of small firms located in isolated areas— are placed at a relative disadvantage in obtaining funds to finance expansion, new technology, or seasonal production processes. In the next three sections, imperfections in specific banking mechanisms and finan cial markets are considered in somewhat 150 CAPITAL AND CREDIT REQUIREMENTS OF AGRICULTURE greater detail to determine whether the flow of funds into rural areas is obstructed. Some thoughts are offered on how procedures might be changed or new mechanisms estab lished to improve the mobility of funds and thereby to increase the potential lending ability of rural banks. First, the efficacy of correspondent bank ing— the present mechanism by which funds are moved within the commercial banking system— is reviewed and appraised. Evi dence from Federal Reserve studies is presented to question whether correspond ent banking makes a net contribution to the flow of funds into rural areas, and to examine whether its contribution could be improved by altering the manner in which most correspondents are compensated for their services. To complement the evaluation of cor respondent banking, which is most highly developed in States with unit banking, the rural lending performance of branch bank ing is also examined. While the conversion of small rural banks into arms of larger in stitutions is shown to present important advantages, several offsetting conditions and circumstances are also noted. The second question raised is whether 151 present central banking mechanisms— the open market and discount operations of the Federal Reserve System— succeed in pro viding an equitable proportion of reserves to rural sectors. An optimum distribution of reserves provided either for seasonal fluctua tions or for long-term growth is thought un likely given the present state of financial markets, with funds reaching small and iso lated banks after considerable lag, if at all. Therefore, a number of suggestions for im proving present markets or compensating for their deficiencies are offered. Third, as a fundamental means for mov ing money market funds into rural lending via the banking system, development of sec ondary markets in rural bank portfolio items is proposed. Organizational considerations and operational methods are briefly out lined for new regional agencies— unified markets— that would make trading in such items feasible by neutralizing certain dis advantages that rural banks face in current money markets. Regional unified markets would provide rural banks with information and trading facilities for all financial instru ments, and thereby place rural banks on a more nearly equal footing with other institu tions in the financial arena of the Nation. VII. CORRESPONDENT AND BRANCH BANKING In the context of this study, correspondent and branch banking constitute existing mechanisms by which the advantages of large banks can potentially be enjoyed in rural areas. These mechanisms are in herently capable of improving the flow of funds, both between the money market and rural banks and among rural banks. In this process the net flow of funds can be either into or away from rural areas. Unfortu nately, national evidence on the net effect is meager. Relevant considerations and some recent research findings are discussed here. One other aspect of correspondent and branch banking deserves special mention: among the mechanisms considered, these arrangements alone provide a means to cope with the problem of farm loans that exceed the legal lending limits of rural banks. C orrespondent banking The correspondent-banking mechanism helps to provide more effective financial serv 152 ices to rural areas. City banks may handle overline loans, provide seasonal credit, ad vise on investment policies, help with ac counting and management problems, execute security transactions, and clear checks. In exchange for such services, country banks maintain deposits with their city correspond ents. This traditional method of payment drains funds from rural areas by tending to offset, or in many cases exceed, funds pro vided through credit services. Our interest in correspondent banking centers on the effectiveness and cost of the credit services rendered to rural areas. In sofar as available information and data permit, the following questions will be in vestigated: (1) Are the credit services re sponsive to rural needs? (2) What is the ratio between funds provided to and drawn from rural areas? (3) Should credit services be paid for by deposit balances? Correspondent-credit services. Intensive in terviews about correspondent relationships were conducted with a number of rural banks in 1966.35 These banks regarded the handling of overline loans as the most im portant credit service rendered by their city correspondents, and nearly all were using or had used their correspondents in this way. In a few instances, a customer had been re ferred to a correspondent, and in some cases the banks had also obtained loans from the city banks. But on the whole, most of the credit was obtained in the form of participations in loans originated by the country bank, and most participations were sought because the loan exceeded the legal lending limit of the country bank. Two earlier but broader surveys confirm this finding. A national survey of correspond ent banking conducted in 1963 found that 35 A total of 29 country bankers in Iowa, Illinois, Colorado, Kansas, and O klahom a were interviewed by personnel of the Federal Reserve Banks of C hi cago and Kansas City. the bulk of correspondent credit was pro vided through participation loans.36 In the 1963 midyear farm credit survey by The American Bankers Association, 84 per cent of country bankers and 74 per cent of city bankers rated participation in overline loans as the most important correspondent bank ing service to agriculture.37 Participations are also used to obtain correspondent credit even when overlines are not involved. For example, one banker has described how his bank obtains sea sonal credit through sale of participation certificates in a block of farm loans. This procedure was found more convenient than the sale of individual notes.38 Extent of overline loan problem. Legal lend ing limits fix the maximum outstanding credit that a bank may extend to an in dividual and are intended to avoid serious financial difficulty should one borrower de fault. For national banks, the legal limit is 10 per cent of the bank’s capital and sur plus, except that loans secured by livestock may go to 25 per cent. Laws governing banks chartered by State governments gen erally also impose lending limits based on similar criteria, although they vary among States. Rapid postwar growth in the size of in dividual farms has resulted in numerous farm loan requests that exceed the legal lending limit of the bank at which they are made. The market value of average assets per farm in the United States, for instance, more than doubled both in 1946-56 and 36 U.S. Congress, House, Subcommittee on D om es tic Finance of the Comm ittee on Banking and C ur rency, A R ep o rt on the C orrespondent Banking System , 88th Cong., 2d Sess., 1964, pp. 2-4. 37 T. P. Axton, “Introductory Rem arks,” C or respondent Agribanking: P roceedings o f the C or respondent A griban kin g Forum (New Y ork: The A m erican Bankers Association, 1963), pp. 5 and 6. 38 R obert L. W alton, “Overcoming Pressure of Seasonal Loan D em and,” A gricultu ral Banking and Finance (N ov.-D ee. 1967), pp. 28-31. CAPITAL AND CREDIT R E Q U IR E M EN T S OF AG RICU LTURE 1956-66. The average amount of credit used per farm more than tripled during each of these periods. The assets and capital of most rural banks did not grow so rapidly. Thus in many areas, a significant proportion of farms grew much faster than the banks by which they were being financed. Therefore, overline loan requests persist even though lending limits of rural banks have been moving upward. Between 1962 and 1968, ABA surveys found that the pro portion of agricultural banks with loan limits under $50,000 declined from 43 to 23 per cent and the proportion with limits of $150,000 and above increased from 25 to 37 per cent. However, the proportion of agricultural banks that received one or more overline requests in the first half of the year increased from about 25 per cent in 1962 to 29 per cent in 1968. In surveys made each year since 1962, this proportion has ranged between 26 and 34 per cent.39 Overline loans are particularly common at rural banks in Western States that pro hibit or severely restrict branch banking. Farms in these States tend to be large, and the rural banks tend to be small. In the 1968 ABA survey, 41 per cent of agricultural banks in the Plains States had received at least one excess farm loan application. The Federal Reserve’s 1966 survey of farm lending confirmed both the widespread occurrence and the geographical concen tration of overline farm loan requests. Of all insured commercial banks, 14 per cent 39 T rends in A gricultu ral Banking: R e p o rt of M id year 1968 A gricultu ral C redit Situation Survey (New Y ork: The A merican Bankers Association, 1968), pp. 11-13, and similar publications for earlier years. The ABA defines agricultural banks as banks with under $5 million in assets having 5 per cent or m ore of their assets outstanding in farm loans, and larger banks with 1 per cent or m ore of their assets in farm loans. In 1968, 840 of these banks participated in the sample survey. 153 had received at least one overline request during the 12 months ending in June 1966. Among banks with capital and surplus be low $300,000, one-fourth had received overline requests. Nationally, there were about 12,000 re quests totaling $330 million. They equaled 0.3 per cent of the number and 3 per cent of the volume of all farm loans outstanding on the day of the survey. At the banks that received the requests, however, the overline requests equaled 1.9 per cent of the number and 15 per cent of the dollar volume of outstanding farm loans. In both relative volume and number, overline requests were about five times more important at small than at large banks.40 In the Northern Plains States— Kansas, Nebraska, and the Dakotas— the dollar volume or overline requests received dur ing the year totaled 7 per cent of farm loans outstanding on the survey date, compared with the national average of 3 per cent. Overline requests also occurred with aboveaverage frequency in the Southern Plains and Mountain States.41 Responsiveness to rural credit needs. The ABA midyear survey of agricultural banks has consistently shown that a high percent age of overline requests has been handled through the correspondent system. In 1968, for instance, 86 per cent of the dollar vol ume of excess loan applications was han dled on a participation basis with correspond ent banks. Another 5 per cent was referred entirely to a correspondent, so that only 9 per cent was lost to other lenders. During 1962-68, the proportion of dollar value handled through the banking system has 40 Em anuel M elichar, “Bank Financing of A gri culture,” Federal Reserve Bulletin (June 1967), pp. 929 and 930. 