View original document

The full text on this page is automatically extracted from the file linked above and may contain errors and inconsistencies.



3 B .2-*30M -10*»1

The attached comes to you two months after your request reached me*
In defense, I can only say that I have used this as an excuse to think through
again — perhaps I could even say to make an "agonizing re-appraisal of” — my
own views on central banking*

What I’ put down here represents, too, des­

pite appearances to the contrary, an effort to state my conclusions briefly* As
you know, for me, that is a time consuming effort*
This memorandum skips little things, even Prof* Wood's interpretation
of the historical record.

He lumps all Federal Reserve officials together too

gives you personally much less than full justice for your independent

views throughout the postwar period, end may well have carried sin so far as to
have ignored our Annual Reports of the past few years and to have skipped a y
own masterpiece in the Williema volume.

I coaid detail some of these, if you

want to mention details in your letter to him, but my hope vould be that you
might prefer to deal with some aspects of the issues discussed in the attached*
Would you want me to circulate this to anyone not mentioned above?
I have taken the liberty of routing copies to them because much of what I have
tried to say here has relevance to other matters that we have been discussing
among ourselves*



October 26, 1954To

Mr. Sproul



Robert V* Roosa

Professor Wood* s Revised

Paper: "Recent Monetary Policies*

Copies to Dr, Williams, Mr, Rouse and Mr. Poelse

On reading this revised, end much more nearly ample, version of
Professor Wood's views, I am frankly disappointed.

His critical perception is

so keen, and his exposure of the limitations and inconsistencies of the present
''norms" is go tolling, that on© is led to expect an incisive, logical formulation
of vhai the System ought to do.

He provides that, to be sure.

But I am afraid

he also illustrates the nakedness of pristine logic in our complicated world.
His view is that Federal Reserve policy should be defined and carried
cut solely in terms of the maturity structure of interest rate®. —
the rate structure for Government securities.


Faced with an "either, or”

I suspect I would take the Wood formula over that of the Ad Hoc Commit­

But I think his foannula, too, is an oversimplification*

several extreme judgments concerning aims and techniques:
for the discount r.echanisia;

It leads him to

that there is no need

that there is no place for fortuitous changes in

and that there is no rignificar.ce to the concept of uncertainty in c a n y •» •

ir.£ out credit control.
I do not see why he has to go so far, and be so dogmatic, in trimming
all System effort down to the determination of a rate curve*

But before turn­

ing to ay ot»n reasons for this doubt, I should, first, outline what seem to me
to be the important zones of agreement between Professor Wood and the views of
r.ost of us here on the underlying principles of monetary policy and action:

Monetary and credit control must be flexible*
The flexibility should be used purposefully, not left to
nature's course.

This means:
Policy concern vith the "degrees of pressure" in all
maturity segments of the market for debt instruments,
Policy action to reach those maturity segments where a
change of pressures clearly farthers the current aim
of policy (restraint, neutrality, or ease).
Conversely, for the System to shrink into one segment of the
market, exclusively, means not merely that monetary control
does less than it could, but even more seriously, that as
practised it may at times makes things worse — like help­
ing nature spread a plague.
The differences begin to arise when Professor Wood formalizes his

Apparently he reasons that all the thinking which goes into policy

determination can be reducod to the selection of a rate curve.

That is, there

must be one rate curve which most nearly embodies all the conflicting considera­
tions and aims that have to be reconciled in setting the single policy appro­
priate for any particular set of conditions.

If the rate curve, and changes

or shifts in it, represent the end results of all policy, and if open market
operations by themselves are capable of bringing about any particular rate
curve that the System might vant to choose, then why bother vith other instru­
ments like the discount mechanism?

Moreover, why leave things fuzzy, or why

stop short of a clear and complete definition of aims, by leaving any zone for
fortuitous changes in rates?

And why, or how, can there be any place for, or

usefulness in, "uncertainty" when thc-re must be one rate curve which best epito­
mizes the aims of policy at any cne time?
This kind of position is difficult to meet on paper, or in debate.
Its logic is appealingly simple.

By contrast, the basis for ny disagreement

is an unsatisfying generalization — that the aims of policy can never be so
clearly specified in advance, nor so fully achieved through changes in interest
rates alone, that ve can rely either on a single measure or on a single tool.

