View original document

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

MAY/JUIMt 1994

ECONOMIC PERSPECTIVES




A review from the
Federal Reserve Bank
of Chicago

Is banking a declining industry?
A historical perspective
Funding small businesses
through the SBIC program

FEDERAL RESERVE BANK
OF CHICAGO

Contents
Is banking a declining industry?
A historical perspective...............................................................................2
G eorge G. K aufm an and
Larry R. M o te

Growth comparisons based on reported assets and
liabilities have created a widespread perception that the
banking industry is declining relative to other financial
institutions. Using alternative measures of bank size or
output to examine the question, this article concludes
that, while there is clear-cut evidence of such a decline
through the early 1960s, the picture is much more mixed
for recent decades. The nature of banks’ business has
clearly changed in a number of ways, including the
unbundling of many services and the entry of banks into
many off-balance-sheet activities. As a result, a growing
number of banking activities are not captured by
traditional measures of bank output.

Funding small businesses
through the SBIC program........................................................................ 22
Elijah B rew er III and Hesna G enay

The Small Business Investment Company (SBIC) program
provides banking organizations with a unique opportunity to
make equity investments in small businesses. This article
compares the financial characteristics, investment patterns,
and profitability of bank-owned and non-bank-owned SBICs
in the 1980s. Using less Small Business Administration
leverage, bank-owned SBICs funded activities that were
difficult to finance using traditional sources.

ECONOMIC PERSPECTIVES
Karl A. Scheld, Senior Vice President and
Director of Research
Editorial direction
Janice Weiss, editor
David R. Allardice, regional studies
Steven Strongin, economic policy and research
Anne Weaver, administration
Production
Nancy Ahlstrom, typesetting coordinator
Rita Molloy, Yvonne Peeples, typesetters
Kathleen Solotroff, graphics coordinator
Roger Thryselius, Thomas O’Connell,
Lynn Busby-Ward, John Dixon, graphics
Kathryn Moran, assistant editor




M a y /J u n e 1994 V o lu m e X V III, Issue 3

ECONOMIC PERSPECTIVES is published by
the Research Department of the Federal Reserve
Bank of Chicago. The views expressed are the
authors’ and do not necessarily reflect the views of
the management of the Federal Reserve Bank.
Single-copy subscriptions are available free of
charge. Please send requests for single- and
multiple-copy subscriptions, back issues, and
address changes to the Public Information Center,
Federal Reserve Bank of Chicago, P.O. Box 834,
Chicago, Illinois 60690-0834, or telephone
(312) 322-5111.
Articles may be reprinted provided source is
credited and the Public Information Center is sent a
copy of the published material.
ISSN 0164-0682

Is banking a declining
industry? A historical
perspective

G eo rg e G. K a u fm an and Larry R. M o te

Regulation has been widely
blamed for contributing to the
decline of the commercial
banking industry in the United
States. Before one can evalu­
ate the truth or falsity of this accusation, it is
necessary to determine whether banking is
indeed a declining industry. To answer this
question, we examine a number of different
measures of changes in the size of the banking
industry in the United States during the twenti­
eth century.
Few industries are as closely associated in
the public image with the growth of modem
economies as is commercial banking. Bankers
have been widely caricatured as pulling the
strings behind the “bosses of industry” and
have been viewed with suspicion or fear in
many quarters. Indeed, it would be difficult to
understand the elaborate set of regulations
intended to restrict the growth and thereby the
power of commercial banks in the United
States without first understanding the wide­
spread distrust of banks and banking dating
back to early U.S. history. In the 1800s,
some states even went so far as to ban banks
altogether.
But the image of bankers as all-powerful
has changed dramatically in recent years, espe­
cially among bankers themselves, their regula­
tors, and the business community. Over the
past decade, banking in particular and deposi­
tory institutions in general have come to be
viewed as declining. This widespread percep­
tion is based primarily on their declining share
of some measure of assets or liabilities for all

 2


financial institutions. An example is provided
by figure 1, which shows the decline since
1952 in the combined total assets of U.S.chartered commercial banks and U.S. offices of
foreign banks as a percentage of the assets of
all financial institutions. Several presentations
at the Federal Reserve Bank of Chicago’s 1993
Conference on Bank Structure and Competi­
tion noted this decline (Federal Reserve Bank
of Chicago 1993).
The common view is that banks are losing
out to a wide range of nonbank competitors
such as finance companies, mutual funds, and
private pension funds that are offering tradi­
tional types of banking products more effi­
ciently, either because technological advances
have eliminated advantages previously enjoyed
by banks or because these competitors are free
of costly regulations imposed on banks. The
source of any decline is important in judging
its welfare implications. If banking were a
declining industry because of market forces, as
was the fate of horse-drawn carriages, the rail­
roads, and coal mining, then it would be of
concern to bankers who lose their jobs but of
little public policy concern.1 Indeed, attempt­
ing to prevent the decline would reduce aggreGeorge G. Kaufman is the John F. Smith, Jr.
Professor of Economics and Finance at Loyola
University of Chicago and consultant to the Fed­
eral Reserve Bank of Chicago. Larry R. Mote is an
economic adviser and vice president at the Feder­
al Reserve Bank of Chicago. The authors wish to
thank Lisa Hardesty for compiling data for the
study. They also gratefully acknowledge the
helpful contributions of Doug Evanoff and Steven
Strongin.

ECONOMIC PERSPECTIVES

gate economic welfare. On the
FIGURE 1
other hand, if the decline were
Commercial banks’ declining share of assets
attributable to excessive regula­
of all financial institutions
tion that prevents banks from
percent
operating more effectively, or
from introducing newer prod­
ucts for which demand is grow­
ing rapidly, then aggregate eco­
nomic welfare would be reduced
and the decline would be a legit­
imate public policy concern.
However, before anyone
writes the banking industry’s
epitaph, it may be useful to look
a little more closely at the evi­
dence. This article examines a
number of data series to deter­
mine whether banking is or is
Source: Board of Governors of the Federal Reserve System,
not a declining industry and, if it
"Flow of funds accounts," various years.
is, whether the decline is the
result of market forces or of
sures of the size of the banking industry are
excessive and discriminatory regulation. Al­
l) assets; 2) employment; and 3) revenues,
though the evidence is not clear-cut, several of
earnings, and value added. We present data for
the alternative measures examined in this arti­
each of these measures in turn, together with
cle suggest that banking may not be declining.
an indication of their strengths and shortcom­
A lte rn a tiv e m e a su re s o f th e size o f the
ings. Because assets and related balance sheet
b a n k in g in d u stry
data have been, by far, the most frequently
There are serious conceptual and practical
used measures of the size of the banking indus­
problems in measuring the output of any indus­
try, we consider them first.
try.2 Because of the intangible nature of the
output provided by service industries general­
A s s e ts
ly, and banks in particular, the problem of
Total assets, earning assets, and total de­
measuring output has occupied banking schol­
posits have all been used at one time or another
ars for decades. The problem arises in a num­
as measures or indexes of banking output.
ber of contexts— for example, in calculating
Such measures accord with the common per­
shares of output in local markets for antitrust
ception of banks as firms that use “inputs” such
purposes, in measuring banking output and
as deposits, labor, and capital to produce “out­
costs for the purpose of determining the rela­
puts” primarily in the form of loans and invest­
tionship between size and efficiency, and in
ments. This more or less commonsense view
calculating the industry’s contribution to gross
has greatly influenced the analysis of banks as
domestic product (GDP). Some of the issues
firms and the measurement of bank output.
related to choosing the most appropriate mea­
It is therefore not surprising that total
sure of the size of the banking industry are
assets or deposits, or some variant thereof, has
discussed in the accompanying box.
long been the most popular measure of the size
Because each of the measures used in past
or output of the banking industry. Balance
studies appears to contain conceptual or practi­
sheet data for banks are readily available at
cal problems—difficulties in obtaining appro­
frequent intervals and serve as the basis for the
priate data, shortcomings in the quality of
widely used flow of funds data published quar­
available data, lack of comparability of data
terly by the Federal Reserve.3 Moreover, in
over time, or a failure of the data to correspond
contrast with most other businesses, the prod­
closely to the theoretical concept they are used
ucts and services of banks have traditionally
to measure— this article analyzes a number of
been closely related to the size and composi­
them. Among the most frequently used mea­

FEDERAL RESERVE



BANK OF CHICAGO

3

tion of their asset and liability portfolios.
Throughout the nineteenth century and the first
half of the twentieth century, the activities of
commercial banks were largely limited to ac­

cepting and processing deposits and making
loans and investments. Indeed, those functions
still account for a substantial if declining pro­
portion of the typical bank’s activities.