41 Ibid., p. 943. 154 ranged from 88 to 97 per cent, and has most generally been at 94 to 95 per cent.42 The 1966 Federal Reserve survey did not ask about the disposition of overline requests, but did show a relationship be tween these requests and outstanding par ticipation loans. Of the banks with over line requests during the year ending in June 1966, one-half had at least one participation loan outstanding on June 30. At banks that had overline requests, participation loans represented a tenth of the outstanding farm loan business, about double the proportion found at all banks. Participation loans were relatively most important in those areas— the Plains and Mountain States— where overline requests were most frequent. The survey found $574 million of farm participation loans outstanding on June 30, 1966, of which the participating banks’ share was $304 million. Participation activ ity was widespread, as 2,500 banks had originated at least one of these outstanding loans, and 1,100 banks were participating in them. Since a similar survey in 1956, the number of originating and participating banks had tripled, and the dollar volume of participation credit had increased by 607 per cent.43 In general, the 29 rural bankers inter viewed in 1966 were pleased with their cor respondent relationships, which echoed at titudes generally expressed by bankers in the 1963 national survey of correspondent banking.44 Several bankers indicated that 42 A gricultu ral Banking D evelo p m en t, 196 2 -1 9 6 7 (New Y ork: The A m erican Bankers Association, 1967), pp. 13 and 14. 43 M elichar, op. c i t p. 937. For another discussion of overlines and participations, see “Lending Limits of Comm ercial Banks,” in Gene L. Swackhamer and Raym ond J. Doll, Financing M odern A gricultu re: Banking's P roblem s and Challenges (Federal Reserve Bank of Kansas City, 1969), pp. 40-53. 44 U.S. Congress, House, Subcommittee on Dom es tic Finance of the Comm ittee on Banking and C ur rency, C orrespondent R elations: A Survey o f B anker Opinion, 88th Cong., 2d Sess., 1964, pp. 6 and 7. their correspondent had always responded favorably when asked to take overlines. However, there were indications that correspondent-credit services may suffer as city banks reach less liquid positions. Several rural bankers had been told to hold credit extensions requiring overline participations to a minimum. Further questioning revealed that the banks that were asked to restrict credit in 1966— a year of general credit tightness— had generally tended to use their correspondents extensively, whereas those re porting no restrictions had never asked their correspondent to take more in overlines or other loans than the amount of their demand balance with the correspondent. The Federal Reserve’s 1966 national survey of farm lending provided additional insights into the resource pressures on city banks that might adversely affect their credit services to country banks. An estimated 83 per cent of participation funds came from banks with loan/deposit ratios of 60 per cent or higher. Also, 27 per cent of these funds were extended by banks that reported difficulty in financing their own farm cus tomers.45 Cost of correspondent-credit services. City correspondent banks are usually “paid” for their correspondent services by having the use of demand deposits that rural banks keep with them. This flow of funds from country to city counters the flow of cor respondent credit from city to country. How do the two flows compare in vol ume? Some indication is provided by com paring farm loan participations and demand balances outstanding on June 30, 1966. Be cause only data on farm loan participations were obtained in this survey, this compari son must be restricted to banks whose lend ing business consists primarily of loans to farmers. Banks with more than one-half of 45 Melichar, op. cit., pp. 940 and 941. 155 CAPITAL AND CREDIT REQUIREMENTS OF AGRICULTURE their total loans in loans to farmers were chosen. The analysis also concentrates on member banks, thereby avoiding the com plicating factor that nonmember banks hold balances in other banks to meet reserve re quirements. (Table 26 provides data for both member and nonmember banks.) At the 855 member banks meeting the farm loan criterion, farm participations averaged 22 per cent of demand balances with other banks, whereas at the 2,069 non member banks, the ratio was 16 per cent. Thus, the balances exceeded the credit re ceived by more than four times, which agrees with other impressions that the net flow of funds is from the country to the city. For the heavily agricultural banks in this analysis, a reasonable allowance for non farm participations and for nonparticipa tion credit would still leave correspondent balances far ahead of correspondent credit. A more detailed analysis of these data reveals wide variation in the ratio of par ticipations to balances among individual banks, indicating that managerial inertia at country banks may be an important factor contributing to the unfavorable direction of the net fund flow. If a country banker is operating at a low loan/deposit ratio, city banks can hardly be faulted for attempting to obtain rela tively large balances from him even though they are not called upon for proportionately large credit services. A distribution of the heavily agricultural member banks by loan/ deposit ratio is particularly revealing. Those banks with loan/deposit ratios under 50 per cent had outstanding participations averaging less than 10 per cent of the bal ances they held in correspondent banks. The ratio of participations to balances rose with higher loan/deposit ratios until it reached 97 per cent at banks with loan/ deposit ratios of 70 per cent and over. These banks kept correspondent balances averag ing only 5.4 per cent of their deposits, com pared with an average of 7.3 per cent for all the member banks in this analysis. A similar though less marked relationship was found at nonmember banks. It is evident that some banks can obtain a relatively high volume of correspondent credit relative to balances. A further break down of the data cited shows that the larger banks among those with high loan/deposit ratios had the higher ratio of participations TABLE 26 COMPARISON OF FARM LOAN PARTICIPATIONS RECEIVED FROM AND DEPOSIT BALANCES HELD IN OTHER BANKS, JUNE 30, 1966 Capital and liquidity status of bank M illions of dollars Farm loan participations as percentage of— Farm loan partici pations Bal ances 58 262 22 7 16 17 18 51 63 82 66 2 1 4 14 16 21 20 36 59 69 56 22 Balances as percentage of— M illions of dollars Deposits Farm loans Farm loan partici pations Bal ances 4.6 7.3 21 91 574 16 13 26 21 27 3.1 5.7 4.4 4.8 8.5 7.5 7.3 6.4 24 22 21 18 52 15 18 6 264 130 123 58 8 3 7 20 29 97 5.9 1.1 1.8 3.9 4.7 10.4 11.5 7.9 7.6 7.3 6.7 5.4 72 33 26 20 16 11 1 1 11 10 44 23 31 80 126 161 116 61 Bal ances Farm loans Member banks Total .......................... Capital and surplus (thousands of d o lla rs): Under 200 . 200-299 ................. 300-499 ................. 500 and over Loan/deposit ratio (per c e n t): Under 3 0 ............... 30-39 ...................... 40-49 ...................... 50-59 ...................... 60-69 ...................... 70 and over ......... Farm loan participations as percentage of— Bal ances Deposits Farm loans 4.7 10.7 30 20 12 15 10 6.3 3.3 4.1 3.0 11.5 9.9 10.3 10.7 32 35 28 29 4 1 9 6 38 38 3.9 .4 3.1 1.8 9.0 7.4 15.2 11.7 10.9 10.3 10.2 10.3 89 49 36 28 24 19 Nonmember banks N ote.— D ata shown are for banks at which farm loans comprised 50 per cent or more of total loans. Farm loans Balances as percentage of— 156 to balances (Table 27). Better managerial skills at larger banks may have played a part in this result, along with the greater power that larger banks presumably have to encourage their correspondents to pro vide more credit service. TABLE 27 FARM LOAN PARTICIPATIONS RECEIVED AS PERCENTAGE OF DEPOSIT BALANCES HELD IN OTHER BANKS, JUNE 30, 1966 L oan/deposit ratio (per cent) Under 50 ............................ 50-69 ..................................... 70 and over ........................ Capital and surplus (thousands of dollars) Under 200 200 to 499 500 and over 18 3 32 3 38 73 1 11 146 N ote.