Full text of Federal Agency Debt and Its Secondary Market : Staff Study

View original document

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

Treasury-Federal Reserve Study of the
U.S. Governinent Securities Market

FEDERAL AGENCY DEBT AND ITS SECONDARY MARKET

Staff Study prepared by
Janice Peskin
Economist, Board of Governors
November lU, 196?

THE
FEDERAL
RESERVE
BANK of
SE LOUIS

Research Library

TABLE OF CONTENTS

-iPage Number

TEXT
Chapter I.

Introduction and Summary of Findings

Chapter II. Characteristics of Federal Agency Debt
A.
B.
C.
D.
Chapter III.

Chapter IV.

Supply
Maturity Structure and Other Characteristics . . .
Yields and Yield Spreads
Demand

1
10
10
17
24
32

Homogeneity of Agency Securities

Homogeneity by Issuing Agency

Homogeneity by Issue Size

Indicators of Market Performance .

A.
B.
C.

The Volume of Trading
Dealers' Positions
Spreads Letvjsen Quoted Bid and Asked Security
Prices

86
100
113

CHARTS
Text
1. Profile of Total Agency Debt Outstanding
2. Profile of Non-guaranteed Agency Debt By Type
3. Profile of Maturity Structure of Non-guaranteed Agency Debt .
4. Profile of Market Yields
. . .
5. Profile of Yield Spreads
6. Yields and Yield Spreads for Various Maturities . .
7. & 7a. Profile of Ownership Structure of Non-guaranteed Agency
Debt • • « . . . . • . * . . . . . . . . . . • . » ; . » «
8. Commercial Bank Holdings of Non-guaranteed Agency Debt,
1954-67 • • . . . . . . . . . • . . . . . . • . . • . • .
9. Yields of Agency Securities, May 31, 1967
10. Yields of Agency Securities, May 31, 1966
11. Yields of Agency Securities, May 1963, 1964 and 1965
12. Yields of Agency Securities, May 31, 1962
13. Yields of Agency Securities, May 31, 1961
14. Yields of Treasury Securities, May 29, 1967
15. Yields of Agency Securities By Issue Size, May 31, 1967 . . .

12
14
19
25
26
31
35 Sc 36
43
65
66
67
68
69
71
81

-iiTABLE OF CONTENTS (cont.)
Page Number
CHARTS
Text
16.
17.
18.

Profile of Market Performance for Federal Agency
Securities Maturing Within 1 Year
Profile of Market Performance for Federal Agency Securities
Maturing After 1 Year
Transactions in Agency Securities and Causal Variables . . .

87
88
92

Appendix
1.

Differences in Yields Quoted by Morgan Guaranty Trust Co.
and First National City Bank

134

TABLES
Text
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.

Ownership of Agency and U.S. Debt by Type of Commercial
Bank
Ownership of Agency and U.S. Debt by Commercial Banks . . . .
Ownership of Agency and U.S. Debt by Nonbank Financial
Institutions
Ownership of Agency and U.S. Debt by Nonfinancial Corporations
Ownership of Agency and U.S. Debt by State and Local
Governments
Ownership of Agency and U.S. Debt by All Other Investors . .
Ownership of Agency and U.S. Debt by Official Accounts . . .
Ownership of FNMA Participation Certificates
Ownership of Participation Certificates
Comparative Ownership Structure, December 31, 1966
Ownership by Issuing Agency
Size and Numbers of Individual Agency Issues . . . . . . . .
Ownership by Size of Agency Issue, March 31, 1967
Gross Dealer Transactions in Comparative Securities' Markets.
Results of Multiple Regressions Explaining Trading in Agency
Securities « . « • « • . . • . . . • « . . . . • « . . « •
Annual Rates of Turnover
Daily Average Trading by Weeks During 1966
Dealer Net Positions in Comparative Securities' Markets . .
Dealer Gross Positions in Comparative Securities' Markets . .
Results of Multiple Regressions Explaining Dealers'
Positions in Agency Securities
Most Typical Spreads in Comparative Securities' Markets . . .

40
41
45
47
49
52
54
56
57
60
74 & 75
79
83
89
91
96
98
102
103
106 & 107
117

-iiiTA3LE OF CONTENTS (cont.)
Pap,e Number
Appendix
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.

Agency Debt Outstanding
Non-guaranteed Agency Debt by Type
Net Expenditures or Receipts of Selected Agencies
Maturity Structure of Non-guaranteed Agency Debt
Yield Spreads
Market Yields at Constant Maturities
Gross Dealer Transactions in Federal Agency Securities . .
Annual Rate of Turnover of Federal Agency Debt
Dealer Net Positions in Federal Agency Securities
Dealer Gross Positions in Federal Agency Securities . . . .
Spread Between Dealers1 Quoted Bid and Asked Prices
on Federal Agency Securities

118 & 119
120 & 121
122
123
124 & 125
126
127
128
129
130
131 - 133

INTRODUCTION AND SUMMARY OF FINDINGS

Federal Agency debt is not a new market instrument.

Federal

Agencies have been selling their securities to the public since before
World War II,

But the size of Agency debt outstanding was only $1-1/2

billion as late as 1950, and its secondary market virtually nonexistent.
The growth in Agency debt during the fifties, and particularly in
the sixties, has been enormous.
just

By mid-1967, Agency debt totaled

under $24 billion, larger than commercial paper, finance

company paper and bankers1 acceptances combined.

Accompanying the

growth in Agency securities has been the development of an active
secondary market.

Trading in Agency issues during the last year has

been averaging $200 million or more each day.

Compared to trading in

U. S. Government securities, which averages over $2 billion a day,
the Agency market appears small, but its secondary market is more
developed than that for any private asset.
With the growing activity and breadth of its secondary
market, and thus the enhanced marketability of the securities,
Agency debt is becoming an increasingly important substitute for
U. S. Government securities in investor portfolios.

And thus

developments in the Agency market should increasingly be felt in the
Government securities market as well.
out by the 1966 experience.

This has already been borne

The sharp increase in Agency debt

during early and mid-1966, absorbed by investors only at successive
new highs in interest rates, was an important element in the ensuing

-2-

near financial crisi30

Not only were the rising Agency yields

directly translated into higher yields on Treasury and private
securities, but the Agency supply situation contributed to the market
feeling of potential crisis.
This paper describes and analyses the secondary market in
Agency securities.

The remaining chapters include (1) The Character-

istics of Federal Agency Debt (including size, risk, maturity, yields,
and ownership); (2) The Homogeneity of Agency Securities; and (3)
Indicators of Market Performance.

For purposes of analysis, considerable

data were drawn together from a wide variety of sources, some less
reliable than others.

Where possible, the analyses included the

fifties to permit temporal comparisons; often, however, data were
available only for the sixties.
A few words on the nature of Agency debt will help to define
and limit the focus of this study.

Debt of all Federal Agencies

can be subdivided into three general types:

direct guaranteed, direct

non-guaranteed, and guaranteed participation certificates.

Agency

securities that are guaranteed as to principal and interest by
the U. S. Government are not considered in this study.

While such

direct guaranteed debt has grown gradually in size, it still totals
only around $500 million; and because single issues are so small it
is by and large not readily tradeable.—^

As of June 30, 1967, only three individual guaranteed issues
exceeded $25 million in size.

-3-

Of the two remaining types, direct non-guaranteed issues
comprise the bulk of Agency debt outstanding--some $18 billion in
mid-1967.

Non-guaranteed securities—^ are the liabilities of six

Federal Agencies.

The Agencies are either supervised, partially2/

owned, or entirely-owned by the U. S. Government.—

These six

Agencies include the Federal Land Banks, the Federal Intermediate
Credit Banks, the Banks for Cooperatives, the Federal Home Loan
Banks, the Federal National Mortgage Association (secondary market
operations function), and the Tennessee Valley Authority.
The expenditures of these Agencies, with the single exception of TVA, are intimately related to the extension of credit
directly or indirectly to the selected sectors of housing and farming.
The Banks for Cooperatives, FICB, and FLB provide loans of varying
maturities to private farm groups.

The Federal Home Loan Banks lend

to savings and loan associations and to other miscellaneous savings
institutions.

Finally, FNMA provides supplementary assistance to

the mortgage market through secondary market purchases and sales.
Their net debt issuance (or repayment) is, in turn,
directly related over the long-run to the net expenditures (or receipts)
of the Agencies.

By and large, net loan extension by the Agencies

will result in a growth in Agency debt outstanding of roughly the
same amount.

Over the short-run, let us say several months, the

While not guaranteed by the U. S. Government, they are of course
guaranteed by the Agencies themselves.
2/ The net expenditures of these Agencies are included as part of
U. S. Government expenditures on a cash budget basis.

-4-

Agencies have available some alternative sources of lendable funds.
The Banks for Cooperatives and FICB at times borrow relatively small
amounts from commercial banks.

The Federal Home Loan Banks have a

sizable portfolio of U. S. Government securities on which they often
draw to supplement financings or to tide them over periods between
financings.

And FNMA may borrow directly from the Treasury, and

often does on an interim basis prior to a debt sale*—^
The second

type of Agency debt instrument considered

in this study is the participation certificate.
instrument, first offered in late 1964.

This is a quite new

Its growth has been rapid,

however, and at mid-1967, some $5.7 billion of fully marketable PC's
were outstanding.

These instruments are participations in pools of

assets, such as VA- and FHA-guaranteed mortgages and Export-Import
Bank, Commodity Credit Corporation or Small Business Administration
loans.

Except for Export-Import Bank and CCC certificates, PC's

are generally called FNMA PC's since FNMA acts as the trustee for
the sales.

PC's are now considered to be fully guaranteed by the

U. S. Government.
By their nature, sales of PC's are at the discretion of
the Federal authorities; their size is not determined by the operating
expenditures of the Agencies involved.

They are in effect a sub-

stitute means of financing the Government's deficit.
\J

While their

This is meant to be only a cursory look at the functions, expenditures,
and financing of these Agencies. A detailed examination would show
considerably more complicated balance sheets than indicated here*
Such detail is presented in D. Hunter, "U. S. Government Agency Financing", a memo from the Federal Reserve Bank of New York, Also see
"Federal Agency Securities," in Monthly Review, Federal Reserve Bank
of San Francisco (September, October, and November, 1963),

-5-

interest cost is higher than that of direct Federal debt, they have
been attractive to the authorities at least partly because they enter
the Federal budget accounts as negative expenditures, thus reducing
the size of the budget deficit (or increasing the budget surplus).—^
The major findings of this study of Agency debt and its
secondary market follow.
(1) The steady growth in Agency debt already noted has,
over the long-run, been accompanied by declining spreads
between yields on Agency and Treasury securities.

This

would indicate an improvement in the Agency market.
(2) The demand for Agency debt has risen with, if not
ahead of, the supply.

From the fifties, there has been

a dramatic improvement in the breadth of the market, as
evidenced by the wide variety of investors who have
fairly recently acquired Agency issues, often while
simultaneously selling U. S. Government securities.

The

larger nonfinancial corporations and state and local
governments appear to participate in the Agency market
in the same degree as in the U. S. Government securities
market.

Commercial banks (particularly reserve city

banks) and the larger nonbank financial institutions
(particularly life insurance companies), however, account
for lesser shares of non-guaranteed Agency debt than

For greater detail on PC's see Lawrence Banyas, "New Techniques in
Debt Management Since the Late 1950's," Treasury-Federal Reserve
Study of the U. S. Government Securities Market, 1967.

-6-

of U.
of PC's.

Government issues and for still lesser shares
The difference is especially evident in the

longer-term maturities.

Apparently, these two investor

groups view at least long-term Agency securities as
less marketable than Treasury issues.

But again, the

difference has diminished during the sixties, indicating
relative improvement in the Agency market,
(3) It is clear that the supply of Agency debt rose too
rapidly during 1966 to enable absorption by investors
without considerable congestion in the Agency, and indeed
in other, markets.

The $5 billion rise in Agency debt

over the first two quarters of the year was accompanied
by drastic rises in yield

spreads between Agency and

Treasury securities, and the spreads had not yet returned
to normal levels by mid-1967.

Thus, while a steady rise

in the supply of Agency debt is a prerequisite for
improvement in the Agency market over the long-run, an
excessive rise in debt can lead to short-run market
deterioration.
(4) There appears to be a single market for all of the
diverse non-guaranteed Agency securities and participation
certificates.

That is, investors apparently view the

securities as homogeneous.

This study found no consistent

or significant differences in market yields or in
ownership of the various Agency securities, including PC's.

-7-

(5) Some evidence was found, however, indicating that
the size of separate Agency issues is an important factor
in their marketability.

Agency issues in 1967 ranged

widely in size, from $20 million tc $535 million (publiclyheld amounts).

The evidence gathered in this study, albeit

limited, showed that quoted yields on the smaller issues
varied quite widely off the yield curve and that they
are to a greater degree than the large issues lodged in
the portfolios of comparatively inactive investors.

The

small size of many issues, particularly long-term maturities,
is probably a major reason for larger commercial banks
and nonbank financial institutions to participate less
actively in the Agency than in the U. 5. Government
securities market.

Moreover, the sharply increased

participation of financial institutions in the new,
larger PC's from their meager participation in the small,
serial PC's is a further indication of the importance
of issue size in the market's development.
(6) The volume of trading in the Agency secondary market
has risen sharply from the early sixties in both shortand long-term maturities,indicating increased ease for
investors to effect buy and sell orders with speed and at
market prices.

The increase in activity resulted from the

rise in Agency debt and the rising volume of gross new issues.
Activity in the Agency market is still only 10 per cent of

-8-

trading in the U. S. Government securities market, though
over the sixties activity in Agencies has grown relative
to Governments.

In the short-term sectors of the markets,

u
a
. *•
,^
,transactionsx
however, Agency trading and turnover (
debt
'

are

at least the equivalent of trading and turnover in Treasury
coupon issues due within 1 year.
(7)

It has sometimes been asserted that trades, and

particularly purchases, of Agency securities are effectuated
primarily during Agency financings, and that the supply of
Agency debt available for trading apart from financings
is limited.

This study found that while activity of course

was higher during financing periods, market activity remained
relatively high at other times.

In particular, excluding

trading during financing periods does not alter the conclusion that the short-term Agency market is at least
the equivalent of the short-term Treasury coupon market.
However, there did seem to be a greater spread between all
trading and trading outside of financing periods in the
longer-term sectors of the Agency market.
(8) Dealers have become more willing to position Agency
securities from early in the sixties.

There has, in fact,

been a three-fold rise in dealers' positions.

The higher

positions have resulted from the greater supply of debt
and market activity and from the rise in gross new issues.
(9)

Spreads between dealers' bid and asked prices in the

short-term Agency market are as low as in the U. S. Government

-9-

securities market.

Quoted spreads on intermediate- and

long-term Agency securities, however, have been around
1 point.

While such issues certainly trade at lower spreads

than the often nominal quotes, it is clear that the spreads
are larger than in the Treasury bond market.

-lo-

ll.

CHARACTERISTICS OF FEDERAL AGENCY DEBT

As a preface to an analysis of the secondary market in Federal
Agency securities this chapter describes the characteristics of Agency
debt.

Such characteristics include supply, maturity structure, risk,

market yields, and ownership.

These supply and demand factors are

integrally related to Agency market performance; they in part influence
and in part reflect the condftLon of the market.
A.

Supply

In mid-1967 Agency debt outstanding totaled $24 billion.—^
Compared to marketable U. S. Government debt, which stood at $211
billion, Agency debt was small.

But the growth in Agency debt during

the fifties and sixties has been extremely rapid.

At the beginning

of the fifties there was only $1-1/2 billion of Agency debt outstanding.
Starting from such a low level, the rise in Agency debt takes on
added significance.

It has, in fact, signaled the institution and

development of a new securities market.
The growth in Agency debt has been virtually continuous,
as the accompanying table illustrates.
roughly doubled every 5 years.

Agency debt has, since 1950,
t

Its increase has been especially

rapid since the end of 1965, with the supply up $8-1/2 billion.
1/

Unless otherwise noted, Agency debt is defined to include nonguaranteed Agency issues * FNMA participation certificates and
fully marketable Export-Import Bank participation certificates.
The $500 million CCC participation certificate issue in April,
1966 and retired in August, 1966 is also included.

-11-

Profile Chart 1 shows non-guaranteed Agency debt outstanding and
participation certificates for the 1S54-1967 period.

These data are

presented in Appendix Table 1.

Federal Agency Debt Outstanding
(Billions of dollars)

Dec. 31, 1950

1.8

1955

3.6

1960

7.9

1965

15.3

1966

21.3

June 30, 1967

23.8

The rise in Agency debt during this period has in part
derived from the introduction of several new types of Agency securities.
The Federal National Mortgage Association did not issue securities
until 1955, nor did the Tennessee Valley Authority until 1960.

Finally,

the introduction and growth of marketable participation certificates
during and after 1964 has accounted for roughly one-half of the rise
in Agency debt in recent years.

But the upward trend in Agency debt

has not resulted solely from the introduction of these new types of
issues.

The already established issues of the Farm Credit Agencies

and of the Federal Home Loan Banks have also trended upward as the
demand for their credit has grown along with the nation's GNP.

CHART 1
PROFILE OF TOTAL AGENCY DEBT OUTSTANDING
(End of quarter datat 195^-67)

1956
Appendix Table 1

-13-

In Profile Chart 2 and Appendix Table 2 non-guaranteed
Agency debt is shown by type of issue, and the upward trend is
evident for all types.

Growth in the debt of FNMA and FHL3 has been

most rapid, so that their issues now total roughly as much as the
issues of the Federal Land Banks.
Changes in the supply of Agency debt show cyclical and
seasonal variations that to some degree are obscured by the pronounced
trend.

These variations are particularly evident when the types of

Agency issues are considered separately.

All five types of non-

guaranteed Agency debt can be seen to have some cyclical component,
though it is most evident for FNMA and FHLB debt.

In general, the

rise in Agency issues is correlated with the degree of monetary
ease or tightness.

When money is tight, Agency debt rises rapidly;

when money is easy, Agency debt rises mere slowly or, in some cases,
declines.

These shifts reflect, of course, increased demands for

the credit of the five Agencies when alternative credit availability
is diminished and when market interest rates are high.
The direct interest rate pressures resulting from a rising
or falling supply of Agency debt thus generally reinforce rate
pressures over the cycle resulting from shifts in monetary policy
and in private credit demands.

To some degree offsetting, of course,

is the concurrent provision of credit by these Agencies to the
selected sectors of housing

and farming.

But even with no net

change in credit flows, a rapid shift in the supply of Agency debt
can have a marked impact on securities market, as borne out by the
experience of 1966.

Severe credit restraint did, and can in the

CHART 2
PROFILE OF NON-*PUARANTEED AGENCY DEBT BY TYPE
(End of quarter data, 195^-67)

SOURCE: Appendix Table 2

-15-

future given the present institutional framex^ork, cause a sharp rise
in demands on all of the Agencies simultaneously and result in an
increase in Agency debt sizable enough to have far-reaching effects
on public and private securities markets generally.

In fact,

continuing growth of all types of Agency debt increases the potential
for even greater

swings in Agency debt in the future.

Non-guaranteed Agency debt also fluctuates seasonally,
reflecting the seasonal nature of credit demands on the Agencies.
The seasonal variation in the debt issuance (or repayment) of the
Agencies can be

seen in Profile Chart 2 and in addition Appendix

Table 3 shows their quarterly net expenditures or receipts.

least during the sixties, the Federal Home Loan Banks have repaid
debt during the first quarter with the repayment of borrowings by
savings and loan associations; FHLB borrowing has tended to be
heaviest during the second quarter of the year.

Debt issuance by

FNMA does not appear to have any seasonal pattern.
Borrowing by the Farm Credit Agencies taken together is
seasonally high during the first half of the year, though the separate
Agencies have partially offsetting borrowing patterns.

The Federal

Intermediate Credit Banks, founded in order to help farm organizations
meet seasonal production and marketing costs, repay their debt during
the fourth quarter and borrow during the remainder of the year,
particularly in the second quarter.

The Federal Land Banks, whose

loans are long-term, show little seasonal variation in their borrowing
but it does appear to usually be largest during the second quarter.
The Banks for Cooperatives, on the other hand, repay debt during the

-16first half of the year and borrow in the market during the second
half, mainly in the fourth quarter.
When aggregated, the seasonal variation in expenditures and
borrowing of these Agencies is quite sharp, as the accompanying table
indicates,

Agency borrowing is concentrated during the second and

third quarters of the year and is highest during the second quarter.
It was no accident, then, that congestion

in the Agency market in

1966 peaked during the summer months, when the cyclically heavy
borrowing needs of the Agencies were superimposed on needs already
at their seasonal peak.
Net Expenditures or Receipts (-) of Federal Agencies^
(Quarterly; Millions of dollars)
I

III

1961

645

604

496

309

1962

401

844

599

130

1963

-1,013

1,131

1,168

512

1964

376

738

459

168

1965

259

1,191

645

1966

111

1,633

955

-547

1967

-1,378

885

1/ Indices FHLB, FICB, Bks. Coops, and FLB.
Source: Appendix Table 3.

