View original document

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


Address before the Forty-seventh Annual Conference
National Association of Supervisors of State Banks
Louisville, Kentucky - September 22, 1948
by KARL R. BOPP, Vice President
Federal Reserve Bank of Philadelphia

Not long ago a friend told me he was sure that on a particular issue
I had been right in the beginning and that I would be right in the end. My
reply was "I am sure only that I am T
right in the middle'.M Just now I feel I
am back in the middle. Your President, Dick Rapport, put me there when he asked
me to discuss the controversial subject of commercial bank reserve requirements.
Nevertheless, I was happy to accept his invitation because it is an
important current topic and because I believe firmly that the best way to
arrive at solutions to controversial public problems is through the democratic
process of frank and open discussion. My own thinking about reserves has been
influenced by many discussions with commercial bankers, bank supervisors, cen­
tral bankers, legislators, and college teachers. It is not their fault that I
still have much to learn about this complex subject, I hope you will feel free
to contribute to the solution of this problem that is so important not only to
all of us here but to the country.
It will help you to keep the main points in mind if I ask three basic
questions at the outset:
JL .


Why should banks be required to maintain reserves against deposits?
What amount of reserves should all banks collectively be required
to maintain?

3, How should each bank's share of the total be determined?
It is easier, of course, to ask than to answer these questions. I do not expect
all of you to agree with the answers I give. I am sure you prefer brevity to
nicety or strict accuracy of expression. At the same time, I want you to know
that I should appreciate it if you will tell me either now or later where I may
have gone astray.

- 2 -


Why should banks be required to maintain reserves against deposits?
The idea of requiring banks to hold at least a minimum of their de­

posits in reserves apparently developed from a desire to increase the liquidity
of banks*

Governments once thought the obvious solution to insufficient cash

was simply to require banks to hold at least a certain minimum of reserves at
all times.

In arriving at this solution, they apparently overlooked the obvious

fact that a bank cannot pay out reserves that it must hold.

Required reserves

can be used to meet a withdrawal only to the extent that they cease to be re­
quired or become freed by the withdrawal itself.

The only way withdrawals could

be met completely out of required reserves would be to fix requirements at 100
per cent of deposits.
Under a fractional reserve system, a drain must be met primarily by
selling assets or borrowing.

If, for example, a bank is required to maintain a

reserve of 20 per cent and holds no excess reserves, it must either liquidate
80 cents of assets or borrow 80 cents for each dollar of deposits withdrawn.
The bank can do this only if someone else stands ready to buy or lend.

So long

as the withdrawals remain modest other private buyers and lenders can usually
be found directly or indirectly.

But if withdrawals become widespread, panic

may ensue.
As you know, one reason for establishing the Federal Reserve System
was to provide an elastic currency and thus prevent panics.

The method adopted

was to authorize the Federal Reserve Banks to create reserves and money.


tunately, the authors of the Act did not grant the System sufficient authority
to discharge this responsibility under all conditions.

Although they expected

the Reserve Banks to provide liquidity, they imposed inappropriate limits on
the amount of notes they could issue, the amount of reserves they could create,
the collateral for notes, and the collateral against which they could lend.

- 3 These restrictions seriously aggravated our monetary problems in the Great

Some of them have since been removed, as in the Banking Act of

1935, which in effect made all sound assets of member banks a potential basis
of advances by Federal Reserve Banks.
One reason for establishing the Reserve System was to provide the
liquidity that requiring banks to hold reserves did not provide*
the only reason.

This was not

The System also has been charged with responsibility for in­

fluencing the supply of money, of which deposits are the largest element.


although a system of fractional required reserves does not provide liquidity,
it is an indispensable tool in regulating the volume of money in a country with
thousands of independent banks.

The reason is that without such requirements

the commercial banks would not be limited in the amount of deposits they could
extend or maintain on a given amount of reserves.

For example, $1 of reserves

will support $5 of deposits if the reserve requirement is 20 per cent but will
support $10 of deposits if the reserve requirement is 10 per cent*

If banks

were not required to maintain minimum reserves, they would be able to increase
the supply of money by deciding to reduce their reserve ratios*
A short answer to our first question, therefore, is that unless com­
mercial banks are required to maintain at least minimum reserves against
deposits, the Federal Reserve System would be unable to regulate the supply of


What amount of reserves should all banks collectively be required to maintain?
Reserve requirements are one of the two factors which determine the

ability of a banking system to extend deposits.
serves available to the banks*

The other is the amount of re­

Control or influence over these two factors are

complementary means of influencing the volume of money*

Adequate administrative

- 4 authority over both permits the Federal Reserve System to develop a flexible
policy adapted to changing conditions.
The System has shifted its emphasis as one or the other appeared in­
adequate or inappropriate to existing circumstances.

