Full text of Financial Industry Studies : December 1993

View original document

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

FINANCIAL INDUSTRY
D e c e m b e r

1 9 9 3

STUDIES
Federal Reserve Bank of Dallas

Banks and Mutual Funds:
Implications for Banking and Monetary Policy
Kenneth J. Robinson
Senior Economist and Policy

Advisor

Eric Franz Joseph Ochel
Summer Intern

The Relationship Between Bank
Lending and Money Growth:
Were Things Different in the 1980s?
Kenneth J. Robinson
Senior Economist and Policy

Advisor

This publication was digitized and made available by the Federal Reserve Bank of Dallas' Historical Library (FedHistory@dal.frb.org)

Financial Industry Studies
Federal Reserve Bank of Dallas
December 1993

President and Chief Executive Officer
Robert D. McTeer, Jr.
First Vice President and Chief Operating Officer
Tony J. Salvaggio
Senior Vice President
Robert D. Hankins
Vice President
Genie D. Short

Industry Studies is published by the Federal Reserve Bank
of Dallas. The views expressed are those of the authors and
do not necessarily reflect the positions of the Federal Reserve Bank
of Dallas or the Federal Reserve System.

Financial

Subscriptions are available free of charge. Please send requests for
single-copy and multiple-copy subscriptions, back issues, and address changes
to the Public Affairs Department, Federal Reserve Bank of Dallas,
P.O. Box 655906, Dallas, TX 75265-5906, (214) 922-5254.
Articles may be reprinted on the condition that the source
is credited and the Financial Industry Studies Department is provided
a copy of the publication containing the reprinted material.

Banks and
Mutual Funds:
Implications for Banking and
Monetary Policy
Kenneth J. Robinson
Senior Economist and Policy Advisor
Eric Franz Joseph Ochel
Summer Intern
Financial Industry Studies Department
Federal Reserve Bank of Dallas

fter several years of financial difficulties, the U.S. banking industry appears
to be on the road to a solid recovery. In
1992, banks enjoyed record earnings, significant increases in capital ratios, and continued improvement in asset quality—positive
trends that continued into 1993- However,
despite the dramatic improvement in the
banking industry's financial condition,
lending activity remains unusually weak.
Adjusted for inflation, bank loans have
declined for the past three years.
A number of explanations have been
offered for this weakness in bank lending.
Loan demand has been held in check as
the result of deleveraging by both households and businesses, reflecting efforts to
restructure balance sheets. Supply-side
factors also are affecting bank lending as
greater regulatory scrutiny and oversight
have increased banks' cost of making loans.
These developments make it increasingly
difficult for banks to compete with their
less regulated counterparts. Finally, the
unusually wide spread between short- and
long-term interest rates that has characterized the yield curve recently also may be
affecting bank lending activity.
Whatever the reasons for the lack of
lending activity, banks have responded in

two ways. First, they have sharply increased
their holdings of U.S. government securities, possibly because banks lack viable
lending opportunities. Banks also may
move into securities if the introduction of
risk-based capital requirements makes it
more costly to extend loans. Second, in an
attempt to maintain market share and
increase fee income, banks have begun
aggressive marketing of mutual funds. From
almost no involvement in the mid-1980s,
banks' offerings of mutual funds have grown
at very rapid rates in the past few years.
These developments in the U.S. banking
industry have important implications for
both banking and monetary policy. Banks'
move into the mutual funds business represents an attempt to diversify the business
of banking into less traditional product
lines. As with previous efforts to expand
their scope, however, the regulatory environment in which banks operate will also
need to evolve if they are to compete
effectively in an increasingly competitive
global financial marketplace. Moreover, as
the second article in this issue investigates,
banks' entry into new product lines could
be a factor affecting money growth, which
has important implications for the nation's
monetary policy.

U.S. Banks: Recovering but Not
Lending Much
Record earnings. After unprecedented
difficulties beginning in the mid-1980s, the
U.S. banking industry appears to have
executed a remarkable turnaround. Record
bank earnings of more than $30 billion
were posted in 1992, easily surpassing the
previous record of $25 billion set in 1988.
Profitability also reached unprecedented
heights in 1992, with an average return on
assets for the year of 0.96 percent, the
highest since the Federal Deposit Insurance
Corporation was created in 1934. Table 1
shows the major contributors to the earnings
position of U.S. banks over the three-year
period ending with the first half of 1993Improvements in net interest margins and
1

Table 1
Major Profitability Components for Insured U.S. Commercial Banks
1991 *
1993*
(Percent of average assets)

Effect on profitability
(Basis points)

Net interest income

3.57

3.96

Noninterest income

1.76

2.08

.92

.51

3.60

3.98

-38

Gains on securities

.04

.09

Taxes

.28

.54

-26

Loss provision
Other noninterest expense

Extraordinary items, net

.04

.11

Net income

.61

1.21

* All figures are for the year as of June 30, annualized.
DATA SOURCE: Report of Condition and Income.

asset quality were major factors behind the
strength in bank profitability, along with
increases in noninterest income.
However, while banks have benefited
from the sustained declines in interest rates
and the widening spread between their
cost of funds and their lending rates that
began in 1990, the current interest rate
environment has resulted in some unusual
trends in the flow of deposits at banks.
After reaching a peak in 1990, small time
deposits at banks have fallen more than 20
percent. Some decline in these deposits is
to be expected, given the sustained decline
in interest rates since spring 1990. However, the magnitude of recent declines in
small certificates of deposit at banks and
other depository institutions was much
larger than expected, based on historical
relationships between small time deposits
and interest rates. Chart 1 shows the movement of small time deposits and the other
components of the M2 measure of the money
supply. These recent movements in retail

deposits have been a source of concern to
monetary policymakers, because the runoff
of small CDs at banks and other depository
institutions is considered by many analysts
to be a major factor behind the current
weakness in the growth rate of M2.1
Lending remains weak. While lower
interest rates can be expected to affect the

Chart 1
Components of M2
Billions of dollars
4.000

Checking accounts
3,500

3.000

2,500

2,000
1,500

1,000
500

1 M2 consists of the sum of currency in circulation,
checking accounts, savings deposits, small time tieposits, and money market mutual funds.

NOTE: 1993 data are for July,
DATA SOURCE: CITIBASE.

growth of time deposits, developments on the
asset side of bank balance sheets may also
shed some light on the unusual behavior
of time deposits over the past few years.
Chart 2 shows how total loans at U.S. banks,
adjusted for inflation, reached a peak in
1989 and have declined by 14 percent since
that time. Both demand and supply factors
could be responsible for the lack of lending activity. Weak demand for bank loans
has been cited by several studies as a significant factor in recent movements in bank
lending. 2 Supply-side factors that have increased the cost to banks of extending loans
are also at work. These factors—largely the
result of the unprecedented banking difficulties of the mid- to late 1980s—include
higher capital requirements, higher deposit
insurance premiums, and increased regulatory oversight and scrutiny. Finally, the
unusually wide spread between short- and
long-term interest rates may also be affecting bank lending activity by reinforcing
incentives for banks to shift into government securities and away from loans. 3
One way banks have responded to this
unusual environment is by increasing their
holdings of government securities. Chart 3
shows how U.S. banks have restructured

Chart 2
Lending Declines at U.S. Banks
Billions of dollars, adjusted for inflation
1,800
1,600
1,400
1,200

1,000
800

Chart 3
U.S. Banks' Holdings of Government
Securities and Loans
Billions of dollars

2,020

2,000

350
1990

1991

1992

1993

DATA SOURCE: Report of Condition and Income.

their assets away from loans and into government securities, with the portfolio share
of securities increasing by more than 5
percentage points over the past two years.
Banks also have responded to the current environment with attempts to bolster
their noninterest income. In fact, as shown
in Chart 4, over the past three years, noninterest income has grown at a fairly steady
rate. Banks' traditional source of earnings
— n e t interest income, or interest income
less interest expense—also has increased
but at a slightly slower pace. The divergence in the growth rates between noninterest income and net interest income at
banks in recent years may indicate that
U.S. banks are relying less on lending and
more on nonlending activities to generate
income. One example of this is the extra
fee income earned from the increasing
presence of banks in the mutual funds

400 -

200 - H
'80

'81

'82

'83

'84

'85

'86

'87

'88

'89

'90

NOTE: 1993 data are for June 30.
DATA SOURCE: Report of Condition and Income.