Export-Import Bank participation certificates have usually
been issued during the first half of the year, specifically in
February and May.

PC issues by FNMA have not followed a set pattern,

and indeed need not.

Once authorized by Congress, sales of PC's

-17-

might be timed to satisfy any number of goals, such as to aid in
meeting Treasury financing requirements, to mesh smoothly with
other Agency financings, or to take advantage of a receptive market,
B.

Maturity Structure and Other Characteristics

The comparative marketability, liquidity and yields of
securities reflect a variety of factors that differ from one market
to another and that differ even within given markets.

Of major

importance is the breadth and depth of the secondary market.

But

of some significance as well are factors such as length to maturity
and risk of default, which are considered briefly in this section.
Maturity structure.
debt is short-term.

The bulk of non-guaranteed Agency

In mid-1967, about two-thirds of such debt

matured within 1 year.

Of the remainder, 22 per cent matured in

1-5 years and 10 per cent after 5 years.

There is a greater con-

centration of Agency debt in the short maturities than is the case
in the U. S. Government securities market.

In mid-1967, fewer than

half of all Treasury issues outstanding were due within 1 year
while almost 25 per cent were due after 5 years.
The growth in Agency debt in the recent past has embraced
all maturities.

Non-guaranteed Agency debt due within 1 year totaled

$12.2 billion in mid-1967 as compared with $5.3 billion in mid-1960.
Over the same period, debt due in 1-5 years rose from $1.7 billion to
\J

The proportion due after 5 years was an even larger 31 per cent
in mid-1965, about the time the Treasury had to stop issuing bonds
because yields had risen above the 4-1/4 per cent interest rate
ceiling.

-18-

$3.9 billion and debt due after 5 years from $1.4 billion to $1.9
billion.

The longer-term issues, however, grew at a slower pace and

the proportion of non-guaranteed Agency debt in long maturities has
declined to its 10 per cent level from 16 per cent in mid-1960 and from
25 per cent in early 1962.

Data on Agency debt by maturity are

shown in Profile Chart 3 and in Appendix Table 4.
The maturity composition of Agency debt differs widely among
the issuing Agencies.

These differences in general reflect the

structure of the Agency's assets.

The loans and discounts of the

Federal Intermediate Credit Banks are, as a rule, short-term and are
restricted to maturities of 7 years or less.

As a result, FICB debt

cannot exceed 5 years and in practice in recent years their debentures
have all been issued with a final maturity of 9 months.

The Banks

for Cooperatives also generally make short-term loans, and all of
their debt has recently been issued with a 6-month maturity.

Loans

of the Federal Home Loan Banks may be relatively long-term as well as
short-term and FHLB debt has sometimes been issued in the 1-5 year
maturity area.

Over the last few years the longest-term new issue by

FHLB carried a 2 year 10 month maturity while most of the issues had
original maturities of about 1 year.
Non-guaranteed Agency debt due in more than 5 years is issued
exclusively by FLB, FNMA and TVA.
of from 5-40 years.

The Federal Land Banks make loans

In mid-1967, FLB bonds were about evenly divided

between within 1 year, 1-5 year, and after 5 year maturities.

the same date, their longest-term issue carried an 11-1/2 year final
maturity.

CHART 3
PROFILE OF MATURITY STRUCTURE OF NON-GUARANTEED AGENCY DEBT
(End of quarter data, I960-67)

-20-

Th e secondary market operations function of FNMA of course
holds assets that consist mostly of VA- and FHA-backed mortgages.
Market obligations of FNMA thus also encompass long-term as well as
short-term maturities.

As of mid-1967, most of the FNMA debentures

outstanding were to mature in 1-5 years, with a relatively few number
of issues due in less than 1 or in more than 5 years.

In addition

to the debentures, FNMA issues short-term discount notes due in 30
to 270 days.
The Tennessee Valley Authority is the sole Agency issuing
non-guaranteed debt whose function is not intimately related to credit
extension, and the bulk of its assets are real as opposed to financial.
TVA debt presently includes several bonds due after 10 years and
short-term discount notes sold at auction.
Participation certificates have a considerably longer average
maturity than do the non-guaranteed Agency issues.

The maturity of

PC's derives from the characteristically long-term nature of the
pooled assets backing the PC's.

In mid-1967, of the $5.7 billion of

marketable PC's outstanding, all but $0.2 billion were due in more
than 1 year and some $2.5 billion were due after 5 years.

The longest-

term PC outstanding carried a final na turity of just under 20 years.
Virtually all of the outstanding balance of PC's has been
issued in the period since early 1965 at a time when the Treasury
has been unable to issue any debt due in more than 5 years because
market yields rose above the 4-1/4 per cent interest rate ceiling set

-21-

by Congress on Treasury bonds.—

The exclusion of Agency debt from

the rate ceiling permits the issuance of some long-term debt by the
Federal authorities during expansionary periods in support of a countercyclical debt management policy.
Risk of default.

Debt of the U. S. Government is as free

from risk of default, as to either principal or interest, as any debt
obligation.

In fact, it is probably viewed by most investors as being

a completely riskless investment.

Agency debt, as obligations of

wholly-owned, partially-owned, or Government-supervised Agencies,
shares in the risk-free nature of direct U. S. debt in varying degrees.
Participation certificates are fully guaranteed by the U. S.
Government.

In the September, 1966 ruling of the Attorney General

it was stated that PC's, which are issued by branches and dependent
Agencies of the Federal Government, "constitute general obligations
of the United States backed by its full faith and credit."

addition, of course, PC's are backed by pools of financial assets.
Other Agency debt, which is by and large issued by Agencies
with only partial Government ownership, is not guaranteed by the
U, S. Government, though it is of course guaranteed by the issuing
Agency.

The fact that the Agencies were created by Congress, are

supervised and in some cases partially owned by the U. S. Government,
and in some cases may borrow directly from the Treasury makes their
1/

At the end of fiscal year 1967, Congress authorized a redefinition
of Treasury notes that extends their maturities out to 7 years
from the previously-defined 5 years. Treasury notes are not
subject to the interest rate ceiling.

-22-

debt in practice almost Government-guaranteed,

Of the six Agencies

that issue non-guaranteed debt, in only two cases--FHLB and FLB--have
the Agencies completely retired stock held by the U. S. Government.
The remainder are partially Government-owned.

Three of the ^six--

FNMA, FHLB and TVA--have the authority to borrow directly from the
Treasury; FNMA may borrow up to $2-1/4 billion and FHLB $1 billion.
Furthermore, the debt of these Agencies, with the exception
of TVA, is backed by financial assets of at least a comparable amount.
The assets of course vary in liquidity, ranging from cash reserves
and U. S. Government security holdings to long-term loans and
mortgages (VA- and FHA-guaranteed).
Other characteristics.
or registered form.

Securities may be in either bearer

Bearer form securities have greater marketability

as they are more quickly and easily traded.

Registered securities

require signatures by owners and other registration procedures that
are time-consuming; in addition, they cannot be transferred over the
Federal Reserve wires.

Non-guaranteed Agency issues may all be

obtained in bearer form.—^

The same is now true of FNMA participation

certificates, though this has been the case only since January, 1967.
This represents a significant improvement in the PC's marketability,
and thus attractiveness to investors.
Also beginning in January, 1967, FNMA participation certificates
were marketed as term issues, with relatively sizable amounts in each
2/
maturity.—

17
2/

Prior to 1967, the FNMA PC's were marketed as serial issues,

In some cases, especially on the shorter-term issues, they are
available only in bearer form.
In the period since then there have been 8 separate FNMA PC issues
marketed, ranging in size from $200-550 million ($150-400 million
offered to public investors).

-23-

with small amounts in a number of issues ranging over a variety of
maturities.

The serial issues ranged in size from only $20-70 million,

and were thus not readily tradeable in the secondary market.
At the same time, Export-Import Bank PC's were made fully
marketable.

In prior offerings, Export-Import PC's were sold only

to restricted groups of investors, primarily commercial banks.

These

PC's are available in bearer as well as registered form.
Another recent development enhancing the attractiveness
of Agency debt has been the institution of repurchase agreements
against Agency issues by the Federal Reserve.

Repurchase agreements

against Agency issues were first made in December, 1966 and have, since
that time, formed a regular part of System Rp operations.

The immediate

impact of the Rp's is to make Agency securities more attractive to
dealers.

But as dealers become more willing to hold Agency debt the

entire market benefits through greater marketability, resulting in
lower spreads between Agency yields and yields on other securities.

-24-

Yields and Yield

Spreads

Agency yields are subject to the same general forces that
determine other market rates.

As Chart 4 shows, Agency yields move

over the cycle in line with other yields.

In 1966, yields in all

of these markets reached post-War record highs, rising to--and in some
cases above--6 per cent.

By mid-1967, yields had declined, though a

glance at the Chart shows that they still remain at high levels.
In comparing yields in the various markets, it can be
seen that Agency yields are generally higher than Treasury yields but
lower than yields on private investments, such as corporate securities
or CD's.

The spread between yields on the various instruments itself

varies within a fairly wide range, since yield levels at any point
in time may reflect supply and demand factors peculiar to a specific
market as well as factors general to all markets.

That yield spreads

can change radically is clear from the 1966 experience.

As shown in

the Profile Chart 5 of Yield Spreads,—^ in 1966 there was a sharp
increase in the spread between Agency and Treasury yields and between
private and Agency yields (and thus between private and Treasury
yields).

Yield spreads between Agencies and other issues rose to

around 50 basis points.
During 1966, the supply of Agency securities rose rapidly,
as did supplies of private issues, particularly corporates.

On the
2/

other hand, the supply of long-term Treasury issues was declining^
and the supply of Treasury bills also declined seasonally over the
jy
2/

These spread data are in Appendix Table 5.
No Treasury debt due in more than 5 years has been issued since
early 1965, and the supply has declined with the passage of time.
In 1966 alone, it declined by $7.2 billion.

CHA.RT h
PROFILE OF MARKET YIEIDS
T^jQiuair^epri
Perce: it r

4r.4r.4rs*

3-MONTH YIELDS
V P & S t

-G0-L
Agency issues
i f A'aWA'A**
h

—

m t X H f i ? .
m

i
4<

-A
m

bills —r

Z
J
zj iZ-

VaTA*.?*
6
x^f4"jf
'A-'.0 AW
4
A'aF.A'A

{

Perce; rtr

\4'aV
1w A T j § ' A W A
\r.4'Ar.pjr,
,04^0 A^aW
r
A**0ArA0.4
\^wjrAr4rA
f
t04 A0.4'AK
A'aPA'aW
irAr.4rr
AWj

10-YEAR YIELDS

fS'Af.

A .

AV.4'a 4'A
r
4'A^.0A a

'A^JTA'aTA
r,04',*
A'a
wjr&s'**
J J
W4'Ar.0jr\
'A

a04^.0A

—

tr n
h

—

L
tzt

—

t z

-—

z
z

L z
zL

i
L

t z

z
z

L
L

tzz

t t L

1
[:

l i t
m

_
__ _ __

r Z

____

~1
11

Z
zj
z

~~j

H
J

_ __ _ ___

_J
z
zj

z
_l

zj
zz
z
z Zj_J
I JJ

_ _

—
—

rr
r Z

z•
1p
•
•
r I Jz I JZ] I
'M I I UN I I IIITttlT
1
1r
1
I
J I
I
1J I 3 I J
Hi Mil 11 lirrrnT
li Ml
-Lypu
196?
'
war
SOURCE: Salomon Brothers ard Hutzler, An Analytical Record of Yields and Yield Spreads. Quarterly averages computed fran m0nthly data.

04'*0ArAn

u u

1 z Jz
|z [Z
L
z
[Z
t\
z
1
1
s
h
!
1
z
^z
z
r
7
p p
[z
[z
(r
t t
t
t
t
r r s
Jz z1zj
J
h*
/ _lz 1
r
L
J
y nr
Z
z
z
L
_
tr
jz
t
jZ
t
1s
t
c Z"
r-r*
h
_i
z
~2
1z
z
~1 z
r~
z —
~J7
L1
Jj N
z
T
n
i
i
li
E
M
s
i
L
I
I
Lfi
s
% -- t - E -w £ - -C- p
- r -- Z- - -t n -u
- h-- - j
J H - 5 t
1_
J
E
s
2
H
n
n
zj J H
zj
|Z
^
III t
I I ud Z I
I Z z z ZJZ Z z
z zz
Z z Z z z z Jz z z z z
t Z
z z z hF z z z f- z Z
—
_
_
— —
^
J
-j
- zji
^
zi
H
z
z zz
z —d
E
1p ^
2
? Ps
A
s ii
1—
F Zjz z z z z z z
F
F
•
z
z
z
z
z
z
z
z
L
rnv >= = I f- S I E:= H2 f z zJ H
H
p
—
— - v-- 9fj- - - - - - - - - "1 - —!- - - - - - - - - - J - - - - - - - - - - - 1 L - - - - Z- - - j j
- 1
Ji—
zL
z n
~
j
z Z]
j
Lj
i~
z p
z
r
z
t~
1 1
zt
~1
r gi aj K E u & szJ
Z] z
z
z
z Z]
zj
z
j
z
r
~
z
1 Z
L
J
zz
q
J
J
z
Zj
z
z
t

mmart'Ar

2=11

KJ-akw,

'J9
J'A'4
r*9JPa9
S a
WATa?*'**
s'jr.s'A*.*
t
w

j
~~j_1
1

'•AWA'aW.4
0 A'.
A"

zn
r
t n
n

\r.0Ar.0jrd

lti

1
1
•

p
•

•

z
ZJ
J

]

CHART 5
PROFILE OF YIELD SPREADS
(Quarterly averages, 195^-67)

Basis points

195^

SOURCES

1956

Appendix Table 5

Basis points

N>
ON

-27-

first three quarters of 1966 when the yield spread with Agencies was
increasing.

Thus far in 1967, some yield spreads have returned to

more normal levels, although there has been no decline in the spread
between Agency and Treasury long-term yields, where comparative supply
shifts remain unchanged.
Looking back over a longer period one can see a marked
secular decline in spreads between long-term Agency and Treasury
yields.

Panel 3 of Profile Chart 5 shows yield spreads of around 50

basis points during 1956-1957 and spreads of around 25 basis points
throughout the 1958-1961 period.

Beginning in 1962 there was a steady

decline in the spread until it was no more than 5 basis points during
1964.

Thereafter, the spread increased.

The decline in the spread

is certainly in part related to an improvement in the breadth* of the
Agency market; larger investors were increasing their share of the
market considerably even at the declining spreads.

However, the

spread decline may also be attributed partly to differing supply
shifts.

During the early sixties, advance refundings were adding to

the supply of long-term Treasury debt while long-term Agency debt
outstanding was declining slightly.

When the supply situation was

reversed in 1965, with the introduction of PC's and the cessation
of long-term Treasury issues, the spread began to increase.
In Panel 4, the spread between Agency and corporate yields
shows virtually no change, apart from its marked rise in 1966.

fluctuated widely in a 25 basis point range in the fifties,cn several
occasions decreasing to zero.

In the early to mid-sixties it remained

in a narrower 10-15 basis point range.

-28-

In long-term markets in general, the quarter-to-quarter
fluctuation in yield spreads diminished considerably during the early
sixties, as the Profile Chart makes evident*

While it is certainly

possible that the diminished spread fluctuations might represent a
growing degree of investor arbitrage among various investments, it
may also simply reflect the greater day-to-day stability in yields
in all markets during the period.
Yield spreads between short-term Agency issues and Treasury
bills, shown in panel 2 of Chart 5, fluctuate widely over the cycle
and to a lesser degree seasonally.—^ Spreads between Agency and
Treasury yields have, throughout the period under consideration,
declined to very low levels during periods of easy money and have
risen during periods of cyclical expansion.

Thus, the spread was

around zero near the troughs of the 1954, 1957-58, and 1960-61
recessions, while it ranged up to around 50 basis points during the
subsequent expansions.

It would appear that the superior liquidity

and marketability of Treasury bills commands a greater premium during
periods of high and rising rates and lessened credit availability
2/
than during periods of easy money.—

The reader will note two curves drawn on the chart, one utilizing
market yields on bills and the other investment yields. Investment
yields reflect the true return on the invested funds. They differ
from market yields by giving the return on the amount invested
rather than on the face amount of the bill at maturity for a 365day rather than a 360-day year. Agency, and other, yields are
always on an investment yield basis.
To a minor degree, the cyclical movement of spreads, at least early
in the fifties, might be related to varying supplies of bills as
versus Agency issues over the cycle. Agency debt, as shown in an
earlier section, rises more quickly when money is tight, i.e.,
during expansions. Treasury debt, on the other hand, during the
fifties often rose more quickly during and just after recessions when
the deficit was enlarged by a drop in receipts with declining GNP.

-29-

Following the 1960-61 recession, the spread remained at very
low levels for several years.

In fact, using investment yields on

Treasury bills the spread was often slightly negative, i.e.,
yields were less than Treasury bill yields.

Agency

This sustained period

of low spreads probably reflected in part the maintenance of upward
pressures on bill yields by official operations of the debt management
and monetary authorities.
With the wide cyclical fluctuation in spreads between short-term
yields, it is virtually impossible to isolate any secular trends.

But

clearly improvement in the market for short-term Agency debt would by
itself have been expected to diminish the yield spread.
During the sixties, the spread between Agency and Treasury
bill yields also shows a consistent seasonal pattern.

The spread rises

in the second and third quarters and generally drops back in the
fourth and first quarters.

It will be recalled that new Agency debt

issues are concentrated in the second and third quarters, and these
quarters have often involved a redemption of Treasury bills.

The

sharp increase in the spread during the second and third quarters of
1966, to a level of 50-60 basis points, thus partly reflected a normal
seasonal rise.
As shown in panel 1 of the Profile Chart, the yield on 3month CD's has, except for one instance, been above the Agency yield.
Generally, the spread has fluctuated in a 5-25 basis point range,
although it was much larger through raid-1967. The sharp decline in
yields on short-term market securities with an easing of monetary
policy in late 1966 did not carry through to CD yields to as great a
degree, leaving a 50 basis point spread in the yield differential.

-30-

Available data on Agency yields are not always as comprehensive
or accurate as would be desirable.

Thus an alternative set of data on

Agency and Treasury yields is included in the paper.

The Treasury

Department, for its own use, on specified dates collects yields at
constant maturities, i.e., yields derived from points on the yield
curve.

These data are shown in Chart 6 and Appendix Table 6 for 1-year,

3-year, 5-year and 10-year maturities, from 1963 to mid-1967.
They show essentially the same spread movements, including a sharp
rise in the spread during 1966 to around 50 basis points on most
maturities.

Prior to 1966, the spread on every maturity fluctuated

in a 10-25 basis point range.

The spread was generally lowest for long--

term issues and highest for 1-year maturities.

The levels of spreads

shown in these data, however, do not always coincide with the Salomon
Brothers data.

For instance, these data do not show a virtual elimina-

tion of the yield differential on 10-year maturities during the 1963-64
period, as do the earlier data.—^
Special market conditions might for a time virtually eliminate
the differential between Agency and Treasury yields, but it is unlikely
that this condition would be sustained over a long period of time.
While Agency issues might be considered in practice as risk-free as
Treasury debt, the more developed market for and greater tradeability
of Treasury debt relative to Agency debt in most sectors should require
1/

Yields used in the two sets of data differ, at least in part
because Salomon Brothers yield data are based on offered quotations (except for bills) and Treasury data on bid quotations.

YIELDS AND YIELD

1963

SOURCE:

196U
Appendix table

1965

1966

CM.RT 6
SPREADS FOR VARIOUS MATURITIES

1967

1963

196U

1965

1966

1967

-32

some yield spread.

But a return to a slower but steady growth in

Agency debt, accompanied by continued development of the secondary
market, should result in a downward drift of the spread.
D.

Demand

The growth in Federal Agency debt outstanding since the
early fifties has been accompanied by a considerable broadening
in its ownership.

At the end of 1950, commercial banks held more than

80 per cent of the $1.8 billion of Agency debt outstanding.

The bank

share had dropped to 50 per cent by 1955 and to a low of 20 per cent
by I960*
Meanwhile, a host of investor groups added Agency debt to
their portfolios.

At the end of 1955, Agency holdings of most large

nonbank investor groups were only nominal.

In the five years after

1955, there was a sharp growth in the Agency holdings of nonfinancial
corporations and of nonbank financial institutions.

Then, after 1960,

state and local governments and individual investors acquired Agency
issues at a rapid pace, as did the smaller commercial banks.