—D ata shown are for member banks at which farm loans comprised 50 per cent or more of total loans. To some extent, fund outflow from rural communities is accentuated by complemen tary deposit accounts that rural bankers keep with city banks that are not called upon for correspondent credit and that in many cases are rarely called upon for serv ices of any kind. Each of the 29 rural banks in the 1966 survey, for instance, was main taining accounts with from 2 to 12 city banks; however, in no case did a bank have more than three active accounts. Some of these banks did report that they were re ducing their number of inactive accounts. However, there is ample evidence that provision of correspondent credit is directly dependent upon maintenance of deposit balances and is related to the amount of such balances. For instance, one Illinois bank, in return for use of an anticipated $300,000 of seasonal participation credit (maximum $800,000), agreed to “keep with the correspondent an average of $150,000 in excess deposits above and beyond that needed to break even on a normal cor respondent relationship.”46 In a normal year, the city bank would apparently provide 46 Walton, op. cit., p. 30. $300,000 for perhaps 6 months, on which it would receive interest “at XA per cent above the prime rate.” In effect, the country bank gave the city bank a yearly average of $150,000 interest-free in return for the priv ilege of borrowing $300,000 at slightly above the prime rate for perhaps 6 months. On a yearly average basis, the city bank’s average commitment of its own funds was zero. If it could invest the balances at the prime rate, its annual earnings from the arrange ment were equal to the prime rate times $300,000. Nevertheless, this country banker was pleased with this correspondent arrange ment. With his bank fully invested, he found it necessary to pay this relatively high price for seasonal credit. City bankers traditionally have viewed deposit balances as additional compensation for provision of participation credit. For example, one banker with a large farm participation business, whose bank was therefore presumably offering participations on terms competitive with other city banks, made this statement in 1963:47 We will not accept an overline from a country bank unless we have a deposit relationship with that bank. W e expect some correlation between the amount of deposit relationship and the amount of overline accommodation. . . . We want the country banker to participate substantially in any loan he asks us to carry. Apparently, farm loans obtained through correspondents were not viewed as a suffi ciently profitable investment, even though the country banks were incurring most of the cost of originating the loans and were sharing the risks involved. Another city banker has more recently affirmed this view:48 47 M orris F. M iller, “O ur Bank’s A gricultural P ro gram ,” C orrespondent A gribanking (New Y ork: The Am erican Bankers Association, 1963), p. 36. 48 R obert E. H am ilton, “F arm Credit— It Should Be Supplied by Bankers, but A griculture M ust C om pete fo r Funds on the Same Basis as Any O ther In dustry,” M id-C ontinent B anker (Nov. 1968), p. 49. CAPITAL AND CREDIT REQ U IR E M EN T S OF AG RICU LTURE W here lies the glam our for M r. City Banker to send funds to F arm er Smith through his banker in w estern Illinois at the prim e rate, w ith no deposit balance? If, on the other hand, the same funds can be placed locally at the same rate with 20% com pensating balances and some good trust business in prospect, w here w ould your stockholders expect the money to go? Similar attitudes on the part of city banks are likely to become even more common as more city banks encounter tighter liquidity conditions, with inherent conflict between the loan demands of their own customers and the credit needs of their country cor respondents. If the latter needs are met, it seems evident that the correspondent system will exact a considerable toll for this service. A proposal to minimize drains on rural funds. Although the use of deposit balances to pay for correspondent services drains funds from rural areas, in a number of common circum stances this means of payment constitutes an efficient use of rural banking resources. For many nonmember banks, correspondent balances also serve to meet State reserve re quirements, and thus the funds would be unavailable for lending anyway. Also, at banks with funds in excess of loan demands, no immediate diminution of lending ca pability results when balances are used to pay for overline participations and other services. Or if a city bank is content to be paid for seasonal credit extensions by balances re ceived only in the rural bank’s off-season, and the rural bank would not be fully in vested at that time in any event, payment through balances may not adversely affect the local credit service of the rural bank. These circumstances are present in many correspondent relationships and may help to explain why a majority of country bankers surveyed in 1963 expressed a preference for balances over fees as means of payment for correspondent services.49 Also, it is natural 49 U.S. Congress, House, op. cit., pp. 10-12. C orrespondent R elations, 157 for these bankers to favor a traditional tech nique to which they are accustomed. But with the increasing shortage of loanable funds and with more sophisticated manage ment, more country banks may question the use of correspondent balances to pay for overline or seasonal participations. In many instances, community needs might be better served if rural banks made additional local loans with the funds they now are using to maintain correspondent balances and used the returns on these loans to pay for cor respondent services on a fee basis. Fees for credit accommodation should prove reason able, as interest rates charged by city banks on participations and other farm loans should be high enough to make them a profitable investment in their own right. If city correspondents are able to adapt to the changed liquidity position of rural banks, and at the same time maintain or ex pand credit services provided to rural areas, they will continue to constitute a useful farm credit mechanism. However, an in evitable conflict may arise as both country and city banks simultaneously approach less liquid positions. Thus at the same time that country banks are interested in reducing correspondent balances and increasing credit services, city banks probably have more in terest in increasing balances and less interest in providing credit. Given these circum stances, it seems unlikely that the traditional correspondent banking system can become the means whereby substantially larger quantities of urban funds are channeled into agriculture.50 However, correspondent bank ing could contribute materially to this flow if city banks prove interested in becoming 50 Further evidence and evaluation of credit flows through the correspondent-banking system have been provided by a Federal Reserve staff task force headed by Ernest Baughman and Dorothy Nichols, Federal Reserve Bank of Chicago. Some data and conclusions from this study are presented by Bernard Shull in R eport on Research U ndertaken in C onnection w ith a System Study, vol. 1 of this series, pp. 60-62. 158 brokers for funds that country banks need to maintain their role in rural finance. Branch banking In the heart of the Nation’s agricultural areas— the western Com Belt, Plains, and eastern Rocky Mountain States— branch banking is prohibited or severely restricted. One-half of the banking system’s farm loans are in this region. Farm loan demands in the area have increased rapidly as crop farm acreages have been enlarged and livestock production increased. In consequence, many agricultural banks have reached high loan/ deposit ratios and some have expressed con cern over their future capacity to finance agriculture. Also, large farming operations are common in this area, so that overline loan requests are frequently received. Con tinued availability of participation credit from larger correspondent banks is essen tial in this environment. In contrast, the 1966 survey of farm lend ing found that overline requests and general farm financing present markedly smaller problems in rural areas served by large branch-banking systems. The most striking evidence was contained in reports from the Pacific States, where very large farms pre dominate. But because the banks there were also relatively large, overline requests were virtually nonexistent, and few banks thought that farm loan demands pressed unduly against their resources. In addition to being less likely to receive farm loan requests exceeding their legal lending limit, large branch banks are po tentially able to improve rural credit serv ices in several other ways. Their greater lending volume can support employment of specialists in farming and farm lending. Their lending practices and terms can there fore stay abreast of modern developments. Over-all management of the bank’s re sources is also likely to be better than that achieved by many small banks. A typical branch system is likely to operate at a higher loan/deposit ratio than the average of an equivalent group of unit banks, partly be cause its geographically diversified lending reduces the over-all risk. More loans can therefore be made from the same banking resources. Within a branch system, funds can be shifted to offices facing the greater loan demand. Consequently, in communi ties where credit needs are greatest loans may easily exceed deposits. Finally, the larger banks are more likely to be able to tap national money markets for additional funds.51 On the other hand, there are reasons to question whether branch banking can be relied upon to improve the flow of bank credit to agriculture. To be most effective, a sys tem should cover an area sufficiently broad and diversified to include both capital sur plus and deficit regions between which funds can be moved. With branching limited to statewide systems at best, it is doubtful that this condition is met in some agricultural States. A more meaningful contribution could be expected if branching were per mitted over a larger economic area or were delineated by national or regional economic sectors rather than State lines. A second major concern is that the man agement of branch systems, because of un familiarity with rural finance, may not implement the policies that would lead to the potential lending improvements cited above. In a branch system covering a di versified area, rural lending may be a less 51 As p art of the over-all discount study, Federal Reserve staff studies of fund flows within branchbanking systems were undertaken by Verle Johnson, H arm on Haymes, and M argaret Beekel. A brief state m ent of the findings is presented by Bernard Shull, op. cit., p. 62. C A PITAL AN D C R E D IT R E Q U IR E M E N T S OF A G R IC U LT U R E important activity than in a unit bank in a rural community. Top managers will prop erly allocate less of their time to this phase of their bank’s business. Nevertheless, such a bank can perform an outstanding rural credit job if top management is interested in developing its rural business along with its other endeavors and employs capable technical staff to work in rural credits. But if such lending is neglected by a large branch system, many rural communities may be adversely affected. The potential 159 limitations should be recognized along with the advantages. Finally, a realistic appraisal must note that State legislation now prohibiting