The full text on this page is automatically extracted from the file linked above and may contain errors and inconsistencies.
Authorized for public release by the FOMC Secretariat on 5/27/2020 EC'D INRECORDS SECTION Ov MAY 29 1968 BOARD OF GOVERNORS '. OF THE FEDERAL RESERVE SYSTEM WASHINGTON, D. C. EDSI .AL """ May 27, 1968 CONFIDENTIAL (FR) TO: Federal Open Market Committee FROM: Mr. Holland Attached is a memorandum prepared by Mr. Peter M. Keir, Assistant Adviser of the Board's Division of Research and Statistics, with the benefit of comments and expressions of view of various members of the Board's staff, on the subject of "Pros and Cons of an RP Rate Independent of the Discount Rate." This memorandum is intended to complement the memorandum of May 22, 1968, from the Account Manager, Alan R. Holmes, entitled, "An Examination of Competitive Repurchase Agreements" and raises some broader issues and considerations. Robert C. Holland, Secretary, Federal Open Market Committee. Attachment Authorized for public release by the FOMC Secretariat on 5/27/2020 May 27, 1968 CONFIDENTIAL (FR) TO: Federal Open Market Committee FROM: Staff Pros and Cons of an RP SUBJECT: Rate Independent of the Discount Rate. The proposal to divorce the rate on repurchase agreements with non-bank dealers from its traditional tie to the discount rate has been recommended as a means of relating the RP rate more closely to prevailing dealer financing costs during periods of tight money. At such times, yields on short-term Treasury securities and the cost of day-to-day dealer financing typically rise to average levels significantly above the discount rate. RP's made at the discount rate, therefore, provide dealers with a substantial windfall. If this new flexible approach to the RP rate is adopted, the Account Manager proposes as a normal practice to set the rate at a level 1/4 of a percentage point below the average Federal funds rate expected by the Open Market Committee when it voted its most recent policy directive. He apparently believes that this standard approach would apply in most situations. But he suggests that at times, to achieve certain System objectives, it may be desirable in dealing with specific market situations to be authorized to vary the RP rate from the standard 1/4 of a point differential. He would not propose to depart from the standard approach, however, "without further consultation with the Committee." The flexible RP rate proposal has raised a number of logical questions, the most important of which are listed below. (1) Is it possible to develop an RP rate formula related to some money market rate other than the discount rate which can in fact be used operationally as a routine instrument devoid of policy significance? If so, what is the appropriate rate series to use? Authorized for public release by the FOMC Secretariat on 5/27/2020 (2) Should the Committee view any RP rate formula considered,strictly as a routine operating tool devoid of policy significance? Or can a case be made for a more discretionary approach, expressly designed to help carry-out short-run System policy objectives? If so, should the Account Manager be required to consult with the Open Market Committee in advance each time he proposes to use the instrument in this way? (3) Since the purpose of the RP instrument is to provide reserves to the banking system at the discretion of theFederal Reserve, is the fact that non-bank dealers sometimes acquire funds at a subsidy rate a really important issue? Does concern about the subsidy stem from significant operational considerations, or is it largely a question of equity between bank and non-bank dealers? (4) Finally, if the granting of interest rate windfalls to non-bank dealers does seem to require some action, are there other means of resolving this problem than the flexible RP approach? For example, would matched purchase-sale contracts with dealers (the reverse of the matched sale-purchase instrument already in use) be a feasible means of adding temporarily to the supply of reserves without raising questions about policy objectives? This memorandum reviews some of the answers that have been offered to these questions. What Is An appropriate RP Rate Guideline? When the Open Market Committee authorized the Account Management to experiment with an RP rate above the discount rate, several Committee members made it clear that the particular RP rate selected should "be based on clear guidelines related to market rates--lest wide discretionary latitude appear to endow the RP rate with unduly great significance."1 / The Manager's proposed standard spread of 1/4 of a percentage point from 1/ Quoted from the Manager's April 25 memo on a flexible RP rate. Authorized for public release by the FOMC Secretariat on 5/27/2020 -3the expected Federal funds rate is obviously an attempt to respond to this Committee preference. Because the purpose of System repurchase agreements with dealers is to supply reserves, the RP rate must be sufficiently attractive relative to the cost of alternative sources of financing to elicit interest from dealers. This suggests that the RP rate must take account of the effective day-to-day cost of dealer financing. But the only precise measure of dealer financing costs available on a regular basis are the dealer loan rates posted by major money market banks in New York City. These are marginal rates which dealers generally seek to avoid (with varying success) by arranging cheaper financing through alternative sources, chiefly nonfinancial corporations and banks outside New York City. While the Trading Desk obtains only impressionistic reports of the prevailing cost of funds being provided by these other sources, such reports are sufficiently reliable to indicate that out-of-town financing costs generally average only slightly above the effective rate on Federal funds. Thus, the funds rate would seem to be as good a day-to-day benchmark as any, of the low end of the range of effective dealer financing costs. An RP rate 1/4 of a percentage point below this benchmark should, therefore, guarantee that the RP instrument will be sufficiently attractive to insure dealer interest whenever the System wants to supply reserves. The key question raised regarding use of the Federal funds rate as a fulcrum for establishing the RP rate is whether it would in practice be viewed by market participants merely as a routine operating tool, or would instead be "endowed