View original document

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

JOINT COMMITTEE PRINT

A STUDY OF THE DEALER MARKET FOR
FEDERAL GOVERNMENT SECURITIES

MATERIALS PREPARED
FOR THE

JOINT ECONOMIC COMMITTEE
CONGRESS OF THE UNITED STATES

Printed for the use of the Joint Economic Committee
UNITED STATES
GOVERNMENT PRINTING OFFICE
62471

WASHINGTON : 1960

For sale by the Superintendent of Documents, U.S. Government Printing Office
Washington 25, D.O. - Price 40 cents




JOINT ECONOMIC COMMITTEE
(Created pursuant to sec. 5(a) of Public Law 304, 79tb Cong.)
PAXJLH. DOUGLAS, Illinois, Chairman
W R I G H T PATMAN, Texas, Vice Chairman
SENATE
J O H N SPARKMAN, Alabama
J. WILLIAM FULBRIGHT, Arkansas
JOSEPH C. O'MAHONEY, Wyoming
JOHN F. K E N N E D Y , Massachusetts
P R E S C O T T BUSH, Connecticut
J O H N MARSHALL BUTLER, Maryland
JACOB K. JAVITS, New York
JOHN W. LEHMAN,

II




HOUSE OF R E P R E S E N T A T I V E S
RICHARD BOLLING, Missouri
HALE BOGGS, Louisiana
H E N R Y S. REUSS, Wisconsin
F R A N K M. COFFIN, Maine
THOMAS B. CURTIS, Missouri
CLARENCE E. KILBURN, New York
WILLIAM B. WIDNALL, New Jersey
Clerk and Acting Executive Director

LETTERS OF TRANSMITTAL
DECEMBER 30,1960.

To Members of the Joint Economic Committee:
For the information of the members of the Joint Economic Committee and others interested there is transmitted herewith an analysis
of the operations of dealers operating in the market for Federal
Government securities.
As you know, the study grew out of the reports collected from the
17 dealers in Government securities in the fall of 1959 to supplement
the information obtained in the committee's hearings which were conducted in connection with the study of "Employment, Growth, and
Price Levels."
This study of the Government securities market is a pioneering one
and should be of permanent value as a source of information in understanding the market for Government securities. No other study of
comparable scope has ever been made of this market. The Federal
Reserve-Treasury study which was reported to the committee during
its hearings in connection with the study on "Employment, Growth,
and Price Levels" provided much information about the Federal
securities market and the developments in the market during 1957
and 1958. The current analysis covers the entire period from 1948
to 1959 and interrelates the factual material from the dealers with
other information obtained by the committee in its hearings and with
the Federal Reserve-Treasury study.
This study will be valuable to those concerned with analyzing the
behavior of this market, to those charged with management of the
ublic debt and with monetary policy, and also should assist in the
evelopment of a uniform and regular system of statistical reporting
for this important market.

S

P A U L H . DOUGLAS,

Chairman, Joint Economic

Committee.

DECEMBER 19,1960.
Hon. P A U L H . DOUGLAS,

Chairman, Joint Economic Committee,
U.S. Senate, Washington, D.C.
DEAR SENATOR DOUGLAS : Transmitted herewith is " A Study of the
Dealer Market for Federal Government Securities" prepared in response to the committee's instructions in its annual report filed with
the Congress February 29, 1960 (S. Rept. 1152, 86th Cong., 2d sess.).
This analysis covers the operations of the 17 dealers in Federal Government securities as set forth in the reports filed with the committee
by the dealers in connection with the committee's 1959 study on "Employment, Growth, and Price Levels." I n connection with that study,




in

IV

LETTER OF TRANSMITTAL

on November 2, 1959, the dealers were requested to furnish information concerning their operations as follows:
(1) Statement of earnings and expenses;
(2) Position and sources of financing on selected dates;
(3) Financial position in securities, 1948-59;
(4) Yearly gross transactions in U.S. Government securities,
1948-59; and
(5) Balance sheets, 1948-59.
A set of four forms w^as supplied to each dealer upon which to record
their reports. The last of these data was received in March of 1961,
the delay being due to difficulties experienced by some of the dealers
in completing the tabulation of the information, particularly for
earlier years. I n completing the study this year, it was discovered
that additional information was needed, including both additional
data and explanations of certain items which were not completely clear.
The request for additional information was made to the dealers in
June. I n July, a number of the dealers were visited in New York
to clarify the meaning of the information supplied.
The tabulations and analyses were prepared and the report was
written by Profs. Allan H. Meltzer and Gert von der Linde, of Carnegie Institute of Technology, Pittsburgh. James W. Knowles of the
committee staff provided the general direction and supervision, designed the questionnaire and carried out the initial phases of the
investigation in connection with the studies of employment, growth,
and price levels during the fall of 1959.
This study has been carried out in complete conformity with Committee Rule No. 23, which provides that the information contained
in any books, papers, or documents furnished to the committee by an
individual, partnership, corporation, or other legal entity shall, upon
the request of the individual, partnership, corporation, or entity furnishing the same, be maintained in strict confidence by the members
and staff of the committee. Under the committee rules the only persons having access to this file have been Messrs. Knowles, Meltzer, and
von der Linde who were formally designated to carry out the study.
I n the report submitted herewith, the confidential character of the
individual responses has been fully protected.
One unpublished and three published sources have been of great
assistance to the authors in the course of this study. The unpublished
volume, "The Market for United States Treasury Obligations," prepared for the National Bureau of Economic Research by Morris Mendelson and Roland Robinson was extremely helpful in corroborating
reports from dealers and in describing the functioning of the market.
The three principal published sources have been acknowledged at
appropriate places in the text. The full citations, however, are given
here:
(1) "The Treasury-Federal Reserve Study of the Government
Securities Market," part I, July 1959; part i t , February 1960; part
I I I , February 1960. This is referred to in the text as the TreasuryFederal Reserve study.
(2) "United States Monetary Policy: Recent Thinking and Experience," hearings before the Subcommittee on Economic Stabilization
of the Joint Committee on the Economic Report, Congress of the
United States, Washington, 1954. This volume contains a copy of



LETTER OF TRANSMITTAL

V

the "Federal Open Market Committee Report of the Ad Hoc Subcommittee on the Government Securities Market, November 12, 1952."
These are referred to as "Hearings, 1954" and "Ad Hoc Report."
(3) "Employment, Growth, and Price Levels," hearings before the
Joint Economic Committee, Congress of the United States, Washington, 1959. Parts 6A, 6B, and 6C. I n the text, these volumes are cited
as "Hearings, 1959."
Numerous interviews or meetings were held with Government securities dealers, the debt management staff of the Treasury, several
officers and staff members of the Board of Governors of the Federal
Reserve System, the Federal Reserve Bank of New York, and the
Federal Open Market Committee. All of those involved contributed
measurably to this report. Without their wholehearted cooperation
the report could not have been completed.
Processing of the data was in part accomplished through the helpful assistance of M. H . Schwartz and Catherine Whistler of the
Division of Research and Statistics, Board of Governors, Federal
Reserve System, after the data had been coded and punched to protect
the confidentiality of the respondents.
Messrs. Meltzer and von der Linde wish me to express, too, their
indebtedness to their colleagues, G. L. Bach and Edwin Mansfield of
Carnegie Institute of Technology for valuable discussion and criticism of ideas and analytical techniques used in this report and to
Mrs. M. Blank whose assistance in preparing the report went far
beyond the obligations of a secretary.
Finally, the committee should note particularly that it was difficult
to arrive at a consolidated picture of the market because of the lack
of uniformity in record keeping practices and reporting in the industry during prior years. This has made it difficult both for the dealers
in supplying data and for the staff in reconstructing reasonable estimates of totals for the market; and it has put certain limitations, noted
in the text, upon the conclusions that can be drawn.
Respectfully submitted.




J O H N W. L E H M A N ,

Clerk and Acting Executive

Director.




CONTENTS
Pace

Chapter I. The role of the dealer market
Introduction
Identifying the dealers
Dealer functions
Definitions
The dealer market and the saving-investment process
Arbitrage in the market
Scope of the study
Chapter II. The dealer market: Trading organization and philosophy of
dealer
firms
Internal organization for trading
Customer services and information
flow
Dealer views of dealer functions
Making markets
Interdealer tradings
Size of transactions
Short sales
Dealers, brokers, and lieutenants
The Federal Reserve and the dealer market
Transactions with the dealers
Clearing securities transactions
Recognizing Government securities dealers
Chapter III. Dealer positions
Definitions and problems in the measurement of dealer positions
What is to be included
_
Government guaranteed issues, agencies, and others
Repurchase agreements
Executed versus commitment reporting
Investment accounts
How should positions be measured
Net versus gross measures
Par versus market versus cost price
..
Timing of the reports
Summary
End of year inventories
Dealer holdings of the public debt
Midyear positions
Individual dealers portfolios
Investment accounts
Findings
Chapter IV. Volume and composition of transactions in U.S. securities
Development of trading in the dealer market
Scope of data and types of transactions
Volume of dealer trading
Composition of dealer trading
Concentration of trading in the dealer market
Findings
Chapter V. Financing the Government securities market
Sources and uses of funds
Comparison of dealer commitments and dealer net positions
Magnitude and distribution of sources and uses of dealers' funds,
Distribution of sources for individual dealers
Rates on dealer loans
Changes in the rate structure over time
Effect of the rate structure on sources of
financing
Leverage and margin requirements
Cash balances
Findings



vn

1
1
2
3
3
5
5
9
11
11
13
15
15
18
19
20
21
22
22
24
25
27
27
27
27
28
29
30
31
31
32
33
34
35
39
41
45
46
47
49
49
50
57
65
66
68
71
72
78
79
82
85
87
89
90
93
94

VIII

CONTENTS
Page

C h a p t e r V I . Legal organization a n d capital base
Legal organization, capital base, and credit worthiness
Legal form of organization of t h e 12 nonbank dealers
Size a n d organization
T a x considerations
T r a d i n g risk a n d limited liability
Dealers capital base
Capital base and cash accounts
Findings
C h a p t e r V I I . Sources of revenue, expenses and earnings
Scope of d a t a and definitional problems
Sources of earnings from current operations
Trading profits, spreads, a n d profit margins
Interest received and interest paid
Other earnings and profits from investment accounts
Current operating expenses
Salaries
Other current operating expenses and special charges on gains
Summary
N e t income before taxes
N e t income before taxes a n d dealer capital
Findings___
C h a p t e r V I I I . Operation of the dealer m a r k e t
S u m m a r y of t h e major
findings

97
97
98
98
99
100
101
103
104
105
105
106
110
118
124
126
126
128
130
132
135
140
141
141

TABLES
Table III—1. Distribution of December 31 by days of t h e week, 1947-58—.
T a b l e I I I - 2 . Aggregate dealer positions a t end of years 1947-58 classified
by m a t u r i t y
T a b l e I I I - 3 . Percentage distribution of dealers position by m a t u r i t y
classes on December 31, 1947, to 1958
T a b l e I I I - 4 . Percentage composition of m a t u r i t y classes of o u t s t a n d i n g
m a r k e t a b l e debt a t 1947-58 yearend
Table I I I - 5 . Dealers' position as a percentage of outstanding m a r k e t a b l e
debt, by m a t u r i t y classes at 1947-58 yearend
Table I I I - 6 . Comparison of net dealer positions, J u n e and December
1948-58
T a b l e III—7. T o t a l net position last Wednesday of December/last Wednesday of J u n e for 14 individual dealers
T a b l e I I I - 8 . Percentage distribution of individual dealer positions
T a b l e I I I - 9 . Dealer investment accounts
T a b l e I V - 1 . Allotments to brokers and dealers on subscriptions for public
m a r k e t a b l e securities, 1953-58
Table I V - 2 . Total transactions volume and r a t e of change in volume by
m a t u r i t y class, all reporting dealers, 1948-5S
T a b l e I V - 3 . Ratios of sales volume to securities outstanding a t t h e end
of calendar years 1948-58
T a b l e I V - 4 . Outstanding public m a r k e t a b l e U.S. securities, end of calendar
years, 1948-58
T a b l e I V - 5 . Percentage composition of public m a r k e t a b l e U.S. securities,
end of calendar years, 1948-58
Table I V - 6 . Composition of t o t a l trading volume, all reporting dealers,
1948-58
Table I V - 7 . Composition of trading volume—annual high, low, a n d
average ratios for individual dealers, 1948-58
T a b l e I V - 8 . Share of m a r k e t transactions, 1948-58
T a b l e V - l . R a t i o of dealer c o m m i t m e n t s t o dealer net positions for 12
nonbank dealers at end of year
T a b l e V - 2 . Sources a n d uses of dealer financing (12 nonbank dealers),
1947-58
T a b l e V - 3 . Percentage distribution of sources for four nonbank d e a l e r s . _
T a b l e V - 4 . Reported rates on loans and R P ' s in percent
Table V - 5 . Annual leverage ratios; nonbank dealers
Table V - 6 . Margin requirements on collateral loans
T a b l e V I - 1 . Legal form of 400 leading investment bankers b y size classes. _
T a b l e V I - 2 . Ratios of positions to net worth, all nonbank dealers, 1 9 4 8 - 5 8 .



33
37
38
40
40
41
44
46
47
56
58
59
60
62
65
66
67
78
80
83
88
91
92
99
101

CONTENTS

IX
Page

Table V I - 3 . N o n b a n k dealer long-term financing and net worth, 1948-58_ _
Table V I - 4 . Long-term debt, for all dealer corporations, 1948-58
Table V I - 5 . Cash positions a n d t o t a l long-term financing, 12 nonbank
dealers, 1948-58
Table V I I - 1 . Composition of gross earnings from current operations and
operating expenses for reporting dealers, 1948-58
Table V I I - 2 . Earnings and expense items as percentages of gross earnings
from current operations, all reporting dealers, 1948-58
Table V I I - 3 . Price quotations for U.S. Treasury bonds and notes
Table V I I - 4 . Earnings a n d expense items in dollars per $1,000,000 of
sales, all reporting dealers, 1948-58
Table V I I - 5 . T o t a l t r a d i n g profit as percentage of gross earnings a n d in
dollars per $1,000,000 of sales, all reporting dealers, 1948-58
Table V I I - 6 . I n t e r e s t received a n d interest paid as percentages of gross
earnings, all reporting dealers, 1948-58
Table V I I - 7 . Interest received and interest paid in dollars per $1,000,000
of sales, all reporting dealers, 1948-58
Table V I I - 8 . Other earnings a n d profits from investment accounts, all
reporting dealers, 1948-58
T a b l e V I I - 9 . Salaries as percentage of gross earnings and in dollars per
$1,000,000 of sales, all reporting dealers, 1948-58
Table V I I - 1 0 . Other current expenses as percentage of gross earnings and
in dollars per $1,000,000 of sales, all reporting dealers, 1948-58
Table V I I - 1 1 . Selected earnings and expense items for diversified nonbank
dealers and all b a n k dealers, 1948-58
Table V I I - 1 2 . Gross a n d net earnings as percentages of net w o r t h a n d
t o t a l long-term financing, all reporting nonbank dealers, 1948-58
Table V I I - 1 3 . N e t income before special charges a n d taxes as percentage
of gross earnings and in dollars per $1,000,000 of sales, all reporting
dealers, 1948-58
T a b l e V I I - 1 4 . R a t e s of r e t u r n for selected industry groups, 1948-58




102
102
103
108
109
111
114
116
118
120
125
127
129
131
133
134
138

CHAPTER I
THE ROLE OF THE DEALER MARKET
INTRODUCTION

Monetary policy and the management of the public debt are among
the most frequently discussed topics in the economic and financial
literature. Public hearings, books, scholarly papers, discussions in
the financial and lay press have analyzed and reanalyzed actions
taken or avoided by the Federal Reserve and the Treasury. Numerous proposals for modification of existing procedures or improvements
in techniques for funding the debt or stabilizing the economy have
appeared. There has been no apparent tendency for the popularity of
these subjects to diminish in recent years. On the contrary, as economic stabilization and debt management problems have become more
acute the discussion has expanded.
Literature about the functioning and performance of the market for
Government securities dealers is in marked contrast. Until recently,
sources of information about dealer operations in this market were
relatively scarce. Quantitative information about this market was
based on intelligent guesswork or illustrative examples—principally
those contained in testimony given at congressional hearings. Detailed
reports of operations which have been collected for many years by one
Federal Reserve bank were submitted to the bank by the dealers with
the understanding that such reports would not be made part of the
public record.
Secrecy about the aggregate operations of Government securities
dealers is surprising in view of their important role in the economy
and the readily available statistics on commercial banks, savings banks,
insurance companies, and other financial institutions which regularly
appear in the Federal Reserve and Treasury bulletins. I t is even
more surprising in view of the fact that these dealers form the largest
securities market in the country and one which is so heavily vested with
the public interest.
Two principal reasons make the operation of this market an important source of information about the working of the economy.
First, monetary policy and debt management operations both condition and are conditioned by the operations which take place in the
dealer market. Second, the role which the dealer market plays in the
savings-investment process merits more detailed examination than has
been possible in the past.
Through the dealer market pass virtually all of the secondary transactions in Government securities. Efforts by the Federal Reserve to
influence the free reserves of commercial banks are most often carried
out through open market purchases or sales. F o r these purposes, the
dealer market is the open market. Efforts by the Treasury to fund
the debt or raise new cash depend in part on the functioning of the
dealers as underwriters of new issues.



1

2

STUDY OF FEDERAL GOVERNMENT SECURITIES DEALERS
IDENTIFYING THE DEALERS

Until recently, few quantified statements about dealer operations
were available. Even the estimated number of active dealers was
subject to some variation in the few publicly available accounts of
Government securities operations. The Federal Open Market Committee report of the ad hoc Subcommittee on the Government Securities Market, prepared in 1952, refers to "about 20 dealers including
some banks with trading departments." * Koosa, writing in 1956,
suggests that there are between 15 and 20 "quasi-specialists." 2
Testimony before the Joint Economic Committee in 1959 identifies
17 Government securities dealers, 5 banks, and 12 nonbanks. 3 Listed
alphabetically, the 17 dealers are:
Bank dealers

1. Bankers Trust Co.
2. Chemical Bank New York Trust Co.
(formerly Chemical Bank & Trust
Co.).
3. Continental Illinois National Bank
& Trust Co. (Chicago).
4. First National Bank of Chicago.
5. Morgan Guaranty Trust Co. (formerly Guaranty Trust Co.).

Nonbank dealers

1. Bartow, Leeds & Co.
2. Briggs, Schaedle & Co., Inc.
3. C. F. Childs & Co., Inc.
4. C. J. Devine & Co.
5. Discount Corp.
6. First Boston Corp.
7. Aubrey G. Lanston & Co.
8. New York Hanseatic Corp.
9. Wm. E. Pollock & Co., Inc.
10. Chas. E. Quincey & Co.
11. D. W. Rich & Co., Inc.
12. Salomon Bros. & Hutzler.

Two of the banks and one of the nonbanks (C. F . Childs) have headquarters in Chicago, but most of the trading is done in New York.
Aside from their transactions in the Government securities market,
there are few similarities among the dealers. Four of the nonbank
dealers are partnerships. Two of the firms are leading underwriters
of new corporate issues and two others participate in many offerings.
Two are members of the New York Stock Exchange; at least two are
active in the acceptance market. Several are dealers in municipal,
agency, and International Bank bonds, and some underwrite the initial
distribution of such issues. Other differences, particularly those
Which reflect on their operations as Government securities dealers, will
be discussed at a later point. Enough has been said to suggest that
the dealer function is performed by a variety of different types of
firms.
From 1944 to 1953, a distinction between dealers was made by the
Federal Reserve Bank of New York. Twelve dealers were recognized
for trading purposes by the Open Market Committee. The remaining 5 dealers—Bartow Leeds, Briggs Schaedle, Pollock, Lanston, and
New York Hanseatic—now participate in open market transactions
with the Federal Reserve on an equal footing with the other 12.4 Since
1953, the terms "recognized" and "unrecognized" have ceased to have
their former meaning in this market.
1
This report was made public In 1954 as a part of "United States Monetary Policy:
Recent Thinking and Experience," hearings before the Subcommittee on Economic Stabilization of the Joint Committee on the Economic Report, Washington, 1954. The quotation
above
is from p. 261 of the hearings.
2
R. V. Roosa, "Federal Reserve Operations in the Money and Government Securities
Market,"
Federal Reserve Bank of New York, 1956, p. 35.
3
"Employment, Growth, and Price Levels," hearings before the Joint Economic Committee, pt. 6B: "The Government's Management of Its Monetary, Fiscal, and Debt Operations." Washington, 1959. p. 1508. This report will be referred to as Hearings 6B.
* Hearings, op. cit, pt. 6B, pp. 1509 and 1515.




STUDY OF FEDERAL GOVERNMENT SECURITIES DEALERS
DEALER

3

FUNCTIONS

One part of dealer operations in Government securities is the underwriting of new issues of Treasury securities. Dealers, along with
many other groups, assist in the distribution of new cash issues or refundings in a variety of ways. If the securities are sold at auction,
dealers bid for their own account and furnish advice on the expected
yield of a forthcoming issue to other bidders. Dealers buy a portion of most new issues which are sold for cash and facilitate the
exchange of old issues for new securities when the Treasury exchanges
issues and when rights are offered.
After a new issue has appeared, the dealers become a principal means
by which secondary distribution is effected. Banks and others buy
for the account of customers and correspondents and assist in the
secondary distribution, but dealers actively buy and sell the new securities in the open market.
Even when dealers do not directly participate in the underwriting,
they usually are willing to trade the issue in the secondary market.
They thereby facilitate the "seasoning" of new issues and the establishment of a yield which reflects the market's opinion of the value
of the new issue in relation to other outstanding securities.
Government securities dealers continue to buy and sell existing
issues for their own account and risk during the life of the security.
Not all dealers trade in all issues as will be seen more fully in chapter
I I , but all Government securities dealers are willing to buy and sell
some range of maturities—frequently in large amounts—at the prices
which they quote over the telephone. As a group, the 17 dealers
listed above are the principal (at times the only) market for outstanding Treasury issues at quoted prices. They thereby permit the
orderly exchange of issues for cash or other issues.
DEFINITIONS

Several terms will be used repeatedly in this study. To avoid repetition, and possible misunderstanding, they are defined here:
The term "Government securities" will be used to refer to the marketable debt available for trading. All of the publicly held marketable issues plus those issues or portions of regular issues held by Treasury trust accounts and the Federal Eeserve are included here. Special issues and nonmarketable issues are excluded.
"Government obligations" is defined as total Government securities
plus nonmarketable and special Treasury issues, plus outstanding
agency issues. Included in the agency issues are Federal home loan
bank bonds, obligations of the Federal intermediate credit banks, the
Federal National Mortgage Association and similar agencies. Maturity classes, as used here, will refer to the following groups unless
otherwise noted:
1. Bills—short-term Government securities (under 1 year) which
when issued weekly have 91 or 182 days to maturity, plus 1-year bills
issued quarterly since April 1959, and special bills issued at irregular
intervals. Bills are issued on a discount basis and are usually offered
by the auction method.
2. Certificates—or certificates of indebtedness—are short-term securities with 1 year or less to maturity when issued. I n the past, the



4

STUDY OF FEDERAL GOVERNMENT SECURITIES DEALERS

Treasury has issued certificates both with and without coupon attached. All of the presently outstanding issues are coupon securities.
A special kind of certificate known as the tax anticipation certificate
has been issued at various times since July 1953. These are particularly attractive to corporations as a medium by which funds accumulated for tax purposes may be invested temporarily. Interest is usually paid beyond the date on which taxes are due.
3. Notes are issued with original maturity of 1 to 5 years. They
are coupon securities issued at par value.
4. Bonds may be issued in any maturity but have usually been
issued with 5 or more years to maturity. Like notes they are coupon
securities issued at par. They may or may not have an optional call
feature which permits the Treasury to retire or refund them prior to
their final maturity. Most often the call feature can be first exercised
5 years before final maturity but both longer and shorter call dates
have been used.
5. "Long-term bonds" as used here, usually refers to issues with 5 or
more years to maturity on a particular date.
6. Under 1-year maturities include all bills and certificates plus
notes and bonds which have less than 1 year to final maturity.
Since the publication of the ad hoc subcommittee report, the phrase
"depth, breadth, and resiliency" has been used in discussions of the
dealer market. This phrase is defined in the subcommittee report and
we quote: 5
In strictly market terms, the inside market, i.e., the market which is reflected
on the order books of specialists and dealers, possesses depth when there are
orders, either actual orders or orders that can be readily uncovered, both above
and below the market. The market has breadth when these orders are in
volume and come from widely divergent investor groups. It is resilient when
new orders pour promptly into the market to take advantage of sharp and unexpected fluctuations in prices.

The market or a particular issue or maturity range is characterized
as "thin" when depth, breadth, and resiliency are not characteristic of
operations in the market, issue, or maturity range. There is a "wide
gap between the prices at which the least firm holders are willing to
sell and potential buyers are willing to purchase." Quotations fluctuate widely "either in response to relatively small buy or sell orders, or
more frequently, as a result of professional
efforts to stimulate interest
by marking quotations up or down."6
The "bills only policy" refers to that policy of the Federal Open
Market Committee adopted and abandoned in the spring of 1953 and
readopted in the fall of 1953 which has led to the practice of executing
purchases7 and sales for the account of the Committee mainly in Treasury bills. While exceptions have been made, this policy has prevailed
since September 1953.
5
Hearings, op. cit., December 1954, p. 265. See also hearings, op. cit., 1959, pt. 6C, pp.
1922^-1926 for the definitions offered by the 17 dealers.
*7 Hearings, 1954, p. 266.
A recent statement of the rationale for this policy is: R. A. Young and Yager, C, "The
Economics of 'Bills Preferably'," Quarterly Journal of Economics, August 1960.




STUDY OF FEDERAL GOVERNMENT SECURITIES DEALERS

5

T H E DEALER MARKET AND THE SAVING-INVESTMENT PROCESS

The Government securities market differs from other securities
markets in a variety of ways. I n several of the chapters which follow
specific characteristics of this market will be considered. We will find
that the size of the market (measured by the value of transactions),
the rates at which dealers borrow, the amount of leverage, the cost of
making transactions, and numerous other operating characteristics
differ substantially from those found in other security markets. I n
combination these characteristics are sufficient to suggest that this
market be regarded as a unique institution.
From the viewpoint of the economy, the particular role which the
Government securities market takes in the functioning of the financial
system is of greater significance than its unique operating arrangements. The functions which are performed by this market—the role
which it plays in the saving-investment process and its use as a
mechanism for transmitting changes in the supply of money to the
economy—make the operation of the market of concern to those who
are not security dealers or specialists.
ARBITRAGE I N T H E DEALER M A R K E T

The dealer market for Government securities operates as a principal
bridge between the money market and the capital markets. A t one
end, Treasury bills are exchanged for cash. As the above discussion
of Treasury bills indicates, they may have as much as 1 year or as little
as 1 week or 1 day to maturity. The outstanding amount of such
securities has an average maturity of less than 3 months. Of all the
outstanding securities, they are the closest substitute for cash. This
is true not only because of their short maturity, or the guarantee of
principal and interest by the Federal Government but also because
they can be sold for cash in the market at most times, in virtually any
amount, with relatively minor price fluctuations.
A t the other end of the maturity range are the Treasury's long-term
bonds. These must directly compete for funds with issues of State
and local governments as well as with long-term obligations issued by
Government agencies, corporate securities, mortgages, and other longterm debt instruments. Each of these instruments must be priced to
attract funds from the capital market at the time it is offered. Thereafter, the capital market reevaluates a particular offering relative to
all other outstanding marketable obligations each time a purchase and
sale is executed.
Many of the securities offered to potential buyers in the capital
market do not trade actively. Some are privately placed and remain
in the portfolios of the original purchaser from the date of issue until
the final maturity is reached. Others are issued in small amounts and
can be sold only in limited quantities or with potentially large price
changes. Still others are traded more frequently; for these, an active
market is said to exist in which prices are realized in transactions and
quoted on a daily basis.




6

STUDY OF FEDERAL GOVERNMENT SECURITIES DEALERS

Because the money creating and taxing power of the United States
is a guarantee that both the interest and principal of Treasury securities will be paid when due and because Government securities dealers
maintain a market for Treasury bonds, purchasers are able to sell
some of their bonds at any time. As a result, such bonds are generally
regarded as more liquid than competing debt instruments. The
greater liquidity and lower risk are reflected in the difference between
the average yield on Treasury bonds and the average yields on obligations of States, municipalities, and corporations. On the whole, a
relatively constant difference in mean yield has been maintained between Treasuries and corporates, while marginal yields, measured by
the yield on new issues, have generally fluctuated in response to
changes in the demand for or supply of money.
The tax-exempt status of State and municipal bonds keeps their
average yield below the yield on Treasuries. However, the increased
stock of tax-exempts has been accompanied by a wider distribution
of outstanding issues. To increase the demand for tax-exempts by
those in other than the highest tax brackets, the average yield on
municipals has risen more rapidly than the average yield on Treasuries and the difference between the two has narrowed during the
postwar period.
Arbitrage between the corporate, municipal, and Treasury markets
adjusts the yields on outstanding issues to general changes in the
economy and specific changes in the supply of outstanding issues.
Since many Government dealers are also active dealers in corporate
and municipal bonds, it is likely that they are among the active arbitragers. Moreover, increases in the stock of corporate and municipal
bonds outstanding have been accompanied by a decline in the absolute
amount of long-term Treasuries which insurance companies, banks,
and some other institutional investors hold. Even if the Treasury is
not issuing new long-term bonds, competition for funds by other issuers will temporarily increase the flow of outstanding Treasuries
to the dealer market and reduce their price.
Thus at one end of the maturity range, there are Treasury issues
which are close substitutes for money, while at the other end there are
securities which compete for funds in the capital market or markets.
I n between there are a variety of other types of securities as we move
from bills to certificates and short-term notes or bonds into intermediate and longer term securities. I n principle, each of these issues
should offer a yield such that none of the owners of a particular
issue would prefer to sell that issue to obtain any other issue at the
existing prices or yields. If a particular issue is priced relatively high
or low when compared to other issues, arbitrage could take place
profitably eliminating the discrepancy in yields.
The Treasury has a larger amount of outstanding debt and a
wider range of maturies than any other borrower in the economy. Of
course there are some gaps in the maturity structure and some special
features attached to particular issues which are not available on other
issues. These create differences which make Treasury securities less
than perfectly homogeneous for reasons other than differences in
length to maturity. Nevertheless, it seems reasonable to believe that
market trading will evaluate these differences and adjust the rates
to compensate holders for call provisions and other special features



STUDY OF FEDERAL GOVERNMENT SECURITIES DEALERS

7

provided that there is an opportunity to buy or sell a comparable
security which does not have the particular feature.
A t the shortest end, the Treasury issues dominate the market.
As the evidence in later chapters will show, "depth, breadth, and
resiliency" are characteristics of this market. Relatively large transactions take place at quoted prices and demand and supply schedules
exist on the books of dealers in all but the most disorderly markets.
The flow of securities to the market in response to rate changes is sufficiently large to permit speculators to quickly arbitrage differences in
yields between the various bill issues, between bills, certificates, and
other short-term Treasury instruments as well as between outstanding
short-term Treasury securities and short-term obligations of other
issuers. This suggests that the Treasury does more than supply a
stock of nonmoney liquid assets when it issues bills and certificates.
I t also provides an effective medium for the rapid adjustment of shortterm interest rates.
Dealers in Government securities arrange a large part of their financing through the money markets. As will be shown in more detail in
chapter V, participants in the dealer market effectively arbitrage between the money and short-term securities markets. Moreover, banks
view the purchase of very short-term Treasury bills as a substitute for
the sale of Federal funds and regard such bills as the closest substitute
for money. From their viewpoint, the Treasury bill market is a part
of the money market.
Arbitrage operations, as has been suggested, relate the average
yields on Treasuries with the average yields on other instruments.
Through the dealer market, the yields on Treasury securities of all
maturities are also related. As will be shown in later chapters,
depth, breadth, and resiliency are far less characteristic of the longterm market. A t times arbitrage operations cannot take place as
rapidly between short- and long-term issues and between long-term
Governments and other capital market issues as they do in the money
and securities market.
Nevertheless, the Treasury has outstanding a larger stock of actively
traded issues than any other institution. Maturity aside, these securities are more homogeneous than the obligations of any other issuer.
I t is likely, therefore, that much of the arbitrage between long, intermediate, and short-term rates of interest—or between the money and
capital markets—is accomplished through the Government securities
market. Certainly this has been true during part of the postwar
period. F o r example, the absorption by the market of a greatly
increased stock of corporate securities has been facilitated by the
reduction in the amount of outstanding long-term Government securities in the portfolios of insurance companies and others. Absorption
of these issues through the dealer market has been accomplished in part
by raising the yield on Treasury bonds from its existing level. Further adjustments, through intermediate and short-term yields, alter the
shape of the yield curve to the new equilibrium level of rates.
Should the Treasury seek to increase the average maturity of its
outstanding marketable debt and thereby affect the relationship between short- and long-term yields? Much of the present practice
and discussion of debt management is predicated on the belief that it
should. But, many fail to recognize that the Treasury has been principally engaged in refunding debt during the postwar years. With
62471—60*
2



8

STUDY OF FEDERAL GOVERNMENT SECURITIES DEALERS

the exception of a few years of recession and the Korean war, new
cash borrowing has been a relatively small part of Treasury debt
operations.
Attempts to increase the average maturity of the debt during ref undings alter the structure of interest rates by increasing the supply
of longer maturities while reducing the supply of shorter maturities.
To extend the maturity of the debt, the Treasury must attract into
the long-term market some of the money released at the short end of
the maturity structure or an equivalent amount of cash from elsewhere. Like corporations which issue long-term bonds, the Treasury
must pay more than the average yield on outstanding issues, if it
is to sell a new issue of bonds. I n periods of tight money, the differential between newr issue rates on Aaa corporates and the average
yield of highest grade corporate bonds has been as much as one-half
or 1 percent. This premium is the price paid to attract additional
funds to the capital markets from other uses. I t rises when money
is tight and short-term rates approach long-term rates.
Judging from the performance of the corporate bond markets, the
differential between the rate paid on new issues and the average yield
on outstanding securities will fall during recession. However, if the
Treasury takes advantage of the opportunity to extend the average
maturity of the debt during recession, it widens the differential between short-term rates and long-term rates and slows the downward
adjustment of interest rates in the capital markets. As a result, it
may reduce the amount of capital investment and hinder the readjustments of the economy.
Since the outstanding marketable debt remains relatively constant
and the Treasury has chiefly refunding operations to consider, the
management of the public debt could be directed toward improving
interest rate adjustments between short- and long-term markets.
Refunding operations release money at one end of the maturity structure and absorb it at other points. No change in the money supply
takes place. With appropriate debt management there need be little
change in the level of interest rates or the differential between short
and long rates as a result of the refunding operation. This could be
accomplished, if desired, by abandoning efforts to increase the average maturity of the debt and concentrating on maintaining a constant
maturity structure.
If the quantity and the maturity composition of the debt are roughly
constant, refunding operations have little impact on the structure of interest rates. Changes which occur will be adjusted by
arbitrage, provided only that the stock of securities in each maturity
class is sufficient to permit "depth, breadth, and resiliency" to be
characteristic of trading in each sector. Moreover, by maintaining
almost constant maturity composition and improving the opportunities for arbitrage, the Treasury makes a greater contribution to the
adjustment of the economy. When the level of interest rates changes,
e.g., as a result of monetary policy, arbitrage along the yield curve will
change the structure of interest rates in response to the demands of
the private economy. If lenders expect inflation, efforts to sell longterm and buy short-term instruments will quickly change the structure
of rates and adjust the yields upward to reflect market considerations.
The dealer market then will transmit interest rate changes up
or down the structure of rates. Through arbitrage between the



STUDY OF FEDERAL GOVERNMENT SECURITIES DEALERS

9

dealer market and the money and capital markets, new savings flow
to the sectors of the capital market most attractive to particular lenders. The dealer market serves as one of the principal mechanisms by
which the interest rate structure more rapidly adjusts the flow of
savings into investment in response to the demands of borrowers and
lenders.
The potential role of present debt operations is in marked contrast
to functions of Treasury debt management during the years of war and
depression. Then the amount of new cash borrowing greatly exceeded
the amount of refunding. The rate which the Treasury offered to the
market had to be sufficient to attract new savings from the economy
or to activate idle balances. The money which the Treasury obtained
was used to pay deficits and to finance public works projects. The
money supply was increased. Under such conditions, the impact on
the level of rates or the question of competition with private investors
were important considerations in deciding upon the maturity of new
Treasury financing. And the role of the Treasury was principally
one of providing funds for new Government investment or new military expenditure.
Increasing the flexibility of interest adjustments requires improvement in opportunities for arbitrage. The extent to which arbitrage
remains imperfect will be considered at a later point. Enough has
been said to suggest that evaluation of the present role of the dealer
market for Government securities in the economy largely reduces to
analysis of its performance in providing a means through which interest rate changes can be rapidly spread from one sector of the financial markets to the others. For it is through arbitrage operations
that the effects of monetary policy are carried up or down the interest
rate structure and rates are allowed to allocate the flow of savings
into particular investment channels in response to the demands of the
private economy, states, municipalities, and others.
SCOPE OF THE STUDY

This study is based on a series of financial reports which the 17
dealers in the Government securities market voluntarily completed
at the request of the Joint Economic Committee. I n addition, several
dealers were able to supplement their own records with information
previously submitted to the Federal Reserve Bank of New York.
Additional material and substantial help in interpretation of accounting records was obtained through interviews with more than twothirds of the dealers as well as with officers of the Federal Reserve
Bank of New York and the staff of the Board of Governors of the
Federal Reserve System.
The main focus of the study is the market in which outstanding Treasury issues are traded, and by far the larger part of the
discussion is a description of the way in which the market operates.
Chapter I I concentrates on market organization, differences between
individual dealer operations, and dealer views of their own operations.
Succeeding chapters consider in turn each of the financial reports submitted by the 17 dealers: position, transactions, sources of financing,
capital, and earnings. A t the end, a brief summary of some of the
principal findings is provided and some questions are posed concerning
the role of the market in the economy.






CHAPTER II
THE DEALER MARKET: TRADING ORGANIZATION AND
PHILOSOPHY OF DEALER FIRMS
The internal organization of dealer firms is closely related to the
size of total operations. I n five cases, Government securities operations are conducted in a relatively small department of a large bank.
For some nonbank dealers other activities—dealing in corporate or
municipal bonds, underwriting new issues, or trading on organized
exchanges—are regarded as the principal occupation of the firm.
Size of the dealer operation in Government securities is a more
important determinant of the manner in which Government securities
are traded than the number or kind of alternative activities in which
the firm engages. Large dealer organizations are more than simple
multiples of smaller firms. I n general, larger firms (1) provide additional services to customers, (2) have greater access to information on
changing market conditions, and (3) perhaps most important of all,
have different philosophies which guide their operations.
For the benefit of those readers who have not had the opportunity
to witness bond trading, a brief description of some aspects of internal
organization is provided. Readers familiar with the operation of
the market may prefer to go directly to the section in which some
of the implications of the differences between large and small firms
are examined. Discussion of differences in the legal form of organization will be found in chapter V I .
INTERNAL, ORGANIZATION FOR TRADING

For the purposes of this study, the dealer market may be regarded as
composed of the 17 firms listed earlier. Five are departments of large
New York or Chicago banks; the remaining 12 are firms whose principal business is the purchase and sale of securities. A t various times
since 1947 at least six additional participants in the market 1 have reported on their operations in Government securities to the Federal
Reserve Bank of New York. Two of the six have been active during
the last 4 or 5 years, but their scale of operations in Government securities has been quite small.
Virtually all transactions in Government securities are conducted
by telephone or teletype. The trader operates from a desk equipped
with a series of direct connections with the trading desk of the Federal Reserve Bank of New York and each of the other dealers. In
addition, there are a number of other telephone lines connecting the
sales department with banks, insurance companies, and other customers. 3 One telephone switchboard and a blackboard, for recording
1
Northern Trust Co., Blair & Co., J. B. Roll, Schroder Rockefeller & Co., Harvey Fisk &
Sons,
and Malon S. Andrus, Inc.
2
In a large firm customer calls are usually handled by a salesman. The "trader" deals
almost exclusively with other traders.




11

12

STUDY OF FEDERAL GOVERNMENT SECURITIES DEALERS

current quotations and direction of price change for the various issues
of Government securities, constitutes all of the equipment used by
a small dealer. I n a bigger operation, a large number of identical
switchboards will be manned by an equal number of specialists each
responsible for trading a particular class of maturities or group of
issues. A "head trader," frequently a principal of the firm, supervises
the operation from a central desk. Large or difficult transactions and
most of the purchases from and sales to the open market desk at the
Federal Keserve will be handled by him. I n large dealer firms, there
are several principals—partners or officers or stockholders as the case
may be. Typically each supervises a particular maturity range.
An electric quotation board may replace the blackboard of the small
dealer. Last prices or quotations on all issues of Government securities, and the highest and lowest quotation of the day are recorded below the closing price for the previous day's trades. While the small
dealer may only post a few selected issues in which he is actively interested or which are close substitutes for those he owns, the large dealer
keeps a current record on the entire list of Government securities and
related issues—agency and other guaranteed obligations.
As trading proceeds, purchases and sales are recorded on a sheet directly in front of the trader. The amount of outstanding repurchase
agreements and the dates on which the securities will be redelivered
are noted there also. I n this way, the trader is kept aware of the
precise amount of securities which he has available, the amount which
he is "short," and the amounts which are due to be delivered on subsequent trading days.
I n large firms, a "money desk" in the trading area is responsible for
arranging the financing of the company's position in Government
securities or in related securities. A copy of each purchase or sale
order is furnished to the money desk. A record is maintained of the
exact amount of additional financing needed to balance the daily cash
flow against the daily flow of securities. Kates and reserve positions
at money market banks and other large banks throughout the country
are checked; repurchase agreements are sought with corporations
throughout the country; interest rate changes in the money market
are noted.
I n a small dealer firm, financing will be combined with trading.
The "money market desk" is reduced to a column on the page which
records the company's net position. As securities are sold or purchased, the "amount remaining to be financed" will indicate the day's
change in cash position as a positive or negative amount.
Frequent "feedback" between the trading desk and the money desk
will influence the course of trading. If money is tight on a particular
day, prices will be shaded in an effort to dispose of securities for cash.
As a result, tight conditions in the money market are reflected in the
securities market through a fall in the price of securities. Through the
constant interplay between the trading desk, the money desk, and the
markets, changes are transmitted to the various markets on which interest rates are formed. Moreover, if trading in Government securities
is done in but one department of a multidepartment securities house,
there must be coordination between the trading in corporate and municipal obligations and the Government securities department. A t
least one of the firms which offers a wide range of securities, coordi-




STUDY OF FEDERAL GOVERNMENT SECURITIES DEALERS

13

nates its trading activities by placing the trading desks for the various
securities—corporate, municipal, and Government—in adjacent areas.
The money desk is used to balance the inflow and outflow of cash on all
markets. I t thus becomes the control center for a firm's cash flows to
or from a variety of securities markets.
Among the bank dealers, the problem of coordination is handled in
a similar way. I n many cases, the bank's portfolio, the dealer function, and the reserve or money position are handled through a single
vice president. He decides on the allocation of the bank's reserve position between the Government securities and portfolio or investment
departments. For most of the bank dealers this officer is the principal
means of coordinating activities in the various security markets. H e
allocates funds internally and borrows or lends Federal funds in the
money market. Differences in sources and availability of funds aside,
it is clear from the above that his function differs little in principle
from the operation of a money desk at a multidepartment dealer firm.
I n these aspects of internal organization for financing and trading,
large firms are simply more complex or more elaborate reproductions
of their smaller competitors. But, as noted above, large firms provide
additional services to customers and engage in operations which
smaller firms avoid. These services play an important part in the
functioning of the market and the transmission of interest rate
changes. They both reflect and condition the policy aims of dealer
firms.
CUSTOMER SERVICES AND INFORMATION FLOW

The trading room of any large dealer appears to the uninitiated as a
lace of hubbub and confusion. Quotations are shouted back and forth
etween traders, salesmen, and clerks. Runners carry messages between the various departments. The sales department and teletype
room, neither of which is generally found in a smaller firm, provide
direct contact with customers seeking to buy or sell securities. Sale
or purchase orders received at branch offices are processed.
Contrast to this the trading room of a small dealer. One, two, or
perhaps four or five men are quietly buying, selling, or quoting securities on the telephone or recording their position on their tally sheets.
There is much less apparent activity. And there are important differences in functions performed.
The large dealer is much more likely to be operating in a national
market. He frequently has large branch offices in various sections of
the country. And he is also more likely to have both telephone and
outside salesmen calling on banks, insurance companies and the treasurers of large nonfinancial corporations outside New York. Moreover, the large dealer will generally have an active and well-staffed
statistical department to analyze the changes taking place in the various parts of the country, their impact on the reserve positions of banks,
on interest rates and the price of Government bonds. I n addition, his
staff people collect itemized portfolios for active participants in the
market. He is thus aware of factors affecting the demand for securities in particular communities.
The information and contacts obtained in these ways not only directly assist the larger dealer to maintain a greater volume, they also
provide information which is useful for two important purposes.

