View original document

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

There is more rigorous evidence
that bank demand for Treasury
securities has not shifted permanently upward in recent months.
In particular, a simple statistical
model of large-bank demand for
such securities estimated over the
January 1972 to November 1982
time period accurately forecasts
bank holdings of Treasuries during
the December 1982 to September 1983 interval.' This suggests
that the long-run decline in commercial bank holdings of these securities has only been temporarily
masked by cyclical and other
special factors.

Summary
Evidence suggests that the recent
increases in bank holdings of Treasuries are not atypical and so do
not indicate a permanent change
in banks' asset preferences. Thus,
there appears to be no reason to
expect that banks are likely to
absorb a disproportionate share of
Treasury debt in 1984 and after.
Banks appear to increase their
holdings of Treasuries only when
private demand for credit is slack.
This trend suggests that banks
can be expected to swallow large
amounts of Treasury debt in 1984
and 1985, only if market rates
rise enough to depress nongovernmental demand for bank credit.

••

BULK RATE
U.S. Postage Paid
Cleveland,OH
Permit No. 385

Commercial Bank
Holdings of
Treasury Debt

Even conservative estimates of
future federal government deficits
suggest that the U.S. Treasury's
borrowing needs will continue to be
large in the months ahead. The
impact of the increase in Treasury
debt resulting from large, persistent deficits on financial market
rates and flows will depend on the
strength of the demands made by
certain classes of investors for such
securities. The absorption of Treasury debt by commercial banks is
of particular concern, because traditionally banks have held more
Treasuries than any private domestic sector except households.'

••

1. Treasuries refers to marketable Treasury
securities, including Treasury bills, notes,
and bonds.

Address Correction Requested: Please send
corrected mailing label to the Federal Reserve
Bank of Cleveland, Research Department,
P.O. Box 6387, Cleveland, OH 44101.

ISSN 042~-1276
January 16, 1984

econo
co

by Gary Whalen

3. For an example of such models, see those in
Stephen M. Goldfeld, Commercial Bank Behavior
and Economic Activity: A Structural Study of
Monetary Policy in the Postwar United States
(North-Holland Publishing Company, 1966).
The model used in this article was based on
Goldfeld's.

Federal Reserve Bank of Cleveland
Research Department
P.O.Box 6387
Cleveland, OH 44101

Federal Reserve Bank of Cleveland

Past Holding Patterns
The share of outstanding Treasury
debt held by commercial banks declined from 20.5 percent in 1972 to
a low of 13.5 percent at the end of
1982 (see table 1). This decline is
evident for all commercial banks
and large commercial banks, which
account for roughly one-third of the
Treasuries held by all banks/
Treasury survey data also show
that banks generally prefer to hold
government debt of relatively short
maturity. A June 1982 ownership
survey revealed that commercial
banks held roughly 9.5 percent of
outstanding marketable Treasury
debt with maturities of less than
1 year, 16.7 percent of debt with
maturities between 1 and 5 years,
and approximately 12 percent of
debt with maturities in excess of
5 years. These data imply that
41.1 percent of the Treasury securities held by banks had maturities of less than 1 year; 50.3 percent had maturities in the 1- to
5-year range; and only 8.6 percent
had maturities of over 5 years. In
1978, just 4 years earlier, banks
held 23.6 percent of the outstanding marketable Treasury debt with
maturities between 1 and 5 years
and 25.4 percent of the debt with

longer maturities. Their 1978 share
of less-than-one-year Treasuries
was essentially equal to the 1982
figure. The 1978 data imply that
85.6 percent of the Treasury debt
held by banks at that time had
maturities of less than 5 years.
The decline in the commercial
banks' share of outstanding Treasury debt has been somewhat
irregular. In the years 1975, 1976,
and 1980, the share of Treasury
debt held by commercial banks
actually rose above the previous
year's level. When all bank and
large bank Treasury holdings are
related to an asset measure of their
size, a similar pattern is evident.
These ratios also show an increase
in 1982 holdings over 1981 holdings. Troughs in the business cycle
occurred in 1975, 1980, and 1982,
suggesting that commercial banks
typically increase their holdings
of Treasury securities late in downturns and early in recoveries; they
then reverse this behavior late in
expansions and early in recessions.