During

the sixties, in fact, many large investor groups increased Agency
debt held in portfolio while selling U. S. Government securities.
The entrance of new investors into the Agency market has not
been related solely to the increased supply of debt, which led to
greater availability of issues as well as widened knowledge of the
market.

It has surely been dependent as well on the development of

the secondary market, and thus improved marketability of Agency debt,
and on the attractive yields on Agency issues relative to other securities.

-33-

But despite the sharply increased Agency holdings of most of
the larger and more active investors, the supply of Agency debt has
been increasing so rapidly in the last few years that the share of
Agency debt held by these investors has declined from 1962-63 levels.
Moreover, comparisons of ownership of Agency and U. S. Government
debt show that two investor groups--commercial banks and nonbank
financial institutions--account for a smaller share of the Agency
market, particularly in the longer-term maturities.

Commercial banks

in late 1966 accounted for roughly 10 per cent less of all nonguaranteed Agency securities publicly-held than of Governments and
this disparity was even more important for reserve city banks alone.
Larger nonbank financial institutions accounted for 10 per cent le ss
of over 5-year Agency securities than of Governments.

This same

disparity was even more true of FNMA PC ownership relative to
Governments.
It seems clear that larger (and probably more active)
investors in general account for a smaller share of Agency than of U. S.
Government debt.

Their smaller participation in the Agency market

reflects in part ,but causes as well ,a poorer secondary market.

The

larger the share of debt held in the portfolios of relatively small
investors, the greater is the likelihood that Agency issues will be
locked into investor portfolios until maturity rather than traded
actively in the secondary market.
Larger investors have increased their share of the Agency
market since the end of 1966, however.

In PC ownership, in particular,

-34-

latger commercial banks and nonbank financial institutions together
increased their share of PC holdings from 10 per cent to 37 per cent.
It would appear that these two investor groups view at least long-term
Agency securities as considerably less marketable, due at least partly
to the small size of individual issues.

It is likely that at least

some part of the sharp rise in their share of the PC market during 1967
relates to the previous-noted improvement in the PC's marketability.
The remainder of this section will examine the ownership
structure in more detail.

Ownership data are presented in Tables 1-10

and are shown in Profile Charts 7 and 7a.

The data on non-guaranteed

Agency ownership are confined to the period beginning in 1960-1961
as it is only for recent years that the Treasury Survey has included
all of the major investor groups.

Securities are classified by maturity

date since assets held by many investor groups are highly concentrated
in particular maturity areas.

The Tables also include data on owner-

ship of U. S. Government securities for purposes of comparison.

Owner-

ship of PC's is considered separately due to data incomparability.
The reader must be warned that the available data do not
present a clear picture of Agency ownership.

The Treasury Survey

does not cover all holders, but only the larger holders in any one
investor class.

The coverage in the Survey for any particular clarss

of investors ranges from an estimated 90 per cent of all Agency
securities held by the particular group to less than 50 per cent
in some cases.
1/

Therefore, the category called "all other investors,"

The estimated coverage for each group is shown in footnotes to Tables
2-5. It also appears that the coverage for any investor group can vary
greatly as between Agency and U. S. Government holdings. In all cases,
reporting investors appear to account for a lower percentage of holdings of Agencies, implying that small institutional investors hold a
larger share of Agency securities than of U. S. Government issues.
This divergence makes comparisons of ownership in the two markets
somewhat tenuous.

CHART 7
PROFILE OF OWNERSHIP STRUCTURE OF NON-GUARANTEED AGENCY DEBT

CHART 7a

Source: Treasury Sirvey of Ownership

.35.

-37-

besides including individuals and other non-reporting groups such as
nonbank Government security dealers and investment companies, includes
non-reporting banks, corporations etc.

If allowance could be made

for these non-reporting investors, holdings of individuals would probably
appear relatively small.

They would certainly not dominate the market

as appears to be the case in the accompanying Profile Charts.
For example, for 1965 and 1966 some data are available for
adjusting the Survey information to give a clearer--though still far
from perfect--picture of the ownership profile of non-guaranteed Agency
debt.

If Survey-reported commercial bank holdings are raised to

include Agency holdings of all commercial banks—^ and corporate pension
trust fund holdings are removed from the all other category, relative
ownership shares are altered dramatically.

As the table shows, at the

end of 1965 bank holdings of short-term Agency issues were slightly
larger than "all other investor" holdings, though this was not the
case in 1966.

These data inadequacies should be borne in mind in the

following discussion.
Holdings of Agency Debt due:*
Within 1 yearjln 1 - 5 years)After 5 years
Dec. 31, 1965
All commercial banks

3.2

3.1

1.5

All commercial banks

3.6

"All other investors"
category

6.0

1.9

1.1

"All other investors"
category
Dec. 31, 1966

*
1/

In billions of dollars.
Available for recent years from Call Reports.

-38-

Non-Ruaranteed Agency Debt
As should be clear from the preceding paragraph, commercial
banks play a more dominant role in the Agency market than the Profile
Charts indicate.

The share of Agency debt held by all commercial banks

is considerably below what it was in the early to mid-1950fs, but it
is not small.

In the last few years banks have held roughly 1/4 to 1/3

of all short-term non-guaranteed Agency issues in public hands.
They are less important participants in the intermediate- and longterm sectors of the Agency market, as is also true of their participation
in the U. S. Government securities market.
In dollar terms, bank holdings of Agency securities appear
minute next to U. S. Government securities held in portfolio.

At the

end of 1966, non-guaranteed Agency securities at all commercial banks
totaled just over $4-1/2 billion while marketable U. S. Government
securities totaled over $57 billion.

Banks account for a larger

share of publicly-owned U. S. debt as well.

Using Call Report data

for the end of 1966,—^ all commercial banks held 22 per cent of publiclyowned Treasury bills, just over 1/2 of coupon issues due in less than
5 years, and 27 per cent of long-term Treasury bonds.

In the Agency

market, on the other hand, banks held 30 per cent of short-term issues,
only 22 per cent of issues due in 1-5 years, and not even 10 per cent of
over 5-year Agency securities*

In both markets, the bank share was

reduced by several years of monetary restraint and would appear somewhat higher in another stage of the cycle*
If

Treasury Survey data, excluding some of the smaller banks, shows
the same comparative pattern (see the lower panel of Table 2).

-39-

There is also a difference between the Agency and U* S*
Government securities markets in the degree of participation of reserve
city as versus other classes of banks.

As shown in Table 1, in 1966

reserve city banks accounted for a greater proportion of all bank
holdings in the U. S. Government securities market while relative
participation of country and insured
the Agency market.

non-member banks was greater in

Country and insured non-member banks accounted for

85 per cent of bank holdings of Agencies but only 65 per cent of U. S.
Governments held by banks.
The growth in bank holdings of Agency debt during the
sixties has in fact been primarily at country and non-member banks.
For years before the mid-19601s, Treasury Survey data must be used,
and they are shown in Table 2*

Bank holdings of Agency issues have

increased thus far during t.:e sixties at an even more rapid pace than
during the fifties.

At the same time, however, banks1 share of

Agency debt outstanding has declined due to the rapid growth in
Agency debt (see the lower panel of Table 2)*
Bank holdings of Agency issues even rose slightly during
1966, when credit restraint was extreme.—^ This absence of selling
of Agencies by banks during the recent tight money period contrasts
with sizable bank sales of U. S. Government securities (also shown
on Table 2).

It contrasts as well with bank sales of Agency issues

Though through November of 1966, the commercial banks reporting
in the Treasury Survey showed a small decline in non-guaranteed
Agency holdings.

Table 1
OWNERSHIP OF AGENCY AND U. S. DEBT BY TYPE OF COMMERCIAL BANK
(Millions of dollars and per cent of all commercial bank holdings)
Nonguaranteed Agency Debt due:
After 5
Within
1-5
years
years
1 year

U. S. Marketable Debt due:
Within 1 year
1-5
After 5
|
Other
years
Bills
years

June 30, 1966

182
5

23
3

5
3

1,229
15

569
7

1,330
6

1,514
11

58
2

17
2

5
3

411
5

188
2

595
3

473
3

371
10

97
11

43
27

1,453
18

1,745
22

5,649
23

3,696
27

1,673
46

401
46

68
43

2,431
30

3,372
43

10,606
44

5,515
40

1,341
37

338
39

37
23

2,588
32

2,062
26

5,937
25

2,721
20

3,625

876

158

8,113

7,936

24,118

13,918

133
4
33
1

27
3
12
2

4
3
3
2

1,797
16
350
3

588
7
181
2

1,253
5
663
3

1,450
12
420
4

355
10

87
11

30
21

2,384
21

2,012
24

5,860
23

1,630
45

384
46

68
47

3,544
31

3,538
42

11,181
44

3,170
27
4,430
38

1,447
40

318
39

40
28

3,299
29

2,082
25

6,466
25

2,193
19

3,599

827

145

11,373

8,400

25,423

11,663

New York City
Chicago
Other Reserve City

%
Country
7c

Insured Non-member
7o

Total-'
December 31, 1966
New York City
7o

Chicago
7a

Other Reserve City
7o
Country
7o

Insured Non-member
7o

Total-'

"JL/ Does not include uninsured banks for which data are not comparable. Uninsured bank holdings of all Agency securities
total about $100 million and of U. S. securities about $350-400 million.
SOURCE: Board of Governors of the Federal Reserve System, Assets a n d Liabilities of Member Banks.

Table

OWNERSHIP OF AGENCY AND U. S. DEBT BY COMMERCIAL BANKSNonguaranteed Agency Debt due:

After 5
years

Within
1 year

1-5
years

995
1,769
2,221
1,990
2,231
2,282
2,299

592
518
525
557
652
597
570

84
122
119
97
91
118
198

23
31
30
27
25
18
21

26
25
25
20
20
16
17

4
5
5
5
6
7
11

Total

U. S. Marketable Debt due:
Within 1 year
1 - 5
After 5
Bills | Other
years
years

Total

Millions of dollars
Dec. 31, 1961
1962
1963
1964
1965
1966
June 30, 1967

1,671
2,409
2,865
2,644
2,974
2,997
3,068

9,962
9,838
9,290
10,969
10,156
8,771
5,844

11,187
10,047
7,413
7,540
7,847
7,067
5,403

30,751
26,348
26,107
23,507
19,676
21,113
24,919

7,174
11,772
12,070
11,737
12,645
10,232
9,359

59,074
58,005
54,880
53,753
50,324
47,183
45,525

25
22
20
23
20
17
14

41
43
39
45
45
44
43

57
53
55
49
45
44
47

19
26
25
24
25
23
23

37
36
34
33
31
30
31

Per cent of publiclyheld debt
Dec. 31, 1961
1962
1963
1964
1965
1966
June 30, 1967

20
24
25
22
21
17
19 ]i

JL/ Banks included in the Survey in 1966 accounted for about 64 per cent of all Agency securities held by banks and
about 83 per cent of all U. So securities held by banks.
SOURCE:

Treasury Survey of Ownership.
i
h-1
I

-42-

during earlier cyclical expansions, as shown in Chart 8.

The fact

that such sales did not materialize during 1965-66 might be due to any
number of factors, including the unprecedented rise in Agency debt
outstanding, the rapid decline in Agency security prices locking
investors into

the issues and a sharp increase in the spread between

Agency and U. S. Government yields.

But in addition it appears to be

due to the continued acquisition of Agencies by the smaller commercial
banks, since reserve city banks as a group did sell Agencies during
1966.

These smaller banks, on the one

hand, tend to be less affected

by monetary tightness and, on the other, were probably responding to a
new awareness of the Agency market coupled with the extremely favorable
spreads between Agency and other yields.—' Whatever the cause, without
the absence of sizable commercial bank selling of Agencies during
1966 the rise in Agency yields would have been even more spectacular
Next in importance to commercial banks as participants in
the Federal Agency market are the nonbanlc financial institutions.
These institutions include mutual savings banks, insurance ccnoanies
(life, fire, casualty, and marine), and savings and loan associations.
As of the end of 1966, non-guaranteed Agency holdings of such
institutions reporting cn the Treasury Survey totaled $1*9 billion,

Agency securities, except for FICB issues, cannot be used as
collateral for borrowing from the Federal Reserve. And many of
the major banks were short of collateral during 1966.

CHART 8
COMMERCIAL BANK HOLDINGS OF NON-GUARANTEED AGENCY DEBT, 195^-67
Billions of dollars

Source:

(End of quarter data)

Treasury Survey of Ownership

-44-

accounting for just over 10 per cent of Agency debt outstanding in
public hands.

Conversely to commercial banks, with liabilities of

longer maturity and greater predictability, they play a more important
role in the intermediate- and long-term sectors of the market.
During the early to mid-fifties these institutions held
virtually no Agency securities.

Beginning in 1956, however, they

began to add Agency issues to their portfolios in some size;

in the

five years ending with 1960 they added roughly $1 billion of such
issues to their portfolios and they had added another $1 billion by
mid-1967*
While Agency holdings of these financial institutions are
considerably smaller in dollar amounts than U. S. Government holdings,
as shown in Table 3, they account for roughly the same
of debt outstanding in both markets.

proportion

It can be seen, iiowever,

that the maturity distribution of their holdings varies considerably *
between the two markets.

In the Agency market, holdings are con-

centrated in the short-term maturities while in the Treasury market
holdings are overwhelmingly intermediate- and long-term.

To some

degree this is a reflection of a variance in the type of financial
institution most important in each market.

In the Agency market mutual

savings banks account for 1/2 of nonbank financial institutions1 debt
holdings and life insurance companies for less than 1 per cent.

But in

the Treasury market insurance companies account for roughly 1/2 of
institutions1 holdings and life insurance companies alone for 1/4 of
the group's holdings.

Among the types of nonbank financial institutions,

Table

OWNERSHIP OF AGENCY AND U.S. DEBT BY NONBANK FINANCIAL INSTITUTIONSNonguar<=mteed Agency Debt due:
Within
1 - 5
After 5
Tnf-i
lOldl1
years
years
1 year

U. S. Marketable Debt due:
Within 1 year
1 - 5
After 5
years
years
Bills | Other

Total

Millions of dollars
Dec. 31, 1961
1962
1963
1964
1965
1966
June 30, 1967

394
571
719
756
920
1,175
1,167

414
341
364
491
534
527
527

323
345
283
257
203
190
227

1,131
1,257
1,366
1,504
1,657
1,892
1,920

778
1,058
1,053
1,167
! 1,387
1,490
886

1,582
1,273
1,196
933
970
784
656

4,631
4,329
4,174
4,636
4,273
4,711
5,062

10,656
11,262
11,586
11,277
11,081
9,588
8,813

17,647
17,922
18,009
18,013
17,711
16,573
15,417

9
10
10
10
10
10
11

18
17
17
18
16
14
16

17
15
13
13
13
11
12

13
12
12
12
12
11
12

2
2
2
2
3
3
2

6
6
6
6
6
5
5

9
9
9
10
10
10
10

29
25
24
23
22
22
22

11
11
11
11
11
10
11

Percent of publiclyheld debt
Dec* 31, 1961
1962
1963
1964
1965
1966
June 30, 1967
1/

j
Includes mutual savings banks, insurance companies and savings and loan associations. Reporting mutual savings
banks and insurance companies account for no more than 90 per cent of all such securities held by these institutions, while reporting savings and loan associations account for only 50 per cent.

SOURCE:

Treasury Survey of Ownership.

i
t r
vn.

-46-

life insurance companies of course hold the longest-term assets.

For

these institutions it seems apparent that yield spreads in favor of
long-term Agency issues have not been large enough to offset their
lesser marketability.
Savings and loan associations account for roughly the same
share of debt in both markets, though again they account for fewer
of the long-term maturities in the Agency market.

Savings and loan

associations hold Agency securities despite the fact that they do
not help to meet legal liquidity requirements.
During 1966, despite declines in savings inflows, nonbank
financial institutions added Agency securities to their portfolios.
In fact, the $235 million rise in their non-guaranteed Agency holdings
during 1966 was larger than in any other year of the sixties.
Simultaneously, these institutions decreased their holdings of U.S.
Government securities by more than $1 billion.
favor of Agencies rose sharply at this time.

The yield spread in
During the first half

of 1967, the same portfolio shifts were in evidence.

Treasury debt

was reduced, again by more than $1 billion, and Agency issues were
acquired, though in only nominal amounts.
Nonfinancial corporations were important holders of shorttern Agency issues early in the sixties, but their participation in
the Agency market, as in the U. S. Government securities market, ha&
declined since about 1964.

The decline in corporate holdings of short-

term Agencies is cleorly shewn in the Profile Charts and in Table 4.

Table 4
OWNERSHIP OF AGENCY AND U. S. DEBT BY NONFINANCIAL CORPORATIONS-'
Nonguaranteed Agency Debt due:
Within
1 - 5
After 5
Total
1 year
years
years

Within ;L year
Bills
Other

U. S. Marketable Debt due:
1 - 5
After 5
Total
years
years

Millions of dollars
Dec. 31, 1961
1962
1963
1964
1965
1966
June 30, 1967

904
902
1,155
677
825
597
318

54
73
49
87
117
103
90

11
11
4
4
11
15
17

969
986
1,208
768
953
715
424

5,466
6,551
6,178
5,043
4,657
3,396
1,900

3,232
2,512
1,493
1,705
1,254
1,334
736

1,747
1,524
2,397
2,001
1,754
1,339
1,194

21
16
16
9
9
5
3

2
4
2
3
4
3
3

1
1

11
10
10
6
7
4
3

14
15
13
10
9
7
5

12
11
8
10
7
8
6

3
3
5
4
4
3
2

102
163
359
387
349
254
191

10,547
10,750
10,427
9,136
8,014
6,323
4,021

Per cent of publiclyheld debt
Dec. 31, 1961
1962
1963
1964
1965
1966
-June 30, 1967
1/

*
*

1
1
1

1
1
1
1
1

7
7
6
6
5
4
3

Includes only general funds. Reporting corporations account for about 50 per cent of all such securities held by
nonfinancial corporations.
Less than .5 per cent.
*
SOURCE-: Treasury Survey of Ownership.
4?rI

-48-

In 1961-62, corporate short-term Agency holdings were about

billion,

accounting for 15-20 per cent of the debt outstanding, but by 1966-67
their holdings had fallen to a $300-600 range, only 3-5 per cent of
Agency debt.

Corporate holdings of Agency issues due in more than 1

year have never been large.
Over the sixties U. S. Government securities held in
corporate portfolios have also declined, from $10-1/2 billion early
in the sixties to a $4-6 billion range currently.

Declines in nonfinancial

corporate holdings of both Agency and U. S. Government securities are
in general related to a shift of corporate liquid assets into time
deposits with the development of the negotiable CD after 1962 and into
other higher yielding short-term assets.

To a lesser degree it probably

relates to the usual decline in corporate holdings in advanced stages
of cyclical expansion and to a reduction in accrued tax liabilities
with the speed-up in corporate tax payments.

Corporations account

for roughly the same share of debt outstanding in both markets.
State and local governments have become increasingly
important investors in the Agency market, particularly in the shortterm maturity sector.

They are especially active in FNMA discount

notes which may be tailored to specific maturity dates.

But state

and local government holdings of securities cover all maturity sectors.
While general fund holdings are primarily in short-term liquid issues,
pension fund holdings are concentrated in long-term issues.

Thus,

in mid-1967, these governments reporting in the Treasury Survey held
about $1 billion of short-term Agency debt and almost $.5 billion of
debt due after 1 year (see Table 5).

While U. S. Government security

Table 5
OWNERSHIP OF AGENCY AND U. S. DEBT BY STATE AND LOCAL GOVERNMENTSNonguaranteed Agency Debt due:
Within 1 1 - 5
After 5
Total
years
1 year | years

U. S. Marketable Debt due:
Within :L year
After 5
1 - 5
Other
Bills
years
years

Total

Millions of dollars
Dec. 31, 1961
1962
1963
1964
1965
1966
June 30, 1967

179
243
246
385
854
887
985

67
48
29
104
205
223
200

167
259
264
328
279
269
273

413
550
539
817
1,338
1,379
1,459

4
4
3
5
9
7
9

3
2
1
4
6
6
6

9
11
12
17
18
15
15

5
5
5
7
10
8
9

2,710
3,282
4,260
3,961
4,574
4,512
4,700

1,264
1,165
1,149
902
997
1,032
923

1,320
1,059
1,618
2,014
1,862
2,166
2,262

5,599
6,210
7,577
8,144
8,274
7,674
6,949

10,893
11,716
14,605
15,021
15,707
15,384
14,834

5
5
6
5
6
6
7

2
2
3
4
4
5
4

15
14
15
16
17
18
17

7
7
9
9
10
10
10

Per cent of publiclyheld debt
Dec. 31, 1961
1962
1963
1964
1965
1966
June 30, 1967
1/

7
7
9
8
9
9
12.