My role in disagreement is not made any easier when I must go farther to admit
that interest rates are probably more important than any other sets of measures
(or guides) for the execution of credit policy, and that open market operations
are likely to be the most important among the System^ tools most of the time*
Perhaps this does suggest, however, that fundamentally my differences with
Professor Wood are differences of degree, and that they may prove to be the kind
of differences that help to invigorate an alliance, rather than the kind of dif­
ferences that drive men into opposing camps*
Dir,count Mechanism
Professor Wood apparently believes that the discount mechanism, in the
fabric of financial institutions in the United States, is virtually useless*
Viihout trying to spell out a full rationale for the discount mechanism — a
task that Mr. Rouse and I are now tackling together, with a hope of completing
something before the end of the year — it does seem to me that bank borrowing
and the discount rate are, to say tho least, necessarily a part of the full use
of open market operations.

Instead of developing a clear-cut, unequivocal view

on what the interest rate curve for Government securities ought to be, at any
uv.rticuler tine, it seems to me the System's task instead should be the determina­
tion of a discount rate appropriate to the circumstances, with open market opera­
tions used, to a large extent, for the purpose of making the discount rate effec­
Though I would not go along with Wood in drawing a whole curve to des­
cribe the appropriate aims of policy, at any particular time, I do think it is
necessary to select one critical point — a point that, as it is moved, tends to
i r p y shifts up or down in the level of the rate curve*

The curve may bend;

on© end or the other may lag, or lead.

I would not see how the "right” curve

could be fully described in advance, and would not see a need to try.


the point — the discount rate — and the synthesising of views that has had to
go into determination of it, I would not then see any need for open market opera­
tions to mark out a whole series of related points along a curve*

Instead, it

should be the job of open market operations to bring about the general degree
of pressure that will, in the different conditions of each week or month, be
consistent with the policy views that have been embodied in the selection of the
discount rate*
In line with the discount rate decision and its policy connotations,
open market operations should help to bring about a roughly corresponding degree
of ease, or of_tightness, in all maturity segments of the debt (or credit) struc­

But there is a wide zone in which borrowing, and the whole discount

mechanism, can do much of the job in developing the ease or tightness intended by
general policy*

Some of the time, changes in line with the policy objectives

may coma about, perhaps not precisely, but adequately, in most maturity segments
without any open market operations*

Even more frequently, limited open market

action in one segment, ordinarily the shortest term, may be enough.

There is

certainly no need to be meddlesome for the sake of meddling, nor for the sake of
pinpointing some pre-deterrained rate curve.
I would argue that there are, indeed, other guides, in addition to the
actual rates of interest on Government securities, which are worth watching* No
doubt all of us in the System would agree that there should be some concern with
the volume of bank reserves.
money supply.

And along with that should go concern over the

There should also be concern with various other market rates of

interest, and with the degree of ease or tightening implied by the level of,

-5 and by changes in the interrelations among, these rates*
What would the approach that I am suggesting mean for periods of ease,
or of restraint?

Very broadly, I would think that in periods of ease the dis­

count rate should be just on the outer fringes of contact with the more sensi- .
tive of the money market rates (Federal funds, dealer loans, and Treasury bills)

and money market rates should be kept within that range mainly through open

market operations.

Borrowing should occur cnly irregularly, and in relatively

small volume, as a safety valve which opens in response to individual bank situa­
tions rather than serving widespread banking needs for funds.

Open market opera­

tions night at tixies, in my "ideal*1 central banking world, exert influences
directly upon the intermediate or long term market, if more decided ease, or a
prompt turn toward ease, seemed to be called for by current economic or credit
the tire.

But action would probably take place within the short area most of
Open market operations should generally provide funds ahead of, or

closely in step with, any general banking need to borrow.

And without producing

.•loppy or disorganised short term markets and rates, open market operations should
keep money market rates generally below, and in an "easing" relationship to, the
the discount rate.

In turn, the discount rate should be the benchmark of the

central bank's policy toward the money market.
In a period of a policy of restraint (and with variations in degree
depending upon the intensity of the need for restraint) the discount rate should
be somewhere close to, perhaps below, the important money market rates —


is, possibly below the rates on dealer loans and Treasury bills, though probably
not Federal funds.

Again serving as a marker, indicating the center of

in the central bank's current policy toward the money market, the discount rate
should be kept effectivo through the aid of open market operations.

This time,

the result should be that banks tend to find themselves in need of marginal
reliance upon borrowing fairly frequently, in relatively large amounts, and that
a significant magnitude of aggregate borrowings should be expected to remain out­
standing much of the time*

There are differences of substance, I believe,

related to the kinds of reserves made available to the banks — and to the strings
that may or may not bo attached to them — which influence the smooth functioning
of the financial mechanism, and the general availability of bank credit* Reserves
are not all homogeneous dollars*

Nor should they be*

In taking advantage of

differences that are qualitative, not purely quantitative, the System may exer­
cise a finer, more sensitive, degree of influence than could be possible solely
through the pluses and minuses brought about through open market operations*
(There can also, of course, be qualitative differences in the uses made of reserves
provided, or taken away, through open market operations — depending mainly upon
the trading methods used — and I would not wish to forswear any of these possi­
bilities, either*)
Fortuitous changes in rates
What then about fortuitous changes in rates?