••

••

2. Specifically, large commercial banks are
weekly reporting commercial banks with domestic assets of $750 million or more as of December 31, 1977.

Gary Whalen is an economist at the Federal
Reserve Bank of Cleveland. June Gates provided
research assistance for this article.
The views stated herein are those of the author
and not necessarily those of the Federal Reserve
Bank of Cleveland or of the Board of Governors
of the Federal Reserve System.

Table 1 Commercial and Large-Bank
In percentages, at year-end

Treasury Holdings

Commercial banks

Year

1972
1973
1974

Share of
outstanding
Treasury debt

Share of
interestbearing
Treasury debt

Treasury
holdings:
total finance
assets

Large-bank
Treasury
holdings:
total assets

19.4

10.6

12.7
11.5

8.1

6.8
5.4

7.0

4.5

14.7

20.5
17.4

1975

16.0
19.4

1976

20.5

15.9

10.2
11.4

1977

14.2

10.1

1978

18.0
15.4

8.8
7.4

1979

14.5

12.2
11.3

8.3
7.5

5.3

1980

15.1
13.7

12.0
11.0

8.0
7.4

5.1

1981

11.2

8.3

4.9

1982

13.5

7.5

5.8

4.6

SOURCES: The first three ratio series were constructed using Federal Reserve flow of funds data and
data from the Treasury Bulletin (U.S. Treasury Department). The last series is devised from balance
sheet data for large weekly reporting commercial banks with domestic assets of $750 million or more
on 12/31177, published in Federal Reserve Bulletin.

The reasons for banks' holding
Treasury securities explain this
general downward trend and the
cyclical deviations from it. Banks
hold some Treasury securities
for investment purposes and for
satisfying certain institutional
requirements. In particular, banks
must collateralize government
deposits and use Treasury securities among other eligible assets to
fulfill this pledging requirement.
Banks may also hold these securities to dress up balance sheets for
year-end reporting.

However, the primary reason
banks hold Treasury securities,
particularly those with relatively
short maturities, is that their risk,
return, and liquidity characteristics
make them ideal assets to hold for
liquidity purposes. Commercial
banks have always needed to hold
some amount of low-risk, interestbearing, highly liquid assets because of the unpredictability of
loan demand and deposit flows, the
relatively short average maturity of
their liabilities, the illiquid nature
of many of their assets, competitive
pressures, and a variety of regulatory constraints. Short-term Treasuries possess all three of the
necessary attributes.
The liquidity motive for holding
Treasuries largely explains the
observed cyclical changes in bank
holdings of these securities. In
the later stages of recessions and
early phases of recoveries, deposit
inflows generally exceed loan
demand. Banks respond by investing excess funds in liquid instruments such as Treasury securities.

During the remainder of the cycle,
when loan demand rises relative
to the supply of available deposit
funds, banks sell off accumulated
liquid assets. In the past, attempts
by banks to lock in high interest
rates at anticipated peaks reinforced this pattern.
Several developments explain
the downward trend in bank holdings of Treasuries over the decade.
Beginning in the early 1960s, constraints on the ability of commercial banks to raise funds in financial markets continuously over the
business cycle have gradually been
eliminated or circumvented. Thus,
in more recent periods of strong
loan demand, banks have been able
to obtain funds by issuing negotiable CDs, or commercial paper, by
purchasing federal funds, or by
tapping the Eurodollar market.
Further, beginning with the introduction of money market certificates in 1978, rates on an everlarger portion of retail deposits
have not been held below market
levels by Regulation Q. Thus,
banks have been able to attract and
hold such deposits, even in periods
of high rates and strong credit
demand. Accordingly, banks have
been able to reduce the amount
of assets held for liquidity purposes. In addition, banks have
increasingly substituted federal
funds, bank CDs, and municipal
securities-assets
with characteristics similar to those of Treasuries-for the Treasury securities
in their portfolios. Finally, banks
have generally reduced holdings of
all types of longer-term, fixed-rate
assets because of volatile interest
rates after 1979.