Issues in the m easurem ent o f bank output
Analysts have tried to measure bank output for
two purposes: to assess economies of scale in bank­
ing, and to calculate banks’ contribution to gross
domestic product (GDP). Earning assets or total
assets were the most widely used measures of bank
output in early studies of the relationship between
scale—that is, size measured in terms of output—and
cost in banking. However, critics pointed out that an
equal number of dollars of credit extended for a
given period of time, as would be reflected in asset
measures, do not necessarily imply equal output in
an economic sense. For example, a given dollar
amount of consumer instalment loans does not neces­
sarily represent the same output as the same dollar
amount of loans to a large corporate customer. A
consumer loan is likely to require much more risk­
bearing, information gathering, credit analysis, and
bookkeeping per dollar of loan principal than a loan
to a large corporate customer (Benston 1965; Greenbaum 1967). Thus, simply adding the dollar amounts
of all the loans on a bank’s books would be adding
apples and oranges. The only dimensions of output
that could be said to be identical for loans of differ­
ent types but equal dollar amounts outstanding are
the amount and duration of the postponement of
consumption by one group of economic units that is
a prerequisite for making a loan enabling another
economic unit to consume beyond its current income.
A similar objection applies to adding the outstanding
values of loans of the same type but of different
sizes.
Another relatively obvious criticism of balance
sheet measures of output is that output is a flow,
measured in quantity or value per unit of time,
whereas assets are a stock at a particular point in
time. Only in banking and related financial indus­
tries have assets been widely used as a measure of
output and relative importance. In other industries,
sales or revenues are the preferred measure for some
purposes, including the calculation of market shares
for antitrust analysis. For most other purposes, there
is fairly general agreement among economists that
the most relevant measure of the size of an industry
is its value added, or contribution to the total output
of the economy. Although there may be issues
affecting the industry for which assets or liabilities

4



or employment are more useful measures—for exam­
ple, changes in the importance of banking as a chan­
nel for monetary policy or banking’s role in creating
new jobs—the contribution of banking to GDP is a
more general measure of the industry’s importance in
the economy.
In the search for a single index of banking
output, considerable progress had been made by the
late 1960s toward achieving consensus that some
variant of bank revenue, rather than assets, was the
preferred measure of final output. Of course, if one
wishes to measure commercial banks’ contribution to
final output as measured by value added, rather than
the value of final output per se, it is necessary to
subtract from revenues the value of purchased inputs.
Nevertheless, the persistence of conflicting views
concerning the nature of the output of financial
institutions led to a continuing debate over which
measure of value added was most appropriate. Seri­
ous questions were raised about the “liquidity princi­
ple” used by the U.S. Commerce Department’s
Office of Business Economics to measure the contri­
bution of banks and other financial intermediaries to
GDP. According to the liquidity principle, bank
output consists only of interest and other services to
depositors, not to borrowers (Hodgman 1969). How­
ever, as Hodgman pointed out,
a closer examination of banking activity and
banking costs will reveal that f i n a n c i a l
s e r v i c e s (rather than deposits or loans) are
the products of banking. . . . When banks
are viewed as financial service firms we see
that the banking product sold to b o r r o w e r s is
not only credit but intermediation and that a
portion of a bank’s interest receipts is paid
by the borrower to cover the costs of inter­
mediation rather than as a payment for
liquidity or consumption foregone by the
ultimate lender. This portion of “interest”
received by banks should be regarded as part
of their gross value product in the national
accounting sense. The remainder of interest
paid to banks will, under competitive condi­
tions, be paid in turn by the banks to the
ultimate lenders who are depositors and

ECONOMIC PERSPECTIVES

Support for the view that banking is de­
clining in relative importance is typically based
on the downward trend in the share of total
assets at all financial institutions (see figure 1),

stockholders. Conceptually, therefore, the
interest received by banks should be
included in gross product originating rather
than set to zero by definition.*

net

But while Hodgman and others were gaining
considerable support for some variant of revenue as
the single index of output in banking, the literature
on bank costs moved in a very different direction.
First, researchers began to estimate separate cost
functions for individual functional areas within the
bank (Benston 1965; Bell and Murphy 1968). Later,
they began to use the translog and related multi­
product cost functions (Benston, Hanweck, and
Humphrey 1982). Neither of these approaches re­
quired using a single index of banking output.
The objections noted above to using assets as
the measure of banking output apply fully only to
attempts to aggregate many different types of loans
or other banking products into a single index of
banking output. As long as each category of loans is
relatively homogeneous—for example, consumer
loans that do not vary greatly in size or riskiness—it
may be unobjectionable to use total loans outstand­
ing as a measure of the output associated with that
category. The reason is that, if all the loans in a
particular category are identical in size, maturity,
risk, and other important characteristics, then the
number of loan accounts, total revenue, and other
alternative measures of output associated with that
category would be proportional to the amount out­
standing. Thus, asset measures may be a reasonable
choice for the estimation of multi-product cost func­
tions that utilize a large number of output categories
rather than a single index of overall output. Indeed,
recent studies comparing the performance of stock
and flow measures of output in bank cost studies
have concluded that there is not much empirical
evidence to favor one over another (Humphrey
1992). But it is still true that this approach finesses
the issue rather than addressing it; there is no pre­
sumption that a dollar of consumer loans represents
the same output as a dollar of commercial and indus­
trial loans.
*Hodgman 1969, p. 191.

FEDERAL RESERVE BANK OF CHICAGO




or particular categories of assets accounted for
by commercial banks or by all depository insti­
tutions. As table 1 shows, the decline in
banks’ share of short-term business credit, the
traditional bread and butter lending activity of
commercial banks, has been even more dramat­
ic than that of banks’ share of total assets.
The data are frequently presented with such a
sense of urgency that one might be led to
believe that the decline in asset share is a
sudden, recent development that requires an
immediate response.
However, a closer review of the evidence
shows that neither this decline nor the concern
over it is of recent origin. A pioneering study
of U.S. financial institutions conducted by
Raymond Goldsmith in the 1950s and 1960s
reported that commercial banks’ share of total
assets of financial intermediaries had declined
from 71 percent in 1860 to 63 percent in 1900
and 32 percent in 1963 (Goldsmith 1958,
1969).4 Table 2 shows commercial banks’
share of the total assets of financial institutions
for selected dates from 1860 through 1993.
Thus, the more recent decline in the market
share of commercial banks should not be over­
ly surprising. Much of it simply reflects the
fact that, because banks were the first major
financial institution in the United States, it
was virtually inevitable that they would lose
market share over time to newer types of finan­
cial institutions offering previously unknown
products, for example, pension funds and
mutual funds.
Nor is evidence of a decline in banks’
market share limited to the United States. As
the data in table 3 indicate, banks’ share of
total liabilities of financial intermediaries in
the United Kingdom also declined between
1913 and 1991. Similar declines have occurred
in most of the 30 major foreign countries ana­
lyzed by Goldsmith (1969).
But even before Goldsmith’s study, bank­
ers lamented that the traditional business of
banking was shrinking and that if banks were
to survive they would have to expand the scope
of their activities. Thus, as corporations relied
increasingly on internal sources of funds and
less on bank loans in the 1920s, banks expand­
ed their lending to include consumer and resi­
dential real estate loans. The same decade also
saw the rapid expansion of banks and bank
securities affiliates into the underwriting and
distribution of corporate securities. Retrospec-

5

TABLE 1

Composition of short-term credit market debt of
nonflnancial corporate business
(1950-92)
1950

1960

Nonbank
finance loans

1980

1990

.......percent—

(..........
Bank loans

1970

1992

■
.........)

91

87

83

71

59

59

6

9

9

14

17

18

sive regulation of banks in combi­
nation with tax and other incen­
tives enjoyed by some nonbank
competitors (Robertson 1968).
Like most other research on the
issue, both the Gurley and Shaw
study and that of the Comptrol­
ler’s office relied on balance sheet
data to support the thesis that
banking was in decline.