'91

'92

'93

2 See Bernanke and Lown (1991) and Gunther, Lown,
and Robinson (1991).

For more on how the yield spread may affect bank
lending activity, see Short, Gunther, and Moore (1993).
3

Chart 4
Noninterest Income and Net Interest Income
at U.S. Insured Commercial Banks
Index, 1990:1 = 100
140

100

1990

1991

1992

1993

DATA SOURCE: Report of Condition and Income.

business. Mutual funds have grown quite
rapidly over the past few years, with banks
becoming important players in this industry.
Banks' involvement in offering mutual
funds is representative of the type of expanded powers that many analysts argue
banks will need to compete effectively in
the global financial marketplace.

Mutual Funds: A Brief Overview
A mutual fund is a company that pools
the funds of its investors by selling shares
to many individuals and then uses the proceeds to purchase a diversified portfolio of
stocks and bonds. Mutual funds have been
available to investors since the first fund
was organized in Boston in 1924. In 1936,
the Securities and Exchange Commission

was directed by Congress to conduct a
study of investment companies (or mutual
funds). That study culminated in the Investment Company Act of 1940, which set the
stage for the rules and regulations under
which mutual funds could operate.
The number of shares in a mutual fund
changes daily as the fund issues new shares
when more money is invested in it and as
the fund redeems shares from its investors.4
Mutual funds offer the advantages of diversification to their shareholders, as well as
the benefit of lower transaction costs that
result from buying large blocks of stocks
and bonds. Originally, mutual funds invested
only in common stocks, but many now
specialize in a wide range of debt instruments as well. In addition, some funds may
invest in a particular industry or may invest
in particular segments of the economy,
such as small businesses. 5
In 1940, total mutual fund assets were
$448 million, with 296,000 shareholder
accounts. Mutual funds reached $1 billion
in assets in 1945 and 1 million accounts in
1951. Since then, mutual funds have recorded impressive growth rates, especially
over the past twenty years. Chart 5 shows
the growth in mutual fund assets since the
mid-1970s. Assets grew from a total of

Chart 5
Mutual Fund Assets
Total net assets, billions of dollars
1,600

An open-end investment company continuously
issues and redeems its shares. Another type of investment company is the closed-end fund, which does
not redeem its shares but usually offers a fixed number of nonredeemable shares that are bought and sold
on a stock exchange.
4

' 7 5 ' 7 6 ' 7 7 ' 7 8 ' 7 9 ' 8 0 '81 ' 8 2 ' 8 3 ' 8 4 ' 8 5 ' 8 6 ' 8 7 ' 8 8 ' 8 9 ' 9 0 ' 9 1 ' 9 2

For a description of the types of mutual funds
available, see Investment Company Institute (1993).
5

DATA SOURCE: Investment Company Institute.

about $50 billion in 1975 to more than $1.5
trillion in 1992. The mutual fund industry is
now the third largest financial industry in
the nation, behind commercial banks and
life insurance companies.
Initially, mutual funds only offered
investors an opportunity to invest in the
stock and bond markets. These so-called
long-term funds are composed of equity
funds that invest in stocks, and bond and
income funds that invest in either corporate
bonds or government bonds. In the early
1970s, though, an important variation—the
money market mutual fund—was introduced. This type of mutual fund invests in
short-term debt instruments of very high
quality, such as U.S. Treasury bills, commercial paper, or bank CDs. A key feature
of these funds is that their shareholders
may write checks on the value of their
holdings, although there generally are
restrictions on both the minimum denomination and the number of transactions
allowed. Tax-exempt money market funds,
whose income is exempt from federal
income taxes, were introduced in 1976.
Chart 6 shows the total net assets in these
four major categories of mutual funds since
1975. All of these categories of mutual
funds have enjoyed rapid growth over the

Chart 6
Mutual Fund Industry Total Net Assets
Billions of dollars

Tax-exempt money market funds
Taxable money market funds
Bond and income funds
Equity funds

1,000 800 -

Chart 7
Distribution of Mutual Fund Assets
by Type of Fund

28%

Equity funds
f~| Bond and income funds

Taxable money market funds

Tax-exempt money market funds

DATA SOURCE: Investment Company Institute.

past two decades. Over the past ten years,
though, there has been a significant shift in
the distribution of mutual fund assets. In
1982, more than two-thirds of the assets in
mutual funds were in taxable money market funds ( C h a r t 7), while the longer-term
mutual funds accounted for approximately
one-fourth of total assets in mutual funds.
By 1992, however, equity funds and bond
and income funds together accounted for
two-thirds of total net mutual fund assets,
with the bond and income funds proving
the most popular with investors and the
taxable money market funds declining in
relative importance.
Mutual funds have grown rapidly over
the past two decades. And within this time
span, investors seemed to prefer the longterm funds that invest in equities and bonds.
Many analysts argue that much of the growth
in mutual funds over the past few years
can be traced to developments in the banking industry. The sharp decline in rates on
retail deposits at banks has probably encouraged savers to seek higher yields through
investments in mutual funds. 6 At the same

400

200

'75 '76 '77 ' 7 8 '79 '80 '81 '82 '83 ' 8 4 '85 '86 '87 '

DATA SOURCE: Investment Company Institute.

6 It is also possible that the activities of the Resolution
Trust Corporation in resolving insolvent thrift institutions contributed to the flow of deposits out of CDs
and into mutual funds. See Duca (1992).

time, the weakness in lending activity may
have altered banks' investment strategies.
To maintain their market share and retain
customer relationships, banks themselves
developed avenues to offer mutual funds.

Banks' Offerings of Mutual Funds
Regulatory issues. The entiy of banks
into the mutual funds business involves
several legal and regulatory issues. Shares
in mutual funds are considered securities,
not deposits. Therefore, both the courts
and regulators have tended to restrict banks
and bank holding companies from certain
activities associated with mutual funds
because of the Glass-Steagall Act.7 Over
the past several years, as banks and their
holding companies have increasingly sought
entiy into the mutual funds industry, some of
these restrictions have been relaxed, while
others have been maintained. The result is
a rather complicated and still-evolving set
of arrangements that govern bank involvement in the mutual funds industry.
Banks' recent move into mutual funds
does not represent the first time they have
sought an interest in this area of financial
services. A 1971 Supreme Court ruling
CInvestment Company Institute v. Camp)

The Glass-Steagall Act, or the Banking Act of 1933,
prohibits commercial banks from underwriting or
dealing in corporate securities. The act, in effect,
separated the activities of commercial and investment
banks.
7

An investment advisor is employed by a mutual
fund to offer professional advice on the fund's investments and asset-management practices.