Until the Great Depression,

the Reserve authorities operated entirely through changes in the amount of re­
serves* which they influenced primarily through open market operations and the
discount rate.
in the statute.

They had no authority over reserve requirements, which were fixed
After the revaluation of the dollar, however, banks acquired

large amounts of reserves primarily through imports of gold over which the System
had no direct control.

Although the System was pursuing an easy money policy at

the time, it was clear that its power over the volume of reserves would be inade­
quate if the flood of gold continued and a strong inflationary movement developed.
To meet this contingency, the System was given authority to change within limits
the minimum reserves that member banks could be required to maintain against de­

The System shifted its emphasis from the volume of reserves to reserve

requirements and increased requirements to the legal maximum before we entered
the war.
During the war it directed its efforts primarily to supporting Govern­
ment securities.

The most important decision was to maintain an established

pattern of rates on such securities.

The System paid for securities with re­

This meant in effect that the decision to support Government securities

implied loss of control over the volume of reserves.

At the end of the war the

System was confronted with the alternatives of (a) continuing support of Govern­
ment securities at the expense of controlling the volume of reserves, (b) regain­
ing control over the amount of reserves at the expense of limiting its support of
Government securities, and (c) acquiring additional authority to immobilize

- 5 -

reserves and thus make them unavailable for further expansion.
has done a little of each.

As you know, it

It has continued to support the long-term 2j per

cent yield level but has withdrawn some of the reserves created in the process
by allowing short-term yields to rise.

It has also increased reserve require­

ments at central reserve city banks - which had been reduced in 1942 - and has
secured limited authority to increase reserve requirements of all member banks.
In addition, of course, a large Treasury cash surplus has been the most important

of restraint.
I have recounted this experience because it indicates that there may

be circumstances in the future when changes in reserve requirements will be an
appropriate instrument of policy and other circumstances when changes in the
amount of reserves will be more appropriate.
The amount of required reserves cannot, of course, be considered with­
out reference to the assets that are counted as reserves.

The definition of

such assets should be related to the purpose that the requirements are supposed
to serve; namely, limiting the volume of deposits.

Obviously - to take an

extreme example - the volume of deposits would not be limited at all if banks
had unlimited access to free reserves.

Those who are charged with responsibility

for the supply of money should have adequate authority over the amount of re­
serves, however defined.
In the United States, the Federal Reserve authorities are responsible
for the volume of money.

To discharge that responsibility, it would be desirable

if the only asset that a commercial bank could count as reserves would be a
liability of a Federal Reserve Bank.

For reasons that I shall give later, it

isn't important what form that liability takes or whether the ccnmercial bank
holds it directly or indirectly.

But it is important that the amount be uniform

- 6 -

and that it be a liability of the central bank.

I need not remind you gentlemen

that some of the assets that many banks may now call reserves provide neither
control nor, in a real pinch, automatic liquidity*
This discussion provides a basis for a brief answer to our second ques­

The amount of reserves that banks are required to maintain should be such

that the reserves available to banks will support the volume of money that is
appropriate to existing economic conditions.


How should each bank's share of the total be determined?
Required reserves may be thought of as immobilized assets that cannot

be further loaned or invested.

The amount of such assets that a bank holds may

be viewed as that bank's contribution to an effective national monetary policy.
The question is how much each bank should be required to contribute.

suppose all agree that the basic standard should be equity - that

each bank should be required to hold its fair share of the total.

It is when we

try to apply this principle - which ultimately must be a formula - that we run
into disagreements.

As a general proposition, it would seem to me that under an

equitable system banks that are alike in general nature of business, size, and
character of deposits should be required to hold the same amount of reserves.
As you know, this is not true under our present structure.

Banks that are alike

in the characteristics I have mentioned may have widely different requirements,
depending exclusively upon their location and the authority that has granted
than charters.

I would be the first to admit that these two characteristics are

very important in many ways*

At the same time, I must confess that I cannot see

how they are relevant in determining the contribution a bank should make to maintaining an effective banking system.

- 7 -

In addition to being equitable, a structure of reserve requirements
should be administratively feasible.

Finally, in the development of a new struc­

ture of requirements, attention should be devoted to conditions as they exist so
that a smooth transition may be made.

One might think of this standard as equity

in the short run.
Views differ, of course, as to the precise structure that best meets
these criteria.

From time to time the System has had members of its staff work

on the problem.