Improving the asset measure
Assets probably give an ade­
quate picture of the size of the
Foreign loans 1
9
9
-banking industry in the nineteenth
Bankers'
century. However, there is reason
acceptances
2
2
2
5
3
2
to believe that even for the first
Total
100
100
100
100
100
100
half of the twentieth century and
certainly for more recent decades,
B illion
324
951
882
dollars
20
43
125
reported assets give a distorted
and incomplete view of the output
Source: Board of Governors of the Federal Reserve System,
Balance Sheets for the U.S. Economy, 1945-92, March 10, 1993.
of the commercial banking indus­
try. The asset figures typically
used in these analyses include
tively, and almost certainly incorrectly, some
only bank-owned or “on-balance-sheet” assets.
blamed the banking collapse of the early 1930s
But banks also manage or otherwise service
on the entry by banks into some of these new
assets owned by others. These activities are
and unfamiliar activities.
referred to as “off-balance-sheet.” The eco­
The 1950s were marked by renewed con­
nomics of banking, as opposed to accounting
cern over banks’ loss of business, this time to
conventions, suggests that banks should be
then rapidly growing nonbank depository insti­
measured by some measure that reflects the
tutions, such as savings and loan associations,
full range of their activities, such as revenues,
which at the time were free of such regulatory
income, or value added. However, because onrestrictions as interest rate ceilings on deposits
balance-sheet assets are the most readily avail­
and reserve requirements. Indeed, the widely
able and frequently used yardstick of the size
discussed Gurley-Shaw thesis held that if regu­
of the banking industry, it may be worthwhile
lation continued to restrain traditional banks
to try to correct banks’ aggregate balance sheet
relative to their nonbank competitors, the result
for a number of failings, in particular its exclu­
would be the development of more and more
sion of important off-balance-sheet activities,
“near monies” such as time and savings depos­
and bring it closer to what might be called an
its at thrift institutions, and the continued
“economic balance sheet.” We will discuss
shrinkage of the banking industry (Gurley and
some of these exclusions and the adjustments
Shaw 1955, 1956, 1960). Eventually, a point
needed to correct for them in the following
would be reached at which monetary policy, if
sections.
it continued to operate only through traditional
Bank trust services
banks, would lose its effectiveness. A quick
Among the most important off-balanceexamination of table 2 shows that, rather than
sheet activities are bank trust services, perhaps
preempting commercial banks, savings and
the oldest off-balance-sheet activities engaged
loan associations and savings banks are them­
in by banks in the United States. Indeed, a
selves now declining rapidly in importance. A
number of banks began as strictly trust compa­
history of the Office of the Comptroller of the
nies providing only trustee or fiduciary servic­
Currency published in 1968 also remarked on
es and expanded into deposit and other banking
the loss of market share by commercial banks
services primarily as an accommodation to
in the postwar period and attributed it to exces­
their customers. Today, few strictly trust comCom m ercial
paper

6




1

2

6

9

12

12

ECONOMIC PERSPECTIVES

TABLE 2

Share of assets of financial institutions in the United States
(1860-1993)
1860

1880

1900

1912

1929

1939

1948

1960

1970

1980

1993

--- ;

-------pe rce nt— .

(---C om m ercial banks

71.4

60.6

62.9

64.5

53.7

51.2

55.9

38.2

37.9

34.8

25.4

U.S.-chartered
banks and bank
holding
companies
U.S. offices of
foreign banks

71.4

60.6

62.9

64.5

53.7

51.2

55.3

37.6

37.2

32.4

21.7

0 .0

0 .0

0 .6

0 .6

0.7

2.4

3.7

14.0

13.6

12.3

19.7

20.4

21.4

9.4

1 1 .8

13.0

15.5

T h rift
in stitu tio n s

Savings and
loan associations

-

-

17.8

22.8

0 .0

-

-

18.2

14.8

2 .2

3.1

3.0

6 .0

4.2

4.7

2 0 .6

15.1

1 1 .8

8 .0

9.2

7.4

6.9

6 .0

0 .0

0 .2

0 .2

1 .1

1.4

\

Savings banks
Credit unions
Insurance
com panies

17.8

--

-

-

-

4.2 J
1.7

7.4a
2 .0

10.7

13.9

13.8

16.6

18.6

27.2

24.3

23.8

18.9

16.1

17.4

1 .8

10.7
3.1

13.6
3.0

14.8
3.8

23.5
3.7

2 0 .6

3.7

19.4
4.4

15.1
3.8

11.5
4.5

1 2 .8

8.9

9.4
4.5

_
_

_
_

„

_
_

2.4'=

1.9*

1.3*

2.9

3.5

3.6

14.9

-

-

-

-

-

-

-

-

-

-

-

-

-

3.5
3.5

-

-

-

-

-

-

-

-

-

3.4
1.5
1.9

14.2

-

2.9
2.9

-

--

-

-

-

-

-

b

-

-

0.0

0.0

0.7

2.1

3.1

-

-

-

-

0.4

0 .8

1 .6

-

-

0 .0

0 .0

0.3

1.3

Finance
com panies

-

0.0

0.0

0.0

2.0

Securities
brokers and
dealers

0.0

0.0

3.8

3.0

M o rtg ag e
com panies

0.0

2.7

1.3

Real es tate
in v e s tm e n t tru s ts

-

-

-

100.0

100.0

100.0

.001

.005

.016

Life insurance
Property/casualty
In vestm e n t
com panies

Mutual funds
Stock and bond
Money market
Closed-end funds
Pension funds

Private
State and local
government

4.6

1 0 .2

4.0
0.7

b

0 .2

9.7

13.0

17.4

24.4

6.4

8.4

12.5

16.7

1.5

3.3

4.5

4.9

7.6

2.2

2.0

4.6

4.8

5.1

4.7

8.1

1.5

1.0

1.1

1.2

1.1

3.3

1.2

0.6

0.3

0.1

b

b

0.4

0.2

-

-

--

-

0.0

0.3

0.1

0.1

100.0

100.0

100.0

100.0

100.0 100.0

.129

.281

.596

1.328

4.025 13.952

T o tal

(percent)

100.0 100.0

T o tal

(trillion dollars)

.034

.123

a
The end of the first quarter of 1993 was the last date for which data for savings and loan associations and
savings banks were reported separately. The figures for that date were: savings and loans, 6 .0 percent;
savings banks, 1.9 percent.
bData not available.
breakdown between open- and closed-end funds not available.
Sources: Data for 1860-1948 from Raymond W. Goldsmith, Financial Structure and Development, Studies in
Comparative Economics, New Haven, CT: Yale University Press, 1969, Table D-33, pp. 548-9. Data for 1960-1993
from Board of Governors of the Federal Reserve System, "Flow of funds accounts, " various years.

______

FEDERAL RESERVE BANK OF CHICAGO




7

either as an agent or as a trustee.5
Trust accounts whose assets are
Share of total liabilities of intermediaries,
managed by the bank are general­
United Kingdom
ly referred to as discretionary,
(1913-91)
while accounts that are in the
B uilding
Insurance
Pension
custody of the bank but managed
Year
Banks
societies
com panies
funds
by others are referred to as non(........... •................... percent-.............................. )
discretionary. At year-end 1992,
bank trust departments, trust
64
4
32
1913
companies, and thrift institutions
1930
61
8
31
-held $1.8 trillion of discretionary
12
1939
55
32
n.a.
assets and $7.7 trillion of nondis1960
43
12
30
14
cretionary assets; the commercial
1970
32
17
27
16
bank share was 87 percent of the
30
20
25
21
1980
former and 94 percent of the latter
28
17
26
26
1990
(Federal Financial Institutions
27
1991
27
18
26
Examination Council 1992). To­
tal trust assets serviced by com­
Note: n.a. indicates data not available.
mercial banks at year-end 1992
Source: Harold Rose, "The changing world of finance and its
problems," working paper no. 167-93, Institute of Finance and
totaled $8.8 trillion, more than
Accounting, London Business School, 1993, p. 29.
2.5
times the assets on the balance
sheets of banks. Moreover, bank
trust assets have expanded rapidly
in recent years, rising from $283 billion at
panies exist. To serve customers who wish to
year-end 1968 to $4.1 trillion in 1985 and $8.8
invest in securities other than bank deposits,
trillion in 1992. As figure 2 shows, the most
many banks have long operated trust depart­
rapid growth in recent years has been in nonments in which they provide fiduciary, invest­
ment, managerial, and custodial services for a
discretionary assets.
Banks face little competition for custodial
fee. Trust department assets are assets that the
bank manages or otherwise services but does
trust services. Few if any financial institutions
not own, and that therefore do not appear on
other than banks or trust companies offer them,
the bank’s balance sheet.
Trust accounts come in vari­
FIGURE 2
ous types and require different
amounts of servicing by the bank;
Assets of bank trust departments and trust companies
accordingly, they generate differ­
trillion dollars
ent amounts of fee income for
banks. Most trusts can be classi­
fied as personal trusts, estates, or
employee benefit trusts. The trust
contracts with the trustee bank for
the kind of services that it re­
quires. Almost all trust contracts
call for custodial and recordkeep­
ing services, including perfor­
mance measurement, timely valua­
tion, portfolio analysis, Employee
Retirement Income Security Act
(ERISA) and other required dis­
closure assistance, benefit dis­
Note: Discretionary and nondiscretionary accounts were not distinguished
bursement, cash management, and
from 1968 to 1977, and purely custodial accounts were excluded; only
discretionary assets were reported from 1978 to 1984. From 1985 on,
proxy monitoring.
nondiscretionary assets include purely custodial accounts.
Source: Federal Financial Institutions Examination Council, Trust Assets
Some banks also provide
of Financial Institutions, 1992.
investment management services,
TABLE 3