In 1982, the Office of the Comptroller of the Currency authorized Security Pacific National Bank to
establish a discount brokerage subsidiary with no geographic restrictions imposed on it. In 1983, the Federal
Reserve Board granted similar authority to bank holding companies by permitting BankAmerica Corporation to acquire the discount brokerage firm of Charles
Schwab and Company. These rulings by regulators
were upheld by subsequent Supreme Court decisions.
9

For more on these issues, see Mack (1993).

held that banks cannot organize an investment company (or what is commonly
referred to as a mutual fund). This ruling
precluded banks from providing these companies with either initial capital or management interlocks.
Despite these restrictions, banks and bank
holding companies were able to become
increasingly active in the mutual funds business. In 1972, amendments to certain Federal
Reserve regulations authorized bank holding companies to act as investment advisors
to mutual funds (later upheld by a 1981
Supreme Court decision, Board of Governors of the Federal Reserve System v. Investment Company Institute). The Office of
the Comptroller of the Currency also has
authorized national banks to act as investment advisors to mutual funds.8
Later decisions by both the courts and
regulators expanded the scope of banks'
brokerage activities by allowing banks and
bank holding companies to establish discount brokerage services that enabled them
to sell mutual fund shares, along with other
types of securities.9 In 1987, the Comptroller authorized national banks to provide
investment advice to their customers regarding mutual funds that are advised by
affiliated investment advisors. In 1992, the
Federal Reserve Board of Governors allowed
bank holding companies or their subsidiaries to provide investment advice and
other services to customers that invest in
any bank-advised mutual fund. In allowing
these activities, however, the Board requires
that a number of disclosures be made to
customers, including the nature of the bank
holding company's relationship to the
mutual fund and the fact that mutual funds
are uninsured investment products.10
Characteristics of bank mutual funds.
Banks have managed to offer mutual funds
while still adhering to the restrictions in the
Glass—Steagall Act. Two types of mutual
funds are available through banks. Nonproprietary funds are those offered by a bank
that is serving as a broker or middleman
for a single fund or a number of different
funds. These types of mutual funds are

created by nonbank organizations that
organize, underwrite, and provide investment advice to the fund. A bank's involvement in offering nonproprietary mutual
funds can range from renting lobby space
to an unaffiliated broker to selling the
shares through a brokerage firm affiliated
with the bank. By offering nonproprietary
funds, banks are able to supplement their
income by receiving a portion of any sales
commission.
The second categoiy of mutual funds
available through banks is known as proprietary mutual funds. In this class of mutual
funds, the bank or an affiliate actually serves
as an advisor and administrator of a fund,
but the fund itself is underwritten by an
unaffiliated distributor and organized by an
unaffiliated third party. The bank then
markets these mutual funds through the
bank's brokerage subsidiary or affiliate.
The officers of these mutual funds cannot
be associated with the bank, although they
may be associated with nonbank affiliates
of the bank holding company.
Despite some regulatory restrictions, commercial banks have become veiy aggressive participants in the mutual fund markets.
Federal regulatory agencies do not collect
any official data on banks' offerings of
mutual funds. However, it has been estimated that in 1992, commercial banks
accounted for approximately one-third of
all net sales of mutual fund shares, and that
mutual funds offered through banks held
11 percent of mutual fund assets, up from
2 percent in 1987." Chart 8 indicates how
banks' offerings of mutual funds varied by
bank asset size. Based on a 1992 survey
of almost 7,000 banks across the United
States, banks' offerings of mutual funds
appear to be positively related to the size
of the bank. 12 About 35 percent of banks
in the sample with assets of $50 million to
$100 million offer mutual funds, while
virtually all banks with assets exceeding
$50 billion offered either proprietary or nonproprietary mutual funds. Chart 9 reveals
that smaller banks tend to offer a greater
proportion of nonproprietary funds, while

Chart 8
Proportion of Banks Offering
Mutual Funds, 1992
Percent

120

$100 million

$250 million

$500 million

$1 billion

$50 billion

Bank asset size
DATA SOURCE: American Brokerage Consultants.

the larger institutions appear to favor the
proprietary mutual funds.
From very little involvement several
years ago, banks are emerging as important
providers of mutual funds. Part of the push
by banks into mutual funds represents an
attempt to maintain market share and to
boost earnings in the face of continued
relative declines in the role of commercial
bank lending in U.S. financial markets.
Whether the source of these lending declines
is related to a lack of demand for bank
loans or to increases in the cost of supplying loans, many analysts have expressed
concern that given the current structure of
the U.S. banking system, banks may continue to play a smaller role in the direct provision of credit to businesses. If these trends

The Economist

(1993, 74).

12 Thrift institutions can also offer mutual funds. However, their involvement has been substantially less
than that of banks. For example, almost half of the
banks in this sample offer mutual funds, while slightly
less than one-fifth of thrift institutions surveyed offer
these services.

Chart 9
Proportion of Banks Offering Proprietary
and Nonproprietary Mutual Funds, 1992
Percent
80

by bank loans has declined fairly steadily
since its peak in the mid-1970s. Increases in
both national and international competition,
brought about in large part by regulatory
restrictions and by advances in technology,
have hurt U.S. banks' profitability and competitive position. Glauber (1993, 34) summarizes this process:
Twenty-five years ago, banks had the
financial services playing field pretty much to
themselves. Today it is very crowded. Automobile companies through finance subsidiaries (for example, GMAC, Ford Credit) offer
auto loans to consumers nationwide. Fidelity
and other mutual fund groups offer nationwide
deposit and checking accounts through money
market mutual funds. Merrill Lynch offers

$100 million

$250 million

$500 million

$1 billion

Bank asset size
•

Proprietary

fl|

Nonproprietary

mortgages nationwide, while General Electric,
through General Electric Credit Corporation,
makes small business loans nationwide. And
Goldman, Sachs offers commercial paper—the

DATA SOURCE: American Brokerage Consultants.

equivalent of bank loans for large, high-quality
corporations—nationwide. Confronting this
competition, banks are prohibited from oper-

continue, they could present important
challenges for economic policymakers with
regard to both banking and monetary policy.

Some Implications for Banking
and Monetary Policy
B a n k regulation a n d b a n k competitiveness. The competitive structure of the
entire financial system has undergone
radical change over the past two decades.
Chart 10 shows one aspect of the changing
nature of U.S. financial markets. The proportion of total private-sector debt financed

13 For more on the FDICIA, see Kaufman and Litan
(1993) and Short and Robinson (1992).
14 See U.S. Department of the Treasury (1991). The
Treasury report was mandated by the Financial Institutions Reform, Recovery, and Enforcement Act of 1989.
The proposal also contained other, related issues,
including the overextension of deposit insurance, a
fragmented regulatory system, and an undercapitalized deposit insurance fund.

ating branches across state lines and generally
from dealing in securities....Structural changes
such as these require structural reform.

Structural reform efforts have been undertaken recently. The Federal Deposit Insurance Corporation Improvement Act (FDICIA)
was passed in December 1991 for the dual
purposes of recapitalizing the deposit
insurance fund and strengthening bank
supervision to prevent a replay of the
banking difficulties of the 1980s. However,
FDICIA has been the source of much controversy. Industry analysts and the academic
community have expressed concern that
the act will further erode the U.S. banking
industry's competitive position by sharply
increasing banks' regulatory burdens in an
effort to constrain risk-taking at federally
insured depository institutions.13
Prior to the passage of FDICIA, the U.S.
Department of the Treasury offered a reform
package designed to improve the competitiveness of the banking system. Included
in the recommendations were three fundamental structural changes. 14 The first was

the repeal of the McFaclden Act restrictions
on interstate branching so that banks could
enjoy the efficiency gains and geographic
diversification benefits associated with
opening branches across state lines. The
second structural change recommended
was repeal of the Glass-Steagall Act and
modification of the Bank Holding Company
Act of 1956 to permit a bank, through the
holding company structure, to affiliate with
firms engaged in offering a broad array of
financial services. Safety and soundness
concerns would be addressed through a
system of "fire walls" that would be erected
between the federally insured bank and
these affiliates. Moreover, only well-capitalized banks would be allowed to engage in
such activities. Finally, the Treasury recommended allowing commercial ownership
of banks as a way to introduce even more
diversification into the banking industry,
which would create strong financial services
companies capable of competing on more
equal footing with financial firms around
the world. Once again, Glauber provides a
succinct summary of what these reform
measures might achieve. 15
The combination of reforms [was] intended
to have two broad effects. First, they would
resize and reshape an industry that is strikingly
and dysfunctionally fragmented. The 11,500
banks in the United States, compared with
200-300 in Britain and Germany, amount to
fifty banks per million people, versus six banks
per million people in the other two countries.
For the United States this has meant recurrent
excess capacity and operating inefficiencies
leading to destructive competition in lending to
ever less credit-worthy customers. Second, the
reforms would allow banks to compete with
other financial services companies for the
profitable business they have lost, including
securities underwriting, mutual fund management, and insurance product sales.