Reporting governments account for about 70 per cent of such securities held by all state and local governments for
the years 1964-67, and 60 per cent for the years 1961-1963.

SOURCE:

Treasury Survey of Ownership.
L .

-50-

holdings were a much larger $15 billion, in both markets these governments account for roughly the same share of debt outstanding.
Investments by state and local governments in both the
Agency and U. S. securities markets have increased considerably over
the sixties, but the relative growth has been greatest in Agencies.
Whereas at the end of 1961 Agency holdings were only 4 per cent of
U. S. debt holdings, by mid-1967 they had risen to 10 per c e n t — a n d
for issues due in 1 year alone, the increase was from 5 to 18 per cent.
Some portion of the growth in Agency holdings results from only
recent legal permission in the case of some state and local governments to invest in Agencies, and a further freeing of governments1
choices should give impetus to future growth in their Agency holdings.
The

all other investors'5 category accounts for 50-60 per

cent of non-guaranteed Agency debt outstanding.

This category

includes investor groups not specified in the Treasury Survey,
such as individuals, foreign investors, nonbank Government security
2/
dealers— , and nonprofit organizations.

Corporate pension trust fund

holdings are also not included in data shown here; at the end of 1966
they held $432 million of non-guaranteed Agency debt, spread over the
full maturity range.

Finally, this category includes holdings of

investors belonging to groups specified in the Survey but not reporting
to the Treasury.

At the end of 1966, non-reporting commercial banks

alone held $1.6 billion of non-guaranteed Agency debt.
1/ If holdings of participation certificates are added to nonguaranteed Agency debt, the percentage is 14 per cent rather
than 10 per cent.
2/ Bank dealer holdings are included in commercial bank data.

-51-

This amalgam of investors has absorbed over one-half of the
rise in non-guaranteed Agency debt since the end of 1961.

The group

accounts for a larger share of outstanding Agency debt than earlier
in the sixties, except in the over 5 year maturity area.

Other

investor holdings have also risen sharply in the U. S. Government
securities market during the sixties, as shown in Table 6.

Such a

rise is typical of periods of tight money.
"All other investors" account for roughly the same share
of short-term Agency issues outstanding as they do of Treasury bills.
They account for a significantly higher share of Agencies than of
Treasury coupon issues, however, particularly in the long-term
maturities (62 per cent as versus 37 per cent).

It would appear that

the larger institutional investors reporting in the Survey find
long-term Agency issues less desirable investments than Treasury
issues, though this difference is diminishing.

It undoubtedly relates

to lesser marketability, in part due to the small size of individual
Agency issues.
The sharp rise in non-guaranteed Agency debt issued to the
public during 1966 was absorbed entirely by "all other investors",
whose holdings rose by almost $4 billion.

This

than 50 per cent rise in the group's holdings.

represented a more
"All other investors"

absorbed a large volume of U. S. securities during 1966 as well.
The rise in Governments held by them was a sizable $5 billion,
although this represented only a 7 per cent rise in their holdings.
One result of this was a sharp jump in Agency yields relative to
Governments.

Table 6
OWNERSHIP OF AGENCY AND U. S. DEBT BY ALL OTHER INVESTORSNonguaranteed Agency Debt due:
After 5
Within
1 - 5
Total
years
1 year
years

Within 1 year
Bills
Other

U. S. Marketable Debt due:
1 - 5
After 5
Total
years
years

Millions of dollars
Dee. 31, 1961
1962
1963
1964
1965
1966
June 30, 1967

1,890
2,302
2,990
3,585
4,286
7,461
6,298

1,147
1,062
1,146
L,505
1,827
2,245
1,957

1,315
1,565
1,561
1,291
1,007
1,202
1,164

4,353
4,931
5,698
6,382
7,121
10,909
9,418

20,596
23,933
25,246
27,541
29,088
32,647
27,339

10,234
8,292
7,700
5,570
6,268
6,038
4,824

15,827
16,121
13,623
15,863
15,784
18,895
19,929

13,780
15,813
17,599
18,367
17,534
16,109
14,617

60,438
64,159
64,167
67,341
68,674
73,689
66,709

43
40
41
49
47
60
57

51
52
54
55
55
61
59

69
68
70
65
63
67
62

51
49
49
53
51
61
58

52
54
55
57
58
64
67

37
36
41
34
36
37
39

29
33
28
33
36
39
37

37
35
36
37
35
37
37

38
40
40
41
43
46
46

Per cent of publiclyheld debt
Dec. 31, 1961
1962
1963
1964
1965
1966
June 30, 1967
1/

"Not including Official Accounts.

SOURCE:

Treasury Survey of Ownership.
i
vn
r\)
I

-53-

It is interesting to note that the need for "all other
investors" to absorb such a large volume of securities derived from
different factors in the two markets.

In the Agency market, all

investor groups reporting in the Treasury Survey with the sole exception of nonfinancial corporations purchased Agency issues on balance
during 1966; the cause of the large rise in "all other investors "
holdings was the sharp increase in Agency debt issued to the public.
In the Treasury market, on the other hand, official purchases more
than absorbed the rise in Treasury debt.

But every reporting investor

group sold Governments during the year which—at higher yields--were
absorbed into "all other investors" portfolios, presumably for the
most part those of individuals.

So far in 1967, with the easing of

credit stringency, "all other investors" holdings of Agency issues,
as well as of U. S. Government securities, have been reduced well
below end of 1966 levels.
Prior to late 1966, only a nominal amount of Agency debt
was held outside of the public's hands.

The Federal Reserve currently

does not own Agency debt outright though it is legally authorized
to do so.

Since late 1966 it has bought Agency issues from security

dealers under repurchase agreements.

Treasury trust fund and agency

acquisitions of Agency issues were undertaken on a large scale beginning
in August, 1966, as a means of alleviating the congestion in that
market.

Since then, Treasury accounts have acquired some $1.7 billion

of non-guaranteed issues, almost entirely through direct allotments
at the time of Agency financings.

These data are shown in Table 7.

Table 7
OWNERSHIP OF AGENCY AND U. S. DEBT BY OFFICIAL ACCOUNTSNonguaranteed Agency Debt due:
After 5
Within
1 - 5
Total
1 year
years
years

U. i3. Marketable Debt due:
Within 1 year
After 5
1 - 5
Bills
Other
years
years

Total

Millions of dollars
Dec. 31, 1961
1962
1963
1964
1965
1966
June 30, 1967
If

29
11
45
1,043
1,169

- -

313
568

—
—

— ~

29
12
45
1,356
1,738

3,932
3,588
5,512
7,795
10,314
13,869
17,866

14,970
15,744
18,912
15,325
15,884
24,277
18,571

10,597
12,259
10,568
15,986
17,253
11,224
18,058

7,866
8,867
10,490
10,084
10,723
9,503
9,672

37,365
40,458
45,482
49,190
54,174
58,873
64,167

Includes Federal Reserve and Treasury trust fund and agency holdings.

SOURCE:

35
- -

Treasury Survey of Ownership.

i
frI

-55-

Participation Certificates
Participation certificates issued by FNMA as trustee
first offered to the public in late 1964*

were

During 1964-1966, since all

PC's were in registered form, FNMA collected ownership data which
were presumably complete as to coverage of specified investor groups.
These data, as of the end of each quarter, are presented in Table 8.
It can be seen that during the past few years commercial banks
accounted for some 10-20 per cent of FNMA PC's outstanding.
end of 1966, their holdings totaled $224 million.

At the

Both nonbank

financial institutions and state and local governments accounted
for a larger share of the PC's outstanding, not surprisingly since
PC's are concentrated in intermediate- and long-term maturities.

holdings of nonfinancial corporations are quite small, aggregating
only about $50 million.

Registered individual holdings are also

quite small, as the Table shows, but a large part of the "all other
investors'1 category is undoubtedly accounted for by individual trust
funds

managed by banks.
These FNMA ownership data cease with the institution of

bearer-form PC's in 1967.

But beginning in December, 1966, the

Treasury Survey of Ownership includes PC's by issuer—FNMA, ExportImport Bank, and CCC.

Unfortunately the Survey carries no maturity

breakdown on the PC's nor a separate listing of each issue, limiting the
usefulness of the data.

There is, for example, no way of isolating

ownership of only the recent large (non-serial) issues of FNMA or the
fully marketable Export-Import issues.
Survey data on PC's.

Table 9 presents the Treasury

Table

OWNERSHIP OF FNMA PARTICIPATION CERTIFICATES—'
(End of quarter data)
Nonfinaneial Corps.
General
Pension
Funds
Funds

Commercial
Banks

Nonbank
Financial
Institutions

66
57
61
152
176
144
234
226
224

67
66
62
165
212
220
328
312
291

8
7
7
7
8
10
28
35
29

5
2
2
10
10
11
9
17
21

17
23
25
83
154
177
321
329
338

2
2
2
4
7
8
37
45
47

135
142
142
405
607
601
1,156
1,110
1,070

300
300
300
825
1,170
1,170
2,110
2,075
2,020

22
19
21
18
15
12
11
11
11

22
22
21
20
18
19
16
15
14

3
2
2
1
1
1
1
2
1

2
1
1
1
1
1
1
1
1

6
8
8
10
13
15
15
16
17

1
1
1
11
1
1
2
2
2

45
48
47
49
52
51
55
54
53

100
100
100
100
100
100
100
100
100

State & Local
All other
Individuals investors 2/
Governments

Total

Millions of dollars
1964 - 4
1965 - 1
2
3
4
1966 - 1
2
3
4
Per cent of publiclyheld PC's
1964 - 4
1965 - 1
2
3
4
1966 - 1
2
3
4
1/
2/

Includes all PC's issued by FNMA as trustee.
Includes partnerships such as investment funds, dealers, and brokers; a major share is held by bank nominees,
including some corporate pension funds.

SOURCE:

Federal National Mortgage Association, Office of Secretary-Treasurer.

Table 1
OWNERSHIP OF PARTICIPATION CERTIFICATES
(End of quarter data)

Commercial
Banks

Nonbank
Financial
Institutions

Nonfinaneial Corps.
General
Pension
Funds
Funds

State & Local
Governments

Official
Accounts

500
900

All Other
investors

Total

FNMA PC1s
Millions of dollars
1966 - 4
1967 - 1
2
Per cent of publiclyheld PC's
1966 - 4
1967 - 1
2
EXPORT-IMPORT PCfs
Millions of dollars
1966 - 4
1967 - 1
2
SOURCE:

92
469
924

103
292
505

26
56
39

79
102
112

227
304
456

5
18
24

5
11
13

1
2
1

4
4
3

11
12
12

271
613
686

32
91
133

60
69

14
18
18

1
56
93

- -

1,493
1,397
1,894

2,020
3,120
4,830

74
53
48

100
100
100

815
747
743

1,135
1,583
1,742

Treasury Survey of Ownership.

i
vn
I

-58For the end of 1966, all specified investor groups show
smaller holdings than in Table 8, and account for a lesser market
share.—^ Focusing on changes in ownership since the end of 1966, the
Table shows a considerable rise in PC portfolios of all investor
groups with the $2.8 billion rise in FNMA PC's outstanding.

To some

degree this rise in holdings of the larger institutional investors
related to the improved marketability of the new FNMA PC's issued
in 1967.
The increases in holdings were largest for the two investor
groups that experienced marked improvements in liquidity over the
period, namely commercial banks and nonbank financial institutions.
As a result, the share of FNMA PC's in public hands held by reporting
commercial banks increased from 5 per cent at the end of 1966 to
24 per cent in mid-1967 while the nonbank financial institutions'
share rose from 5 per cent to 13 per cent.

Shares of nonfinancial

corporations and state and local governments remained virtually
constant, so that the share of "all other investors" dropped from
75 per cent to just under 50 per cent*
Comparisons of FNMA PC ownership with the ownership
structure of non-guaranteed Agency issues and also Treasury issues
can be made, but not with any degree of precision.

Comparative

The single exception is corporate pension trust funds which show
larger holdings on the Survey than on FNMA data. This is because
banks manage some of the pension funds and they are to this degree
included in the all other category of the FNMA data.

-59-

ownership data as of December 31, 1966 are shown in Table 10 but differences among the markets in the percentages of debt held by the various
investor groups may not be meaningful.

In tihe first place, the

percentages will vary with the average maturity of debt in each
market.

Secondly, the share of debt accounted for by non-reporting

institutions differs among the markets.

Allowing for these complica-

tions however, there do appear to be some significant differences
in the degree to which some investors participate in the various
markets.
Table 10 shows the relative ownership of reporting commercial
banks to be less in PC's than in non-guaranteed Agency issues and
sharply less than in U. S. debt (lines 1-3).
reflects the longer average maturity of PC's.

To a small degree this
Utilizing data for

all commercial banks does not alter the general pattern.

Holdings

of all commercial banks at the end of 1966r^ accounted for 36 per cent
of publicly-held U. S. marketable debt, about 25 per cent of Agency
debt, and only 11 per cent of FNMA PC's.

It is clear, however, that

banks hold a larger share of the new, more readily tradeable PC's,
It is not clear whether all nonbank financial institutions
hold relatively fewer PC's than Agencies or Governments,

It is

apparent from the Table, however, that the larger, reporting
institutions hold considerably fewer PC's.,

They account for 5 per

cent of the PC's outstanding, as versus about 10 per cent in the other
1/

Call Report data.

Table 10
COMPARATIVE OWNERSHIP STRUCTURE, DECEMBER 31, 1966

Item

(1)

U. S. Marketable 1/
Mil $

All
Others

Total Held
by Public

6,323
4

15,384
10

73,689
46

159,152
100

2,997
17

1,892
11

715
4

1,379
8

10,909
61

17,893
100

92
5

103
5

26
1

227
11

1,572
78

2,020
100

FNMA PC's 2/
Mil $
7o

224
11

291
14

29
1

338
17

1,138
56

2,020
100

1/
2/

State &
Local
Govts.

16,573
10

(2) Non-guaranteed 1/
Mil $

(4)

Nonbank
Nonfinancial
Financial
Corporations
Institutions (general funds)

47,183
30

(3)

Commercial
Banks

FNMA PC's 1/
Mil $

Treasury Survey Data.
FNMA Data

-61markets; and given the longer maturities of the PC issues one would have
expected them to account for a greater market share.

As noted earlier,

these institutions increased their PC holdings sharply during the first
half of 1967 and they accounted for 13 per cent of outstanding PC's
by June, 1967.

It would appear that the small size of the serial

PC's issued before 1967, and perhaps their registered form, detracted
considerably from their appeal to the large institutions.
Nonfinancial corporate and state and local government
investors hold roughly the same relative share of each debt instrument,
allowing for maturity differences.

It is interesting to note, however,

that the smaller, non-reporting corporations appear to hold virtually
no FNMA PC's, as indicated by the similarity of their holdings in
lines 3 and 4 on the Table, whereas such corporations apparently
account for 1/2 of U. S. debt holdings of corporations.
Finally, because banks and nonbank financial institutions
account for a smaller share of the market for PC's, the "all other
investors* category is considerably larger.

Using all commercial

banks, "all other investors" accounted for some 72 per cent of PC's
outstanding at the end of 1966, compared with 52 per cent of nonguaranteed Agencies and 40 per cent of U. S. Governments.
Treasury trust funds and agencies acquired FNMA PC's for
the first time during January-June 1967.
$900 million of the PC's.

At mid-1967, they held

As with their non-guaranteed Agency debt

holdings, the issues were acquired through direct allotments at the
times of financings rather than through market purchases.

-62-

Treasury Survey data on the ownership of Export-Import Bank
PC's (see Table 9), as noted earlier, include all PC's outstanding,
of which only $900 million are fully marketable.

Prior to 1967,

Export-Import PC's were sold to commercial banks, who in turn could
distribute them to specified institutional investors.

As a result,

the bulk of these PC's are held by banks and some investors held none
of the issues, such as nonfinancial corporations and state and local
governments.

The $900 million of Export-Import PC's offered since

February 1967 are fully marketable.

It is probably not far wrong

to assume that the rise in holdings of nonfinancial corporations
(general funds), state and local governments and nonbank financial
institutions since the end of 1966 has been in these fully marketable
issues.

If so, at mid-1967 these groups, respectively, accounted for

8 per cent, 10 per cent and 11 per cent of the new Export-Import PC's.
These market shares

are very close to FNMA PC shares, except for

nonfinancial corporations who hold a relatively greater amount of the
Export-Import PC's.

-63-

III.

HOMOGENEITY OF AGENCY SECURITIES

A question of particular importance to participants in the
Agency market--lenders and borrowers alike--is whether there is a
single homogeneous market for all Agency securities or whether, conversely,
the securities of each Agency form smaller, distinct markets.

A related

question is whether single Agency issues with small amounts outstanding
differ from larger Agency issues in a manner that might imply less
marketability.
To answer these questions data were assembled on comparative
yields and ownership, by issuing Agency and by issue size.

These

data show the Agency market to be homogeneous from the standpoint of
securities of individual Agencies, that is, no consistent and significant
differences were found in market yields or in ownership of the separate
Agency securities.

With respect to Agency issues of small size, it

would appear--on the basis of limited data--that their quoted yields
often vary quite widely off the yield curve and that they are to a
greater degree than the large issues lodged in the portfolios of
comparatively inactive investors.
A.

Homogeneity by Issuing Agency

Even within a homogeneous market differences in yield and
ownership will exist among various issues, depending on their maturity
and on less important attributes such as their coupon rate.

As pointed

out earlier, the maturity characteristics of Agency debt differ widely
by issuing Agency:

all debt of the Banks for Cooperatives and of

the Federal Intermediate Credit Banks matures within 1 year, the Federal

-64-

Home Loan Banks issue some debt with maturities of more than 1 year,
and debt of the Federal National Mortgage Association and Federal Land
Banks is more heavily weighted in the intermediate- and long-term
maturity sectors.

To abstract from these maturity differences,

ownership data are classified by maturity category and curves relating
the yield of every Agency issue outstanding to its maturity date
are plotted for selected periods.

Yields.

The accompanying Charts 9-13 shew plots for Agency yield

curves in which the securities of the separate Agencies are differentiated.
The curves include all outstanding Agency issues, shown as of the end
of May for the years 1961 through 1967.

The market yields that were

utilized are those published in the daily quote sheets of the Morgan
Guaranty Trust Company.—' For the 1966 and 1967 dates, curves were also
drawn showing after-tax yields to corporations in order to adjust for

2/
relative coupon size.—
Inspection of the Charts yields several general impressions:
(1) the yield curves are rela tively smooth, although less so than the
yield curves for U. S. Government securities; (2) the degree of smoothness varies considerably over time and by maturity area; (3) the yield
differences among issues of comparable maturity are, as often as not,
among issues of the same Agency and there are not consistent differences
1/ Differences among dealers in published Agency yield quotations on
specific issues are quite sizable. On one observation date they
rangedup to 40 basis points (see Appendix Chart 1).
2/ When market security prices are below par, given the same market yield
and maturity issues with high coupons are less attractive to investors
than those with low coupons because capital gains are taxed at a
lower rate than interest income. Issues with relatively low coupons
thus generally carry lower before-tax market yields.

CHART
YIELDS

AGENCY

SECURITIES

flAY

JLf

1967

tourJ!$

Morga^%uaranty?Sl?rust Company,

quo

CHART 10
YIELDS

OF AGENCY

SECURITIES,

» T 3 1 t

1966

CHART 11
YIELDS OF AGENCY SECURITIES, MAY 1963, 1964 and 1965

CHART

YIELDS OF AGENCY SECUBITIES, MAY 31,1962

PERMIT

i «
3.50

3.00~

t:i t

f - z :

W
L...

w
i

n a s i

2.00

i
ON
CO

nay
1963
Source:

> i t it. ill i ifi i i>
9

»65

'66

'67

Morgan Guaranty Trust Company, daily

'68

'69

quotation sheets.

'70

•71

•72

•73

•74

•75

CHAffT 13
YIELDS

AGENCY

SECUHITIES,

MAY 31,

1 9 &

PI i

BI3
4.50-

4.00-

3.5L>-

3.0(r

2 , 5 ( r

xh:

i ±
- r w x t i
j

•

2.CX!-

" '

1 , 1

1,1

Source:

1,1

•"

Morgan

ii It

ii f
i
r. i
f
f
•65
66
'6?
68
Guaranty Trust Company, daily quotation sheets•

'69

•70

iii in iii
•72

'73

-70-

over time in the yields on the securities of one Agency as versus
those of another Agency.
The divergence among yields on Agency issues of similar
maturity has on several of the observed dates been as large as 50
basis points, and has not uncommonly been around 25 basis points.
Such a divergence persists, and to an even greater degree on longerterm issues, when after-tax yields to corporations are utilized.
The dispersion of Agency yields around the yield curve is
clearly greater than in the U. S. Government securities market.
In Chart 14, before- and after-tax yield curves for U. S. Government
securities are plotted for the end of May, 1967, utilizing Morgan
Guaranty quote sheets.—^

When the low coupon issues are excluded,

the yield curve for Governments is smoother than for Agencies.
The maximum divergence of any specific issue from the curve is no
more than 15 basis points and is usually considerably less.