8




ECONOMIC PERSPECTIVES

TABLE 4

Ten largest bank trust departments, personal and employee benefit accounts
(1992)
Tru s t assets
Discre­
tio n a ry

N ondiscretio n a ry

T o tal

( ------------- ............b ill io n d o lla r s ----------

State Street Bank
(Boston)
Morgan Guaranty
(New York)
Bank of New York
Citibank (New York)
Northern Trust (Chicago)
Mellon Bank (Pittsburgh)
Bankers Trust (New York)
Chase Manhattan
(New York)
Boston Safe Deposit
Bank of America
(San Francisco)

Bank
assets

-------- )

D iscretion ary
tru s t assets/
bank assets
<...................... r a tio

T o tal
tru s t assets/
bank assets

------------ )

113

1,165

1,278

16.5

6.8

77.5

38
30
23
37
37
128

695
685
399
341
323

76.7
36.5
163.8
11.9
29.6
55.8

0.5

9.6
19.6

3.1
1.3
2.3

31.8

222

733
715
422
378
361
351

17
19

339
217

356
236

74.5
8.3

0.2
2.3

4.8
28.4

112

107

219

133.4

0.8

1.6

0.8
0.1

2.6
12.2
6.3

Source: Federal Financial Institutions Examination Council, Trust Assets of Financial Institutions, 1992.

and only a few trust companies are not char­
tered as banks. The ten largest bank trust de­
partments according to assets in personal and
employee benefit accounts are listed in table 4.
Two of the institutions— State Street Bank and
Boston Safe Deposit—are basically trust com­
panies rather than banks, although both have
bank charters. As the data in the table make
clear, the trust assets held by each of these
institutions greatly exceed the assets on its
balance sheet.
Banks also provide corporate trust servic­
es. Such services include serving as trustee for
the holders of corporate and municipal securi­
ties and as registrar, paying agent, transfer
agent, and recordkeeper for publicly issued
securities, including mutual funds. As trustee
for the debt security holders, the bank trust
department monitors scheduled payments for
timeliness and represents the holders’ interests
in disputes. The largest bank trust departments
in each corporate trust activity are listed in
table 5. Only as mutual fund transfer agents
do commercial banks appear to face serious
competition.
When personal trust assets held by bank
trust departments are added to balance sheet
assets for the years since 1900, the share of
assets held by banks increases somewhat, but

FEDERAL RESERVE



BANK OF CHICAGO

the downward trend is basically unaltered. For
the period since 1968, adding personal trusts
increases commercial banks’ share of total
assets by an amount ranging from 4.5 percent­
age points to 9 percentage points. However,
the downward trend remains and is in fact
intensified in percentage terms, since the ratio
of banks’ personal trust assets to total assets of
financial institutions fell by 50 percent over
that period, whereas banks’ share of balance
sheet assets fell only about a third. As figure 3
shows, essentially the same conclusion holds
when other assets are included over which
bank trust departments exercise managerial
discretion. These assets, which include rough­
ly one-third of employee benefit trust assets,
were nearly three times as large as personal
trust assets at year-end 1992 but have grown at
roughly the same pace in recent years. Thus,
while their inclusion substantially increases
banks’ average share of the market over the
period, it does little to moderate its downward
trend. Including trust assets over which banks
do not exercise managerial discretion would
moderate the decline, but because a narrower
range of services is provided in conjunction
with such accounts, they should not receive the
same weight as discretionary assets.

9

TABLE 5

Largest bank providers of corporate
trust services
(1992)
C orporate
and municipal
security trusteeship

Securities,
principal am ount

(billion dollars)

Citibank (New York)
First National Bank (Chicago)
Bank of New York
Chemical Bank (New York)
Bankers Trust (New York)
Texas Commerce (Houston)
Chase Manhattan (New York)
State Street (Boston)
Bank of America (San Francisco)
United States Trust (New York)
Stock or bond
transfer agent

PNC National
(Wilmington, DE)
Investors Fiduciary
Trust (Kansas City)
Firstar (Milwaukee)
Putnam Fiduciary (Boston)
Investors Trust (Boston)
NationsBank (Dallas)
Norwest Bank (Minnesota)
Wells Fargo (San Francisco)
Fifth-Third Bank (Cincinnati)
W ilmington Trust

197
160
149
132
99
94
92
91
87

Num ber of issues

Citibank (New York)
Chemical Bank (New York)
Bank of New York
Bankers Trust (New York)
Seattle-First National
Ameritrust Texas (Dallas)
American National (St. Paul)
Security Pacific (New York)
First Chicago Trust (New York)
First National of Boston
M utual fund
transfer agent

222

16,030
12,109
8,124
2,961
2,849
2,360
2,223
1,905
1,542
1,347

Num ber of issues

427
241
132
71
44
32
23
22

8
7

Source: Federal Financial Institutions Examination
Council (1992).

The reentry o f banks into securities activities
Primarily through the nonbank subsidiar­
ies of their parent holding companies, banks
have also been entering or reentering areas of
activity long considered off-limits to banks, at
least since the enactment of the Glass-Steagall
Act in 1933. Although banks’ own aggressive­

10



ness and inventiveness have been the driving
force in this development, much of it would
have been impossible without a series of rul­
ings by the Comptroller of the Currency and
the Board of Governors of the Federal Reserve
System (Kaufman and Mote 1990). As of
today, banking organizations, subject to some
quantitative restrictions that are more onerous
for smaller institutions, may serve as fullservice or discount securities brokers, may
underwrite and deal in a full range of munici­
pal and corporate debt, futures, options, swaps,
and other derivative securities as well as corpo­
rate equities, and may manage or broker (but
not underwrite or sponsor) mutual funds.
In recent years, commercial banks have
made significant inroads into the underwriting
of new securities. In 1993, two bank holding
companies—J. P. Morgan and Citicorp—
ranked among the top 15 underwriters of all
new domestic securities sold in the United
States. The remaining 13 were investment
banks. Three banks ranked among the top 15
underwriters of both investment-grade and
junk bonds and also among the top five under­
writers of asset-backed securities. It is of inter­
est to note that only one commercial bank
ranked among the top 15 underwriters of mu­
nicipal revenue bonds, most of which they
were not permitted to underwrite until recent
years. But this is the same number of banks
that rank among the top 15 underwriters of
municipal general obligation bonds, which
banks have always been permitted to under­
write. As we have noted elsewhere, it is only
since the late 1970s that banks have become
aggressive in pursuing securities underwriting
activities (Kaufman and Mote 1990). In part
this may reflect differences in corporate culture
between these activities and more traditional
commercial banking activities.
Banks and mutual funds
Mutual funds are one of the newer and,
since the late 1970s, more rapidly growing
types of financial institutions. As figure 4
shows, mutual funds have increased their share
of assets of all financial institutions from 1.8
percent in 1977 to 14.2 percent in 1993. This
rapid increase is the result of both a rapid in­
flow of new funds into mutual funds, in part
reflecting the introduction of money market
funds in the early 1970s, and the sharp increase
in stock and bond prices in recent years. Ex-

ECONOMIC PERSPECTIVES

in the pool and are generally
valued at the day-end net asset
Commercial banks’ share of assets adjusted
price of the asset portfolio. The
for discretionary trust assets
fund stands ready to buy and sell
percent
shares continuously at this price.
The sponsor investment company
may manage the fund by provid­
ing investment advice, provide
the necessary back-room opera­
tions including recordkeeping,
custodial, and transfer services,
and/or market and sell the shares
directly to the public, or it may
hire one or more third parties to
do so. Thus, mutual funds consist
of a sponsor, investment manager,
share distributor, and operations
agent. These four functions may
Sources: Board of Governors of the Federal Reserve System,
be conducted by a single entity,
"Flow of funds accounts,’ various years; and Federal Financial Institutions
Examination Council, Trust Assets of Financial Institutions, various years.
four different entities, or some­
thing in between.
Commercial banks are tradi­
tionally portfolio investors that raise funds by
cept for money market funds, mutual funds are
selling primarily debt instruments (deposits) to
valued at market prices. In contrast, the assets
second parties. Thus, unlike the case with
of depository institutions, insurance compa­
nies, and finance companies are typically mea­
mutual funds, most bank investors are creditors
sured by book value.
rather than owners, whose returns are fixed.
But many bank customers also wish to invest
Mutual funds are open-ended investment
in securities offering greater risks and hopeful­
funds sponsored (organized) by an entity called
ly higher returns than can be obtained on bank
an investment company that sells shares to
deposits. This has been especially true in re­
raise a third-party pool of funds for investment
cent years as households have become wealthi­
in securities. The shares represent an interest
er and older and have placed
increasing emphasis on saving for
FIGURE 4
retirement through pension plans.
As indicated above, banks have
The rise in mutual funds’ share of assets
of all financial institutions
long provided some of these ser­
vices through their trust depart­
percent
ments. It has been common
practice for trust departments to
commingle trust accounts for
investment purposes in order to
reduce transaction costs and real­
ize operating economies.
In 1965, however, the Comp­
troller of the Currency permitted
the First National Bank of New
York, the predecessor of Citi­
bank, to commingle its managing
agency accounts and to advertise
them to the general public. Cus­
tomers would receive participa­
Source: Board of Governors of the Federal Reserve System,
’ Flow of funds accounts,’ various years.
tion units in the pool. This
change was challenged by the