branch banking seems likely to be changed quite slowly, if at all. While a gradual nationwide trend toward reduction of restrictions can be detected, it has made little or no progress in many nonmetropolitan States. Meanwhile, it may prove desirable to implement other measures to improve the ability of the bank ing system to finance agriculture. VIII. FEDERAL RESERVE CREDIT Since the early 1930’s, the Federal Reserve System has relied mainly on open market operations to provide reserves to support a growing volume of money and bank credit, as well as to offset seasonal and other fluc tuations affecting bank reserves. Recently, only about 2 per cent on average of total Reserve Bank credit has been supplied through the discount window. Making Reserve Bank credit available through open market operations and letting market forces determine its allocation has much appeal, should distributive mechan isms in the financial markets enable a near optimum allocation to be achieved. Imper fection in present mechanisms, however, may lead to less-than-optimum distribution of new reserves among different economic and geographic sectors or may cause long lags before an optimum allocation is at tained. Empirical evidence tends to verify these fears.52 52 A fter empirical work th at included comparisons of portfolio behavior at reserve city and country banks, Goldfeld concluded th at in operations pro viding for reserve growth, “There is . . . no assurance that reserves generated by open-m arket purchases will find their way to banks in need of funds . . . openm arket operations are likely to affect country banks only indirectly” (p. 52). Similarly, in operations off- To compensate for inequities that thus arise from the present structure of banking and of financial markets, more Reserve Bank credit could be provided directly to rural banks. Through the discount mechan ism, member banks in need of reserves have had access to credit for short periods of time— too short in most cases to provide effective support for farm lending. Data on intrayear flows of loans and de posits show that many rural banks could each year use adjustment credit provided for the entire length of a farm production sea son. Such credit could be supplied through a discount mechanism redesigned to incor porate a seasonal borrowing privilege. The extent to which alternative types of such privileges might improve farm credit avail ability at banks is examined in this section. The analysis also reveals that more than half of member banks in relatively tight setting reserve losses, “there is no assurance th at the reserves created by open-m arket purchases will be distributed am ong m em ber banks in proportion to the reserve losses which they are intended to re place” (p. 183). A dditional evidence is presented on pp. 149 and 150. See Stephan M. Goldfeld, C o m m ercial Bank B ehavior and E conom ic A c tiv ity : A Structural Study of M on etary P olicy in the P ostw ar U n ited States (A m sterdam : N orth-H olland Publish ing Company, 1966). 160 liquidity positions face a year-round rather than seasonal strain on their lending re sources. Furthermore, many rural banks are not members of the Federal Reserve System, and if they do not belong to the System, they would not be eligible for sea sonal discount credit. Two possible Fed eral Reserve actions to assist both groups of banks are examined: (1) longer-term credit through the discount or open market mech anisms and (2) improvement of markets for assets and liabilities of rural banks. Seasonal discount credit Banks in nonmetropolitan areas frequently experience a seasonal squeeze on funds through simultaneous withdrawal of deposits and expansion of loan demands. Because of their small size and geographic isolation, rural banks frequently are ill equipped to utilize their resources effectively in a highly seasonal environment. During the off-peak season of the year, funds that might other wise have been committed to financing in termediate-term rural needs instead tend to be maintained in short-term Government securities, city bank accounts, or other forms that provide a high degree of liquidity, but that represent inefficient use of the financial resources of the community. Twenty, or per haps even 10, years ago, when most banks had ample stocks of liquid assets even at seasonal peaks, this situation caused little concern. Now, however, many banks are hard pressed to meet the loan demands of their area. With discount policy revised to allow rural banks to borrow a substantial portion of the funds required to meet sea sonal outflows, these banks would have more funds for meeting community needs and would be able to handle their invest ment portfolios more satisfactorily. The existing regulation permits exten sion of short-term discount credit for sea sonal requirements “. . . beyond those which can reasonably be met by use of the bank’s own resources.” This regulation has usually been interpreted to mean that a bank is ex pected to meet seasonal outflows of his torically average amplitude through its own portfolio adjustments. And when borrowing for seasonal needs has been permitted, the assistance has usually been of shorter term than the period of the bank’s need. A help ful revision of the rule would permit Federal Reserve Banks to establish seasonal borrow ing privileges for their member banks for meeting a portion of normal seasonal needs, with maturities geared to the length of need. This recommendation again seeks to remedy partially the inability of small and isolated rural banks to tap national money markets effectively for short- and intermediate-term funds. The following discussion demon strates the scope of the seasonal lending problem at such banks and the extent to which assistance through the discount win dow might be helpful. Example of seasonal fund flows at rural banks. Not all rural banks experience seasonal loan demand and deposit withdrawals that are large in relation to the size of the bank. But at some banks, principally those in crop production areas, such fund outflows can be violent. To illustrate a situation of this kind, actual data for three small banks in Nebraska, each with large seasonal flows relative to its size, were averaged to obtain data for a composite rural bank (Table 28). More than three-fourths of the loans at this bank consisted of loans to farmers, both in December and in June. But farm loans increased by 64 per cent between De cember 1965 and June 1966. In the same interval, deposits of individuals, partner ships, and corporations (IPC ) decreased by 17 per cent. Even after adjustment for an CA PITA L A N D C R E D IT R E Q U IR E M E N T S OF A G R IC U LT U R E Community consequences of large seasonal flows. The composite rural bank described TABLE 28 SEASONAL FUND FLOWS AT A COMPOSITE RURAL BANK 1 Amounts in thousands of dollars Amount outstanding Bank’s portfolio of — Farm loans ............ Total l o a n s .............. Deposits, IPC ........ Balances at other banks .................... U.S. Govt, securities Loans as per cent of total deposits . . . . June 1965 Dec. 1965 June 1966 1,867 2,414 2,684 1,295 1,664 3,549 2,124 2,485 2,952 324 687 555 1,847 398 611 77.8 40.4 74.9 Trend-adjusted January-June change as percentage of December total deposits 22 25 —23 -6 -3 8 i Average of data for three rural banks experiencing relatively large seasonal flows of funds. upward growth trend in deposits and loans, the combined semiannual fund outflow from increases in all loans and withdrawal of IPC deposits was equal to 48 per cent of the December level of total deposits. Other data in Table 28 indicate that as this bank received an inflow of deposits in the second half of the year, the money was placed primarily in U.S. Government securi ties and to a lesser extent was held as bal ances at other banks. In the spring, deposits flowed out of the community, and in addi tion farmers borrowed for production pur poses. To accommodate these seasonal de mands, the bank sold the securities that had been purchased the previous fall and also drew down its balances with other banks. The combined trend-adjusted semiannual change in these two assets totaled 44 per cent of total December deposits, almost equal to the relative semiannual outflow of 48 per cent. This bank thus financed the seasonal demands of its community from its own resources; that is, from the resources of the community. Funds deposited in the fall were merely stored in anticipation of the certain outflow of the following spring. The bank had a loan/deposit ratio of 75 per cent in June, at which time its resources were al most fully employed. In December, however, its loan/deposit ratio was only 40 per cent. 161 above had nearly a maximum year-round level of loans consistent with meeting the indicated seasonal outflow from its own re sources. If it had additional year-round loan demand, such as for farm machinery and equipment purchases or from nonfarm busi nesses, it could in theory operate in a dif ferent fashion and still meet the same sea sonal outflow: it could commit its own funds to the additional year-round loans and bor row an equivalent sum during the spring and summer to meet the seasonal demand. However, with the present structure of money markets, this could be difficult for a small bank in Nebraska. On the other hand, if this bank were able to obtain seasonal funds from its Reserve Bank in sufficient quantity to cover a signif icant portion of its seasonal outflow for the entire period of the outflow, it could operate in just that fashion. It could increase its year-round lending for legitimate commu nity needs with complete assurance that funds would be available for the vital sea sonal demands. In addition, this bank, in spite of the large seasonal outflow, is possibly not meet ing the full seasonal loan requirements of its customers. Faced with increasing demand for both year-round and seasonal credit, perhaps the latter is being curtailed instead of, or in addition to, the former. In this event, seasonal borrowing from the Federal Reserve Bank would enable the bank to meet more adequately the complete sea sonal needs of the community. After several years the real seasonal pattern would be evident, and the bank could obtain still greater seasonal sums from the Federal Re serve, thereby releasing the community’s own resources for additional year-round loans. 