The Treasury's reform proposals would
have offered banks greater opportunities to
compete on a more level playing field with
their nonbank competitors. Although FDICIA
did not implement these proposals, recent

Chart 10
Ratio of Bank Debt to Total Private Debt
Percent

DATA SOURCE: CITIBASE.

statements by the comptroller of the currency indicate that legislation may be introduced that includes proposals for interstate
branching and some expanded powers for
banks. 16 Continued efforts to address the
secular decline in the nation's banking
industry should serve to increase the flow
of credit, especially to small businesses,
whose primary source of financing is
banks. Moreover, attempts to strengthen
the competitive position of the nation's
banks may have important implications for
monetary policy, because it is possible that
a less competitive banking system has
presented some difficulties in the Federal
Reserve's conduct of monetary policy.
Bank competitiveness a n d m o n e t a r y
policy. For the past several years, growth
in the M2 measure of the money supply
has been unusually weak. Until recently,
the Federal Reserve has used this measure
of the money supply as its guide to conducting monetary policy, because it bore
a close relationship to overall economic
activity. However, that relationship now

See Glauber (1993, 36).

See Rehm (1993).

appears to have broken down in the sense
that M2 growth has been much weaker
than what would be expected based on
historical relationships between M2, economic activity, and interest rates. Many
analysts have cited the unusual conditions
in the U.S. banking industry as factors in
the puzzling behavior of M2. In particular,
the sharp decline in small time deposits
and the concomitant rapid increase in mutual
funds are factors that may be affecting
monetary policy. Because these developments reflect, at least in large part, the lack
of lending activity at banks, interest has increased in the extent to which developments
on the asset side of banks' balance sheets
can affect monetary policy. The next article
in this issue addresses this topic.

Conclusions
In many respects, 1992 was a banner
year for the U.S. banking industry. Banks
earned record profits and strengthened
their capital positions, while at the same
time recording large reductions in their

troubled assets. But despite these healthy
trends, bank lending activity remains
depressed. Both demand- and supply-side
factors lie behind these lending declines,
which appear to have accelerated the
relative decline in the importance of banks
in financial markets. In an attempt to maintain market share, banks have responded
by offering mutual funds to their customers.
Despite a somewhat cumbersome regulatory structure that makes it more difficult
for banks than other institutions to offer
these types of financial services, banks'
role in the mutual funds industry appears
to be growing quite rapidly. These developments call attention once again to efforts
to reform the U.S. banking industry, with
the goal of maintaining safety and soundness while enhancing banks' competitiveness in an increasingly integrated financial
marketplace. Moreover, these developments in the banking sector appear to have
played a role in affecting the course of
money growth, making it more difficult to
interpret movements in current measures
of the money supply.

References
Bernanke, Ben, and Cara S. Lown (1991),
"The Credit Crunch," Brookings Papers on
Economic Activity, no. 2, 204-39Duca, John V. (1992), "The Case of the
Missing M2," Federal Reserve Bank of
Dallas Economic Review, Second Quarter,
1-24.
The Economist (1993), "Up and Up Until It
Popped," August 14, 73-74.
Glauber, Robert R. (1993), "FDICIA: The
Wheels Came Off on the Road Through
Congress," in Assessing Bank Reform:
FDICIA One Year later, ed. George G.
Kaufman and Robert E. Litan (Washington,
D.C.: Brookings Institution), 33-41.
Gunther, Jeffery W., Cara S. Lown, and
Kenneth J. Robinson (1991), "Bank Credit
and Economic Activity: Evidence from the
Texas Banking Decline," Federal Reserve
Bank of Dallas Financial Industry Studies
Working Paper no. 5—91 (Dallas, December).
Investment Company Institute (1993), Mutual
Fund Fact Book (Washington, D.C.).

Kaufman, George G., and Robert E. Litan,
eds. (1993), Assessing Bank Reform: FDICIA
One Year later (Washington, D.C.: Brookings Institution).
Mack, Phillip R. (1993), "Recent Trends in
the Mutual Fund Industry," Federal Reserve
Bulletin 79 (November): 1001-12.
Rehm, Barbara A. (1993), "Ludwig Plans
Legislation on Banking This Year," American Banker, September 21.
Short, Genie D., Jeffery W. Gunther, and
Robert R. Moore (1993), "Governments,
Deficits and Banking," Federal Reserve
Bank of Dallas Financial Industry Issues,
Second Quarter.
, and Kenneth J. Robinson (1992),
"Banking Conditions and Legislation,"
Federal Reserve Bank of Dallas Financial
Industry Issues, Third Quarter.
U.S. Department of the Treasury (1991),
Modernizing the Financial System: Recommendations for Safer, More Competitive
Banks, February.

The Relationship
Between Bank Lending
and Money Growth:
Were Things Different
in the 1980s?
Kenneth J. Robinson
Senior Economist and Policy Advisor
Financial Industry Studies Department
Federal Reserve Bank of Dallas

rowth of M2 has been unusually weak
in recent years. In fact, this broad
monetary aggregate has been either at or
below the lower level of its targeted growth
path since early 1992 ( C h a r t I). 1 At the
same time, bank lending activity has also
exhibited unprecedented weakness. In
inflation-adjusted terms, the total value of
bank loans has declined by about 7 percent
since the end of 1989- Bank loans have
continued to decline even though the U.S.
recession ended in the spring of 1991Many analysts have pointed to these bank
lending declines as evidence of a credit
crunch that stems from the difficulties experienced by the banking sector from the
mid-1980s until fairly recently. The unusual
weakness in bank lending activity, coupled
with continued weak growth in the M2
measure of the money supply, has raised
the possibility of a connection between
these two economic variables. Has the lack
of bank lending activity affected the course
of money growth in the U.S. economy?
Two contrasting views about the effect
of bank lending on money growth have
emerged. One view states that increases in
bank lending are not a causal factor in M2
growth but rather, just the opposite. In the
traditional monetarist view, increases in
economic activity that are generated by

increases in money growth ultimately lead
to an increase in bank lending activity.
Therefore, increases in lending do not lead
to increases in M2 but are instead the result
of prior increases in the money supply.
An alternative view of how bank lending activity and money growth are related
stresses the role of financial deregulation.
In this view, the elimination of interest rate
restrictions on various types of consumer
deposits in the late 1970s and early 1980s
allowed bank lending to affect money
growth. In particular, a large share of the
components of M2—small time deposits—
can now be treated as managed liabilities
in the same way that large time deposits
(those over $100,000) have been since their
interest rate restrictions were removed in
the early 1970s. By varying the yield offered
on small time deposits, banks can now
actively manage components of M2 in response to their funding needs. If banks are
now using components of M2 as part of
their managed liabilities, then the supply of
M2 could be affected by the amount of bank
lending activity. In addition, the emergence
of widespread banking difficulties in the
mid-1980s could be expected to heighten
this effect of bank lending on M2. An adverse financial shock to the U.S. banking
system may have made banks increasingly
unable or unwilling to extend loans, leading
banks to bid less aggressively for deposits.
In this deregulated environment, then, monetary policy could be affected by developments on the asset side of the banking
system's balance sheet.
The periods before and after financial
deregulation provide an opportunity to
investigate these two viewpoints about
bank lending activity's effect on money
growth. We might expect that before the
advent of deregulation in the early 1980s,
bank lending activity would not have

1 M2 consists of the sum of currency in circulation,
checking accounts, savings deposits, small (less than
$100,000) time deposits, and money market mutual
funds.