FEDERAL RESERVE



FIGURE 3

BANK OF CHICAGO

11

securities industry and ultimately struck down
by the Supreme Court, which ruled that it vio­
lated the provisions of the Glass-Steagall Act
separating commercial and investment bank­
ing. The court ruled that commingling manag­
ing agency accounts and selling participation
shares in them was in effect dealing in securi­
ties, which was prohibited. The court conclud­
ed that such a “bank investment fund finds
itself in direct competition with the mutual
fund industry” (Fischer, Gram, Kaufman, and
Mote 1984). The decision temporarily stalled
banks’ efforts to offer a competitive invest­
ment product. However, a 1972 decision by
the Board of Governors of the Federal Reserve
System that explicitly permitted banks to act as
investment managers for mutual funds, while
prohibiting them from brokering such funds,
helped banks to enter this market.
Although the Glass-Steagall Act prohibit­
ed banks from dealing in private securities for
their own account, it did not prohibit them
from purchasing and selling private securities
without recourse upon order of their customers.
While some banks offered brokerage services
as an accommodation to their customers, few
viewed them as a profitable activity. Indeed,
in 1936, the Comptroller of the Currency ex­
plicitly authorized national banks to offer bro­
kerage services, but only as an accommodation
to their customers and not on a profit-making
basis. The increase in securities activities and
the end of fixed commissions on the New York
Stock Exchange in 1975 caused banks to re­
consider their interest in brokerage activities.
In 1981, BankAmerica Corporation announced
its intention to acquire Charles Schwab, the
country’s largest discount broker. Shortly
thereafter, Security Pacific National Bank
initiated a cooperative arrangement with the
Fidelity Group to broker securities, including
mutual funds, to its customers and then orga­
nized its own discount broker as a subsidiary
of the bank. Both activities were undertaken
with the approval of the regulatory agencies.
Thus, banks could broker mutual funds either
directly through the bank or bank holding
company or indirectly through a cooperative
agreement with a third-party broker. Some
banks began to offer their customers “privatelabel” mutual funds managed by others. At the
same time, some banks also started “propri­
etary funds” that were managed by the organiz­

12



ing bank but distributed by others. In 1992,
the Federal Reserve liberalized its regulations
to permit banks and bank holding companies to
broker funds that they also managed. Thus,
banks could effectively engage in all aspects of
mutual fund operations except sponsoring and
distributing (underwriting) the shares directly.
Banks have moved relatively slowly into
the mutual fund business and were not overly
aggressive in lobbying the regulators to lower
the barriers. Not until the substantial runoff of
time deposits in search of higher yields when
market interest rates declined sharply in the
early 1990s did many banks awaken to the
possibilities of offering money market and
other mutual funds to their customers. Never­
theless, by 1992 more than 90 percent of all
banks offered mutual funds in some way, more
than double the proportion in 1985. Data on
bank-managed and proprietary mutual funds
since 1983 are presented in table 6. As late as
1987, banks managed less than 5 percent of all
mutual fund assets, and by early 1993 this had
increased to only 11 percent. Banks made
much more substantial gains in money market
funds, managing 23 percent of the assets of
such funds in 1993, compared with only 6
percent of stock and bond mutual funds. When
brokered private-label and other funds are
included, banks sold more than one-third of the
dollar volume of all mutual funds in the first
half of 1992, nearly all of which were money
market funds. The ten banking organizations
that managed the largest amounts of mutual
fund assets in 1993 are shown in table 7.
In recent years, some banks have tried to
increase their participation in the mutual fund
industry by acquiring large mutual fund invest­
ment companies or entering into exclusive
joint agreements with them. In 1993, for ex­
ample, Mellon Bank, the twelfth largest bank
in the country, announced its intention to pur­
chase the Dreyfus Funds, the third largest
sponsor of money market funds and tenth larg­
est sponsor of other mutual funds. At the same
time, NationsBank entered into a partnership
that gave Dean Witter Financial exclusive
rights to market proprietary NationsBank funds
as well as other funds to bank customers from
locations in the bank’s offices. On the other
hand, Chemical Bank and Liberty Financial
broke off their attempted joint venture.

ECONOMIC PERSPECTIVES

TABLE 6

Bank-managed mutual funds: dollar amount, number of funds, and percent of industry
(1983-93)
Money market
Assets

Other funds

Number

Assets

$■
1993
1992
1991
1990
1989
1988
1987
1986
1985
1984
1983

%b

#

%b

$a

134

23.1
19.4
16.9
13.1
11.5
11.3
10.3

461
382
316
256
191
154
109
80
56
48
42

39.9
36.0
32.8
33.1
28.8
26.2

85
47
27
13

4.6
3.4
2.4

10

1 .8

111

95
67
50
38
31
28
19
17
14

1 0 .1

8.3
7.8
8 .2

2 1 .8

19.0
15.1
15.4
15.0

6

4
4

%b
6 .0

1.3
0.9
0.9

2

0 .8

1

0.7

1

0 .8

Total

Number

#
954
502
359
271
213
166
104
65
52
39
24

%b
2 0 .2

14.7
12.7
11.4
9.5
8 .0
6 .2

4.9
4.9
4.7
3.6

Assets

$«
219
158
122

80
60
44
35
32
21

18
15

Number

%b
1 1 .0

9.9
10.3
7.9
7.0
5.4
4.6
4.5
4.3
5.0
5.2

#

%b

1,415
884
675
527
404
320
213
145
108
87

24.2
19.9
17.9
16.7
13.9

66

1 2 .0

9.8
8.3
7.6
7.6
7.0

aBillion dollars.
bPercent of industry.
Source: C ourtesy of U p p e r A nalytical Services.

each of the eleven years for which data are
available reduces the decline in banks’ share
of assets over the past decade by nearly 2 per­
centage points. Taking account of both trust
department and bank-managed mutual fund
assets would further reduce the downward bias
in asset measures of banks’ share of financial
institutions’ output over the past decade.
Unfortunately, because the trust
asset data include a large but not
TABLE 7
precisely determinable portion of
Banking organizations with largest managed
the assets of mutual funds man­
mutual funds
aged by banks, the two sets of
(1993)
data cannot be simply added.
In addition to trust, securi­
Assets
ties, and mutual fund activities,
M on ey
Bank holding com pany
m arket
O ther
Total
banks also engage in a number of
(-------bilion dollars-------)
other activities either directly or
through nonbank subsidiaries of
PNC (Pittsburgh)
18.0
2.7
20.7
their parent holding companies
NationsBank (Charlotte)
8.3
13.7
5.5
that are reported in the “Flow of
BankAmerica (San Francisco)
1 1 .8
1 .1
12.9
funds accounts” as part of other
Wells Fargo (San Francisco)
2.5
8.1
5.6
financial industries. These sub­
4.1
7.4
Banc One (Columbus)
3.3
sidiaries include consumer and
5.7
Northern Trust (Chicago)
1 .2
6.9
commercial finance companies,
3.4
NBD (Detroit)
2.3
5.7
mortgage companies, and savings
State Street (Boston)
1 .8
3.3
5.1
associations. For example, in
Chase Manhattan (New York)
2 .1
3.0
5.0
September 1993, bank holding
4.7
Norwest (Minneapolis)
3.9
0 .8
companies owned 154 thrift insti­
Source: Y v e tte D. K an tro w , "B an k-m a n a g e d fun ds g re w by 34% in
tutions with assets of $107 bil­
1993," American Banker, February 9, 1994, p. 14.
lion. Citicorp operated Citibank

Although the flow of funds data incorpo­
rate the assets of mutual funds managed by
banks, they do not attribute those assets to the
commercial banking sector. Rather, assets of
all mutual funds, regardless of their managers,
are listed under a separate mutual funds sector.
Adding the data on bank-managed mutual
funds from table 6 to banks’ total assets for

FEDERAL RESERVE BANK OF CHICAGO




13

TABLE 8

Nonbank assets held by large bank
holding companies
(1992)
Billion
dollars

Percent

Securities brokerage
and underwriting

77

36

Thrift institutions 3

34

16

Mortgage banking

19

9

Commercial finance

16

8

Consumer finance

12

6

6

3

Small business
investment companies

4

2

Data processing

2

1

A ctivity

Leasing

Insurance underwriting
and insurance agency
Other nonbank
T o ta lb

2

1

41

19

212

100

E x c lu d e s in stitutio ns supervised by th e Federal
D eposit Insurance C o rp o ra tio n , such as statechartered savings banks
bC olu m n s m ay not total because of roun din g.
Source: Board o f G overnors of th e Federal
R eserve System .