E




14

STUDY OF FEDERAL GOVERNMENT SECURITIES DEALERS

First, when a customer is seeking to purchase or sell a particular longterm issue which is not actively traded, knowledge of the present
holders of the issue and an awareness of their portfolio requirements
is a prerequisite to servicing the order. I n the present long-term market, many issues are not actively traded for reasons discussed in more
detail below. Easier access to information permits the larger dealer
to service the order by offering to exchange securities with some other
customer, borrow the securities which a particular customer desires,
or otherwise arrange to fill the order. Through these means some of
the disadvantages of a "thin" long term bond market are overcome.
Dealer information on customer portfolios eliminates some of the
risk involved in holding "thin" issues and helps to prevent infrequently traded issues from selling at a price which is low relative to
comparable maturities.
Second, greater access to market information permits the large dealer
to take advantage of temporary situations which arise in other parts of
the country. Through his branch qfEces and his more frequent contacts with firms throughout the country the large dealer is made more
quickly aware of minor but important changes in local or national
conditions. F o r example, the forthcoming large sale of a particular
issue by a pension fund would temporarily reduce the price for that
issue. The dealer may therefore wish to dispose of his position in
advance. A result of this has been a reduction in the number of dealers
who actively trade longer term securities for which markets are
"thin" and the risk involved in holding a given dollar amount of securities is great.
One small dealer summarized his role in the market by stating that
he was generally unable to assume the risks of a long-term position. If
he tried, he felt sure that he "would end up 'holding the bag' for one
of the bigger dealers. H e [the large dealer] can find out what's happening in Cleveland, Detroit, or on the west coast and unload his
position on me long before I know what is going on." As a result,
some smaller dealers operate much less or not at all in long-term
maturities.
More important is the fact that the small dealer tends to think
of his operation as concentrated principally but not exclusively in
New York. The large dealer is more likely to regard his operation
as national in scope. There is no rule which clearly delineates the two.
Even small dealers were anxious to point out in interviews that they
had long established borrowing arrangements with the X Bank in
Seattle or San Francisco. But, in discussing their trading activities
they were much more likely to choose examples based on interdealer
trading or sales to and from New York money market banks and
financial institutions. Moreover, questions about the scope of activities generally indicated that the New York market was relatively
more important for small than for large dealers.
The foregoing does not imply that the market is "dominated" by
the large firms or that the small firms are without important functions.
The role of smaller firms in this market is discussed later. However, we wish to suggest here t h a t : (1) the functions performed by
dealer firms differ, (2) size of firm is a rough indicator of this difference, and (3) large firms incur additional costs and undertake
additional risks in the interest of increasing their trading volume,
profits and service to customers.



STUDY OF FEDERAL GOVERNMENT SECURITIES DEALERS

15

DEALER VIEWS OF DEALER FUNCTIONS

Making markets
Conversations with dealers and written material about the Government securities market are replete with references to the dealer function of "making markets." Usually this refers to the well-known fact
that the market for Government securities is an over-the-counter
market, and that securities dealers clear the market by buying (absorbing securities into their position) or selling (eliminating securities
from position or, alternatively, shorting the market). When interviewed many dealers seemed proud of the way in which this function
is performed. Descriptions of their relationships with customers
often refered to the alleged fact that "we make good markets" or "we
are willing to stand on our market." Care was taken to make a distinction between dealers who as principals buy and sell securities for their
own account and "take positions" (hold an inventory of Government
securities) and brokers who arrange transfers between buyers and
sellers but do not take positions.
But the repetition of such phrases aside, there was much disagreement about the function of making markets when dealers were asked
to describe their function in detail. At one extreme there was a group
of firms, generally the larger firms, who asserted that markets were
made almost exclusively for customers. If a security is quoted to a
customer, the dealer will be willing to buy or sell at the given "bid"
and "asked" prices. However, the extent to which these quotations
may be regarded as firm commitments appears to depend directly on
the dealer's present or desired inventory position and inversely on
the years to maturity of the issue. I n the case of short-term securities
most large dealers describe themselves as willing to "stand on their
market" for almost any amount that a customer offers. But, in
the event that a customer desires to sell a relatively large amount
of a long or medium-term issue, e.g., $5 million, a large dealer will
normally buy some proportion of the offer, say $y2 or $1 million. He will then try to dispose of the remaining amount through
interdealer trading or after inquiry, among those customers who
are likely to have a portfolio need for the issue or maturity range.
The dealer will normally not take the remaining $4 million or
$414 million into position until he is able to arrange for the sale of
these securities to a customer. On occasion an infrequently traded
issue is offered. Even the largest dealers indicate an unwillingness
to take such "thin" issues into position until an offsetting sale has
been arranged.
Large dealers also emphasize what they describe as a service
function. If a customer desires a particular issue, the dealer will
scour the known holders of the issue and try to arrange for an exchange or "swap" with a holder in order to release the securities
desired by the customer. Or if the security is actively traded, the
dealer will sell the security short or hedge the transaction. To do
this, he borrows the security and "goes short" to his customer. He
restores his inventory with an available issue in the same maturity
range. Thus the dealer services the customer by temporarily furnishing the market with a security which is in short supply at the existing
market price. At a later date, the hedge transaction is reversed. The



16

STUDY OF FEDERAL GOVERNMENT SECURITIES DEALERS

dealer purchases the security which was sold short and sells the
security which was purchased.
While there was some consistency in the views expressed by the
large dealers, there was far less unanimity of opinion among the
smaller dealers with respect to the dealer's function in the market
and the definition of "making a market." Three points of view cover
the range of expressed opinions. Two of them reveal a greater orientation toward the dealer market than to the service of customers.
One dealer described himself as "the dealer's dealer." This does not
mean that he refuses to sell to nondealers but he was quite explicit
that he views himself as a specialist who makes a market for other
dealers. His method of operation consists primarily in bidding for
bills at the weekly Treasury auction, taking them into position and
disposing of his award during the week. On occasion he buys bills
from other dealers and sells them to customers or vice versa. But his
principal occupation seems to be winning an award of longer term—
180 days or 1 year—bills and selling them off during the course of the
week to other dealers. H e views his function as a means of providing
a portfolio of bills available to other dealers who are more directly
engaged in servicing customers.
One long established dealer describes his primary function as making markets for other dealers. He denied emphatically that he "stands
on markets" for customers. Moreover, he does not view the bank dealers as an integral part of the market and does not feel it necessary to
make markets to them. H e indicated that in his firm generally the
customer is asked about the dollar amount of securities which the
customer wishes to purchase or sell. The trader then decides whether
or not to quote a price. The decision seems to be based principally
on the dealer's current and desired inventory.
A third small dealer describes himself as a "merchant of securities."
I n general, he does not take very large positions in Government securities. Practically all of his trading is with customers rather than
with other dealers and most of this is done in bills. Like the larger
dealers, he views his firm as a service organization. But unlike them,
the process of "making a market" is described as one in which customers are contacted, their needs are discovered, and securities are
purchased to fill their demand. Aside from the small position which
is normally carried by the firm and the alleged fact that the firm never
charges commission or other fees, there is little difference between
the "merchant of securities" and a broker.
The bank dealers do not differ substantially in their operations
from other dealers. Banks with large dealer departments tend to
hold larger portfolios in medium- and long-term securities; the smaller
departments are concerned more directly on the average with bills,
certificates, and other short maturities. However, it is likely that
transactions with commercial banks are a larger part of total trading
for bank dealers than for nonbank dealers. For some of the bank
dealers, the opportunity to attract funds from banking correspondents
outside New York or Chicago makes their dealer operation attractive even during periods of "tight money" when numerous alternative
uses of their reserve positions are available. Moreover, some of the
bank dealers seem convinced that their dealer operation attracts
large corporate depositors since the department enables the bank to



STUDY OF FEDERAL GOVERNMENT SECURITIES DEALERS

17

effect securities transactions for corporations more rapidly than can
banks which do not have dealer departments.
Smaller firms tend to concentrate on a particular range of maturities; larger firms are more likely to take positions over the whole
range of maturities. Some of the small firms have become bill specialists, others have concentrated on intermediate and long-term securities or agency and guaranteed obligations. I n part, this reflects
the problem, discussed in greater detail in chapter I I I below, of
maintaining a position in a wide variety of issues when capital is
limited. I n part, it is a reaction to a previously unprofitable operation over the whole range of the Government securities market. A
variety of other reasons could be cited. I n general they fall into
three main classes:
1. Information.—Large
dealers have large sales organizations, frequently have branch offices throughout the country, and are generally much more aware of the conditions and operations throughout the country. This point was discussed above and need not be
elaborated further.
2. Risk.—The existence of long bonds in the portfolio means that
the dealer is carrying a greater risk per dollar of portfolio. This
point will be discussed in more detail in chapter I I I .
3. Cost.—Average size of transaction varies inversely with the maturity of the issue. Most dealers prefer to operate in a market in
which the average size of transaction is relatively large. Where large
dealers are more willing to incur the cost of handling small orders, it
is in the hope of building up long-term customer goodwill. Bank
dealers handle small orders to build goodwill among their correspondent banks and customers. Small dealers are somewhat less willing to handle small lots.
What then is the function of the small dealer in "making markets"? By taking positions, at least over some range of maturities,
they assist in clearing the market, establishing an equilibrium price
for Government securities and hence a structure of interest rates. As
active participants in the market, they seek profitable arbitrage transactions between dealers and thereby assist in adjusting the spreads
between bid and asked prices in various maturity classes. They "hit"
the bid and offer quotations of the larger dealers, e.g., by selling short
against a competing dealer's offers to buy or buying against his offer
to sell. I n this way, small dealers help to create an active, centralized
market in which prices are less subject to erratic fluctuations.
I n addition, the smaller dealers are particularly active in the "unwinding" or "unravelling" of positions. Transactions in Government
securities often are made in a variety of issues with a single customer.
F o r example, in the process of making a purchase, a dealer frequently
acquires several different issues. Some of these may go into his position; some may have been purchased to complete a sale; others will
increase his inventory of particular issues beyond the level which he
wishes to maintain. A dealer can attempt to dispose of the latter
group of issues through interdealer trading. Several additional
transactions may be required to dispose of the securities and arrive at
a desired portfolio composition. This process of portfolio adjustment, known as "unwinding," takes place principally within the
dealer market and is accompanied by much negotiation. Small deal-




18

STUDY OF FEDERAL GOVERNMENT SECURITIES DEALERS

ers assist in the unwinding process by taking positions in the issues
offered or by selling issues which are required to complete a sale.
Thus, at the risk of greatly oversimplifying, we view the function of
"making markets" as composed of twTo parts: (1) servicing customers
throughout the country and (2) absorbing or offering securities in
response to changes in price so that the flow of securities is equilibrated between buyers and sellers. Smaller dealers appear to function principally but not exclusively in the latter area. I n addition,
they help to "turn the market" by allowing sellers to exchange securities for cash and buyers to exchange cash for securities within relatively small price ranges.
Interdealer trading
Estimates of the magnitude of interdealer trading are not available
in the records which have been examined. Crude approximations obtained during interviews with some of the large and small dealers
varied considerably as the foregoing sections might suggest. These
estimates ranged from a minimum of 10 to 15 percent in the case of
one large dealer to a maximum of 90 percent for one of the small
dealers. These estimates tend to support the suggested difference in
orientation between large and small dealer firms.
The amount of interdealer trading will, of course, vary over time
as market conditions change. One such change in recent years was
the elimination of the 100 bond agreement in 1956. Under this arrangement, the dealers agreed to stand willing to buy and sell to
each other at least 100 bonds ($100,000) at the prices which they
quoted. As a result, there is reported to have been a large amount
of telephoning between dealers and frequent requests to "run the
list" (quote bid and offer prices on all outstanding issues). Several
dealers viewed these frequent quotations as a form of harassment
which tied up substantial amounts of their traders' time. Moreover,
in the present "thin" long-term bond market, there wrere many issues
which dealers did not wish to hold in their positions. Both large
and small dealers complained in retrospect that competitors would try
to sell these issues to them and that much time would be wasted in
the process of acquiring and disposing of relatively small amounts
of long-term issues. Furthermore, abuse of agreements increased
turnover and tended to have an exaggerated effect on the prices of these
issues.
A general agreement of this kind is not likely to be revived in the
near future. None of the dealers who were asked about the prospects
for a new agreement felt that present market conditions make such
agreements viable. Some felt that cooperative arrangements of this
kind focused too much attention on interdealer trading, used a substantial amount of their traders' time and resulted in poorer service to
customers. Some indicated that they now participate in trading agreements limited in scope. Others suggested that such agreements were
incompatible with the active use of monetary controls and the accompanying large relative changes in interest rates and Government bond
prices. Most dealers are unwilling to relinquish the degree of control
over their own positions that such arrangements imply. Finally, many
dealers feel that the function which such agreements perform could
be better left to the market.




STUDY OF FEDERAL GOVERNMENT SECURITIES DEALERS

19

Size of transactions
Descriptions of the dealer market usually note that dealers are
wholesalers of securities. This means that the dealers trade principally with banks, insurance companies, pension funds, governmental
institutions, large corporations, or other dealers and brokers. But
while it is true that sales and purchases are made almost exclusively
with such institutions, the implication that banks and others act as
retailers is not always true. Frequently, the dealer carries out both
the wholesaling and retailing function, and the financial institution
acts as a salesman. 3 Illustrative examples of this occurred in 1959 and
1960 when the 4%'s of 1963 and the '5's of 1964 were first issued.
Many individuals, influenced no doubt by the spread in rates between
these Treasury notes and alternative rates at savings and loan associations, savings banks, etc., desired to obtain some of these new notes for
their portfolios. The dealers were asked to fill a large number of
relatively small orders which were passed on to them by banks, stock
exchange brokers, or savings institutions. I n general, these institutions did not attempt to pool the orders into larger units.
Many dealers were unable to estimate the number or average size
of transactions in which they participated during a year. And there
was little agreement about the definition of an "odd lot" though several dealers used this term to describe their operations. Nevertheless,
a rough indication of typical transactions can probably be gleaned
from the following:
1. Bills: $1 million is regarded as a small transaction; $25 to
$30 million is regarded as large.
2. Notes and bonds with less than 5 years to maturity: $1 million would be considered a "good" trade; $5 million is considered
a large trade.
3. Longer-term bonds: $% million (or perhaps less) is a typical
trade; $y2 to $% million would be considered a large transaction.
One dealer estimated that between 70 and 80 percent of his total
transactions by number w^ere for purchases or sales of $100,000 or less.
To defray some of the cost of handling these transactions, he now
charges an odd lot fee of $7 per ticket plus a fractional addition to
the price per bond when selling (or subtraction when buying). 4 Several factors suggest that this volume of small sales is atypical.
Nevertheless, it does indicate that dealers often engage in both a wholesale and a retail business contrary to common belief.
Most dealers deplore the existence of small transactions. Some
view them as a service which the dealers should perform. Others
stated explicitly that they make every effort to avoid them. I t is in
part as a result of the latter attitude that from time to time charges
arise that the dealers are unwilling to process small orders for customers or that some customers are unable to purchase or sell Government securities in the market. The recent growth in the use of odd
lot fees will probably eliminate or at least change the direction of many
of these complaints in the future.
3
Or, small buyers may execute some transactions through commercial banks. Many
banks
make no charge to customers for purchases of new issues.
4
Other dealers also have reported that "odd lot" fees are now charged, for example, %2
per $100. We did not inquire into the extent or specific nature of such charges, but there
is some evidence that they have not become standardized.




20

STUDY OF FEDERAL GOVERNMENT SECURITIES DEALERS

Short sales
I n the process of establishing market clearing prices, short sales
occur for a variety of reasons. First, such sales are undertaken by
dealers as one part of an arbitrage transaction. This may arise as part
of a customer service operation when a particular security is not available in the market at quoted prices. Or, they reflect a dealer's view
that a particular security is priced above others in the proximate
maturity range. Second, in the course of the day, a trader will often
sell a customer more of a particular issue than he currently has in
position, or he may sell securities which he does not have to complete
a sale of several issues. If the security is actively traded, he knows
that it can be replaced after the sale has been made, or hedged, through
buying an equal amount of a nearby maturity. Occasions arise on
which the replacement has not been made before the close of the
trading day or "unwinding" has not been completed. Short sales of
this kind are referred to as "technical shorts" and generally do not
last longer than 1 or 2 trading days. Third, dealers attempt to anticipate changes in the price of outstanding issues or changes in the
level of shape of the yield curve. The expectation of a fall in the price
of Government bonds (or a general rise in interest rates) will induce
some dealers to sell short on some issues. These can be labeled "speculative shorts."
Most dealers allege, and the statistics tend to confirm,5 that the
volume of dealer "speculative shorts" is not very large. Frequently
dealers suggested that selling short was "too expensive." Several
argued that they not only have to pay one-half of 1 percent to borrow
securities but that interest charges accrue against them daily. Hence
they reason that the expected fall in price has to be large enough to
compensate for the daily cost of borrowing the securities plus the
accrued interest. Only in the event of a very rapid fall in the price
of securities would this be likely to occur.
When securities are sold short, cash is obtained which may be used
to finance the dealer's position or to purchase collateral required by
the lender of the securities which are sold short. If the marginal
interest rate at which the dealer finances his position exceeds the cost
of carrying the short position, the net cash flow increases the dealer's
cash balance. Just as in the case of speculative long positions, the
net cost of carrying "speculative shorts" depends on the difference
between the interest rate on the securities held and market carrying
costs and the extent to which the dealer correctly anticipates the expected price change. But, at the long end, the carrying cost for speculative short sales generally is higher than in the case of speculative
long positions as the following example illustrates.
Suppose the dealer sells short on a long bond and invests the cash in
a bill or certificate. The rate on the long bond will almost always
exceed the rate on the bill or certificate and will rarely ever be less
than the rate of the bill by an amount equal to the cost of borrowing
securities (one-half of 1 percent). Thus the net interest cost to the
dealer will be positive in almost all cases. If the dealer invests the
cash obtained from the short sale in a long-term security at a rate
higher than the bill yield, he is performing an arbitrage transaction
along the existing yield curve and not a "speculative short sale."
5
See, for example, the "Treasury-Federal Reserve Study," pt. II, pp. 138-139, for the
week-to-week changes in aggregate dealer positions.




STUDY OF FEDERAL GOVERNMENT SECURITIES DEALERS

21

A more fundamental reason for the absence of "speculative shorts"
reflects the dealers' view of their function in the market. Genererally, they appear to be much less concerned with correcting or influencing the long-term trends in the bond market than with smoothing
short-run, day-to-day changes in prices. Their principal concern is
the servicing of customers or other dealers and the elimination of
erratic short-term movements in price. While there are clear exceptions to this generalization, the discussion of dealer investment accounts in chapters I I I and V I I will show that most dealers do not
take either speculative long or short positions for the purpose of
making capital gains. The combination of financing practices, which
have been developed to service the market, and tax considerations
mitigate against large-scale dealer speculative positions. Moreover,
the difficulty which dealers encounter when they attempt to borrow
securities, particularly long-term securities, further restricts the use
of "speculative short positions."
DEALERS, BROKERS AND LIEUTENANTS

I n recent years, increased use of Government securities brokers has
facilitated trading, particularly at the long end of the market. One
obvious reason for their growth was the breakdown of the "100 bond
agreements." Closely related is the thinness of the long-term market.
Brokers provide two services for dealers in the intermediate- and
long-term market. First, they permit interdealer trading to take
place without revealing the identity of buyer or seller. Second, they
provide a quotation service for dealers.
Dealers assert that brokers trade exclusively with dealers but do
not take positions in Government securities. Instead, for a reported
fee of one sixty-fourth they negotiate trades between buyer and seller.
The results of these trades are then reported to all other dealers in
the long-term market. I n these ways they fulfill many of the functions formerly achieved by interdealer trading agreements. And they
avoid the practice of introducing false bids into the market to obtain
information about the intentions of competitors, 6
I n the present "thin" long-term market, it is not surprising that
"false" bids or offers are made, Several dealers referred caustically
to the practice of using "lieutenants" for this purpose. I t is claimed
that this practice occurs most frequently when a dealer wishes to unwind a long- or medium-term position. The position which he has
acquired may have been originally offered by the customer to several different dealer firms. The firm which has executed the transaction is aware that as soon as it makes an offer of the securities,
whether directly or through a broker, the rest of the market will know
the composition, the amount, and the price of the issues, By employing a "lieutenant," often a stock broker, rather than a Government
securities broker some firms hope that their competitors are convinced
that the stock broker is handling an issue for a nondealer. Since
dealers usually quote a wider spread to customers than to other dealers
or brokers, the dealer holding the securities may be able to realize
a higher price from his competitor in this way.
«Brokers mentioned as active in interdealer trading included the following: Biggs &
Moffit, E. H. Hoffman, Rutberg, and G. A. Winter. Probably as a result of their activities it is possible to complete transactions at the long end of the market with smaller
fluctuations in the price of Government bonds.



22

STUDY OF FEDERAL GOVERNMENT SECURITIES DEALERS
T H E FEDERAL EESERVE AND THE DEALER MARKET

Some of the principal operations of dealers and the dealer market
involve the Federal Reserve System, particularly the Federal Reserve
Bank of New York. Here we are concerned primarily with the way
in which the market is organized to carry out its functions.
Among the more important services provided by the Federal Reserve System are (1) the auctioning of bills and the exchange of issues
when they act as fiscal agent for the Treasury, (2) a source of financing
of dealer positions through the use of repurchase agreements, (3) the
purchase and sale of securities for the open market account and for
the account of others, (4) one means of clearing or settling securities
transactions, and (5) the recognition of Government security dealers
with whom the Federal Reserve deals. The first two points have been
discussed in detail elsewhere and need not be further elaborated here. 7
Transactions with the dealers8
The Federal Reserve Bank of New York engages in two distinctly
separate types of transactions in Government securities. On the
one hand, it acts as agent for foreign central banks, the U.S. Treasury, and others. The purchases and sales must be effected without
interfering with the aims of monetary policy, and they must be made
at rates which reflect current prices in the market. On the other hand,
transactions for the System's open market account are designed
mainly to affect the availability of bank reserves in accordance with
the directives of the Open Market Committee. At times, the volume
and size of both types of transactions are relatively large and capable
of substantially changing the existing market rates. Care must be
exercised to assure that the two functions do not conflict. Moreover,
efforts are required to assure that certain transactions which are made
as agent are not regarded as indications of a change in monetary
policy.
A brief hypothetical example may illustrate the point: Assume that
an order to purchase $5 million of 5-year notes is to be executed
for a Treasury investment account at the end of a period of rapidly
falling prices; e.g., the fall of 1958. This might be interpreted as a
departure from the "bills only" or "bills usually" policy. How can
an explanation be given to the market without establishing a tradition
for such explanations ? How can the possibly unsettling actions resulting from such purchases be countered in a market which has just experienced a sharp price break ?
I n practice the Federal Reserve Bank of New York separates its
transactions as agent for others from its transactions for the System's
open market account. While it does not inform the dealers about the
origin of each transaction, this is accomplished by three devices all of
which are familiar to the dealers and unlikely to cause confusion.
7
F o r a discussion of t h e auction m a r k e t and t h e exchange or r e t i r e m e n t of issues see
W. L. Smith, "Debt Management in t h e United S t a t e s , " "Study of Employment Growth
and P r i c e Levels," 19, J o i n t Economic Committee, Washington, 1960, and hearings before
t h e J o i n t Economic Committee, J u l y 1959, pt. 6A, pp. 1148 ff. A discussion of repurchase
agreements is contained in "The Federal F u n d s Market," Bjoard of Governors of t h e Federal
Reserve System, Washington, May 1959, and R. V. Roosa, "Federal Reserve Operations In
the Money and Government Securities M a r k e t , " New York, Federal Reserve Bank of New
York,
J u l y 1956.
8
This discussion is based on practices prevailing in mid-1960. A discussion of pre-1953
operations is contained in the ad hoc committee report. Cf. Hearings before t h e J o i n t
Committee on t h e Economic Report, December 1954.




STUDY OF FEDERAL GOVERNMENT SECURITIES DEALERS

23

First, usually different people make the two types of transactions
at the trading desk in New York. Traders are arranged around a
horseshoe-type desk.^ On one side, sales and purchases are made
for the open market account; on the other, transactions in which the
Federal Reserve Bank of New York is acting as agent are effected.
Since all purchases and sales are made through the dealer market,
it is not surprising that the traders in the dealer houses become familiar
with the type of transaction which prompts the call. Moreover, if a
large transaction is to be made for the System's account, the managers
of the open market account will often call for the principal of the
dealer firm, since it is he who will ultimately decide on the price to
be quoted.
Second, the telephone conversation is different in each case. The
trader who is acting as agent is interested in buying or selling securities at the current market. A glance at the quotation blackboard at
the front of the room tells him the most recent quotations for each
issue given several times daily by four or five leading dealers. With
this information in mind, he calls one of the 17 dealers to effectuate
a transaction in much the same way as any other customer might call.
Over time, he attempts to distribute his calls among the dealers to
the extent that this is possible, given the differences in the composition
of their portfolios.
Purchases and sales for the open market account originate in
what is known as a "go-around." All dealers are invited to quote
and are asked to "stand on their market" for approximately 20
minutes; i.e., until the trading desk has decided upon the issues and
amounts to be purchased and sold. Dealers will quote not one but a
variety of issues and prices; e.g., several issues of 91-day or 182-day
bills under the "bills only" policy. Quotations are compared and the
managers of the account select those issues and prices which are lowest
if they are buying or highest if they are selling. Again, the dealer
need not guess to distinguish between transactions as agent and those
involving the System's account.
Finally, when acting as agent, the Federal Reserve Bank of New
York asks the dealer to calculate the aggregate amount of the purchase
or sale and to forward the memorandum to its offices within the day.
Transactions for the System's own account do not contain this provision.
To state that dealers can distinguish between the two types of transaction does not imply that as a result they always are able to identify
changes in the direction of monetary policy. The open market account
engages in a variety of offsetting transactions and occasionally diverges
from the above procedures. I t may wish to give temporary relief
in a technical situation arising perhaps from a sudden heavy flow of
reserves into or out of New York. Or, the managers of the account
may decide to execute a "turnaround" buying on the early days of the
week and selling in the latter part of the week. This operation may
reflect the desire to mask a forthcoming change in System policy or
to probe the extent to which quoted prices are firm on the bid and
9
A picture of the a r r a n g e m e n t mentioned here may be found in Carl H. Madden, "The
Money Side of 'The Street,' " New York, Federal Reserve Bank of New York, September
1959, p . 85.
62471—60
3




24

STUDY OF FEDERAL GOVERNMENT SECURITIES DEALERS

asked sides of the market. Numerous additional possibilities will no
doubt occur to the reader.
The separation of the two accounts for trading purposes is designed to eliminate some incorrect expectations which might arise as
a consequence of the Federal Reserve Bank of New York acting as
agent for others. This is particularly true when relatively large
transactions must be put through the market. The market will interpret the sale of $75 million in bills by a foreign central bank in a somewhat different way than a similar sale by the Open Market Committee.
As an extreme example, the sale by the foreign central bank might be
viewed as an expression of belief by foreigners that U.S. inflation
will continue and that foreign balances should be taken home in gold;
a similar sale by the Open Market Committee might be interpreted as
a sign of "tighter money." Both of these actions would lead to a
decline in the domestic money supply (or its rate of increase). However, dealers would be likely to view the former as temporarily destabilizing and the latter as a stabilizing influence on the day-to-day
movement of prices in the securities market.
Large transactions are more likely to be troublesome. I t is for this
reason that all of them are not put through the dealer market despite
occasional policy statements to the contrary. Relatively large transactions may be offset if the opposite transaction is available at the
Federal Reserve trading desk. The dealers believe that offsetting
transactions of this kind occur infrequently and arise only when
relatively large amounts must be traded in the market.
Clearing securities transactions
Since early 1948 telegraphic transfer of all outstanding marketable
debt can be made through the various Federal Reserve Banks if the
transfer is necessary to complete a sale. No fee is charged for Treasury bills, certificates, or notes and bonds with less than 1 year to
maturity. Nominal fees (e.g., $10 per transfer of a single issue whose
face amount exceeds $50,000) are charged in other cases. Most dealers
make use of this service,10 but at least one has found that it is an expensive way of delivering odd lots, particularly when several issues
are purchased by a single customer.
The general manner in which a transaction between a dealer and
customer is processed is illustrated in the following example. The
customer notifies the dealer by telephone and confirms by wire that
he wishes to make a purchase of $1 million in a particular bill issue.
The dealer quotes a price which if satisfactory is accepted. If the
customer is located outside the second Federal Reserve district, payment is transferred through the Federal Reserve System to the Federal Reserve Bank of New York which usually credits the account
of the Manufacturers Trust Co. Meanwhile, the bills have been
delivered to Manufacturers Trust Co. or the Marine Midland Bank
and by them to the Federal Reserve Bank of New York. The
bills are retired in New York and a wire is sent to the Federal Reserve bank in the purchaser's home district which reissues the bills
to the purchaser. If the purchaser resides within the second Federal
Reserve district, the Federal Reserve Bank of New York delivers the
securities directly.
w There are limitations on? the use of the facility in connection with new issues and
redemptions and on interest payment dates.



STUDY OF FEDERAL GOVERNMENT SECURITIES) DEALERS

25

A number of variants on this procedure are common. The customer may prefer to have the securities delivered to a bank of his
choice in New York. Or, the customer may make a repurchase agreement instead of an outright purchase. Transfer of securities and
payment in Federal funds may be made on the day of the transaction;
i.e., a "cash" basis, or on a regular, i.e., overnight, basis. 11 Other
transactions may call for delayed delivery if the customer wishes to
avoid payment for a fewT days. The number and variety of transactions within this general framework is large and testifies to the
agility and skill with which dealers and their customers have
developed methods for accommodating a multitude of differing
requirements.
The Manufacturers Trust Co. acts as clearing bank for most—but not
all—nonbank dealer transactions. They record the debits and credits
to the dealers' accounts in both securities and money and arrange for
transfers. For this service they collect a clearing charge—currently
$10 per million for bills, $15 per million for certificates, $35 per million
for longer maturities, and $0.15 per bond for municipals. More
recently two competitors—the Marine Midland Trust and the Irving
Trust—also have entered the clearing business.
Generally bank dealers and some of the larger nonbank dealers
arrange for at least some of their own clearing. As we shall see in
greater detail in chapter V, clearing is closely connected with the
financing of dealer positions since the transfer of securities is matched
by a flow of funds. Bank dealers prefer to have the debits and credits
at the Federal Reserve made directly to their own accounts. Moreover, their financing problems differ from those of the nonbank dealers
and particularly those of the smaller nonbank dealers.
Recognizing Government securities dealers
I n principal, anyone may become a dealer in Government securities
if he has clients who are willing to execute their purchase and sales
orders through his firm. In practice, entry into the business is
feasible but somewhat more difficult.
One important restriction comes from the financial side. I t is not
essential that a prospective dealer invest a relatively large sum in his
business, but it is essential that he have established arrangements for
financing his position. This restriction largely reduces the set of
prospective entrants to those who are "known" in the money market.
As a result most of the entrants to the market have come from two
sources: those who were established traders of a similar instrument,
e.g., municipal bonds, and those who have acquired their experience
as a trader for one of the established Government securities dealers.
The latter group appears to have furnished more of the entrants over
the years.
A second limiting factor comes from the Federal Reserve. New
dealers can, of course, trade without any "approval" from the Federal
Reserve, but to do so they must forgo important privileges. First,
they are not invited to trade with the System account and as a result
sacrifice a volume of transactions in bills which is reliably reported to
average as high as 5 percent of the aggregate bill volume of the market
during some years. Moreover, for a small bill specialist, volume with
ii Qf "The Federal F u n d s Market," op. cit., especially pp. 4 5 - 4 8 .




26

STUDY OF FEDERAL GOVERNMENT SECURITIES DEALERS

the Federal Reserve's trading desk has amounted to as much as 10
percent of his total annual transactions in bills. Second, they would
not have the privilege—reserved to nonbank dealers—of entering into
repurchase agreements with the Federal Reserve. This privilege is
not negligible and has provided as much as 70 percent of total financing for a nonbank dealer in recent years. 12
Federal Reserve requirements are not designed to restrict the number of dealers. Their purpose appears to be that of assuring that the
firms admitted to trading with them are financially able to fulfill their
commitments and that they are willing to undertake relatively large
commitments. To assure dealer solvency, up-to-date financial statements must be submitted periodically. More recently the additional
requirement has been added that dealer firms must report in detail
their volume of transactions, repurchase agreements, positions, and
borrowing on a daily basis.
The third and probably most important restriction on the number
of dealers is the profitability of the dealer business. This point will
be discussed in greater detail in chapter V I I . There we present some
data on dealer earnings during recent years and compare the rate of
return for nonbank dealers with those which occur in some other
branches of the securities business. The evidence seems to suggest
that prior to 1957 the relative profitability of the present dealers in
Government securities was not sufficient to attract sizable financial
commitments to the business from dealers in corporate or municipal
bonds. The availability of equally profitable or more profitable opportunities in other branches of the securities business combined with
the high degree of risk inherent in trading in Government securities
limits the expansion of the number of dealers. This is probably more
important than the more frequently cited reason—lack of trained
personnel.
12

See table V-3 below, p. 83.




CHAPTER III
DEALER POSITIONS
I n the discussion of dealer philosophy, it was noted that central
to the dealer view of his function was the fact that dealers hold inventories, that is, take positions. While there are divergent and changing views about the extent to which this should be done, only
a few instances were found during the years covered by this study
(1948-58) in which there were dealers with negative or zero inventories. I n this chapter the available data on dealer positions are presented after a discussion of the problems associated with these aggregates. The relation of dealer inventories to the outstanding marketable debt and to the rate of interest are then considered separately.
I n a concluding section, an attempt is made to summarize and assess
the changes which have taken place in the way in which dealers perform their function of making markets.
DEFINITIONS AND PROBLEMS I N THE MEASUREMENT or DEALER
POSITIONS

There are, unfortunately, few well-defined and generally accepted
accounting conventions for this market. Divergent opinions and practices are far more common than agreements. Some of these differences
apparently reflect disputes about appropriate accounting practices;
others mirror differences in performance and operation. Such differences are reflected in the reports of position which dealers supplied
to the Joint Economic Committee and in the supplemental data which
several dealers furnished from the records which they had voluntarily
submitted on a daily basis to the Federal Eeserve Bank of New York.
WHAT IS TO BE INCLUDED?

Government guaranteed issues, agencies, and others
Obligations of the U.S. Government greatly exceed the total amount
of Government securities. There are nonmarketable issues which are
direct obligations of the Treasury and there are marketable issues
guaranteed by the Treasury but issued by others—e.g., Panama Canal
bonds. I n addition there are outstanding nonguaranteed obligations
issued by Federal Home Loan Banks, Federal Intermediate Credit
Banks, the Federal National Mortgage Association, to name but a few.
Several dealers in Government securities regularly take positions in
these issues as well as in International Bank bonds. Many of these
obligations are close substitutes for Government securities. Nevertheless inventory holdings of these obligations have not been included in
the total position in governments reported here.
Municipal and State obligations have likewise been excluded. However, dealer positions in municipals and nongovernments are reported




27

28

STUDY OF FEDERAL GOVERNMENT SECURITIES DEALERS

separately below for purposes of comparing the changes over time in
dealer positions in these issues with changes in dealer inventories of
Government securities. For our purposes dealer positions in agency
and guaranteed issues are included with their position in nongovernment securities.
These figures must be read with care, however, since the group of
dealer firms included in this study is quite heterogeneous. One firm
for example is a large underwriter of corporate issues and takes positions in these issues. Such issues often are held in the underwriting
department but could not be effectively separated from trading positions so were included in the nongovernment total. Bank holdings
of corporate or municipal issues are not generally included in the
dealer department and were not included in the total holdings of nongovernment securities.
Another firm is a large dealer in a variety of nongovernment
securities. They elected to separate agency and other Government
obligations from the corporate and other holdings. Hence, their contribution to the aggregate figure for any year differs conceptually
from the figures supplied by some other dealers. Differences in
dealer accounting did not permit a uniform measure to be used for
all dealers. Instead, we attempted to provide a measure which was
consistent over time. Since dealer positions probably reflect the fact
that the amount of outstanding corporate debt has increased faster
than the amount of outstanding Government securities during this
period, it is likely that there is an upward bias in the proportion of
total nongovernment securities to Government securities presented
below.
Repurchase agreements
Dealers were asked to include repurchase agreements as part of
their position on the forms filed with the committee. However this
was not always done—given the dealers' existing records and the differences between bank and nonbank dealers. Moreover, there are differences of opinion about the appropriate ways to treat repurchase
agreements or E P ' s as they are generally known and there are different
kinds of EP's.
Some agreements permit the dealer to substitute another issue for
the one originally placed under repurchase. Most E P ' s are shortterm arrangements, often for overnight or over-the-weekend financing,
but some are for periods of 30 or more days. (The latter are often
a means by which corporations invest funds accumulated for tax
payments and seemingly vary with the proximity of tax dates.)
E P ' s are often referred to as "buy-backs" if they involve the use of
separate purchase and sale contracts. I n addition, there are reverse
E P ' s or "sell-backs," but these are never included in the dealer's position and hence need not concern us here.
There are three principal methods of handling E P ' s when positions are recorded: (1) All E P ' s regardless of maturity or other
characteristics are carried as part of the dealer's position. Several
dealers regarded all E P ' s as loans. They did not distinguish them
from collateral loans or other securities which were held in position.
They argue that all dealer loans are secured by collateral and that
such collateral is part of the dealers' position. F o r them, securities
held under repurchase agreements are analogous to the collateral post


STUDY OF FEDERAL GOVERNMENT SECURITIES DEALERS

29

ed as security for collateral loans. (2) A t the other extreme were
several bank dealers and one nonbank dealer. Repurchase agreements were never included in their position. I n the case of bank
dealers the amount of repurchase agreements is relatively small or
nonexistent. For the nonbank dealer, notes to the balance sheet indicated the amount of repurchase agreements and they were included in
the position figures given here. (3) An intermediate position is taken
by some dealers. An accounting difference may be based on the time
dimension of the R P contract or on the source of funds. The Federal Reserve and some dealers make an arbitrary distinction between RP's with 15 days or less and those with 16 days or more to
maturity. The former are treated as a source of financing to the
dealer and included in the dealer's position; the latter are considered
a sale and purchase by the dealer and are included in the dealer's
transactions but not in his position.
One large dealer defended this dichotomy on the grounds that
there is some point at which the dealer effectively loses control of
the security for trading purposes. I t is no longer available for delivery to a customer and does not enter into the trader's plans. Hence,
he argued, it does not belong in the firm's position.
Another dealer made a distinction based on the source of funds.
If the repurchase agreement is made with the Federal Reserve bank,
it is included in the firm's position and looked upon as a method of
financing the dealer's position. Since the Federal Reserve never
makes R P ' s for more than 15 days, this is a distinction as to time as
well as source of funds. But for this dealer, all other RP's, irrespective of the time of the contract, are regarded as purchases and sales
and hence excluded from the firm's position during the period of
repurchase agreement.
Fortunately, the dollar value of repurchase agreements which run
for 16 or more days is relatively small. The total amount of repurchase agreements which are excluded from the position as a result of the differences in dealer accounting practices are also small.
But there is an underestimate in the aggregate position figures particularly for more recent years as a result of these omissions. Since
the bulk of repurchase agreements are made in Treasury bills, the
absolute magnitude of the bias is undoubtedly largest in the bill
inventory.
Executed versus commitment
reporting
When should a dealer consider a particular security as part of his
position? When should securities be taken out of the dealer's position ? The answer given by most dealers is that securities available to
be sold are part of a dealer's position. Whether a dealer has obtained
physical possession of the securities is not, in this view, the important
question. The securities may have been purchased to cover a technical
short position, or they may be sold shortly after the dealer has made a
commitment to purchase. Most dealers include commitments to buy as
part of their position and exclude from their position securities which
they have agreed to sell. This method of recording purchases and
sales is referred to as "commitment reporting."
If the firm using commitment reporting draws up a balance sheet
at the end of a particular day, securities which they had agreed to




30

STUDY OF FEDERAL GOVERNMENT SECURITIES DEALERS

purchase would be included in their position. They would be offset by a liability called "securities purchased but not yet received."
When the securities are delivered, this account and the cash account
are reduced by the amount of the purchase. Sales are treated in an
analogous fashion. When the sale is made, the position is decreased
and the, asset account called "securities sold but not yet delivered"
is increased. The receipt of cash increases the cash account and decreases the "securities sold" account.
One of the large dealers records increases or decreases in position
only when the securities are delivered. His balance sheet has neither
an asset nor a liability account for securities which the firm failed
to deliver or receive. Thus his position figures are not recorded
on a basis consistent with those of his competitors. We will refer
to his method of reporting as the "execution basis" since only completed transactions affect his inventory.
Two points must be borne in mind. First, the difference between
commitment and executed reporting is not relevant for all transactions. A large percentage of dealer operations are for Federal
funds and are completed during the course of a single trading day.
Only "regular," that is, next day delivery transactions or delayed
deliveries are a source of difference in practice between the two methods of reporting. Second, for most years securities which dealers
have failed to deliver exceed the amount of securities which dealers
have failed to receive. 1 Hence, the execution basis of reporting results in an overestimate of the position of one large dealer and of the
aggregate dealer position. For most years, we would estimate that
this bias in the aggregate dealer position would be between 1 and 5
percent.
However, the dealer who reports on an executed basis is the same
dealer who excludes repurchase agreements—other than those with
the Federal Reserve—from his position report. The extent to which
these two differences in his method of reporting cancel is not clear.
But, it is clear that they work in opposite directions and thereby
reduce the overall error in the position data.
Investment accounts
The discussion above has been concerned principally with the
measurement of the dealer's trading position. But, some nonbank
dealers also take positions in securities with the object of establishing
a base for long-term capital gains. Securities held for this purpose
are segregated into an investment account which is carefully separated
from the dealer's trading account. Banks, of course, hold Governments in their investment accounts to a much greater extent than
nonbank dealers, but this is not a part of their dealer function and
thus is beyond the scope of this study.
The extent to which investment accounts are found on individual
dealer balance sheets varies with the profitability of dealer operations,
with the difference between the "yield" and "carry" of an individual
issue and the expected change in the price of bonds. Corporate dealers
who have experienced operating losses in past years find little or no tax
advantage in long-term gains, since prior losses can be carried forward
1

tee.

This is true for the dates on which dealers reported their balance sheets to the commitWe have not ascertained the extent to which it is true on other dates.