162 Community benefits from a seasonal dis count privilege can thus be expected in situ ations where banks (1) are experiencing a significant seasonal outflow of funds rela tive to their size, and (2) are operating at a relatively loaned-up position at the peak of the seasonal outflow. The latter condition indicates that term an d /o r seasonal loan de mands are not being fully met or that such a situation may soon develop. The following discussion attempts to measure the preva lence of these circumstances among rural member banks and to estimate the impact that seasonal discount arrangements might have on farm lending at these banks. Prevalence of large relative seasonal outflows. Fund flow data for all banks, similar to those shown for the composite rural bank, indicate that banks involved in financing agriculture to the extent of at least 25 per cent of their total loan volume (hereinafter referred to as agricultural banks) are more likely than other banks to have semiannual TABLE 29 DISTRIBUTION OF M EMBER BANKS BY RELATIVE SEMIANNUAL FUND OUTFLOW AND BY IMPORTANCE OF FARM LENDING, 1965-66 Relative semiannual fund outflow (per cent) Total All banks .............. Under 5 .............. 5 to 9 .................. 10 and o v e r ........ 6,151 3,398 1,784 969 Farm loans as percentage of total loans Under 1 1 to 24 25 and over Number of banks 1,388 867 344 177 2,713 1,665 786 262 2,050 866 654 530 Percentage distribution (by relative outflow) All b a n k s ................ Under 5 .............. 5 to 9 .................. 10 and o v e r ........ 100 55 29 16 All b a n k s ................ Under 5 .............. 5 to 9 .................. 10 and over ........ 100 100 100 100 100 62 25 13 100 61 29 10 100 42 32 26 (by farm loan ratio) 23 26 19 18 44 49 44 27 33 25 37 55 fund outflows. As Table 29 indicates, 26 per cent of agricultural banks experienced a semiannual fund outflow equal to at least one-tenth of deposit volume. Only 11 per cent of other banks had relative outflow of this magnitude. Also, an additional 32 per cent of agricultural banks had semiannual fund outflows of from 5 to 9 per cent of deposits, still a slightly higher proportion than found among other banks. As of June 1966, agricultural banks com prised one-third of all member banks. But of member banks with relative semiannual outflow equal to 10 per cent or more of deposits, 55 per cent were agricultural banks. Of banks at which outflow com prised 5 to 9 per cent of deposits, 37 per cent were agricultural banks. Thus, rela tively large seasonal fund outflows were more prevalent among agricultural banks— many of which were precisely the banks unable to cope with such seasonal flows ex cept by keeping their own resources avail able for this use. Potential impact of specific seasonal discount proposals. In any arrangement that permits banks to borrow from Federal Reserve Banks to meet part or all of seasonal out flows, specific rules would be needed to guide the definition and measurement of seasonal outflows, and to indicate the proportion of outflows that could be met by borrowing. Formulation and execution of such rules could, and undoubtedly would, employ more detailed banking data than the semiannual statistics shown thus far. The particular regulations adopted would influence the total amount of seasonal credit extended by the Federal Reserve System, as well as the amount that could be obtained by agricul tural banks. An indication of the proportion and amount of borrowing that might be done by agricultural banks, however, can be ob tained from the semiannual data that is now readily available for all banks. Suppose, therefore, that seasonal outflow is defined as in the example involving the composite rural bank and that banks are allowed to C A PITA L A N D C R E D IT R E Q U IR E M E N T S OF A G R IC U LT U R E borrow funds equal to all outflow exceeding either 5 or 10 per cent of average de posits. The extent to which agricultural banks could participate in seasonal borrow ing and the relationship between their po tential borrowings and their volume of farm lending can be calculated to show the poten tial impact of the seasonal discount credit on farm lending. Selected data of this kind are shown in Table 30. TABLE 30 ESTIMATED MAXIMUM SEASONAL BORROWING AT AGRICULTURAL AND OTHER BANKS UNDER ALTERNATIVE DISCOUNT PLANS, 1965-66 Seasonal discount plan (level of deductible) Total Farm loans as percentage of total loans Under 1 1 to 24 2^ d Number of banks eligible 10 per cent ........ 5 per c e n t ........ 969 2,753 177 521 262 1,048 530 1,184 Distribution of banks eligible (per cent) 10 per cent ........ 5 per cent ........ 100 100 18 19 27 38 55 43 Borrowings (millions of dollars) 10 per cent ........ 5 per cent ........ 461 2,130 157 1,022 185 801 120 307 Distribution of borrowings (per cent) 10 per cent ........ 5 per cent ........ 100 100 34 48 40 38 26 14 Borrowings as per cent of deposits at eligible banks 10 per cent ........ 5 per cent ........ 4.1 3.0 3.0 2.6 4.8 3.1 5.5 5.3 Borrowings as per cent of farm loans at eligible banks 10 per cent ........ 5 per cent ........ 59.4 87.3 97.0 82.4 20.5 21.9 Farm loans at eligible banks as per cent of farm loans at all insured banks 10 per cent ........ 5 per cent ........ 7.0 22.3 4.0 20.1 9.9 23.8 Borrowings as per cent of farm loans at all insured banks 10 per cent ........ 5 per cent ........ 4.2 19.5 3.8 16.6 2.0 5.2 Under the 10 per cent plan, 16 per cent of member banks could borrow, and 55 per cent of these would be agricultural banks. The 5 per cent plan would broaden potential borrowing to 45 per cent of mem 163 ber banks, of which 43 per cent would still be agricultural banks. Although dominant in numbers, the agri cultural borrowing banks would tend to be smaller than other borrowing banks. Thus, of total potential seasonal borrowings of $461 million under the 10 per cent plan, agricultural banks would obtain 26 per cent; under the 5 per cent plan they would obtain 14 per cent of total borrowings of $2,130 million. Although agricultural banks would get a smaller portion of the total credit extended under the latter plan, they would obtain a much larger sum than under the 10 per cent plan— $307 million versus only $120 million. Under either plan, however, the potential borrowings would have more impact on the agricultural banks than on the other banks, reflecting the fact that seasonal outflows are proportionately greater at agri cultural banks. Borrowings by agricultural banks under the 10 per cent plan could po tentially equal 5.5 per cent of deposits at the eligible banks, against only 3.0 per cent at eligible banks with few or no farm loans. The 5 per cent plan yields the same differ ence in potential impact. Thus, seasonal bor rowing arrangements would not only benefit a greater proportion of agricultural banks than other banks, but would also be of relatively greater importance to the agricul tural banks among the banks eligible to borrow. At the eligible agricultural banks, poten tial borrowings under either plan would equal about one-fifth of present farm loan volume. The proposal could thus have a significant impact on farm lending at these banks. The impact on total farm lending by all insured commercial banks would be much smaller, but still potentially significant, es pecially under the 5 per cent plan. Mem ber banks eligible to borrow under the 10 164 per cent plan hold 7.0 per cent of the total farm loans outstanding at all insured banks, whereas under the 5 per cent plan the proportion rises to 22.3 per cent. Mem ber agricultural banks eligible to borrow under the alternative plans hold 9.9 per cent and 23.8 per cent, respectively, of total farm loan volume at all insured agri cultural banks. Potential borrowings are equal to 2.0 per cent and 5.2 per cent, respectively, of the total farm loan volume at all insured agricultural banks. The po tential impact on total farm lending is re strained because (1) agricultural banks with large seasonal outflows tend to be small banks, and (2) two-thirds of all agricul tural banks, as well as of agricultural banks with large seasonal outflows, are nonmem ber banks that would not be eligible for seasonal discount credit from the Federal Reserve System unless they became mem bers, or unless a basic legislative change per mitted borrowing by nonmembers. Impact of bank liquidity on potential borrow ing. It seems reasonable that banks with little liquidity, particularly at the peak of seasonal outflows, would be most likely to utilize a seasonal borrowing arangement to advantage. The 1966 farm loan survey indi cated that banks began to experience sig nificantly increased difficulty in financing their farm borrowers when loan/deposit ratios exceeded 60 per cent. Table 31 in dicates that about 34 per cent of agricul tural banks eligible to borrow under the 10 per cent plan, and a slightly smaller portion of those eligible under the 5 per cent plan, were illiquid to this degree at their seasonal peak. Another 30 per cent had loan/deposit ratios in the 50 to 59 per cent range, indi cating that they might soon be more seri ously concerned with liquidity, and perhaps might already be able to benefit from some seasonal borrowing. At more than one-third of agricultural member banks with large relative seasonal outflows, however, loan/deposit ratios are apparently under 50 per cent— at some banks, under 40 per cent— even at the sea sonal peak. Although some of these banks might exercise a seasonal borrowing priv ilege and perhaps thereby improve their farm lending service, their present liquidity would permit them to do so even in the ab sence of such arrangements. An analysis performed by the Federal Reserve Bank of Kansas City showed that many such banks do not seem to lack farm lending opportuni ties in their communities. Many had neigh boring banks in the same or adjacent towns with much higher loan/deposit ratios, some of which were expressing concern about their inability to meet the legitimate loan demands of the area. Greater educational and other efforts to overcome the apparent managerial inertia at the banks with low loan/deposit ratios would be of service to the communities affected. TABLE 31 DISTRIBUTION OF AGRICULTURAL MEMBER BANKS WITH FUND OUTFLOW IN JANUARY-JUNE 1966 By relative size of outflow and by loan deposit ratio on June 30, 1966 Loan/deposit ratio (per cent) Relative fund outflow (per cent) Total Under 40 40-49 50-59 60-69 70 and over Total 1,663 602 568 493 304 118 102 84 376 143 138 95 472 156 170 146 40-49 50-59 60-69 70 and over Percentage distribution of banks Number of banks All banks .......................... Under 5 .......................... 