As you know, I have been chairman of the staff committee that

developed a plan of uniform reserve requirements.

Your President has asked me to

review that plan, which was presented to and published by the Joint Committee on
the Economic Report.

I should like to emphasize that the plan has not been ap­

proved by the policy-making officials of the System,

Establishment of the plan

would involve a number of changes in the Federal Reserve Act,

Although the plan

deals only with member banks, the staff committee recommended that consideration
be given to the desirability of prescribing uniform reserve requirements for all
commercial banks.

Counsel has advised us that such an extension would clearly

be constitutional,

I have given you reasons for my conviction that only such a

system would be equitable,

I am convinced also that it would not destroy the

dual banking system.
The plan itself consists of five basic points.

The first point id nega­

tive in that it would abolish central reserve city and reserve city designations
of banks.

The Board of Governors now classifies banks into three categories:

central reserve city banks, reserve city banks, and banks not in reserve cities and establishes reserve requirements for each category within the limits prescribed
by the Act.

Classification of cities was a method of identifying banks that were

eligible to receive reserve deposits of other banks under the National Bank Act.

- 8 The obvious intent was to require reserve depositary banks to carry larger re­

The law, however, subjected all eligible banks to the higher require­

ments, whether or not they actually held reserve deposits of other banks*


method of basing reserve requirements on the location of a bank rather than the
character of its business has resulted in inequities*
arise when some banks in a city
and others do not.

Inequities are bound to

hold substantial amounts of interbank deposits

The only choice before the Board of Governors is to classify

the city as a reserve city or as a non-reserve city.

If it does the former, it

penalizes - relative to banks doing similar business elsewhere - the banks with
little or no interbank deposits*

If it does the latter, it favors - relative to

banks doing similar business elsewhere - the banks with such deposits.


inequities have been mitigated slightly by the qualification that the Board may
designate outlying banks in central reserve and reserve cities as country banks;
but not all inequities can be eliminated because the adjective "outlying1 also
relates to location, not to character of business.

I need not remind you, who

are bank supervisors, of the headaches involved in the administration of a law
that is inherently inequitable.
The second point of the plan is that, for purposes of assessing reserve
requirements, deposits be classified into interbank, other demand, and time de­

Many a theoretical hair has been split in disputes over the classifica­

tion of deposits.

The compelling practical objection to treating all deposits

alike is that, depending on the level set, launching such a system would create
enormous excess reserves in central reserve city banks, enormous deficiencies in
non-reserve city banks, or both.

The compelling practical objection to a com­

prehensive system of classification is that it would be impossible to administer.

- 9 -

Any classification is somewhat arbitrary.

Advantages of the proposed classifi­

cation are that, by and large, the three classes of deposits are used for dif­
ferent purposes, are readily identifiable, have traditionally been treated dif­
ferently, and differential treatment would minimize initial disturbances while
yet retaining effective over-all control.

The staff committee recommended that

initial requirements be established at 30 per cent against all interbank deposits,
20 per cent against other demand deposits, and 6 per cent against time deposits.
Several factors were taken into account in selecting these particular ratios.
They were chosen wholly pragmatically after considerable discussion and analysis
of conditions as they existed several months ago*
were tested.

Many different combinations

Feasible combinations that produced total requirements roughly

equal to existing requirements were then applied to individual banks.

On the

basis of tests made at the time, the suggested initial requirements seemed to be
most satisfactory*

should like to emphasize that the initial requirements of any new

system should be established with particular reference to total existing require­
ments at the time and with respect to the impact of the change on individual banks
rather than with reference to any preconceived or established notions.

In general,

the proposed requirements would hit hardest those banks outside reserve and cen­
tral reserve cities - and those subject to "country" bank reserve requirements which nevertheless do a type of business similar to that done by most large banks
in such cities, particularly those holding substantial amounts of interbank de­

Included also are banks that hold a large proportion of demand deposits,

are "loaned up", and have relatively small amounts of vault cash and balances
due from correspondents.

Conversely, banks with relatively large amounts of

vault cash and balances due from correspondents and relatively large proportions

- 10 -

of time deposits would experience reductions in their required reserves.


the standpoint of more effective control of bank credit, this would appear to be
a desirable result.
The third point of the plan is that the appropriate System authorities
should be authorized to change the requirements within limits established in the

I have already discussed the desirability of enabling System authorities

to change reserve requirements from time to time within prescribed statutory
limits in order to prevent injurious credit expansion and contraction.

It is a

modem instrument of central banking policy which is discussed with approval in
virtually every textbook on money and banking.