Professor Wood recoils

froiu the kind of mysticism, or perhaps it is romanticism, that would leave
interest rates to "natural forces1•

He sees the humbug in any pretense that

there can be a genuinely "free" money market, so long as a central bank exists
for the purpose of influencing the general availability of money and credit*
But he lets this wholesome skepticism cany him to another extreme — that there
is no purpose in "continual minor changes in rates which are purely fortuitous
and serve no purpose whatever from the standpoint of credit T t f i a M t w (p. 10,
underlining mine).

He cays further, "A change of rates should be a -■Hgmii 0f

the intentions of the authorities" (also p* 10)*

He indicates that Federal

Reserve officials defend these fortuitous changes on two grounds, first, that
they generate an uncertainty concerning future rates that may be helpful in
gaining System policy objectives, and second, that they reflect the natural
conditions of supply and demand in the market.
discussed further below.

What about the second?

The role of uncertainty is
Is there no tillable ground

between the "natural forces" point of view, and Professor Wood,s

scar observa­

tion that when "changes occur in a fortuitous way it merely means that the
authorities have taken their hands off the controls"(pp. 11 and 12)?
It seems to me that instead of being haphazard, occurring only because
of a System failure to carry out policy with exactness, rate changes "in the
market" serve a fundamental purpose.
funds —

They serve to guide the distribution of

to some extent among maturity sectors, but even more among kinds of

u-es within each sector.

Within a boundary set by the policy implications of

the discount rate itself, there may be a great number of combinations in the
inter-relationships aaong rates.

In the short-tern market, for example, whether

the Treasury bill rate is 1 per cent or 1

l/u per cent, in relation to a dis­

count rcte of 1 .1/2 per cent, nay be of less significance in describing the pre­
vailing decree of pressure than whether dealer loan rates are above or below
the Treasury bill rate, or the Federal -Tunis rate is above or below, or whether
"buy-bacl:" funds are'
available outside New York at rates above or below the
Treasury bill rate.

And there are many possible combinations and differences

of decree among these, and other, variables in the short-term market —


of producing, with rough similarity, the differing degrees of general ease or
pressure that may be sought by credit policy.
This constellation of short-term market rates and techniques migjht
possibly adapt itself, and work reasonably well, if the Treasury bill rate were

sto become a movable peg, such as Wood contemplates for all rates on Government

Bat that would at best be a gamble*

Why pat on the Sjystem a

responsibility to pick out, in advance, the bill rate consistent with the judg­
ments embodied in selection of the discount rate?

Why not instead let the mar­

ket help to find the Treasury bill rate that gives the desired degree of ease or
tightness when taken in combination with other freely moving market rates, in the
circumstances of any given period?
This is not to say that the System should be unconcerned with the .
Treasury bill rate, or with rates on any other maturities of Government debt*
It is to say, though, that concern need not be identical with certainty as to
where those rates ought to be*

And that the System should watch rates, and

rate movements, not as ends in themselves but as symptoms of the prevailing
availability of credit, and as guides to whether more funds, or perhaps less
funds, are needed to fulfill t e general intentionr. of policy for increasing or
hreducing the availability of credit*
To some extent, it seems to me that fortuitous changes among rates
may take the place of deliberate System action.

'Where that is not the case,

they can at least provide useful guides to help direct that action.

Even if,

in given conditions, a rising Treasury bill rate should actually mean that an
unwarranted degree of tightness was building up in the short-term market, and
that System buying should be undertaken to provide additional reserves and
relieve the pressure, there need be no presumption that the {System ought just
to go on purchasing until the bill rate returned to some specified point* Quite
the contrary.

Having helped to signal the need for relief, the Treasury bill

rate may, once System buying begins, become merely a relatively fixed point
around which other rates begin to realign themselves.

To force the Treasury bill

-9 rate itself back down night, then, be to overdo Systran intentions*

It would be

important to watch the Treasury bill rate, and appraise the significance for
general policy of changes in the constellation of rates and credits surrounding


But tko "fortuitousness" of market forces should be relied upon to indi­

cate whether or not it was sufficient merely to lean against developments by re­
leasing a few reserves, or whether sustained baying would be needed to restore
a desired degree of ease.