5 to 9 .............................. 10 and over .................. U ^ er 355 125 116 114 156 60 42 54 100 100 100 100 18 20 18 17 23 24 24 19 28 26 30 30 21 21 20 23 9 10 7 11 CA PITA L A N D C R E D IT R E Q U IR E M E N T S OF A G R IC U LT U R E Latent seasonal demands. The preceding evidence on the potential assistance of sea sonal discount credit to farm lending neces sarily shows only how it could help banks to cope with the seasonal loan demands that they have actually filled in the past. It is probable, however, that some— perhaps many— banks have seasonal loan demands in their communities that have not been met because of lack of funds. Again, at progres sive banks this situation is likely to occur more often as liquidity is exhausted. Although banking statistics alone cannot reveal latent seasonal demands, some indica tions of their existence and relative signifi cance have already been noted in Section VI, where the trend in the amount of sea sonal credit provided by banks was com pared with trends in seasonal operating expenses (Table 22) and in seasonal credit provided by PCA’s (Table 24). Expenditures for current farm operating expenses with significant seasonal compo nents increased by 3.3 per cent annually during 1950-68. Over the same period, the semiannual variation in total institutional non-real-estate debt rose at an average yearly rate of 4.4 per cent. Although this comparison is far from complete, the evi dence is nevertheless consistent with the hypothesis that increased farm seasonal credit demands are being met by institu tional lenders. During the same period, however, the amount of the semiannual variation in bank non-real-estate loans rose at an average an nual rate of 3.0 per cent, whereas at PCA ’s the average annual gain was 5.5 per cent. PCA’s probably provided more additional seasonal credit than did banks over this period. In 1950-54, for instance, the aver age January-June loan increase at banks amounted to $309 million, whereas at PCA ’s the amount was $170 million. By 165 1965-68, the amount at banks had risen to $407 million, or by 32 per cent, whereas the amount at PCA’s had doubled to $343 million. These data are consistent with the hy pothesis that banks have encountered greater difficulty in financing seasonal credit demands of farmers, presumably because increased year-round loan demands have re duced liquidity from which seasonal de mands could be met. Increased seasonal de mands upon PCA’s were readily financed by short-term borrowings in the central money market, whereas rural banks could not easily tap this source for significant amounts. It is conceivable that some farm borrowers switched from banks to PCA ’s primarily because PCA ’s were more inclined to meet their seasonal requests— not be cause PCA’s had a more favorable attitude toward the wisdom of such borrowing, but simply because they were much better able to cope with these demands. A seasonal dis count arrangement for member banks would restore their ability to compete with PCA’s in seasonal lending. Supplemental adjustment credit. To encour age rural banks to take advantage of their eligibility for seasonal discount credit under any plan that is implemented, such a plan should clearly indicate that banks using seasonal credit remain equally eligible for additional short-term adjustment credit should circumstances make use of the latter advisable. Otherwise, at least until other financial mechanisms are improved, rural banks might be reluctant to make full use of the seasonal privilege for fear of unex pectedly finding themselves in an illiquid position. Rural banks on the whole have not made effective use of present short-term adjust ment credit available through the discount window. Clarification and simplification of 166 the terms of and Federal Reserve attitudes toward this privilege would be desirable to promote such use, as conditions that present no problem to sophisticated money market banks may have deterred many rural bank ers. Similarly, any new regulations to gov ern seasonal credit extensions should be comprehensible and suitable to the rural bankers that these arrangements are in tended to serve. Longer-term credit Many rural banks face year-round rather than only seasonal strains on their lending resources, according to loan/deposit ratios that have been examined (Table 31). Struc tural factors at present limit the access of these banks to financial markets in which larger and less isolated banks in the same circumstances are able to obtain funds by selling their assets and liabilities. To com pensate for the market imperfections, the Federal Reserve System conceivably could provide reserves directly to the rural banks on a long-term basis. One might propose that particular rural banks be allowed to borrow at the discount window for indefi nite periods, or that the Federal Reserve purchase certain assets or liabilities of these banks, such as farm loans, debentures se cured by farm loans, or certificates of de posit. However, in contrast to seasonal or other temporary assistance, provision of long term Federal Reserve credit directly to spe cific banks presents severe operational and conceptual difficulties. In principle, given the situation outlined above, the Federal Reserve could try to provide the quantities of funds that rural banks might obtain if they had better access to financial markets and could try to charge the rate of interest they would have to pay in the market. But what these quantities and rates might be and how they might be altered from time to time by changes in general monetary con ditions and other factors would not be easy to determine within acceptable limits. Nevertheless, if implemented, a program of direct compensatory assistance would undoubtedly improve the availability of bank credit to farmers, obviously a goal of its proponents. Paradoxically, this effect, al though it might be in the public interest, creates a fatal conceptual difficulty. As a principle of sound monetary policy, the Fed eral Reserve will not knowingly enter upon programs in which its credit-creating powers are used for the special benefit of a particu lar sector of the economy or in which it is called upon to allocate credit among specific uses. Through the years, Congress has rein forced this view of the proper role of the Federal Reserve by turning to or creating nonbank financial institutions to augment credit supplies for specific economic sectors judged to be in need thereof, rather than by asking the Federal Reserve to deliberately influence the allocation of credit to these uses. Because it would be difficult to deter mine the point at which compensation for market imperfections ends and favoritism toward farm credit begins, it is also difficult to visualize the Federal Reserve adopting a program of direct long-term assistance. This conclusion about direct long-term credit, however, does not negate the fact that a central bank can obtain equitable and satisfactory results in supplying reserves mainly through open market operations only if financial markets are well developed, as they generally are in the United States. Thus, the Federal Reserve has an implicit stake in the development and maintenance of financial markets that serve all sectors of the economy. It should work toward per fection of markets on which the fairness and success of its procedures depend, and it has C A PITA L A N D C R E D IT R E Q U IR E M E N T S OF A G R IC U LT U R E done so on numerous occasions. In the face of decreased rural bank liquidity, and given that unit banking is required in many pri marily agricultural States, the Federal Re serve should now undertake to secure im provement of secondary markets for assets and liabilities of rural banks, including agricultural paper and debentures secured by agricultural paper. The Federal Reserve has the knowledge and resources to take an active role in the development of secon dary markets such as those outlined in the next section. If necessary, for example, the Federal Open Market Committee could ex tend material support to an embryonic mar ket through a controlled volume of trading in its instruments, similar to the manner in which the FOMC has helped to establish the market for bankers’ acceptances.53 The 53 M uch of the Federal Reserve System’s rationale for purchases of bankers’ acceptances and its pro cedures and experience in this m arket appear trans ferable to the proposed dealings in rural bank paper or an instrum ent secured by such paper. F or instance, in describing operations in bankers’ acceptances, Roosa states “the Federal Open M arket Committee, in recognition of the potentialities fo r further use of 167 Federal Reserve System and its FOMC should not ignore the structural imperfec tions in financial markets and instruments that discriminate against small and isolated banks. bankers’ acceptances that m ay be inherent in the expanding role of the U nited States in financing world trade, and for other reasons, decided to re sume the acquisitions of a portfolio in bankers’ ac ceptances for the System itself. . . . The Federal Reserve has not, as a m atter of practice, sold accept ances out of its portfolio. . . . there are alm ost al ways some acceptances m aturing every day, and in a relatively short time m aturities alone can run the holdings down as far as might be appropriate in conform ing to the direction of other credit policy action. . . . the job of the acceptance clerks is . . . one of . . . verifying the negotiability . . . as well as inquiring, under some circumstances, into the credit standing of the business concern drawing the acceptance. Because of the nature of this paper, however, the principal reliance as to its soundness is placed upon the nam e of the accepting bank and the added endorsem ent which the acceptance carries. C ur rent lists are m aintained of all banks in the U nited States engaged in extending