Chart 1
M2 and Target Growth Cones
Billions of dollars

DATA SOURCE: Federal Reserve Statistical Release H.6.

played an important role in M2 growth.
Banks, at least in the short run, would be
more or less forced to accept the deposits
that their customers chose to supply at the
regulated rates. If, however, banks now
use components of M2 as managed liabilities, then we might expect to find bank
lending to be a significant factor in affecting M2 growth after the early 1980s. Moreover, the relationship between bank lending
and the narrow monetary aggregate Ml
should not be affected by financial deregulation, because Ml does not contain any
managed liability components. 2
Monthly data for 1959-92 are used in
an attempt to estimate the impact of bank
lending on money growth, measured as Ml
and M2. I first examine the relationship
between bank lending and M2 growth in
the period before financial deregulation
and then look for any changes in this relationship following deregulation. I also
examine movements in bank lending and

2 Ml consists of currency in circulation plus checkable deposits, which are not generally used by banks
as part of their managed liabilities. See Boyd and
Gertler (1993) for a discussion of the characteristics of
bank deposits.

money growth since the onset of banking
difficulties to judge whether this development heightened any impact of bank lending activity on Ml or M2 growth.
The results indicate that before deregulation, bank lending did not greatly affect
money growth. Even after the phase-in of
financial deregulation, I do not find bank
lending affecting M2 growth significantly.
However, I find some evidence that after
the onset of widespread banking difficulties, bank lending activity emerged as an
important factor in M2 growth. Finally, in
none of the periods I examine is Ml growth
affected by bank lending activity. These
results suggest that the combination of
banking-sector difficulties and the increase
in banks' ability to compete for funds in
M2 may have had an impact on the growth
rate of the M2 measure of the money
supply. But, as expected, developments
on the asset side of the banking system's
balance sheet do not appear to be a factor
in Ml growth.

Bank Lending and Money Growth
Two views regarding the relationship
between bank lending and money growth
can be identified. The first of these views is
a traditional monetarist interpretation of the
connections between money growth and
bank lending activity. According to this view,
the Federal Resetve's ability to control the
growth rate of the money supply is not related to banks' lending activity. As Friedman
(1992) points out, even in an extreme case
in which banks use reserves provided by the
Federal Reserve only to purchase government securities, these securities are matched
by increases in deposits on the liability side
of the balance sheet, just as an increase in
loans would be matched by increases in
deposits. Friedman sums up this view by
stating that "banks' unwillingness or inability
to make loans does not hinder the Fed from
increasing M2. (To give a historical example:
M2 rose by 26 percent from 1934 to 1936;
loans by commercial banks fell a trifle.) In
any event, increases in loans by banks are

generally a result of increases in spending
stimulated by higher monetary growth,
rather than a cause of the greater spending."
Thus, under this interpretation, bank lending would not be expected to play much of
a role in explaining money growth. Rather,
money growth is viewed as the cause of
bank loan growth, because increases in the
money supply lead to increases in overall
spending, which then lead to an increase in
loan demand.3
A contrasting explanation of the relationship between money growth and bank
lending stresses that the financial deregulation of the late 1970s and early 1980s has
altered the relationship between lending
activity and M2. Before mid-1978, all of the
time and savings deposits included in M2
were subject to Regulation Q restrictions in
the form of ceilings on the interest rates
banks could offer on these deposits. Then,
beginning in mid-1978, banks were allowed
to offer small certificates of deposit with
interest rates tied to the U.S. Treasury bill
rate. The Depository Institutions Deregulation and Monetary Control Act of 1980
hastened the deregulation process by calling
for the gradual phase-out of Regulation Q
ceilings on all deposit accounts (with the
exception of demand deposits). Thus, by the
early 1980s, the deregulation of deposits
included in M2 was in full swing. Some
economists have found evidence that banks
then began to treat retail deposits in M2
more like managed liabilities, or more like
large CDs have been treated since their
interest rate restrictions were removed in
the early 1970s. That is, banks began to
actively manage their small time deposits to
fund lending activity, rather than passively
accept the amount depositors chose to
offer at regulated yields.4 As a result, M2
growth in the postderegulation environment
may now be more affected by developments on the asset side of the banking
system's balance sheet than was the case
prior to elimination of Regulation Q interest
rate ceilings.
Banking difficulties experienced beginning in the mid-1980s could also be ex-

pected to enhance the effects of bank
lending on M2 growth. Since the mid-1970s,
there has been a sustained downward
trend in the amount of intermediation
through banks, with little discernible effect
on M2 growth. However, the declining role
of banks in the intermediation process was
accelerated by the banking difficulties of
the 1980s and has manifested itself in a
marked slowdown in bank lending activity.
Weak loan demand has been an important factor behind the recent sharp slowdown in lending activity by U.S. banks. 5
But higher capital requirements, higher
deposit insurance premiums, and increased
regulatory oversight and scrutiny—all outgrowths of the unprecedented banking
problems of the 1980s—have frequently
been cited as important factors in banks'
declining role in financial markets.
If banks are playing a lesser role in the
credit channeling process, due to either
supply-side or demand-side factors, then
they would tend to depress deposit rates
and issue fewer deposit liabilities.6 The
Federal Reserve summarized this process
in its 1993 Monetary Policy Report to the
Congress. "Banks' unaggressive pricing of
deposits reflected substantial paydowns of

For an examination of the relationship between the
money supply and economic activity, see Friedman
and Schwartz (1963). Mishkin (1992) provides an
overview of the role of banks in the money-supply
creation process.
3

See Judd and Trehan (1987), Motley (1988), and
Higgins (1992) for the changing nature of small time
deposits in M2 before and after deregulation.

5 See Gunther, Lown, and Robinson (1991) and
Bernanke and Lown (1991) for the role of economic
activity in explaining bank lending declines.

One factor that could lie behind weak money
growth is anemic reserve growth. However, reserves
have grown at a considerable pace. Since our current
economic recovery began in spring 1991, total reserves have increased at a compound annual rate of
more than 10 percent, while the monetary base has
increased more than 8 percent, much faster than in
previous economic recoveries.
6

bank debt by households and businesses,
which kept loan demand low and banks'
need for funds to finance them quite limited.
In addition, banks and thrift institutions
have been discouraged from going after
deposits by the rising cost of issuing deposits to make loans; among the factors
accounting for this increase have been increases in deposit insurance rates and higher
capital ratios occasioned by market and
regulatory forces." 7
The periods before and after the introduction of deposit interest rate deregulation
provide an opportunity to test whether the
asset side of the banking system's balance
sheet can affect money growth. To investigate this issue, I examine the relationship
between bank lending activity and money
growth, after accounting for other factors—
such as economic activity and interest
rates—that may affect these two variables.
If the traditional monetarist proposition
holds, then bank lending should not be
important in explaining subsequent money
growth either before or after deregulation,
because according to this view, money
growth is the cause of bank loan growth.
However, if variations in bank lending are
found to be important in accounting for

Board of Governors (1993b, 186). Duca (1993)
presents a theoretical argument of how a less competitive banking system can affect money growth.
7

Lending activity at thrift institutions would also be
expected to play a role in the money supply process
in a manner analogous to bank loans. However, data
on thrift loans were not available until relatively
recently.
8

As in Judd and Trehan (1987), the postderegulation
sample begins in July to avoid the portfolio readjustments that might have arisen due to the introduction
of nationwide NOW accounts. Also, dummy variables
were included in the estimation for the time period
December 1982-February 1983 to account for the
introduction of money market deposit accounts and
Super-NOW accounts. Starting the postderegulation
period in July 1981 also avoids the possible distortions
arising from the change in Federal Reserve operating
procedure in 1979 and the imposition of credit controls in March-July 1980.
9

movements in the money supply after the
period of financial deregulation, then this
would be consistent with the view that
banks are treating components of M2 more
as managed liabilities, which implies that
bank lending can affect a broad monetary
aggregate such as M2.
The effect of the banking sector's difficulties on M2 growth can also be examined
by investigating the relationship between
bank lending and M2 growth since the
emergence of banking difficulties in the
mid-1980s. The effects of developments on
the asset side of banks' balance sheets on
M2 growth could be enhanced during a
period of unusual financial-sector distress.
Finally, as a check on the robustness of my
results, I estimate the relationship between
bank lending activity and Ml growth. Because the accounts in Ml are not used as
managed liabilities, I do not expect to find
a significant effect on Ml from bank lending in either the prederegulation or postderegulation period. For the same reason,
I do not expect bank lending to affect Ml
growth during the period associated with
banking-sector difficulties.