STUDY OF FEDERAL GOVERNMENT SECURITIES DEALERS

31

and charged against current profits as in any business. Thus the past
earnings of dealer firms have in several years reduced the number
of firms which held investment accounts. Moreover, tax considerations probably lead some corporate dealers to carry investment accounts on a personal rather than corporate basis.
A positive difference between "yield" and "carry" assures the dealer
that the cost of borrowing funds will be more than offset by the yield
of the securities in the investment account, A difference of one-fourth
of 1 percent is reported to be sufficiently large to indicate the possibility of investment account operations. Finally, a dealer must be
convinced that a change in rates will take place during the next 6
months.
Securities which are placed in the investment account are not available for trading purposes. However, a number of investment account
transactions are terminated before the end of the 6-month period
largely as a result of the failure of expected price changes to develop.
The securities may then be mingled with those in the trading account.
Despite the uncertainty which arises from the possibility just mentioned, we have kept the investment accounts and trading accounts
separate in the position statements reported here. Moreover, investments have been excluded from the statistics on dealer inventories of
outstanding Government securities. The latter decision is based on
the observed fact that investment in Government bonds (or speculation on movements in the rate structure over time) is not a function
which is peculiar to dealers. The investment account is of interest
principally as a measure of the extent to which dealers speculate on
future price movements, the maturities which are selected for this
purpose and the timing of dealer speculative activity.
HOW SHOULD POSITIONS BE MEASURED?

Net versus gross measures
Dealers were asked to report long, short, and net positions by maturity classes. With only a few exceptions, where short positions in
bills and certificates were not available, they complied with this request. However, the data presented here are based only on net positions.
A dealer's record of long and short position in an issue is not the
same as a record of gross position. The amount recorded as "long"
or "short" has in many cases already been "netted out." An example
will indicate how this occurs.
Suppose that on a given day a dealer has $2 million in a particular
bill issue out as a "buy back" subject to repurchase on the following
day. H e may sell all or part of the issue during the day to one or
more customers. Assume that a single sale for the full $2 million is
made with delivery to be made on the following day. If the position
statement is drawn that evening and if he uses either the execution or
commitment method of reporting, his statement would show a zero
balance for the long, short, and net position in the issue.2 Dealers
2
The dealer on the execution method would regard none of these transactions as completed or executed, hence they would be carried as notes on the trader's blotter. Even
if the gross long exceeded the gross short (or vice versa) he would not record a position.




32

STUDY OF FEDERAL GOVERNMENT SECURITIES DEALERS

recording on a commitment basis would record the gross long and
gross short position separately. Similarly, if the amount of the issue
out on the "buy back" exceeds the amount of the sale, the difference
would be carried as a part of the long position and conversely.
Thus short and long positions which are recorded on the reports
submitted and on other dealer records are summaries of the total net
long position in particular issues of the maturity class or the total
net short position in other issues of that maturity class. The net
position in the maturity class is of course the difference between the
two and is unaffected by the dealer's failure to record gross positions.
F o r this reason, only net position figures are presented.
Gall versus maturity date
Ideally we would like to have all issues classified according to their
"true" maturity. This, of course, cannot be done in advance of the
actual call. Failing this it would be desirable to have all dealer
records on a consistent basis for aggregation. But again dealer accounting practices differ and render the task impossible on the basis
of available records.
Several dealers record maturity at nearest call date when the securities are at a premium and at final maturity date when they are at a
discount in the market. This practice is based on the provisions of the
Internal Kevenue Code relating to the definition of taxable net income on securities. 3 But, since these provisions do not apply to dealers'
incomes, some dealers consistently record all of their securities according to maturity date (with the exception of the partially tax-exempt
2%'s of 1960-65) ; others record all issues to call date.
I n the early years of the period covered by this study, bonds rarely
sold at a discount. But, more recently, the difference in reporting
practice makes for some inconsistencies in the record of maturity
classifications both in comparisons over time and in the aggregation of
individual dealer reports. The substitution of call dates for final maturity results in some slight downward bias in the average maturity of
dealer portfolios in recent years.
Par versus market versus cost price
An additional source of difference in dealer reporting for recent
years results from differences in valuation procedures. I n the reports
for the earlier years of this study, bond prices did not deviate from
par values to the extent that they have in the recent "tight money"
periods. Moreover, given the rapid turnover of most dealer holdings,
and the heavy concentration in bills and other short-term maturities,
differences between cost and market are probably not very large in the
aggregate.
Differences between par and market or cost values are likely to
be more important particularly at the long end of the maturity range.
I n the data presented below there are two sources of difficulty. First,
most dealers recorded their position at par value, but a few used cost or
market. Second, some dealers who reported positions at par furnished
incomplete records for certain years. In some cases, data was supplied on a cost or market basis from balance sheets. These dis8
A taxpayer may elect to amortize all bonds of a particular class which have been
bought at a premium. If the taxpayer elects to amortize, taxable income is the coupon
income earned on securities bought at a discount but the amortized income earned on
securities bought at a premium.




STUDY OF FEDERAL GOVERNMENT SECURITIES DEALERS

33

crepaneies in the method of reporting valuation result in an understatement of dealer positions particularly in the 5-year-and-over maturity class. However, it is unlikely that the totals reported are in
aggregate biased downward by more than 1 or 2 percent from the
par valuation since most of the large dealers value their positions
at par.
TIMING OF THE REPORTS

The greatest single source of difficulty in arriving at an accurate
estimate of dealers' positions occurs as a result of the reporting date.
Dealers were asked to report the detailed statements of their position
for December 31. Two principal problems resulted. First, not all
dealers maintain their records on a calendar year base. Second, there
were some particular problems associated with end of calendar year
statements. These factors must be considered before we can turn to
an analysis of the data.
I t should be recalled that even if economic conditions remain the
same there would be year-to-year differences in position reports as a
result of the weekly Treasury bill auctions. Bills represent an important part of total dealer positions and the day of the week on which
bill awards are made is likely to show a larger position than any other
day of the week, other things equal.
Dealers on a commitment basis record their receipt of bills on the
day the award is made. Both 91-day bills and 182-day bills are
awarded on Tuesday. This imparts an upward bias to the bill position for years which end on Tuesday. On the other hand, as noted
in chapter I I , there may be some tendency for dealers to sell off their
bill positions in anticipation of the weekly auction. This would
impart a downward bias to those years which ended on Monday.
The auction of 182-day bills was not put on a regular weekly basis
until 1958 so that only data for the last year covered by the study is
likely to show a strong upward bias.
TABLE I I I - l . -—Distribution

of Dec. 31 by days of the week,

Number of
occurrences

Day

Monday
Tuesday
Wednesday.
Thursday
Friday i
J

_
_

_

19^7-58

_

.

2
1
3
1
5

Years
occurring
1951, 1956.
1957.
1947,1952,1958.
1953.
1948,1949,1950,
1954, 1955.

Years ending on Saturday or Sunday are included here.

The table suggests that as a result of the weekly auctions there will
be some understatement of the bill position in 1951 and 1956 and probably some overstatement in 1957. The extent to which dealers are successful bidders at bill auctions varies as does the size of the auction,
but on the average, dealer awards amount to about 20 percent of the
weekly total. This would indicate probable limits to the overstatement
of the bill position of from $200 to $400 million in 1957 and less in
1958.
Most dealers reported their positions on a calendar year base. However, several dealers were able to supply fiscal year information only
and in some cases the fiscal year changed during the period. F o r all



34

STUDY OF FEDERAL GOVERNMENT SECURITIES DEALERS

but two of the dealers, we were able to correct this source of error by
supplementary information obtained from records which most of the
dealers had voluntarily submitted to the Federal Reserve Bank of New
York. 4
Finally, several dealers indicated that end of the year reports of
position overstate net position. This results from several factors: the
taking of capital gains and losses, balance sheet "window dressing,"
dividend payments by corporations, and other seasonal demands for
cash by banks and corporations. Doubtless there is some tendency toward a seasonal pattern in this market particularly at the short end.
Moreover, it is likely that the pattern is similar to the well-known seasonal change in the demand for money since short-term instruments
are a close substitute for cash. Banks undoubtedly sell some of their
bill portfolio to dealers to meet seasonal needs. However, some of the
adjustment in the banking system is made through "runoff," i.e., by
exchanging maturing short-term instruments for cash and hence is not
wholly reflected in dealer positions.
Unfortunately, the only available data come from the Treasury-Federal Eeserve study. These show weekly dealer positions for 62 Wednesdays from October, 1957 through December 1958.5 However, the
weekly changes derived from their table are probably larger than
average changes in other years for two reasons. Positions rose very
rapidly after the reduction in the discount rate in November 1957, and
reported positions fell $2.5 billion in the 17 weeks following the drop
in prices in June 1958. These fluctuations are probably much greater
than average and thus prevent us from estimating the extent to which
December positions are biased upward. Nevertheless, the data presented there show that the yearend total position for 1958 exceeds the
annual average by less than $300 million. Moreover, in this period
with rapid changes in economic conditions, more than two-thirds of
all week to week changes are less than ±$300 million, and one-half the
changes are within 10 percent of the annual average total dealer position given there.
A similar analysis of changes in weekly bill positions shows that
about half of the deviations are within ±$150 million and that twothirds of the changes are within 20 percent of the annual average.
However, the yearend bill positions exceed the annual average by
more than 50 percent in 1958 and exceed the 10-week average by approximately 30 percent in 1957. Dealer interviews suggest that such
yearend increases in the bill portfolio are not unique but are probably
larger than usual.
Summary
The weight of the conflicting tendencies described in this section
seems on balance to overestimate net dealer positions. This is probably more true of bills than of other maturities owing to the substantial increase in bill positions at calendar-year end. However, it is unlikely that the annual estimates given here exceed the annual net positions of dealers by more than 10 percent.
4
However, the supplements were not always prepared in a manner consistent with the
method of reporting to the committee, e.g., maturity may have been by call date on one
report
and maturity date on the other.
8
Cf. Treasury-Federal Reserve Study * * *, op. cit, II, 138-139.




STUDY OF FEDERAL GOVERNMENT SECURITIES DEALERS

35

E N D OF YEAR INVENTORIES

Aggregate dealer inventories by maturity classes are shown in
table I I I - 2 based principally on year end data. Table I I I - 3 , computed from these aggregates, presents portfolio composition by maturity classes. Together these data suggest some of the ways in which
monetary policies and changing economic conditions affect dealer positions. Several tendencies shown by the data should be noted:
1. Dealer positions have fluctuated greatly over the years since December 1948. They have ranged from approximately $750 million at
the end of 1950 to $1.7 billion in December 1953 and 1957. Economic
conditions seem to account for the general direction of the movements
displayed. Positions are highest (in the years following the accord
of March 1951) when interest rates are relatively high and are expected to fall; conversely, when interest rates are relatively low and are
expected to rise, dealer positions are smaller. For example, in 1957,
we observe the dealer position less than 2 months after the reduction
of the discount rate. Dealers appear to have increased their portfolios
in anticipation of a rise in security prices. On the other hand, 1955
and 1956 are years in which interest rates were expected to rise. F o r
those years, dealer positions are lower. Aggregate positions for 1947,
and to a lesser extent 1950, are informative in this regard. The 1947
position is extremely small even when we consider that only 13 dealers
reported in that year. Dealer positions for that year end probably
reflect the unpegging of the short-term rate which had been announced
earlier. The observation for 1950 reflects the uncertainly engendered
by the increase in the discount rate in August and the many rumors
that the Federal Eeserve and Treasury disagreed about the extent to
which Federal Eeserve operations should be directed at supporting
Government bond prices at par. I n both cases, increasing interest
rates were anticipated and dealer positions were reduced.
2. The accord of 1951 seems to have led to an increase in aggregate
dealer positions. This is suggested by a comparison of 1949, 1953,
and 1957—years in which recessions started. And this is consistent
with the data in chapter I V which shows that the volume of transactions increased in the years following the accord.
3. There is some slight evidence that the application of the "bills
only" policy increased dealer positions in bills. The fact that bill
positions are larger in 1957 and 1958 than in any of the years since
1952 in part reflects the fact that 1952, 1957, and 1958 are years which
end on Tuesdays or Wednesdays. As noted above, the bill position on these dates should be larger by the unsold part of the weekly
award to dealers at the auction. Similarly, if dealers generally sell
out a large part of their bill position in anticipation of the weekly
auction, 1951 and 1956, which end on Monday, or 1954 and 1955,
which end on Friday, should present lower bill positions than might
be observed during the last Tuesdays or Wednesdays of the same
years. Nevertheless, it is unlikely that dealer awards at auction alone
can account for all of the steady increase in bill positions since 1954.
As we shall argue below, the effect of the "bills only" policy has probably been indirect, but nonetheless effective in altering the dealers'
portfolio composition.




36

STUDY OF FEDERAL GOVERNMENT SECURITIES DEALERS

4. There does not appear to be any observable trend in long-term
(over 5 years) dealer positions. This is not surprising since only one
observation per year during a period of relatively rapid changes in
economic conditions is given here. And the transactions data below
seem to indicate that there has been a declining trend in activity in
the long-term market. These declines have been interrupted from
time to time by an increase in transactions in years with sharp swings
in the price of long-term bonds. But, the absence of any upward
trend in dealer bond positions and the decline in transactions in the
over 5-year maturity class seem to imply that the "bills only" policy
has failed to provide the much advertised "depth, breadth, and resiliency" in the long-term market for Government securities. 6
5. Dealer positions in the municipal bonds and "other" issues (principally Government obligations other than direct Treasury obligations) have shown a tendency to increase during the period of observation. Like holdings of Government securities, they tend to fluctuate
with changes in interest rates and economic conditions. As noted
above, the stock of outstanding issues increased markedly during the
12 years covered by this study, and this must be recalled when the
dealer positions are assessed. Nevertheless, it would appear likely
that the failure of the long-term Government securities market to
develop into a more active trading market in part accounts for the
increased interest in other long-term issues by Government securities
dealers. This point is reinforced somewhat by the increased number of dealers who now take positions and actively trade these other
obligations.
0
The reader may wish to compare the annual data presented here with the weekly data
on dealer positions available from the Treasury-Federal Reserve study. Unfortunately
the two are not strictly comparable. As noted above, we have attempted to report
positions on a commitment basis. Our data, therefore, differ from the Treasury-Federal
Reserve data principally by the amount of "securities sold but not yet delivered" which is
included in their totals but not in ours. Their figures overstate dealer inventories whether
computed on the commitment or execution method of reporting since they include both the
items which dealers have agreed to purchase and those which they have agreed to sell.
The amount of their overestimate is in the neighborhood of $1 billion at the end of
1957 and 1958 and is concentrated principally in the bill account.




H
d
T A B L E III-2.—Aggregate dealer positions at end of years 1947 to 1958 classified by maturity

l

o

[Millions of dollars]

b
Years

Item
1948

1947

Total bills
Certificates of indebtedness
Notes and bonds under 1 year

_

_
_

Total under 1 year
Notes and bonds 1 to 5 years
Bonds over 5 years
Total position in Government securities.
Municipals
Others l
_
Number 3 of dealers reporting position in Governments
1

1954

1953

1952

1956

1955

1957

1958

81.5
102.6
19.4

49.7
359.3
57.8

255.3
311.7
42.7

236.8
25.0
146.1

466.8
177.0
141.1

767.2
125.4
101.2

627.0
333.7
356.7

226.9
190.2
285.5

268.0
79.9
121.2

458.0
67.4
151.4

762.1
283.7
251.8

917.8
154.0
122.6

203.5
26.2
62.3

466.8
146.0
188.9

609.7
283.5
50.1

407.9
251.1
93.9

784.9
111.3
23.4

993.8
159.6
120.7

1, 317. 4
264.5
118.6

702.6
233.0
-12.5

469.1
364.5
95.9

676.8
95.8
55.6

1, 297. 6
304.6
70.0

1,194. 4
125.3
51.6

292.0

801.7

943.3

752.9

919.6

1,274.1

1,700. 5

923.1

929.5

828.2

1,672. 2

1,371. 3

8.2
36.8

16.8
34.2

34.7
50.9

29.9
64.6

65.5
50.9

70.2
111.1

51.3
94.1

100. 0
112.7

62.2
115.9

41.9
93.0

74.4
122.5

110.7
54.7

13

15

16

17

17

17

17

17

17

17

17

17

For discussion of use of call dates versus maturity dates, description of "others" and
reporting on calendar versus fiscal year, see text.




1951

1950

1949

a
o
<

i
%
ui
tel

o
d

2

The number of firms reporting municipals or "others" ranges from 5 (1947) to 1
(1953-55) for the former and 8 (1947) to 13 (1957 and 1958) for the latter.

02

00

o
T A B L E III-3.—Percentage distribution

of dealers' position

by maturity

classes at Dec. 31, 1947, to 1958

l

Years

W

Item
1947
Total bills
Certificates of indebtedness
Notes and bonds under 1 year
Total under 1 year
Notes and bonds 1 to 5 years
Bonds over 5 years
Municipals, total GovernmentsOthers, total Governments
Total municipals plus others
1

See notes to table I I I - l .




- --

1948

__-

27.9
35.1
6.7

6.2
44.7
7.2

--

69.7

58.1

8.9
21.4

18.2
23.5

2.8
12.5

2.1
4.2
6.3

1949

1950

1951

1952

1953

31.5
3.3
19.5

50.8
19.3
15.3

60.2
9.8
7.9

36.9
19.6
21.0

64.5

54.3

85.4

77.9

77.5

30.1
5.3

33.5
12.4

12.0
2.5

12.5
9.5

15.6
6.9

3.6
5.4

4.0
8.5

7.1
5.5

5.5
8.7

3.0
5.5

9.0

12.5

12.6

14.2

8.5

27.0
33.0
4.5

1954

1955

1956

1957

1958

a
o
<

28.8
8.6
13.1

55.4
8.2
18.3

45.6
17.0
15.1

67.0
11.2
9.0

76.1

50.5

81.9

77.7

87.2

i

25.2
-1.3

39.2
10.4

11.6
6.7

18.2
4.2

9.2
3.8

%

10.8
12.2

6.7
12.5

5.1
11.4

4.4
7.3

8.1
4.0

23.0

19.2

16.5

11.7

12.1

24.6
20.6
30.9

Ui

O

STUDY OF FEDERAL GOVERNMENT SECURITIES DEALERS

39

DEALER HOLDINGS OF THE PUBLIC DEBT

One source of fluctuations in dealer holdings of various maturities
which has not been considered is the change in the maturity composition of the debt. The average of outstanding maturities may decline
with the passage of time or may reflect a Treasury policy of increasing,
maintaining, or reducing average maturity. Changes in the composition of outstanding maturities must be considered in appraising the
changes in the maturity structure of dealer positions. I n table I I I - 4
the statistics on the maturity structure of the outstanding marketable
debt held by the public and the Federal Eeserve banks are presented.
As is well known, the average maturity of the debt has shortened in
the postwar years. More than 45 percent was due or callable within
1 year at the end of 1958.7
Dealer positions are more heavily concentrated in the short end
of the maturity structure than the total debt. I n particular, their
bill inventories and operations far exceed the relative importance of
these items in the marketable public debt. Notes and bonds with
less than 1 year to maturity occupy approximately the same relative
position in dealer inventories as they do in the outstanding marketable
debt. As the years to maturity or first call lengthen beyond 1 year
the proportion of dealer holdings in these maturity classes declines
below the relative position of such issues in the composition of the debt.
These results are not surprising in the light of the transactions data
below and the differences in function which buyers seek when they
purchase differing maturities. Short-term debt is principally a means
of holding liquid balances. As such, it is expected to turn over
rapidly. Longer term issues are often purchased by pension funds,
state retirement plans, and insurance companies among others. These
groups are said to be "investors" or "holders" and it is claimed that
they are often willing to sacrifice yield to some extent in order to
obtain a desired time distribution of cash flow. Hence much of the
longer term debt does not become available to the dealer market and
there is less reason for dealers holding as large a share of the outstanding issues.
These observations are also reflected in the data of table I I I - 5
where the percentages of particular maturity classes and of the total
marketable debt which dealers hold are presented. I n the period
during which interest rates were pegged, much of the short-term debt
and particularly low-yielding bills moved into the Federal Eeserve
banks. Debt holders chose the secure and liquid longer term issues for
their portfolios. By 1949 dealers had begun to hold a larger proportion of outstanding bill issues than of any other maturity. And, in
general, their share of short-term issues exceeds their share of longer
term issues since that time. This becomes more evident when it is
recalled that approximately 15 percent of the marketable public debt
used to compute these percentages is held by the Federal Eeserve.
Eoughly 80 percent of the Fed's holdings in recent years has been in
the under-1-year maturity class. Therefore, exclusion of the Federal
7
As noted earlier, dealers as a group do not consistently use either call date or final
maturity date in classifying their own portfolio. We have used call dates as the method
of classifying the outstanding public debt since that is the more common dealer practice
during the years considered.
62471—60
4




40

STUDY OF FEDERAL GOVERNMENT SECURITIES DEALERS

Reserve's inventory from the denominator would increase the dealer's
share of marketable short-term debt far more than it would increase
their share of the last two maturity classes.
Dealers' share of total issues in general seems to be largest early in
recession after interest rates have risen and are expected to fall. Conversely, when rates are low and expected to rise, dealers hold much
smaller proportions of the outstanding debt. This observation is similar to our findings about total dealer position and is subject to at
least two interpretations. One view is that the dealers use their positions to speculate on changes in interest rates. Another is that dealers
prepare for the changes that are about to take place in the market by
adjusting their portfolios to the expected demand and supply.
These two views need not be, and in fact are not, mutually exclusive.
As our description of the dealer function implies, dealers make markets to their customers by purchasing or supplying securities which
the customer wishes to buy or sell. I n the process, they may find that
they are purchasing against a falling price or selling against a rising
price. Such transactions may have a transitory effect on dealer position, i.e., may be quickly reversed. The more fundamental question
raised by critics of the dealer market is that dealers hold inventories
off the market during periods of rising Government securities prices
and tend to be a destabilizing influence by forcing wider swings in
interest rates than would otherwise occur.
The evidence from the annual data and from the share of outstanding issues is insufficient to answer the question raised. However
additional evidence in the form of intrayear changes in position sheds
some light on the dealer methods of operation.
TABLE III-4.—Percentage composition
by maturity classes1 of
marketable debt2 at 194^-58 year end

Bills
Certificates
Notes and bonds under 1
year
Total under 1 year..
Notes and bonds 1 to 5
years. _
B o n d s 5 years a n d o v e r . —

outstanding

1947

1948

1949

1950

1951

1952

1953

1954

1955

1956

1957

9.1
12.8

7.8
16.8

7.9
19.1

8.9
3.5

13.7
20.4

14.6
11.2

12.6
17.1

12.4
18.0

13.7
9.6

15.7
11.9

16.4
21.0

16.9
20.7

8.7
30.6

6.6
31.2

9.3
36.3

25.6
38.0

18.0
52.1

24.1
49.9

19.4
49.1

9.4
39.8

17.7
41.0

19.0
46.6

11.3
48.7

8.7
46.3

30.1
39.3

28.0
40.9

22.6
41.0

21.9
40.0

20.1
28.9

20.3
29.7

17.2
33.6

24.9
35.3

26.5
32.6

28.7
24.7

27.6
23.6

28.5
25.2

1958

1

Held publicly and at Federal Reserve banks.
Classified to maturity or first call, whichever is earlier.
Source: Treasury Bulletins.

2

TABLE III-5.—Dealers'" position as a percentage of outstanding
marketable
by maturity classes at 1948-58 year end2

Total bills
Certificates
Notes and bonds under 1 y e a r . . .
N o t e s a n d b o n d s 1 to 5 years
B o n d s 5 y e a r s a n d over
T o t a l position
.
1
2

1948

1949

1950

1951

1952

1953

1954

1955

1956

1957

0.04
.14
.06
.03
.03
.05

0.21
.10
.03
.08
.01
.06

0.17
.05
.04
.08
.02
.05

0.26
.06
.06
.04
.01
.06

0.35
.08
.03
.05
.03
.08

0.32
.13
.12
.10
.02
.11

0.12
.07
.19
.06
0
.06

0.12
.05
.04
.08
.02
.06

0.18
.04
.05
.02
.01
.05

0.28
.08
.14
.07
02
.10

Defined as publicly held marketable debt plus Federal Reserve held marketable debt.
See notes to table III-li




debtx

1958
0.31
.04
.08
.02
.01
.08

STUDY OF FEDERAL GOVERNMENT SECURITIES DEALERS

41

MIDYEAR POSITIONS

I n table I I I - 6 dealer net positions on December and June dates are
compared. These data are based on the reports for only those firms
which were able to submit records on a consistent basis for both sets of
dates. They indicate that over the 11-year period June positions have
averaged less than December positions by approximately $275 million
and that in most years dealer positions fell from December to the
following June.
When some allowance is made for the absence of three firms from
the data of table I I I - 6 , the range of dealer positions appears to be between 450 million and $1.8 billion with a mean position in the neighborhood of $950 to $1,000 million.
Large changes in position occur within several years, most notably
from June 1949 to June 1950, December 1952 to December 1953, June
1954 to June 1955, and from June 1957 to June or December 1958.
These swings in dealer holdings roughly coincide with changes in the
level of economic activity. Their relationship to changes in interest
rates is shown in chart I I I - l .
TABLE III-6.—Comparison of net dealer positions, June and December 1948-58
1948

Number of firms
June
December
_.

1949

1950

14
14
13
877.1 1, 233.9 458.2
790.2 891.4 693.0




1951

1952

1953

1954

14
14
14
14
624.7 1,261.6 404.7 1,377.8
724.7 1,166.1 1,505.7 748.8

1955

1956

14
393.9
764.1

14
14
14
532.8 541.4 1,415.4
660.3 1,511.2 1,048.4

1957

1958

CHART I I I - l
M I L L I O N S O F DOLLARS

to
32

c!
o
feJ

>
| ( O00

O
O
<1

INTEREST
RATES

w

4.0

3-0

a
tr1

w
Ww'Vn




&

A9 %>

<v^ 'Vso m.

'Vsi

6

/*z '*M Vst '¥53 fsv

n/s* &/f$ 'Ysr 6/st <Yr*°/si °fsi */s? «/*r

DATES

STUDY OF FEDERAL GOVERNMENT SECURITIES DEALERS

43

I t is clear from the chart that dealer positions frequently move in
the opposite direction to interest rate changes. From this some might
conclude that dealers speculate on changes in the rate and adjust their
positions accordingly. Such a conclusion might imply that dealers are
able to anticipate Federal Eeserve operations and can thereby profit
from them. I t is undoubtedly true that dealers attempt to do this,
but as we shall see in chapter V I I , their profit records do not indicate
that they are particularly successful except when wide swings in the
interest rate take place at the start of recessions. Moreover, the available evidence suggests that as a group they are no more successful in
anticipating changes in interest rates than other professional traders—
most notably commercial bankers. F o r if we consider the 17 semiannual changes in interest rates and dealer positions shown in the
chart, we find that 11 of the 17 changes in the medium-term rate are
opposite to the change in dealer position but 6 are in the same direction. And 7 of the 17 semiannual changes in the bill rate are in the
same direction as the change in position. We may assume that differences in seasonal pattern or fluctuations in position within the
semiannual periods account for some of the observations. Nevertheless, the relatively large reductions in interest rates from December
1953 to June 1954 and from December 1957 to June 1958 were accompanied by reductions of approximately 20 percent in net position.
Other evidence (1) that there is disagreement among individual
dealers about the direction of rate changes; or (2) that dealers are not
in general successful forecasters of changes in Government securities
prices is shown by the data of table I I I - 7 . The ratios of December
to June positions indicate that in only 2 years—1953 and 1957—all
dealers changed their positions in the same direction over 6-month
periods. I n both of the years relatively large decreases in interest
rates accompanied the start of recession. Moreover, the weekly data
in the Treasury-Federal Eeserve study seems to suggest that the bulk
of the increase in dealer positions which we observe in the latter part
of 1957 did not occur until after the announcement of a reduction of
the rediscount rate by the Federal Eeserve Board.
From June to December 1954 the aggregate position of dealers fell
more than $600 million. Again in 1958, total net position was reduced
by nearly $400 million. During both of these periods, interest rates
rose and bond prices fell. Nevertheless, 5 of the 14 dealers increased
their holdings during the latter half of 1954 and 6 of the 14 added to
their positions during the last 6 months of 1958. Similar results hold
for other periods. These considerations seem to suggest the conclusion discussed earlier—that dealer operations are principally conditioned by the relatively short-term view which the individual dealer
takes and that with the exception of periods like the fall of 1953 or
1957 and the spring of 1958, most dealers do not attempt to influence
or follow the movement of the yield curve. Most often they are more
concerned with day-to-day trading activities and arbitrage operations.
Furthermore, the data presented in table I I I - 7 reveal that there are
marked differences between individual dealers in the amplitude of the
swings in net position. F o r example, dealers Nos. 4 and 5 have approximately the same average net position. And they tend to agree
on the direction of change in position for most of the years examined
here. But the average size of the changes for dealer No. 4 is much
smaller than the average change for dealer No. 5. Similarly, dealers



44

STUDY OF FEDERAL GOVERNMENT SECURITIES DEALERS

Nos. 7 and 9 have approximately the same average position for the 11year period. But while the average December position for all of the
dealers included in this table is 33 percent greater than the corresponding J u n e position, dealer No. 7 has a smaller than average
change from J u n e to December and dealer No. 9 has an average change
which is substantially larger than the change for the aggregate of all
dealers.
I t should be recalled that these figures are aggregates and that the
risk which a dealer undertakes tends to increase directly with the
average maturity of the dealer's position as well as with its size.
Thus, while we have attempted to compare dealers whose operations
are reasonably similar, even for them day-to-day or week-to-week
fluctuations in position are accompanied by opposing directions of
changes in portfolio composition which go in opposite directions.
Such changes alter the average risk assumed by the individual dealers.
As a result, a reduction in position from a particular June to December
may be accompanied by an increase in longer term maturities. A
particular dealer may therefore be in a better position to profit from
changes in the prices of Government securities even though he has
reduced his total commitment. While there is little doubt that this
happens, there is also very little evidence that it is a general
phenomenon.
TABLE

I I I -7.— Total

net position last Wednesday of December/last
of June for 11^ individual dealersx

Wednesday

Firm number
Year
1
1948
1949
1950
1951
1952
1953
1954
1955
1956
1957
1958
Average _

2

3

4

5

6

0.82
.88
1.72
2.13
.87
5.45
.83
2.02
.81
3.22
.98

1.14
1.40 0.57 2.22 2.64
.92
.79
.03
3.40 1.18
1.11
.76 - 1 . 2 5 - 6 . 0 4
.57
.96
.23 1.13 2.49
.76
.68 2.18 1.02
.63
.58
1.32 10.95 1.72 20.32 4.36
.28
.15 1.21
1.37
1.59
6.42 6.82 13.26
.20 1.02
.92 1.16 2.36
1.48
1.79
1.00 3.69 1.31 45.53 5.32
.19 1.18
.63
.82 1.27

1.80

1.17

2.21

1.33

6.91

3.13

7

8

9

10

11

12

1.08 0.59 ( 2 )
1.42
0.42 1.84
.41
.83 1.84
.08
.99
.33
.31
1.06
1.31
.19 1.50
.96
8.00
.35 1.04
.74
0
.93
.31
.68
.48
.72
.28 2.25
1.24 4.50 40.00 - 4 . 2 0 11.35
1.25
.37
.12 4.08
.49
.48 1.26
.76
.75 1.63 2.52 0
.80
.98
1.49 - . 4 1
1.11 8.50 - 2 . 5 0
1.39
1.74 - 1 . 4 7 13.34 9.32 1.37
.69
.32
1.62
1.50
1.86
1.30
1.10

.54

2.64

6.10

.64

2.01

13

14

0.82
1.00
1.07
1.82
.74
9.27
.30
3.10
.64
1.28
.42

0.30
.98
.50
1.02
.87
1.63
.47
3.30
5.49
2.20
.52

1.86

1.57

t The average for aggregate dealer positions is 1.33 for the period covered.
Not available.

Finally, there is some additional evidence that dealer operations are
more likely to be concerned with short-term changes in the money and
securities markets. Some of this evidence will be discussed in chapter
V when the sources of dealer financing and changes in the cost of
carrying securities are examined. Other evidence is considered below
in the section on dealer investments. F o r the present, it would appear
that there is little evidence in the data which have been examined that
indicates that dealers in the aggregate exercise a destabilizing influence in the market or that they tend to accentuate the amplitude of
interest rate fluctuations. And there is some evidence that at the start
of recession periods, they substantially increase the demand for Government securities and profit from assisting in the process by which
interest rates are adjusted downward.



STUDY OF FEDERAL GOVERNMENT SECURITIES DEALERS

45

INDIVIDUAL DEALER PORTFOLIOS

Examination of the portfolio composition of individual dealers
reveals a great deal of heterogeneity. To give some indication of the
differences between individual dealers, the percentage composition by
maturity classes for each dealer has been computed by dividing the
sum of a dealer's 12-year inventory of a particular maturity by the
sum of his total position for the 12 years. The resulting averages are
presented in table I I I - 8 . While these averages do not indicate the
relative size of the individual portfolios, they help to suggest the differences in approach taken by individual dealers. 8
I n chapter I I , we stated that some dealers specialize in particular
maturity classes while others make a market throughout the entire
range. This is again suggested by the data of table I I I - 8 . Bill positions represent as little as 21 percent and as much as 61 percent of an
individual dealer's inventory over the period. Likewise, several dealers on the average hold more than 80 percent of their portfolio in
very short-term maturities; for others, only three-fifths of their inventory was concentrated in under-1-year maturities on the average.
Similar differences occur in the long- and medium-term maturity
classes.
These data are subject to many of the qualifications discussed earlier.
Moreover, as noted in the table, complete records were not obtained
from all firms for the entire period. Nevertheless, it seems reasonable
to group some of these firms for purposes of comparison. Dealers C
and O, for example, hold about the same proportion of both bills and
long-term bonds; dealers K and M have very similar distributions by
maturity class on the average and hold similar views of the dealer
function. The distribution of their portfolio suggests that these
firms took positions over the entire maturity range during 1947-58.
But, when we consider a more recent period, 1953-58, we find that
dealers K and M have substantially reduced the proportion of their
position in the over-5-year-maturity class while increasing the proportion of bills which they hold. Since the adoption of the "bills only"
policy, the management of these firms apparently has decided to concentrate in shorter maturities either as a means of reducing the risk
inherent in their position or in recognition of the failure of the longterm market to expand. Similar results hold for dealers F , G, and
O ; all of whom formerly held a relatively large proportion of their
inventory in bonds with 5 or more years to first call. As may be seen
from the table, only three dealers (A, C, and D) increased the proportion of long bonds which they held.
Conversely, 14 of the 17 dealers (or 12 of 15 dealers whose positions
are shown) increased the proportion of bills in their total position
during recent years. For firms G, L, and N bills represented more
than three-fifths of their total inventory during the 6-year period
1953-58.
8
The letters assigned in place of the dealer names are the result of a haphazard selection process. As indicated earlier, dealers released their records under an agreement
which barred publication of records which might reveal the position of their firm. Two
small dealers whose operations are substantially different from any of the others shown
here have been omitted from table III-8.




46

STUDY OF FEDERAL GOVERNMENT SECURITIES DEALERS
TABLE III-8.—Percentage distribution of individual dealer positions
AVERAGES FOR DEC. 31, 1947-58 i BY MATURITY CLASS
Dealer firms

Items
A

B

C

D

E

F

G

H

I

J

K

L

M

N

O

27.6 41.0 26.3 38.0 36.5 30.8 56.6 53.1 51.8 34.8 37.4 61.4 37.8 56.9 21.1
20.6 10.8 13.4 23.8 21.5 18.5 14.9 20.4 15.1 25.4 20.6 8.6 23.6 19.0 23.8

Bills
Certificates
Notes and bonds
under 1 year

12.0 15.4 17.8 10.7 22.0 19.7

2.8

7.7 19.8 13.5 15.2

8.0 14.5

7.6 23.3

Total under 1 year_ 60.2 67.2 57.5 72.5 80.0 69.0 74.3 81.2 86.7 73.7 73.2 78.0 75.9 83.5 68.2
Notes and bonds 1 to
5 years
31.5 27.6 32.2 20.8 17.0 21.3 14.3 12.7
Bonds 5 years and
8.3 5.3 10.2 6.7 2.9 9.5 11.4 6.1
over

6.8 21.9 15.1 21.1 14.2 11.2 20.6
6.4

4.1 11.5

.9 10.0

5.2 11.2

D E C . 31, 1953-58
Bills
Over 5 years

_

._

37.0 48.7 17.6 51.9 41.1 34.2 66.1 41.0 54.8 36.1 43.0 63.9 41.6 63.3 19.1
9.0 3.6 13.0 8.5 2.8 6.9 6.0 2.8 3.0 3.2 6.6 - . 8 5.5 2.9 4.5

1
Based on differing number of obsei vations: 11 firms in 1947; 13 in 1948; 14 in 1949; 15 thereafter. 2 small
dealers have been excluded to prevent possible recognition.

The data on individual dealer positions and the changes over time
just discussed seem to suggest that most dealers have reduced their
relative holdings in the long-term market in recent years. 9 As chapter I V shows, this has been accompanied by a relative decline in
transactions in this sector of the market except in years of sharp decreases in interest rates. As a result, transactions in the long-term
market have become more concentrated among the few remaining
dealers who are willing to undertake the risk of maintaining sizable
positions in that area.
INVESTMENT ACCOUNTS

W e have indicated several times that most dealers do not take large
speculative positions in Government securities for capital gain. I n
the data which we examined, bank dealer investment accounts were
excluded since these are carried as part of the bank's portfolio. Of
the remaining 12 nonbank dealers, only 6 report investment accounts
in 1 or more years and there is no year for which more than 4 dealers
reported holding Government securities for capital gain. I n aggregate, the investment accounts for all reporting nonbank dealers exceeded $100 million in only 1 year—1957.
I n part, the absence of dealer investment accounts reflects the fact
that for some of the dealers, there is little tax advantage in attempting
to hold securities for capital gain rather than for corporate or partnership income. Several dealers had sustained losses from their operations which could be carried forward; others are in relatively low tax
brackets. More important are two characteristics of the dealer market
which were suggested earlier: first, most of the dealers take a relatively short-term view when they consider the changes taking place
in the market; second, there is lack of available sources willing to
finance dealer positions for long periods at rates which are below the
9
T h e testimony of several dealers tends to confirm this conclusion,
1959, pp. 1845-1848.




Cf. H e a r i n g s 6 C

STUDY OF FEDERAL GOVERNMENT SECURITIES DEALERS

47

yield of the securities. Further, as noted above, the investment accounts of some corporate dealers probably are carried in the name of
the individual principals rather than as a part of the corporation's
assets.
TABLE III—9.—Dealer investment

Year

1947
1948
1949
1950
1951
1952
1

__ .

Number of
dealers
reporting
investments

accounts

Amount
reported
(millions)

1
0
3
3
0
1

(0

0
73.5
14.2
0
0)

Year

1953
1954
1955
1956
1957
1958

Number of
dealers

4
3
4
1
4
3

Amount

43.7
28.4
68.5
0)
236.0
96.8

Not available under the rules covering the submission of the data.

More than half of the total investment shown in table I I I - 9 was
held by one dealer; approximately four-fifths was held by but two
dealers. Both of these firms are large and active traders of long-term
securities and have active research staffs. Most of their investment
was confined to years such as 1949, 1953, and 1957 and coincides
reasonably well with peaks in total dealer trading accounts.
Finally, it should be noted that none of the totals show negative
net positions for the dealers reporting. This is equally true of the
individual dealer records and reflects both the cost and difficulty of
borrowing securities and the accrual of interest charges during the
period in which the securities are sold short. If a difference of onequarter percent between yield and carry is required to induce dealers
to buy securities for their investment account, a difference of threequarters percent would be required to induce short sales given the onehalf percent cost of borrowing securities.
FINDINGS

Most dealers now recognize the weakness or "thinness" of the longterm market for Government bonds. This has not always been true.
Until recently spokesmen for investor groups privately accused the
dealers of being overoptimistic in their expectation that the long-term
market would revive. Failure of the market to become more active and
the closely related failure of dealers to take larger positions in bonds
result from two principal weaknesses: (1) most important has been
the failure of the Treasury to market long-term issues of sufficient
size to permit arbitrage transactions to take place and (2) the failure
of the "bills only" policy to restore the now famous trinity—depth,
breadth, and resiliency—to the long-term market. These are not
unrelated problems.
Two shortcomings of the dealer market are reflections of these
weaknesses in monetary and debt-management policy. First, arbitrage
operations between issues at the long end become difficult or in some
cases impossible. I n part this occurs because of the small size of many
of the outstanding issues; in part because of the statutory or policy
requirements of many institutional investors which prevent lending of
securities. This results in a series of price quotations for many of the




48

STUDY OF FEDERAL GOVERNMENT SECURITIES DEALERS

small issues which are not directly based on market prices. Instead,
dealers quote prices to the financial press and others by inferring what
the price of the issue "should be." Attempts to execute transactions
at the quoted price often cause the price to change substantially.
Second, arbitrage operations over time have been important only when
the Federal Reserve has given a clear indication of a policy of ease.
We found little evidence that at other times dealers as a group attempt to use their positions to anticipate changes in the shape of the
yield curve; e.g., by simultaneously selling in the short-term market
and buying long-terms or by taking relatively large positions at the
long end in anticipation of a change in interest rates.
The inability of the debt managers to increase or maintain the average maturity of the debt is not the only source of difficulty. For, whatever its relative merits for other purposes, a "bills only" policy, or any
monetary policy which is to be successful in adjusting the interest rate
on long-term investment, must raise or lower the intermediate or longterm rate either directly or by encouraging arbitrage operations. I t
is clear that recent monetary policies have not operated directly on
long maturities and it seems equally clear that they have not succeeded
in encouraging arbitrage between short-term and long-term maturities by assisting in the establishment of a free, active trading market.
Any assistance which the Federal Reserve might render which would
increase the opportunities for dealers and others to take positions in
the intermediate or long end of the market for purposes of conducting
arbitrage operations would assist the economy to adjust interest rates
more rapidly to meet changing economic conditions.
The foregoing does not imply that there is no arbitrage or that there
is little connection between the short- and long-term markets. I t suggests that the opportunities for these types of transactions have not
increased in recent years, in part because monetary and debt-management policies have not been designed to increase the marketability of
securities in the long-term market. Since buyers or sellers are unable to carry out purchase or sales exceeding one-quarter of a million
dollars (or less) in many issues without substantially changing the
price, one of the primary advantages of a Government bond—its relative liquidity—is destroyed. As a result, the increased relative
liquidity and higher yields available on large corporate issues, or State
and municipal bonds, augment the attractiveness of these issues to
institutional investors. I n this way, the "bills only" policy further
reduces the liquidity and hence the attractiveness of long-term Governments and contributes to the difficulty which the Treasury has in marketing future long-term issues.