Although the chief purposes of

authorizing changes in reserve requirements is to influence total requirements,
experience has demonstrated that discretion should be granted as to each require­
ment as well as to requirements as a whole.

In this connection, it should be

pointed out that various groups of member banks could be variously affected by
selective use of changes in the requirements against different classes of

Thus, combinations of changes in requirements on the three classes of

deposits could be utilized to exert differential influence on banks doing differ­
ent types of business.

For example, because of the new treatment of balances

due from banks, an increase in the requirement against interbank deposits would
result in increases in required reserves of banks with an excess of "due to other
banks" over "due from other banks", while at the same time causing increases in
excess reserves of banks with an excess of "due from other banks" over "due to
other banks".

If all commercial banks were subjected to the requirements, there

would be no change in total requirements.

An increase or decrease in the require­

ment for either nonbank demand deposits or time deposits would, of course, affect
»11 banks alike in proportion to their holdings of such deposits.

- 11 -

The fourth point is that banks be allowed to count vault cash as legal

The role of vault cash in the banking system has changed fundamentally

in the past half century.

Before the Federal Reserve System was established,

vault cash was the ultimate reserve of the banking system, since it alone was
available to meet cash withdrawals*

The Federal Reserve Banks, however, have

been empowered to create additional reserves or cash when needed.

The use of

vault cash as reserves would not impair the System's influence over the volume
of bank credit, provided initial requirements are established at appropriate
levels to offset the change*

From the point of view of credit control, System

authorities need not be concerned as to the form of Federal Reserve Bank liabili­
ty - whether against Federal Reserve notes or reserve deposits - that a member
bank prefers to hold as reserves*

The transition to the new system of reserve

requirements would be facilitated by permitting banks to count vault cash as
legal reserves*

Establishment of the suggested uniform requirement against other

demand deposits would increase required reserves of country banks*

Since, how­

ever, such banks hold somewhat larger amounts of vault cash, relatively, the in­
crease in their total requirements would be offset in part by permitting them to
count vault cash as legal reserves.
The fifth and last point is to permit a bank to count as reserve that
portion of its balances due from other member banks which those banks, in turn,
are required to hold as reserves against such balances.

The relationship between

correspondent balances and reserves is a knotty problem with a long history*
After many discussions the committee came to the conclusion that correspondent
balances ought to be related to reserves in such a way that (a) a shift of funds
by member banks into or out of Hdue from banks" would not affect the total volume
of excess reserves in the system as a whole;

(b) "reserve credit" would be

allowed for precisely the portion of "due from banks" that is on deposit with

- 12 -

Federal Reserve Banks (by way of the reserve requirement imposed on deposits due
to banks); and (c) correspondent bank relationships and interbank balances would
be recognized as an established part of our banking system.
designed to accomplish this result.

The fifth point is

So long as the rate at which the "country"

bank or the reserve city bank is allowed reserve credit for its "due from" bal­
ances is equal to the rate at which depositary banks are required to maintain
reserves on interbank deposits, a given reserve will support the same volume of
nonbank deposits irrespective of whether the owner-bank keeps all of its reserve
with its Federal Reserve Bank, or keeps a portion of it on deposit with a cor­
respondent and therefore indirectly with a Federal Reserve Bank.

In either case,

only vault cash and balances which are directly or indirectly on deposit with
Federal Reserve Banks would constitute legal reserves.
We are now in position to give a brief answer to our third question.
An equitable, economically defensible and administratively feasible system of
reserve requirements can be based on the three ma.ior classes of deposits, irre­
spective of the location or the enfranchising authority of the individual bank,
provided the reserve, whether held directly or indirectly« is a liability of the
central bank.

Concluding Comments
In banking, as in all phases of life, we are torn between the forces
of continuity and of change.

We want a banking system suited to our changing

We will not maintain such a system if we resist adamantly all change.

On the other hand, our banking history includes numerous instances - such as the
failure of reserve requirements to assure liquidity - in which seemingly obvious
solutions have proved not to be solutions at all.

Since the study of history

- 13 does not provide a basis for assuming we are noticeably more intelligent than
our forebears, it makes for humility of spirit.
The study of history does, however, encourage us to work cooperatively
in solving problems of mutual concern.

That is why I appreciate so greatly the

opportunity to expose these thoughts on reserve requirements to your critical

I have no pride of authorship in the specific system of requirements,

which contains no single item for which I could conscientiously claim credit.
As chairman of a hard-working committee, I have collected praise for the work
that others have done.
directed to me.

It is only fair, therefore, that criticisms likewise be