acceptance credits, and the condition of each such bank is periodically re viewed.” See R obert V. Roosa, F ederal R eserve O p erations in the M on ey and G overn m en t Securities M arkets (Federal Reserve Bank of New Y ork, 1956), pp. 87-90. Recent year-end Federal Reserve holdings of bankers’ acceptances have approached $200 mil lion. IX. UNIFIED MARKETS TO SERVE RURAL BANKS Unified markets— in which small and rural banks could obtain market information and conduct trading in a wide variety of port folio items in units that correspond to their needs— would improve the flow of funds to rural areas. These institutions could provide rural banks with both market information and trading facilities for purchases and sales of Federal funds and Government securi ties, placement and secondary marketing of certificates of deposit, and secondary mar keting of loan paper. With these services centralized in one location, rural bank man agers would have the market options now effectively available only to larger banks, as well as the information necessary for proper decisions— for example, whether to raise funds by selling bonds, discounting loan paper, or participating in a certificate of deposit issue. More transactions would become profitable— some are not now economical because of the small amounts involved and the numerous telephone calls to different markets required— and rural banks would have an enhanced ability to respond to changing loan demands and other conditions. Structural and operational aspects of a unified market are considered in this sec tion. First, some ideas for its basic organi 168 zation are advanced. Next, the approach the agency might use in providing a second ary market for rural bank paper is analyzed in some detail. Although the unified mar ket would be most effective if all major types of commercial bank loan paper were traded, only non-real-estate agricul tural loan paper is covered herein, because the primary concern is with availability of credit to farmers. However, much of the analysis also applies to trading in other types of paper. Finally, the prospective role of the unified market in trading in other instruments— Federal funds, certificates of deposit, and bonds— is briefly discussed.54 Organization The cardinal principle in organization of a unified market should be to provide rural bankers with a maximum amount of infor mation and service for a minimum of ex pense and effort on their part, just as present money markets are organized to invite and expedite trading by large institutions. Rea sonably convenient facilities, adequate capital, and a knowledgeable operating staff are essential, as are a competent research staff and appropriate facilities for gathering and disseminating information. To attain these goals in the most effec tive and efficient way, operations of regional unified markets should be coordinated and supervised by a national agency. Given present and foreseeable developments in communications and computer technology, a national network of unified markets can constitute a practical and desirable addition to the Nation’s financial mechanisms. 54 F or additional discussion of the unified m arket concept, see Raym ond J. Doll, “Unified M arkets for R ural Banks,” Banking, Journal o f the A m erican Bankers A ssociation (Jan. 1969), pp. 63-65; and “Unified M arkets to Facilitate Exchange of Bank Assets and Liabilities,” Bank N e w s M agazine (June 1969), pp. 13-18. A secondary m arket for rural bank loans Successful secondary markets for loans made by rural banks would materially in crease the banks’ ability to finance rural com munities. Development of such outlets would be a primary goal of unified markets. There are two basic ways in which a mar ket for such paper could be provided. First, the unified market could simply bring to gether buyers and sellers of the notes. Or the market, acting as an agency, could sell debentures and use the proceeds to purchase rural bank paper. By either method, if the market is effective, a bank that is loaned up could obtain funds by selling notes from its portfolio. It could then use these funds to make additional loans. Trading in loan paper. Direct sale of loan paper to investors avoids the more com plicated process of issuing debentures, with the market itself becoming directly in volved with questions of risk. However, the market for such paper might prove quite thin, as most of the notes are small and fre quently in odd amounts and maturities. But even if direct sales were restricted to the larger farm notes of borrowers for whom financial and credit ratings are readily avail able, significant sums might be obtained and rural banks would be especially encouraged to provide adequate financing for the larger farms and other firms located in their com munity. To increase the marketability of loan paper, the unified markets could provide or arrange for some form of insurance that would reduce or eliminate the risk of loss to the purchaser of an individual note (altern ative insurance plans are discussed later). In so doing, however, the markets would probably become involved in risk determina tion to about the same extent that they would if they had bought the paper themselves in a debenture operation. CA PITA L A ND C R E D IT R E Q U IR E M E N T S OF A G R IC U LT U R E Sale of debentures. The alternative method — sale of debentures secured by loans— would resemble the present operations of the cooperative farm credit system, par ticularly those of the Federal intermediate credit banks. These banks have been able to raise funds in national capital markets and use the proceeds to discount agricultural paper of the production credit associations. This process has proved efficient, and much of the experience would be transferable to the operations of unified markets. Also, favorable investor experience with these is sues should improve initial marketability of unified market debentures. The unified markets could logically use both approaches. They could act as direct brokers, where feasible, in bringing together buyers and sellers of rural bank paper, and in addition could issue debentures to raise funds for purchase of such paper from com mercial banks. These debentures should be joint obligations of all unified markets— with only the paper purchased by these uni fied markets being used as security. The primary advantage of debentures is that they would enjoy a much broader mar ket than individual notes because they would be issued in standard sizes, have more diversified security than individual notes, and would not require a new investigation by the potential investor for each purchase. Thus, they could undoubtedly be sold in larger volume and at lower interest rates than individual notes. Insurance mechanisms. The attitudes of bankers and bank examiners make it un likely that significant amounts of discount ing can be done if bankers must retain the risk on the paper sold. It seems desirable, therefore, that all sales be made on a non recourse basis, with controls established to prevent bankers from ignoring the credit risks. One such control is insurance. Sellers 169 could be required to buy insurance on each note sold to the markets, with the rate de pending on the note’s risk classification, but high enough to build up an adequate re serve.55 Such insurance could be funded by the markets themselves or handled by pri vate insurance companies. The markets could underwrite the insurance by acting through a central body to achieve geo graphical diversification. While the princi ple of insurance is applicable whether loan paper is traded or debentures are issued, in surance would be of particular benefit— or be virtually required— in the former case, where it would reduce risk differentials and greatly increase the probable number of market participants. Alternatively, the markets could provide for risk differences by varying the offering price according to the risk classification of each note. Prices could be adjusted so that, after allowance for probable losses, the rate of return on all notes would be the same. Over the long run, the price differences would exactly compensate for losses. Ad ministrative costs of insurance would be saved, but the risk classification process would entail some costs and difficulties. The insurance problem might be better handled by a third alternative, the establish ment of a reserve account for each bank. For instance, if a bank’s actual losses aver age 0.5 per cent, payment into its reserve account might proceed at the rate of 1 per cent of new loans sold until the reserve equaled 2.5 per cent of total loans sold and still outstanding. Payments into the account would then cease until there was either a net increase in the bank’s activity or a loss 55 Available evidence indicates th at default losses on bank agricultural loans average less than 0.5 per cent. Average insurance rates would, of course, have to be slightly higher to cover other insurance costs, although the insurance rate charged for high-quality loans might still be less than 0.5 per cent. 170 on one of its notes— in which case, they would be resumed at a rate of 1 per cent of new sales. The relative size of the reserve would be varied according to losses experi enced over an appropriate period. Losses larger than the reserve account would be borne by the markets, so the pro cedure would be equivalent to sales on a limited-liability basis. Bankers who sold very high-quality paper would be rewarded by very low insurance costs once the reserve was established. Among the possible dis advantages is the likelihood that some bank ers might be reluctant to change an estab lished volume of loan sales because this would require additional payments into their reserve accounts. The accounts also would require continual supervision, but total administrative costs might well be lower than in the preceding alternatives be cause individual notes would not have to be evaluated for risk. Education. To realize the full potential of secondary marketing of loan paper, a major educational program would initially be de sirable to demonstrate the need for and benefits of secondary markets to bankers and their customers. Such mutual under standing would help preclude damage to customer relationships when banks market loan paper. Unfortunately, the mere