CHAPTER IV
VOLUME A N D COMPOSITION OF TRANSACTIONS IN U.S.
SECURITIES
DEVELOPMENT OF TRADING I N THE DEALER MARKET

The present market for U.S. securities is a part of the over-the^
counter securities market. I t s institutions, trading practices, and
transaction volume have developed over the past 180 years in response
mostly to Treasury financing needs. The Treasury's demand for borrowed funds derived mostly from war finance. During the first half
of the 19th century relatively small amounts of Federal Government
bonds were outstanding. Not until the issuance of Federal debt during the Civil W a r and the use of U.S. Government bonds as backing
for the note issue of national banks did the market for U.S. securities
attain national importance.
The organized auction market of the New York Stock Exchange
provided an adequate trading place during the periods of little activity
in U.S. securities. Increases in the Federal debt and in the volume of
trading in Governments focused attention on the weaknesses of an
auction market in handling the large transactions typical of the trading in U.S. securities. A t times of heightened activity in this market
during and immediately after the Civil War, around the turn of the
century, and again after the First World War, trading in Governments
moved from the organized exchange into a negotiated market composed of specialized security dealer firms and a few banks. These
institutions acted like the specialists on the organized exchanges.
In order to provide a stable and continuous market for U.S. securities
they purchased and sold large amounts into and out of their own
inventories. The impact of temporary excess market supply or demand—large orders to buy or sell—was thus smoothed and erratic
price fluctuations in U.S. securities were reduced.
I n an auction market—or organized exchange—temporary discrepancies between orders to buy or sell result in periods of inactivity and
uncertainty or in large price fluctuations. Chiefly for this reason,
sellers and purchasers of U.S. securities prefer a negotiated market.
Volume of trading in Governments on the organized exchange has
been replaced by increased over-the-counter volume.
This development occurred during the early 1920's despite a
Treasury policy of favoring trading of U.S. securities on the New
York Stock Exchange. The Treasury reversed its attitude during the
midtwenties and since then has helped to strengthen the negotiated
over-the-counter market in U.S. securities. A t present, trading volume in Governments on the organized exchange is negligible in relation to the volume of transactions in the dealer market.




49

50

STUDY OF FEDERAL GOVERNMENT SECURITIES DEALERS

The rapid growth of the Federal debt during the thirties and forties,
the increasing use of money market instruments (Treasury bills and
certificates) in Federal debt management, and the rise of open market
policy as the major monetary control instrument in the hand of the
Federal Keserve System again heightened the need for an efficient,
competitive national market for the secondary trading of U.S. securities. Thus the dealer market has developed into its present size and
organization in response to (1) Treasury financing needs, (2) the
economy's demand for highly liquid, easily marketable debt instruments, and (3) the role of open market policy which is conducted exclusively through this market. An analysis of the size of
transactions, their composition, and of the concentration of trading
activity in the dealer market has considerable significance in appraising the market's efficiency and performance as part of the monetary
control system.
This chapter will deal with changes in volume and composition of
transactions in the dealer market during the period from 1948 to 1958.
The concentration of trading activity will be analyzed, and the trading
volume in this market will be related to the stock of marketable securities, to Treasury debt management actions, and to monetary policy.
To make these discussions meaningful, a description of the scope of
the data and definitions of transactions are presented first.
SCOPE OF DATA A N D T Y P E S OF TRANSACTIONS

The discussion of dealer transactions is based on the dealers' answers
to the questionnaire on gross transactions in U.S. Government securities. Most of the dealers were able to report on their purchase and sale
volume for all of the 11 years by types of securities and by maturity
classes. I n some cases, missing information was supplied by the
dealers, from the daily reports of trading volume submitted in the past
to the Federal Reserve Bank of New York. Two-thirds of the dealers
were interviewed to establish the comparability of the reported data. 1
Varying dealer reporting methods and differences in the definition of
transactions necessitate a detailed discussion of the types of transactions included in the reported data and of the methods of valuation
and classification used by the dealers.
Valuation and classification by maturities
Transaction reports could be given on the basis of market value
or par value. I t appears that most dealers have reported volume on
the basis of par value, rounded to the nearest hundred thousand. F o r
the purposes of this study, volume data have been rounded to the
nearest million dollars.
A more critical problem is presented in the classification of securities
by maturities. The questionnaire asked for classification of notes
and bonds by final maturities. Many of the dealers have kept their
records on the basis of classification by earliest call date or final
maturity. The choice often depended on the market price of the
securities. Issues traded at a discount often were reported on final
maturity; issues traded at a premium were classified according to
1
I t was pointed out earlier t h a t t h e r e had been no concerted effort a t obtaining agreem e n t on reporting s t a n d a r d s for the dealer market until the recent Treasury-Federal
Reserve study and the present revamping of dealers' reporting methods by t h e Federal
Reserve Bank of New York.




STUDY OF FEDERAL GOVERNMENT SECURITIES DEALERS

51

earliest call date. For the types of issues without call dates, i.e., for
bills, certificates, and notes, these differences are irrelevant and would
not affect comparability of the data. For bond issues, however, variations in method of classification may lead to differences in reports
between individual dealers. Where earliest call date had been used,
trading volume in bonds of under 1 year and of 1 to 5 years will be
overstated relative to trading in the longer maturities. Since bonds
have been selling at discounts more frequently during the last 3 years
than in earlier years, the differences in reporting will probably be less
important for the years 1956 to 1958.
I t is impossible to estimate with accuracy the understatement of
trading in longer maturities which is introduced into the aggregate
data by the reporting of maturities on the basis of earliest call date.
I t appears that most of the dealers, reporting the largest part of the
aggregate trading volume, have reported to the Committee and to the
Federal Reserve on the basis of earliest call date whenever applicable.
Fiscal years and calendar years
Most of the dealers reported their trading volume on a calendar year
basis. Some of the dealers, however, reported on fiscal years different
from calendar years. No adjustments for these reporting differences
could be made except in a few cases, where data from the Federal
Eeserve were supplied on a calendar-year basis. Fiscal years overlap
with calendar years for at least 8 months in all cases, and for as much
as 10 and 11 months in others. Morover, most of the larger dealers
reported on a calandar-year basis. The data are sufficiently consistent
to allow for the analysis of interdealer differences and of year-to-year
fluctuations in aggregate trading volume.
Another timing difference in reporting transactions is introduced
with the distinction between recording transactions on a commitment
basis versus an executed basis. This problem was discussed earlier.
I t is largely irrelevant in this context because annual volume data are
analyzed. The delayed reporting of transactions by dealers on an
executed basis would be critical only in the analysis of daily or weekly
fluctuations of transactions volume.
Types of transactions to he included
The dealers were asked to report on purchases and sales of U.S.
securities exclusive of agency bonds, broken down by type of security
and by maturity classes for notes and bonds. Repurchase and resale
agreements were to be excluded from transactions. From interviews
with the dealers it became apparent that there exist numerous interpretations of the types of transactions which should be included
in, or excluded from, the dealer reports of purchases and sales of U.S.
securities.
Purchases and sales of Governments can be classified in a number of
ways. First, one can distinguish them by type of customers who are
parties to the purchase or sale contracts. I n order of importance in
the generating of dealer trading volume, these customers are commercial banks, other financial institutions, nonfinancial corporations,
the Federal Reserve Bank of New York, and other U.S. securities
dealers. All outright purchases from, and sales to, these institutions
constitute transactions to be included in volume reports. Allotments
of new securities by the Treasury to the dealers are treated as pur


52

STUDY OF FEDERAL GOVERNMENT SECURITIES

DEALERS

chases by most of the dealers. There are some exceptions, however,
which will be discussed in connection with the trading in rights below.
A second basis for distinguishing between different types of transactions is their economic function as viewed by the dealers and
their customers. Dealer purchases may constitute allotments from
the Treasury of new security issues; they may represent actions of
monetary restraint initiated by the open market desk; they may be
part of arbitrage or tax-swapping transactions or they may be completions of repurchase agreements. As noted earlier, dealers differ
about the character of repurchase agreements. The following discussion of the major types of transactions will also describe their
treatment in the reported data on transaction volume.
1. Outright sales to and purchases from customers
Commercial banks and other financial institutions are the most important customers of the dealers. Outright purchases and sales are
made by these customers for their investment portfolio and in adjustment of their secondary liquidity reserves. A considerable part of this
business appears to be conducted with commercial banks for tax reasons. This type of activity is known in the market as "tax swapping."
During profitable periods commercial banks may sell U.S. securities
at a loss and offset this capital loss against their taxable net operating
income. If the reserves acquired from the sale are used to purchase
other securities, the operation is called a tax swap. Since the loss is
fully deductible and any future gain is taxed at the capital gains rate,
banks with high annual income in a particular year are encouraged
to swap.
Nonfinaneial corporations have become increasingly aware during
recent years of the investment opportunity which short-term U.S. securities offer for temporarily idle corporate cash balances. I n addition to lending money to the dealer market under repurchase agreements, nonfinancial corporations are investors in bills and certificates.
The Treasury has offered special tax anticipation issues to attract
more short-term commitments of corporate funds to the Government
securities market. Moreover, some large corporations have their own
staff of money market experts who use corporate funds to engage in
arbitrage transactions.
The Federal Reserve Bank of New York appears among the customers of dealer firms in two roles. I t may act as fiscal agent for
foreign or Treasury accounts and it may buy or sell for its own open
market account. I n the execution of both functions, the Federal
Reserve conducts outright purchases and sales with the dealers. Some
available estimates support the conclusion that the Federal Reserve in
its buying and selling for the open market account alone has contributed roughly 5 percent of dealers' total transaction volume on the
average.
Two other categories of transactions are in evidence. (1) I n order
to fill customers' demands for purchases or sales of particular issues,
the dealers must often look for the securities in the market or find
another purchaser in the market if the dealer does not want to take
the issue into his position. One large order to sell from an insurance
company, for instance, may cause the dealer to first locate another customer who wants to buy this issue. Or, if the dealer is willing to take
the offered securities into his position, the addition may increase his



STUDY OF FEDERAL GOVERNMENT SECURITIES DEALERS

53

position in a certain maturity range beyond the amount he desires to
hold. He will then engage in transactions designed to readjust the
composition of his portfolio.
One customer transaction may thus give rise to a number of position adjustment transactions. The dealers take great pride in their
ability to handle large orders without disturbing the market or letting
the rest of the dealers know the size and direction of their activity in
the particular issues concerned.
Most of the interdealer trading seems to result from such position
adjustments connected with the handling of large orders. The amount
of trading volume conducted with other dealers varies between firms.
I t ranges from a low estimate of 10 percent of total volume for one
of the larger nonbank dealers to a high of 90 percent for one of the
smaller dealers. For all dealers as a group interdealer trading has
ranged probably from 15 to 25 percent of all trading during the
period.
(2) Theoretically all groups of customers mentioned above may
engage in arbitrage transactions to gain from price differentials
between comparable security issues.
Arbitraging speeds the transmission of monetary policy action from
the short end of the maturity structure to the long end and facilitates
the adjustment of the rate structure to changes in demand and supply
in segments of the money and capital markets. To bring about these
adjustments, only marginal sales and purchases, of course, are necessary. Not all holders of U.S. securities have to engage in arbitrage for
rapid adjustment of interest rates to be realized. I t is not clear which
institutions actually engage most heavily in this type of activity.
Certain financial institutions are in fact barred from engaging in
arbitrage transactions by their own stated policies or by legal restrictions. In the language of the market, these institutions are said to
be "locked in." The extent to which holders are locked in is an
empirical question which, together with the volume and characteristics
of arbitrage transactions, deserves further study.
All outright purchases and sales, regardless of the party to the
contract and of the economic function of the transactions pose no
serious problem of classification. They all add to the transactions
volume of the dealers. Only those sales and purchases which are
contracted under special terms are difficult to classify. The most important of these special transactions is the repurchase agreement in its
various forms.
2. Repurchase agreements
The dealers carry positions which are large in relation to their
own capital funds. They can employ their capital profitably only
by financing their positions with borrowed funds. Leverage enables
them to earn a competitive rate of return despite the low trading
profit margins prevailing in the market. Two principal methods of
financing are available to the dealers: (1) They can use the securities carried in their position as collateral for a loan from a commercial
bank; (2) they can sell the securities to a supplier of funds under an
agreement to repurchase the securities on a specified date at a predetermined price.
I n most dealer firms the second type of financing arrangement is
recorded as a sale and as a commitment to purchase the securities



54

STUDY OF FEDERAL GOVERNMENT SECURITIES DEALERS

at a future date. A sales and a purchase ticket are written and
add to the transactions volume of these dealer firms. If the dealers record their transactions on the commitment basis, the securities, which are no longer in the firm's physical possession, are still
carried as part of the dealers' positions. The legal obligation to
repurchase the securities sometime hence is strong enough to let
most dealers consider securities out on repurchase agreements as a
part of their position. This practice results in the recording of sales
and purchases which never influence the dealers' aggregate position.
I t appears that the character of E P ' s changes with their duration.
The longer the time span of the agreements, the more likely they are to
be viewed as an investment by the suppliers of the funds and as a sale
by the dealer. And the longer the agreement the smaller the influence of the securities on the dealers' day-to-day trading and position
decisions. The Federal Reserve Bank of New York requests the
dealers to separate RP's with a duration of more than 15 days from
those with shorter durations. Some of the dealers have suggested that
all HP's be treated as bona fide sales; others want to distinguish between RP's as sales or loans on the same time basis as the Federal
Reserve. A third opinion is that all RP's should be treated as loans.
One can take the position that, if short-term RP's are added to sales
and purchases, all collateral loans should be treated equally as part of
the dealers' trading volume. I t really does not make any difference for
the functioning of the market how RP's are classified. But it may
be important to the supplier of funds. National banks, for instance,
found the definition of RP's as sales and investments helpful before
the Comptroller of the Currency ruled that R P ' s are loans and subject
to the limitations under which national banks can lend money against
U.S. securities. Some State chartered banks continue to classify R P ' s
as purchases and investments. New York State for example modified
its ruling only recently to permit this interpretation. 2
F o r the purposes of this study the dealers were requested to omit repurchase and resale agreements from their reported purchases and
sales. I n some cases and especially for the earlier years of the reporting period, dealers were not able to separate RP's from transactions.
However, for the years since 1953 during which RP's increased rapidly, most dealers were able to omit volume generated by RP's. 3
Some data were available for the relationship of repurchase agreements to total trading volume for one of the larger nonbank dealers.
2
One solution for this general problem has been adopted by the Federal Reserve. Repurchase agreements are reported separately from purchase and sales and are broken down by
the maturity of the agreements. On the basis of this information any interested analyst
could rearrange the data for his particular purposes. Whether HP's are sales and investments or merely borrowing and lending is not answered by this method of reporting.
But similar difficulties are encountered in other financial areas where financing instruments are defined and separated on the basis of gradual transitions over time. Thus
the distinction between money and capital markets, between short-term, intermediateterm, and long-term instruments as well as the age-old discussion over whether time deposits are a part of the money supply, pose the same definitional problem as the classification of repurchase agreements. The solution has to be arbitrary.
In a later chapter the major suppliers of funds under repurchase agreements are described. The best discussion of the large variety of forms of repurchase and resale agreements can be found in the study of the Federal funds market conducted by the Federal
Reserve System referred to earlier.
P For 1958, the total reported trading volume in this study amounted to $353 billion.
The Treasury-Federal Reserve study gives a figure of $346 billion for the same period.
The difference may result from the failure to eliminate some $7 billion of repurchase
agreement transactions in the data supplied by the dealers for 1958. This introduces a
bias of about 2 percent and should be reflected mostly in an overstatement of trading
volume at the short end.




STUDY OF FEDERAL GOVERNMENT SECURITIES DEALERS

55

E P transactions amounted to 21 percent of the dealer's total sales and
purchases in 1956, 27 percent in 1957, and 19 percent in 1958. If this
relationship is typical, the volume of E P transactions for the dealer
market probably ranged from $36 to $50 billion annually during the 3
years. I t is interesting to note that not only bills and certificates but
also substantial amounts of short-term notes and bonds were sold
under E P ' s by this dealer.
3. Dealer underwriting and trading in rights
The dealers participate as underwriters in Treasury offerings of
both cash and exchange issues. Table I V - 1 summarizes the dealers'
role as underwriters of new issues of certificates, notes and bonds during the years 1953-58. The shares of the new issues which were
allotted to dealers and brokers annually is given for cash issues and
exchange issues separately. I t is assumed that the U.S. security
dealers account for almost all of the allotments to dealers and brokers
which are published in the Treasury Bulletin.
Dealers play a more important role in the underwriting of bonds
and notes than in the underwriting of certificates. They were allotted
a higher percentage of bond issues than of note issues and appear to
have been more active in exchange issues than in cash issues. The
following schedule relates dealer sales of allotted new securities to
total dealer sales volume in certificates and in notes and bonds. Dealer
volume derived from their underwriting activities is small in relation to their total sales of certificates, notes, and bonds. New bill
issues appear to be a more important source of transactions. The
scanty evidence available suggests that on the average the dealers
acquired about 20 percent of the regular wreekly 91-day bill issues. If
this ratio is applied to the Treasury bill issues for the years 1953 to
1958, the dealers have drawn down between $15 and $18 billion of new Treasury bills annually in the bill auctions. These
Sales of allotted

new securities

as percentages
1953

Certificates
Notes and bonds
Total...

-

of total dealer sales,
1954

1955

1956

1953-58
1957

1958

5.8
3.0

3.3
3.6

5.0
2.7

1.6
.6

5.2
2.5

7.3
2.6

8.8

6.9

7.7

2.2

7.7

9.9

figures, related to reported dealer sales of Treasury bills, lead to the
estimate that sales of new bills by the dealers have accounted for
roughly 20 to 25 percent of their total sales of bills.

62471—60——5




56

STUDY OF FEDERAL GOVERNMENT SECURITIES DEALERS

TABLE IV-1.—Allotments to "brokers and dealers
on subscriptions for public
marketable securities,11953-38
Dealer
allotment

Dealer
allotment
Total
issues
Amount

Certificate issues:
1953
1954
1955
1956
1957
1958
N o t e issues:
1953
1954
1955
1956
__ _
1957
1958
B o n d issues:
1953
1954
1955
1956
1957
__
1958

Dealer
allotment
Cash
issues

Exchange
issues
Percent

Millions
Millions
26,386
$727
28, 463
634
18, 951
698
140
19, 023
671
44,326
1,045
36,365

2.8
2.2
3.7
.7
1.5
2.9

11,172
9,257
27,094
14,165
8,705
9,233

230
839
776
248
665
507

2.0
9.1
2.9
1.8
7.6
5.5

6,213
21, 738
2,745

597
872
407

9.6
4.0
14.8

1,311
15, 588

131
1,972

10.0
12.6

Amount

Percent

Millicns
Millions
$612
20, 484
442
24, 729
462
10, 569
122
15, 802
666
40,956
941
32, 798

3.0
1.8
4.4
.8
1.6
2.9

11,172
2,897
24, 562
14,165
4,620
4,078

230
276
714
248
363
136

2.0
9.5
2.9
1.8
7.8
3.3

1,748
21, 738
1,924

269
872
354

15.4
4.0
18.4

12,969

1,691

13.0

Amount

Millions
Millions
5,902
$115
3,734
192
8,382
236
3,221
18
3,370
5
3,567
104

Percent

1.9
5.1
2.8
.6
.1
2.9

6,360
2,532

563
62

8.8
2.4

4,085
5,155

302
371

7.4
7.2

4,465

328

7.3

821

53

6.4

1,311
2,619

131
281

10.0
10.7

1

Excludes the issuance of 1% percent Treasury notes available in exchange to holders of nonmarketable
2% percent Treasury bonds, investment series B, 1975-80, and special allotments to Government accounts
in 1957.

Most of the dealers reported the allotment of new securities as a
purchase. This seems to have held true for all allotments of Treasurybills and other securities issued for cash. The treatment of the transactions involved in exchange issues has not been so uniform. I n a
refunding the maturing issue carries with it the right to subscribe to
the issue offered in exchange. Investors holding these rights often
are not interested in adding the exchange issue to their portfolios.
They have the option of demanding cash for the maturing issues or
selling the maturing securities to investors interested in acquiring
the issue offered in exchange or the rights.
Rights trading often takes place in advance of Treasury announcements of refundings and reaches its peak during the subscription
period. The purchasing of rights by the dealers is normally paralleled
by sales of the new exchange securities on a when-issued basis immediately following the Treasury announcement and during the subscription period.
The dealers act as quasi-underwriters in providing a market for
rights; i.e., the maturing securities. This dealer action may help the
Treasury to increase the proportion of the expiring issue exchanged.
Successful execution of this function depends on the condition of the
money market at the time and on the attractiveness of the terms of the
new securities. If the market views the terms of the new issue as
more favorable than those on existing comparable instruments, the
maturing issue will go to a premium. Holders of the rights who
perceive better alternatives will sell their rights. Through arbitrage
in the dealer market the price differentials between the maturing, new,
and comparable existing issues are reduced or eliminated. If the terms
of the new issue offer no advantage over existing comparable issues,
the rights will not go to a premium unless the rate on the maturing




STUDY OF FEDERAL GOVERNMENT SECURITIES DEALERS

57

securities exceeds market rates on money market instruments of the
same maturity.
There are three possible courses of trading in rights issues. (1)
The dealer purchases rights and holds them for exchange into the
new issue. He then sells the new issue after allotment. (2) He acquires rights and offsets the purchase with a sale of the new security
on a when-issued basis. (3) The dealer acquires the rights and resells
them to an institution (or individual) interested in acquiring the new
issue. The course of action chosen will depend on the dealers' expectations about short-term changes in the money market and the
relative prices of the rights and of the new issue.
I n this context it should be noted that some dealers have made a
practice of recording the exchange of rights for the new issue under the
trading courses (1) and (2) described above as both a sale and a
purchase of securities. As a result they would record and report four
transactions where most dealers would have reported only two: the
purchase of the rights and the disposition of the new security after
allotment or on the when-issued basis. I n view of the relatively small
proportion of sales of new exchange issues to total dealer sales, this
recording and reporting difference does not affect the comparability
of the data significantly.
VOLUME OF DEALER TRADING

The preceding sections described the types of transactions included
in the reported dealer trading volume which are summarized in table
IV-2. Aggregate dealer trading volume has tended to rise over the
entire 11-year period. This upward trend was interrupted in only
3 years, 1949,1951, and 1955. The decrease in trading volume in 1949
is small, but the other 2 years show substantial declines; 1953 and 1956
show only small increases in volume. Among the other 5 years in
which trading activity increased, 3 years stand out because of substantial increments over preceding years: 1950, 1954, and 1958. The
following analysis of the determinants of fluctuations in trading
volume will focus on the experience during the 5 years with rapid
downward or upward changes in trading volume.
Five explanations of the secular increase and of year-to-year changes
in trading volume will be discussed.
(1) Trading volume may move
in proportion to the stock in trade, in this case of course with the
amount of public marketable U.S. securities outstanding. (2)
Changes in the composition of the U.S. public marketable debt, i.e.,
in the average maturity of the debt, may increase or decrease the
trading volume in the dealer market. (3) The underwriting activity
of the dealers in U.S. securities may add to their trading volume in
years of large financing activity by the Treasury. (4) Increases or
decreases in the amount of securities outstanding may induce trading
activity designed to achieve desired portfolio adjustments by investors
and dealers. I n other words, retirements, exchanges, or cash offerings may cause trading volume to increase or decrease temporarily.
(5) Changes in monetary policy, in the level and structure of interest
rates and in the demand and supply situation in the money and capital
markets may cause trading volume to increase or decrease through the
encouragement or discouragement of portfolio readjustments, arbitrage operations and speculative activity.



a
o

TABLE IV-2.—Total transactions

volume and rates of change in volume by maturity class, all reporting dealers, 1948-58]

>

[In millions of dollars and in percentage of 1948]
1

Total transactions reported
Bills
Certificates _._
_
Notes and bonds, all
__
Under 1 year_
1 year to under 5 years.
5 years and over..
All securities under 1 yearTotal transactions reported

1948

1949

1950

1951

1952

1953

1954

1955

1956

1957

! 177,489

176,116

228, 753

194,460

217, 555

219, 225

285, 341

262, 358

262, 994

286,999

100
100
100
100
100
100
100
100

108
102
82
68
75
93
102
99

122
102
144
245
129
102
130
129

131
36
112
241
68
75
116
110

153
88
84
98
91
71
128
122

156
68
90
160
81
59
132
124

176
103
154
191
170
123
157
161

165
76
143
108
203
117
134
148

181
49
140
205
183
86
145
148

221
70
102
155
145
46
171
162

1958
353,005
218
78
187
238
251
117
181
199

a
o
<
W

o
1

Total transactions are available for all dealers; detail is based on 15 dealers only.




a

>

STUDY OF FEDERAL GOVERNMENT SECURITIES DEALERS

59

The relationship between securities outstanding and trading volume
can be expressed as the ratio of purchases or sales to the stock of securities outstanding. For this purpose, total trading volume 4 figures have
been divided in half in order to obtain a measure of turnover. Table
I V - 3 presents the turnover ratios of outstanding U.S. public marketable securities by type of instrument and by maturity classes.
TABLE IV-3.—Ratios of sales volume to securities outstanding at the end of
calendar years 1948-58

AU
securities

Bills
(quarterly
sales)

Certificates

AU
notes
and
bonds

Notes and bonds Notes and bonds
under 1
1 to u n d e r 5

Cain

1948
1949
1950_
1951
1952
1953
1954
1955
1956
1957___
1958

__

_
___
__
._

Averages of
t h e 11 ratios.

Final
maturity

Cain

Final
maturity

B o n d s 5 years
a n d over

Call*

Final
maturity

0.34
.36
.51
.42
.42
.41
.57
.50
.48
.47
.64

0.81
.86
.89
.72
.70
.79
.89
.73
.71
.82
.73

0.69
.64
.35
.23
.97
.48
.67
.89
.47
.37
.39

0.26
.22
.33
.36
.23
.25
.43
.35
.37
.30
.52

0.69
.34
.45
.67
.19
.38
.92
.27
.45
.59
1.12

1.96
.77
.65
129.00
.38
.42
1.12
.34
.56
.85
2.63

0.22
.20
.37
.22
.29
.29
.41
.45
.38
.30
.48

0.31
.18
.29
.14
.23
.38
.55
.51
.42
.30
.46

0.22
.21
.24
.26
.23
.16
.31
.30
.30
.17
.37

0.16
.19
.22
.21
.18
.13
.25
.25
.24
.15
.32

.46

.79

.56

.33

.55

2.97

.33

.34

.25

.21

1
The stock of outstanding notes and bonds is classified according to final maturity or earliest call date
under
"Call," according to final maturity only under "Final maturity."
2
The 1951 ratio was omitted in the calculation of the average ratio.

For notes and bonds, two turnover ratios are presented. One is
based on the classification of the securities according to final maturity
or earliest call date where applicable; the other is based on the classification according to final maturity only. Table I V - 4 summarizes
the outstanding public marketable debt under both classifications. As
was pointed out earlier, the classification by maturity or earliest call
date seems to be the more appropriate one for purpose of comparison
with the volume data reported by the dealers. But since some of
the dealers reported volume in bonds on the basis of final maturity,
both turnover rates are given in table I V - 3 to establish a relevant
range within which the true turnover rate will fall.
4
Purchases and sales for the dealers in the aggregate as well as for individual dealers
have each contributed almost constantly 50 percent of total t r a d i n g volume. Any excess
of purchases over sales would indicate a buildup of positions, and any excess of sales
over purchases would be a reflection of t h e liquidation of positions. Minor recording
and reporting discrepancies might also a t times account for small deviations from the
basic equality of purchase and sale volume for a year. T h e organized securities exchanges
t r e a t purchase and sale of a security as one t r a n s a c t i o n . The same practice has been
followed here.




o
1-3

d
o

TABLE IV-4.—Outstanding public marketable U.S. securities, end of calendar years 1948-58

»=!

[In millions of dollars]
1951
All securities
Treasury bills
Certificates
Notes and bonds, all
Due or callable:
Under 1 year
1 to under 5 years.
5 years and over..
Final maturity:
Under 1 year
1 to under 5 years.
5 years and over..
All securities under 1 year:
D u e or callable
Final maturity




H
1952

1953

175, 586
29,748
36, 364
109,474

30, 030
26, 675
52, 028

14, 865
39, 267
55, 736

28, 828
43,199
53,170

30, 550
45, 963
39, 659

18, 621
45, 367
38, 792

15, 227
50,013
44,234

27, 338
20,106
61, 289

12, 254
29, 606
68, 008

22, 576
38, 307
64, 314

24, 354
41, 021
50, 797

12, 957
46, 513
43, 310

6,504
52,319
50,651

80, 032
74,368

81,339
72,616

148, 581
21,713
16, 712
110,156

154, 630
19, 511
26, 386
108, 733

10,329
44,053
64, 352

14, 431
35, 067
63,671

29, 013
33, 378
61,059

25, 610
28, 678
41, 218

35, 836
30,196
44,124

3,648
30,105
84,981

6,343
38,983
67, 843

27, 017
42, 082
64, 351

102
45, 322
50, 082

18, 613
37, 628
53, 915
74, 261
57, 038

1958

164,191
26, 857
34, 554
102, 780

142, 686
18,102
29, 078
95, 506

72, 790
47, 282

1957

160, 374
25,179
19,023
116,172

152, 450
13, 627
5,373
133, 450

58, 013
46, 017

1956

163,251
22, 313
15,741
125,197

155,124
12,319
29, 636
113,169

56, 386
48, 298

1955

832
506
458
868

157,483
12,224
26, 525
118, 734

49,078
42, 397

1954

75, 927
73,235

157,
19,
29,
109,

62, 829
60, 218

66, 882
60, 630

74, 752
68, 556

o
o
<!
St

CO

o
d
tfi

5
GO

STUDY OF FEDERAL GOVERNMENT SECURITIES DEALERS

61

Trading volume in all securities has ranged from a low of 0.34 percent of the total amount of securities outstanding in 1948 to a high
of 0.64 percent in 1958. The variation around the mean of the annual
ratios (0.46 percent) is too large to allow for the acceptance of the
hypothesis that trading volume changes in proportion to the stock
of outstanding securities. The widest variations from the mean
occurred during the years which show the largest increases in trading
volume. For the 2 years of large decreases in volume, the turnover
rates for all securities fell short of (1951) and exceeded (1955) the
median ratio for all years while the stock of securities decreased during
1951 and increased in 1955. During years of large increases in volume
(1950, 1954, and 1958) the turnover of all outstanding securities increased over preceding years and exceeded the average rate of turnover for all years. I n 1950 the increase in the rate of turnover was
offset partially by a decrease in the stock of outstanding securities;
in 1954 and in 1958 increased turnover was reinforced by a rise in
the public marketable debt.
This evidence leads to the conclusions that (1) trading volume does
not change in direct proportion to the amount of public marketable
securities outstanding. Frequently the stock of securites and trading volume move in opposite directions. (2) Over the 11-year period
both rates of turnover and amount of securities available for trading
have increased. Causes of the observed variations in turnover during
the 5 years of abrupt increases or decreases in trading activity and the
secular increase of the rate of turnover for all securities remain to be
explained.
Table I V - 5 gives the percentage composition of the outstanding
public marketable debt at the end of the calendar years 1948-58, classified by both final maturity and earliest call where applicable. The
data reflect the shortening of the average maturity of the public debt
from 1950 to 1953 and, again, after 2 years of interruption, from 1955
to 1957. The reduction in the proportion of bonds with over 5-year
maturities is especially noticeable. This development must be compared with the data in table I V - 3 which show that rate of turnover
of classes of outstanding securities decreases with the increase in number of years to maturity. Bonds with a maturity of 5 years and more
have turned over during the period at an average rate of 0.25, while
certificates showed a turnover rate of 0.56, and bills turned over at
the rate of 0.79.




62

STUDY OF FEDERAL GOVERNMENT SECURITIES DEALERS

TABLE IV-5.—Percentage composition of public marketable U.S. securities, end of
calendar years, 19^8-58

All securities
Treasury bills
_.
Certificates
Notes and bonds, all__ __.
Due or callable:
Under 1 year.
1 to under 5 years
5 years and over ._.
Final maturity:
Under 1 year
1 to under 5 years
5 years and over
All securities under 1 year:
Due or callable._ _._
Final maturity

1948

1949

1950

1951

1952

1953

1954

1955

1956

1957

100
8
17
75

100
8
19
73

100
9
4
88

100
13
20
67

100
15
11
74

100
13
17
70

100
12
18
70

100
14
10
77

100
16
12
72

100
16
21
62

100
17
21
62

6
28
41

9
23
41

26
22
40

18
20
29

24
20
30

19
17
34

9
25
35

18
26
33

19
29
25

11
28
24

9
28
25

2
19
54

4
25
44

18
28
42

1
32
35

12
25
36

18
13
40

8
19
43

14
23
39

15
26
32

8
28
26

4
30
29

31
27

36
31

39
31

51
34

50
38

49
48

39
38

42
38

47
43

48
45

47
42

1958

The rate of turnover for Treasury bills has remained fairly stable
over the period. The annual rates for certificates have fluctuated
more widely, probably because of the wide fluctuation in the stock
of certificates outstanding and the large holdings of these instruments
by the Federal Reserve System. I n contrast, the rates of turnover
for notes and bonds of all maturities have not only shown wdde annual
fluctuations but have also tended to increase during the last 5 years
of the period over the level which prevailed in the earlier years. The
secular increase in the rate of turnover of all securities was apparently caused by the shift in maturity composition of the outstanding public marketable debt toward the short end and by the higher
rates of turnover for notes and bonds during the 5 years 1954-58.
The proportion of allotments of new securities to dealers' total
trading volume is not large enough to account for the observed variations in trading volume. This conclusion would not be changed if
the allotments of Treasury bills wTere added to the allotments of new
securities. Other factors must be considered.
The size of the stock of outstanding securities and its maturity
composition were not the principal causes of changes in trading
volume. But this conclusion holds only for the distributions of maturity which have been observed during the postwar years. F o r it
is doubtlessly true that a large permanent change in the maturity composition of the debt, e.g., a major refunding into longer term maturities, would alter both the distribution of the debt among holders and
the trading characteristics of particular maturities. Under these
conditions, longer term securities would become somewhat more
marketable than they presently are. Changes in turnover ratios for
particular maturity classes would be likely to follow. Moreover, the
effect of changes in the size and composition of the stock and their
effect on trading volume in a given year has not yet been discussed.
The stock of U.S. public marketable securities increased in 6 of the
11 years and decreased in 1948 through 1951 and in 1956. There is
no correlation between the absolute changes in the stock of securities
and the changes in trading volume as a comparison between tables
I V - 4 and I V - 2 shows. However, in most postwar years, changes in
the stock have been small relative to the total and the methods by
which the stock of marketable securities was changed and the types



STUDY

OF FEDERAL

GOVERNMENT

SECURITIES

DEALERS

63

of securities retired or issued may have important influences on the
rates of turnover and the resulting trading volume.
The first observed sizable change in trading volume took place in
1950. This is also the first postwar year with large Treasury refunding operations in intermediate maturities. Roughly $39 billion
of Treasury notes were issued in exchange during the year. This
suggests some relationship between Treasury financing operations and
the dealers' trading volume. Nineteen fifty-four is another year with
a large amount of financing operations. $45.5 billion of certificates,
notes, and bonds were issued for cash ($10 billion) and in exchange for
maturing securities ($35.5 billion) ; 1954 also shows one of the large
increases in dealers' trading volume.
I n 1958 $11 billion of securities other than Treasury bills were
issued for cash, and $28 billion was exchanged. I n 1955, large Treasury financing was undertaken but trading volume declined during
the year by a substantial amount. Another exception to the apparent
rule that the amount of Treasury financing influences changes in
trading volume is the experience of 1953. The Treasury issued almost
$31 billion of securities during the year, but trading volume remained
virtually unchanged. On the other hand, trading volume declined
substantially during 1951 while Treasury activity was relatively low.
I t appears that the amount of financing undertaken by the Treasury
is related to the increases or decreases in dealers' trading volume
in some years and is a leading cause of changes in the volume of trading. The distribution of Treasury offerings between cash and exchange issues, however, does not help to explain the failure of trading
volume to respond to large financing activity in 1953. Cash offerings
were attractive to many commercial banks because of the advantages
of subscription through the tax and loan accounts during a recession.
This practice may have resulted in the selling off by the banks after
the interest advantage was realized and this could have increased
volume subsequent to Treasury cash issues. On the other hand, exchange issues invited speculation in rights often well before the announcement of the issue and caused active rights trading during the
subscription period. I t is likely that exchange issues induced more
transactions in the dealer market relative to their size than did cash
issues. But the difference in ratio of cash issues to exchange issues
between 1953 and 1954 is not sufficient to explain the differences in
trading volume during the 2 years. 5
Both 1954 and 1958 were periods of monetary ease. Large Treasury
financing activity coincided with lowered discount rates and an increased supply of funds to the money market by the Federal Reserve, The expectation of further antirecessionary measures and the
provision of longer term securities by the Treasury invited purchases
of the new issues for price appreciation. Moreover, offerings of
alternative investments decreased.
The dealers felt these effects doubly. Their own expectations led
them to increase their portfolios to take advantage of the expected
wider spreads between acquisition and sale prices. The demand for
U.S. securities facing them in the market led to increasing levels of
sales which in turn necessitated additional positioning of securities
5
An extrenie example of t r a d i n g activity induced by an exchange occurred in J u n e
1958. This is considered extensively in t h e Treasury-Federal Reserve Study, P a r t I I
and in Hearings, 1959 P a r t s GA, B & C.




64

STUDY OF FEDERAL GOVERNMENT SECURITIES DEALERS

on the part of dealers. 6 Policies of monetary ease thus reinforced the
impact on dealers' trading volume of the issuing of new securities by
the Treasury.
I n periods of monetary restraint the opposite occurs. Dealers
hesitate to take large positions in periods of generally declining interest rates. This situation seems to have prevailed in 1953 and 1955,
when large Treasury financing operations coincided with a practically
unchanged trading volume and a sharp decrease in the volume of
dealer trading, respectively. I n these 2 years and during other periods
of monetary restraint virtually no bond issues were offered by the
Treasury; during 1954 and 1958 large bond issues were brought on the
market.
The offering of shorter maturities in prosperity further reduces
total speculative interest because of smaller price fluctuations which
accompany interest rate changes at the shorter end of the maturity
spectrum. While there seems to be considerable speculative interest
in U.S. securities during times of falling interest rates both within
and outside of the dealer market, this interest diminishes sharply
during periods of rising interest rates. Also, as we reported earlier,
dealers view speculative short sales in Governments as relatively risky
and expensive.
Many of the investors purchasing new issues during periods of
rising interest rates probably buy them for liquidity reasons. The
terms offered by the Treasury are not attractive enough to divert
investor demand from other segments of the capital markets. The
offering of short-term securities and the demand for liquid instruments combined to cause the decreases in dealer trading volume during
the years 1951 and 1955. As a result, 1951 was a low profit year for
the dealers, and in 1955 the dealers as a group lost money. This may
have further discouraged them from taking positions or from shading
prices quoted to customers to prevent further increases in position.
I n conclusion it can be stated that changes in the rates of turnover
and dealers' trading volume are related to a number of factors. Among
these the direction of change in monetary policy probably ranks first
in importance. Treasury financing actions rank second and act as a
reinforcing factor. The size of the stock of outstanding public marketable securities and its maturity composition are also important but
more for the secular trend in trading volume than for the changes in
turnover from year to year. Subsidiary, but nevertheless important
influences, are the failure of the Treasury to compete effectively with
other segments of the capital markets during tight money periods and
the growing sophistication of an increasing number of market participants about the magnitude of changes in bond prices. The last force
was felt especially during 1958, the year showing the largest absolute
and relative increase in dealer trading volume. Attractively priced
new issues and the expectation of continued antirecessionary measures
induced many old and new market participants to take positions in
the U.S. securities market. With unchanged debt management
policies, similar forces can be expected to become increasingly effective
during periods of tight money policies in the future.
6
Through the process of arbitrage, price appreciation was carried throughout the
maturity spectrum and invited speculation in old issues as well as in the new securities
offered by the Treasury.




STUDY OF FEDERAL GOVERNMENT SECURITIES DEALERS

65

COMPOSITION OF DEALER TRADING

The composition of trading volume in the dealer market is determined by the same factors responsible for the changes in trading
volume. Only the emphasis changes from the influence of monetary
policy to the reactions to it by the Treasury in form of its debt management operations. Table I V - 6 presents the composition of trading
volume by types of securities and maturity classes. The strongest
trends over the entire period are the increase of Treasury bill volume
as a proportion of total dealer transactions and the decreasing importance of long-term Treasury bonds. These developments appear
to be a direct reflection of changes in debt management practices.
Despite the pronounced goal of lengthening the maturity of the
public debt, the Treasury has not been able to realize its objective in
the presence of prevailing monetary policies and developments in the
capital and money markets.
TABLE IV-6.—Composition of total trading volume, all reporting dealers, 19J/8-58
[Items as percentages of total transactions]
1948

1949

1950

1951

1952

1953

1954

1955

1956

1957

44.6 49.2 43.4 55.9 59.2 60.6 51.3 53.3 57.9 66.4
T r e a s u r y bills
9.8
7.1 15.7 12.4 14.0 11.4
7.3
Certificates
20.7 21.8 16.9
6.3 11.0
8.4
Notes a n d b o n d s u n d e r 1 y e a r . . . . 8.1
6.8 11.3 10.1
5.7 15.7 18.6
N o t e s a n d b o n d s over 1 year,
8.2 11.0
u n d e r 5 years
7.0
8.5
7.6 11.9 15.7 14.0 10.4
10.7
4.9
B o n d s over 5 years
15.9 15.2 12.9 11.4
9.8
8.2 12.8 13.4
9.7
100.0
100.0
100.0
100.0
100.0 100.0 100.0 100.0 100.0 100.0
T o t a l transactions
All notes a n d b o n d s over 1 y e a r . . 26.6 23.4 23.9 18.4 18.3 15.8 24.7 29.1 23.7 15.3
73.4 76.7 76.0 81.6 81.7 84. 3 75.4 71.0 76.2 84.6
All securities u n d e r 1 year
.1
.1
.1
.1
.1
.1
Discrepancy
.1
15
15
15
15
N u m b e r of dealers reporting
15
15
15
15
15

1958
54.7
9.0
10.8
15.1
10.4
100.0
25.5
74.5
15

The distribution of trading volume between all securities with maturities under 1 year and notes and bonds maturing in 1 year and
more has remained fairly stable over the period. Trading in notes
and bonds of maturities up to 5 years took the place of trading in
long-term bonds. The volume of trading in certificates has declined
generally, mostly because of the large holdings of these instruments
by the Federal Eeserve. Large annual changes in the stock of certificates outstanding have accounted for the fluctuation in their proportion of dealer trading over the 11-year period.
The aspects of composition of dealer trading which have not been
covered by the discussion in the preceding section of this chapter are
the interdealer differences. Table I V - 7 presents individual dealers'
ratios. The table shows the ratio of trading volume in a particular
maturity to total trading volume. The data presented are for dealers
with the highest and lowest annual ratios for each maturity class.
Averages for the group as a whole are also shown.