Savings, which is the eighth largest savings
association in the country. Similarly, at yearend 1992, twelve of the fifty largest finance
companies were owned by bank holding com­
panies. These include CIT, the
ninth largest company, which is
owned by Dai-Ichi Kangyo Bank
(Japan). The total assets of non­
banking subsidiaries owned by
reporting large bank holding com­
panies (which are estimated to be
roughly 95 percent of those of all
bank holding companies) were
$212 billion in 1992.6 As table 8
shows, over one-third of these are
in securities brokerage and under­
writing subsidiaries. Despite their
absolute importance, the nonbank
assets of bank holding companies
are dwarfed by the reported assets
of banks. If the total assets of the
nonbank subsidiaries of bank hold­
ing companies are added to report­
ed bank assets, the market share of
banks in 1992 increases only from

14




25.8 percent to 27.5 percent. But even after
one adjusts reported assets for assets either
owned by subsidiaries of bank holding compa­
nies or managed by banks but reported under
other institutions in the “Flow of funds” data,
there is still a growing volume of bank activi­
ties which are unrelated to either owning or
managing assets but which generate income for
banks, for example, lines of credit, letters of
credit, and futures, options, and swaps.
Noninterest income
Although the growing importance of offbalance-sheet activities is not captured by
traditional asset measures, it does show up in
the growth of noninterest or fee income. While
fee income has received much attention over
the past two decades, it is not of recent origin
or importance, as figure 5 indicates. When
loan demand collapsed and interest rates fell to
extremely low levels in the 1930s, commercial
banks’ ratio of fee income to interest income
increased sharply. However, because of the
steady rise in interest rates in the post-World
War II era, the growing importance of fee
income was obscured until the early 1980s.7
The trend towards an increase in fee in­
come relative to interest income is present not
only in the United States but in nearly all de­
veloped countries. The percentages of gross
bank income derived from fees in fifteen major
countries for selected years from 1980 to 1990
are shown in table 9. (Note that these data are

ECONOMIC PERSPECTIVES

TABLE 9

Fee income as a percent of gross income
of banks, 15 major countries8
(1980-90)
Countries

United Statesb
Japan 60
Germany6
France6
Italy
United Kingdom 6
Canada0
Australia 0
Belgium 0
Finland
Netherlands
Norway
Spain 6
Sweden
Switzerland

1980-82

1984-86

1990

31.4
24.6
28.6
15.3
30.3
36.9
23.7
33.5
23.4
58.3
24.7
35.2
18.1
33.5
47.5

38.0
35.9
34.9
24.9
26.8
41.1
31.0
34.0
23.0
46.9
29.7
25.9
22.3
26.2
49.1

30.0
20.4
30.6
14.6
26.0
28.5
2 1 .6 d
32.1
19.6
48.8
25.0
27.3
15.7
29.8
46.6

aS hare o f n o n in tere st in co m e in th e gross incom e
o f co m m e rc ia l banks; th e data are not fu lly
c o m p a ra b le across countries.
bLarge co m m e rc ia l banks.
°Fiscal years.
d1982.
Source: Bank fo r In tern a tio n a l S e ttle m e n ts , Annual
Report, 1992, p. 196.

not fully comparable with the data for U.S.
banks described above nor across countries.)
In all countries except Finland, the importance
of fee income increased during this period.
Although fee income is relatively more impor­
tant in the United States than in most other
countries, it is considerably less important than
in Switzerland or Finland and somewhat less
important than in the United Kingdom.
The unbundling and securitization o f
financial services
The rise in fee income is in part a conse­
quence of another phenomenon. The 1970s
witnessed an acceleration of a trend that had
been evident for some time, namely the “un­
bundling” of financial services. Unbundling is
the separation of complex banking services,
including such fundamental and traditional
banking services as real estate and commercial
lending, into their component steps or func­
tions and the performance of some of those
functions by separate entities. The oldest and
most obvious example of unbundling was

FEDERAL RESERVE BANK OF CHICAGO




separating the origination and servicing of
residential mortgage loans from the portfolio
investment function through the sale of the
mortgage from the originator to an institutional
investor. Pioneered by mortgage companies
decades before, this practice has since been
adopted by banks and other mortgage lenders.
A major development in this unbundling
was the introduction of the mortgage-backed
security by the Federal Home Loan Mortgage
Association and the Government National
Mortgage Association. This was also the first
step in the now familiar process of “securitiza­
tion,” the issuance of securities whose princi­
pal and interest payments reflect the behavior
of a pool of underlying assets. The 1980s saw
an enormous enlargement of the scope of secu­
ritization, which now encompasses automobile
loans, credit card receivables, and other con­
sumer credit, and is even making inroads into
commercial loans, a type of asset that is much
more difficult to securitize because of the
greater heterogeneity of loan agreements and
covenants. The banks receive fees for origina­
tion and possibly servicing but frequently do
not hold the asset in their portfolios and thus
do not receive interest revenue from it.
Sanford Rose, a former associate editor of
the American Banker, argued vigorously in the
early 1980s that costly regulation, inadequate
compensation for lending risks, and the futility
of trying to outguess the market regarding
increasingly volatile interest rate movements
were bringing about a fundamental transforma­
tion of the banking environment (Rose 1981).
He asserted that the most prudent strategy for
banks was to reduce their emphasis on portfo­
lio investment, hedge or sell off their interest
rate risk, and rely on origination and servicing
fees to provide the bulk of their earnings. In­
deed, he argued that mortgage companies,
which have long operated in this manner, were
the model for the financial firm of the future.
In the years since this analysis appeared, com­
mercial banking organizations have come more
and more to resemble Rose’s vision: They
originate a large volume of loans—although
even here they have lost ground to other insti­
tutions—and sell off a growing proportion of
them. They also use their financial expertise,
reputation, and capital to provide guarantees of
financial performance, mostly in the form of
standby letters of credit, but increasingly en-

15

compassing a growing variety of
new and exotic instruments.

FIGURE 6

Share of commercial bank financial intermediation
The Boyd-Gertler approach
percent
Two somewhat different ap­
proaches to the adjustment of bank
assets for off-balance-sheet activi­
ties were recently presented by
Boyd and Gertler (1994). Both
involved developing estimates of
the asset equivalents of bank offbalance-sheet and fee-for-service
activities. The first approach ad­
justed for loan commitments and
letters of credit, two of the most
important types of off-balancesheet guarantees offered by banks,
using the risk weights developed in
Note: Both adjusted series include undercounted oft-shore loans, as well as
off-balance-sheet credit equivalents.
the Basel risk-weighted capital
Source: John H. Boyd and Mark Gertler, “Are banks dead? Or are the
reports greatly exaggerated?* unpublished paper, Federal Reserve
standards. These weights were
Bank of Minneapolis, March 1994.
used to calculate the level of assets
that would represent the same risk
exposure to the bank as the off-balance-sheet
Boyd and Gertler’s two alternative methods, is
activities. These asset equivalents were then
shown in figure 6.
added to each institution’s on-balance-sheet
Summary o f asset measures
assets to obtain a more complete asset measure
This section has described a number of
of banks’ market share. The shortcoming of
approaches to adjusting data on bank assets to
this approach is that it takes account only of
take account either of assets that are managed
loan commitments and letters of credit and
by the bank but do not show up on its account­
omits such important activities as trust services
ing balance sheet or of activities that are done
and mutual funds.
for a fee and are not associated with assets
Boyd and Gertler’s second procedure was
either owned or managed by the bank. One of
to convert all noninterest income— from loan
the problems with trying to “fix” banks’ bal­
servicing, asset management, and other servic­
ance sheets is that the problems associated with
es (including trust and securities activities), as
them are not limited to banking. For example,
well as off-balance-sheet guarantees— into a
life insurance companies also engage in a large
balance sheet equivalent. Using net interest
volume of off-balance-sheet activities. Thus,
income (interest income less interest expense
to obtain a meaningful measure of banks’ rela­
and loan losses) as a measure of the return
tive importance in the financial system, it
from on-balance-sheet assets, and assuming
would be necessary to perform similar adjust­
that the same rate of return is earned in offments on the balance sheets of other financial
balance-sheet activities, the authors capitalized
industries. Together with the conceptual short­
fee income at that rate to generate “imaginary”
comings of assets as a measure of output—that
asset equivalents. They then added the asset
is, it is a stock rather than a flow measure, and
equivalents of the noninterest income to ondifferent levels of output may be associated
balance-sheet assets to obtain a more compre­
with the same value of assets of different
hensive measure of bank output for the years
kinds—this suggests the desirability of also
since 1971. When they did so, virtually all
looking at alternative measures of the relative
evidence of a downward trend in banks’ share
size of the banking industry.
of financial institutions’ assets over the period
1957-1990 disappeared, although there was
E m p lo y m e n t
some decline from the 1974 peak. Commercial
A second measure of banking’s size or
banks’ share of the assets of all financial insti­
importance is the number of employees in the
tutions, both unadjusted and as adjusted by
industry. For some purposes, employment may