Table 2 Large Commercial Bank Holdings of Treasury Securities:
1982 vs. 1975
Recession

Total Treasuriesl
total assets

1975
1982

Cumulative change in
Treasuries from trough,
billions of dollars

At trough"

Three months
after trough

Six months
after trough

Nine months
after trough

5.16
4.79

6.15
5.31

6.78
5.87

7.47
5.60

4.60
5.82

7.11
10.97

11.40
9.08

1975
1982

47.72
29.72
19.25
Cumulative percent
1975
21.48
26.02
13.72
1982
change in Treasuries
from trough
a. The trough of the 1975 recession occurred in March 1975. The trough of the 1982 recession occurred
in November 1982.
SOURCE: Balance sheet data for large weekly reporting commercial banks with domestic assets of
$750 million or more on 12/31177, published in Federal Reserve Bulletin.

Recent Changes
Commercial banks added $16.1 billion of Treasury securities to their
portfolios in the last quarter of
1982, $18.1 billion in the first
quarter of 1983, and an additional
$16.2 billion in the second quarter
of 1983. Representing a 43.8 percent increase over bank holdings
at the end of the third quarter of
1982, these additions pushed banks'
share of total outstanding Treasury debt to 15.4 percent and their
share of interest-bearing debt to
12.7 percent. At the end of the first
half of 1983, Treasuries constituted 10.1 percent of total financial
assets held by banks. Large commercial banks increased their holdings of these securities by $16.5 billion (roughly 43 percent) over the
same three-quarter period (fourth
quarter of 1982 through second
quarter of 1983), pushing their
Treasury holdings to about 6 percent of total assets at the end of
the second quarter of 1983.

Maturity class data indicate that
at least large banks (similar data
are not available for all banks) continue to prefer short-term Treasury
debt. Of the $16.6-billion increase
in large-bank Treasury holdings,
$8.0 billion was in the less-thanI-year maturity range and $8.3 billion was in the 1- to 5-year range.
Thus, large-bank demand for longterm Treasury debt did not rise
over this period.
Can banks be expected to continue to hold this share of Treasury
debt, or even increase their share
in the months ahead? Would such
an increase be largely a cyclical phenomenon? Or would it be caused
by special circumstances and probably be reversed? The latter explanation appears to be more likely. A
comparison of large-bank behavior
in recent months with such behavior during the 1974-75 recession
suggests an answer to this question (see table 2).

Ups and Downs of
Treasury Debt Demand
Banks tend to increase their holdings of Treasuries around cyclical
lows. Data in the tables confirm
that this occurred in both recessions. However, the data also
clearly show that bank accumulation of Treasuries in the 1982-83
period was smaller and briefer than
in the earlier period. Despite the
large absolute size of the increases
in bank holdings of Treasuries in
recent months, the ratio of Treasury securities to total assets
remained well below the levels
reached in 1974-75. Further, bank
Treasury holdings rose for only
seven consecutive months after the
recession trough in November 1982
before declining; by contrast, these
holdings rose for ten consecutive
months in the earlier recession. As
a result, the ratio of Treasuries to
total bank assets peaked only six
months after the trough in the
recent period and then began to
decline. In 1975, this ratio rose
steadily for 13 consecutive months
after the trough to a peak of
8.26 percent.
Indeed, a large part of the run
up in bank holdings of Treasuries
seen in late 1982 and in 1983 may
have been caused by two transient,
noncyclical factors. Some of the
increase that took place in late 1982
may simply reflect balance-sheet
window dressing by banks, which
recurs in the fourth quarter of
every year. An even greater portion
of the buildup in both 1982 and
early 1983 may have been caused
by the unexpectedly large deposit
inflows to banks as a result of the
introduction of money market deposit accounts in December 1982.