66

STUDY OF FEDERAL GOVERNMENT SECURITIES DEALERS

TABLE IV-7.—Composition of trading volume—Annual high, low, and average
ratios for individual dealers, 1948-58

As p e r c e n t a g e s of t o t a l t r a n s a c tions:
Bills:
High
__
Low
Average
Certificates:
High
„
Low
_.
Average
Notes and bonds, under 1
year:
High
._
Low
Average
N o t e s a n d b o n d s , 1 to u n d e r
5 years:
High
Low
„
Average
B o n d s over 5 y e a r s :
High
Low
Average

1948

1949

1950

1951

1952

1953

1954

1955

1956

1957

73.7
1.1
39.4

77.2
.7
43.6

73.8
15.4
38.1

90.8
29.4
54.4

99.8
23.7
58.5

97.3
41.7
61.9

92.8
35.7
53.5

95.2
19.8
54.7

97.4
37.0
62.5

96.3
56.6
69.7

92.7
42.4
57.6

27.4
11.8
20.0

33.8
8.4
20.2

29.0
6.3
15.8

18.7
.9
5.8

23.2
.2
13.2

15.8
1.7
11.2

21.2
4.2
11.1

15.2
2.6
8.8

11.0
.3
6.0

13.0
1.9
8.3

11.2
2.6
7.7

11.9
3.7
6.6

14.9
1.5
5.7

25.0
3.8
13.9

27.4
5.2
16.0

10.8
.1
6.0

18.6
.5
9.5

14.7
5.7
9.5

12.4
.3
6.5

18.6
.3
9.4

12.8
.5
6.4

16.9
1.8
11.1

22.2
5.5
12.0

16.5
3.6
9.5

18.9
8.0
12.2

16.2
1.2
9.0

13.3
4.2
7.9

14.3
.4
7.3

18.7
2.3
12.2

32.5
1.5
15.8

21.5
1.1
12.9

18.6
1.2
10.3

21.2
.9
12.5

56.2
5.2
20.6

68.3
5.0
21.0

68.5
3.9
19.3

85.3
2.0
19.4

63.2
3.4
16.0

32.7
.1
10.0

34.0
.6
13.7

32.8
.3
14.3

18.5
.9
9.1

11.4
.1
5.2

30.3
2.0
11.8

1958

Dealers specializing in particular classes of securities show the high
ratios for these securities in most of the 11 years. One dealer, for
instance, accounts for 10 of the 11 annual high ratios of bill volume
to total volume. Another dealer shows the lowest ratio of bill trading
to total trading of all dealers during 5 of the 11 years. The average
ratio of bill volume to total trading indicates however, that bill trading accounts for most of the trading volume for the majority of
dealers. Six of the dealers derived more than 60 percent of their total
volume in at least 5 of the last 6 years from trading in bills—among
these six dealers are four of the five bank dealers.
Similar specialization is apparent for other types of securities. The
annual high ratios of certificate volume to total trading for the last 5
years for instance are those of one larger nonbank dealer. The same
firm also accounts for four of the high annual ratios in trading of notes
and bonds maturing in less than 1 year. Some smaller nonbank dealers appear as specialists in long-term bonds. One of these firms had
the highest ratio of trading in long-term bonds to total trading in 8 of
the 11 years.
F o r the dealers as a group as well as for individual dealers the table
shows that the bill volume ratio increased during periods of tight
monetary policies and decreased during periods of monetary ease. The
opposite can be observed for the trading ratio in the long-term bonds
which decreases with rising interest rates and increases with a fall in
the level of interest rates. This again suggests that monetary policy
in combination with the Treasury's financing policies and actions is
an important determinant of the composition of dealers' trading volume as well as of changes in total volume.
CONCENTRATION OF TRADING I N THE DEALER MARKET

One frequently used measure of concentration in an industry is the
percentage of total industry sales accounted for by the largest firms.
Total transactions of individual dealers are used here as the measure



STUDY OF FEDERAL GOVERNMENT SECURITIES DEALERS

67

of sales. Table I V - 8 gives the share of market accounted for by
the three, five, and eight largest dealer firms. Besides the annual
share of total transactions executed by these three groups of firms, a
further breakdown by types and maturity classes of securities is
provided.
TABLE IV-8.—Share of market transactions, 19Jt8-58

T o t a l transactions:
3 largest firms
5 largest firms
8 largest firms
N u m b e r of r e p o r t i n g
dealers
Transactions in—
Bills:
3 largest firms
5 largest firms
8 largest firms
Certificates:
3 largest firms
5 largest firms
8 largest
firms
.__
Notes and bonds, under
1 year:
3 largest firms
5 largest firms
8 largest
firms
___
Notes a n d bonds, 1 to
u n d e r 5 years:
3 largest firms
5 largest firms
8 largest firms...
B o n d s , 5 years a n d over:
3 largest firms
5 largest firms
8 largest firms
N u m b e r of reporting dealers

Average
194858

1948

1949

1950

1951

1952

1953

1954

1955

1956

1957

1958

46.3
66.0
87.3

42.6
62.0
84.9

42.5
62.1
85.1

43.1
61.3
82.3

43.3
60.9
81.8

45.0
60.6
81.7

43.8
60.3
80.4

43.7
61.8
81.9

41.8
59.7
81.4

43.7
61.1
81.5

41.8
61.2
82.1

16

16

16

16

16

16

16

16

16

16

16

43.4
65.3
87.4

45.1
67.5
86.2

49.3
71.1
89.5

47.3
68.4
87.6

44.9
64.9
88.5

47.1
64.0
86.3

42.7
62.2
84.4

43.5
66.7
85.4

40.6
61.8
83.6

45.7
64.6
84.4

47.0
64.8
84.0

45.1
65.6
86.1

47.1
67.6
88.4

44.9
65.6
87.5

47.2
67.1
87.2

63.1
77.9
91.8

53.5
70.2
91.7

51.5
69.2
88.2

58.8
74.1
91.4

57.6
74.8
91.9

53.1
69.8
90.8

58.1
73.7
90.3

53.1
72.6
89.7

53.4
71.1
89.9

55.8
71.5
88.2

55.6
72.5
89.9

52.3
69.7
89.9

50.8
69.0
89.7

53.0
71.9
90.4

51.9
69.7
88.2

47.1
67.1
88.2

45.1
65.2
87.5

59.0
74.3
90.4

62.3
77.7
92.4

48.2
67.9
87.3

52.8
70.6
89.3

50.8
69.5
87.3

44.6
63.2
84.8

43.7
63.4
87.6

40.2
61.1
85.0

55.8
72.4
91.1

56.2
71.5
87.9

46.0
66.1
86.4

50.8
68.9
88.2

53.2
69.6
88.9

49.1
65.9
87.5

55.1
70.4
90.0

49.6
67.4
87.7

53.7
73.7
88.0

51.9
72.4
87.9

52.4
73.9
88.1

53.9
74.3
87.9

51.1
70.2
85.9

57.2
72.0
87.0

54.1
71.7
88.3

50.6
70.7
89.0

54.4
72.0
89.9

57.2
74.9
89.1

49.0
66.4
85.2

53.2
72.0
87.8

15

15

15

15

15

15

15

15

15

15

15

43.4
61.5
82.8

The eight largest dealers accounted for more than 80 percent of all
transactions in the dealer markets in all 11 years; the largest five for
more than 60 percent and the largest three for over 40 percent in all
years. The members of the three groups of dealers are not the same
in all 11 years, however. This suggests greater mobility in the dealer
market than is encountered in most highly concentrated industries
where the same firms appear as the largest firms for long periods of
time. The following schedule summarizes the number of firms appearing among the leading group over the period.
Number of dealers appearing among the
largest—
3 firms
Total transactions
Treasury bills
Certificates
Notes and bonds:
Under 1 year
1 year to under 5 years.
5 years and over




5 firms

8 firms
10
11
9
10
10
11

68

STUDY OF FEDERAL GOVERNMENT SECURITIES DEALERS

From the data in table I V - 8 it also appears that that the degree
of concentration of trading was higher for all individual classes of
securities than for total trading transactions. The number of dealers
appearing among the largest firms in any one of the five maturity
ranges exceeded the greatest number of dealers represented in each
of the three groups given in the schedule above. Among the three
leading firms in any of the security classes during the 11 years,
eight separate firms appeared. Nine firms were included among the
leading 5 at various times and 13 among the leading 8. Two firms
were not represented in the data.
This evidence supports the conclusions reached earlier in the discussion of dealer specialization. The degree of specialization differs
between the five classes of securities. I t is highest in certificate
trading and lowest in Treasury bill trading. Interesting is the relatively large participation by small dealer firms in the trading of longterm bonds. These relatively small dealers specialize in particular
instruments to the extent that their trading volume in bonds, for
example, often is as large as the volume in bonds generated by some of
the large and well-diversified dealers.
The share of trading volume of the three largest firms has decreased
in years of falling interest rates and increased in years during which
the Federal Reserve System pursued tight monetary policies. This is
noticeable also in their share of trading in long-term bonds. But it
shows up most clearly in their share of certificate trading. This
reflects reliance by the Treasury on certificates in periods of tight
money and the large dealers' specialization in certificate trading. The
fact that the degree of concentration in long-term bonds lessened
during years of falling interest rates is another indication of the
heightened trading activity and increased position taking of a larger
number of market participants in periods of monetary ease.
The data presented in table I V - 8 strongly suggest that six of the
firms form the permanent core of the dealer market. Together they
are active over the entire maturity range though some of them
specialize to some extent. They are also the dealers with the largest
sales staff, and with few exceptions, they have the largest number of
branch offices throughout the country. However, over the 11-year
period their share of the U.S. securities market has decreased.
FINDINGS

Trading volume in the U.S. securities market is larger than the
volume of transactions in any of the organized securities exchanges.
Transactions on the New York Stock Exchange, for instance,
amounted to $32.7 billion in common and preferred stock and $1.4
billion in bonds during 1958. This total N Y S E trading volume compares with sales of U.S. securities through the dealer market of $176
billion in 1958. Data on the volume of transactions in corporate and
municipal securities in the over-the-counter are not available but it is
safe to assume that they would show a much smaller transactions
volume in those markets than the 17 dealers have reported for the
U.S. securities market.




STUDY OF FEDERAL GOVERNMENT SECURITIES DEALERS

69

Secular and cyclical changes in trading volume and rates of turnover
of outstanding stocks of U.S. securities are caused mostly by the interaction of debt management practices and the changes in direction of
monetary policy pursued by the Federal Reserve System. The varying degree of concentration during tight and easy monetary policy
periods also reflects the impact of monetary policy on the volume of
transactions in the market. While six firms have accounted for the
major part of the trading volume over most of the period, no tendency
toward increased concentration is reflected in the data.







CHAPTER V
FINANCING T H E GOVERNMENT S E C U R I T I E S M A R K E T
Financing inventories probably poses more problems for the individual dealers than any other single activity. This is particularly
true during "tight money" periods when dealers must search the entire
country for sources of funds. Through this search process dealers
locate excess reserves of banks or idle cash balances of corporations
and bring them into use. And partly through this process the effects of
Federal Reserve policies are transmitted to banks and to nonbanking
organizations throughout the country.
Historically, broker and dealer loans by large banks in New York
have been the principal source of financing for the Government securities market. Such loans were generally both callable and collateralized
and provided the banks with a secondary reserve. The introduction of
Treasury bills in 1929, the increase in Government debt during the
thirties, the long-run effects of a movement of reserves out of New
York, and later the increased demand for business and consumer
loans, all contributed to a relative decline in the supply of funds
available for broker and dealer loans. Banks turned to short-term
Government securities—principally bills—as the prime source of
secondary reserves. Moreover, with the advent of tight money in
recent years, the alternative cost—and hence the rate at which banks
offered broker and dealer loans—has increased. This has resulted
in a further reduction in the supply of funds.
For the Government securities market, financing net position alone
requires that on the average $1 billion must be obtained from all
sources. Of this amount, roughly $250 to $300 million is the average
net position of dealer banks and is financed almost exclusively through
internal allocation of funds. Since we have little information about
the decision process within banks, we are forced to ignore the bank
dealers throughout most of this chapter.
Reputation as a trader is the principal basis for borrowing. Margins
are low and leverage ratios are extremely high, higher than the deposit/capital ratios of commercial banks and higher than the leverage
ratios of other securities dealers. Since the turnover of dealer portfolios is also greater than turnover in other securities markets, an additional constraint is imposed. Borrowing must be confined principally
to call loans and other very short term arrangements which can be
cancelled or allowed to expire when a dealer's position declines.
This chapter is concerned with the sources and allocations of dealer
financing. I t describes part of the interaction between the money and
securities market. But before appraising the effectiveness of financial
arrangements and the interaction of these two markets, we must first
consider some definitional problems and describe some institutional
features.
62471—60

6




71

72

STUDY OF FEDERAL GOVERNMENT SECURITIES DEALERS
SOURCES AND USES OF F U N D S

A dealer's allocation of funds includes all of his commitments. As
we indicated in chapter I I I some difficulties arise when the dealer
accounting statements are aggregated to compute total position. Similar problems arise when a sources-and-uses-of-funds statement is
prepared for the 12 nonbank dealers. However, the total sources
or total uses of funds exceed the total net position. Several of the
difficulties which were described earlier are eliminated in the data
of this chapter as a result of aggregation. Hence, it is likely that the
error in these observations is proportionally much smaller than the
estimated error in total net dealer positions.
Uses of funds
Four principal uses of funds appear on the balance sheets of most
dealer firms and have been included in the items which dealers finance.
Presented here in the order of their relative size they are: Total Government securities owned; securities sold but not yet delivered; cash
deposits for securities borrowed; and agency, municipal, and other
bonds. We consider each in turn.
1. Total Government securities owned is generally the largest single
item on a dealer's balance sheet. As indicated earlier, the amount
included here is not a gross figure since some canceling of gross long
and gross short positions takes place before this figure is obtained.
F o r present purposes, total investment accounts are included in this
sum.
2. The account "Securities sold but not yet delivered" is similar to
an account receivable in that payment is due from a customer. However, the securities remain in the dealer's possession. The amount included in this account varies greatly between dealers.
Dealers whose business is most heavily concentrated in the bill market sell a large proportion of their securities for cash and make deliveries on the day of the sale. Some bills and most other securities are
sold on a "regular" basis which requires delivery on the next trading
day. Overnight the value of these securities would be included here
if the dealer uses the commitment method of reporting. 1 Some sales do
not require delivery for several days. F o r example, when "corporations and others have large blocks of funds becoming available from
capital market financing, they may begin to buy Government securities several days before the new money becomes available in order to
spread out the market impact of the transactions. Another example
is the use of delayed delivery sales of new Treasury issues to institutional investors. I n the February 1955 refunding, some of these dealer
contracts to sell 3-percent bonds of 1995 ran as long as 4 months." 2
3. "Cash deposits for securities borrowed" appears in the asset
accounts when the dealer borrows securities to sell short or to complete a transaction. For most years, the aggregate amount of this
item is relatively small, but it exceeded $30 million in several years.
Larger dealers who operate more extensively in the long-term market
make relatively more use of this transaction than the smaller dealers
because it is more difficult to borrow long-term securities.
* If he uses the execution method, they are reported in his gross position. When total
commitments
are summed, part of the bias which we discussed in ch. I l l disappears.
a
"Treasury Federal Reserve Study * * *," op. cit., I. 23.



STUDY OF FEDERAL GOVERNMENT SECURITIES DEALERS

73

The transaction which gives rise to this asset account is relatively
expensive from the dealer's standpoint and occurs only when securities
cannot be borrowed against a pledge of other securities. I n pledging
cash, the dealer sacrifices the yield on those securities which might
have been purchased with the cash. H e must pay the interest accruing
on the securities sold short and a fee of one-half of 1 percent for the
privilege of borrowing securities. These costs undoubtedly account
for the infrequent appearance of this account on most of the balance
sheets wilich were examined.
At first glance, it might appear that this item is a part of the cash
account and should not be included among a dealer's commitments.
But, as noted, the cash which is included in this account is obtained
when securities are borrowed and sold short. More commonly, collateral (other securities) is pledged against securities borrowed.
Such collateral is considered part of the dealer's position. The alternative to holding cash in this way is not that of having the cash available in the cash account. The cash is obtained in the transaction
which gives rise to the dealer's short position. If he had pledged
securities in place of cash, the cash obtained would be used to purchase either the securities pledged as collateral or some other issue in
his portfolio of equivalent dollar value. 3
4. Municipal and other bonds include agency and other Treasury
obligations as well as a few corporate and other issues. The detail
of this account is described in the chapter on dealer positions and
need not be repeated here. However, this item is included among the
dealer commitments so that meaningful comparisons may be made
between total sources and uses of funds. Exclusion of non-Government obligations could not be made without some estimate of differences in margin requirements on loans and differences in sources of
funds for the various types of securities. However, bankers acceptances have been excluded since the amount of loans used to finance
them was indicated on the balance sheets.
Sources of funds
Four sources of funds can be identified: securities purchased but
not yet received, securities sold but not yet purchased, collateral loans,
and repurchase agreements. Again, we will consider each of these
in turn.
1. Securities purchased but not yet received are analogous to an
account payable in a commercial enterprise. The securities are included in the dealer's position if they have not yet been resold, but
the dealer has not yet paid the seller. Variations in the amounts included here reflect changes in the volume of transactions, the proportion of bill to total transactions, and the proportion of cash to regullar transactions. Differences in dealer accounting, referred to previously, result in an understatement of the aggregate amount included here, but for the 11 nonbank dealers on a commitment basis
the total exceeds $300 million on the end of the year balance sheets
for several years.
8
The total of this account has been reduced by "cash liability for securities loaned"
which appears when dealers lend securities, principally to other dealers, against cash
pledged as collateral. In the data used here, the amount subtracted is generally quite
small and the net amount is always positive.




74

STUDY OF FEDERAL GOVERNMENT SECURITIES DEALERS

2. "Securities sold but not yet purchased" is the net amount
of securities sold short as we indicated in the discussion of net position earlier. The aggregate amount of dealer short sales included on
end of the year balance sheets fluctuated between a low of approximately $50 million in 1949 and a 1958 high of $364 million for the
12 nonbank dealers. The following observations give some indication
of the variation of the short positions over recent years:
12 nonbank dealers
[In millions]
End of year

1954
1955
1956
1957
1958

-_

.

.
_

Reported
short
positions
$32C
21?.
203
85
364

Cash

$25
23
21
22
30

Fluctuations in the short position of one large dealer dominate these
totals. His share of the aggregate short position of all dealers has
varied from 33 to 50 percent of the amounts shown above and corresponds to his attitude toward risk and his view of the dealer function.
Short sales finance dealer positions by providing cash which may be
used to purchase additional securities. Alternatively, the dealer may
sell short and hold cash in anticipation of a fall in securities prices.
The data above suggest that there is little or no relation between
changes in short positions and changes in cash accounts. Hence, no
problem is created by the inclusion of all short positions as sources
of funds to finance dealer commitments. And the data again suggest
that most dealer short sales may be viewed as arbitrage transactions.
3. Collateral loans were formerly the principal source of funds for
financing dealer positions. I n the earlier years covered by this study,
most of these loans were made by clearinghouse banks in New York,
but more recently many banks in other parts of the country have participated in financing the Government securities market. Several foreign banks which have agencies in New York and large nonfmancial
corporations also extend loans to the nonbank dealers.
Collateral loans may be made in either Federal funds or clearinghouse funds. Since almost all short-term (under 1 year) transactions
in Government securities now are settled in Federal funds, dealers
try to arrange financing in this form. Small dealers will often employ
the services of a money or Federal funds broker—Garvin, Bantel &
Co.—to assist them in arranging loans with banks outside New York.
F o r a reported fee of one-eighth percent the broker arranges the transfer of deposit balances from the reserve account of the lending bank
to the reserve account of the bank which clears transactions in the
Government securities market—most often the Manufacturers Trust
Co.
Loans made in Federal funds have the advantage of being available
to the dealer on the same day. Most of these loans are 1 day or overnight loans and are repaid in Federal funds on the following day.
Loans of clearinghouse funds do not become available until the following day when clearinghouse balances are settled. They cannot be
used to pay for cash purchases unless they are exchanged for Federal



STUDY OF FEDERAL GOVERNMENT SECURITIES DEALERS

75

funds in the market. As a result, dealers try to arrange their financing
in Federal funds early in the day (usually before 1 p.m.) and are more
likely to resort to clearinghouse loans later in the day when shortages
must be covered.
New York City banks make dealer loans available, with only a few exceptions,
in clearinghouse funds. The rate charged by New York City banks on dealer
loans is frequently higher than the cost of financing in other cities; it is also
often above the yield on the dealer's portfolio, especially in periods of tight money.
Thus, dealers are often under pressure to borrow out of town or from nonbank
sources, both to obtain lower rates and to acquire Federal funds to pay for
securities against which they are borrowing. If they borrow from the New York
City banks, dealers are required to pay the current rate of interest on the loan
plus the cost of the Federal funds for 1 day or for 3 days if they borrow on
Friday. When money is tight in other parts of the country, dealers rely more
heavily on the New York City banks, thus exerting direct pressure on the New
York Federal funds market. In such instances, Federal funds are supplied by
the clearing bank and by other New York City banks, even though borrowing
from the Federal Reserve bank is necessary.4

When transactions are cleared through the Manufacturers Trust,
credit and debits to a dealer's account are recorded by the bank. The
bank makes day loans to cover shortages of Federal funds but makes
no charge other than the nominal clearing charge for this service.
Over time, it expects shortages and overages to balance out for each
dealer.
The dealer maintains his own record of "money position" and attempts to balance his inflows and outflows of funds during the day
either by finding sources of financing or by selling for cash. If, after
searching the market, his money position is "net short" near the close
of the trading day, he can often get some overnight accommodation
in Federal funds from the clearing bank. I n contrast to day loans,
overnights are made at rates equal to those charged by the other New
York banks. However, there is no commitment on the p a r t of the
clearing bank to act as lender of last resort.
The amount of overnight loans made at the clearing bank apparently vary substantially over time and between dealers. Some small
firms indicated that they preferred not to rely on this source of
financing because of the difference in rates between New York and
other borrowing centers. Some suggested that it was usually less
costly to arrange a loan through Garvin, Bantel & Co. A small number of dealers supplied detailed distributions of their financing by individual banks. These data indicate that for certain dealers the
clearing bank was the largest single source of funds and had supplied
as much as 15 to 20 percent of these dealers' total financing at the end
of particular years.
Some of the larger dealers do not clear all of their transactions
through the Manufacturers Trust. Instead, some use other banks
as clearing agent or arrange for the transfer of some longer term
securities through other means. One large firm clears all of its own
transactions. To clear transactions, "day loans" or credit lines, available in the morning from New York banks, are used to repay expiring
overnight loans. During the course of the day, net purchases of
securities for cash are paid for either by drawing against the day loan
or by Federal funds purchases. By the end of the day, new financing
* "The Federal Funds Market," op. cit., pp. 47-48.




76

STUDY OF FEDERAL GOVERNMENT SECURITIES DEALERS

must be obtained to cover a net short money position and to reimburse the New York banks for advances made during the day.
Some of the larger dealers do not have the difficulties in obtaining
financing from the New York banks that many of the smaller dealers
report. One large dealer has an arrangement with two banks which
take turns supplying financing in alternate weeks. However, the rates
charged do not appear to differ from those available to other dealers.
As a result, this arrangement supplies a larger proportion of the particular dealer's funds when rates are low and New York banks are
more willing to lend to other dealers. 5
4. Eepurehase agreements have become an increasingly important
source of financing in recent years. F o r the dealer, the E P has several advantages over the collateral loan: First, it always provides
Federal funds with which the dealer may pay for his purchases of
short-term securities. Second, it can often be obtained at a rate lower
than the rate at which banks are willing to extend collateral loans.
Third, it may have a maturity of several days or several weeks and
may be used to reduce the dealer's daily borrowing requirements
during a period of relatively "tight money." I n recent years, some
of the bank dealers have also recognized some of these advantages
and have used E P ' s to finance a part of their dealer position.
The four principal sources of E P ' s for the dealer market are nonfinancial corporations, commercial banks outside of New York, the
Federal Reserve Bank of New York, and the Federal home loan banks.
Other institutions which supply funds to the dealer market by these
means include state and local governments, foreign agency banks,
New York Clearing House banks, mutual savings banks, and savings
and loan associations. The latter group of institutions was considerably less important in the aggregate; their role will not be described
in detail.
Nonfinancial corporations acquire temporary deposit balances at
commercial banks. Legally, banks may not pay interest on such deposits. E P ' s offer the corporation an opportunity to invest their
funds in maturities tailored to meet their requirements. The funds
are invested under a contract which specifies both the price at which
the corporation buys securities (or lends funds) and the price at
which it resells the securities (or is repaid). The interest rate is
fixed in advance and the transaction is collateralized by Government
securities which are usually deposited at a bank selected by the lender.
Thus, the corporation converts non-interest-yielding assets into interest-yielding assets with a minimal increase in risk.
Commercial banks also enter into repurchase agreements with
dealers. They may use an E P instead of a collateral loan for one
of several reasons. First, the yield on an E P often exceeds the yield
on a Federal funds transaction. Second, small banks may wish to
sell Federal funds but may not have the minimum unit in which the
market trades, $1 million. Third, they may prefer the greater security of principal in a secured E P transaction to the outright sale
of Federal funds even if the E P rate is temporarily below the Federal
funds rate. Or, they may prefer to invest surplus funds at an agreed
5
Statements by each of the dealer firms on the financing of their commitments is available in part 6C of Hearings, 1959 * * * op. cit. See especially pp. 1848 ff.




STUDY OF FEDERAL GOVERNMENT SECURITIES DEALERS

77

upon rate and allow them to remain in use rather than risk daily
changes in the Federal funds rate.
A third major source of E P ' s is the Federal Reserve Bank of New
York. The bank may lend on securities with up to 15 months to maturity for a maximum period of 15 days. F o r the bank, the E P is a
relatively simple device for correcting temporary reserve shortages or
imbalances. The resulting increase in bank reserves is reversed when
the E P expires. Dealers usually pay the rediscount rate for the privilege of borrowing but the bill rate may be charged at the discretion
of the bank.
F o r the nonbank dealers the Federal Eeserve is a next to last source
of funds during tight money periods. After they have searched
elsewhere, but often before they try to borrow at the relatively high
rates charged by the clearinghouse banks, the dealers will inquire at
the Federal Eeserve about the possibility of making repurchase agreements. The Federal Eeserve has available the data on reserve positions of principal banks and seems to base its decision principally
on the distribution of available free reserves and the objectives of
its policy. E P ' s are probably easier to arrange with the Federal
Eeserve when money is being loosened than when it is being
tightened. However, temporary accommodation is also given during
a tight money period, particularly on Fridays, holiday weekends, and
at the end of December when dealer positions are increased in response
to yearend balance sheet adjustments and corporate dividend payments.
Although the Federal Eeserve Bank of New York may lend for
periods up to 15 days, many E P ' s are made for shorter periods and
all are subject to a call provision. Only nonbank dealers have the
privilege of borrowing in this way. However, bank dealers may
rediscount for Federal funds at Eeserve banks using eligible paper
or bills as collateral. 6
The fourth source of E P ' s is more heterogeneous and includes several different types of lenders. Federal home loan banks and savings
banks became more important sources of funds for the dealer market
during the tight-money period. Because their operations differ from
those of commercial banks, Federal home loan banks are often able
to make commitments for longer periods. They thereby provide an
alternative source of financing wThich assists the dealers to hold individual issues for longer periods. These longer term (for example,
30 day) funds are also helpful in financing dealer positions during
certain times of the year. Particularly during the months in which
quarterly corporation tax payments are due, the supply of funds
available for repurchase agreements from nonfinancial corporations
falls while the supply of Government securities offered on the
dealer market rises. Longer term repurchase agreements seem to have
a more important role in financing the resultant temporary large
increases in dealer positions.
For the economy as a whole, the E P is a particularly interesting
method of financing. Through its use, the maturity of any particular
issue of Government securities is modified to provide whatever maturity is required by the lender. This can also be accomplished
6
F o r a more complete discussion of Federal Reserve practices, see Roosa, op. cit., pp.
83—87.




78

STUDY OF FEDERAL GOVERNMENT SECURITIES DEALERS

through purchases and sales of relatively low risk instruments, for
example Treasury bills. But the risk element in a repurchase agreement based on bills is even smaller than the risk in owning a short
term bill which is subject to price fluctuations between purchase and
sales dates.
The combination of low risk, tailored maturity, and relatively high
rate of return in recent years has undoubtedly resulted in a mobilization of idle cash balances through the dealer market and has increased
the velocity of money. Some of these balances would have been transferred to the money market through the bill market and the Federal
funds market, but the K P undoubtedly attracts additional funds from
those who wish to reduce risk.
When the dealer market has accumulated surplus funds, for example, during the latter part of 1958, reverse repurchase agreements
have been made. Reverse HP's are a means by which some dealers
lend Federal funds to banks which require additional reserves. Some
dealers continue to sell all of their excess Federal funds in the market.
Others have taken an active and increasing role in supplying reserves
to banks under reserve HP's.
COMPARISON OF DEALER COMMITMENTS AND DEALER NET POSITIONS

Before presenting the data showing the sources and uses of funds
in the Government securities market, we wish to compare the size of ;
commitments which dealers must finance with the data on net dealer
positions from chapter III. Table V - l presents the ratio of total
commitments/total net position for the 12 nonbank dealers.
I t is clear from the table that commitments substantially exceed
net positions in all years. Moreover, the data suggest that in years
when net positions are relatively large, the ratio is relatively small
and vice versa. This implies that the total dealer commitments
fluctuate proportionally less than total net positions. I n part, this
results from the inclusion of gross positions in the commitments
data and in part from the inclusion of securities which have been
sold but not delivered.
TABLE V-l.—Ratio of dealer commitments
to dealer net positions
dealers at end of year
Ratio in
percent

Year

1948
1949-__
1950
1951
1952
1953

_

__

_--

133.3
147.2
178.1
176.5
184.3
126.9

Year

1954
1955
1956
1957
1958

for 12

nonbank

Ratio in
percent
238.0
193.6
222.3
153.9
172.7

"Fails" (securities not delivered on the due date) are relatively
more important in the intermediate- and long-term markets. Hence
dealers who sell short and those who operate more extensively in
the bond market should have higher ratios of commitments to positions. This is confirmed when the ratios are computed for each of
the 12 dealers.




STUDY OF FEDERAL GOVERNMENT SECURITIES DEALERS

79

Dealers who operated principally in bills and short-term maturities
had ratios of commitments to net position as low as 105 or 108 percent. For them, total commitments and net position are approximately equal. Dealers operating more heavily in the intermediateand long-term markets showed total commitments far in excess of
their net positions. For example, one dealer who carries a relatively
large position had commitments which on the average were more
than three times as large as his net position. The problem of financing commitments clearly differs in magnitude for the two types of dealers described in these examples. And this suggests that constraints
which arise on the financial side may be an additional source of limitation on dealer operations in the over-5-year maturity range.
M A G N I T U D E A N D D I S T R I B U T I O N OF SOURCES A N D USES OF DEALERS' F U N D S

The Treasury-Federal Reserve accord of 1951 and the return of a
more flexible monetary policy increased total dealer commitments.
This is shown in table V-2 (line 3). Since the end of 1952, the 12
nonbank dealers found it necessary to finance commitments of $1.25
to $2 billion on the dates for which data were available. If the
ratio of total net position of bank dealers to total net position of
nonbank dealers may be used as a guide, the total financing required
by this market from all sources must have varied from a low in the
neighborhood of $1.6 billion to a high of $3 billion at the end of
recent years. 7
As the table shows (lines 4 and 5), most of the financing comes
from the sources described earlier in this chapter. These sources in
the aggregate have rarely supplied less than 95 percent of the total
amount required. Year-to-year variations are remarkably small and
may in part be accounted for by differences in the accounting practices
of individual dealers. Even when allowance is made for margin
requirements on collateral loans, the amount of financing which dealers must supply from their own capital remains small.
The ratios of sources to uses for individual dealers show very little
dispersion. Only one dealer had an average ratio below 95 percent
for the entire period; two had ratios of 96 percent; two were at 97
percent; two were at 98 percent; five had ratios of 99 percent. Variations between individual dealers seem to reflect two things: the extent
to which the dealer operates in markets other than the market for
Government securities and the relative importance of bills in the
dealer's portfolio.
7
Treasury-Federal Reserve study, vol. I, pp. 4 2 - 4 3 , shows $2.7 and $1.8 for Dec. 31,
1957 and 1958, respectively.




00

o
CO

d
o
o
T A B L E V-2.—Sources and uses of dealer financing
1947
(1) Total collateral (loans+RP's).
._N Loans+RP's .
(2)
-T7-7
rr.— m percent
x
Net position
^
(3) Total commitments (uses)
(4) Total financing (sources)
(5) Sources/uses in per cent
,„. Loans+RP's
Total uses .
<6> Loans+RP's
n ^ f i „ c n , m percent
in percent(7)
Total sources
Number of firms.

261.1

1948

1949

600.8
562.7
93.6

718.5
111.4
860.1
825.2
95.9

43.4

83.5

767.1
124.0
911.0
873.5
95.9
84.2

46.4

87.1
11

87.8
11

NOTE.—Items (1), (3), and (4) are in millions of dollars; others are in percent.




1950

{12 nonbank dealers), 19J+7-58

1951

551.2
126.0
779.0
739.1
94.9
70.8

702.3
129.6
956.5
921.5
96.3

74.6
12

76.2
12

1952

1953

1,237.1
133.7
1, 706.0
1, 670. 9
97.9
74.0
12

1954

1955

1956

1957

1, 215.4
90.2
1, 709.3
1, 670. 6
97.7
71.1

158.9
1,475.4
1, 429.1
96.8
66.7

845.6
117.8
1, 389. 2
1, 340. 9
96.5
60.9

696.7
121.6
1, 274.4
1,227.8
96.3
54.7

1,445.1
113.8
1, 953. 9
1, 898.5
97.2
74.0

847.8
96.6
1, 516.9
1,452.4
95.7
55.9

72.8
12

68.9
12

63.1
12

56.7
12

76.1
12

58.4
12

1958

>
a
o

3
3
GO

o
d
GO

«
>

t- 1
feJ
GO

STUDY OF FEDERAL GOVERNMENT SECURITIES DEALERS

81

Loans and RP^s
The total amount of loans and repurchase agreements which dealers
reported on their year-end financial statements are shown in line (1)
of the table. These totals do not agree with the available totals in
the Federal Eeserve bulletins principally because some firms could
not supply calendar year information. Moreover, differences in accounting practices referred to earlier and the failure of several dealers
to separate collateral loans from repurchase agreements on their balance sheets prevents any meaningful separation of the aggregate
figure. As a result, discussion of changes in the composition of line
(1) will be deferred to the next section where individual dealer reports are presented.
When collateral loans and HP's are grouped, they appear to be the
largest source of dealer funds (line 7) and to finance the larger part
of dealer commitments (line 6). Changes in the ratio of loans and
HP's to dealer commitments have generally been in the same direction
as the change in total dealer commitments. However, the downswings
in the ratio have been proportionally larger than the upswings. As
a result, the ratio has fallen in recent years while total commitments
have risen relative to the early years of the period. Other sources—
"shorts" and purchases not received—have increased relatively.
The ratio of loans plus HP's to total net position (line 2) indicates
that in all but 2 years the total amount borrowed by dealers exceeded
their net inventories. While this ratio has fluctuated substantially
over the 12-year period, there is no indication of the downward trend
observable in the data of line (7). One explanation for this difference
wras suggested above when net positions and total commitments were
compared. Total commitments or total sources of funds were found
to fluctuate less than total net positions. Another was suggested in an
earlier chapter when total holdings of Government securities were
compared to non-Governments in dealer positions.




82

STUDY OF FEDERAL GOVERNMENT SECURITIES DEALERS
DISTRIBUTION

OF SOURCES FOR I N D I V I D U A L

DEALERS

Table V-3 was constructed from records submitted by 4 of the 12
nonbank dealers. F o r two dealers, the data presented are based on
end of the calendar year reports; for the other two, dates in the second
quarter of the year were chosen. I t is likely that the calendar year
reports overstate the importance of repurchase agreements with the
New York Federal Reserve Bank and understate the proportion of
financing obtained through repurchase agreements with commercial
banks and nonfinancial corporations. The noncalendar year data
probably err in the opposite direction; the proportion of R P s with
the Federal Reserve may be understated and the proportion of R P s
with commercial banks and nonfinancial corporations may be overstated in comparison with the financing practices of the dealer community as a whole.
Nevertheless, the table is suggestive of trends which have taken
place during recent years. These trends reflect dealer responses to
relative changes in interest rates as well as changes in the available
supply of funds offered by lenders. At the interest rates which have
prevailed during recent years, the total amount of collateral loans
available to the 12 nonbank dealers from all sources has exceeded $1
billion only rarely. Most often, the total appears to be well below
that figure. Repurchase credit, excluding the amounts advanced by
the Federal Reserve Bank, is probably about $1.5 billion at its
maximum.
The two firms for which year-end data are presented continue to
finance the major part of their commitments with collateral loans if
R P s at the Fed are excluded. New York banks remain the largest
single source of private funds for both dealers, but in years when
money is tight, the relative contribution of New York banks seems to
decline. When money rates in New York are high, dealer W seems
to rely on banks outside of New York and to a lesser extent on nonfinancial corporations to supply financing. Dealer X depends more
heavily on foreign agency banks to lend funds.




STUDY OF FEDERAL GOVERNMENT SECURITIES DEALERS
TABLE V-3.—Percentage distribution

of sources for Jt nonbank

83

dealers

DEALER W
Collateral loans

December:
1948
1949
1950
1951
1952
1953
1954
1955_
1956
1957
1958

_

__

-

Total
collateral
loans

(1)

100
100
100
55
33
82
68
39
35
53
74

Repurchase agreements

New
York
banks

Other
commercial
banks

Other

Total
RP's

(2)

(3)

(4)

(5)

96
77
86
41
28
60
44
25
20
16
65

4
23
14
8
5
20
24
14
15
26
9

6
2

11

45
67
18
32
61
66
48
26

Other
commercial F e d e r a l
b a n k s Reserve
and
Bank
nonfi- of N e w
nancial Y o r k
corporations
(6)

(7)

Other

(8)

45
67
18
32
21
6
16
10

40
60
32
16

DEALER X
December:
1948
1949
1950
1951..
1952
1953 _ _
1954
1955
1956..
1957
1958.

100
90
72
63
29
50
100
47
81
59
100

_

_

94
83
66
45
22
34
81
32
62
50
100

6
2
6
2
1
5
6
8
6

5
16
7
15
14
9
14
3

10
28
37
71
50

10
3
9

53
16
41

18
20

44
78
99
84
100
90
92

44
78
96
74
88
86
89

28
6

28
6

29

29

90
87
87
64
85

65
85
51
46

5

25
28
71
41

4

35
16
21

DEALER Y
Non-December:
1948
1949_
1950
___
1951
1952..
1953
1954
1955
1956_.
1957
1958
_

100
100
100
100
56
22
1
16
0
10
8

_

100
100
100
100
56
22
1
16
10
8

9
5
4

4
1
7
3

DEALER Z
Non-December:
1948
_,
1949
1950
_
1951
1952
_.
1953
1954
1955
a
1956
1957
1958

-.
_-

_

—
_

72
94
100
100
71
100
10
13
13
36
15

72
94
100
100
71
100
10
13
13
22
15

NOTE.—Detail may not add to 100% due to rounding.




14

90
22
2
13
39

84

STUDY OF FEDERAL GOVERNMENT SECURITIES DEALERS

Dealers Y and Z have used repurchase agreements as their principal
source of funds since the middle fifties. I n comparing the distribution
of their financing with the similar distribution for dealers W and X ,
it should be recalled that dealers are required to absorb a large flow of
securities at year-end dates. Other things equal, changes in the supply
of repurchase agreements should have a seasonal pattern opposite to
that of the demand for money. This may in part account for the
relatively large proportion of repurchase agreements shown for dealers
YanclZ.
Both dealers Y and Z have financed principally through commercial
banks outside New York and nonfinancial corporations during recent
years. However, dealer Z has borrowed from (or entered into KPs
with) "other" sources to a much greater extent than dealer Y. Like
the differences in the relative importance of sources for dealers W and
X , these arrangements suggest that all dealers do not compete for
funds from all lenders. Established contacts with particular lenders
and the size of individual dealers seem to influence the number and
type of source from which the dealer obtains financing.8
Table V-3 suggests that important procedural changes have taken
place in the money market in recent years. If dealers rely principally
on the New York Clearing House banks to supply funds, the responsibility for finding idle excess reserves or corporate balances is left to
the clearinghouse banks. However, when the nonbank dealers undertake to secure a large portion of their financing from banks outside of
New York, corporations, and others, the job of bringing excess reserves and idle balances into use, devolves upon the dealers. By offering rates above the rate for Federal funds sales, dealers induce a flow
of funds into the New York money market. I n the process, changes in
monetary policy are spread throughout the country.
Table C-3 of the Treasury-Federal Reserve study provides some
evidence of the changing composition of the ratio of loans to repurchase agreements during a period of relatively wide swings in the
interest rate. I t suggests the extent to which mobilizing reserves
has become part of the dealer function. Their data include estimates of the use of bank funds by dealer banks in New York and
Chicago. If these funds are excluded from the totals shown there,
the amount of collateral loans available to the nonbank dealers must
have been extremely small or nonexistent during most of October and
November 1957 and 1958. Moreover, it would appear that the nonbank dealers were using reverse repurchase agreements to supply Federal funds to banks in New York and Chicago.
Dealer commitments in the spring of 1958 were probably as large
as they had been in any preceding period. If this is true, it would
indicate that the maximum amount of loans and E P s available to
the market under present financing arrangements and the then prevailing interest rate structure may be in the neighborhood of $3 to $3.5
billion. This amount was divided almost equally between the two
financing methods. Assuming again that the bank dealers relied
principally on collateral loans from their own banks, the data suggest
that as much as 80 percent of the total nonbank dealer borrowing
must have been completed through the use of E P s . This estimate
8
Additional evidence on the distribution of sources of funds can be found in 1959 hearings * * *, op. cit., 6-C, p. 1858.




STUDY OF FEDERAL GOVERNMENT SECURITIES DEALERS

85

compares reasonably well with the data for 1958 in table V-3
(non-December).
KATES ON DEALER LOANS

The crux of the dealer financing problem is the rate which the
dealer pays. When money is easy, a larger proportion of his commitments can be carried with collateral loans from New^ York banks.
Otherwise, alternative sources must be found or total commitments
must be reduced. I n searching for funds, dealers must compete with
banks and others, but several factors combine to give the nonbank
dealers a comparative advantage in obtaining temporarily idle
balances.
First, since the alternatives for dealers and banks differ, dealers can
pay higher rates. F o r banks, the purchase price for Federal funds
will exceed the rediscount rate only if they have been denied the privilege of additional rediscounts, have profitable loan or investment opportunities available to them, and have no securities with yields below the Federal funds rate available for sale. What little evidence
has been collected on this point suggests that this combination has not
occurred in recent years. 9 For dealers, the relevant comparison is between the yield on the additional securities which the dealers will
position and the cost of carrying those securities. F o r example, if bill
yields exceed the Federal funds rate, a dealer will be willing to pay
for repurchase agreements a rate higher than the rate on Federal
funds. The difference between the yield on bills and the rate paid will
contribute to net income. If he anticipates that the price of securities
will rise, that spreads between buying and selling prices will widen,
or if he wishes to build good will with certain customers, the dealer
may position securities when the borrowing cost exceeds the yield.
Thus, for banks, the relevant comparison is most often the rediscount
rate; for dealers the alternative to paying the higher rate is the loss
of profit on the sale of additional securities taken into position.
Second, by paying a higher rate, the dealer will attract funds which
might otherwise not be available. The quantity supplied increases as
the interest rate rises. By offering higher rates, dealers induce commercial banks outside New York to reduce precaution ar}T balances or
to borrow from the Federal Reserve banks in their districts. These
loans are made at the rediscount rate (usually equal to the Federal
funds rate) and lent in New York at a higher rate. New York banks
could attract funds for dealer loans in this way, but would have no competitive advantage. As long as the rate which dealers pay outside of
New York is below the rate in New York by the additional cost of locating and transferring funds, the dealers can reduce their borrowing
cost by making collateral loans or E P ' s in other parts of the country.
This analysis may connote an overly defensive attitude. Some
dealers view the E P as an alternative to a sale of securities rather than
as a borrowing arrangement. For them, differences between the yield
on securities and the cost of funds are an indication that they should
aggressively exploit the difference by purchasing securities and making repurchase agreements. This is particularly true for longer term
repurchase agreements and suggests the reason for the distinction be9

Cf. "The Federal Funds Market," op. cit, pp. 98-104.