16




ECONOMIC PERSPECTIVES

be the most relevant and useful
FIGURE 7
measure. This is most obviously
Commercial bank employees as a percent of total
true in regard to the industry’s
for financial, insurance, and real estate sector
impact on the economy of a partic­
percent
ular city or region. However, be­
cause employment is a measure of
input rather than output, it is much
less appropriate as a measure of the
size or competitiveness of an in­
dustry relative to other industries
producing similar products or ser­
vices. Moreover, employment does
not adjust for differences in pro­
ductivity between sectors of the
economy or changes in productivi­
ty over time. Nevertheless, it may
serve as a useful check on the accu­
racy of other measures.
Sources: U.S. Department of Labor, Bureau of Labor Statistics,
As figure 7 indicates, between
E m p lo y m e n t, H o u r s , a n d E a r n in g s , U n it e d S ta te s , 1 9 0 9 -9 0 , Bulletin 2370,
Vol. 2, March 1991 and 1 9 8 1 -9 3 , Bulletin 2429, August 1993; and Federal
1934 and 1977, employment in the
Deposit Insurance Corporation, H is t o r ic a l S ta t is t ic s o n B a n k in g , 1992.
commercial banking industry more
than kept pace with that in the
entire financial, insurance, and real estate sec­
that productivity in banking has declined over
tor. Thereafter it declined by roughly oneany extended period in recent years.
fourth through 1992. As a percentage of total
R evenu es, e a rn in g s, and v alu e a dded
employment in the private nonfarm economy,
A third set of measures of the importance
employment in commercial banking continued
of banking is based on revenue and earnings
to rise through 1983, when it peaked at 1.67
data that reflect the full range of services of­
percent. Since then, that number has fallen as
fered. Such measures have the advantage of
well. The absolute level of employment in the
being flow rather than stock measures of out­
industry continued to rise through 1986, peak­
put. Indeed, in most industries, market share is
ing at 1.56 million. By 1992, it had fallen to
typically measured by revenues, sales, or value
1.48 million.
added rather than assets. These measures are
The decline in employment in the banking
also used by the Department of Justice in anti­
industry in recent years is not surprising given
trust actions. Revenue and value added mea­
the large number of bank closings and consoli­
sures are available for banks and other deposi­
dations in the 1980s and the acceleration of
tory institutions from data reported to the bank
consolidation in the early 1990s. However, the
regulatory agencies in their periodic Reports o f
rise over the preceding decades suggests two
Income and Dividends or from data reported on
possibilities: either commercial bank produc­
a regular basis to the Internal Revenue Service
tivity was falling continuously over that peri­
(IRS). Virtually since the IRS was established
od, as the declining ratio of bank assets to
in 1916, it has published annual compilations
employment would suggest,8 or total assets is
of income and expenses of corporations and
an inadequate measure of financial institution
individuals. For the earliest years, these re­
output. Although a decline in the productivity
ports were based on the universe of federal
of banking extending over five decades cannot
income tax returns; more recently, they have
be dismissed as a logical possibility, it seems
been based on a sample. The advantage of the
inconsistent with the increased use of comput­
IRS data is that they can be obtained on a rela­
ers and other advanced technology by banks
tively uniform basis for all categories of finan­
and with the continued rise of productivity in
cial institutions.
the economy as a whole. Moreover, the great
A measure of the size of the banking in­
expansion of off-balance-sheet activities in
dustry based on IRS data that takes account of
banking described in detail in the preceding
both lending and off-balance-sheet activities is
sections casts further doubt on the hypothesis


FEDERAL RESERVE


BANK OF CHICAGO

17

cause of changes in definitions
and reporting categories and the
Banks’ share of total receipts of the financial sector
amount of detailed information
percent
published by the IRS, the measure
is available only from 1938 on.
As figure 9 shows, this mea­
sure of commercial banks’ share
of the total output of the financial
sector gives a considerably differ­
ent picture than reported asset
measures. Rather than declining
monotonically over the entire
period like the asset measure, it
averages around 25 percent in the
late 1930s, rises to the low 30
percent range in the 1940s and
early 1950s, declines to just over
Note: A break in the series for the financial sector occurred between 1957
15 percent in the 1960s, rises
and 1958 because of a redefinition of receipts of life insurance companies
to include premiums and other income as well as investment income.
above 20 percent in the midSource: U S. Treasury Department, Internal Revenue Service,
Statistics of Income, Part 2, Corporation Income Tax Returns, 1938-89.
1970s and again in the early
1980s, and declines to about 16 or
17 percent by the late 1980s.
There was clearly a decline in the banking
simply total receipts or revenues. Figure 8
industry’s share of the output of all financial
shows the ratio of total receipts for banks to
institutions through the mid-1960s, although
those for the entire financial sector, including
the greater part of the apparent sharp decline
insurance, for the years 1938-82. Unfortunate­
between 1957 and 1958 was spurious, reflect­
ly, this measure is strongly influenced by
ing a change in the reporting of revenues of life
movements in the general level of interest
insurance companies. However, there has been
rates, and its volatility tends to obscure the
no obvious trend since then.9
basic trend in the data. Moreover, as was sug­
Relative to the entire economy, the output
gested in the earlier discussion of the conceptu­
of both banks and the entire financial sector
al problems in measuring the importance of
has increased over the past half-century. As
banking in the financial services industry, there
is much to be said for using a mea­
sure of value added—the value of
FIGURE 9
the products sold by an industry
Banks’ share of value added of the financial sector
less the value of intermediate goods
and raw materials purchased by it.
percent
The IRS data permit the calcu­
lation of commercial banks’ share
of a variable that closely approxi­
mates a measure of value added for
total financial institutions proposed
by Donald Hodgman (see box).
This is the difference between total
receipts, including interest re­
ceived, and interest paid. The
netting of interest received and
paid greatly reduces but does not
eliminate the enormous variation in
bank revenues stemming from
changes in the level of market
Source: U.S. Treasury Department.
interest rates. Unfortunately, be­


18


FIGURE 8

ECONOMIC PERSPECTIVES

environments, such as the United
Kingdom. However, when one
Value added of banks and the financial sector as
analyzes other measures of the
a percent of gross domestic product*
size of the banking industry, such
percent
as employment, revenues, and
value added, the same conclusion
does not always emerge. In part,
this is because the nature of bank
activities has changed drastically
over the past several decades and
many of the newer activities are
not reflected in balance sheet
assets. When asset figures are
adjusted to incorporate some
measure of the new activities,
they show either no decline or a
much attenuated rate of decline
for banking in recent years.
Note: A break in the series for the financial sector occurred between 1957
In summary, the evidence
and 1958 because of a redefinition of receipts of life insurance companies
to include premiums and other income as well as investment income.
does not clearly support the wide­
*1938-1958 is gross national product. 1959-1989 is gross domestic product.
Source: U.S. Treasury Department.
spread perception that banking
has declined, either absolutely or
relative to the financial services
figure 10 shows, the value added of the finan­
industry or the entire economy, since the early
cial sector as a proportion of GDP rose from
1960s. Nonetheless, this conclusion is consis­
7.5 percent in 1938 to 23.5 percent in 1989,
tent with the belief that banking has not grown
while that for commercial banks increased
as rapidly as it might have if banks had not
from 1.9 percent to 4.0 percent.
been constrained from providing new products
quickly in response to changes in market con­
C o n c lu s io n
ditions. Unfortunately, we do not know how
Is banking a declining industry? There is
rapidly banking would have grown under alter­
a widespread perception that the size of the
native regulatory regimes, or what the social
commercial banking industry relative to that of
costs and benefits of those alternatives would
all financial institutions in the United States
have been. Nor do our measures of the relative
has been declining rapidly in recent years.
size of the banking industry shed light on
Restrictive regulations imposed primarily in
whether the regulations, by preventing individ­
earlier years when banking was relatively more
ual banks from expanding or opening branches
important are often blamed as contributing to
across state boundaries, have restricted the
the decline. Most of the evidence for the belief
efficiency of banks and thereby increased the
that banking is declining consists of data on
cost of banking to consumers. Those, howev­
commercial banks’ share of reported assets or
er, are the types of questions that need to be
of specific categories of assets, such as com­
answered in order to improve public policy
mercial loans. Similar results obtain for banks
towards banking.
in countries with greatly different regulatory
FIGURE 10