Empirical Results
Sample periods analyzed. I investigate
the relationship between money supply
growth and bank lending activity using
monthly data for the period 1959-92. 8
Following Judd and Trehan (1987), I estimate this relationship over two different
time periods. The prederegulation period is
1959—77, because during this period Regulation Q ceilings were in place on the components of M2. The postderegulation period
extends from July 1981 through 1992 and
includes the period when Regulation Q
ceilings were being eliminated on various
types of deposit accounts. 9 I also examine
how M2 and bank lending might be related
since the onset of banking difficulties,
using the period 1985-92. It was during
this period that U.S. banks began to experience increased financial difficulties, first on
a regional basis in the Southwest, then on

a more widespread basis as difficulties
surfaced in the Northeast and on the West
Coast. The number of banks on the FDIC's
"problem list" peaked in 1987 at 1,575, representing 11 percent of insured U.S. banks.
Problems at larger institutions also emerged
during this time, with the assets of commercial banks on the problem list topping
out at $610 billion in 1991.10 Recently, these
problems have abated, and the U.S. banking industry has been enjoying a sustained
recovery.
Statistical model: Does m o n e y predict
loans o r vice versa? In an effort to investigate the relationship between the money
supply and bank lending, I construct statistical models designed to estimate the extent
to which past movements in bank loans
are important in accounting for money
supply growth and, conversely, the extent
to which past movements in the money
supply are important in explaining bank
lending activity.
If the strict monetarist position holds,
then in both the prederegulation and postderegulation periods, prior money growth,
as measured by M2, should be important
in explaining bank lending activity, since
monetarists believe increases in the money
supply stimulate economic activity and
loan demand. However, I would not expect
to find that prior growth in bank loans is
important in predicting M2 growth. Alternatively, because of regulations in place during
the prederegulation period that prohibited
the components of M2 from being used as
managed liabilities, I would not expect
developments on the asset side of banks'
balance sheets to affect M2. However, if,
in the postderegulation period, banks are
using the components of M2 (especially
small time deposits) as managed liabilities,
then I would expect to find that bank
lending activity is an important factor in
explaining subsequent M2 growth. I also
expect that bank lending activity is important in accounting for M2 growth during
the period of banking-sector difficulties.
Finally, an examination of the relationship
between lending and the narrow monetary

aggregate Ml might also shed some light on
the role of financial deregulation in money
growth. I would not expect the relationship
between money growth, as measured by
Ml, and bank lending activity to be affected
by the ability of banks to offer market rates
on deposits, because banks do not tend to
use Ml as part of their managed liabilities.
In estimating the relationship between
bank lending activity and money growth, I
also need to control for other factors that
may affect these two variables. In the statistical tests that follow, I include a measure
of economic activity and interest rates, in
addition to bank lending and the money
supply. The economic activity measure is
U.S. personal income, while the interest
rate used is the federal funds rate.11 The
top panel of Table 1 shows the statistical
significance of these variables in predicting
M2 growth. The entries in the table are
marginal significance levels derived from
the results of testing the hypothesis that
past values of the variable in question are
not important in explaining current money
growth. A small number implies that I am
fairly confident, in a statistical sense, that
the hypothesis is rejected. A marginal significance level of 0.05 or less is usually indicative of statistical significance. For
example, in the prederegulation period,
1959-77, the marginal significance level of
lagged values of M2 in predicting current
M2 is less than 0.01. This implies that I am
more than 99 percent confident that I can
reject the hypothesis that lagged values of
M2 are not important in helping to predict
current M2. This result is not too surpris-

10 Duca (1992) examines how the activities of the
Resolution Trust Corporation in resolving insolvent
thrifts may have affected M2 growth.
11 Augmented Dickey-Fuller tests indicate that each
variable used in the statistical tests is differencestationary. Therefore, all variables are expressed as
the first-difference in the logs of the series, with the
exception of the interest rate variables, which are
expressed as the difference in levels. The lag length
for each equation is determined using a chi-square
test suggested by Sims (1980).

Appendix Table
Relationship Between Bank Lending and M2

Marginal significance levels of variables
predicting M2
Variables
Income
M2
Loans
Fedfunds

1959-77

1981-92

1985-92

.58
.00
.91
.00

.37
.00
.57
.00

.46
.00
.00
.00

Marginal significance levels of variables
predicting loans
Income
M2
Loans
Fedfunds

.12
.00
.00
.03

.27
.11
.02
.00

.71
.13
.08
.00

Lag lengths: 1 9 5 9 - 7 7 = 8 lags
1 9 8 1 - 9 2 = 8 lags
1 9 8 5 - 9 2 = 4 lags
NOTES: Income is total U.S. personal income; M2 is the broad monetary aggregate. Loans are
total loans and leases at commercial banks, and fedfunds is the federal funds rate. All
data are obtained from Citibase, with the exception of the loans series prior to 1973,
which was obtained from the Statistical Abstract of the United States.

ing, because it is often the case that one of
the best predictors of a current economic
variable is prior values of that very series.
The first column of numbers in Table 1
also shows that movements in the federal
funds rate are statistically significant in predicting M2, while income is not. More important for our purposes, as expected, prior
movements in bank lending activity are not
statistically significant in accounting for M2
growth during this prederegulation period.
The bottom panel of Table 1 shows the
marginal significance levels of these same
variables in predicting bank lending. In the
prederegulation period, prior movements
in loans, M2, and the federal funds rate are
important in predicting lending activity,
with at least a 95 percent certainty or greater.
Thus, in the 1959-77 period, bank lending
was not an important factor in subsequent
M2 growth. Rather, as Friedman suggests,
18

during this period, it was increases in M2
that contributed to lending activity.
The second column of numbers in Table
1 shows marginal significance levels of
those variables in predicting M2 and loans
in the postderegulation period. Because
banks could offer market interest rates on
the components of M2 during this period,
it might be expected that the broad monetary aggregate would be more influenced
by bank lending activity. However, the
marginal significance level of loans in predicting M2 is only 0.57, indicating that
lending activity was not a statistically important factor in predicting subsequent M2
growth during this period. Also, from the
bottom panel, I am now only about 90
percent confident that past values of M2
are significant in explaining bank lending
activity. These results do not offer much
support for the hypothesis that the ability to

treat components of M2 as managed liabilities has significantly affected M2 growth.
But it also appears that the causal role of
M2 growth in loan activity was diminished
during the postderegulation period.12
Did banking difficulties play a role?
Bank lending activity is not found to be
statistically significant in affecting M2 growth
in the postderegulation environment. I can
make use of these same statistical techniques to judge the extent to which the
emergence of banking difficulties may have
heightened bank lending's role in money
growth. Whatever the underlying causes
of lending declines at banks, even though
banks are now free to offer market rates on
their retail deposits, with a smaller volume
of lending activity to fund banks have not
pursued retail deposits as aggressively as
in previous periods, possibly making M2
more susceptible to variations in bank loans.
The third column of numbers in Table 1
reveals which variables are statistically significant in explaining M2 and loans with
the onset of banking difficulties.13 In this
period, we see that lagged values of M2
and the federal funds rate are statistically
significant in explaining current M2. Even
more importantly, we can see that past
bank lending activity is now a significant
factor in predicting subsequent M2 growth.
That is, I am more than 99-percent confident that prior movements in bank loans
are a significant factor in predicting current
M2 growth during the period associated
with banking-sector distress. In the bottom
panel, prior movements in the federal funds
rate, and possibly in loans, would be considered statistically important in explaining
lending at banks. These results suggest that
the onset of banking difficulties in a period
in which banks can bid for retail deposits
has resulted in a statistically significant
effect of bank lending on growth in the
broad monetary aggregate.
I can also estimate the same model, but
with an interest rate spread variable in
place of the federal funds rate. The spread
variable is defined as the difference between
the rate on ten-year Treasury bonds and