86

STUDY OF FEDERAL GOVERNMENT SECURITIES DEALERS

tween E P ' s of 16 or more days and those of 15 or fewer days to
maturity.
Third, the nonbank dealers are a more permanent source of demand
for short term funds than banks. Even at the end of 1956 and the
beginning of 1957, there were occasions on which the Federal funds
rate fell below the rediscount rate. This suggests that temporarily the
supply of Federal funds in New York exceeded the demand, and probably reflects erratic shifts in deposit balances. The demand for Federal funds by dealers fluctuates with the size of their commitments and
the maturity composition of their transactions. Day to day fluctuations in dealer demand are probably smaller and somew^hat more
predictable. Hence it is probable that many banks and nonfinancial
corporations would prefer a more or less permanent arrangement with
a particular dealer even if the average rate over time is the same as
the average Federal funds rate.
Fourth, the minimum transactions in Federal funds is usually $1
million. Dealers are often willing to borrow smaller amounts. I n
some cases, dealers have financed commitments of $100 million or more
with a few large loans—that is, $50, or $25 million—and many loans
or E P ' s of $100,000 or less. I t is unlikely that transactions of that size
could be made in the Federal funds market on a regular basis.
From the preceding analysis, it would appear that differences between the bill rate and the Federal funds rate should be eliminated.
Since New York banks operate in both markets, we would expect that
banks would exploit differences in rates and, in particular, differences
between the rate on bills maturing within the week, the Federal funds
rate, and the rediscount rate. The available data suggest that this
happens less frequently than might be supposed. Legal restrictions
on the size of loans which a bank may make to one individual or corporation and differences in relative liquidity between bills and Federal
funds have been suggested as explanations of those differences which
persist. More detailed discussion of this point and some additional
suggestions may be found in the Federal Eeserve System monograph
referred to earlier.
Furthermore, our analysis suggests that when money is tight the rate
on collateral loans and repurchase agreements outside New York
should approach or exceed the yield on securities which dealers carry
and should exceed the discount rate at the New York Federal Eeserve.
Evidence for the former is presented in a later chapter when the net
interest earnings of dealers are discussed. There we find that in several recent years, interest expense was greater than interest income. 10
10
Some of the difference probably arises from the fact that the yield on securities for
a nondealer is the yield discounted to maturity. For the dealer, the yield to maturity
is not an approppriate yardstick since the security will remain in position for a very short
time. Dealers compute the yield on a security as the ratio of the coupon to the purchase
price—that is, simple interest. Thus there is a difference between the dealers' rate of
return and the alternative cost as seen by the supplier of funds.




STUDY OF FEDERAL GOVERNMENT SECURITIES DEALERS

87

CHANGES I N THE RATE STRUCTURE OVER TIME

The relationship between the various rates paid by dealers and the
rediscount rate at the Federal Keserve Bank of New York is shown in
table V-4. These data are based on reports submitted by dealers and
are suggestive of the relationships w^hich prevailed at various points
in time. But it should be noted that all dealers did not submit rate
information and others could supply this information for recent years
only.
At times, there is a great deal of spread in the rates on collateral
loans prevailing in a particular sector. These spreads appear to be
largest when rates are adjusting to new levels. F o r example, there is
a difference of 1% percent between the lowest and highest rates at
New York Clearinghouse banks in December 1954 and December
1957; at other times, the spread is only one-half or three-fourths percent. Such differences probably reflect opposing variations in the reserve positions of particular banks and in the time of day at which
the loans are made. Most often, the range of rates reflects the charges
made to individual dealers by different banks rather than a difference
in the rates paid by small and large dealers.
The rate on collateral loans charged by the New York banks rarely
falls below the discount rate at the Federal Reserve. This is particularly true for the December data. For non-December dates the lending rate in New York is below the discount rate on several occasions,
but in 1956 and 1957 even the minimum rates reported are above the
rediscount rate. Since many of these loans are in clearinghouse funds,
the cost of financing commitments in Federal funds is higher than the
rate reported by the cost of converting to Federal funds.
The loan rates charged by commercial banks outside of New York
are frequently below the rates charged in New York and lower than
the rediscount rate. However, as the discussion above suggests, when
money is tight or dealers are carrying relatively large positions, rates
outside New York rise and equal or exceed New York rates. Examples
of this are found in the reports for December 1953, 1957, and 1958
as well as for non-December dates in 1956 and 1957. On the latter
date, the ranges reported were the same for commercial banks in and
out of New York and were three-fourths percent higher than the rediscount rate in New York.

02471—60

7




88

STUDY

OF FEDERAL

GOVERNMENT

TABLE V-4.—Reported

rates

on loans

Collateral loans
New
York
banks
December:
1948
1949
1950
1951
1952
1953
1954
1955
1956
1957
1958
Non-December:
1948
1949
1950
1951
__
1952
1953
1954
__
1955
1956
1957
1958

1
i

-m
-m

m-m
1/2-2
1H-2H
1^-3
3 -3H
3 -3%
3 -m
2M-3
1 -1H

I -m

i
-m
mr2H
1H-2H
2 -2%
i
-m
1/2-2%
3 -3/2
3%-4
lH-2^

Other
commercial b a n k s

1 -1H
1 -m
m-2Vz
m-2H
m-s
m-zu
m-2%
2 -m
2/-4H
3H-4H
2^-4
1 -W
1 -1/2
1^-2/
iH-2^
1/2-3
2 -3H
1H-2/2
1H-2H
3 -3^
3%-4
%-3/2

SECURITIES

DEALERS

and

percent1

RP's

in

R e p u r c h a s e agreements

Other

NonfinanOther
commer- cial corporations
cial b a n k s

Federal
Reserve

Other

Discount
rate
NewYork
Federal

1
i

-m

m

m-2
m-2
m-2

m-m

2/-3
3 -3H
2 -2Yi
1

m

m-2
2

m

1/2-1%
3
3H
H-1H

0.80-0.90

m
m
m
m-m m -m
m-m
2M

2 - 2 % 2 / -3H
3
2%
3 -m
.80
2H-2&A

m
m
m
i%
1U-2H
3
3
2/

m-m
m
1/4
2 -2H
3
3/
2- 2%
1

1%

1 -2H

m -2%

2H
2%-3
/-1/2

2H
3
U -l

m

m

i%
2

m

2Vi
3
3
2H
1

1/2

m -m

m

m
m
i%

1
2%
3

i Based on reports for approximately the same dates as those used in table V-3.
transactions completed by a small number of dealers.

%s-m
m-m

2/2-2^
3^-3%
1^

m
m
m
2m
m
m
2%
3

m

Rates shown are for

Eates charged on E P ' s with commercial banks appear to be equal
to the rate on E P ' s with nonfinancial corporations and to be lower
than the rates charged by commercial banks for collateral loans.
Comparison of the minimum loan and E P rates suggests that onefourth percent is frequently sufficient to equate these alternatives.
The rates on E P ' s with nonfinancial corporations which are presented
in the table are based on a small number of observations in several
years. However, they appear to support the conclusions arrived a t
above: I n periods of tight money, the maximum reported rates are
above the discount rate at the New York Federal Eeserve Bank but
below the maximum rates charged for collateral loans at New York
banks.
Again, we can compare the data which dealers reported to the Joint
Economic Committee with the more comprehensive sample for 1958
published in the Treasury-Federal Eeserve study. 11 T h e two sets of
data are similar for the spring of 1958. Both show approximately the
same range of rates for dealer collateral loans at commercial banks and
the dealer E P ' s at commercial banks. Moreover, the Treasury-Federal Eeserve data indicate that the average interest rate paid by Government securities dealers was lower for E P ' s than for collateral loans.
Furthermore, the data there suggest that the repurchase agreement
is a more important source of financing to Government dealers than to
other brokers and dealers or to New York Stock Exchange firms.
This would imply that Government securities dealers are more active
in bringing reserves from other parts of the country into New York.
I n part, of course, this is simply a reflection of the kind of securities
on which repurchase agreements are most often made. A n d this is
11

See especially tables A - 1 2 a n d A - 1 3 in vol. I I , p p . 1 2 0 - 1 2 1 .




STUDY OF FEDERAL GOVERNMENT SECURITIES DEALERS

89

also reflected in the relatively low average rates which Government
securities dealers pay for E P ' s .
EFFECT OF THE RATE STRUCTURE ON SOURCES OF FINANCING

When the distribution of financing sources for the four dealers
(table V-3) is compared to the rates which dealers reported (table
V - 4 ) , the data supports the conclusion that relative changes in interest
rates charged by the various lenders alter the composition of dealer
sources of funds. But, even when there are substantial rate differences between E P ' s and collateral loans, e.g., in 1956 and 1957, some
collateral loans have been used as sources of funds. I n part, this
reflects the upper limit on the amount of E P ' s available to the dealers
from nonfinancial corporations. Given the existing rate structure,
there is no opportunity for corporations to borrow and enter into repurchase agreements. However, it suggests that banks outside of New
York may not take full advantage of their opportunities to borrow at
their district Federal Reserve banks.
Nevertheless, the two tables do suggest that dealer operations are an
important element in spreading the effects of changing interest rates
throughout the economy. Variations in the proportion of commitments financed with collateral loans and the proportion financed with
HP's tend to follow differences in rates. Similarly, the proportion
financed in New York tends to rise when rates are low and fall when
rates are high. However, several exceptions can be found; e.g., dealer
Y in 1954 presumably paid as much as 2 % percent for E P ' s when
collateral loans were available in New York at 1% percent. Such
exceptions probably arise because of incomplete reporting of rates
charged and an absence of information on rates which would be
charged for additional transactions. A t times slightly lower rates in
New York may reflect the potential cost of converting clearinghouse
funds into Federal funds.
Finally, it should be noted that the existing system of borrowing
operates to the advantage of the bank dealers during periods when
money is tight. At other times, nonbank dealers can obtain funds from
the clearinghouse banks and others at rates which are in the neighborhood of the discount rate and in quantities generally sufficient to
finance most of their commitments. During tight money periods,
however, the rates which dealers pay and their sources of financing
seem to indicate that nonbank dealers must offer substantially more
than the rediscount rate or reduce their positions. 12 Access to repurchase agreements at the Federal Eeserve, a source of funds which
is closed to the bank dealers is an offsetting advantage. But, such
funds are uncertain and available only at the discretion of the Eeserve
bank.
Bank dealers have the option of borrowing at their Federal Eeserve banks. As long as this privilege is restricted to banks, bank
dealers have a rate advantage. However, their dealer position is but
one part of their operation. If the supply of rediscounts is not unlimited, the advantage of borrowing at the Federal Eeserve need not be
translated into an advantage in the Government securities business.
& Testimony of several nonbank dealers to this effect may be found In Hearings * * *
pt. 6C, pp. 1848 ff. However, other nonbank dealers did not agree.



90

STUDY OF FEDERAL GOVERNMENT SECURITIES DEALERS

Put otherwise, the dealer department of a bank will be able to
finance its commitments in Government securities at lower rates only if
the rediscount window is open and the rate of return in the dealer
department is expected to exceed the return in other departments. Indications that the rates which banks pay at the Federal Reserve are
lower than rates which nonbank dealers pay to their sources of funds
is not sufficient to indicate that bank dealers have an advantage over
their nonbank competitors.
Some evidence on the share of total net position held by bank dealers is available. I t shows that during tight money periods, the proportion of the aggregate inventory held by bank dealers was larger
than during periods of easy money. While this conclusion is based
on a relatively small number of observations, it suggests that the
difference in rates may be converted into an advantage in positioning
securities. If this is so, it implies that the expected rate of return
from additional dealer operations is higher than the perceived rate on
other alternatives available to the bank but less than the cost of additional collateral loans to nonbank dealers. At such times, if net interest income is negative for nonbanks and positive (or zero) for
banks, banks would be more willing to undertake additional commitments. However, the evidence is also consistent with the hypothesis
that there are differences between dealers in the expected rate of return to be derived from additional commitments.
LEVERAGE AND MARGIN REQUIREMENTS

Earlier in this chapter, we presented the ratio of sources to uses of
funds. From those data, it is clear that dealer commitments exceed
dealer financing by a relatively small amount. I t is, therefore, not
surprising to find that the dealers operate with relatively high ratios
of borrowed to own funds.
The reasons for high leverage have also been suggested by the
earlier discussion. Almost all loans or other financing arrangements
are collateralized or guaranteed by a deposit of Government securities. Risk of loss of principal is slight or nonexistent in most periods.
Moreover, the gross profit per dollar of invested capital is relatively
low in most years as will be seen in chapter V I I .
Looked at positively, high leverage ratios testify to the efficiency
of the market as measured by unit costs. They permit dealers to make
transactions at very low unit costs. I t is clear that any advantage
which would stem from a reduction in leverage or an increase in
margins would be accompanied by a rise in average unit costs. Lower
leverage ratios arrived at through increased capital in the dealer market also might result in a reduction in the number of dealers.




STUDY OF FEDERAL GOVERNMENT SECURITIES DEALERS

91

TABLE V-5.—Annual leverage ratios; nonbank dealers

Year

T o t a l loans
plus R P ' s
divided b y
t o t a l longt e r m financing
(1)

1948
1949
1950
1951
1952
1953
1954
1955
1956
1957
1958

_

_

13.17
13.26
10.38
13.54
22.99
20. 54
15. 34
12. 21
11.36
21. 47
11.49

T o t a l sources T o t a l sources
of funds
of funds
divided b y
divided by
t o t a l longtotal net
t e r m financworth
ing
(2)
15.12
15.10
13.92
17. 77
31. 06
28. 23
22. 20
19. 36
20.01
28.20
19.68

(3)
18. 56
16.59
15.41
19.70
34.96
32.56
25.12
22.79
24.40
33.77
23.44

Three different leverage ratios are shown in table V-5. The differences in numerators were explained earlier in the chapter. The
denominators are both measures of dealer capital. I n columns (1)
and (2), capital is measured by the amount of long-term dealer funds
invested including long-term borrowing and senior securities; the
denominator of column (3) excludes long-term borrowing and some
other items. A complete discussion of these measures is found in
chapter VI.
Both columns (2) and (3) indicate that dealer capital has varied
between 3 and 7 percent of total dealer financing. Changes in the
ratio roughly conform to changes in dealer commitments and suggest
that a capital requirement of approximately 3 percent or a leverage
ratio of 35 to 1 are close to the respective minimum and maximum for
dealers as a group. Such ratios are undoubtedly higher than those
of other classes of security dealers. They greatly exceed the depositcapital ratios of commercial banks.
Several firms have ratios considerably higher than the average.
Five dealers have leverage ratios for the 11-year period which equal or
exceed 35 to 1 and three dealers have leverage ratios above 50 to 1
if the ratio in column (3) is used as a measure. 13 F o r three dealers,
the average leverage ratio has been much lower and has averaged
less than 20 to 1.
Differences between dealers reflect a number of differences in operation. Dealers who specialize in bills and tend to finance extensively
by means of repurchase agreements are more likely to have high levIf col. (2) is used a s t h e m e a s u r e of leverage, one of t h e t h r e e dealers is eliminated.




92

STUDY OF FEDERAL GOVERNMENT SECURITIES DEALERS

erage ratios since the risk in a given dollar amount of commitments is
lower and the margin required on repurchase agreements is smaller.
Two firms which are members of the New York Stock Exchange must
allocate capital against their commitments as specified in the rules of
the exchange. Since these "capital charges" are slightly higher than
the margin requirements generally charged by banks, the average
leverage ratio for the two firms is intermediate if dealers are ranked
from high to low. Moreover, New York Stock Exchange member
firms and other dealers who have relatively large positions in corporate, municipal, or agency securities tend to have lower leverage
ratios because of the higher margins required when these securities
are used as collateral. Such securities are less frequently used in
making repurchase agreements.
Despite the relatively high leverage ratios, there have been no reports of dealers failing to repay their obligations. As indicated
earlier, most collateral loans are for very short periods; many are
overnight loans. The combination of low risk collateral, short duration, and the margins reported in table V-6 is probably sufficient to
protect lenders.
TABLE V-6.—Margin requirements on collateral loans

Maturity:
Bills
Certificates
Notes and bonds under 5 years
Bonds 5 to 10 years
Bonds over 10 years
Agencies
1

Margin
% of 1 percent.
y2 of 1 percent.
1 point.1
2 points.
3 points.
5 points.

1 point equals 1 percent of the value of the securities.

The margins reported here are those in effect during the summer
of I960.14 Small dealers report that they meet the same margin requirements as large dealers, and as indicated earlier, appear to have
access to credit at the same rates. I n fact, since many of them
concentrate most heavily in the short-term market, their leverage is
often higher than that of the larger dealers. However, the TreasuryFederal Reserve 15 study indicates that margin requirements are variable on both collateral loans and on repurchase agreements. Some of
this variation results from aggregating all maturities together but
some must result from differences in regulations at individual banks.
Additional evidence on other dealer margin requirements can be
obtained from the balance sheets of the dealer firms. The margin
requirement for borrowed securities, the ratio of collateral deposited
for securities borrowed to the value of securities borrowed, averages
about 2 or 2*4 percent. When cash is pledged in place of securities,
the margin appears to be as high as 3 to 4 percent, but this may reflect
differences in maturity as well as differences in lender attitudes.
Margin requirements on repurchase agreements vary with the source
of the E P . Many banks now require that the value of the securities
deposited against the E P exceed the value of the contract. Others,
particularly nonfinancial corporations, did not require any margin
14
There is no indication that they change over time, but slightly different rates are
reported in the testimony of Mr. Girard L. Spencer. Cf. hearings, op. cit., pt. 6B<, pp.
1558-1559. Capital charges for New York Stock exchange firms are also reported by
Mr. Spencer.
10
Footnote II, p. 62, and tables A-9, A-10, and B-6.




STUDY OF FEDERAL GOVERNMENT SECURITIEiS DEALERS

93

in the past, and some do not require any now. However, several dealers report that they encourage their corporate lenders to ask for some
additional collateral and often deposit more than is requested.
The ratio of securities pledged for RP's to the value of the R P ' s was
computed from the balance sheets submitted by several dealers. Collateral deposits in excess of 1 percent were rarely observed. I n several
years, individual dealers carried $20 or $30 million in securities on a
margin of one-tenth or two-tenths of 1 percent. Margins below onehalf of 1 percent were far more common than those above. And the
Treasury-Federal Reserve study shows that it is not uncommon to have
zero margin on repurchase agreements even when the securities under
repurchase have more than 10 years to maturity. That study shows
no margin requirement for three-fourths of the RP's made against
$67 million in bonds due in 5 to 10 years and five-eighths of the $115
million in RP's against bonds with more than 10 years to maturity.
Where margins were reported for the former, they averaged only 0.15
percent; for the latter, they were slightly higher but less than one-half
of 1 percent.
Some dealers have suggested that repurchase agreements are sales
and purchases rather than loans. F o r this reason, they might regard
margin requirements on such transactions as an anomaly. There is
some merit to this point of view. However, it is clear that they differ
in two important respects from regular purchases and sales: (1) the
completion of the transaction depends upon a particular dealer rather
than on the marketability of the instrument and (2) the rate at which
the loan is made assumes the absence of all but minimal risk. The
rapid falls in Government security prices in May 1953 and June 1958
suggest that the risks may be underestimated at certain times.
The margin requirements which dealers now pay for collateral loans
could be extended to repurchase agreements. Margins wTould eliminate potential risks to lenders—particularly to those small banks and
nonfinancial corporations which are unfamiliar with the risk involved
in RP's when security prices fall. This is particularly important for
R P ' s which are made against securities which have more than 1 year to
maturity. Administration of the margin requirements would not be
difficult now that ail dealers report their repurchase agreements to the
Federal Reserve Bank of New York on a daily basis. 16
CASH BALANCES

The cash assets of dealers vary much less from year to year than
other asset items despite their very high velocity. As we have seen,
net position turnover for the 17 dealers has been as high as 1% times
per trading day and has averaged at least one per day. Given the
small cash balances and the high leverage ratios which are typical
of dealer operations, the cash balance turnover is undoubtedly much
higher. Estimates by the New York Clearing House suggest that average cash balances of $20 million turn over about 45 times per day or
11,000 times per year on the average. 17 This estimate is consistent
16 The Federal Reserve Bank of New York requires a margin on dealer repurchase agreements. Cf. hearings, op. cit., pt. 6B, p. 1564.
« Hearings, op. cit., pt. 6B, p. 1511.




94

STUDY OF FEDERAL GOVERNMENT SECURITIES DEALERS

with our previous estimate of an average net position of $950 million
and an average position turnover of one per trading day.
The New York Clearing House estimate of mean dealer cash balances
is slightly lower than the year-end cash shown on dealer balance
sheets. The difference is approximately $2 to $3 million on the average for the years 1948-58 inclusive and is larger in recent years than
in earlier years. This probably reflects a geographical redistribution
of dealer deposits which has accompanied the shift in sources of funds.
I n several interviews dealers were asked if they maintained minimum cash balances. Most of them indicated that they kept balances at
those banks from which they normally borrowed but felt no formal
obligation required them to do so. One dealer indicated that at the
request of particular banks he maintained minimum balances which
varied with the activity of his account. H e also stated that in periods
of stress, e.g., during a Treasury refunding, he would be able to reduce
the amount of this balance without any comment from the bank.
FINDINGS

The financial arrangements which permit the dealer market to
operate are an important part of the mechanism by which reserve
positions of banks are adjusted and idle balances are mobilized. In
the process of increasing their profits (or reducing their costs) dealers
must search for idle balances which can be obtained at rates sufficiently low to permit them to carry their commitments. Through
these efforts interest rate changes and monetary policies are spread
throughout the country.
The growth of repurchase agreements during the postwar years
has played an important role in making the money market more efficient. Through the R P , the existence of higher interest rates is
brought to the attention of small banks and nonfinancial corporations.
Undoubtedly this has led to a more rapid adjustment by the private
economy to the effects of changes in the supply of money and credit,
and to a reduction of the timelag between the introduction of a policy
and the effectiveness of the policy.
The evidence on the financing of the market is as yet too scanty to
come to more than a few preliminary conclusions on the effectiveness
of possible changes. A more complete report, prepared by the clearinghouse banks has been prepared but has not been released to the
public. This report, or a similar study, should be made available to
permit a broader understanding by the interested public of the ways
in which monetary policy alters the distribution of the money supply.
High on the priority list should be a study of the effects of higher
margin requirements on loans to nondealers. The available information seems to indicate that nondealers deposit slightly higher margins
than dealers on the average, but at times the cash or collateral deposited has been insufficient from the view of lenders. Margin calls
have led to forced sales and rapid decreases in the price of Government securities. Both the magnitude of present margin requirements
and the expected effects of higher margins on the pricing of Government securities should be investigated before a conclusion can be
reached on this subject.
I n addition, more information Is needed about the availability of
securities which can be borrowed. This issue arises principally when



STUDY OF FEDERAL GOVERNMENT SECURITIES DEALERS

95

dealers want to sell long-term issues short. Trusts, pension funds,
and governments hold a large share of many of these issues. Their
rules often prevent them from lending securities. Mutual savings
banks, which formerly were a principal lender of securities, have decreased their portfolios of governments in recent years and no longer
have an available supply of many issues. The absence of an ability
to borrow reduces the opportunities for dealers to arbitrage or to
supply securities to the market. 18 An improvement in arrangements
for borrowing securities would improve the marketability of longterm issues at least marginally and wTould strengthen this important
segment of the bond market.
Further investigation of the available supply of funds for financing dealer portfolios of long-term and intermediate term bonds would
be desirable. I n this study, we have not ascertained the extent to
which funds are available for carrying longer term securities in
amounts and at rates similar to those quoted above. Many of the
sources of dealer financing are short-term and hence geared to securities which turn over rapidly. Longer term securities often must
be held for longer periods of time before sales can be arranged. Conversations with several dealers indicated that much of the time it is
difficult to finance longer term issues at rates below or equal to the
yield of the securities. If this is so, it is an additional reason for
the weakness of the long-term bond market and hence a matter of
public interest.
Two other conclusions seem warranted:
First, margin requirements on repurchase agreements by dealers
could be increased. On this point, the evidence that dealers deposit
substantially lower margins for KPs than for collateral loans is reasonably good. There does not appear to be any overriding consideration which makes lower margins a necessary accompaniment of financing through KPs. The difference in rates at most times is sufficiently
large to permit higher margins without substantially altering the
composition of dealer financing. The Federal Reserve Bank of New
York could assure that such margins are maintained by requiring that
the appropriate information be submitted as part of the dealers daily
reports. To the extent that lower margins are available on collateral
loans made outside of New York, this difference should be eliminated
also. We believe that this requirement would reduce the probability
of margin calls, forced selling, and market disorganization.
Second, loan rates at commercial banks seem to be far less flexibile
downward than the rates on many of the securities which banks buy.
As a result, nonbank dealers are at times less willing to take securities
into position at existing interest rates. Particularly, in the long- and
medium-term market where transactions are relatively small, this
has a temporarily disturbing influence on the rate structure. Dealers
must take the cost of carrying securities into consideration when prices
are quoted. A more flexible interest rate policy at commercial banks
and particularly at clearinghouse banks would assist in the adjustment of the economy which takes place through the dealer market.
18
This is one of the reasons for the increase in the "failed to deliver" account in recent
years.







CHAPTER VI
LEGAL ORGANIZATION AND CAPITAL B A S E
LEGAL ORGANIZATION, CAPITAL BASE, AND CREDIT WORTHINESS

Dealers in Government securities can fulfill their functions only if
they carry a position in the securities for which they are making
markets. Trading in and out of positions becomes profitable over the
long run only because of the dealers' ability to finance their securities
inventory with borrowed funds. If they were forced to finance commitments with their own capital funds, that is, to forego any leverage
effects, either trading would become so unprofitable that it would probably disappear or spreads between buying and selling prices would
become impossibly large. The dealers' ability to borrow for the purpose of carrying Government securities in their positions is thus a
crucial condition for the functioning of this important market and
for profitable dealer operations. Dealers' ability to borrow funds at
rates which will allow them to operate profitably at prevailing spreads
is determined by two sets of factors, (1) the liquidity of the securities
which can be used as collateral, and (2) the credit worthiness of the
individual dealer firms.
Government securities as collateral offer the highest liquidity of
any security market collateral except cash, with liquidity increasing
as the time to maturity of the securities decreases. The credit worthiness of the dealer firm is determined by the lending institutions on
the basis of a number of considerations. Among the considerations
is the requirement that dealers must report their financial condition
to the Federal Eeserve Bank of New York. The submission of these
financial statements is one of the conditions imposed on a dealer firm
if it wants to do business with the open market desk. This can be
considered as an important screening process which other lenders may
take into account when they evaluate the credit worthiness of the
dealers.
Lenders generally consider the legal form of organization and the
capital cushion represented by a borrower's equity among the important determinants of a debtor's credit worthiness. Neither factor seems to carry much weight in appraising the credit standing
of dealers in U.S. Government securities. I n relation to the large
dollar amounts involved in day-to-day financing, dealer capital is
quite small, while the protection offered to the lender by collateral arrangements is substantial. However, dealers' personal
integrity and past experience—measured by their past performance
record—seem to be of great importance to lenders.
Little or no difference is noticeable in the ease with which both
dealer-partnerships and dealer-corporations maintain their credit
standing. This appears to hold true despite the fact that a dealer




97

98

STUDY OF FEDERAL GOVERNMENT SECURITIES DEALERS

partnership offers not only the firms' capital but also the assets of the
general partners as added security cushion in case of financial failure.
The legal form of a dealer firm becomes important for this study, however, in comparisons of the capital base of dealer-partnerships and
dealer-corporations, and in the treatment of their income for income
tax purposes. [See ch. V I I . ]
LEGAL FORM OF ORGANIZATION OF THE 12 NONBANK DEALERS

The 5 bank dealers among the 17 firms are departments or divisions
of their banks which, of course, are incorporated. Except for an
attempt to estimate the share of capital which the bank dealers have
committed to their Government bond departments, the discussion in
this chapter will be concerned with the 12 nonbank dealers. Comparisons between dealers on the basis of legal form is made difficult by differences between individual firms. As noted earlier, some dealers confine their operations to Government securities while others are broadly
based securities firms. Moreover, at the end of 1958, 4 of the 12 nonbank dealers were organized as partnerships while 8 were incorporated.
One of the larger firms had changed during the 12-year period, 194758, from a corporation to a partnership.
While one-third of the nonbank dealers in U.S. securities chose the
partnership form of organization, this is not typical of securities
dealers in general. Bettei4 than 90 percent of the member firms of the
New York Stock Exchange were partnerships at the end of 1958; only
53 out of a total of 657 N Y S E member firms were incorporated at that
time. I t appears that the Government dealers have a greater preference for the corporate form than member firms of the N Y S E , but this
inference is based on a comparison of the population of 12 dealers with
that of all N Y S E member firms.
Several factors suggest that Government security dealers may differ
from other security dealers in their attitude toward incorporation.
The corporate form was available to Government security dealers at
an earlier date. Until recently, member firms of the N Y S E could not
be incorporated. Thus a trend toward incorporation among this group
may be now underway. Since 10 of the 12 Government dealers are
not members of the N Y S E they were not restrained from incorporating.
Three other possible explanations for a dealer preference of the corporate form which require more detailed analysis are (1) size and
organization of dealer firms, (2) tax considerations, and (3) trading
risk and limited liability.
SIZE AND ORGANIZATION

The size of dealer firms does not appear to influence their choice of
legal form. Both very large and very small firms are found in both
groups. 1 For comparison, the distribution of corporations, partnerships, and sole proprietorships by size classes for the 400 largest
investment banking firms in the United States is given in table V I - 1 .
1

"Finance," Mar. 1$, 1960.




STUDY OF FEDERAL GOVERNMENT SECURITIES DEALERS

99

TABLE VI-1.—Legal form of 400 leading investment bankers by size classes
Corporations
Size of firm capital i n t h o u s a n d s

P a r t n e r s h i p s a n d sole
proprietorships

Number
of firms
i n class
Number

Percent

Number

Percent

Over $5,000
$1,001 to $5,000
$501 to $1,000
$251 to $500
$100 to $250

65
154
61
68
52

12
50
29
44
34

18.5
32.5
47.5
64.7
65.3

53
104
32
24
18

81.5
67.5
52. 5
35.3
34.7

All firms

400

169

42.2

231

57.8

Source : Finance, Mar. 15, 1960.

These data show that the probability that an investment bank
will incorporate decreases with size throughout the range. This is
not true when a similar array is constructed for the 12 nonbank dealers. Of the two dealers with less than $500,000 capital at the end
of 1958, one was incorporated and one was a partnership. Five of
the dealers had more than $5 million in capital funds. Two of these
were partnerships. Of the dealer firms with capital funds ranging
from $500,000 to $5 million, only one was a partnership. Thus,
there is no evidence from the small population of Government security
dealers that size of the dealer firm determines the choice of legal
form.
Most of the incorporated dealer firms seemed to be managed on
the same basis and along similar organizational lines as the dealer
partnerships. Senior stockholders or officers assume the role of
senior partners with only rare exceptions. The "corporate manager"
with no stake of his own in the business was not very much in evidence. A combination of experience and reputation as a trader and
capital investment seem to be the prerequisites for leading positions
in the nonbank dealer firm.
TAX

CONSIDERATIONS

The corporate form of organization can have important tax advantages to owners of the business if they are or expect to be in very
high personal income tax brackets. A security dealer firm, organized as a partnership, does not have the option under the 1954 Internal
Eevenue Code to elect to be taxed as a corporation. Even if this
option were available to the dealer partnerships, it is unlikely that
they could reap the most important tax advantage of the corporation which adhere from the retention of earnings after payment of
the 52 percent (maximum) corporate income tax. The reinvestment
of these earnings in the business and the resulting appreciation in
value of the business can be realized by the owners through sale of
their share of the business at a later date. Any gain realized is then
taxable only at the capital gains tax rate which cannot exceed 25
percent of the gain. 2
Two conditions have to be met to make this tax aspect of the corporation worthwhile to the owners of a business: they must be in a 643
A more complete discussion of the effects of capital gains taxation is found in chapter VII.




100

STUDY OF FEDERAL GOVERNMENT SECURITIES DEALERS

percent or higher personal income tax bracket and their share in the
business must be easily marketable. While there is little or no evidence about individual partners' incomes, general marketability does
not seem to exist. Although there have been some exceptions, the
usual way of transferring interests in the dealer firms consists of
bringing in promising younger men with experience and increasing
their stake in the business gradually. Interest in the ownership is
rarely sold to outsiders or nonemployees.
This argument appears valid even though retained earnings have
been by far the most important source of capital for the eight dealer corporations during the period 1948-58. Their aggregate net
worth including preferred stock and certain reserve accounts increased by $8.7 million; their retained earnings or earned surplus accounts increased by $14.7 million. The difference between increases in
retained earnings and additions to total net worth was caused by the
retirement of some preferred issues and the reduction in capital accounts designed mostly to make up for losses.
Total long-term investment of the eight firms increased even less
than their net worth; namely, by $7.3 million. Since long-term investment figures include the firm's indebtedness, it is clear that the incorporated dealers as a group were able to apply their retained earnings toward retirement of some of their debt. This occurrence, however, was not peculiar to the corporate dealers. Similar developments
took place in the partnership firms and are reflected in increases of
net worth and, more specifically, "partners' free balances" and "partners' individual balances."
TRADING R I S K AND LIMITED LIABILITY

As has been pointed out repeatedly (see chs. I l l and V ) , the major
trading risk derives from the holding and financing of large positions.
Here again the U.S. Government securities dealers stand out among
other security dealers in both the over-the-counter and organized exchange markets. I n "The Over-the-Counter Securities Market," 3
Winn states that the relationship between the long security positions
and "net capital" of 3,081 registered broker-dealers during the September-November period of 1949 showed a definite variation with the size
of firm. H e found from questionnaire responses that the ratio of average positions to "net capital" was 203 percent for the largest size class
(50 firms with "net capital" of over $3 million), 101.7 percent for the
medium size class (175 firms with "net capital" of over $500,000), and
97.1 percent for the smaller size class (2,856 firms with "net capital"
of less than $500,000).
The use of the "net capital" concept results in an upward bias in
the relationship between dealer inventory and dealer net worth as
noted by Winn. The comparison of 1949 position with 1947 capital
further overstates the relationship. Yet these figures serve as a basis
for comparison with similar data for the 12 nonbank dealers in Government securities covered in this study.
3
1 . Friend, G. W. Hoffman, M, J. Winn, "The Over-the-Counter Securities Market,"
McGraw-Hill Book Co., Inc., New York, 1958, pp. 68-69 and 297 ff.




STUDY OF FEDERAL GOVERNMENT SECURITIES DEALERS 101
TABLE VI-2—Ratios of positions
1948
1949
1950
1951
1952
1953

to net worth, all nonbank

Times
14. 51
11. 68
9.12
11. 59
19. 36
26. 25

1954
1955
1956
1957
1958

dealers,

1948-58
Times
10. 89
12. 20
11. 39
22. 59
14.17

Moreover, the Winn data include his estimates of the position and
"net capital" of the Government securities dealers. If these were eliminated, the ratios shown by Winn would be reduced and the difference
in the ratio of capital base to positions would exceed the differences
shown here.
Even though the inventories of dealers and brokers in corporate
securities are subject to much wider price fluctuations than those of
dealers in U.S. securities, the data bear out the statement that the 12
dealers in this study are faced with substantial risks deriving from
the relationship between their capital and their total commitments.
The fact that profitable operations in the Government securities
market require a relatively high ratio of position/capital suggests
that the predominance of the corporate form may be more closely
related to risk rather than to other considerations often pointed t o :
tax advantages, organizations, size of firm, or credit worthiness.
DEALERS' CAPITAL BASE

I n table V I - 3 the net worth of the 12 nonbank dealers is summarized
for the years 1948-58. During the 11 years, aggregate net worthy of
nonbank dealers (col. 1) declined during 3 years and increased during
7 years. Declines and increases are, of course, caused first of all by
fluctuations in earnings and changes in retention policy.
Net worth as defined here includes all surplus accounts, all common
and preferred stock, and certain reserve items which displayed stability over time and appeared to be segregated surplus reserves rather
than liability reserves. F o r the partnership dealers, classification at
times was more difficult, especially with regard to "Partners' free
balances" and "Individual partners' free balance." F o r the compilation of net worth presented here, these special accounts have been
omitted. I t appears, however, that the funds represented by these
accounts are committed to the firms for long periods of time. Since
the balance sheets on which this study is based are mostly year-end
statements, they may contain accrual accounts on which the partners
may draw during the course of the year. Considering their size, it
seems unlikely that they would be completely eliminated during the
normal course of events. F o r the purpose of calculating rates of
return on capital, these items were added back to net worth together
with a few relatively small long-term debt items. The resulting
amounts of total long-term financing (capital base) are also presented
in column (2) of table V I - 3 .




102

STUDY OF FEDERAL GOVERNMENT SECURITIES

TABLE VI-3—Nonbank dealer

long-term

financing

DEALERS

and net worth,

1948-58

[In thousands of dollars]

Net worth

Long-term
financing
other t h a n
net worth

Long-term
financing
plus n e t
worth

(2) as percent
of (1)

(1)

(2)

(3)

(4)

Year

1948
19491950
1951.
1952
1953
1954
1955
1956
1957
1958

44,465
52.886
47,955
46, 777
47, 797
51,308
56, 901
58, 832
50,317
56,223
61,958

_

54,566
58,099
53,090
51,865
53,802
59,183
64,183
69,260
1,360
7,312
3,777

10,101
5,213
5,135
5,088
6.005
7,875
7,282
10,428
11,043
11,089
11,819

22.7
9.8
10.7
10.9
12.6
15.3
12.8
17.7
21.9
19.7
19.1

Long-term debt has been relatively unimportant in recent years.
I t appears on the balance sheets of only some of the eight dealer corporations. I t shows high volatility from one year to the next, and the
impression is gained that much of this debt may consist of loans to
the firms by insiders or persons closely related to the firm. Table
V I - 4 summarizes the aggregate long-term debt of the eight corporate
nonbank dealers during the 11 years.
TABLE VI-4.—Long-term debt, for all dealer corporations,

19^8-58

[In thousands of dollars]
Long-term
debt

T o t a l longt e r m financing

Col. 1 as percent of col. 2

(1)

(2)

(3)

Year

1948
1949
1950
1951
1952
1953
1954
1955
1956
1957
1958

3,926
1,488
1,705
1,697
1,708
1,710
1,685
928
1,725
1,528
177

48,265
43,447
38,844
38,476
38,647
41,911
42,833
44, 518
40,256
44,464
48,437

8.13
3.42
4.39
4.41
4.42
4.08
3.93
2.08
4.28
3.43
.36

Since long-term debt constitutes such a small part of dealers' total
long-term financing, in most cases, it was not segregated from this
amount for the purpose of calculating rates of return. The data presented in this section allow for any desired reconciliation on the part
of the reader.
The 12 nonbank dealers increased their net worth by $17.5 million
from 1948 to 1958 with only 3 years of decreases. This aggregate
performance was achieved despite substantial temporary and some
permanent reductions of net worth of individual dealer firms during
the period. The permanent reductions of net worth for instance total
$2.5 million for the 11 years. This reflects losses realized by this group
of firms and some withdrawals of capital by a few dealers.
Total long-term financing of the 12 firms during this period increased by $19.2 millon. Here again a group of firms experienced a
reduction of total long-term financing of $1.9 million resulting from



STUDY OF FEDERAL GOVERNMENT SECURITIES DEALERS 103
losses and from retirement of outstanding issues. Accumulation of
partners' balances and additional use of outside funds supplemented
the capital provided through increases in the dealers' net worth.
A t the end of 1958, the operations of the 12 dealers were conducted on a total long-term capital base of $73.8 million. However,
several of the nonbank dealers had substantial operations in corporate
and municipal securities during the 11 years. Ideally one would like
to allocate part of these dealers' capital base to their other activities.
This task is difficult in the absence of more specific information about
their activities in other markets. F o r the purposes of this discussion
no such allocation will be attempted. But, despite the lower ratio of
positions to dealer capital in other securities markets, the relatively
small positions which Government dealers take in non-Government
securities (see ch. I l l ) suggest that only a small proportion of aggregate dealer capital is allocated to activities outside the Government
market.
Theoretically at least, one could add to the nonbank dealers' capital
the amount of the dealer banks' capital which would be necessary to
support their operations in U.S. Government securities if they were
nonbank dealers. F o r this purpose, the average rate of capital turnover (ratio of dealer transactions volume to long-term financing) for
the 12 nonbank dealers was compared with the transactions volume of
the bank dealers. This method leads to an estimate of bank dealer
funds committed to support trading operations of the bank Government bond departments of between $24 and $30 million or a total
commitment of approximately $100 million by all dealers in the Government securities market.
CAPITAL BASE AND CASH ACCOUNTS

Dealer cash is a very small part of dealer assets averaging no more
than 2.5 percent for the entire period. However, over the 11 years the
12 nonbank dealers have held a rather large and constant portion of
their total capital funds in the form of cash balances. Table V I - 5
presents the relationship between total long-term financing and the
cash balances on hand or held with banks for the years from 1948 to
1958. F o r the entire period cash balances averaged 38.1 percent of
total long-term financing, ranging between a high of 41 percent in
1953 and a low of 33.5 percent in 1957. The decrease of the cash
balance ratio during the years 1954 to 1957, reflects tight monetary
policy and the increase in the cost of funds as well as a substantial
addition to earned surplus in 1957.
TABLE VI-5.—Cash positions

Cash positions
(in thousands
of dollars)

Year

1948.-. —
1949
1950
1951
1952
1953

20,408
22,891
20,701
21, 111
21,705
24,267

__

S2471—60

8




and total long-term
1948-58
Cash positions
as percent of
total long-term
financing
1954
37.4
39.4 ! 1955
39.0
1956
40.7
1957
40.3
1958
41.0

Year

financing,

12 nonbank

dealers,

Cash positions Cash positions
(in thousands
as percent of
total long-term
of dollars)
financing
24,817
23,278
21,409
22,554
29,918

38.7
33.6
34.9
33.5
40.6

104

STUDY OF FEDERAL GOVERNMENT SECURITIES DEALERS

P a r t of the total long-term financing which is not tied up in cash
balances, is used to finance relatively unimportant assets like prepaid
expenses and office equipment. Most of the remaining portion of the
long-term financing is available and necessary to maintain margins on
borrowed securities and on collateral loans. I n addition, the cash
balances serve to adjust any imbalances between such current asset and
current liability accounts as receivables from dealers and customers,
accrued interest receivable and payable, and other temporary aberrations in the flow of funds.
FINDINGS

The single most important criterion for the choice of legal form
appears to be the protection from the risk of carrying positions which
are large in relation to the capital invested. The extent to which tax
advantages influence the decision to incorporate is an empirical question which cannot be conclusively answered. I t seems that lack of
marketability of ownership rights in these firms denies the opportunity
of taking full advantage of the capital gains potential under the corporate form.
The dealer market operated at the end of 1958 on the basis of an
estimated capital commitment of between $97.8 and $107 million. Of
this estimated total, the 12 nonbank dealers accounted for $73.8 million. This capital base increased over the 11-year period despite
temporary reductions which were mostly due to losses and, to some
extent, capital reductions and retirement. The earned surplus accounts were the most volatile items accounting for most of the net
worth and capital base reductions during the years 1950, 1951, 1955,
and 1956, as well as for the overall increase in the capital base during
the 11-year period.
The increase in total long-term financing over the period amounted
to 35.2 percent of the 1948 capital base. Total transactions more than
doubled during the same years and total position was 70 percent larger
in 1958 than in 1948. A number of explanations can be given for these
discrepancies: (1) The dealers had an unnecessarily large amount of
capital in 1958; (2) financing techniques have been improved and less
capital is needed; (3) trading has changed from activity in long-term
securities to the short-term end of the market where financing is easier
and lower margins are required; (4) not enough capital can be attracted into the market because rates of return are insufficient for the
risk of exposure of capital funds.
Explanations (1) and (2) have been dealt with in the chapter on
financing of positions (see chapter V) ; point (3) was discussed in the
transactions and financing chapters (see chapters I I I and I V ) . The
following chapter on the earnings and expenses of the dealer firms will
deal at length with the issues surrounding the fourth explanation.