Table 1 Commercial and Large-Bank
In percentages, at year-end

Treasury Holdings

Commercial banks

Year

1972
1973
1974

Share of
outstanding
Treasury debt

Share of
interestbearing
Treasury debt

Treasury
holdings:
total finance
assets

Large-bank
Treasury
holdings:
total assets

19.4

10.6

12.7
11.5

8.1

6.8
5.4

7.0

4.5

14.7

20.5
17.4

1975

16.0
19.4

1976

20.5

15.9

10.2
11.4

1977

14.2

10.1

1978

18.0
15.4

8.8
7.4

1979

14.5

12.2
11.3

8.3
7.5

5.3

1980

15.1
13.7

12.0
11.0

8.0
7.4

5.1

1981

11.2

8.3

4.9

1982

13.5

7.5

5.8

4.6

SOURCES: The first three ratio series were constructed using Federal Reserve flow of funds data and
data from the Treasury Bulletin (U.S. Treasury Department). The last series is devised from balance
sheet data for large weekly reporting commercial banks with domestic assets of $750 million or more
on 12/31177, published in Federal Reserve Bulletin.

The reasons for banks' holding
Treasury securities explain this
general downward trend and the
cyclical deviations from it. Banks
hold some Treasury securities
for investment purposes and for
satisfying certain institutional
requirements. In particular, banks
must collateralize government
deposits and use Treasury securities among other eligible assets to
fulfill this pledging requirement.
Banks may also hold these securities to dress up balance sheets for
year-end reporting.

However, the primary reason
banks hold Treasury securities,
particularly those with relatively
short maturities, is that their risk,
return, and liquidity characteristics
make them ideal assets to hold for
liquidity purposes. Commercial
banks have always needed to hold
some amount of low-risk, interestbearing, highly liquid assets because of the unpredictability of
loan demand and deposit flows, the
relatively short average maturity of
their liabilities, the illiquid nature
of many of their assets, competitive
pressures, and a variety of regulatory constraints. Short-term Treasuries possess all three of the
necessary attributes.
The liquidity motive for holding
Treasuries largely explains the
observed cyclical changes in bank
holdings of these securities. In
the later stages of recessions and
early phases of recoveries, deposit
inflows generally exceed loan
demand. Banks respond by investing excess funds in liquid instruments such as Treasury securities.

During the remainder of the cycle,
when loan demand rises relative
to the supply of available deposit
funds, banks sell off accumulated
liquid assets. In the past, attempts
by banks to lock in high interest
rates at anticipated peaks reinforced this pattern.
Several developments explain
the downward trend in bank holdings of Treasuries over the decade.
Beginning in the early 1960s, constraints on the ability of commercial banks to raise funds in financial markets continuously over the
business cycle have gradually been
eliminated or circumvented. Thus,
in more recent periods of strong
loan demand, banks have been able
to obtain funds by issuing negotiable CDs, or commercial paper, by
purchasing federal funds, or by
tapping the Eurodollar market.
Further, beginning with the introduction of money market certificates in 1978, rates on an everlarger portion of retail deposits
have not been held below market
levels by Regulation Q. Thus,
banks have been able to attract and
hold such deposits, even in periods
of high rates and strong credit
demand. Accordingly, banks have
been able to reduce the amount
of assets held for liquidity purposes. In addition, banks have
increasingly substituted federal
funds, bank CDs, and municipal
securities-assets
with characteristics similar to those of Treasuries-for the Treasury securities
in their portfolios. Finally, banks
have generally reduced holdings of
all types of longer-term, fixed-rate
assets because of volatile interest
rates after 1979.