Inclusion

of the low coupon issues (the 2-1/2 per cent bonds) increases the
dispersion of yields around the curve--before-tax yields of such
issues are below the curve and after-tax yields above it.
The smoothness of the Agency yield curve has varied rather
widely over the 1961-1967 period.
smooth.

During 1961-64 it was comparatively

As Agency yields began to rise more rapidly in 1965 the

dispersion of yields around the curve increased, particularly in
the short-term maturity range.
1/

By May of 1966, when congestion in

It appears that the yield curve for U. S. Government securities is
less smooth using the yields quoted by one dealer than when composite
yield quotations from all dealers are utilized.

CHART
YIELDS OF TREASURY SECURITIES, MAY 29, 196?
(Morgan ftnaranty Trust Co. daily quote sheets^
PEftGEN2
W! i- !
i !

jSHEjES^n&XIZIEGD rs
DEk & A A . XI T- Cl LI

!! : I
ir
i !

1**5CL

:33C

M i l

-I i I. Mil

it at

i J'

jio t
r -

i
ELL

• ' 1 n. I ITT Mil
May
1968
•69

»i in ii
•70

uLL
•72

'73

•7^

•75

•76

•78

•79

•8o

-72-

the Agency market was nearing a peak, the dispersion of yields was
marked.

With the improvement in the Agency market in late 1966- early

1967 the yield curve was, by May of 1967, again relatively smooth.
In general the Agency yield curve has been smoothest for
issues due within 1 year, particularly after allowance is made for
differences in coupons.

This is despite the fact that it is only in

the within 1 year maturity range where all the Agencies issue securities.
Differences in yields on Agency issues of comparable maturity,
as often as not, arose among securities of the same issuing Agency.
No consistent differences among yields on securities of the separate
Agencies were evident during the sixties.

And in 1967 there was no

divergence between after-tax yields on participation certificates
(FNMA or fully marketable Export-Import Bank PC's) and on regular
Agency issues of comparable maturity.—^
At any one point in time there have been differences among
yields on the various Agency securities*

Such deviations can be

related primarily to differences in coupons.

Particularly in the

shorter-term maturity area, coupons on the various Agency securities
can diverge widely when current interest rate levels are high or low
relative to past interest rates, since short-term issues of some
Agencies (FNMA, FLB and to some degree FHLB) often originated as longterm issues that have approached maturity with the passage of time
while other Agency debt issues (primarily FICB and Banks for Cooperatives)
are always marketed close to current interest rates.

The participation certificate yields studied were only those of the
large and fully marketable PC's, first issued in early 1967.

-73The comparative yield differences that stand out in the
accompanying Charts can usually be traced to such a movement in
interest rates and coupons.

The shorter-term portion of the yield

curve was relatively smooth in the early sixties.

But as Agency

yields rose to fairly high levels relative to earlier periods, the
yields on FNMA and FLB issues, carrying comparatively low coupons,
moved below other short-term Agency yields.

Such a divergence first

appeared in the yield curve for 1964 and it became more pronounced
in 1965 and 1966.

Using after-tax yields for 1966, of course,

the divergence was erased.

And by May, 1967, when yields and coupons

on newly-issued Agencies were considerably reduced, the yield curve
was again relatively smooth even on a before-tax basis.
The yield curve for 1967, before tax adjustments, shows
market yields of regular FNMA debt to be generally above comparable
FLB debt and yields of participation certificates to be consistently
above regular Agency debt of similar maturity.

Both of these

seeming divergences represent, at least in part, coupon differences,
1/
however, and after-tax yields of these issues are equalized.Ownership.

In a homogeneous market one would expect to

find the different ownership groups holding

roughly the same proportion

of the various securities outstanding, abstracting from maturity differences.
Were this not the case, securities of some Agencies would at times be
subject to interest rate pressures that differed or were absent for
other Agency issues.

Table 11 presents the percentage of publicly-

owned debt cf the various Agencies held by the large investors in the

1/ The after-tax yields shown overstate the true impact of couoon
rates on yields since they utilize the corporate tax rate which is
higher than the marginal tax paid by many investors.

Table 11
OWNERSHIP BY ISSUING AGENCY
(Per cent of Publicly-held Debt)
Issues Due in 1-5 Years

Issues Due in 1 Year

Commercial Banks
Dec* 31, 1961
1962
1963
1964
1965
1966
Mar. 31, 1967
Nonbank Financial
Institutions
Dec. 31, 1961
1962
1963
1964
1965
1966
Mar* 31,. 1967
Nonfinancial
Corporations
Dec. 31 # 1961
1962
1963
1964
1965
1966
Mar. 31, 1967

Bks. for
Coops

FIC3

FHLB

FNMA—^

26 o 9
39.0
30.1
28.3
28s 6
24.8
23.8

22.4
30o7
29.5
26 e 8
27.3
22,0
23.7

27.7
313 4
31«2
27.4
23.4
17.0
15.5

6.7
8,3
10 o 5

8c0
7.7
8C2

7.6
6.6
7o7
8.0
8C4
7.9
901

26.7
14.7
14 06
10.2
9.7
407
30 6

19.1
14«0
13.0
902
8.3
3.7
3c4

80O

FLB

FHLB

FNMA

FLB

10o5(17el)
12.2(23.2)
23.3(23.3)
17.7(12.2)
1536(21.4)
130i(17ol>
11.3(16.1)

21.4
35o6
27.9
25.1
25«6
21.7
21.7

32.0
35.4
26c 7
17.7
24,0
15.3
21.7

22.8
24.4
21.8
1SC6
13.7
14.2
14.6

100 4
12.4
10e 3
12.4
12*3
11.3
12.0

11*3(18.0)
8.7(15.6)
11.9(11.9)
8o5(16*3)
7.0(14.9)
7o7( 8 C 0)
10,8( 9.2)

8.1
10.8
12 .2
8.8
7.7
80O
8C2

31.2
18.3
28.3
23„5
19 c 3
14.0
16.3

24.4
18.3
19.3
10,4
9.1
4,1
3.3

22.0(
l6 o 0(
4*i<
3»1(
13o6(
8.1(
2.8(

5.5
7*9
</
;©•*-i
2.5
7,1
5o3
3„9

0.8
8.6
4.3
5.2
6.5
3.1
2.1

4c9)
4.0)
4d)
~ )
7.5)
4.9)
4.2)

Issues Due After £ Years
FNMA

FLB

27.1
24,4
25.2
21.9
19.5
17.9
18.8

5.3
5o9
4,8
4.4
3.6
3.9
5.0

3,9
5.4
6„6
6.3
8.9
8.8
14.6

19.6
20ol
23.1
24.7
20.5
17.9
17.9

14.6
14.3
12,5
13.8
13.3
11.9
11.6

22.9
19.6
16.0
16 „ 3
15.0
13.7
14.9

9.6
10.2
9.8
10,0
11.4
9.3
12.6

3.8
4.7
1.3
1.9
2.2
3.7
4.1

1.8
2.1
1.9
2.6
2.6
2.0
2.2

0.5
0.4

0.2
0.6
0.4
0.3

—

0.1
1.6
1.5
1.8

—

0.6
0.6

State & Local Govts.
Dec, 31, 1961
1962
1963
1964
1965
1966
Mar. 31, 1967

Other Public Investors
Dec. 31, 1961
1962
1963
1964
1965
1966
Mar. 31, 1967

1.1
2.0
2.4
5.4
5.9
6.0
8.7

3.0
4.7
5.1
5.6
10.0
10.2
7.7

2.2
2.1
2.2
3.5
6.5
4.1
3.7

11„4( 5.1)
17.4(19.2)
9,9( 9.9)
28.1( 9.2)
39.4( 1.0)
15.2( 5,9)
22.1( 3.4)

5.7
2.3
2.4
3.3
4.3
4.8
5.7

3„2
0C6
0.8
2.7
3.0
3.4
1.8

3.0
3.7
1.6
2.1
5.0
4.2
4.5

2.8
1.7
1,5
4.7
7.8
8.5
8.4

5.8
7.6
8.4
10.7
12.0
12.5
11.7

9.4
10.2
10.5
17.7
15.6
12.0
10.9

36.3
36.0
40o4
47.5
46.7
56.9
55.7

46.9
43.9
44.3
50.2
45.3
56.2
56.1

34.7
35.7
36.8
46.3
48.3
63.6
65.5

44,2(54.6)
45.8(38.0)
50.9(50.9)
42.7(62.2)
24.3(55.2)
55.9(64.0)
53.0(67*1)

59.2
43.3
52.4
59.9
55.4
60.2
60.6

32.8
37,1
39.5
51.0
47.3
64.1
58.0

50.8
47.2
52.1
52.7
58.6
60.0
58.9

53 o 8
57.4
58.9
57.1
56.8
59.7
59.0

65.6
66.5
70 C 8
68«5
67.9
68.5
66.7

76.8
73.6
72.7
65.6
64.1
69.3
61,3

I
Percentages in parentheses were computed excluding the FNMA discount notes, issued to investors on demand.

Source:

Treasury Survey of Ownership*

i
i

-76-

major ownership groups:

coitimercial banks, nonbank financial institutions,

nonfinancial corporations, state and local governments, and all others.
The ownership percentages are shown for three maturity groupings
(debt due within 1 year, in 1-5 years, and after 5 years) and for
the years 1961 through 1967.
The patterns of ownership show a remarkable similarity.
Moreover, variations in ownership were greater in the early sixties
than during the last few years.
Small differences in the ownership percentages should be
disregarded for a number of reasons.

In the first place, small

percentage variations in ownership in many of the categories involve
only minor dollar amounts.

Secondly, the maturity categories are

quite broad and the average maturity of issues of the separate
Agencies will differ within any of the maturity groupings.

Thirdly,

holdings of a specific Agency security by an ownership group may at
times be less than what the group desires to hold if the Agency was
retiring debt when the ownership group was able to buy and issuing
debt when the ownership group did not have funds available for
investment.

It may be for this reason, for example, that commercial

bank holdings of FHLB issues were relatively low in December, 1966
and March, 1967.

Finally, since the percentage of debt held by

certain ownership groups varies considerably by maturity (for
example, commercial banks hold considerably more short-term debt
than long-term), issues that were originally long-term but that have
passed with time into the shorter-term maturity categories should

-77-

have ownership patterns that vary from original short-term issues.
The variance in ownership of such issues should be biased in the
direction of the ownership percentages for long-term Agency issues.

this regard, the ownership percentages for FLB and FNMA issues in the
1-5 year sector should differ from 1-5 year FHLB issues in the
direction of the over 5 year percentages, and the ownership
percentages of FLB, FNMA and FHLB in the within 1 year maturity
sector would likewise vary somewhat from those of FICB and Bks. Coops.
After allowing for some variation in the percentages for the
above reasons, there remain two cases in which differences in ownership among Agency securities may be significant, though even in these
cases the differences are small.

Nonbank financial institutions

hold a larger share of FNMA and FHLB issues than of other Agency
securities.

This is not surprising, however, in view of the relation-

ship of the Federal Home Loan Banks to savings and loan associations
and of the active participation of most financial institutions in the
mortgage market in which FNMA also plays an active role.

Secondly,

FNMAappears to be held to a lesser degree by commercial banks than
are other Agency securities—^; even this difference could be related
to the fact that FNMA did not issue a single security over the entire
1962, Q4-1965, Q3 period,
B.

Homogeneity by Issue Size

The size of individual issues in the Agency market ranges
widely.

On May 31, 1967, single non-guaranteed Agency issues had

anywhere from $60 million to $700 million outstanding and from $60
1/

The relevant ownership comparison is for FNMA issues not including
discount notes, that is, the figures in parentheses on Table 11.

-78-

million to $535 million held by the public.

With but one exception

the issues of the Federal Intermediate Credit Banks and of the Banks
for Cooperatives had amounts publicly-held in a $243-403 million range.
Federal Home Loan Bank issues ranged in general from $250-535 million.
All of the Agency issues with less than $100 million outstanding were
obligations of the Federal National Mortgage Association and of the
Federal Land Banks.

While FNMA and FLB issues range up to $400 million

(publicly-held) in size, the large number of small issues makes the
average issue size of these two Agencies significantly less than for
the other Agencies.
The size of individual Agency issues is considerably less
than in the U. S. Government sedurities market, and there are more
individual Agency issues outstanding than there are Treasury coupon
issues.

As of the end of May, 1967, there were 49 Treasury coupon

issues outstanding (excluding the 1-1/2 per cent notes), as compared
with 70 separate issues of the five large Agencies (not including
PC's).

The average size of Treasury issues was $3.1 billion in terms

of total outstanding and $1.1 billion in terms of publicly-held portions.
There were only five Treasury issues for which the amount held by the
public was below $1 billion.
Table 12 shows the number of Agency issues outstanding,
their average issue size and the range of issue size for selected dates.
It is clear from the Table that the sizable growth in Agency debt from
the mid-1950's has been through a growth in the number of issues as
well as in their size.

In the five years from May, 1955, to May, I960,

Table

SIZE AND NUMBERS OF INDIVIDUAL AGENCY ISSUES

FICB

Bks. for Coops

FHLB

FNMA

FLB

All
Aqencies

May 31, 1967
No. issues outstanding
Average issue size 1/
Range of issue size 1/

4
9
2/
257
365(343)—
236-465(161-403) 243-275

11
17
465(400)
196(168)
185-700(185-535) 63-550(63-400)

29
158(152)
60-341(60-341)

70
248(226)

May 31, 1960
No. issues outstanding
Average issue size 1/
Range of issue size 1/

9
174
137-210

3
110
92-138

6
199
105-351

17
164
90-797

21
102
60-154

56
146
60-797

May 31, 1955
No. issues outstanding
Average issue size 1/
Range of issue size 1/

4
65
41- 91

3
37
30- 40

2
71
60- 81

1
570
570

7
151
71-228

17
126
30-570

Millions of dollars.

Figures in parentheses show amounts held by the public.

Source:

Morgan Guaranty Trust Company, daily quotation sheets.

60-700(60-539

-80-

the number of individual Agency issues rose from 17 to 56 while the
average issue size increased only slightly--and in fact decreased
for FNMA and FLB.

Between 1960 and 1967, cn the other hand, the growth

in the number of issues tapered down while the average issue size
increased significantly, from $1146 million to $2i|8 million.
Since Agency debt is sure to grow further, any differing
market characteristics of the small as versus larger issues should
be of interest to the Agency debt managers, as well as to
private and official investors.

The evidence gathered for this

study suggests that there is indeed a difference in the marketability—^
of the small issues when compared with larger Agency issues.

the basis cf one observation in 1967, it appears that for the small
issues (1) quoted yields vary more widely off the yield curve—

and

(?) the more active market participants hold them to a somewhat lesser
degree.
Chart 15 shows for May 31, 1967, the plots for before- and
after- tax yield curves differentiating the Agency debt by issue size.
Issues of the Federal Intermediate Credit Banks and of the Banks for
Cooperatives are not included since they vary only slightly in size.
The Chart shows clearly that, at least on this one date, the issues
whose quoted yields verged most widely off the yield curve were issues
of $100 million or less.
1J

This stands out particularly after allowance

Less marketability of the small issues need not imply diminished
liquidity. If small issues are held tightly in investor portfolios
the offered side of the market would be weak bur. the bid side
need not be, i.e., a ready market might exist for sales of the issues.
In addition, spreads in quoted yields among dealers appear, for
the debt of some Agencies, to be greater on the small Agency issues.

CHART 15
YIELDS

AGENCY

SECURITIES

ISSUE

SIZE,

HAY

31t

PERI

JyUTU
rii ISLSIftrLl lEI

5.00

tsl

1967

D
' x i \

Jf.50
H i )

4.00

3.50 \Sh
per<WS!4

fii

3.00

w l

n li j '

a M K M S

i t
1

lajiir

:i m -

c>f
H i

2.50

JD2J

2.00
n

I 111 I

il III I III I III I III Mil

III ll

Lit

JCSt

•

Li-

J
l
lt£9B

IS:

H
I
11 I III I III I III Mil Mil l III Mil 1 III I in III Mil I 111 Mil Mil I li

ll III 111* III III III It,

May

Source ^Morgan $?aranty ?rust Companyf daiiy quotatZJn sheetZ^

Nay

•77

•78

'79

'80

-82-

is made for differences in coupons.

Yields on the small issues lay

above as well as below the yield curve, and accounted for a sizable
portion of the lack of smoothness of the yield curve.
Ownership data also appear to point to a lesser tradeability
of the small Agency issues.

In Table 13, the ownership of FLB and

FNMA debt on March 31, 1967, is shown differentiated by issue size.
In all but one case (FNMA debt due within 1 year) the larger commercial
banks and nonfinancial corporations reporting in the Treasury Survey
held a greater share of the large Agency issues than of the small issues.
Investor groups who are probably less active on the buy and sell side
taken together--nonbank financial institutions and state-local governments --sometimes held greater and sometimes lesser portions of the
small issues.

And other investors, considerably less active in the

market than the above groups, held higher--and sometimes sharply
higher--shares of the small issues.
With relatively large institutions often holding

only

minimal amounts of the small Agency issues, it is indeed likely that
buy orders for these issues--particularly of any size--would be filled
only with difficulty, and probably at sharply rising prices.

The

volume of trading in the market and dealers' positions (particularly
gross short positions) would also be smaller than for a comparable
amount of larger-sized issues.

It seems clear, then, that Agency

debt could be made more attractive to at least some investors were
the size of individual issues increased.—^
1/

It will be recalled that financial institutions increased their
portfolios of PC's sharply in 1967 with the introduction of
relatively large-sized issues.

Table 13
OWNERSHIP BY SIZE OF AGENCY ISSUE, MARCH 31, 1967
(Per cent of Publicly-held Debt)
Commercial
Banks

Nonfinaneial
Corporations

Nonbank Financial
Institutions

FLB Debt
Due within 1 year
Issues of < $100 million
Issues of $100-199 million
Issues of > $199 million

8.8
20<>S
23,5

0.7
3.4
5.1

llc6

Due ia 1«5 y^ars
Iss*
of < $100 million
Issues of $100-199 million
Issues of > $199 million

llo9
21.5
24.3

Due after 5
Issues of
Issues of
Issues of

years
< $100 million
$100-199 million
> $199 million

FNMA Debt
Due within 1 year
Issues of < $100 million
Issues of $100-199 million
Issues of > $199 million

Due in 1~5 years
Issues of < $100 million
Issues of $100-199 million
Issues of > $X99 million
Due after 5
Issues of
Issues of
Issues of

years
< $100 million
$100-199 million
> $199 million

State and Local
Governments

Other
Investors

8.4

13.6
6.5
4.2

65.3
6U2
58 e9

1.4
2.8
2.4

9.5
12.9
11.6

17*3
5.6
2.9

59.7
57.1
53.8

8o0
26,7

0.8
0.2

903
18.6

13.9
5.4

67,9
49.5

21c8
24.7
13.4

5o7
4i0

14.9
10.0
8.3

2.3
2,7
3.5

55.2
58o0
70.8

13c

71.3
57.6
52.4

66.7

502
16 c 3

4 , 0

O "

U o<£>

18 o a

1.4
3.2
6 c6

20o2

9.3
4.1
2.0

5.2

lc8

14.6

11.7

1 8 o 8

i
CO
w
I

-84-

At the present time, the size of newly offered individual
Agency issues is by and large determined by the operating needs of the
particular Agency in any month.

Should an Agency make the overt

decision to increase the size of its issues, it would probably
necessitate borrowing ahead of need and/or following a program of
reopening issues already outstanding.

Objections might be raised

by the Agencies to borrowing ahead of need, since in the short-run
it might cost them money.

However, another proposal has been advanced

that would eliminate this problem, namely the establishment of a
centralized borrowing authority for all, or some group, of the
Agencies.

Were centralized borrowing instituted, it would permit

larger--and more marketable--issues, it would reduce the number of
financings per month, and it would also enable a better co-ordination
of financings as between the Agency and U. S. Government securities
markets.

-85-

IV.

INDICATORS OF MARKET PERFORMANCE

The performance of the secondary market in Agency securities
may be evaluated in terms of several indicators.

They include the

volume of trading, the size of dealers1 positions, and the spread between
quoted bid and asked security prices.

In general, the marketability,

and liquidity, of a security is related to these indicators.

A large

volume of trading implies that investors are able to execute the transactions they desire at reasonable speed; and in addition it implies
that the market is "broad" since a sizable trading volume probably
reflects a large volume of orders on the dealers1 books from a wide
spectrum of investors.