CHAPTER VII
SOURCES OF R E V E N U E , E X P E N S E S A N D EARNINGS
Dealer functions were described in detail in earlier chapters. The
dealer's functions of market making, market stabilization, transmission of changes in monetary policy as well as of day-to-day technical
adjustments initiated by the open market desk take place in a competitive market. I n a competitive market, w^here participants are expected to utilize their capital in the financing of large positions in securities, the necessary capital funds will be attracted and remain committed only if "competitive" rates of return can be realized in the long
run.
Dealers render services to their customers. P a r t of the dealers'
gross earnings can be considered as the price customers in the aggregate had to pay for these services. Net earnings realized by the
dealers, on the other hand, constitute their payment for services rendered plus or minus any gains or losses incurred in exercising their
general economic or market functions.
I n this chapter, gross earnings will be presented and analyzed as
to their origin and their fluctuations over time. Earnings arising
directly from trading activity will be compared with the transaction
volume from which they result to provide a measure of the cost of
trading and of market efficiency. This will in turn be compared with
similar measures for other important national securities markets to
appraise the relative profitability of the dealer market. Impact of
changes in monetary policy and general economic conditions on the
amount and pattern of dealer earnings are also discussed. Finally,
the relationship between earnings and dealer capital will be presented and compared with rates of return on capital funds of other
financial institutions and industrial corporations.
SCOPE OF DATA AND DEFINITIONAL PROBLEMS

The questionnaire on which this study is based asked for a
detailed breakdown of the income and expenses of all 17 dealers for
the fiscal years 1948-58. As noted in the chapters on positions and
transactions, accounting standards and agreement on terminology are
conspicuous by their absence in this market. This difficulty is compounded by the complexity and variety of individual dealer operations
and their cost allocation systems. Dealers responded fully to the
questions relating to the sources of gross earnings from current operations. This part of the analysis of earnings can be based on complete and detailed information for most years. The data received on
various expense items show less completeness and uniformity. Nevertheless, detailed expenses were reported by 16 of the 17 dealers
for most of the 11 years and part of the analytical task in this chapter
is to point to interdealer differences in charges made to gross earnings.




105

106

STUDY OF FEDERAL GOVERNMENT SECURITIES DEALERS

Another difficulty arises from the fact that some dealer firms are
organized as partnerships while the majority are corporations. Therefore available income statements are comparable only to the item "net
income before taxes." Most of the net profit comparisons and calculations wTill be based on this reported item rather than on net income
after taxes.
I n comparisons of dealer profits over time, differences in their fiscal
years become important. Only a small number of dealers, however,
have fiscal 3^ears apart from calendar years so that the bias introduced
in estimating aggregate annual earnings will be very small. Fortunately where fiscal years differ from calendar years, transactions and
capital figures are usually available for the same period. Furthermore, fiscal years overlap calendar years at least for a period of
8 months. Thus the data are comparable for the determination of
individual dealers' profitability and only slightly biased for purposes
of analysis of annual changes and differences in rates of annual change
between dealer firms.
The available data are summarized in tables showing aggregate
income and expenses for the dealer market. I n aggregating the data
to compute ratios the most important problem arose from the absence
of some numerators or denominators. To compensate for this, some
of the reported figures had to be omitted from some of the tables.
However, table V I I - 1 presents a summary of all reported earnings and
expense items. Tables V I I - 2 , and 4 through 10, are based entirely
or in part on the reports of those dealers who gave complete information on earnings as well as on expenses and transactions. The three
bottom rows of table V I I - 2 indicate the scope of the available data.
The number of dealers included never falls belowx 8 and is 13 for the
last 6 years. This dealer group accounts for no less than 77 percent
of all dealer transactions and earnings during the years 1950-58. Information on all reported data is given where possible in the form of
high and low values and as averages of individual dealer ratios. Any
omissions of reported data are footnoted in the tables and/or discussed
in the text.
SOURCES OF EARNINGS FROM CURRENT OPERATIONS

Dealers' gross earnings are derived from five principal sources.
(1) Trading with customers takes place, moving securities in and out
of the dealers' positions. Any excess of sales prices over purchase
prices increases gross earnings; if securities are sold at prices below
acquisition costs, losses result. (2) On securities in which dealers are
holding long positions, interest accrues. U.S. securities are not traded
"flat," that is, accrued interest is paid by the purchaser in addition
to the contracted price. 1 (3) Dealers may realize a gross profit or loss
1
Treasury bills, of course, constitute an exception to this rule because they are traded
on a discounted basis, where the discount reflects the rate of interest or, better, the yield
to maturity.




STUDY OF FEDERAL GOVERNMENT SECURITIES DEALERS 107
from securities held in their investment accounts. As was pointed
out in chapter I I I (positions), only a few of the dealers reported
holding securities in investment accounts. Correspondingly few report on gains or losses resulting from these accounts. Some dealers
stated in interviews that at times they might have held a position in
investment accounts but that they were not able to report for prior
years what portion of their total trading profits they had earned on investment accounts. Despite these reporting inperfections it appears
from the available data that investment accounts and investment account earnings have been relatively unimportant during the period.
Nevertheless, these earnings have been included in total trading profits.
(4) Some dealers reported rather small earnings from "other" sources.
These include such items as service charges, commissions, underwriters' fees, collection and exchange charges, et cetera. Where earnings appear under this classification they might have been derived
from the subscription to and the subsequent sales of new U.S. securities issues. Only the individual reporting firms could clarify this
point exactly. The reported figures under this item are small but are
included. (5) Only one of the dealers reports on earnings in the
form of interest received on loans. Even for this firm these earnings
constitute only a very small part of total gross earnings from current
operations and appear only for part of the period. For purposes of
this analysis, interest received on loans can be ignored. I t is included
in total gross earnings from current operations.




o
00

d
o
TABLE VII-1.—Composition of Gross Earnings from current operations and operating expenses for reporting dealers, 1948-58
[In thousands of dollars]
W

Gross earnings from c u r r e n t operations
I n t e r e s t received
T o t a l t r a d i n g profit
R e p o r t e d profit from i n v e s t m e n t a c c o u n t s
T r a d i n g profit o t h e r t h a n o n i n v e s t m e n t a c c o u n t s
Other earnings..
I n t e r e s t paid—.
__
Salaries
O t h e r i n v e s t m e n t expenses
N e t i n c o m e from operations
N e t i n c o m e before taxes

_

1948

1949

1950

1951

1952

1953

1954

1955

1956

1957

11,681
7,455

28, 241
16,552

17, 708
12,564

19,880
10,474

26,149
14,126

40,753
17,135

48,771
24,668

29,392
19, 733

37, 580
22,392

81,701
31,501

95,666
34,537

3,891

11,390

4,700

8,978

11,354

23,002

23,346

8,696

13,933

48,563

59,984

3
3,888

140
11,250

368
4,332

(398)
9,376

556
10,798

(183)
23,185

897
22,449

(109)
8,805

(470)
14,403

847
47,716

11,200
48,784

335
4,217
2,807
2,291
1,520
1,519

298
10,452
4,053
3,195
8,014
7,369

445
8,245
3,595
3,797
1,041
1,110

432
8,040
4,006
4,026
2,916
3,201

671
10,149
4,385
4,565
5,150
5,037

616
13,020
5,283
5,597
15,037
14,805

756
15,105
5,845
6,628
15,419
15,102

964
14,741
5,480
6,775
494
569

1,257
16,781
5,914
7,033
2,903
3,013

1,635
29,484
7,497
7,843
27,996
27,815

1,143
26,406
8,969
9,482
33,525
33,737

NOTE.—Reported expenses plus income from operations do not add to gross earnings because fewer dealers reported on expenses than on earnings.
ported items are reflected in table VII-2. Sources of earnings may not add to gross earnings because of rounding.




1958

Full adjustments for nonre-

$
&
O
<

O

B

T A B L E VII-2.—Earnings and expense items as percentages of gross earnings from current operations,
1948
T o t a l t r a d i n g profit
I n t e r e s t received
O t h e r earnings

__.

Gross earnings

___ .

Interest paid
Salaries
__
_
O t h e r c u r r e n t expenses_
Local taxes a n d statistical d i s c r e p a n c y
N e t income before special charges a n d income taxes.._

_
__

N u m b e r of dealers r e p o r t i n g
_
T r a n s a c t i o n s i n c l u d e d in t a b l e as percentage of t o t a l for all
dealers
_.
Gross earnings included in table as percentage of total earnings
for all dealers
__
__. _




1949

1950

1952

1951

1953

all reporting
1955

1954

dealers,
1956

1948-58
1958

1957

32.5
65.3
2.2

39.4
59.5
1.1

29.1
68.4
2.5

48.9
48.8
2.3

44.8
52.1
3.1

57.4
49.9
1.7

47.6
50.4
2.0

31.3
64.7
4.0

41.9
54.0
4.1

63.8
33.8
2.4

67.2
31.4
1.4

100.0

100.0

100.0

100.0

100.0

100.0

100.0

100.0

100. 0

100.0

100.0

42.3
26.1
21.2
1.4
9.0

44.4
15.4
11.7
1.2
27.3

56.4
21.4
22.9
1.1

48.3
20.8
20.9
.8
9.2

47.6
18.3
19.0
.9
14.2

37.1
13.5
14.3
.6
34.5

39.3
13.8
15.8
.5
30.6

57.5
18.3
21.8
.7
1.7

45.0
10.5
10.8
.6
33.1

35.4
11.2
11.9
.9
40.6

8

10

11

12

12

13

13

13

13

13

13

59.5

67.3

78.6

79.0

78.3

85.6

83.5

81.4

78.9

78.1

79.3

84.4

81.6

81.0

83.7

81.5

85.8

78.3

82.9

77.2

79.6

76.9

(1.82

60.3
20.2
25.2
.5
(6.4)

110

STUDY OF FEDERAL GOVERNMENT SECURITIES DEALERS
TRADING P R O F I T S , SPREADS, A N D P R O F I T

MARGINS

Data on the amount and composition of reported aggregate dealer
earnings from current operations are presented in table V I I - 1 ; the
percentage distribution of aggregate earnings and expenses adjusted
for missing information is given in table V I I - 2 . I n 3 of the 11
years, 1953, 1957, and 1958, trading profits constituted the single
largest source of aggregate dealer gross earnings. They were second
only to interest received on securities in the other 8 years. The performance of the dealers, their traders, and the existence of competitive
market pressure can be appraised by comparing prevailing spreads
with the profit per dollar of transactions reported by the dealers.
Dealer pricing is one of the most important aspects of dealer
strategy. Telephone publicity on competitors' runs and active shopping by customers around the dealer circuit exert considerable pressure
on the trader to quote competitive prices. He must avoid widening the
spreads between quoted bids and asked prices or the customer will buy
from or sell to another dealer. On the other hand, to avoid losses, he
does not want to shade his prices too much. Annual trading profits will
result only from successful pricing over the course of a year. Thus
the policy of the firm, the skill and experience of traders, plus the size
of spreads between purchase and sale prices allowed by the market
determine the amount of trading profit a dealer firm can realize.




STUDY

OF FEDERAL

GOVERNMENT

TABLE VII-3.—Price quotations

SECURITIES

for U.S. Treasury

DEALERS

bonds and

111

notes

A m o u n t o u t s t a n d i n g (in
millions)

Coupon

Maturity

Bid

Asked

Spread

Yield

(1)

(2)

(3)

(4)

(5)

(6)

(7)

$5,264.
$3,452_
$9,561.
$278—
$3,806.
$1,485.
$144—
$4,078.
$2,136.
$2,239.
$332—
$6,962.
$647—
$1,435.
$551—
$2,211.
$2,109.
$158...
$590—
$1,143.
$3,971.
$533—
$1,743.
$6,755.
$506...
$3,011.
$2,815.
$3,854.
$457—
$4,933.
$3,895.
$3,738.
$2,316.
$490—
$4,195.
$3,812.
$6,896.
$4,691.
$58—
$2,113.
$2,938.
$1,484.
$1,806.
$2,716.
$3,633.
$320—
$1,276.
$654—
$470...
$1,600.
$884—
$1,135.
$1,727.
$2,727.

2H J u n e 15,1962/592H Dec. 15,1962/594% A u g . 15, I 9 6 0 — .
1H Oct. 1,1960
2H N o v . 15,1960--.
2% Dec. 15,1965/60A p r . 1,1961
May 15,1961—
A u g . 1,1961
4
Sept. 15,1961 . . .
2% Oct. 1,1961
1H N o v . 1 5 , 1 9 6 1 - . 21/2
F e b . 15,1962—_
F e b . 15,1962—.
A p r . 1,1962
4
M a y 15,1962—.
4
J u n e 15,1967/622H A u g . 1 5 , 1 9 6 2 —
4
Oct. 1, 1962-—
1H N o v . 15, 1962—
F e b . 15, 1 9 6 3 - . A p r . 1, 1 9 6 3 - . 2% M a y l 5 , 1963. _.
A u g . 15, 1 9 6 3 - 4
2H Oct. 1, 1963
N o v . 15, 1963—
Dec. 15, 1968/63.
4% F e b . 15, 1964--.
2H A p r . 1, 1964
3
M a y 15, 1964—
M a y 15, 1964_._
J u n e 15, 1969/64.
4% A u g . 15, 1964—
21/2 Oct. 1, 1964
5
N o v . 15, 1964
1H Dec. 15,
4% F e b . 15, 1965
2H M a r . 15, 1970/65.
A p r . 1,1965
M a y 15,1965—.
21/2 M a r . 15, 1971/66.
1H A u g . 15,1966—
5% J u n e 15, 1972/67.
2H Sept. 15, 1972/67
3
Dec. 15, 1972/672H M a y 1 5 , 1 9 6 8 —
2H Oct. 1, 1969
2H N o v . 15, 1974—
M a y 15, 1985/75.
4
J u n e 15, 1983/78.
3:
Feb. 15, 1980.-4 H M a y 15, 1985—
3J4 F e b . 15, 1990-.4
F e b . 15, 1995-.3J4
3;
3

m\
m

m

m

m

m

m

1

97-31
97-20
100-6
99-20
99-23
99-31
98-30
100-10
100-29
99-13
98-2
98-29
100-16
100-30
97-8
100-29
92100-28
96-10
100-17
97-22
95-12
100-30
96-27
94-14
103-23
90-2
97-27
93-14
103-22
100-6
89-12
104-23
92-14
104-13
88-30
95-20
88-20
91-4
103-20
87-16
96-18
87-8
8787-4
100-4
101-10
99-12
102-16
91-20
100-8
91-20
93-4
87-2

98-3
97-24
100-8
99-24
99-25
100-3
99-6
100-13
101-1
99-17
98-10
99-1
100-20
101-2
97-16
101-1
92-8
101-4
96-18
10 0-21
97-26
95-20
101-2
96-31
94-22
103-27
90-10
97-31
93-22
103-26
100-10
89-20
104-27
92-22
104-17
89-6
95-24
88-28
91-12
103-24
87-24
96-26
87-16
87-8
87-12
100-12
101-18
99-20
102-24
91-28
100-16
91-28
93-12
87-10

4
4
2
4
2
4
8
3
4
4
8
4
4
4
8
4
8
6
S
4
4
8
4
4
8
4
8
4
8
4
4
8
4
8
4

84 1
8
8
4
8
8
8
8
8
8
8
8
8
8
8
8
8
8

3.29
3.22
1.78
2.67
2.77
2.73
2.65
3.13
2.99
3.16
2.92
3.25
3.22
3.31
3.00
3.41
3.78
3.43
3.12
3.46
3.51
3.19
3.60
3.58
3.25
3.64
3.85
3.60
3.31
3.67
3.66
3.88
3.71
3.38
3.73
3.88
3.64
3.89
3.50
3.77
3.91
3.59
3.81
3.82
3.79
3.82
3.79
3.91
4.00
3.78
3.96
3.75

Parts of a price-quotation sheet published by one of the dealers on
July 13, I960, are presented in table V I I - 3 . Bid and asked prices
are given for U.S. securities arranged by maturity classes of under
1 year, 1 to under 5 years, 5 to under 10 years, and 10 years and
over. Column 6 contains the spreads between the bid and asked
prices measured in thirty-seconds of a point. For maturities under 1




112 STUDY OF FEDERAL GOVERNMENT SECURITIES DEALERS
year spreads vary between two and four thirty-seconds. Only
one spread amounts to eight thirty-seconds. This issue, the 1%'s
of April 1, 1961, is very small in comparison to the dollar
amounts of comparable issues outstanding at the same time.
For the longer maturities spreads become at least four thirty-seconds, and with increasing frequency, eight thirty-seconds. I n the 1
year to under 5 year maturity class, 13 spreads reach eight thirty-seconds. Eight of these issues have dollar amounts outstanding of less
than $600 million. This is a reflection of the thinness of the market in
these issues and is consistent with observations made in earlier chapters
(see chs. I l l and I V ) .
If two transactions, i.e., a purchase and sale, in a 4-year maturity
could be made on the same day for cash settlement at a spread of oneeighth of a point, the dealer would earn a gross profit of $1,250 on a
$1 million sale. One point on a $1,000 bond is $10. Certificates, notes,
and bonds are quoted in terms of prices, and points are expressed
normally in thirty-seconds, sometimes in sixty-fourths, and occasionally one-hundred-twenty-eighths. Prices and spreads in table V I I - 3
are outside (small lot, published) quotations. For purposes of inside
(interdealer or large lot) quotations spreads normally will be considerably narrower if the dealer wants to trade in any of the securities
on his list. The gross profit on the hypothetical transaction described
above is, therefore, likely to be higher than profits typically found in
trading. I t would be further reduced by the costs of carrying a position in the security. 2
The size of inside and realized spreads varies between individual
issues, within maturity classes and with changing money market and
economic conditions. At the same time, traders' expectations as to
changes in money market rates and, consequently, changes in bond,
note and certificate prices, determine the positions to be carried.
Purchasing into a position at the beginning of a decline in interest rates
will boost the spread realized on subsequent sales out of position.
During such periods, realized spreads may be larger than spreads
on inside bid and asked quotations. During periods of rising interest
rates the converse will often be true; realized spreads may be negative
and/or smaller than quoted spreads.
Treasury bills are quoted on a yield basis. Spreads between the
bid and asked yields rarely become larger than 6 to 8 basis points until
the bill approaches maturity. This means that a bill with 83 days to
maturity in July 1960 would be quoted at 2.46 percent bid and 2.42
percent asked. The spread in this case is 4 basis points. I t would represent a gross trading profit of $113.30 to a bill dealer who bought and
sold $1 million of the bills at the quoted prices during the same day
for cash delivery.
As Treasury bills approach maturity, the spread increases as the
prices of the security, expressed as yields, decrease. A bill with two
days to maturity acquired at a bid price of 1.70 percent, and sold at the
asked price of 1.40 percent, that is, with a 30 point spread, would bring
the dealer a gross profit of $10 on a $1 million transaction. This
gross profit would be consumed by the clearing charge of $10 per $1
million transaction which the clearing agent charges to the dealer.
a
F o r detail on interest received and interest paid out see the next section of this
chapter.




STUDY OF FEDERAL GOVERNMENT SECURITIES DEALERS 113
During the 11 years, spreads on long-term securities appear to have
widened mostly as consequence of the deterioration of the long-term
market and the unpegging of interest rates. During the late forties
a spread of one thirty-second on long term bonds is reported to have
been prevalent. Spreads on Treasury bills also seem to have widened
somewhat. However, 2-point spreads on bills still are encountered
rather frequently.
The widening of spreads is probably important in analyzing the
basic tendency of the long-term market to become thinner and less
supported by position-taking and arbitrage operations. A t the shorter
end, narrowing and widening of spreads indicate short-term changes
more than any long-term trends. The direction of changes in the
money market appears more important for the profitability of dealer
operations than changes in spreads between quoted prices. This last
point is supported by the following analysis of the relationship between trading profits, gross earnings, and transaction volume.
Judging on the basis of the data presented in table V I I - 2 , one
could conclude that widening of quoted spreads has caused trading
profits to gain in relative importance as a source of earnings.
From 1950 on, with the exception of 1955, trading profits increased as
a percentage of total earnings. This argument may be fallacious,
however, because it is based on the ratio of trading profits to gross
earnings which themselves are subject to wide fluctuations. A better
measure is presented in table V I I - 4 , where trading profits and other
earning and expense items are related to dollars of sales. Using this
measure, increases in profits which are proportional to increases in
sales are eliminated. Moreover, fluctuations in trading profits are
separated from other gross earnings to permit comparison of changes
in the composition of gross earnings per dollar of sales.
Four peak profit years stand out: 1949, 1953, 1957, and 1958. The
other years between 1949 and 1958 show considerably smaller ratios
of trading profit to sales or decreases from preceding peak ratios. I n
some part of the 4 peak years, there were recessions and in all of
them the Federal Eeserve System initiated or continued to pursue a
policy of monetary ease. Securities prices rose and dealers were stimulated to increase their commitments in the form of larger positions and
were able to sell at increased realized spreads.
Even in the recession of 1949, when the Federal Eeserve discount
rate was not lowered, prices of all maturities of Governments appreciated because of a policy of monetary ease. The effect proved
to be in the same direction as in 1957, when the discount rate was lowered and additional reserves were made available to the banking system
through open market operations. I n 1953, the Federal Eeserve began
to supply more reserves before the discount rate was lowered early in
1954. The resulting price appreciation of Governments began during
the second half of 1953;




114

STUDY OF FEDERAL GOVERNMENT SECURITIES DEALERS

TABLE VII-4.—Earnings and expense items in dollars per $1,000,000 of sales, all
reporting dealers, 1948-58
1948
Total trading profit
Interest received
Other earnings

1949

1950

1951

1952

1953

1954

1955

1957

1958

60.65 153.11 46.40 104.15 112.06 213. 99 152. 58 71.48 117. 22 !370.04j 353. 46
121.86 231.22 109.06 103.93 130. 32 152.48 161. 55 147. 76 151.07 196.04 165.16
6.33
7.76
9.14 11.47 13. 92 7.36
4.11 4.27 3.98
4.90
6.41

Gross earnings.... 186. 62 388. 60 159. 44 212. 98 250.14 372. 80 320. 54 228. 38 279. 76 580.00 525. 98
Interest paid
Salaries
Other current expenses.
Local taxes and statistical discrepancy
Net income before special charges and income taxes
Number of dealers reporting

78.94 172. 54
48.71 59.84
39.56 45.47
2.61

4.66

16.80 106.09
10

89.92 102.87 119.07 138.31 125. 97 137. 71 160. 86 261.00 186.20
34.12 44.30 45.78 50.33 44.23 46.13 51.20 60. 90 58.91
36.51 44.51 47.53 53.31 50.64 57.55 60.99 62.64 62.59
2.24

1.76
(2.87) 19.59
11

12

2.23

35.52 128. 62
12

13

1.62

1.61

1.95

3.481

4.73

(14. 62) 4.76 191.98 [213. 55
13

13

13

13

The poor trading profits of 1955 can be explained by the opposite
policy pursued by the Federal Reserve. The discount rate in New York
rose more steeply in this year than during any other calendar year between 1948 and 1958. I t climbed from iy2 percent early in the year
to 2y2 percent in December of 1955. By comparison, the rise in the
rate during 1950, another poor trading profit year, amounted to only
one-quarter of 1 percent. The period of late 1949 to early 1950, however, was the starting point for the substantial depreciation in security prices caused by tighter monetary policies and two hikes in
the discount rate.
The evidence suggests strongly that dealers' trading profits are most
heavily influenced by monetary policy changes. While they take advantage of periods of monetary ease by increasing their positions and
selling at wider realized spreads, the opposite response does not occur
during the beginning of monetary restraint. Short selling is only
rarely conducted in anticipation of downward movements of prices.
Eeasons for this were discussed in detail in earlier chapters. The pattern of trading profits is added evidence for the conclusions arrived
at there.
Changes in monetary policy appear to be mostly responsible for the
peaks and troughs in trading profitably. Other causes for the generally higher level of trading profits since 1953 must be discussed. (1)
The composition of trading volume by maturity classes may have
changed in a manner which influenced trading profits. (2) Wider
spreads may have been quoted and realized in recent years. F o r this
discussion it is helpful to drawT on the available information on
changes in the composition of trading volume and on trading profit
ratios of individual dealers. Table V I I - 5 presents these two measures
of trading profitability and shows high, low, and average ratios of
trading profit to gross earnings and to sales.
The average of the ratios for individual dealers tend to be higher
in most years because they give equal weight to small and large firms,




STUDY OF FEDERAL GOVERNMENT SECURITIES DEALERS 115
and extreme fluctuations in earnings occur in small firms more often
than in larger, more diversified dealerships. Individual dealer
margins vary substantially over time. Four dealers account for the
11 high ratios of trading profits to sales volume. Three of the same
four dealers also showed the lowest trading profit margins (or highest
loss margins) for 6 of the 11 years. The other five low figures are accounted for by one of the bank dealers. One dealer stands out by providing the highest profit margin in 4 years and the lowest margin
(or highest loss margin) in 4 other years.
During the period from 1948 to 1951, the highest profits and losses
seem to have been realized by those dealers operating mostly at the
long end. 1950, 1951, and 1958, all years with breaks or declines in
security prices, produced the largest losses for dealers in longer maturities. The high profit ratios for 3 of the last 4 years were reported
by firms trading and holding positions mostly at the short end of the
market. This same group of dealers was less affected by the reverses
of 1955 than were dealers w^ho operate more at the long end or over
the whole range of the market.
Dollar trading volume in Treasury bills has increased steadily over
the period with only two interruptions in 1955 and in 1958. Bill trading as a percentage of total transactions has fluctuated more over
the 11 years, but around an increasing trendline. The decline of the
share of certificate trading was compensated by increased bill trading
and by increases in relative volume of 1- to 5-year maturities. I t
appears that this changing composition of total trading volume helps
to explain the heavy impact of monetary policy on individual dealers,
especially on those operating at the long end of the market. However, in chapter IV, we found only two strong trends in the composition of trading volume, namely, increase in bill volume and decrease
in trading of maturities over 5 years. I t is doubtful that these changes
have had the effect of increasing realized spreads at the short end.
The profits realized from trading a larger volume of bills have not
been offset by losses which would have occurred on substantial positions
in long-term securities had these positions not declined. 3
One suggested explanation for a higher level of profit margins in
recent years is the widening of spreads between quoted bid and asked
prices. The evidence on this issue is scanty. I t appears doubtful,
however, that the dealers can increase their realized spreads by giving
outside and inside quotations with wider spreads. Pressure from
experienced and mostly large customers would soon enforce competitive pricing even if the dealers by some means attempted to maintain
a wider spread. Among the customers there is, of course, the Federal
Reserve Bank of New York which at any given time during the trading
day is well informed about all dealers' price runs.
3
One additional note of caution: the data go back only to 1948. The level of profits
during the years 1948-52 may appear lower than the average profit margins over the last
6 years only because the early years were not typical. Profit margins before 1948 and
before the Second World War may have been as high or higher than those of the last 6
years for which data are here available.




(73

d

T A B L E VII-5.—Total trading profit as percentage of gross earnings and in dollars per $1,000,000
1948
Aggregate t r a d i n g profit as percentage of
gross earnings
T r a d i n g profits as percentage of gross earnings, for all i n d i v i d u a l dealers:
High
Low
Average
Aggregate t r a d i n g profit i n dollars per
$1,000,000 of s a l e s . . .
Aggregate t r a d i n g profit p e r million as
m u l t i p l e of H2 ($312.50 per million dollars)
T r a d i n g profits i n dollars per $1,000,000 of
sales, for all i n d i v i d u a l dealers:
High
Low
Average

1949

1951

1950

1952

1953

of sales, all reporting dealers,

1954

1955

1956

1948-58

1957

1958

>
32.5

39.4

29.1

48.9

44.8

57.4

47.6

31.3

41.9

63.8

67.2

54.7
19.7
136.7

70.0
2.2
40.1

47.8
(46.2)
25.9

81.6
2.9
143.3

72.1
(7.4)
146.1

86.0
39.3
58.4

59.5
28.3
50.2

69.0
(11.3)
133.6

75.9
(31.7)
148.1

82.4
22.3
56.9

86.7
(60.1)
59.3

$60.65

$153.11

$46.40

$104.15

$112.06

$213.99

$152.58

$71.48

$117.22

$370.04

$353.46

.19

.49

.15

.33

.36

.68

.49

.23

.38

1.18

1.13

750.00
100.00
272.00

780.00
76.00
234.00

O

240.00
(460.00)
1 96.00

360.00
8.00
164.00

74.00
(386.00)
1 44.00

274.00
(1,774.00)
1110.00

492.00
(8.00)
U72.00

152.00
(52.00)
174.00

386.00
(54.00)
1 156.00

1,290.00
42.00
380.00

702.00
(940.00)
1 390.00

1 Firms with negative trading profits were omitted in the calculation of the average ratio. Firms with both negative gross earnings and trading losses were omitted in high, low
and average ratios.




b

o

<
W

O

STUDY OF FEDERAL GOVERNMENT SECURITIES DEALERS 117
The only widening of quoted and realized spreads appears to have
taken place at the long end of the market, especially for issues which
do not trade actively or for which the amount outstanding is relatively
small. Often quoted prices will show large spreads because the dealer
cannot increase his bid price without attracting securities which the
dealer does not want to position.
Moreover, it is at the long end of the market where some of the
dealers have increasingly acted as an agent for their customers. When
the dealer acting as agent does not charge a commission, his reward
is the difference between the price his customer is willing to pay and
the price at which he can obtain the securities by shopping around. If
a commission is charged, it amounts typically to one sixty-fourth of
a point. I t is conceivable that shopping around for a customer may
produce a wider realized spread for the dealer in the long run—but
mostly because he reduces the risk represented by a position in securities and avoids any holding cost. I n other words, a dealer can increase his realized spreads by acting as a broker and thereby avoiding
losses from price declines.
Thus, the substitution of bill positions for positions in very long
term securities and the transactions conducted on an agency rather
than a dealer basis may help to explain the widening of realized
spreads. The primary cause is the structure of the outstanding marketable debt and the resulting thinness of the long-term market which
makes positioning of long-term securities very risky. These factors
reinforced the effects of changing monetary policies on dealers' trading profits over the last 6 years.
I n table V I I - 5 , aggregate dealer trading profits also have been expressed as multiples of one thirty-second of a point. I n only 2 years,
1957 and 1958, trading profits amounted to more than one thirtysecond, that is $312.50 on a $1 million sale. I n 8 of the 11 years, trading profits amounted to less than one sixty-fourth. Translated into
percentage points, this means profit margins of .03125 and .0156 percent respectively. While the average margins over the last 6 years
are higher than for the first 5 years, they still are considerably smaller
than commission rates charged in other securities markets.
I n the over-the-counter securities markets, Friend et al. give information on commissions and gross profit margins 4 for a variety of
security transactions. F o r the period September through October
1949, they report gross profit margins on U.S. Government securities
traded by over-the-counter dealers of 0.1 percent, 0.4 percent on
State and municipal securities, 0.4 percent on corporate bonds, and
on corporate common stock 1.7 percent (all for transactions ranging
from $50,000 to $99,999). This compares with a generally accepted
figure for the average cost of transacting business with member firms
of the New York Stock Exchange of 3 percent. All these cost figures
are for purchases or sales only.
There are good reasons, of course, why one would expect the cost of
trading U.S. Government securities to be much lower than that of
trading in municipal and corporate securities. The risk of holding
a given dollar amount of Governments is much less than that of holding the same value in corporate securities. The average size of trans* Op. c i t , pp. 342r-409.




118

STUDY OF FEDERAL GOVERNMENT SECURITIES DEALERS

actions in Governments is much larger
actions in corporates and municipals,
handling and administrative charges.
flected in the marketing of Government

than the average size of transthus reducing the burden of
These considerations are resecurities.

I N T E R E S T RECEIVED A N D I N T E R E S T PAID

U.S. securities are not traded flat. Interest accrued during the time
a security has been held is paid in addition to the contracted price for
all maturity classes other than Treasury bills. This interest is shown
separately as a source of current operating income by all reporting
dealers. I t represents an important offset against the dealers' interest
expenses and should be viewed in conjunction with the cost of financing
positions which is reflected in these data by the item interest paid.
For 8 of the 11 years, the differential between interest received and
interest paid was positive and added to net income before taxes.
TABLE

VII-6.—Interest received and interest paid as percentages of gross
earnings, all reporting dealers, 1948^58

Interest received.
Interest paid
_
I n t e r e s t differential
I n t e r e s t as percentage of gross
earnings for all i n d i v i d u a l
dealers:
I n t e r e s t received:
High
Low
_
Average
Interest paid:
High
Low
Average

1948

1949

1950

1951

1952

1953

1954

1955

1956

1957

1958

65.3
42.3
23.0

59.5
44.4
15.1

68.4
56.4
12.0

48.8
48.3
.5

52.1
47.6
4.5

40.9
37.1
13.8

50.4
39.3
11.1

64.7
60.3
4.4

54.0 33.8
57.5 45.0
(3.5) (11.2)

31.4
35.4
(4.0)

80.3
45.3
60.9

95.6 93.9 97.1
29.1 44.0 18.4
59.0 172.0 155.3

98.8
27.9
55.7

60.7
12.9
40.6

66.4 92.9 71.5
25.4 21.8 24.1
47.6 164.1 148.5

77.7 75.2
17.6 12.0
41.2 138.9

88.0
20.7
46.5

97.9
28.7
51.6

81.4
22.4
40.5

90.7 251.2 157.1
17.6 18.5 21.4
41.9 88.2 2 73.8

99.6
18.8
46.9

56.3
23.7
2 38.5

96.3 94.0
36.0 22.5
67.2 2 55.8

2

81.2
15.6
39.3

i Firms with negative trading profits and/or negative gross earnings were omitted.
Firms with negative gross earnings were omitted.

2

Table V I I - 6 shows interest received and interest paid as percentages of total gross earnings. Row three gives the interest rate differential, showing the excess of received interest over interest paid. The
fact that this differential is positive during the period from 1948 to
1955 and negative thereafter, is another indication of the shift in
maturity composition of positions held by the majority of the dealers
and of the important changes in monetary policy during the period.
U p to 1956, the yield on long-term Governments stayed roughly 1




STUDY OF FEDERAL GOVERNMENT SECURITIES DEALE.RS

119

percent above the rediscount rate.5 Yields on 5- to 10-year bonds
maintained a smaller differential above, and bills and certificates
moved in close accordance with the rediscount rate. This traditional
set of relationships was changed during the period beginning in late
1955 and continuing through 1957. During this period, dealers held
more of their positions than previously in shorter maturities which
they had to finance at rates often well in excess of the rediscount rate.
This resulted in the negative "carry" expressed bjr the excess of interest
cost of carrying a position over the interest received on the securities
held.
Table VII-7 presents interest received and interest paid in dollars per $1 million sales. Data in this table eliminate the distorting influence of fluctuations in gross earnings. The interest differential, expressed in dollar margins, is of course still negative for the
last 3 years. Insight into the reason for this reversal can be gained
from consideration of individual dealers' ratios.
e
See comments on relevant rates of return on securities in dealers position later in this
chapter.

62471—60

9




fcO

o

O

a
T A B L E VII-7.—Interest received and interest paid in dollars per $1,000,000

I n t e r e s t received
Interest paid
___
I n t e r e s t differential
_ _
Aggregate interest in dollars p e r $1,000,000 of sales for all indiv i d u a l dealers:
I n t e r e s t received:
High
Low
Average,
Interest paid:
High
Low.
__
Average-_




of sales, all reporting dealers,

1948-58

>

1948

1949

1950

1951

1952

1953

1954

1955

1956

1957

1958

121.86
78.94
42.92

231. 22
172. 54
58.68

109.06
89.92
19.14

103.93
102. 87
1.06

130.32
119.07
11.25

152. 48
138.31
24.17

161. 55
125. 97
35.58

147. 76
137. 71
10.05

151.07
160. 86
(9. 79)

196. 04
261. 00
(64. 96)

165.16
186.20
(21. 04)

979.00
77.00
210.00

2,817.00
114.00
431.00

1,222.00
61.00
248. 00

935.00
29.00
179.00

324.00
95.00
169.00

395. 00
70.00
181. 00

748.00
47.00
246.00

441.00
44.00
195.00

473. 00
37.00
184.00

659. 00
39.00
268. 00

928. 00
51.00
284.00

486.00
50.00
136.00

1, 450. 00
152.00
328.00

738. 00
64.00
192.00

692. 00
54.00
176.00

374. 00
32.00
158.00

444. 00
48.00
194.00

492. 00
58.00
366.00

526. 00
44.00
220.00

674. 00
50.00
252.00

1, 716. 00
56.00
394.00

1, 020.00
50.00
302.00

O

o
<
w
g
GO

O

d

o
>
H

pa

w

STUDY OF FEDERAL GOVERNMENT SECURITIES DEALERS 121
The annual high, low, and average dollar interest margins among
individual dealers are also presented in table V I I - 7 . One firm trading in long-term securities accounts for the high interest-received
margins in 9 of the 11 years. The same firm accounted for most of
the annual high and low trading profit margins. Apparently this
dealer benefited from the higher yield of the long-term securities in
his position but lost as heavily on price depreciation as he gained from
price appreciation. The same reasons account for the relatively high
interest-received ratios of some of the other dealers who emphasize
trading at the long end.6
A number of factors in combination lead to variations over time and
between dealers in the interest-received dollar margin: (1) The higher
the proportion of long-term securities in a dealer's position, the higher
will be the amount of interest received. (2) F o r the dealers individually and as a group, interest receipts fluctuate with prevailing
money market rates and yields. This factor tends to act as an offset
against rising dealer trading profits during periods of monetary ease
and to increase their trading profits during periods of monetary
restraint. The effects of changing monetary policies on interest receipts will be greater when and where the dealers' holdings of longterm bonds amount to a large part of their position. (3) Since the
interest-received ratio is composed of interest as numerator and dollar
sales as denominator, it will be higher for firms with lower turnover
of positions, other things equal.
Most of the variations in firms' interest-received margins can be
explained on the basis of the above three factors. I n some cases, however, a firm trading at the long end shows a high turnover of position
which results in some atypical variations in interest margin. The
benefit of rising yields when money is tightened is normally weakened
by the reduction of dealer positions, especially of positions in longterms. However, shifts in the maturity composition of positions are
of minor importance for the dealers as a group though they help to
explain interdealer variations. Atypical interest received margins
for firms that would otherwise fall into the high or low categories
because of their transactions characteristics most often result from
exceptionally high or low position turnover rates.
The available evidence does not support so clearly a set of explanations of observed variations in dealers' interest-paid margins. I n
tables V I I - 6 and V I I - 7 these measures are presented for all dealers
who furnished complete reports, and again, high, low, and average
ratios for all individual dealers are given. One fact stands out clearly.
Of the five bank dealers only three report on interest paid. To them
this interest cost is based on an internally charged rate because they use
their own funds for financing of positions. I n most of the years this
rate appears to have been lower than the rates paid by the reporting
nonbank dealers. I n only 3 years, 1955, 1956, and 1958, are the interest-paid margins reported by any of the banks above the average for
all dealers. Bank dealers account for the lowest interest-paid margins
in 8 of the 11 years.
I n this context the question arises what the real interest cost is to
dealer banks who finance their positions with their own funds. Some
6
As noted earlier, the publicly quoted yields on U.S. securities are not the relevant rate
of return to the dealers. They do not hold instruments long enough to take advantage of
the capital gain to maturity.




122

STUDY OF FEDERAL GOVERNMENT SECURITIES DEALERS

of the nonbank dealers stated that the banks 5 cost was substantially
lower than that borne by nonbank dealers. This : argument implies
that the bank dealers could not employ the funds in some alternative
way, e.g., by selling Federal funds, investing or lending. If the
bank dealers use the rediscount privilege to carry a position without
foregoing other investment opportunities, the rediscount rate would
be the relevant cost of financing.
The nonbank dealers have been able to borrow from New York
banks at rates lower than the discount rate only on rare occasions.
They have met this competitive disadvantage by mobilizing funds
from commercial banks and nonfinancial institutions all over the
country—and often borrowed from these sources at rates below the
New York discount rate as we have seen. I t is also likely that the
bank dealers find funds more easily available than nonbank dealers.
The fact that bank dealers' internally charged interest rates appear
to be below the rates paid by the nonbank dealers does not necessarily reflect a real cost differential but results in an understatement of interest paid for the group of dealers whose reported data
are contained in tables V I I - 6 and V I I - 7 . If all interest received and
interest paid by the reporting bank dealers were excluded from these
figures, a negative interest differential would appear for the years
1951,1952, and 1955.
The interest-paid ratio, expressed in dollars per $1 million sales, is
influenced by a greater number of factors than the interest-received
ratio. I n connection with the analysis of variations in individual
dealers' ratios, the following general influences will be discussed: (1)
Changes in the level and structure of interest rates raise or lower interest payments necessary to finance a given position in securities. (2)
Interest rates charged by lenders may be influenced by size of firm,
performance of dealers, or by the degree of leverage employed in
financing a position. (3) The mix of financing methods utilized by
the dealers as a group or by individual dealers may raise or lower
interest expenses as we saw earlier. (4) The composition by maturity classes of positions financed by dealers over time has changed
in favor of more short-term securities and has been different at identical times for different individual dealers. This factor is important
only if it is more expensive to finance longer maturities than it is
to finance a position of short-term securities. (5) The treatment of
the costs of repurchase agreements is neither clear nor consistent.
Some dealers who consider repurchase agreements as loans would
include the cost of R P ' s in interest paid. Where E P ' s are treated
as outright sales and purchases of securities, their cost would be
reflected in a reduction of trading profits. (6) Finally, the relationship between sales volume and size of positions will be reflected in
the interest-paid margin. Again, as for interest-received ratios, firms
with relatively higher turnover should show lower interest paid
margins.
The most easily identifiable causes for changes in interest-paid ratios
are the level of interest rates and changes in monetary policy. The
trend in interest-paid ratios for all dealers has been upward—following the increase in the level of interest rates since 1949-50. There are
three interruptions of this upward trend; in 1950, 1954, and in 1958.
I n two of these periods, monetary policy changed from restraint to ease
to stave off a recession. I n 1950, no change occurred in the policy pur


STUDY OF FEDERAL GOVERNMENT SECURITIES DEALERS 123
sued; on the contrary, the discount rate was raised late in 1950. The
explanation lies in the drastic reduction of dealer positions during
the year and in the large increase in total transactions reported for
the year. As a result of these two factors, interest-paid ratios decreased from the previous year for all but one of the dealers.
I n 1954 and in 1958, the effects of lowered discount rates, open
market operations, and a substantial increase in reported sales combined to lower the interest-paid margin. I n these 2 years, the impact
was not as uniform as in 1950. During 1954, five of the dealers
showed higher margins of interest paid than in 1953, and three
had 1958 margins above those reported for 1957. Of the five firms
who behaved in this way during 1954, three had increased their
total positions; three had changed from a long position in 1953 to
a short position in bonds of over 5-year maturities. During 1958,
only one of the three exceptions among dealers increased total position, but all three of them added to their 1957 holdings of long-term
securities. A t the same time, their transactions increased or, in one
case, remained roughly at the 1957 level. Thus, the conclusion still
appears valid that changes in monetary policy and the general level of
interest rates mostly determined variations in interest-paid margins
over time.
Other factors are not as easily traced in their effects on interest
costs. I t is difficult to evaluate the influence of size of firm and dealer
performance. The available data indicate that the interest-paid ratios
tend to be higher for the smaller dealers and probably reflect the payment of brokerage fees to the money broker. Other explanations derive from a comparison of interest-paid margins with position turnover rates for individual dealers. Among the five dealers with the
highest interest-paid margin are four with the lowest position turnover ratios. The opposite is true for the five dealers with the lowest
interest-paid margins; four of them also show four of the five highest
turnover rates.
Another reason for deviation from the typical pattern of interestpaid ratios occurs as the result of financing a position in non-Governments. The cost of carrying the non-Government part of the position
will be reflected as interest expense. For the calculation of the turnover rate, only the position in U.S. Treasury securities was used.
This seems to apply to at least two dealers who have had sizable
transactions in municipals, corporates, and acceptances.
Dealers may differ among themselves in their efficiency of locating
funds and finding the cheapest financing mix on any given day. Their
very efforts in this direction, however, tend to spread the intended
impact of monetary policy, and, at the same time, tend to eliminate
differences in individual dealers' financing costs. Accordingly, interdealer variations in interest-paid ratios can be expected to be largest
around the times of changes in monetary policy and to disappear
largely with time and increasing effectiveness of monetary actions
taken.
The effect of maturity mix of positions again cannot be isolated because of a lack of relevant information. I t is likely, however, that
financing costs increase with the time to maturity of securities
financed. This effect is known in part. The higher margins often
required on collateral loans on other than short-term securities, for
instance, result in slightly higher effective borrowing costs. More



124 STUDY OF FEDERAL GOVERNMENT SECURITIES DEALERS
information on loan rates must be gathered before any further conclusions can be made with accuracy.
The same applies to the treatment of the cost of repurchase agreements. Where E P ' s are considered as ordinary sales and purchase
transactions, they increase sales volume rather than, or in addition to, increasing the amount of financing contracted during the
course of the years. The cost of E P ' s may be treated as interest cost
or may appear in dealer trading profits as two realized spreads. One
spread would be negative on a straight E P and would be the difference between sales price and repurchase price. The second spread
would arise as the positive or negative difference between repurchase
price and final sales price. Firms which treat E P ' s as sales and purchases and add the cost of some or all E P ' s to interest costs have
lower interest-paid margins than those firms charging interest on
E P ' s but omitting them from sales volume. Finally, dealers who
do not add cost of E P ' s to interest expense but increase their sales
volume by the amount of E P ' s , show even lower interest-paid
margins.
The effect of differences in position turnover rates on the interestpaid margin was discussed in connection with the analysis of firm
size and dealer performance earlier in this section. To summarize
here briefly: The higher the rate of turnover achieved with a given
position, the lower will be the interest-paid margin. Some distortions,
however, arise from the differences in reporting of positions on an
executed versus a commitment basis, and the treatment of E P ' s which
may or may not appear in a dealer's position. This and the other problems raised in this chapter will be important in future preparations
of consistent statistical series on the dealer market.
Valuable assistance in improving reports can be accomplished
through action by the Federal Eeserve Bank of New York in concert
with the accountants of the dealer firms and their auditors. Through
the cooperation of these two groups, minimum acceptable standards
can be achieved which will greatly facilitate the current collection
of information now underway at the New York bank.
OTHER EARNINGS AND PROFITS FROM I N V E S T M E N T

ACCOUNTS

Other earnings and profits from investment accounts amounted to a
relatively small part of gross earnings over the reporting period. The
following table summarizes the available information.