Table 2 Large Commercial Bank Holdings of Treasury Securities:
1982 vs. 1975
Recession

Total Treasuriesl
total assets

1975
1982

Cumulative change in
Treasuries from trough,
billions of dollars

At trough"

Three months
after trough

Six months
after trough

Nine months
after trough

5.16
4.79

6.15
5.31

6.78
5.87

7.47
5.60

4.60
5.82

7.11
10.97

11.40
9.08

1975
1982

47.72
29.72
19.25
Cumulative percent
1975
21.48
26.02
13.72
1982
change in Treasuries
from trough
a. The trough of the 1975 recession occurred in March 1975. The trough of the 1982 recession occurred
in November 1982.
SOURCE: Balance sheet data for large weekly reporting commercial banks with domestic assets of
$750 million or more on 12/31177, published in Federal Reserve Bulletin.

Recent Changes
Commercial banks added $16.1 billion of Treasury securities to their
portfolios in the last quarter of
1982, $18.1 billion in the first
quarter of 1983, and an additional
$16.2 billion in the second quarter
of 1983. Representing a 43.8 percent increase over bank holdings
at the end of the third quarter of
1982, these additions pushed banks'
share of total outstanding Treasury debt to 15.4 percent and their
share of interest-bearing debt to
12.7 percent. At the end of the first
half of 1983, Treasuries constituted 10.1 percent of total financial
assets held by banks. Large commercial banks increased their holdings of these securities by $16.5 billion (roughly 43 percent) over the
same three-quarter period (fourth
quarter of 1982 through second
quarter of 1983), pushing their
Treasury holdings to about 6 percent of total assets at the end of
the second quarter of 1983.

Maturity class data indicate that
at least large banks (similar data
are not available for all banks) continue to prefer short-term Treasury
debt. Of the $16.6-billion increase
in large-bank Treasury holdings,
$8.0 billion was in the less-thanI-year maturity range and $8.3 billion was in the 1- to 5-year range.
Thus, large-bank demand for longterm Treasury debt did not rise
over this period.
Can banks be expected to continue to hold this share of Treasury
debt, or even increase their share
in the months ahead? Would such
an increase be largely a cyclical phenomenon? Or would it be caused
by special circumstances and probably be reversed? The latter explanation appears to be more likely. A
comparison of large-bank behavior
in recent months with such behavior during the 1974-75 recession
suggests an answer to this question (see table 2).

Ups and Downs of
Treasury Debt Demand
Banks tend to increase their holdings of Treasuries around cyclical
lows. Data in the tables confirm
that this occurred in both recessions. However, the data also
clearly show that bank accumulation of Treasuries in the 1982-83
period was smaller and briefer than
in the earlier period. Despite the
large absolute size of the increases
in bank holdings of Treasuries in
recent months, the ratio of Treasury securities to total assets
remained well below the levels
reached in 1974-75. Further, bank
Treasury holdings rose for only
seven consecutive months after the
recession trough in November 1982
before declining; by contrast, these
holdings rose for ten consecutive
months in the earlier recession. As
a result, the ratio of Treasuries to
total bank assets peaked only six
months after the trough in the
recent period and then began to
decline. In 1975, this ratio rose
steadily for 13 consecutive months
after the trough to a peak of
8.26 percent.
Indeed, a large part of the run
up in bank holdings of Treasuries
seen in late 1982 and in 1983 may
have been caused by two transient,
noncyclical factors. Some of the
increase that took place in late 1982
may simply reflect balance-sheet
window dressing by banks, which
recurs in the fourth quarter of
every year. An even greater portion
of the buildup in both 1982 and
early 1983 may have been caused
by the unexpectedly large deposit
inflows to banks as a result of the
introduction of money market deposit accounts in December 1982.

There is more rigorous evidence
that bank demand for Treasury
securities has not shifted permanently upward in recent months.
In particular, a simple statistical
model of large-bank demand for
such securities estimated over the
January 1972 to November 1982
time period accurately forecasts
bank holdings of Treasuries during
the December 1982 to September 1983 interval.' This suggests
that the long-run decline in commercial bank holdings of these securities has only been temporarily
masked by cyclical and other
special factors.