three-month Treasury bills. Some analysts
argue that the term structure of interest
rates may affect growth in the broad monetary aggregate. This may be especially true
in the recent past, given the unusually
wide spread between long- and short-term
interest rates. It is also possible that this
steep yield curve may be an important factor
in bank lending activity if it encourages
banks to substitute long-term government
securities for loans.14 Table 2 reports marginal significance levels of the variables
that might predict M2 growth and bank
loans, but with this spread variable used as
the interest rate measure. As before, from
the top panel of Table 2, prior growth in
loans is not statistically significant in predicting M2 growth in either the prederegulation or postderegulation environment.
However, in contrast to the prior results, in
the period coinciding with banking-sector
difficulties, lending is not a statistically
significant variable in M2 growth. Questions about the importance of the spread
variable, though, arise from the fact that
this variable is insignificant in this period,
although it is significant in the earlier time

12 The introduction of money market certificates in
June 1978 and small saver certificates in July 1979,
with interest rates tied to short-term U.S. Treasury
securities, represented the first time that banks could
bid for retail deposits. These instruments proved very
popular when introduced, especially among mediumand small-sized commercial banks (Simpson and
Parkinson 1984). However, deregulation continued
into the early 1980s through such avenues as lower
minimum denominations and shorter maturities. By
October 1983, all interest rate restrictions and minimum denomination requirements on time deposits of
more than seven days were eliminated. Therefore, I
also estimated the models beginning in November
1983, but could not find a statistically significant effect
of bank lending on money growth.
13 To judge the robustness of the results with regard to
banking difficulties, I also estimated the models beginning in 1987 and 1988, which did not affect the
conclusions.

See Feinman and Porter (1992) and Short, Gunther,
and Moore (1993).
14

Appendix Table A2
Relationship Between Bank Lending and M2
Marginal significance levels of variables
predicting M2
Variables
Income
M2
Loans
Spread

1959-77

1981-92

1985-92

.88
.00
.85
.00

.12
.00
.46
.03

.15
.00
.37
.43

Marginal significance levels of variables
predicting loans
Income
M2
Loans
Spread

.20
.00
.00
.43

.53
.22
.00
.07

.95
.02
.01
.10

Lag lengths: 1959-77 = 8 lags
1981-92 = 7 lags
1985-92 = 3 lags
NOTES: Spread is the difference between the yield on ten-year Treasury bonds and the yield on
three-month Treasury bills. For other variables, see notes to Table 1.

periods.15 And, from the bottom panel of
Table 2, the spread variable is not statistically significant in the conventional sense
in any of the time periods.
How are bank loans and Ml related?
I find some evidence that financial deregulation coupled with banking difficulties appears
to have provided an avenue for lending
activity at banks to affect money growth, as
measured by M2. Further evidence on the
robustness of my results can be obtained by
examining the relationship between bank
lending activity and the narrow monetary
aggregate, Ml. Although Regulation Q ceilings
no longer apply to most accounts in Ml,
because the components of the narrow

15 1

also estimated the same models found in Tables 1
and 2, but with the addition of total reserves in the
statistical models. My conclusions are not affected, as
revealed in the tables in the Appendix. The same
conclusions are reached if the monetary base rather
than reserves is included in the statistical tests.

monetary aggregate are mostly transactionstype deposits and less sensitive to interest
rate variations, I would not expect banks to
use Ml as part of their managed liabilities.
As a result, lending activity should not be a
factor in affecting Ml growth, either before or
after deregulation or in periods of financialsector difficulty.
Table 3 reports marginal significance
levels from statistical tests of the relationship
between bank lending and Ml growth. As
before, I also control for economic activity
and interest rates in these tests. From the
top panel of Table 3, prior movements in
bank loans are not an important factor in
predicting growth in the narrow monetary
aggregate, either before or after the introduction of deregulation. Moreover, in the
period coinciding with banking difficulties,
I do not find that lending activity played a
significant factor in accounting for Ml growth.
The bottom panel of Table 3 shows that in
the 1959-77 period, prior Ml growth was

Appendix Table
Relationship Between Bank Lending and M2

Marginal significance levels of variables
predicting M2
Variables
Income
M1
Loans
Fedfunds

1959-77

1981-92

1985-92

.00
.00
.53
.00

.15
.00
.41
.00

.60
.00
.68
.03

Marginal significance levels of variables
predicting loans
Income
M1
Loans
Fedfunds

.48
.00
.00
.79

.61
.36
.00
.00

.82
.76
.00
.05

Lag lengths: 1959-77 = 6 lags
1981-92 = 4 lags
1985-92 = 7 lags
NOTES: M1 is the narrow monetary aggregate. For other variables, see notes to Table 1.

important in predicting loan growth, as were
past values of bank loans. However, the
importance of Ml disappears in the later
time periods. As expected, financial deregulation did not appear to alter the effect
of bank lending activity on Ml growth.16
Comparing Tables 1 and 3 though, reveals
that the impact of Ml growth on loan activity
did appear to change more dramatically
after deregulation than was the case with
M2, perhaps reflecting the changing relationship between Ml and overall economic
activity beginning in the early 1980s.17

Policy Implications
Some of the statistical tests employed for
this article suggest that a combination of
financial deregulation and banking-sector
difficulties altered the relationship between
bank lending activity and M2 growth. I find
some evidence consistent with the hypothesis that adverse shocks to the banking
system, coupled with the ability to treat

retail deposits in M2 as managed liabilities,
have been factors in the slowdown in M2
growth.18 Until recently, the Federal Reserve has emphasized the use of the M2
measure of the money supply as a guide in
its conduct of monetary policy. M2 was the
primary information variable for monetary
policy because its overall relationship with
economic activity was fairly stable and predictable, at least until several years ago.
Recently, that relationship has become

16 1 also estimated the same model as in Table 3, but
with the addition of total reserves. My conclusions are
not affected, as revealed in Appendix Table A3.
17 For more on the changing relationship between Ml
and economic activity, see Friedman (1988).
18 A more conclusive test of this hypothesis would
entail examining how M2 would have responded to
banking difficulties in the period before deregulation.
However, no comparable banking-sector shocks
occurred during this earlier time period, making such
an investigation impossible.

much more problematic.19 While growth in
M2 has been very weak, overall economic
activity has continued to expand. Thus, the
current slow rate of M2 growth does not
appear to reflect a moderation or tightening
in the overall flow of credit to the economy.
Rather, it is likely that weak M2 growth has
resulted from the continued diversion of
credit flows from banks and other intermediaries to the capital markets, through
which households appear to be investing a
larger share of their wealth.20
Some of the factors that have affected
the growth rate of M2 may be temporary in
nature. The banking system now appears
to be recovering. Portfolio adjustments by
both households and businesses to the debt
buildups of the last decade are proceeding.
These developments would be expected to
lead to increases in lending activity in the
future and to increases in M2. However, if
the U.S. banking system finds itself at a competitive disadvantage in the provision of
credit, it could continue to see its role in
the intermediation process decline. In such
a scenario, the value of M2 as a policy guide
would continue to be questionable. In light
of these potential developments, several
proposals to alter the definition of the broad

19 Prior to the early 1980s, the Federal Reserve concentrated on the Ml monetary aggregate in its conduct of
monetary policy. By 1987, however, increasingly difficult to interpret movements in the narrow monetary
aggregate led the Federal Reserve to discontinue setting a target range for Ml. Moreover, in congressional
testimony in July 1993, Federal Reserve Chairman
Alan Greenspan stated that the Federal Reserve was
temporarily downgrading measures of the money
supply as guides to its policy formulations and instead
would rely on a variety of data, including real interest
rates, in setting monetary policy.