STUDY OF FEDERAL GOVERNMENT SECURITIES DEALERS 125
TABLE VII-8.—Other earnings and profits from investment accounts, all reporting
dealers, 1948S8

1948.
1949
1950
1951___
1952___
1953
1954...
1955
1956
1957
1958-..

_
„
_._
_

T o t a l , 11 years
1

_—
._

__ __

Other
earnings
( I n $1,000)

N u m b e r of
dealers
reporting

Profits from
investment
accounts
( I n $1,000)

N u m b e r of
dealers
reporting

(1)

(2)

(3)

(4)

333
298
445
432
671
606
732
937
1,223
1,442
982

3
6
6
5
8
6
6
6
7
6
6

8,101

0)
0)
368
(398)
556
(183)
897
(109)
(470)
11,200
4,542

1
1
2
2
5
4
5
5
5
4
2

16,666

Not available under the rules covering submission of the data.

The source of "other earnings" is not clear. The bulk of them are
reported by two of the larger nonbank dealers and by two bank
dealers. The figures have been added into gross earnings from current operations and are therefore reflected in net income before taxes.
Dealer firms reporting these figures have had some operations in addition to dealing in Treasury securities. Other earnings reported may
have resulted from such operations. They are small enough in relation to total gross earnings to omit them from further discussion.
The interesting fact about reported profits from investment accounts
is that four of the five firms reporting in any one year are partnerships.
Two explanations can be given for the apparently predominant use of
investment accounts by partnerships.
First, partners in high tax brackets list their share of the partnership's long-term capital gain on their individual income tax returns.
They may choose to have the gain taxed as ordinary income or have
it taxed separately at a maximum 25-percent rate. F o r partners in
marginal tax brackets higher than 50 percent, it pays to choose the
25-percent rate. If a partner is in an 80-percent tax bracket, this
method would save him approximately 55 percent of his income from
capital gains—the excess of tax under method 1 over the tax in
method 2.
A corporation has the same alternatives. The highest marginal
corporate tax bracket is roughly 52 percent. This would represent an
advantage of 27 percent of method 2 over method 1. To the stockholder of the corporation, however, another tax is unavoidable, if




126

STUDY OF FEDERAL GOVERNMENT SECURITIES

DEALERS

he wants to get the income from capital gains out of the firm. If
dividends are disbursed, they will become taxable to him at his personal income tax rate. I n order to take advantage of the more favorable capital gains tax rate, he would have to dispose of part or all of
his interest in the firm. The latter procedure is unlikely to be chosen
until retirement, and will depend even then on the marketability of
his share in the firm. If dividends are declared, the residual gain
from the investment account will be further reduced below the gain
he could have realized had he not been a stockholder but a partner
or individual subject to only one tax. Capital gains are therefore
more attractive to the partners of a firm than to the stockholders of
a corporation.
Second, the 1954 Internal Revenue Code provides that security
dealers have to separate clearly within 30 days after purchase any
securities held for investment purposes from securities held for trading purposes. I t is more difficult for principals in a partnership to
establish this separation than it is for individual stockholders who
purchase investment securities for their own account.
F o r both reasons, tax advantage and difficulty of separation of securities, it is to be expected that partnerships would show investment
accounts more often than corporate dealers. I t is likely, however, that
some of the principal stockholders at times have held investment
accounts in their own name which do not appear on the corporate
dealers' books and therefore are not in evidence in these data.
CURRENT OPERATING EXPENSES

Dealers reported on the following expense items separately: Salaries, interest paid, local taxes, and other current expenses. Interest
paid was discussed in detail in the last section of this chapter. Local
taxes were very small and will be ignored even though they have been
subtracted from income before income taxes. The following section
will deal with interdealer differences in salaries and other current
expenses charged to gross income.
SALARIES

Salaries paid by the dealers might be analyzed by breaking them
down into three categories. Wages and salaries paid to members of
the dealers' staff, salaries paid to the partners or stockholders of the
firm, and, finally, payments to principals in excess of average executive
salaries paid in the industry, and bonus payments. However, no such
breakdown is available.
Table V I I - 9 summarizes salaries in relation to gross earnings, and
expressed in dollars per $1 million sales. The ratio of salaries to
gross earnings fluctuates over the 11 years in accordance with the
rapid changes in gross earnings. The salary dollar margin is again
a measure which allows for more meaningful interdealer and over time
comparisons.
For the period from 1948 to 1958, salaries in relation to dollar
volume have remained remarkably stable. The ratio varies between
a low of $34.12 in 1950 and a high of $60.90 in 1957. I n these 2 years,
absolute dollar amounts paid in salaries by reporting dealers also
reach a low and a high, respectively.



STUDY OF FEDERAL GOVERNMENT SECURITIES DEALERS 127
TABLE VI1-9.—Salaries as percentage of ffross earnings and in dollars per
$1,000,000 of sales, all reporting dealers, 1948^58

Aggregate salaries as percentage of gross earnings
26.1 15.4 21.4
Salaries as percentage of
gross earnings for all individual dealers:
High
34.0 52.6 53.4
Low
8.4
8.8
6.0
Average
U6.6 24.1 23.2
Aggregate salaries in dollars
per $1,000,000 of sales
59.84 34.12|
Salaries in dollars per $1,000,000 of sales for all individual dealers:
High
336 620 248
Low
16
18
10
Average
82
110
58

1952

1953

1954

1955

1956

1957

1958

20. i

18.3

13.5

13.8

20.2

18.3

10.5

11.2

40.0
8.7
U9.2

33.8
6.3
15.2

31.6
2.5
13.2

74.6
5.5
26.5

85.7
74.6
2.9
5.5
26.5 124.0

46.13] 51.20| 60.90] 58.91

44.30| 45.78! 50.33!

476
12
84

146
16
60

160
20
64

36.8 30.7
1.1
2.8
12.8 i 13.5

170
14
58

1300
12
64

154
14
58

226
14
76

170
12
70

Firms with negative gross earnings were omitted.

Salaries show an upward trend, which is not surprising in view of
generally increasing wages and salaries in most areas of the economy
during the period under consideration. Furthermore, the volume of
business transacted in the dealer market has doubled during the 11
years. Part of the increase in salary payment has been caused by expansion of the dealers' work force. Average weekly earnings in banks
and trust companies, for instance, have increased by roughly 50 percent from 1948-49. If the dealer market expansion required an addition to work force of 50 percent of the 1948-49 employment, salaries
should have roughly doubled from 1948-49 to 1958.
The argument might be forwarded that, during adjoining years, salaries have fluctuated more than could be explained by a rising standard of living and expansion of dealer operations. Between 1953 and
1956, for instance, total salary payments had leveled off between $5
and $6 million (see table V I I - 1 ) . They jumped by almost $3 million
between 1956 and 1958—an increase of about 3 3 % percent over the
base year. P a r t of this increase is probably bonus payments.
If estimated bonus payments are compared to total net income for
the last 2 years, it is apparent that neither net income nor net profit
margins would increase substantially. Bates of return on invested
capital would be larger by at most 10 percent of the margin or rate
reported.
I n table VII-12, below, all salaries paid have been added back into
net income before taxes to show in the extreme what the effect on
profitability would be if all salaries were part of the profit. Such an
assumption of course overstates the profit figure by the total wage and
salary bill for clerks, secretaries, administrative help, traders, and
executives. But it does put an upper limit on the dealers' rate of
return.
The aggregate dollar payment of salaries has exceeded aggregate
net income before taxes in 6 of the 11 years. I t has been increasing
every year. This is an indication that the high reached in 1958 will
probably be maintained even in future low income years as long as the
level of dealer operations does not decline substantially. For individual dealers, dollar amounts of salaries paid and salary margins




128 STUDY OF FEDERAL GOVERNMENT SECURITIES DEALERS

do not display the same stability. Eight of the dealers report reduced
salary payments during years of low earnings or losses—1950-51 and
1955-56. Several dealers decreased salary payments in 1956 and 1958.
The salary margin decreases even more often for some of the dealers
because of fluctuations in sales volume. With a relatively constant
amount of salaries paid, any increase in sales volume will lower the
margin, any decrease in sales will increase the salaries paid to sales
margin.
One of the smaller nonbank dealers shows the highest of all annual
salary margins in 10 of the 11 years. Four other small dealer firms
account for the four next highest salary margins over the 11-year period. Bank dealers have paid the lowest salaries in relation to their
sales volume. Their salary margins have changed very little over the
period; they are either at the 1948 level or below. This is surprising in
view of the report that some large security dealers are considering
entrance into the U.S. securities markets as dealers but cannot do so
because of lack of trained personnel.
Finally, there is some evidence of economy of scale in the data for
salaries per dollar of transactions. Sales volume for some of the banks
has increased more rapidly than dollar amounts of salaries. And,
large dealers usually generate higher sales volume per dollar of wage
and salary payments.
OTHER CURRENT OPERATING EXPENSES AND SPECIAL CHARGES OR GAINS

Other current operating expenses reflect dealers' outlays on rent, office equipment and supplies, wire service fees, telephone charges, etc.
In table VII-10, these expenses are expressed as percentages of gross
earnings and in dollars per million dollars of sales. The ratios of
other expenses to gross earnings give an impression of the average percentage of earnings absorbed by items in this residual expense category.
Between 1948 and 1956, from 11 to 26 percent of gross income was
charged off as current expenses. The rapid increase in earnings during the last 2 years makes other current expenses appear deceptively
low. The margin of dollar current expenses in a million dollar sales
indicates a much more stable relationship between transactions volume
and other current expenses. Only in 1950 and 1955 was the increasing
trend of the expense margins for the group of dealers interrupted.




d
o
o
TABLE VII-10.—Other current expenses as percentage of gross earnings and in dollars per $1,000,000 of sales, all reporting dealers, 1948-58
1948
Aggregate other current expenses as percentage of gross earnings
Other current expenses as percentage of
gross earnings, for all individual dealers:
High
__
Low.
Average
Aggregate other current expenses in dollars
per $1,000,000 of sales
Other current expenses in dollars per
$1,000,000 of sales, for all individual
dealers:
High
Low
Average

21.2

1949
11.7

1950
22.9

1951

20.9

1952

19.0

1953

14.3

1954

15.8

1955
25.2

1956
21.8

1957

1958

10.8

11.9

31.4
11.1
U9.1

27.4
6.0
12.8

39.8
11.2
22.8

36.2
10.1
120.1

27.6
8.7
17.8

25.6
3.2
13.0

24.9
5.2
12.7

77.5
7.5
30.2

95.9
4.6
*25.2

22.6
2.3
11.9

$39. 56

$45.47

$36.51

$44.51

$47.53

$53.31

$50.64

$57.55

$60.99

$62.64

$62.59

288.00
16.00
66.00

430.00
20.00
80.00

332.00
10.00
64.00

602.00
14.00
80.00

146.00
16.00
64.00

160.00
18.00
56.00

158.00
18.00
56.00

310.00
22.00
70.00

172.00
20.00
60.00

206.00
20.00
68.00

154.00
20.00
66.00

1

47.8
3.2
13.8

>
O
O

<

w
1

Firms with negative gross earnings were omitted.




130

STUDY OF FEDERAL GOVERNMENT SECURITIES DEALERS

The variations from year to year for individual firms are minor.
Again, one dealer firm accounts for all of the individual dealers' annual high margins. The bank dealers, with one exception, appear to
allocate less of their overall operational current expenses to their
Government departments than most nonbank dealer firms actually
paid out in relation to sales volume. Firms with the largest sales
organization and the highest number of branch offices show most of
the higher annual margins. Smaller dealer firms, with one exception, account for lower margins than most of the large nonbank
dealers.
Special charges made to income and special gains added to income
before income taxes ; are presented in the following schedule. The
reporting dealers did not indicate what gave rise to these entries.
The net income figures used in the following section of this report
are adjusted to reflect these items. They are not important in relation to both gross and net income and will therefore be omitted from
further discussion.
Special charges 1 and gams, all reporting 6 ealerst 1948-58
[In thousands of dollars]

1948
1949
1950
1951
1

O I 1952
(645) 1953
69 1954
285 11955

(113) 1956
(232) 1957—
(317) 1958
75

110
(181)
212

Items in parentheses are not special charges.
SUMMARY

Only one of the three important components of total current operating expenses was strongly influenced by monetary policies and dealer
strategies in financing and positioning of securities, namely; interest
paid. Both salaries and other current expenses displayed a much
higher degree of stability. Where variations occur for the dealers
over time and between individual dealers, they are reflections of the
expansion of dealer activities and interdealer differences in trading
philosophy, methods of operations and size of firm. Where firms had
experienced adverse financial developments, effects on salaries showed
more strongly than effects on other current operating expenses.
The group of nonbank dealers with substantial operations in other
securities markets and the bank dealers had to allocate part of their
current expenses to their Government bond departments. On balance
and in comparison with the other nonbank dealers, this allocation
process seems to have resulted in an underallocation of diversified
firms' expenses relative to the expenses of dealers deriving more
than 90 percent of their gross earnings from operations in U.S. securities. These latter dealers were not asked to report a breakdown
of expenses related to operations in non-Government securities.




STUDY OF FEDERAL GOVERNMENT SECURITIES DEALERS 131
Table VII-11 presents selected total earnings and expense data for
the 10 diversified dealers. All items in the table are expressed as
percentages of gross earnings generated from all operations because
no volume figures for these dealers' total transactions are available.
From a comparison of table V I I - 1 1 with the data in tables V I I - 9 and
10 it becomes clear that salary and other current expense ratios for
the diversified group of dealers are higher and more stable than those
for all dealers reporting on earnings and expenses resulting from
trading in U.S. securities. This is further evidence in support of the
conclusion derived earlier from the analysis of variations in individual
dealer expense ratios. I t was found there that the bankdealers had
lower salary and other current expenses to sales margins and that they
charged to their bond departments less interest in relation to sales volume than most of the other dealers. The result of these relative
underallocations would be to overstate the net income figure for the
dealers as a group.
TABLE VII-11.—Selected earnings and expense items for diversified
dealers and all hank dealers, 1948-58

nonbank

[Items as percentages of gross earnings]
1948
I n t e r e s t received
_
T r a d i n g profits
Gross earnings
Salaries
Interest paid
O t h e r c u r r e n t expenses

1949

1950

1951

1952

1953

1954

1955

1956

1957

38.0 37.6 36.7 29.4 27.2 26.0 27.4 26.1 20.6 17.5
6.0
6.5
5.1
6.9
5.0
9.7
8.0
4.8
6.8
6.4
__ 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0
28.9 27.7 26.6 26.5 24.8 24.1 24.1 24.7 23.3 22.0
22.5 22.3 22.0 20.0 19.3 19.6 19.5 18.9 18.7 18.3
45.2 45.2 47.1 49.6 51.5 52.1 51.9 51.6 51.7 53.5

1958
21.5
9.4
100.0
22.5
20.5
51.6

I t has been pointed out that salaries paid might contain bonus payments to principals which could be added back to profits. If such a distribution of earnings is present it is small. Some evidence for this
conclusion is given by the income data for all incorporated securities
dealers in the "Statistics of Income." All incorporated securities
dealers spent 16.3 percent of their total compiled gross receipts during
the 11-year period on compensation of officers. This compares with
the salary ratios for U.S. securities dealers given in table V I I - 9 ,
which fluctuated between 26.1 percent of gross earnings in 1948, and
11.2 percent in 1958. I n 6 years the salary ratio for the Government
dealers exceeded the compensation of officers ratio for security dealer
corporations. I t must be kept in mind that the salaries paid by the
Government dealers contain all wages and salaries paid whereas all
incorporated dealers in the "Statistics of Income" reported on officers'
compensation only.




132

STUDY OF FEDERAL GOVERNMENT SECURITIES DEALERS
N E T INCOME BEFORE TAXES

Net income before taxes has been chosen as the basis for measuring
dealer profitability to avoid the distorting tax influences introduced
by the presence of both partnerships and corporations. Table V I I - 1 3
presents net income before special charges and income taxes as a percentage of gross earnings and in dollars per $1 million of sales. Fluctuations in the ratio of net income to expenses are paralleled roughly
by the changes in the ratio of dollar earnings per million dollar sales.
Both ratios show that 1950 and 1955 were years during which the reporting dealer group incurred losses.
Table V I I - 1 2 relates the net income figures of reporting nonbank
dealers to nonbank dealers' capital investment. Some more detail is
shown in this table, and the number of dealers showing losses is given.
1955 was a loss year for three of the bank dealers. F o r this reason, the
rate of return on capital of nonbank dealers in table V I I - 1 2 is positive in 1955, while all dealers combined (table VII-13) incurred an
aggregate loss. Table V I I - 1 2 also reflects the effect of special charges
and gains which have been omitted from table V I I - 1 3 .
Further analysis of the net income margins presented in table
VII-13 shows the strong impact of monetary policy changes on dealers' net income. The largest numbers of dealers reporting losses is
concentrated in years during which monetary policy changed from ease
to restraint, specifically 1950 and 1955. Next in number of dealer
losses are 1956 and 1951, years during which interest rates continued
to rise. On the other hand, the largest dealer profits appear in years
during which a policy of ease was introduced or pursued; 1953-54 and
1957-58 stand out as examples.




w

cl
O

o
T A B L E VII-12.—Gross and net earnings as percentages of net worth and total long-term
1948

1949

1950

1951

financing,

1952

1953

all reporting
1954

nonbank

1955

dealers,

1948-58

1956

1958

f
Ratio of gross earnings to:
Net worth.
_
Total long-term financing...
Ratio of salaries and net income before taxes to:
Net worth
Total long-term
financing
_
Ratio of net income before special charges or gains and taxes to:
Net worth
_
_
Total long-term financing
Median of the ratios of net income after special charges, before
income taxes, to total long-term financing for individual
dealers
_
Ratio of net income after special charges or gains before taxes to:
Net worth
Total long-term financing
Number of nonbank dealers reporting
_
Number of dealers reporting net losses (including dealer banks).




45.4
40.2

50.8
44.7

72.9
61.5

101.5
82.0

100.0
83.7

11.2

17.7
15.5

23.8
20.1

47.5
38.3

44.7
37.5

73.3
55.4
18.4
13.9

20.3
17.1

(.7)
(.6)

4.9
4.3

9.3
7.9

32.1
25.9

29.1
24.4

3.2
2.4

2.2

21.5

(12.0)

3.0

12.3

20.8

27.4

(3.4)

1.8
1.2
7
1

20.2
17.0

4.9
4.3

9.2
7.8
10
2

31.7
25.6
10
0

28.8
24.2
10
0

3.1
2.3
10
10

51.0
34.5

77.1
65.0

16.9
11.5
1.7
1.1

10
5

Q
O
<J

104.0
72.9

193.7
144.4

189.2
143.9

21.0
14.7

78.8
58.7
57.6
42.9

99.5
75.7
77.1
58.6

29.2

46.2

CD

56.9
42.4
10
3

76.4
58.1
10
1

Q
Cj

(9.9)

W

3

fei
GO
O

H
02

o
T A B L E V I I - 1 3 - -Net inccme before special charges and taxes as percentage of gross earnings and in dollars per $1,000,000
dealers, 1948-58

of sales, all

reporting
H
ft)

1948
Aggregate income as percentage of gross earnings
Incomes as percentages of gross earnings, for all individual
dealers:
High...
Low...
_
_
_.
Average
Aggregate income in dollars per $1,000,000 of sales
Aggregate income before taxes per $1,000,000 of sales as multiples
of # 2 ($312.50 per $1,000,000)
Incomes in dollars per $1,000,000 of sales, for all individual
dealers:
High
_
Low
__
__
Average

1949

1950

1951

1952

1953

(1.8)

1954
30.6

1955

1956

(6.4)

1957
33.1

1958
40.6

O

o

<
59.0
1
(1.0)
2 21.7
$16.80

71.9
(24.3)
2 35.7
$106.09

0.05

0.34

$102
($474)

$4,388
($92)

68.2
(59.1)
2 21.3
($2. 87)

70.0
i (30. 7)
2 22.3
$19. 59

74.8
(37.1)
2 26.2
$35. 52

75.3
9.3
35.7
$128. 62

1.5
35.8
$98.08

0.11

0.41

0.31

$278

$452
$50
$172

$708
$8
$170

0.06

$94
($494)

$136
($2, 528)

79.4

63.4
70.9
(238.1) » (238.8)
2 34.7
2 35.3
($14. 62)
$4.76

($616)

84.8
(3.5)
2 42.0
$191.98

0.02

0.61

$164
($426)

$588
($30)

81.8
U0.9
2 37.8
$213. 55
0.68

$522
($912)

O

d
i Some firms incurred negative gross and net earnings and were omitted from high, low, and average ratios.
> Loss ratios omitted in the calculation of average ratios.




02

©
M

STUDY OF FEDERAL GOVERNMENT SECURITIES DEALERS 135
The variation in individual dealers' net income margins is considerable in most years, as indicated by the range between high and
low margins in table VII-13. In 9 of the 11 years, the low margin is
provided by dealers incurring losses. Total net income for the 11-year
period was related to their total sales volume over the same time for
each dealer. Analysis of this individual dealer average profit margin
shows that among the six firms with the highest average net income
margin for the entire period were two bank dealers, one small and
three large nonbank dealers.
The lowest average profit margin reflects an overall loss for one of
the dealers during the 11 years. Of the six firms with the lowest
average individual profit margins were three small and one larger nonbank dealers, and two of the bank dealers. Thus, there does not appear
to be a definite correlation between size of firm and the magnitude of
profit margins over the entire period.
The profit margins discussed here are a measure of relative efficiency
of dealer operations. Differences between individual dealer's margin
during any given year can be traced back to the contribution of interest received and trading profits, and the charges represented by current expense items. If suffices, however, to state that most of the
interdealer differences are caused by differences in their trading profits
and interest received—interest paid ratio. All of these are influenced
most heavily by changes in monetary policy, but also by dealer efficiency in arranging for relatively low cost financing, turnover rates of
positions, composition of position by maturity classes, and, conceivably
by discrepancies in dealer accounting procedures.
NET INCOME BEFORE TAXES AND DEALER CAPITAL

Net income is related to dealers' capital to compare dealer earnings
with the profitability of other lines of business. I n table V I I - 1 2
dealers' gross and net earnings are expressed as percentages of their
net worth and of their total long-term financing. These two terms
were defined and discussed in chapter V I . I t was suggested there that
for the purpose of calculating rates of return on capital, the total longterm financing concept should be used. Both measures of rate of
return have been presented in table VII-12, but the rest of the discussion in this chapter will be based on the rate of return on total longterm financing or "capital base" as it is frequently called.
The number of nonbank dealers 7 for whom information on both net
income and capital was available appears at the bottom of table
VII-12. I n one case no allocation of expenses to operations in U.S.
securities was provided; the second omitted nonbank dealer had such
sizable operations in other securities markets that inclusion of his net
income from operations in U.S. securities and of his total capital would
have resulted in a considerable understatement of dealers' rates of
return.
7
The scope of the information used in table VII-12 is narrower than that of the data
contained in table; VII-13 for a number of reasons. The bank dealers had to be excluded
from the rate of return calculation because only rough estimates are available of the
amount of the banks' total capital which could be considered as committed to their
operations in U.S. securities.
62471—60
10




136

STUDY OF FEDERAL GOVERNMENT SECURITIES DEALERS

Four smaller dealers also reported income from operations in other
markets. During most of the years, net income from trading in U.S.
securities provided the majority of these firms' net income. All of
their capital has been included while only their net income derived
from operations in Governments appears in the aggregate net income
used as numerator in table VII-12. The following schedule reflects
all earnings of these four firms.
Schedule of the ratios of net income after special charges and gains, to net worth
and total long-term financing, before income taxes, all reporting dealers,
1948-58
Net worth

1948
1949
1950
1951
1952
1953
1954

2.0
19.8
(1.9)
5.3
9.4
31.9
28.8

Total
long-term
financing
1.3
16.7
(1.7)
4.6
8.0
25.8
24.2

Net worth

1955
1956
1957.. . .
1958
Average of annual
ratios:
48 to 57
48 to 58

_
-.

Total
long-term
financing

5.4
2.2
58.7
77.0

4.1
1.6
43.8
58.6

16.2
21.7

12.8
17.0

These adjustments result in ratios only slightly different from the
ratios in table VII-12. They are higher than the ratios in table
V I I - 1 2 in 6 years, lower in 4 years, and are exactly equal in 1 year.
The averages presented in the schedule are the averages of the annual
ratios. They are somewhat lower than the ratios of total income to
total capital over the period. But on the basis of this evidence, the
method chosen in compiling table V I I - 1 2 appears acceptable.
All ratios in table V I I - 1 2 reflect the predominant influence of
changes in monetary policy on dealer profitability. Gross earnings
as well as net income ratios show large decreases during years of
monetary restraint and large increases as results of changes from
monetary restraint to monetary ease. This is also reflected in the
median of individual dealers' ratios of net income to total long-term
financing given for each year. I n 3 years, 1950, 1955, and 1956, the
median is a loss ratio. Individual dealers' annual ratios of net income
to capital base show considerable variations in poor profit years as
well as in high profit years.
F o r the 11-year period, one dealer's rate of return was negative.
The 16 dealers with a positive return on capital base showed average
rates of return for the entire period ranging from a low rate of about
3 percent to a high rate of about 28 percent. Three dealers realized
average rates of return on capital base above 20 percent. Larger
dealers were more profitable on balance than smaller dealers, indicating a more efficient use of capital resources as well as the beneficial
effect of large and diversified operations which protect the larger firms
to some extent from the adverse influences of monetary policy changes.
The ratio of net income plus salaries to capital base, given in table
V I I - 1 2 , merits some further explanation. I t provides an upper limit
on the profitability of the dealers included in this table if all salaries
paid were actually part of the dealers' profit. This assumption is, of
course, unrealistic. As stated earlier, salaries paid by the U.S. Gov-




STUDY OF FEDERAL GOVERNMENT SECURITIES DEALERS 137

ernment securities dealers include all wages paid to their staffs and
their executives. Only if bonuses were paid in some relation to net
income could they be considered a part of the firms' profits.
All incorporated securities dealers included in the data provided by
Statistics of Income paid an average of 7.7 percent of their total
long-term investment over the period 1948-58 to their officers as compensation. This does not include payments to their other employees.
Salaries paid by the U.S. securities dealers ranged from 10.7 to 17.6
percent of their capital base and averaged aoout 13 percent. If
the assumption is valid that officers of the Government bond houses
received compensations comparable to the officers of incorporated securities dealers, this would leave about 6 percent of their capital base
for the rest of their payrolls during an average year.
The ratios of net income after special charges and gains, but before
income taxes, to net worth and capital base, from table VII-12, also
appear at the top of table VII-14. They are presented together with
selected rates of return for other industry groups. Toward the bottom of table VII-14 rates of return for three other types of financial
institutions appear: (1) all insured commercial banks, (2) all incorporated securities dealers, and (3) financial corporations, including
commercial banks, real estate firms, and insurance companies. The
ratios available for all insured commercial banks were for net income
before taxes. They have been adjusted on the assumption that the
average tax rate applicable has been 40 percent.




T A B L E VII-14.—Rates of return for selected industry

1948

Ratio of net income before taxes, reporting nonbank dealers, to net
worth
Total long-term financing
- _ __
Ratio of net income before taxes to total equity, selected industry
groups: *
1. Apparel
2. Textile,
_
___
3. Leather
_
4. Food_
_
5. Lumber
_
6. Scientific instruments
__
7. Ordnance
8. Chemical
_
_
9. Electrical equipment and machinery
10. Motor vehicles
11. Incorporated securities dealers
12. Financial corporations (including insurance and real estate)..
Ratio of net income
after taxes to total capital funds, all insured commercial banks 8
_
E s t i m a t e d before t a x (40 p e r c e n t ) r a t i o

i Source: "Statistics of Income, 1948-57."
* Source: Moody's.




1949

1950

1951

1952

groups,

1953

1948-58

1954

1955

1956

1957

1.8
1.2

20.2
17.0

(0.9)
(.8)

4.9
4.3

9.2
7.8

31.7
25.6

28.8
24.2

3.1
2.3

0.3
.2

56.9
42.4

16.0
27.0
14.0
16.1
26.6
21.2
18.4
21.8
24.5
30.3
2.2
13.0

8.9
11.6
9.5
15.0
12.7
15.9
8.5
19.6
17.9
32.5
3.8
13.3

15.3
20.1
16.4
17.6
26.7
23.5
27.4
33.0
32.1
49.1
6.3
14.6

8.6
15.2
11.1
14.5
20.4
28.4
30.6
29.0
31.1
36.4
4.4
14.3

9.1
8.3
12.0
13.8
13.8
25.3
28.9
21.6
29.7
32.0
6.2
13.7

8.5
8.1
11.4
15.3
11.5
25.1
34.0
22.2
27.8
28.8
5.9
13.8

7.7
5.1
11.5
14.3
12.3
24.2
26.4
20.1
20.9
27.2
11.0
14.5

10.2
8.9
15.0
16.6
17.2
23.4
21.5
26.8
20.5
46.1
8.3
13.6

10.9
9.1
12.8
15.7
11.1
23.5
18.5
22.7
19.7
24.8
8.1
12.3

9.0
7.4
12.7
14.6
6.3
20.0
19.2
21.2
21.6
31.7
7.4
11.6

7.5
12.5

8.0
13.3

8.5
14.2

7.8
13.0

8.1
13.4

7.9
13.2

9.5
15.8

7.9
13.2

7.8
13.0

8.3
13.8

1958

76.4
58.1

A v e r - Average,
age,
1948-57 1948-58

17.1
13.6

24.2
19.1

10.4
12.0
12.6
15.3
15.8
23.0
23.3
23.8
24.5
33.8
6.4
13.4
9.6
16.0

8.1
13.5

8.3
13.7

STUDY OF FEDERAL GOVERNMENT SECURITIES DEALERS 139
The last two columns in table V I I - 1 4 give the average rates of
return for the entire period for which data were available. During
the period from 1948 to 1957, the U.S. securities dealers realized approximately the same rate of return as the insured commercial banks
and the financial corporations. They showed better returns on capital
than the incorporated securities dealers even though in 5 of the 10
years the U.S. securities dealers had lower rates of return. During
the same 10-year period, the dealers in Governments had a higher
average rate of return on capital base than three of the manufacturing
groups listed in table V I I - 1 4 ; i.e., apparel, textile, and leather.
The manufacturing groups chosen are those with the five lowest and
five highest rates of return on total equity (including preferred stock
and all surplus accounts in addition to common stock). The dealers
thus achieved an average rate of return on their capital base between
1948 and 1957, which was higher than that realized by the three manufacturing industries with the lowest rates of return of all manufacturing industries. Two of five manufacturing industries with the lowest
rates of return had higher rates than the dealers.
This picture changes with the inclusion of 1958. Data for this year
raise the average rate of return for the dealers over the 11-year period
substantially. Only the top five manufacturing industries realized
higher profit rates between 1948 and 1957 than the dealers between
1948 and 1958. Over the 11-year period the dealers also realized
higher rates of return than the other financial institutions, especially
higher than the rates shown for the incorporated securities dealers.
None of the other industry groups, however, show the drastic fluctuations in rates of return which are characteristic of the dealer market.
I t is also interesting to observe that the cyclical fluctuations in the
rates of the manufacturing corporations reach peaks when the rates of
the U.S. securities dealers are at a low point. The automobile industry, for instance, had its best profits years in 1950 and 1955. Both
years show losses or very low profits for the dealers. This again
supports the conclusion reached earlier, that dealer profitability is
determined mostly by changes and direction of changes in monetary
policy. They stand to gain from the relaxing of monetary controls
during recessions. They realize their lowest profits and highest losses
in the late stages of a business boom when monetary controls tighten.




140

STUDY OF FEDERAL GOVERNMENT SECURITIES DEALERS
FINDINGS

The factors influencing dealer profitability have been analyzed and
discussed. Changes in monetary policy have been found to be the
most important causes of fluctuations in dealer profits. One basic
question still remains; namely, is the rate of return on dealers' capital
competitive enough to keep capital in this market or to attract new
funds into it ? For the years 1948-56, this question would probably
have been answered negatively. The dealers had experienced five
poor profit or loss years over the period. The 3 profitable years had
not allowed sufficient net income to bring the dealers' average rate
of return even close to those realized by commercial banks and all
financial corporations. I t was competitive, at the 9-year average
of 9 percent, only with the rate of return on the capital funds of
incorporated securities dealers, most of whom are small firms specializing in corporate securities.
The increased average profitability of the U.S. securities dealers
began in 1957. The recession of that year, appropriate monetary
policy, and new cash financing by the Treasury combined to increase
dealer activity and dealer earnings during 1957-58. Some dealers
took advantage of the change in monetary policy by purchasing securities into accounts more heavily. All dealers profited to some extent
by the unprecedented sharp drop in money market rates from the high
points in late 1957 to the lows of 1958. These changes produced
trading profits much higher than those realized by the dealers in
1953. The performance of 2 years improved the dealers' rate of return
for the entire 11-year period. If this level of profitability should or
could be maintained, other firms and new capital will probably enter
the market. There is, however, no assurance that the level of earnings
will stay as high in future years, and potential investors will look at
dealers' adverse experiences in earlier years as well as at their high
profits in the last 2 years.




CHAPTER VIII
OPERATION OF THE DEALER MARKET
The analysis of the data submitted by the 17 Government securities
dealers and the description of dealer operations is completed, but
the problem of evaluating the performance of this market remains.
That problem can be subdivided into four questions.
First, does the market fulfill its obligations to the public at large,
and to holders of the Government debt, in a responsible and efficient
manner and one consistent with the public interest? To assist in
arriving at a consensus about operations within the dealer market,
the major findings of this study are summarized below and some
questions are raised about possible changes in operating procedures
which may lead to an improvement in the manner in which the
market functions.
However, as chapter I suggests, a study of the Government securities market should be viewed in terms of the interaction of this
market with the economy. F o r as noted earlier, the market for
Government securities is a particularly important source of information about the working of monetary policy and debt management
and is a major influence in the saving-investment process. The
second, third, and fourth questions are therefore posed within the
broader context of the interrelationship between the dealer market
and economic stabilization policies.
Can Treasury debt management practices be altered to eliminate
some of the observed weaknesses in the dealer market? Can the
effectiveness of monetary policy be increased by strengthening the
intermediate and long-term market for Government securities?
Would such changes improve present policies designed to promote
economic stability ?
These broader questions are not considered here. But it is likely
that fruitful suggestions for improving the functioning of the market can be obtained by examining debt management practices and
monetary policies. Some possibilities along these lines were noted
in chapter I I I and in the introduction to the study.
SUMMARY OF THE MAJOR FINDINGS

The market for U.S. Government securities is composed of 17
over-the-counter dealers. Measured by the value of total purchases
or sales, it is the largest security market in the country. I n 1958
alone, the volume of transactions in the dealer market was more than
five times larger than the total trading volume on the New York
Stock Exchange. The major part of this volume was handled by
6 of the 17 firms during most of the period covered by this study.
However, there has been no apparent tendency for concentration,
measured by share of market, to increase.




141

142

STUDY OF FEDERAL GOVERNMENT SECURITIES DEALERS

Aside from their common activity—providing a market for the
Government's marketable debt—differences between dealer firms are
more notable than similarities. Five of the dealer firms are banks;
the others range from multidepartment securities houses to small
specialists in Government issues. Even within the Government
securities market, differences between dealers are readily observed.
Some dealers tend to specialize in particular maturities. Some are
quite active in the intermediate-term market for notes and bonds;
a few appear to engage extensively in the long-term bond market.
But almost all dealers actively trade in the short-term market and
bid in the weekly Treasury bill auctions.
Dealers make markets by quoting firm prices or spreads at which
they are willing to buy or sell. But small dealers often concentrate
on interdealer trading and transactions with the trading desk at the
Federal Eeserve Bank of New York. Large dealers are both relatively and absolutely more active in trading with institutional investors and other holders of the publicly owned marketable debt.
I n addition to size of firm, several suggested explanations related to
the internal operations of the firm were advanced to account for differences in dealer positions and composition of trading volume.
Differences in dealer philosophy and attitudes toward risk were noted
in several sections. The latter point is particularly important in
view of the increase in the risk of holding a given dollar amount of
securities as maturity increases and in view of the relatively high
leverage ratios of most dealer firms. Other proposed explanations
of differences between individual dealers were based on factors external
to the firms. Principal among these were those difficulties which
reflect the problems of debt management and economic stabilization
policy.
I n chapter V I , it was estimated that dealers had committed $100
million in aggregate capital to the market by 1958. Capital increased
over the 11-year period of observation despite temporary reductions
due to losses and retirements. But capital per dollar of transactions
or per dollar of net position declined.
A principal reason for the relative decline in capital was discussed
in chapter V I I where the earnings and expenses of dealer firms were
presented. For the years from 1948 to 1956, the rate of return on
dealers capital was low relative to the amount of risk involved and
was less than the average rate of return earned by commercial banks
and all nonfinancial corporations. As a result, retained earnings were
relatively low and new capital was not attracted.
The profitability of the dealer market increased in 1957 and 1958
with the advent of recession. As in previous postwar recessions, realized spreads increased as interest rates rose. Some dealers purchased
securities for their investment accounts when monetary policy changed.
The earnings for these 2 years (1957-58) increased the dealer's rate
of return for the period 1948-58 to the level of some of the more profitable manufacturing industries and above the level of all commercial
banks. However, in the light of the evidence of chapter V I I , it is
likely that the combination of relatively low rate of return, high risk,
and small profit per dollar of sales in most years discourages new firms
from entering the Government securities business.




STUDY OF FEDERAL GOVERNMENT SECURITIES DEALERS 143
The search for borrowing rates below the interest rate earned on their
commitments is a critical part of any effort by dealers to increase their
profitability. During tight-money periods, some dealers are particularly adept at finding temporarily idle balances available at rates
lower than those charged in New York. The regrowth of repurchase
agreements has been one of the principal means by which this has been
accomplished. I n the process dealers bring idle balances to the money
market and spread the effects of monetary policy to other parts of
the country.
Nevertheless, the available information suggests that financing commitments remain the primary problem of dealer operations. Improvements in this area which do not result in larger increases in the money
supply than would otherwise occur would speed the adjustment of
interest rates to new economic conditions and facilitate the positioning
of securities in dealer inventories. Smaller dealers in particular
would probably benefit from the establishment of a lender of last
resort willing to lend at market rates of interest. Additional study
of this subject and the release of the report by the clearinghouse banks
would make possible the development of improved financing for this
market.
Furthermore, investigation of arrangements to improve opportunities to borrow securities for short sales should be made. And changes
can be made to prevent rapid falls in the prices of securities which
occur as a result of forced selling and margin calls. Higher margin
requirements on repurchase agreements to dealers and investigation
of the effective margin on collateral loans and repurchase agreements
to nondealers were suggested in chapter V as possible means of
strengthening the market.
I t is difficult to summarize the overall performance of the market
succinctly since changing economic conditions make any conclusion
about operating procedures inapplicable at particular times. Interest rates, the present and expected level of economic activity, the
availability of bank reserves, the advent of new Treasury offerings,
and a host of other factors must be considered in addition to those
mentioned above. However some general conclusions can be drawn
with respect to the various maturity classes.
At the short end of the market, the evidence strongly suggests that
active competition prevails. Securities turn over rapidly at narrow
price spreads. Dealer trading profits per dollar of sales are small.
Many informed buyers and sellers take positions on either side of the
market as prices change throughout the day. A sudden fall in the
price of short-term securities quickly brings new buyers into the
market; a rise in the price attracts new sellers. Dealer operations
permit this active trading to take place. They are willing to hold
large inventories and to execute transactions of $50 million or more
for a single customer at their own risk. I n short, the dealers are at
the center of a highly competitive, efficient trading market.
I n the 1 to 5 year maturity range, the performance of the market
continues to be good. A large volume of outstanding securities is
available. Many of the dealers actively trade and are willing to take
positions, long or short, to accommodate the requirements of buyers or
sellers. Spreads remain relatively small and competitive conditions
prevail.




144

STUDY OF FEDERAL GOVERNMENT SECURITIES DEALERS

The further one moves through the maturity range toward the long
end, the lower the probability of finding the conditions just described.
Fewer dealers operate in the longer term markets. Those that do
often feel that they must act as brokers rather than dealers. A t times,
only one or two large dealers are willing to take the risk of quoting
prices and making markets. The size of orders which can be executed
at quoted market prices becomes small. Quoted price spreads widen;
activity declines; the efficiency of market operations falls; price discontinuities are common.
The basic weaknesses of the longer term market are principally reflections of the debt management problem and do not appear to be
capable of correction solely through the efforts of the dealers. F o r
example, in chapters I I I and IV, the relationship between dealer
positions and transactions on the one hand and debt management on
the other was noted. The available data there suggest that the small
size of many long-term issues and the frequency of Treasury refundings reduce the size of dealer positions and the marketability of these
Government securities. Other Treasury practices, e.g., maintenance
of extremely small average cash balances, probably lead the Treasury
to be unduly concerned with the amount of "attrition" (exchanges for
cash) when it is refunding debt. Heavy attrition on a particular refunding and a low cash balance requires the Treasury to return to the
market with an additional offering at an early date.
Solution of the major problems of the dealer market would require
going beyond the question of the performance of the dealers. As
noted in the introduction, it is difficult to study the dealer market
separate from the performance of debt management and monetary
policy. F o r where the marketing of Government bonds is handled
in a well organized manner, the secondary market performs well.
But when the marketing process for Government securities encounters
difficulties, secondary trading of existing Government securities becomes less efficient.




o