The existence of dealers who take positions

in securities is a crucial aspect of the market, enabling investor
orders to be translated into transactions with speed, in size, and at
prices close to the market.

Finally, a small spread between bid and

asked security prices indicates dealers1 willingness to make markets
and, at the same time, induces investor participation in the market.
The remainder of this chapter analyses these three indicators
of Agency market performance, in isolation and in comparison with other
securities markets.

The analysis is in general confined to the

period beginning in 1960 due to the lack of available data prior to
the sixties.—^

Data are generally classified by term to maturity, since

the indicators are quite different in magnitude for short-term and longerterm Agency debt.

The data include fully marketable participation certif-

icates for the relevant periods.
1/

Some data on Agency trading and positions are available for the
1958-1959 years but they are not consistent with later data.

-86-

The Volume of Trading

Daily average trading in Federal Agency securities has
risen sharply from the early sixties.

By 1966-67 the volume of trading

stood at levels 3-4 times as high as in 1960-61.

Daily trading in

Agency securities due within 1 year averaged $56 million in 1960-61
and some $150 million in 1966-67.

Daily trading in Agency debt due

after 1 year, while considerably smaller, increased even more in
percentage terms--from $18 million on average in 1960-61 to $64
million in 1966-67.

Charts 16 and 17, Profile of Market Performance,

illustrate the marked secular rise in trading of Agencies; quarterly
data are presented

in Appendix Table 7.

The growth in trading activity in the Agency market has
outpaced that in other securities' markets, and by 1966-67 activity
in both the short-term and longer-term sectors of the Agency market
had surpassed trading volume in bankers' acceptances and in some
maturity sectors of the Government securities market.

Table 14

presents dealer transactions in Federal Agency debt, U. S. Government
debt and bankers' acceptances.

As the Table shows, trading in Agency

issues maturing within 1 year was, by 1966-67, some three times as
large as trading in bankers' acceptances or in Treasury issues maturing
after 10 years, somewhat larger than trading in Treasury issues
maturing within 1 year or in 5-10 years, and somewhat smaller than
trading in 1-5 year Treasury maturities.

Activity in Treasury bills

of course far outpaces any other market.

Trading in Agency issues

maturing in more than 1 year had, by 1966-67, risen to a level roughly
comparable
debt.

to trading in bankers' acceptances or in long-term Treasury

CHART 16
PROFILE OF MARKET PERFORMANCE FOR FEDERAL AGENCY SECURITIES MATURING WITHIN 1 YEAR
Quarterly data, 1960-67

Millions of dollars

2nds

I960

1961

1962

1963

196^

1965

1966

1967

1/ The spread as shown is for FHLB, FICB and Bks. for Coops, issues; spreads on PNMA and FLB issues are larger.
SOURCE: Appendix Tables 7-11

CHART 17
PROFILE

MARKET PERFORMANCE FOR

FEDERAL

AGENCY

Quarterly data, 1960-67

SECURITIES

MATURING

AFTER

YEAJR

Table 94
GROSS DEALER TRANSACTIONS III COMPARATIVE SECURITIES1 MARKETS
(Averages of daily data in millions of dollars)

Date

If
If

T
J. Se Government debt due:
Federal Agency debt duej
After
1 Within 1 year
1 - 5
5 - 10
Within
1 year
1 vear
I Bills | Coupon issues
years
years

After
10 years

Bankers1
Acceptances

1960-^

818

137

253

1961

1,035

168

265

1962

1,229

171

225

120

1963

1,200

122

216

141

1964

1,293

219

126

1965

105

1,402

195

102

1966

156

1,586

121

242

110

1967*'

140

1,645

289

Based on data for July-December.
Based on data for January-June*

-90-

To what factors can this sharp rise in trading in the Agency
market be attributed?

To answer this question, multiple regressions

were calculated relating dealers1 transactions in Agency securities
to the independent variables that are significant determinants of
transactions in the U. S. Government securities market.—^

The simple

least squares regressions related daily average dealer transactions
in Agency securities to the independent variables for quarters of the
1960, Q3-1967, Q1 period.

The results are presented in Table 15.

Three variables explain 97 per cent of the variance in Agency
transactions during the 1960-67 period:

Agency debt held by public

investors, gross new Agency issues, and the level of free reserves.
All three variables were found to be highly significant (at the 1 per
cent level) as determinants of transactions.
The secular rise in transactions is explained by a secular
rise in Agency debt and, to a lesser degree, in gross new issues.
These independent variables, as well as transactions, are shown in
Charts 18.

The Chart also makes evident that many of the quarterly

peaks in transactions can be related to the volume of new Agency
issues marketed in that quarter.

As shown in Table 15, a rise of $12

million in daily average transactions was associated with a $1 billion
rise in Agency debt, and a $17 million rise in transactions with a
$1 billion rise in gross new Agency issues.

The pertinent regression results for the U. S. Government securities
market are in Louise Ahearn and Janice Peskin, "Market Performance
as Reflected in Aggregative Indicators," Treasury-Federal Reserve
Study of the U. S. Government Securities Market, 1967.

Table 15
Results of Multiple Regressions Explaining Trading in Agency Securities—

| tf

.97

Durbin-IJatson
ratio

2.08#

Constant

-64.52

Net regression coefficients and standard errors
Free Reserves
Agency D E B T held by
Gross new Agency issues
public (.Billions of $) 2/ to public (Billions of $) 3/ (Million of $) 4/

11.53**
(•89)

17.01**
(2.20)

. 04**
(.01)

**
#

Significantly different from zero at 1 per cent level.
No positive serial correlation (Theil and Nagar's Table, 5 per cent significance level for rejecting null
hypothesis of residual independence).

The dependent variable is gross dealer transactions in Agency securities (all maturities and including
participation certificates), quarterly averages of daily data in millions of dollars. The data are shown
in Appendix Table
. Regressions were run for the 1960, Q3 - 1967, Ql period.
Agency debt (including PC's) held by public investors; quarterly average of end-of-month data for four months
in and closest to the quarter.
Gross new Agency debt (including PC's) issued to public investors; total during quarter.
Quarterly average of monthly averages.

2/
3/
4/

VO
H
i

CHART 18
TRANSACTIONS

AGENCY

SECURITIES

1960-1967

AND

CAUSAL

VARIABLES

-93-

The volume of trading was also related to the degree of
monetary tightness or ease, as measured by the level of free reserves,—
The relationship, as measured, involved a $4 million rise in the volume
of trading with a $100 million rise in free reserves (or decline in
net borrowed reserves).

The negative relationship between trading

and monetary tightness probably reflects several factors.

In the

first place, holders of Agency debt may become less willing sellers
when sales involve capital losses--the loc.ked-in effect-- and more
willing sellers when capital gains can be realized.

Secondly, dealers1

positions decline when money is tight and with low positions the
translation of investor buy orders into purchases becomes more
difficult.
These same variables were generally found to be significant
determinants of trading volume in the U. S. Government securities
market during the sixties, as discussed in Chapter 2 of the AhearnPeskin study, but the magnitudes of the relationships often differ
from those found in the Agency market.

Trading in U. S. Government

securities was found to be more responsive to shifts in monetary
ease or tightness than was Agency trading.

A $100 million rise in

free reserves, for example, was associated with increases of $12
million in trading in intermediate-term Treasury coupon issues and
of $58 million in trading in Treasury bills, as compared with only
1/

In an alternative equation, Agency transactions were significantly
related to the level of interest rates on Agency debt. In this
case, the volume of trading increased by $19 million with a 1
percentage point decline in Agency interest rates.

-94-

$4 million in Agency trading.

The associated elasticities—

were

also somewhat higher for Treasury issues, ranging from .08-.19
compared with .05 for Agency issues.
The response of trading to increases in debt in the Agency
and Treasury markets varies, but with no discernible pattern.

The

rise in trading for a $1 billion rise in debt is larger for Agency
issues then for Treasury coupon issues but smaller than for Treasury
2/
bills, as the accompanying table shows in line 1.—

The elasticity

of trading with respect to debt (line 2) varies between .8 and 2.1,
and stands at 1.2 for the Agency market.

The elasticity for the

Agency market indicates that a percentage rise in debt held by
public investors causes a somewhat greater percentage rise in the
volume of trading.

Only for short-term Treasury coupon issues is

the elasticity less than 1.

Item

Trading and Debt Relationships in Selected Markets
Treasury Market
Agency
Coupon Issues due:
Ma rke t Bills
Within 1 year
5-10 years

Rise in trading (millions
of dollars) per $1.0
billion rise in debt 1/
Elasticity:
.A trading
debt
. .
^A debt
" trading —

1.2

1.6

2.1

1/
2/

Coefficients of multiple regression studies.
At mean values (1960, Q3 - 1967, Ql) of debt and trading.

Defined as the ratio between the percentage change in trading and
the percentage change in free reserves, or ^T
F
AF * T
where T -- trading
**
F -- free reserves
The elasticities noted here were taken at the mean values of F and T
for the 1960, Q3 - 1967, Ql period.
Debt was defined as debt held by the public for the Agency and Treasury
coupon markets but as total debt outstanding for the Treasury bill
market where official holdings are actively sold and purchased.

-95-

There appears to be no pattern in the differences in
elasticities among the markets.

The differences obviously do not

relate to the maturities of the debt.

Nor do they relate to the

relative amounts of debt outstanding in any sector since the greatest
amount of debt is in Treasury bills and the least in Agencies.

There

is also no variance in the elasticities that might be traced to
comparative debt turnover (trading/debt) in the various market
sectors; the highest elasticity (5-10 year Treasury issues) is in
the sector for which turnover is the lowest, but the second highest
elasticity (Treasury bills) is where turnover is the greatest.
The amount of debt outstanding was generally established,
by this study as well as the Ahearn-Peskin study, to be the aost
important determinant of trading activity in a securities market.
It is of some use to be able to visualize changes in market activity
after making rough allowance for the sizable shifts that have occurred
in debt outstanding.

To enable this, Appendix Table 8 and the

Market Profile Charts, pages 87 and

88, present a series on the

annual rate of turnover of Agency debt, defined as daily average
trading multiplied by 249 (the number of trading days in most years)
and divided by the average debt held by the public.

During the sixties

as a whole, the annual rate of turnover of Agency debt averaged 2.5.
As is particularly evident in column 1 of Table 16 there was a clear
upward trend in the Agency turnover rate, especially after 1963.

This

upward trend undoubtedly reflects the rising volume of gross new Agency
issues.

-96Table 16
Annual Rates of Turnover
Treasury Debt 4/
Coupon Issues due:
Within
5 - 10
1 - 5
1 year
years
years

Date

Agency
Debt

19 60-'

2.22

5.45

1.73

1.04

.79

.39

1961

2.38

6*46

1.60

1.24

.68

.37

1962

2.37

6.98

1*52

1.15

1.30

.45

1963

2.34

6.12

1.50

1.12

1.11

.70

1964

2.41

6.14

1.34

1.16

.95

.53

1965

2.56

6.19

1.27

1.08

.78

.62

1966

2.75

6, 62

1.82

1.32

.97

.46

19672/

2.71

6.35

1.71

1.40

.88

.43

1/
2/
2/
4/

Treasury
Bills 3/

After 10
years

Based on data for July-December,
Based on data for January-June,
Debt held by the public was the denominator for all classes ©XCCpt
Treasury bills, where total bills outstanding was used.
Turnover rates for Treasury debt were derived from Appendix Table 2
of the Ahearn-Peskin study.

The Table also includes comparative turnover rates for
maturity sectors of the U. S. Government securities market.

In the

Treasury market, the turnover rate increases with the nearness to
maturity of the coupon issues and is considerably larger for Treasury
bills than for coupon issues.

Turnover in the Agency market is

greater than for short-term Treasury coupon issues, especially considering the inclusion of long-term Agency issues in the Agency turnover rates

Turnover in the Agency market is less than half that in

the Treasury bill market, however.

-97-

Besides explaining at least some portion of the upx^ard
trend in turnover in the Agency market, the volume of gross new Agency
issues undoubtedly accounts for some portion of the greater turnover in
the Agency market when compared with the Treasury coupon market.

New

Agency securities are issued every quarter, and in considerable size
(see Chart 18), enlarging both the volume of trading and turnover
of Agency securities.

For example, had there been no new Agency debt

issues during the 1966-67, Ql period the daily volume of Agency trading,
instead of averaging $211 million, might have been only about $150
million; and the turnover rate an average 1.99 instead of 2.79.—^
In comparing the Agency market x^ith other securities markets, the
comparative volume of new issues can thus be an important source of
trading differences.

In some sectors of the U« S. Government securities

market, to note a particular case, financings are considerably less
frequent.
The importance of gross new issues to trading activity raises
the question of just how low inter-financing trading is in the Agency
market.

It has sometimes been alleged, for example, that effecting

trades--particularly purchases—apart from financings is very difficult.
For this reason, an attempt was made to isolate the volume of trading
excluding financing periods.

Since Agency financings occur so frequently,

Actual gross new Agency issues per quarter averaged $4.0 billion
over this period and for every billion dollars of new issues trading
was estimated to be $17 million higher. The estimated turnover rate
of 1.99 uses the actual level of debt though without any new issues
debt would have been lower.

-98however, there are not a great number of such inter-financing periods—^
to observe.

Such data for 1966 and 1967 indicate a trading volume

quite similar to the $150 million of daily trading noted in the
preceding paragraph (when all gross new issues were excluded).
Daily average trading for weekly periods during 1966 shows
a wide variance;

from a low of $68 million to a high of $259 million

for Agency issues due within 1 year and from $12 million to $191
million for longer-term Agency issues.
during financings.

The highs of course occurred

But when daily trading for all weeks in which

there were no financings is averaged for the year, it shows no
startling drop from average trading on all days in the short-term
sector.

There is a decline from $156 million to $121 million (lines

3 and 4 in Table 17).
however.

Trading in long-term issues drop6 considerably,

It will be noted that Agency trading in both maturities

combined excluding financing periods averages $149 million a day
during 1966.
Table 17
Daily Average Trading By Weeks During 1966
(Millions of dollars)

(1)
(2)
(3)
(4)

Highest week
Lowest week
Average for year
Average excluding
financing periods*
(5) Range excluding
financing periods*
*

Agency securities maturing:
Within 1 year
After 1 year
259
191
68
12
156
47
121
68-181

28
12-54

Number of weeks without financing periods was 12 for short-term
Agencies and 27 for longer-term issues.
Excludes the period from the offering date through the payment date
for every new Agency issue, usually several weeks in duration.

-99-

In short-term Agencies, there were only 5 weeks in 1966 during
which trading averaged $100 million a day or less.

In three-quarters

of the weeks, trading was in a $100-200 million range, and was above
$200 million in 8 weeks.

In longer-term Agencies, trading averaged

less than $20 million a day in 5 weeks, and was most often in a $20-60
million range.
In the first half of 1967, the number of Agency financings
dropped considerably with the Home Loan Banks out of the market.
Trading on non-financing days during February, March and April of 1967
was analysed. —^ For short-term Agency issues, it showed average daily
trading of $125 million.

The lowest trading day was $39 million

and the highest $197 million; on 8 of the 30days, trading was less
than $100 million.

For Agency issues due after 1 year, daily trading

averaged $57 million, with a low of $24 million and a high of $109
million.

Again, the exclusion of trading on financing days does not

radically alter one's impression of overall market activity, at least
for short-term issues.

Total trading averaged $140 million a day

during the first half of 1967, only somewhat above the $125 million
on non-financing days.
Despite some exaggeration of Agency trading relative to
trading in certain other securities markets, it seems clear--from
both the data and from dealers1 comments--that activity in the shortterm Agency market is equivalent to, if not greater than, that in the

The number of such non-financing days was 30 for short-term
Agencies and 32 for long-term issues.

-100-

market for short-term Treasury coupon issues.

It is evident, moreover,

that there has been a marked improvement in recent years in the
performance of the Agency market as evidenced by the volume of
trading and turnover.

The growth in activity suggests increased

"breadth" of the Agency market; it implies, as well, a greater marketability of Agency securities, that is, the execution of investor buy
and sell transactions with greater speed, in greater size, and probably
at prices closer to the market.
B.

Dealers1 Positions

Dealers1 daily average positions in Agency securities have,
like transactions, risen sharply since the early 1960's.

On average

during the first half of 1967 dealers held daily net positions of
$264 million in Agency issues due within 1 year and of $165 million
in longer-term Agency issues.

These net position levels are, respec-

tively, some three and nine times larger than in 1961.

Agency positions

are shown in the Market Profile Charts, pages 87 and 88, and in Appendix
Tables 9 and 10.
As the Charts show, dealers1 gross short and long positions
have also risen considerably.

Gross long positions of

dealers include

securities owned outright and gross short positions are securities
borrowed and sold.

The rise in gross positions indicates an increased

willingness and ability of dealers to both buy and sell Agency securities.
Dealers' net positions in Agency securities due within 1 year
have in recent years been of the same order of magnitude as dealers'
net positions in short-term Treasury coupon issues and bankers'

acceptances.

Net positions in longer-term Agency issues have been not

unlike net positions in any one maturity sector c£ the longer-term
Treasury niarket*

If all Treasury issues due after 1 year were lumped

together, Agency positions would of course appear small in comparison.
These comparative data on net positions in various securities
markets are shown in Table 18.

The sharpupward trend in Agency positions,

as well as in positions in bankers1 acceptances, stands out clearly.
Positions in Treasury coupon issues, on the other hand, show no trend,
except perhaps for over-5year maturities.
While gross long positions in Agency securities are, like
net positions, roughly comparable to those in specific maturity
sectors of the Treasury coupon market, gross short positions in
Agency issues are decidedly smaller.
in Table 19.

Gross position data are presented

Gross short positions in Agency securities due within

and after 1 year each

averaged about $35 million a day during 1966

and 1967, considerably less than in the Treasury market.

The small

size of Agency gross short positions indicates a weakness on the
offered side of the Agency market, at least when compared with the
U. S. Government securities market.^
The low level of gross short positions probably reflects
two factors.

First, dealers may have difficulty finding investors

willing to lend Agency securities in any reasonable volume.—
1/

Secondly,

The small size of gross short positions probably limits the size of
dealer gross long positions, thus reflecting back on the bid side of
the market. In periods of expectations of falling security prices,
an inability of dealers tohedge long positions by selling short
would probably lead to a greater cut back in gross long positions
than would otherwise have occurred.
The small size of many outstanding Agency issues adds to the difficulty, since it probably decreases the amount of any specific
issue held by one, or a few, investor(s).

Table

DEALER NET POSITIONS IN COMPARATIVE SECURITIES1 MARKETS
(Averages of daily data in millions of dollars)

Date

Federal Agency debt due:
After
Within
1 year
1 year

U. S. Government debt due:
Within 1 year
5-10
1-5
Bills
Coupon issues
years
years

After
10 years

Bankers1
Acceptances

1961

1,921

438

337

1962

163

2,424

499

273

1963

196

2,542

334

383

103

1964

212

2,636

265

309

131

206

1965

233

104

2,629

186

140

193

197

208

1966

361

1,925

335

142

284

1967-'

264

165

2,874

329

520

117

404

Based on data for January-June,

tr1
O
ro

Table 108
DEALER GROSS POSITIONS IN COMPARATIVE SECURITIES1 MARKETS
(Averages of daily data in millions of dollars)
Federal Agency debt due;
Within
After
1 year
1 year

U. S. Government debt due :
Within 1 year
5-10
1-5
Bills
Coupon issues
years
years

After 10
years

€ross Long
1961

104

2,044

484

516

1962

172

2,604

540

417

158

1963

214

2,709

361

547

239

1964

224

2,808

297

503

277

130

1965

253

127

2,865

262

353

276

236

1966

396

111

2,268

412

315

163

1967—/

299

206

3,117

391

632

170

127

1961

125

180

1962

177

144

1963

165

164

141

1964

175

194

147

1965

235

213

1966

343

172

108

1967-/

244

112

Gross short

Based on data for January-June.

-104-

investor sell orders may not be in sufficient size to guarantee that
dealers are able to buy any specific Agency issue at a future date in
order to enable return of the borrowed security.
As with dealer transactions, multiple regressions were
run on daily average net, gross long and gross short positions for
quarters of the 1961, Ql - 1967, Ql period.

The results provide at

least tentative answers to several important questions:

(1) What

factors have caused dealers1 positions to rise during the sixties?;
(2) Are Agency positions responsive to the same factors as positions
in U. S. Government securities, and in the same magnitudes?
A model of position determination must include several
groups of variables, each of which influences dealers1 profits and
thus the level of positions dealers desire to hold.—^
(or losses) may be categorized as:
(3) interest carry.

Dealers1 profits

(1) speculative; (2) trading;

Speculative profits (or losses) result from capital

gains and losses on the securities held in position as security prices
fluctuate.

When security prices rise, capital gains are realized on

gross long positions and capital losses are experienced on securities
sold short.

The opposite

is true when security prices decline.