Summary
Evidence suggests that the recent
increases in bank holdings of Treasuries are not atypical and so do
not indicate a permanent change
in banks' asset preferences. Thus,
there appears to be no reason to
expect that banks are likely to
absorb a disproportionate share of
Treasury debt in 1984 and after.
Banks appear to increase their
holdings of Treasuries only when
private demand for credit is slack.
This trend suggests that banks
can be expected to swallow large
amounts of Treasury debt in 1984
and 1985, only if market rates
rise enough to depress nongovernmental demand for bank credit.

••

BULK RATE
U.S. Postage Paid
Cleveland,OH
Permit No. 385

Commercial Bank
Holdings of
Treasury Debt

Even conservative estimates of
future federal government deficits
suggest that the U.S. Treasury's
borrowing needs will continue to be
large in the months ahead. The
impact of the increase in Treasury
debt resulting from large, persistent deficits on financial market
rates and flows will depend on the
strength of the demands made by
certain classes of investors for such
securities. The absorption of Treasury debt by commercial banks is
of particular concern, because traditionally banks have held more
Treasuries than any private domestic sector except households.'

••

1. Treasuries refers to marketable Treasury
securities, including Treasury bills, notes,
and bonds.

Address Correction Requested: Please send
corrected mailing label to the Federal Reserve
Bank of Cleveland, Research Department,
P.O. Box 6387, Cleveland, OH 44101.

ISSN 042~-1276
January 16, 1984

econo
co

by Gary Whalen

3. For an example of such models, see those in
Stephen M. Goldfeld, Commercial Bank Behavior
and Economic Activity: A Structural Study of
Monetary Policy in the Postwar United States
(North-Holland Publishing Company, 1966).
The model used in this article was based on
Goldfeld's.

Federal Reserve Bank of Cleveland
Research Department
P.O.Box 6387
Cleveland, OH 44101

Federal Reserve Bank of Cleveland

Past Holding Patterns
The share of outstanding Treasury
debt held by commercial banks declined from 20.5 percent in 1972 to
a low of 13.5 percent at the end of
1982 (see table 1). This decline is
evident for all commercial banks
and large commercial banks, which
account for roughly one-third of the
Treasuries held by all banks/
Treasury survey data also show
that banks generally prefer to hold
government debt of relatively short
maturity. A June 1982 ownership
survey revealed that commercial
banks held roughly 9.5 percent of
outstanding marketable Treasury
debt with maturities of less than
1 year, 16.7 percent of debt with
maturities between 1 and 5 years,
and approximately 12 percent of
debt with maturities in excess of
5 years. These data imply that
41.1 percent of the Treasury securities held by banks had maturities of less than 1 year; 50.3 percent had maturities in the 1- to
5-year range; and only 8.6 percent
had maturities of over 5 years. In
1978, just 4 years earlier, banks
held 23.6 percent of the outstanding marketable Treasury debt with
maturities between 1 and 5 years
and 25.4 percent of the debt with

longer maturities. Their 1978 share
of less-than-one-year Treasuries
was essentially equal to the 1982
figure. The 1978 data imply that
85.6 percent of the Treasury debt
held by banks at that time had
maturities of less than 5 years.
The decline in the commercial
banks' share of outstanding Treasury debt has been somewhat
irregular. In the years 1975, 1976,
and 1980, the share of Treasury
debt held by commercial banks
actually rose above the previous
year's level. When all bank and
large bank Treasury holdings are
related to an asset measure of their
size, a similar pattern is evident.
These ratios also show an increase
in 1982 holdings over 1981 holdings. Troughs in the business cycle
occurred in 1975, 1980, and 1982,
suggesting that commercial banks
typically increase their holdings
of Treasury securities late in downturns and early in recoveries; they
then reverse this behavior late in
expansions and early in recessions.

••

••

2. Specifically, large commercial banks are
weekly reporting commercial banks with domestic assets of $750 million or more as of December 31, 1977.

Gary Whalen is an economist at the Federal
Reserve Bank of Cleveland. June Gates provided
research assistance for this article.
The views stated herein are those of the author
and not necessarily those of the Federal Reserve
Bank of Cleveland or of the Board of Governors
of the Federal Reserve System.