existence of a secondary market for rural bank paper would not eliminate the managerial inertia that exists in some rural banks. However, these banks would be placed under more pressure than at present, from their competitors and customers, to improve their credit services to their communities. Other services of unified m arkets Federal funds. Inclusion of Federal funds activity in the unified markets would assure rural banks of greater access to the funds market, particularly on the buying side. At present, participation by small banks is largely dependent on the willingness of city correspondent banks to act as brokers or dealers in Federal funds. Accommodation hinges on whether the correspondent has complementary reserve needs or can match the wishes of two country correspondents. Under other circumstances, correspondent banks appear much less willing to accommo date small transactions in Federal funds. By acting as a dealer, a unified market could give small banks access to the Federal funds market on a basis that is continuous, cer tain, and independent of a correspondent. Unified markets probably could provide most effective service in Federal funds by taking dealer positions. This operation would enable banks to accommodate trans actions of differing size and would allow them to offset net buying or selling by their customers through trading in the national market— in effect, by acting as wholesalers of Federal funds. In addition, a dealer op eration would stimulate trading because a selling bank would not have to concern it self with the solvency of a different small bank each time it sold or to establish re strictive lists of banks to which it would sell. The minimum trading unit needs to be relatively small if banks serving rural areas are to be able to participate effectively Also, small banks might arrange to have thv markets buy or sell funds for them for specified periods of time on some automatic basis. For example, a bank might place a standing order for purchases or sales when ever its excess reserves vary by one trading unit from a specified amount. Another more sophisticated approach would rely on daily computer analysis by the unified market of each bank’s reserve account, with decisions about whether and how much to trade being CAPITAL A N D C R E D IT R E Q U IR E M E N T S OF A G R IC U LT U R E based on recent patterns of its reserves and of Federal funds rates, the stage in the settle ment period, and the existing Federal funds rate, as well as the bank’s current reserve position. The unified market would need ready access to the most recent information about each bank’s position for this approach to be most effective. Arrangements could probably be made, with authorization from the commercial banks concerned, for uni fied markets to obtain current reserve status data directly from Federal Reserve Banks at which these accounts are kept.56 Certificates of deposit. Unified markets could further improve the geographical dis tribution of credit by facilitating the issue of certificates of deposit by rural banks. Rural commercial banks have been excluded from the market for negotiable CD’s primarily be cause the standard size of those traded is so large. For a bank with $5 million to $10 million in deposits, a $1 million CD— for that matter, even one for $100,000— simply is not a satisfactory instrument; it is too large, relative to the bank’s needs and re sources, to be attractive either to the bank or to potential investors.57 A unified market could enable smaller banks to compete for time deposit money by offering certificates in which a number of affiliated banks participate. Such certificates, of course, would be only partly insured by the Federal Deposit Insurance Corporation under existing regulations, and so a prospec tive purchaser might need to investigate a number of banks in order to evaluate the 56 F or more inform ation on the present structure of the Federal funds m arket and some other sugges tions for improvement, see Parker B. Willis, A Study of the M arket fo r Federal Funds, vol. 3 of this series. 57 Description and evaluation of the present sec ondary m arket for negotiable certificates of deposit and review of num erous suggestions for im prove m ent are provided by Parker B. Willis in The Sec ondary M arket fo r N egotiable Certificates o f D eposit, vol. 3 of this series. 171 total risk inherent in a given certificate. To make such certificates marketable, it might be necessary for the unified market to accept liability for them. With proper supervision, unified markets should be able to guarantee such instruments with minimum, risk. If in surance or guarantees were secured, the certificates almost certainly could be traded in the existing market. But if the unified markets cannot guaran tee certificates issued jointly by small banks, it might still be possible to establish a new market for such issues. The certificates would be classified as nonprime and thus expected to carry a slightly higher rate of interest than prime-name certificates. Also, many relatively small certificates would likely be sold to allow issuing banks to ob tain maturities of desired length and diver sity. With sufficient effort, a new group of investors might be attracted to these higheryielding small issues, including smaller cor porations, banks, other financial institutions, and even individuals. The volatility of demand for small certifi cates could prove less than that experienced in the present large-certificate market, thus making these instruments a more appro priate source of funds for small banks. And in particular, banks with well-established seasonal patterns in local deposits and/or loans could meet seasonal outflows by tim ing the maturity of certificates to coincide with periods of loan repayment or deposit inflows. The development of a strong demand for these small, joint-issue CD’s will be depend ent on a good secondary market for them, thus making it important for the unified markets to act as brokers in resales as well as in original issues. The Federal Reserve System could contribute to market develop ment by making its wire facilities available for transfer of certificates. If offices of the 172 unified market also stored and redeemed certificates, costly mail transfers would be avoided and marketability thereby en hanced. Bond services. Another activity valuable to participating banks would be information and brokerage services in U.S. Government securities and municipal bonds. Unified mar- kets could provide up-to-the-minute bond quotations along with analysis of bond mar ket trends and conditions. Using this and other information provided by the markets, bankers could choose the alternative for raising or investing funds that is best suited to their specific situation in each instance. X. CONCLUDING COMMENTS If rural banks are to finance rural capital investment effectively in the future, they must increasingly assume the role of inter mediaries that facilitate flows of funds from money market centers. But present banking and money market mechanisms are ill-suited to the needs of progressive rural bankers who undertake this task. With many rural banks encountering liquidity problems today— and such situa tions likely to intensify as well as multiply in the future— the Federal Reserve System should act promptly to provide more re serves directly to such banks, while simul taneously seeking to perfect market mechan isms. Rediscount procedures should be im mediately revised to provide a greater vol ume of seasonal credit on a more appro priate basis than heretofore. Discount proce dures in general should be revised as neces sary to encourage and facilitate use of this source of funds by rural banks. These measures, promptly instituted, would buy time during which financial m ar ket mechanisms could be improved to ac commodate the needs of small and isolated banks. The Federal Reserve System, and particularly its Federal Open M arket Com mittee, should face up to indications that such banks are unable to compete for funds with money market banks and other agen cies. Federal Reserve distaste for providing long-term discounts to disadvantaged banks, or for purchasing their securities in the open market, is justified only if financial markets are structured to permit such banks to compete for available funds. The Federal Reserve System thus has both an obligation and a stake in securing market perfections that make more significant and equitable participation by small banks possible. One of the more effective ways to over come present deficiencies might be through establishment of a network of unified mar kets to handle transactions in the assets and liabilities of small banks. A device to per mit these banks to market farm and other notes should constitute a vital part of the services provided by such markets. In this and other financial instruments, unified mar kets could provide one-stop information and service to small banks. As these various measures are taken, a considerable number of rural bankers would, as judged from present liquidity levels and trends, be waiting to utilize them. However, perhaps an equal number, judging from the same banking statistics, are not now serving the loan demands of their communities as well as their present liquidity status would permit. Federal Reserve Banks could render valuable service by conducting educational programs aimed at overcoming or minimiz ing managerial inertia at such banks, both C A PITA L A N D C R ED IT R E Q U IR E M E N T S OF A G R IC U LT U R E now and especially as improved sources of funds are established. As knowledge of the improved sources spreads, more community pressure on inert banks could also be ex pected. The Federal Reserve System can make a real contribution to rural finance by helping to achieve the legislative, regulatory, and market changes required by these recom mendations, as well as by arousing private 173 individuals and institutions to face the chal lenges presented. The proposals are revolu tionary in their implications for city cor respondent banks, for rural banks char acterized by managerial inertia, and for the discount officers and the Federal Open Market Committee of the Federal Reserve System, but no more so than the sweeping changes in rural economies that have made them necessary.