Thus

expectations of near-term rises in security prices should cause dealers
to increase gross long positions and cut gross short positions.
1/

What follows is only a minimal description of the theoretical model,
which was described in detail in Chapter 3 of the Ahearn-Peskin
study of the U. S. Government securities market. The reader is
urged to first read the Chapter mentioned, including sections on
measurement difficulties.

-105-

Trading operations contribute to dealers1 profits.

Trading

profits depend on the spread between bid and asked security prices,
the volume of securities traded, and trading costs.

Interest carry

on securities held in position is also an important item in the
dealers' profit outlook.

Nonbank dealers finance their positions by

short-term borrowings, and at the same time they earn interest on the
securities they hold.

When interest paid on borrowings is greater

than interest earned, there is a "negative carry"; when the interest
earned is greater, there is a "positive carry."

A rising positive

carry or a falling negative carry should induce dealers to hold
larger positions.
In addition, dealers underwrite the sale of new Agency
issues, and their gross and net long positions should rise with the
size and frequency of Agency financings.

In general, official accounts

have not purchased and sold Agency securities in the market and thus
their transactions did not form part of the model tested.
The regression results are shown in Table 20.

The model

tested explains 85 per cent of the variance in dealers' net Agency
positions, 91 per cent of the variance in gross long positions and
69 per cent of the variance in gross short positions during the
sixties.

The model accounted for a larger proportion of the variance

in long positions of Agency securities during the sixties than did
essentially the same model for positions in the U. S. Government
securities market.

But only two of the variables tested--gross new

Table

111
1

Results of Multiple Regressions Explaining Dealers

Dependent
variable
(Millions
of $)

D-W
ratio

Gross new
Agency issues
Cons tant to public 2/
(Billions
of $)

Met regression
Average volume
j
of trading in
Agencies, two
preceding
quarters 3/
(111 1 lions of $)

Positions in Agency Securities—

coefficients and standard errors
Agency debt Change in
Change in
free
Agency yield
held by
public 4/ reserves 5/ preceding
quarter 6/
(Millions
(Billions
(Basis points)
of $)
of $)

Interest carry
on Agency
securities 7_/
(Basis points)

Net
Positions i
(i)

.85

1. 7 6# -39.46

90.75**
(11.61)

.71*
(.31)

(2)

.85

1.39

-31.12

84.69**
(12.01)

.64
(.50)

(3)

.85

1.91#

-56.68

83.39**
(13.31)

<*)

.91

1.31

-34.86

90.14**
( 9.93)

.88**
(.26)

1.18
(.61)

(5)

.91

1.38

-22.13

90.17**
(10.14)

.77
(.43)

lc 07
(.70)

(6)

.92

1.5 2#

-57.31

82.10**
(10.96)

1.27
(.82)

.07
(.44)

9.57*
(3,99)

Gross
Long
Positions

1.03
(.61)

11.88**
(3.40)

/-N /-N
CO -sj

Gross
Short
Positions

i
M
0
as
1
.64

1. 75# -13.48

.69

1. 66# -20.68

Footnotes are on following page.

-.12
(.37)

.43**
(.06)
4.64**
(.64)

-.060
(.029)
-.034
(.025)

Table 20-2
FOOTNOTES

**
#

Significantly different from zero at 1 per cent level* * At 5 per cent level*
No positive serial correlation (Theil andNagar's Table, 1 per cent significance level for rejecting
null hypothesis of residual independence)®
1/ The dependent variable is dealers1 positions in Agency securities (all maturities and including
participation certificates), quarterly averages of daily data in millions of dollars. The data
are shown in Appendix Tables
. Regressions were run for the 1961, Ql - 1967, 01. period.
2/ Gross new Agency debt (including PC's) issued topublic investors; total during quarterc
3/ Daily average gross dealer transactions; average for two preceding quarters.
4/ Agoncy debt (including PC's) held by public investors; quarterly average of end-of-month data for
four months in and closest to the quarter.
5/ Change in free reserves, based on quarterly averages of monthly averages; excludes all such
quarterly changes less than $50 million.
6/ Change during preceding quarter in quarterly averages of monthly yields on 3-month and 6-month Agency
issues averaged, Scurce of basic yield data was Salomon Brothers and Kutzler, "An Analytical Record
of Yields and Yield Spreads", Part III®
7/ Interest carry is interest earned on Agency securities held in position less financing costs.
Interest earned was measured by averaging coupon rates on most outstanding Agency issues for the
mid-month of the quarter.
The series on financing costs was average posted rates for new loans
in federal funds at the major New York City banks. Financing costs were then subtracted from
interest earned; a plus indicates positive carry and a minus negative carry.

i
M
0
-J
1

-108-

Agency issues arid the volume of Agency trading or, alternatively,
Agency debt--were significant determinants of Agency positions.
Neither interest carry nor expectations of future security prices
(as measured by the change in Agency yields last quarter,—^ the
change in free reserves, or the change in the discount rate) were
found to be significant.

Table 20 presents the regression results for net and gross long
positions including this variable. The sign of its coefficient is
of some in terest. In every instance there is a positive relationship between long positions and the preceding quarter's change in
Agency yields. While the variable is not significant at the 5 per
cent level, still the probability that the true coefficient is
zero or negative is only 9-12 per cent. In the U. S. Government
securities market, on the other hand, positions in intermediateand long-term issues were related negatively (and significantly)
to the precedingqjarter1s change in Treasury yields; and with a
coefficient large enough to indicate virtually no chance of the
true coefficient being positive* Several points can be made with
respect to this apparent difference in behavior in the two markets.
(1) This appears to some degree to be a difference between shortterm and long-term markets in general. Thus, in the short-term
U. S. Government securities market the change in yields last
quarter was not significant as a determinant of positions and
while its coefficient was always negative the standard errors
were very large. This might indicate that dealers generally
expect movements in long-term yields to continue direction and
move relatively smoothly over the cycle but that they expect
yields on short-term securities to move in a more erratic manner
as with the season or near-term money market conditions. It could
also indicate that dealers project short-term yield movements in
a more sophisticated manner than long-term yields. (2) The
difference does not arise from divergent movements in Agency and
Treasury yields. During the sixties yield changes on Agency
issues and on intermediate- and long-term Governments (the areas
where there was a significant difference in coefficient signs)
were positively related--the simple correlation coefficient was
about 55 per cent. (3) To the degree that these relationships are
truly measuring dealers' expectations, it would appear that dealers
expect Treasury yield movements (at least on long-term issues) to
continue direction but Agency yield changes to reverse direction
over quarterly periods.

-109-

For gross short positions, however, the change in free
reserves fell just short of being significant in one equation (equation 7).

In this case, gross short positions were negatively associated

with changes (positive) in free reserves, as expected.

The coefficient

relating short positions and free reserve changes was smaller for
Agency securities than for U. S. Government securities.

Positions

as well as transactions in Agency securities thus appear to be consistently less responsive to changing monetary conditions than is the
case for U. S. Government securities.
This lesser responsiveness of trading and positions in Agency
securities to monetary developments during the sixties is not without
foundation.

As pointed out in Chapter II, until about 1965 Agency

yields were declining relative to Treasury yields, i.e., the yield
spread was being reduced.

Probably of even greater importance was

the absence of any selling of Agency securities by commercial banks
with the tightening of monetary policy as the sixties progressed,
whereas bank sales of Treasury issues were quite heavy.
The most important determinant of dealers1 gross and net
long positions in Agency securities during the sixties has been gross
new issues of Agency debt.

For every rise of $1 billion in new

Agency issues daily long positions (net and gross) were some $82-91
million higher on average during the quarter
This response of positions to gross new issues in the
Agency market was virtually identical with that found for Treasury
bills and greater than that found for Treasury coupon issues, where

-110-

the coefficients ranged between $14 and $51 million.

However, it is

impossible to draw conclusions about dealers1 underwriting in the
various markets from these regression results because the size of the
financing coefficients depends on the number of financings in the
quarter, the number of separate issues offered per financing and the
dates of the financing(s) within the quarter.

And these factors vary

sharply between the Treasury bill, coupon and Agency markets, ranging
from an average one financing a quarter in Treasury coupon issues to
some 15 financings a quarter in Treasury bills.

Positions, as well

as transactions, in Agency securities are of course enlarged relative
to the Treasury coupon market by the greater overall volume of new issues.
There is a basic difference in the source of underwriting
profits in the U. S. Government and Agency markets.

The method of

marketing new Agency issues is to distribute them to large selling
groups who receive commissions ranging from about $.50 to $3.50 per
$1,000-^ of issues.

In the U. S. Government securities market, there

are no such commissions, and the new issue is made attractive to
investors by pricing it below comparable outstanding issues.

While

price discounting of necessity occurs with the Agency issues as well,
in view of the commission to underwriters such discounting may be less
than on Treasury issues.

At least one dealer in interviews voiced his

opinion that underwriting of Treasury issues was in fact more profitable
than underwriting of Agency issues, due to the more attractive pricing
of Treasury issues.
1/

The size of the commission increases with the maturity of the new
issue.

-111-

The second independent variable found to be a significant
determinant of positions in Agency debt T.;as a measure of trail ing
activity.

Trading activity was measured by the average volume of

daily trading during the preceding two quarters—^ or, alternatively,
by the amount of publicly-held Agency debt outstanding.

Equations 1

and 4 show a rise of $.71 million in net positions and of $,88 million
in gross long positions with a $1 million dollar rise in trading
2/
volume.—

Gross short positions were also positively related to

trading, by $.43 million per $1 million trading rise (equation 7).
While the coefficient is only half the size of that for long positions,
it represents a much greater percentage rise with trading activity
than for long positions.
As an alternative measure of trading activity, the amount
of Agency debt outstanding was also significantly and positively
associated with positions.

Equations 3,6, and 8 show increases of

$9.6 million in daily average net positions, of $11*9 million in daily
average gross bng positions, and of $4.6 million in daily average gross
short positions with a $1 billion rise in Agency debt.

These coefficients

imply a similar--perhaps slightly greater--rise in positions with
trading then do the trading coefficients themselves, given the relationship found earlier of an $11.5 million rise in trading with a $1 billion
V
rise in debt.—'
1/ Because the relationship between trading and positions may be two-way,
with trading to some degree dependent on positions, and because the
volume of new issues causes trading and positions to rise concurrently,
the volume of trading was used for preceding quarters rather than for
the same quarter to avoid a bias in the coefficients.
2/ The reader will note smaller coefficients in equations 2 and 5, where
trading was not significant. The lack of significance was caused by
multicollinearity between trading and interest carry and, to a lesser
degree, between trading and yield changes in the preceding quarter.
3/ See Table 15, page 91.

-112-

These results thus indicate generally a somewhat less than
proportional rise in positions with trading, i.e., a $1 million rise
in trading causes a less than $1 million rise in positions.

In the

U. S. Government securities market, on the other hand, positions rose
by somewhat more than the trading rise.

However, data inadequacies

make this finding only tentative.
While trading and new issues were the only independent
variables found to be significant determinants of Agency positions,
several other variables difficult to measure may well alter dealers1
desires to hold Agency issues relative to U. S. Government issues.
In the first place, the interest carry on Agency debt is greater than
on U. S. Government securities since Agency interest rates are higher.—^
Secondly, the risk of capital loss on Agency debt held in position
may be greater for dealers than on U. S. Government securities.
might be the case for any number of reasons, including:

This

(1) greater

fluctuations in Agency security prices; (2) greater difficulty in
forecasting movements in Agency yields and prices; (3) diminished
ability to alter gross positions in response to expectational stimuli;
and (4) the frequency of new financings.
1/

To some degree the higher interest earned on Agency securities might
be counterbalanced by higher dealer borrowing costs on such securities.
Higher average borrowing costs would result if dealers found it more
difficult to sell Agency issues under repurchase agreements. In this
respect, it was not until late in 1966 that the Federal Reserve was
given the authority to purchase (outright or under RP) non-guaranteed
Agency debt. Moreover, responses of institutional investors to a questionnaire (see Joseph Scherer, "Institutional Investors and the Uo S.
Government Securities Market," Treasury-Federal Reserve Study of the
U. S. Government Securities Market, 1967--page 23) showed the number
of investors who enter into repurchase and resale agreements to be
considerably less in Agency securities than in U. S. Government securities
(29 as versus 55 per cent).

-113-

For whatever reasons, the profitability of Agency operations
is certainly implied by the sharp rise in dealers1 positions, absolutely
and relative to positions in Treasury issues.

The higher positions

lend added support to the observation that the performance of the
Agency market has indeed improved in recent years.
C•

Spreads between Quoted Bid and Asked Security Prices

Spreads between quoted bid and offered prices are a key
factor in the functioning of securities' markets.

In a general

sense, the size of spreads is indicative of the degree of "depth,
breadth, and resiliency11 characterizing a particular market.

specifically, small spreads would indicate a willingness of dealers
to operate on both sides of the market.—^

In addition, small spreads

engender a broad investor participation as the cost to the investor
of transactions is diminished.
In the Profile Charts, pages 87 and

88, and in Appendix

Table 11, spreads are shown for the various Agency securities by
maturity category.

A note of caution must be introduced in interpret-

ing these data, which are derived from published quotations of one
particular dealer (Morgan Guaranty Trust Company).

As is true

quoted spreads in the U. S. Government securities market, the published
quotations overstate the size of the spread for all preferred customers,
1/

In a healthy market, spreads must be subject to some minimum level
consistent with dealer profitability. A reduction in spreads reduces
dealers1 trading profits unless the volume of trading rises correspondingly. Trading profits may be especially important when
other dealer profits are limited by either high carrying costs
or steadily rising interest rates.

-114-

whose trades take place at "inside" quotations.

Additional sources

of error in spread data on Agency issues may arise from the use of only
one dealers' price quotations and from the potential inaccuracy
of price quotations due to the relative trading inactivity in some
longer-term Agency issues.

Nevertheless, the published quotations

are the only available source of spread data.
Data on published Agency spreads show some differences among
the issues of the various Agencies as well as among the different
maturities.

Quoted spreads on the short-term issues of FICB, FHLB,

and Bks. Coops„ have in general fluctuated between 1/32 and 2/32
since 1958.

Spreads on the short-term issues of FNMA and FLB, on

the other hand, have more generally ranged between 2/32 and 4/32.
This difference in spreads is probably indicative of lesser activity
in the FNMA and FLB issues, at least in part due to their smaller
average size.
Quoted spreads on Agency issues bearing maturities of 1 to
5 years have, since 1958, ranged between 4/32 and a full point (32/32).
To some degree, the movement in these spreads over time reflects shifts
in the maturity structure of issues within the 1-5 year category; the
shorter-term issues of course carry the smaller spreads.

Over and

above such a maturity difference, the FHLB issues have in recent
years carried somewhat lower spreads when compared with FNMA and FLB
issues.

On the long-term Agency securities, issued by FNMA and FLB,

quoted spreads have been at one-half point or at one point over the
entire period.

-115-

Spreads on participation certificates have generally been
the same as on FNMA and FLB issues in the 1-5 year and after 5 year
maturities.

Since 1966, they have been at one-half or one point on

both FNMA and Export-Import Bank PC's.
The interviews conducted with dealers, summarized in
Appendix I, disclosed some information on the spreads at which
Agency issues actually trade, at least for the larger customers.—^
In general, the dealers interviewed said that short-term Agency issues
trade at a 1/32 spread, but that the spread could be as low as 1/64.
It was noted, however, that short-term FNMA issues trade at a larger
spread, probably 4/32.

It was less clear at what spreads the longer-

term issues trade but one dealer pointed to around a 4/32 spread on
2-3 year issues and another noted a fair amount of business done at
a 1/4 point spread on the longer-term Agency issues.

In every case,

these spreads are less than the quoted spreads shown in the accompanying Tables.
In only one case, that of the shortest-term issues, does
there appear to have been any secular decline in quoted spreads with
the rapid growth in trading activity and debt in the Agency market.
For FICB, Bks. Coopse and FHLB issues due within 1 year the quoted
spread during 1962-64 was 1/32 as compared with a 2/32 spread in prior
years.

One interviewed dealer also noted a decline in trading spreads

in recent years.

With the sharp rise im interest rates in 1965 and

1966, quoted spreads on all Agency issues increased.

Odd-lots trade at greater spreads.

-116-

Quoted spreads on short-term Agency issues compare favorably with those on Treasury bills and short-term coupon issues.

anything, the quoted Agency spreads are smaller, as shown in Table 21.
With a lengthening of maturities, however, there is a widening disparity
among quoted spreads on Agency and U. S. Government issues.

In the

intermediate-term sector, Agency spreads have in recent years ranged
from 4/32 to a full point, compared with a 4/32 spread on Treasury issues.
Where for the longer-term Agency issues quoted spreads stood at 16/32
or one point, Treasury spreads were 8/32.

It is probable that the

spreads at which intermediate- and long-term Agency issues actually
trade are considerably below those quoted, and also closer to the
spreads at which Treasury issues trade.

It is unlikely, however,

that they are as low as those on Treasury issues.
The evidence, possibly misleading, shows no indication of
any secular decline in spreads over the sixties on Agency securities
other than the shortest-term issues.

Nevertheless, the small size

of spreads on shorter-term Agency issues indicates a strong market.
On the other band, the relatively large quoted spreads on long-term
and on certain short-term issues points to a lesser tradeability of
issues in some specific sectors of the Agency market.

Table 21
HOST TYPICAL SPREADS IU COMPARATIVE SECURITIES1 MARKETS
(In 32nds or in basis points for Treasury bills)

Date

Federal Agency debt due:
1 - 5 years
Within 1 year
FICB
FNMA
FNMA
FLB
FHLB
FHLB
FLB
PC's
Bks.Cocps

After 5 yrs
FNMA
FLB
PC's

U. S. Government debt due:
Within 1 year
3 - 5
5-10
6-13 mo.
3-mo.
years
years
bill
Coupon issues

After 10
years

1953

16-32

3-4

1959

16-32

I960

1961

o
o

1962

o
o

1963

o
o

1964

16-32

1965

8-16

1966

1-2

16-32

i
M
'-a
I

Appendix Table 11-2

-118-

AGENCY DEBT OUTSTANDING
(End of quarter data in millions of dollars)

Quarter

Hon - sua ran teed—^

Participation
Certificates H

Total

1954 •- 1
2
3
4

1,071
1,155
2,092
2,064

1,871
1,155
2,092
2,064

1955 •- 1
2
3
4

2,615
2,931
3,172
3,575

1956 -- 1
2
3
4

3,681
O o r> n
J,OUU
4,129
4,010

1957 •- 1
2
3
4

4,415
5,016
5,205
6,220

1958 •- 1
2
3
4

6,327
5,422
5,637
5,722

6,327
5,422
5t 637
5*, 722

1959 •- 1
2
3
4

5,902
6,707
7,519
7,917

1960 -- 1
2
3
4

7,753
0,403
7,726
7,910

3,681
oo
,0ooo
3

4,129
4,018

-119Appendix Table 11-2

Quarter

Non-guaranteed—^

Participation^ i
Certificates —

Total
7,429
7,765
8,312
3,574

1961 - 1
2
3
4

7,429
7,765
8,312
8,574

1962 - 1
2
3
4

3,995
9,332
9,883
10,133

1963 - 1
2
3
4

9,267
10,192
10,870
11,705

1964 - 1
2
3
4

11,133
11,865
11,996
12,127

300

11,133
11,865
11,996
12,427

1965 - 1
2
3
4

12,246
13,460
13,965
14,086

300
30 C
825
1,170

12,546
13,760
14,790
15,256

1966 - 1
2
3
4

15,055
17,626
10,396
19,249

1,170
2,610
2,075
2,020

16,225
20,236
20,471
21,269

1967 - 1
2

18,604
18,026

3,620
5,730

22,224
23,756

- -

- - - -

- -

3,995
9,332
9,833
10,133
9,267
10,192
10,870
11,705

Non-guaranteed Agency debt includes debt of i:he Federal Intermediate
Credit Banks, Federal Land Banks, Banks for Cooperatives, Federal
Heme Loan Banks, Federal National Mortgage Association and Tennessee
Valley Authority. Source of these data is the Treasury Survey of
Ownership.
Includes only fully marketable participation certificates: all
FNTIA PC's, Export-Import Bank PC's issued in and after February
1967, CCC certificate issued in April, 1966 and retired in August 1966.