Board of Governors (1993a). One example of this type
of household behavior is the explosive growth of mutual funds, especially those now being offered by banks.
20

For more on this proposal, see Duca (1992).

Board of Governors (1993a, 843).

Motley (1988) and Judd and Trehan (1987).

monetary aggregate have been made. One
proposal is to add components of mutual
funds to M2 to account for the importance
of nonbank intermediaries in the provision
of credit.21 Comparing the quarterly growth
rates of M2 with the sum of M2 and bond
and stock funds over the past two and a
half years reveals that the latter variable
has grown at nearly a 5-percent annual
rate, compared with less than a 2-percent
annual rate for the current M2 measure.
However, even after making this adjustment,
the new total remains volatile, indicating
that other forces are also at work affecting
either M2 or mutual funds or both.22
Another proposal is to classify the components of the money supply based on the
distinction between those components that
have a stated maturity term versus those
that do not. This proposal would alter the
M2 definition by excluding small time deposits from M2 and combining them with
the large time deposits found in M3 (plus
other accounts in M3 such as term repurchase agreements and term Eurodollars) to
create a new aggregate "term M3." In an era
of deregulated deposit interest rates, this
new combination of monetary assets could
be a more reliable guide for monetary policy
because it contains a more similar grouping
of financial assets—those primarily used
by banks as managed liabilities.23
During the 1980s, financial deregulation
and adverse banking-sector conditions
combined to make the M2 measure of the
money supply more sensitive to developments on the asset side of the banking
system's balance sheet. As a result, a number
of serious empirical difficulties have arisen
in attempting to predict the growth rate of
M2 and also in the use of M2 as a guide for
the conduct of monetary policy. Unfortunately, it is still not known whether these
developments are temporary in nature or
are due to fundamental changes in the
financial structure, with a diminishing role
for banks. For these reasons, the Federal
Reserve has placed less emphasis on M2,
while economists continue to search for a
reliable guide for monetary policy.

Appendix Table A1
•:
Relationship Between Bank Lending and M2
Marginal significance levels of variables
predicting M2
Variables

1959-77

1981-92

1985-92

.21
.00
.72
.00
.53

.54
.00
.62
.01
.85

.37
.00
.02
.00
.16

Income
M2
Loans
Fedfunds
Reserves

••fit > f

-•

Marginal significance levels of variables
predicting loans
Income
M2
Loans
Fedfunds
Reserves

.26
.00
.00
.62
.66

.41
.15
.02
.01
.88

.61
.24
.11
.02
.57

Lag lengths: 1959--77 = 6 lags
1981--92 = 8 lags
1985--92 = 4 lags
NOTES: Reserves are total reserves of depository institutions. For other variables, see
notes to Table 1.

;

Appendix Table A2
Relationship Between Bank Lending and M2
Marginal significance levels of variables
predicting M2
Variables

1959-77

1981-92

1985-92

.88
.00
.84
.00
.59

.16
.00
.58
.06
.86

.34
.00
.56
.53
.56

Income
M2
Loans
Spread
Reserves

Marginal significance levels of variables
predicting loans
Income
M2
Loans
Spread
Reserves

.26
.00
.00
.44
.90

Lag lengths: 1959-77 = 8 lags
1981-92 = 7 lags
1985-92 = 4 lags
NOTES: See notes to Appendix Table A1

V"' . -

.67
.19
.00
.14
.76

.54
.17
.04
.12
.18

Appendix Table A3
Relationship Between Bank Lending and M1
Marginal significance levels of variables
predicting M1
Variables

1959-77

1981-92

1985-92

.00
.00
.64

.46
.18
.77
.01
.97

.53
.16
.89

Income
M1
Loans
Fedfunds
Reserves

.00

.32

.00

.81

Marginal significance levels of variables
predicting loans
Income
M1
Loans
Fedfunds
Reserves

.42
.02
.00
.72
.80

Lag lengths: 1959-77 = 6 lags
1981-92 = 8 lags
1985-92 = 8 lags
NOTES: See notes to Appendix Table A1.

.39
.26
.00
.02

.76

.66
.63
.03
.09
.55

References
Bernanke, Ben S., and Cara S. Lown (1991),
"The Credit Crunch," Brookings Papers on
Economic Activity, no. 2, 204-39.
Board of Governors of the Federal Reserve
System (1993a), "Monetary Policy Report
to the Congress," Federal Reserve Bulletin
79 (September): 827-45.
(1993b), "Monetary Policy Report
to the Congress," Federal Reserve Bulletin
79 (March): 167-87.
Boyd, John H., and Mark Gertler (1993),
"U.S. Commercial Banking: Trends, Cycles,
and Policy," NBER Working Paper Series,
no. 4404 (Cambridge, Mass.: National
Bureau of Economic Research, July).
Duca, John V. (1993), "Regulation, Bank
Competitiveness, and Episodes of Missing
Money," Federal Reserve Bank of Dallas
Economic Review, Second Quarter, 1-23.
(1992), "The Case of the Missing
M2," Federal Reserve Bank of Dallas Economic Review, Second Quarter, 1-24.
Feinman, Joshua N., and Richard D. Porter
(1992), "The Continuing Weakness in M2,"
Federal Reserve Board Finance and Economics Discussion Series, no. 209 (September).
Friedman, Benjamin M. (1988), "Lessons
on Monetary Policy from the 1980s," Journal of Economic Perspectives 2 (Summer):
51-72.

University Press for the National Bureau of
Economic Research).
Gunther, Jeffery W., Cara S. Lown, and
Kenneth J. Robinson (1991), "Bank Credit
and Economic Activity: Evidence from the
Texas Banking Decline," Federal Reserve
Bank of Dallas Financial Industry Studies
Working Paper no. 5-91 (Dallas, December).
Higgins, Biyon (1992), "Policy Implications
of Recent M2 Behavior," Federal Reserve
Bank of Kansas City Economic Review,
Third Quarter, 21-36.
Judd, John P., and Bharat Trehan (1987),
"Portfolio Substitution and the Reliability of
Ml, M2 and M3 as Monetary Policy Indicators," Federal Reserve Bank of San Francisco Economic Review, Summer, 5-29.
Mishkin, Frederic S. (1992), The Economics
of Money, Banking, and Financial
Markets
(New York: HarperCollins Publishers).
Motley, Brian (1988), "Should M2 Be Redefined?" Federal Reserve Bank of San
Francisco Economic Review, Winter, no. 1,
33-51.
Short, Genie D., Jeffery W. Gunther, and
Robert R. Moore (1993), "Governments,
Deficits, and Banking," Federal Reserve
Bank of Dallas Financial Industry Issues,
Second Quarter.

Friedman, Milton (1992), "Too Tight for a
Strong Recovery," Wall Street Journal
October 23, A12.

Simpson, Thomas D., and Patrick M. Parkinson (1984), "Some Implications of Financial
Innovation in the United States," Board of
Governors of the Federal Reserve System
Staff Studies, no. 139 (September).

, and Anna Jacobson Schwartz
(1963), A Monetary History of the United
States, 1867-1960 (Princeton: Princeton

Sims, Christopher A. (1980), "Macroeconomics and Reality," Econometrica 48
(June): 1-48.

Full text of Financial Industry Studies : December 1993

FRASER