with unduly great significance." Authorized for public release by the FOMC Secretariat on 5/27/2020 -4Experience with the new approach to date suggests that it probably would be difficult to keep market participants from reading policy implications into its use. Since the System moved to a 5-1/2 per cent discount rate, all of the experimentation with the new approach has been at percentage point below the Federal funds rate expected a rate 1/4 of a by the Committee. If this approach were continued, once it became known (as it undoubtedly would) that the Desk normally sets the RP rate at a level 1/4 of a percentage point below the funds rate expected by the Committee, the RP rate would tend to be viewed as an implicit indicator of the Committee's money market forecast. Moreover, since the Desk has the power to make the Committee forecast on Federal funds rates come true, the RP rate would also tend to be interpreted as an indirect indicator of what the System wanted in the way of money market conditions. In these circumstances, any change in the RP rate would immediately suggest to the market that the System had changed its immediate money market objective, and speculation would develop as to what this meant for general monetary policy. This would be so, even though no change were made in the standard 1/4 of a percentage point spread between the RP rate and the expected funds rate. To try to de-emphasize the policy significance of a flexible RP rate related to the expected Federal funds rate, the Desk might simply calculate the RP rate by substituting a fixed and unvarying spread from some past measure of the rate on Federal funds--such as the average effective rate for the preceding statement week, or even for the preceding day. Over-time an automatic formula of this type would undoubtedly Authorized for public release by the FOMC Secretariat on 5/27/2020 -5soon be accepted as such by the market. But given the volatile swings that often occurs in the funds rate, there would clearly be times when such an automatically determined rate would be inappropriate to the Desk's operational requirements. In such situations, the Desk would either have to depart from the fixed formula approach in calculating the RP rate or turn to outright transactions in lieu of RP's. One has the impression the Desk settled on the expected funds rate as a basis for the RP rate, expressly to avoid the arbitrary results likely to follow from an RP rate tied automatically to some measure of the actual funds rate. Moreover, as has been noted, even when using the expected funds rate as a guideline, the Account Manager would expect some occasions to arise when continued adherence to the normal 1/4 of a point differential would not adequately promote System objectives. Examples of situations in which departures from the standard RP rate guildeine might occur are described below in sentences quoted from the Manager's April 25 memo. "Generally, when Treasury bill rates were moving below a level that the Committee felt was desirable, a higher rate on repurchase agreements could help to achieve the System's objectives. On the other hand, if the Federal funds rate were to move significantly higher in relation to the discount rate, then consideration could be given to a repurchase agreement rate at a spread greater than 1/4 per cent from the Federal funds rate. Otherwise the repurchase rate could chase the Federal funds rate higher in an escalating spiral that might not accord with System objectives. Generally, when Treasury bill and other market rates are under strong upward pressure, or when unusually high dealer financing costs pose a particular problem to the functioning of the market, the Committee might wish to approve a lower than normal rate for repurchase agreements as a means of dealing with these conditions." This suggests that occasions on which departures from the standard 1/4 of a percentage point guideline would seem appropriate might not be so rare. And experience with the standard guideline thus far seems to confirm Authorized for public release by the FOMC Secretariat on 5/27/2020 -6this suggestion. For example, during the statement week ending May 15 (in which dealers and others were preparing to make settlement on the new issues offered in the Treasury's May financing), the effective rate on Federal funds consistently averaged well above the 6 per cent level contemplated by the Committee. Despite substantial System injections of reserves through short-term repurchase agreements--expressly intended to moderate the prevailing tightness in the money market--the high level of Federal funds rates persisted. Apparently this was partly so because market participants interpreted the RP rate--1/4 of a percentage point above the discount rate--as evidence that a relatively high funds rate was one of the characteristics of money market conditions that the System wanted to achieve. In these circumstances a dropping of the RP rate back toward the discount rate would have given a signal to money market participants that the Desk viewed the tightness then prevailing to be somewhat excessive. In short, there is reason to believe that a flexible RP rate related to the expected Federal funds rate would generally be viewed by market participants as a significant indicator of policy intentions. logical questions follow from this conclusion: Two (1) Is there a money- market rate other than the funds rate, less dominated by System operations, which might serve more effectively as a fulcrum for a routinely determined RP rate less endowed with policy significance? (2) Would an RP rate endowed with policy significance necessarily be bad? If the Account Manager possessed the necessary discretion to vary the RP rate in relation to the funds rate, might not this instrument serve as a useful addition to the System's kit of policy tools? Authorized for public release by the FOMC Secretariat on 5/27/2020 -7The market yield on 90-day Treasury bills is one obvious money market benchmark whose level is not so strongly influenced by day-to-day Desk action. If the RP rate were set in some fixed spread relationship close to or above the 90-day bill yield, instead of in relation to the expected funds rate, market participants might be more likely to accept this as an objective market-determined tool for setting the RP rate on an automatic basis. The bill rate would provide only a rough measure of relevant dealer financing