-120Appendix Table 11-2
NCI • -GUAPuANTEED AGENCY DEBT BY TYPE
(End of quarter data in mi llions of dolla rs)
Banks for
Cooperative s

FICB

FLB

1954 - 1
2
3
4

110
72
120
120

617
353
776
641

940
679
1,017
1,030

204
51
179
273

1955 - 1
2
3
4

120
110
110
110

699
793
324
657

1,035
1,117
1,133
1,263

141
341
535
975

570
570
570
570

1956 - 1
2
3
4

110
133
143
143

702
834
361
705

1,321
1,322
1,437
1,437

878
929
918
963

670
670
770
770

1957 - 1
2
3
4

135
132
207
222

767
924
948
386

1,519
1,552
1,600
1,599

724
738
765
326

1,220
1,620
1,685
2,687

1953 - 1
2
3
4

191
199
232
252

971
1,159
1,205
1,116

1,625
1,646
1,687
1,743

476
456
616
714

3,064
1,962
1,397
1,897

1959 - 1
2
3
4

253
234
320
364

1,206
1,456
1,524
1,356

1,792
1,888
1,936
1,936

699
992
1,402
1,774

1,947
2,087
2,287
2,437

1960 - 1
2
3
4

360
330
346
407

1,416
1,600
1,665
1,454

2,047
2,137
2,137
2,210

1,293
1,255
1,167
1,266

2,637
3,081
2,411
2,523

Quarter

FHLB

FNilA

TVA

- - -

- - - -

- -

...
- - -

- - -

- -

Total
1,371
1,155
2,092
2,064
2,615
2,931
3,172
3,575
3,681
3,838
4,129
4,018
4,415
5,016
5,205
6,220
6,327
5,422
5,637
5,722
5,902
6,707
7,519
7,917
7,753
8,403
7,726
7,910

Appendix Table 2-2

Quar ter

FICB

FLB

FHLB

Fr-1IA

TVA

Total

1961 - 1
2
3
4

404
302
304
435

1,519
1,723
1,732
1,585

2,210
2,357
2,431
2,431

329
1,055
1,335
1,571

2,416
2,193
2,281
2,453

50
50
100
100

7,429
7,765
3,312
3,574

1962 - 1
2
3
4

452
430
475
505

1,644
1,855
1,930
1,727

2,495
2,550
2,596
2,628

1,602
1,797
2,257
2,7C7

2,658
2,556
2,481
2,422

145
145
145
145

3,995
9,332
9,383
10,133

1963 - 1
2
3
4

400
459
473
539

1,842
2,133
2,233
1,952

2,661
2,725
2,796
2,834

2,014
2,770
3,299
4,363

2,126
1,960
1,899
1,787

145
145
170
180

9,267
10,192
10,370
11,705

1964 - 1
2
3
4

506
493
538
686

2,069
2,315
2,424
2,112

2,886
2,973
3,102
3,169

3,627
4,201
4,132
4,369

1,735
1,698
1,571
1,601

180
130
130
190

11,133
11,865
11,996
12,127

1965 - 1
2
3
4

723
687
708
797

2,206
2,462
2,603
2,235

3,293
3,532
3,612
3,710

4,090
4,757
5,046
5,221

1,739
1,797
1,756
1,834

190
225
240
240

12,246
13,4^0
13,965
14,036

1966 - 1
2
3
4

819
844
882
1 ,074

2,470
2,353
2,991
2,736

3,813
4,105
4,295
4,335

5,060
6,309
6,765
6,359

2,648
3,269
3,178
3,300

245
245
285
345

15,055
17,626
13,396
19,249

1967 - 1
2

1 ,113
1 ,042

2,944
3,297

4,450
4,611

5,741
4,585

4,010
4,078

345
415

13,604
18,026

Source:

Banks for
Coop eratives

Treasury Survey of Ovmership.

-122Appendix Table 3
NET EXPENDITURES (' ) OR RECEIPTS (-) OF SELECTED AGENCIES
(Quarterly- data in millions of dollars)
Banks for
Cooperatives

FICB

FLB

1961 - 1
2
3
4

- 3
- 25
4
50

64
204
57
-196

- 2
149
77
- 3

704
276
353
453

645
604
496
309

1962 - 1
2
3
4

13
- 21
45
29

58
212
75
-204

64
57
47
32

541
596
432
273

401
844
599
130

1963 - 1
2
3
4

- 24
- 21
14
116

114
291
100
-282

34
64
71
39

-1,137
797
983
639

-1,013
1,131
1,168
512

1964 - 1
2
3
4

- 3
- 89
40
149

118
246
109
-311

52
87
129
69

543
494
181
261

1965 - 1
2
3
4

37
- 36
20
87

94
256
139
-266

129
234
80
103

519
737
406
99

- 259
1,191
645
23

1966 - 1
2
3
4

23
24
38
193

135
384
138
-205

99
292
191
101

146
933

111
1,633
955
- 54 7

1967 - 1
2

39
- 71

159
354

53
161

-2,129
-1,329

Quarter

SOURCE:

Monthly Statement'of Receipts and
S tates Governmen t•

FHLB

coo
OO

636

Total

376
730
459
16C

-1,878
- 885

Expenditures of the United

-123Appendix Table 4
MATURITY STRUCTURE OF NON-GUARANTEED AGENCY DEBT
(End of qucr ter data in millions of dollars)

Quarter

Within 1 year

Debt iiau uring:
In 1-5 years
After 5 years

Total

1960 - 2
3
4

5,326
4,720
4,414

1,717
1,446
1,786

1,360
1,560
1,710

0,403
7,726
7,910

1961 - 1
2
3
4

4,225
3,953
4,256
4,400

1,493
2,060
2,154
2,272

1,710
1,752
1,902
1,902

7,429
7,765
0,312
0,574

1962 - 1
2
3
4

4,690
4,932
5,438
5,703

2,003
2,090
2,143
2,043

2,302
2,302
2,302
2,302

0,995
9,332
9,883
10,133

1963 - 1
2
3
4

4,773
5,363
6,220
7,360

2,229
2,586
2,4C8
2,113

2,264
2,233
2,233
2,232

9,267
10,192
10,370
11,705

1964 - 1
2
3
4

6,780
7,310
7,104
7,406

2,213
2,503
2,833
2,742

2,132
2,044
1,979
1,979

11,133
11,865
11,996
12,127

1965 - 1
2
3
4

7,729
3,475
9,136
9,164

2,764
3,114
3,163
3,335

1,754
1,871
1,665
1,590

12,246
13,460
13,965
14,006

1966 - 1
2
3
4

10,127
12,435
12,670
13,446

3,214
3,307
3,932
4,008

1,713
1,803
1,794
1,794

15,055
17,626
13,396
19,249

1967 - 1
2

12,469
12,236

4,225
3,912

1,909
1,079

10,604
10,026

SOURCE:

Derived from the Treasury Survey of Gunership.

Appendix Table 2-2
YIELD SPREADS(Quarterly averages in basis points)

Quarter

3-month Ai^ency yield less:
3-month bill,
3-month bill,
market yield
investment yield

3-month
CD

10-12 year Agency yield .less:
10-year
10-year Aa
Treasury new corporate 2/

1954 - 1
2
3
4

24
27
2
8

22
25
0
6

1955 - 1
2
3
4

5
22
32
26

2
19
28
21

1956 - 1
2
3
4

30
43
24
29

25
38
19
23

25
34

-13
-24

1957 - 1
2
3
4

18
25
23
47

12
18
15
40

51
35
36
71

- 4
-26
-22
- 2

1958 - 1
2
3
4

5
7
5
2

1
5
1
- 4

53
42
9
29

- 9
-11
-23
-18

1959 - 1
2
3
4

15
22
16
37

9
16
7
26

29
31
18
20

-14
-14
-25
-21

1960 - 1
2
3
4

35
23
16
26

25
17
11
21

31
27
27
20

0
- 8
-18
-25

- -

„_
- - -

- -

-125Appendix Table 11-2

Quarter

1/
2/

3-month Agency yield less:
3-month
3-month bill,
3-month bill,
CD
market yield
investment yield

1961 - 1
2
3
4

15
3
12
9

10
- 2
7
4

1962 - 1
2
3
4

2
5
6
3

1963 - 1
2
3
4

10-12 year Agency yield less
10-year Aa
10-year
Treasury new corporate 2/

- -

24
33
37
21

-22
-10
-13
-19

- 4
0
0
- 3

-23
-22

22
13
17
8

- 9
-13
-11
-10

5
14
19
13

- 1
8
12
5

-22
- 6
4
-13

12
6
9
1

-11
- 9
-14
-16

1964 - 1
2
3
4

5
17
26
27

- 3
9
18
18

-24
-16
-12
- 7

5
3
1
5

-14
-15
-15
-18

1965 - 1
2
3
4

12
23
24
23

2
13
15
13

-15
-17
-21
-15

7
14
20
15

- 9
-18
-15
-16

1966 - 1
2
3
4

29
42
62
33

18
31
50
19

-10
-30
-24
-19

18
32
29
39

-10
-17
-37
-54

1967 - 1
2

26
23

15
14

-52
-53

49
52

-30
-25

Quarterly averages of monthly data.
offer.
Equipment trust certificates.

SOURCE:

Bills quoted at bid, other issues at

Based on data in Salomon Brothers and Hutzler, An Analytical Record of
Yields and Yield Spreads, Parts I and III.

Appendix Table 6

-126-

1/
MARKET YIELDS AT CONSTANT MATURITIE S-'
(Quarterly data in per cent)
Quarter

Spread

Agency

3-years
| Treasury

| Spread

3.24
3.26
3.76
3.88

3.06
3.09
3.48
3.68

,18
.17
.28
.20

3.59
3.62
3.95
4.04

3.43
3.45
3.76
3.92

.16
.17
.19
.12

1964 - 1
2
3
4

4.07
4.10
4.00
4.08

3.93
3.90
3. 77
3.86

.14
.20
.23
.22

4.26
4.26
4.13
4.15

4.15
4.11
3.97
4.02

.11
.15
.16
.13

1965 - 1
2
3
4

4.27
4.35
4.47
4.58

4.04
4.03
4.31
4.39

.23
.32
.16
.19

4.30
4.35
4.51
4.68

4.12
4.09
4.27
4.47

.18
.26
.24
.21

1966 - 1
2
3
4

5.23
5.53
6.06
5.87

4.97
4.92
5.67
5.38

.26
.61
.39
.49

5.28
5.47
6.09
5.67

5.01
4.98
5.74
5.35

.27
.49
.35
.32

1967 - 1
2

5.00
4. 70

4.60
4.32

.40
.38

5.07
4.97

4.62
4.56

.45
.41

Agency

i
5-years
Treasury

Spread

(

Agency

10-years
| Treasury

| Spread

1963 - 1
2
3
4

3.89
3.89
4.08
4.13

3.70
3.72
3.87
4.01

.19
.17
.21
.12

4.08
4.08
4.18
4.24

3.94
3.97
3.99
4.15

.14
.11
.19
.09

1964 - 1
2
3
4

4.30
4.33
4.25
4.29

4.16
4.11
4.06
4.06

.14
.22
.19
.23

4.35
4.33
4.32
4.35

4.23
4.19
4.20
4.18

.12
.14
.12
.17

1965 - 1
2
3
4

4.33
4.42
4.54
4.71

4.14
4.15
4.27
4.47

.19
.27
.27
.24

4.34
4.44
4.56
4.69

4.20
4.23
4.31
4.46

.14
.21
.25
.23

1966 - 1
2
3
4

5.23
5.26
5.86
5.48

4.97
4.89
5.51
5.18

.26
.37
.35
.30

5.08
5.16
5.66
5.32

4.83
4. 78
5.22
5.00

.25
.38
.44
.32

1967 - 1
2

5.07
5.19

4.64
4.78

.43
.41

5.07
5.20

4.56
4.81

.51
.39

Yields di>rived from yield curves drawn on selected dates. Quarterly yields
are averages for all dates within the quarter if yield curves were drawn on
more than one date. Based on bid quotations.

SOURCE:

1-year
Treasury

1963 - 1
2
3
4

Quarter

Agency

Treasury Department.

Appendix Tr.bi.2 7

-127

GROSS DEALER TRANSACTIONS IN FEDERAL AGENCY SECURITIES(Averages of daily data in millions of dollars)
Quarter

Securities due
within 1 year

Securities due
after 1 year

Total

1960 - 3
4

58
53

11
20

69
73

1961 - 1
2
3
4

58
55
54
53

16
23
20
15

74
78
74
73

1962 - 1
2
3
4

62
67
72
71

25
23
16
21

87
90
88
92

1963 - 1
2
3
4

64
63
94
34

18
25
13
16

82
93
107
100

1964 - 1
2
3
4

82
116
94
85

10
16
27
24

92
132
121
109

1965 - 1
2
3
4

73
121
116
103

23
47
23
47

101
168
144
150

1966 - 1
2
3
4

121
181
152
171

40
59
55
35

161
240
207
206

1967 - 1
2

146
133

95
69

241
202

Transactions include dealer purchases and sales but exclude allotments
of new issues, maturities, exchanges and repurchase agreements.
Classification is by final maturity date. Averages are based on the
number of trading days in the quarter; participation certificates are
included.

Source:

Market Statistics Division, Federal Reserve Bank of New York.

-128Appendix Table 11-2
ANNUAL RATE OF TURNOVER OF FEDERAL AGENCY DEBT*

QUARTER

1960 - 3
4

2.12
2.32

1961 - 1
2
3
4

2.41
2.58
2.32
2.19

1962 - 1
2
3
4

2.47
2.47
2.27
2.26

1963 - 1
2
3
4

2.11
2.44
2.55
2.24

1964 - 1
2
3
4

2.03
2.87
2.53
2.21

1965 - 1
2
3
4

2.04
3.19
2.49
2.50

1966 - 1
2
3
4

2.55
3.28
2.56
2.60

1967 - 1
2

2.97
2.45

The annual rate of turnover equals daily average gross dealer
transactions multiplied by 249 divided by Agency debt (including
participation certificates) held by the public.

-129Appendix Table

11-2

DEALER NET POSITIONS IN FEDERAL AGENCY SECURITIES*
(Averages of daily data in millions of dollars)

Securities due
within 1 year

Securities due
after 1 year

3
4

142
86

19
26

161
112

Quarter
1960

Total

1961

1
2
3
4

77
109
93
105

19
31
18
5

96
140
111
110

1962

1
2
3
4

115
177
171
191

11
50
24
36

126
227
195
227

1
2
3
4

156
255
212
168

22
44
38
34

178
299
250
202

1
2
3
4

177
234
215
220

10
9
42
69

187
243
257
289

1
2
3
4

194
301
238
200

61
128
138
90

255
429
376
290

1966

1
2
3
4

280
591
213
352

26
125
77
77

306
716
290
429

1967

1
2

288
239

192
137

480
376

1963

1964

1965

Data are on a commitment basis and include securities sold by dealers
under repurchase agreement. Securities are classified by final maturity
date, and include participation certificates. Averages are based on the
number of trading days in the quarter.
Source: Market Statistics Division, Federal Reserve Bank of New York,

-130Appendix Table 11-2
DEALER GROSS POSITIONS IN FEDERAL AGENCY SECURITIES*
(Averages of daily data in millions of dollars)

Quarter

Securities due
within L year
Gross
Gross
Ions
short

1961 - 1
2
3
4

86
114
101
116

9
5
7
11

27
41
31
18

8
10
13
13

114
154
132
134

17
15
20
24

1962 - 1
2
3
4

123
184
181
198

9
6
10
7

29
57
34
45

17
7
10
10

151
241
215
244

26
13
20
17

1963 - 1
2
3
4

162
266
236
192

9
11
25
25

39
58
54
49

17
13
16
15

201
324
290
240

26
24
42
40

1964 - 1
2
3
4

193
244
227
230

16
10
12
9

29
36
58
79

19
26
16
9

222
279
285
309

35
37
28
18

1965 - 1
2
3
4

207
316
257
232

14
15
19
32

74
148
156
130

13
23
19
40

281
464
413
362

27
37
37
71

1966 - 1
2
3
4

310
625
267
380

30
29
51
29

71
156
118
99

46
33
40
22

381
781
385
479

76
62
90
51

1967 - 1
2

313
285

26
44

224
187

30
50

537
472

56
94

Securities due
after 1 year
Gross
Gross
long
short

jLotai
Gross
long

Gross
short

Data are on a commitment basis, are classified by final maturity date, and
include participation certificates. Gross long positions include securities
sold under repurchase agreement®.
Source: Market Statistics Division, Federal Reserve Bank of New York,

Appendix Table 11
SPREAD BETWEEN DEALERS' QUOTED BID AND ASKED PRICES
ON FEDERAL AGENCY SECURITIES*
(Most typical spreads in 32nds)

Quarter

Banksfor
Cooperatives
Debentures 1/

Federal Intermediate
Credit Bank
Debentures 1/

Federal Home Loan Bank
issues due:
Within 1 year In 1 - 5 yearj

1958 - 1
2
3
4

4
2
2-3
2

2
2
3
2

1959 - 1
2
3
4

2
2
2
2

8
8
8
16

1960 - 1
2
3
4

2
2
2
2

16
16
16
8

1961 - 1
2
3
4

2
2
2
2

8
8
8
8

1962 - 1
2
3
4

1
1
1
1

4
2/
2/
4

1963 - 1
2
3
4

1
1
1
1

4
4
8
4

1964 - 1
2
3
4

1
1
1
2

4
4
4
4

1965 - 1
2
3
4

2
2
2
1

4
4
4
4*

1966 - 1
2
3
4

1
1
2
2

8
4
8
8

1967 - 1
2

2
2

8
8

2/
4
8
8

-132Appendix Table 11-2

quarter

Federal National Mortgage
Assn/., Debentures due:
In 1 - 5
Within
After
1 year
years
5 years

1958 - 1
2
3
4

4
2
2-3
2-3

4
8
8
8

8
8-16
16
16

1959 - 1
2
3
4

2
2
2
2

8
8
8
8

1960 - 1
2
3
4

2
2
4
4

1961 - 1
2
3
4

Federal Land Bank
Bonds due:
Within
In 1 - 5
1 year
years

After
5 years

2-4
2
2
2

8
8
8
8

32
32
32
32

16
16
16
16

2
4
4
4

8
8
8
8

32
32
32
32

16
16
16
8

32
32
32
16

4
4
4
4

16
16
16
8

32
32
32
32

2
2
2
2

8
8
8
8

16
16
32
32

2
2
2
2

8
8
16
16

32
32
32
32

1962 - 1
2
3
4

2
2
2
2

4
4
8
4

32
32
32
32

2
2
2
2

8
8
8
8

32
32
32
32

1963 - 1
2
3
4

2
2
2
2

4
4
8
8

16
16
16
16

1-2
1-2
2
1-2

8
8
8
8

16
16
16
16

1964 - 1
2
3
4

2
2
2
2-4

4
4-16
16
16

16
16
32
32

2
2
2
4

8
8
8
8

16
16
32
32

1965 - 1
2
3
4

4
2-4
4
4

4-16
16
4-16
16

32
32
32
16-32

2-4
4
4
4

8
8
8
8-16

32
32
32
16-32

1966 - 1
2
3
4

4
4
4
4

16-32
32
32
32

2-4
2-8
4-8
4-8

16-32
16-32
16-32
16-32

32
32
32
32

1967 - 1
2

4
4

32
32

4-8
4

16-32
32

32
32

-133-

Appendix Table 11-3

FNMA Participation Certificates due:
Quarter

Within
1 year

In 1 - 5
years

After
5 years

1965 - 1
2
3
4

2
2-4
4
4

1966 - 1
2
3
4

8
8
16
.16

16-32
32
32
32

1967 - 1
2

16
16

16-32^
32

3/
16-32-'
32

1/
2/
3/

4
8
8
16

Export-Import Bank
Participation Certificates due:4,
In 1 - 5 years

After 5 years

16
32

16
16
16
32

All such debentures mature within 1 year.
No issues outstanding in this maturity.
In the first quarter of 1967, the spread of 16 was on the new, larger FNMA
PC's while the spread of 32 was on the earlier serial issues of PC's.
Prior to 1967, only serial PC issues were outstanding.
4/ Includes only the fully marketable Export-Import Bank PC's, first issued in
February, 1967.
* The quarterly series were derived from observations on the last trading day
of each month. Generally, the typical spread is the one which existed in
two out of the three months.
Source: Morgan Guaranty Trust Company, Government Bond Department, daily
quotation sheets*

APPENDIX
DIFFERENCES

YIELDS

QUOTED

MORGAN

GUARANTY

CHART

TRUST

CO.

1
AND

FIRST

NATIONAL

CITY

BANK

BASIS'

prapswniii:

bi S -

10
Ay Iti L - Ilk.

10
20

V
r

x;
30

U . r r A -

HCLUoxaJI

City

Icier

ii
BASIS*

££22.
Nai ;:.<n ia] Qity

G
>7

10
liCJ^LKL IfiuaaiiGyE

iwt

rn»mir(7riiRfvmiv«iRi3iiiii
lISL

i^hrlrbxncr
ii III III III lll l Ii

May

1968

'69

>70

'71

•72

'73

'7^

'75

•76

•77

'78

*79

'80

Full text of Federal Agency Debt and Its Secondary Market : Staff Study

FRASER