The observation of rates is an essential part of

that decision making process;

tut this does not mean that rates are necessarily,

however, the common demoninator in which all parts of that process can be ex­
pressed, unequivocally and unambiguously, and determined in advance, on the
basis of independent criteria.
Professor Wood has good reason to doubt that fortuitous changes in
rates should be the end-all of policy in "free" money markets.

He need not

swing all the way from that criticise, however, to the opposite view that no
uoful zone e:d.3ts for any frscdon of rate action outside the purposeful efforts
of the central bank itself.
litertainty as a racomir.abli influence upon credit availability
Professor Wool does not feel that "uncertainty" has any legitimate
place o”on^ the influences used, or exerted, by monetary control.

He recognizes

that "unco.'tiinty of future rates" may "rcduce(s) the moneyness of debt instr-uncuts" (p.ll).

Eat he then observes that any such influence can be had

t'-.rvx’b r^n actual and direct ohanse in the yield itself.

Thei-efore, he implies,

v - y bother with "uncertainty”?
Hero i ; one place ;.+iore, I think, a little use, even loosely, of the
mathematician*s language might help, oven though it appears to be an academic

-1 0 digression.

For I think Professor Wood might agree, as a logician, that*

within reasonable limits, a change which widens the range of variance around an
expected value, may take the place of a change in the expected value itself —
in achieving a given effect*

Translating, this means that comparisons between,

say, tho present rate on 10 year Government bonds, and the rate expected on such
bonds a year from now, can have vastly different meanings, depending on the
degree of certainty attached to the guess concerning the rate next year.


ring other considerations, an investor might be quite ready to bay a 2 l/2 per
cent 10-year bond in the market today, for example, if he expects the yield on
a comparable maturity a year ahead to be 2 1 4 per cent.

But the whole deci­

sion turns (still leaving other considerations aside) on how confident this
investor, or the consensus of the market, may be concerning next year's yield*
If the best estimate — 2 l/A per cent — haa a relatively small variance, that
is, for example, if investors think it quite unlikely that the rate will
deviate by more them l/8 either side of 2 l/4 per cent, then buying of 2 l/2*s
now should be heavy.

But if there is felt to be a good chance that next year's

rate may be anywhere from 1 per cent above to 1, or possibly l/2, per cent below
the "mean expected value” (2 1 4 per cent), then there may well be fewer inves­
tors ready to plunge into the 2 1/2’s today.
Without trying to build this oversimplified example into a real,
flesh-and-blocd illustration, perhaps I can leap on to the implications*


long as there can be differences in the "band of uncertainty” surrounding any
expectations of future rates, even tomorrow* 3 rates, not just those for next
year, then any given interest rate curve, as projected ahead, can have a wide
range of different possible meanings.

It may be as important to influence the

-1 1 uncortainty band, by narrowing or widening it, as to influence the rate itself*
And if for any reason, of a good and practical nature, it proves impracticable
or impolitic at times to exert influences that will lead to sizeable changes in
the rates themselves, then by working on the uncertainty band instead the cen­
tral bank may be able to maintain something like the desired degree of pressure
upon the availability of credit.
In logic, and no doubt in practice, Professor Wood is right in saying
that for every change in uncertainty (i.e*, every increase or decrease in the
variance —

and possibly every alteration in the symmetry -- of the probability

distribution that surrounds each expected mean value) there is a specific change
in the rate itself which could achieve the same effect*
always doing it one way?

But why insist on

At times the hazards of life may simply make large

actual rats charges unacceptable.

Why not then at least make room for uncer­

tainty as a substitute for actual rate change?
r.uch nore often, changes in the

3 tate

At other times, and probably

of uncertainty may just come about, per­

haps without any specif!.' inducement from central bank action*

If those

ianges should be in tha direction of reinforcing current policy, the central
bar;’ would be ungracious indeed (or perhaps oven stubbornly doctrinaire) to
insist that the effect aavjt be achieved another way —

tHough engineering

actual further changes in the rates themselves*
Professor ’
.food's rejection of uncertainty, like his insistence on the
definition of all aims at any given time in terns of a clearly drawn rate
curve, implies some lack of feeling for what 3eeras to me to bo the "art" in­
volved in effective central banking*

let his references to the need for

gauging policy in terms of degrees of pressure (e*g* p. 7) suggest that he
could, without changing much that is fundamental in his paper, concert himself


into an exponent of many of ths positive views on the nature and
monetary controls that have been advanced here.



Federal Reserve Bank of St. Louis, One Federal Reserve Bank Plaza, St. Louis, MO 63102