costs, however; much of the time an RP rate close to the bill rate would provide a bigger element of subsidy than the 1/4 of a point spread from the funds rate. But at other times, as in the recent period of pessimism about fiscal action, an RP rate linked automatically to the bill rate would be too high. Thus, with a bill rate guideline too, the need for departures from the norm would probably arise, posing some of the same types of problems associated with an RP rate related to the funds rate. The Flexible RP Rate as a Policy Instrument The preceding discussion suggests that an RP rate related to the expected funds rate would inevitably be endowed with policy significance. For this reason any continued use of the proposed funds rate guildeine might on occasion require some departure from the normal 1/4 of a point differential,merely to counter-balance an overreaction of the money market to developments set in motion by earlier Desk use of the guideline. Beyond these occasions, however, one can readily postulate situations in which changes in the RP rate would serve a useful policy function. Authorized for public release by the FOMC Secretariat on 5/27/2020 -8Most recently, for example, the accentuation of upward pressures on money market rates created by the rather widespread rumor of an impending increase in the discount rate might have been reversed by dropping the RP rate from 5-3/4 per cent back to the prevailng discount rate. On the other hand, on occasions when the 90-day bill yield was tending to sink below the range contemplated by the Committee, upward pressure could be exerted on this level by raising the RP rate. In situations like these the Desk would, of course, have to stay particularly alert to changing market tendencies and attitudes in order to be prepared to modify the RP rate again, once it became clear that the given policy purpose had been accomplished. While there is little question that a flexible RP rate used in this way could have a significant and subtle impact on money market conditions, the fact that it would be endowed with policy significance raises a question how closely and in what ways the Committee might want to control its use. If the Account Manager were to be required to consult with the Committee before any departure from the normal 1/4 of a point spread, this would seem to require a special telephone conference, since situations in which a departure might be desirable would be likely to develop rather suddenly and require rapid action to be fullyeffective. This procedure could, of course, be streamlined somewhat if approval of only the Executive Committee, rather than the full Committee, were required. Need for a Flexible RP Rate Thus far, the case for a flexible RP rate has been based principally on the fact that RP's made at the discount rate often give non-bank dealers an unwarranted windfall in their financing costs. But the financing of Authorized for public release by the FOMC Secretariat on 5/27/2020 -9dealers is, of course, only incidental to the primary RP purpose of facilitating reserve management objectives. In many areas of public policy, subsidies to private firms are used to facilitate the realization of policy objectives. Hence, it seems relevant to ask, (1) whether the subsidy element in RP's to non-bank dealers is needed, makes no difference, or makes it more difficult to achieve primary objectives; and (2) if not needed, whether steps taken to reduce the subsidy will on balance be a net benefit or hindrance to the attainment of overall policy goals. It is not clear from the Account Manager's April 25 memo whether he is seriously concerned that RP's made at the discount rate during periods of tight money will give dealers enough of a windfall on this marginal part of their financing to complicate the efforts of the Desk in attempting to realize Committee goals. Reference to the occasional difficulty of keeping the bill rate from sinking below the range expected by the Committee does suggest, however, that this is one of the considerations he has in mind. The April 25 memo also seems to imply that it is undesirable on equity grounds for non-bank dealers to have a financing advantage relative to bank dealers (who do not have access to System RP's). But neither the operational nor the equity considerations involved in the subsidy question have been spelled out in any detail. Without further analysis of these operational and equity considerations, answers to the final question whether steps taken to eliminate the subsidy would be a net policy benefit or a net hindrance are not easy. If one wishes to avoid an RP rate guideline endowed with too much policy significance, he may be inclined to conclude that the disadvantages of Authorized for public release by the FOMC Secretariat on 5/27/2020 -10trying to eliminate the subsidy outweigh the advantages. But if one is inclined to press for a variable RP rate as an additional policy tool, he may be inclined to urge this approach quite strongly, for reasons of greater priority than elimination of the subsidy to non-bank dealers. Matched Purchase -Sale Contracts Among the Committee members who have questioned the desirability of a flexible RP rate, Governor Robertson has proposed matched purchasesale contracts as an alternative. The ostensible advantage of this technique is that by letting dealers bid for funds to be supplied by the System on a temporary basis, the rate is market determined; and changes in its level lose their policy significance. The technical difficulties with matched purchase-sale transactions described by the Account Manager seem to rule out in his May 22 memo to the Committee would any near-term use of this instrument. However, if the Committee were to decide that matched purchase-sale transactions offer considerable promise in principle, further study might profitably be given to whether a more precise system of pricing could be developed for this type of instrument and readily adapted to processing quickly as needed on a Desk-type computer similar to some of those already in use by dealer firms that have complicated price decisions to make on short deadlines. The competitive repurchase agreements alternative suggested by the Account Manager. as one technique which would be technically feasible for auctioning temporarily available System funds is an interesting one with a number of potentially useful applications which seem to bear further consideration.
@FedFRASER