FOOTNOTES
'It has been argued that a major decline in the size o f the
banking industry, regardless o f its cause, would create
problems for the implementation and effectiveness o f
monetary policy. This article does not attempt to address
this issue.
2The National Bureau o f Economic Research has spon­
sored a number o f conferences on this and related issues
(National Bureau o f Economic Research 1961, 1969).


FEDERAL RESERVE


RANK OF CHICAGO

3Board o f Governors o f the Federal Reserve System,
“Flow o f funds accounts.”
4A recent book by Robert E. Litan (1987) also presents
data for 1835 (figure 2.2, p. 18). These data were ob­
tained from a Census publication (U.S. Bureau o f the
Census 1975).

19

5The major difference between an agent and a trustee is
that in an agency relationship the principal (customer)
retains legal title to the assets, whereas in a trust relation­
ship, legal title passes to the fiduciary. In addition, a trust
relationship will involve more duties and responsibilities
on the part o f the fiduciary even in the absence o f specific
written authority and, unlike an agency relationship,
which terminates on the death o f the principal, may
continue beyond the death o f the grantor o f the trust.
6The data on nonbank subsidiaries o f bank holding com ­
panies reported here were obtained from the F R -Y 11Q
and FR-Y 1 IA S reports, which are filed with the Federal
Reserve by all bank holding companies with consolidated
assets o f more than $1 billion and by those with assets o f
more than $150 million that have nonbank activities
exceeding specified levels. These figures are larger than
reported in table 8 because they include FDIC-supervised
savings banks that are excluded from the data used to
construct the table.
7In the early 1960s, many o f the larger banks sought to
increase the variety and volume o f services that they
offered on a fee-for-service basis. A series o f favorable
rulings by then Comptroller o f the Currency James Saxon
encouraged national banks in their efforts to expand their

activities. The services that they, or subsidiaries o f their
bank holding companies, began to offer included such
relatively minor extensions o f existing activities as pro­
viding investment advice, payroll accounting, data pro­
cessing, armored car and courier services, and insurance
agency services. Because the courts eventually disal­
lowed many o f these activities as violating the National
Banking Act, the Banking Act o f 1933 (Glass-Steagall
Act), or the Bank Holding Company Act, the activities did
not contribute greatly to banks’ fee income.
8Bert Ely o f Ely Associates, Inc., a financial institutions
consulting firm, has argued that, largely as a consequence
o f regulatory constraints, the efficiency o f the entire
financial system has declined and that many widely
heralded “innovations” in finance represent nothing more
than “regulatory arbitrage” (Ely 1992).
9This result accords with that o f Boyd and Gertler (1994),
who also presented data on value added. They found a
slight upward trend in the share o f value added o f bankrelated industries relative to that for the entire finance,
insurance, and real estate sector over the period 1947 to
1990. However, their “bank-related industries” category
contained all depository institutions.

REFERENCES

Bank for International Settlements, Annual
Report, 1992.

____________ , “Flow of funds accounts,”
Washington: Board of Governors, 1952-93.

Barth, James R., R. Dan Brumbaugh, Jr.,
and Robert E. Litan, The Future o f American
Banking, Armonk, NY: M. E. Sharpe, Inc.,
1992.

Boyd, John H., and Mark Gertler, “Are
banks dead? Or are the reports greatly exag­
gerated?” unpublished paper, Federal Reserve
Bank of Minneapolis, March 1994.

Bell, Frederick W., and Neil B. Murphy,
Costs in Commercial Banking: A Quantitative
Analysis o f Bank Behavior and its Relation to
Bank Regulation, Federal Reserve Bank of
Boston, Research Report No. 41, April 1968.

Ely, Bert, “Commercial banks are not obsolete
and the federal government should stop trying
to make them so,” Credit Markets in Transi­
tion, Proceedings of a Conference on Bank
Structure and Competition, Chicago: Federal
Reserve Bank of Chicago, 1992, pp. 356-90.

Benston, George J., “Economies of scale
and marginal costs in banking operations,”
National Banking Review, Vol. 2, June 1965,
pp. 507-49.
Benston, George J., Gerald A. Hanweck,
and David B. Humphrey, “Scale economies
in banking: a restructuring and reassessment,”
Journal o f Money, Credit, and Banking, Vol.
14, November 1982, pp. 435-56.
Board of Governors of the Federal Reserve
System, Balance Sheets fo r the U.S. Economy,
1945-92, March 10, 1993.

20



Federal Deposit Insurance Corporation,
Historical Statistics on Banking, 1992.
Federal Financial Institutions Examination
Council, Trust Assets o f Financial Institutions,
various years.
Federal Reserve Bank of Chicago, FDICIA:
An Appraisal, Chicago: Federal Reserve Bank
of Chicago, 1993.
Fischer, Thomas G., William H. Gram,
George G. Kaufman, and Larry R. Mote,
“The securities activities of commercial

ECONOMIC PERSPECTIVES

banks,” Tennessee Law Review, Vol. 51,
Spring 1984, pp. 467-518.
Goldsmith, Raymond W., Financial Interme­
diaries in the American Economy since 1900, a
study by the National Bureau of Economic
Research, Princeton, NJ: Princeton University
Press, 1958.
____________ , Financial Structure and Devel­
opment, Studies in Comparative Economics,
New Haven, CT: Yale University Press, 1969.
Greenbaum, Stuart I., “A study of bank
costs,” National Banking Review, Vol. 4, June
1967, pp. 415-34.
Gurley, John G., and Edward S. Shaw, “Fi­
nancial aspects of economic development,”
American Economic Review, Vol. 45, Septem­
ber 1955, pp. 515-38.
____________ , “Financial intermediaries and
the saving-investment process,” Journal o f
Finance, Vol. 11, May 1956, pp. 257-76.

Litan, Robert E., What Should Banks Do?
Washington, DC: The Brookings Institution,
1987.
National Bureau of Economic Research,
Output, Input and Productivity Measurement,
Studies in Income and Wealth, John W. Ken­
drick (ed.), Vol. 25, Princeton, NJ: Princeton
University Press, 1961.
____________ , Production and Productivity in
the Service Industries, Studies in Income and
Wealth, Victor R. Fuchs (ed.), Vol. 34, New
York: Columbia University Press, 1969.
Powers, John A., “Branch versus unit bank­
ing: bank output and cost economies,” South­
ern Economic Journal, Vol. 36, October 1969,
pp. 153-64.
Robertson, Ross M., The Comptroller and
Bank Supervision, Washington, DC: Office of
the Comptroller of the Currency, 1968.

____________ , Money in a Theory o f Finance,
Washington, DC: Brookings Institution, 1960.

Rose, Harold, “The changing world of finance
and its problems,” working paper no. 167-93,
Institute of Finance and Accounting, London
Business School, 1993.

Hodgman, Donald R., “Discussion,” Produc­
tion and Productivity in the Service Industries,
National Bureau of Economic Research Studies
in Income and Wealth, Vol. 34, New York:
Columbia University Press, 1969, pp. 189-95.

Rose, Sanford, “‘De-intermediation’: a word
for the ’80s,” The Future o f the Financial
Services Industry, Proceedings of a conference,
Atlanta: Federal Reserve Bank of Atlanta,
1981, pp. 157-64.

Humphrey, David B., “Flow versus stock
indicators of banking output: effects on pro­
ductivity and scale economy measurement,”
Journal o f Financial Services Research, Vol.
6, August 1992, pp. 115-35.

U.S. Bureau of the Census, Historical Statis­
tics o f the United States, Colonial Times to
1970, Part 2, Washington, DC: U.S. Govern­
ment Printing Office, 1975.

Kantrow, Yvette D., “Bank-managed funds
grew by 34% in 1993,” American Banker,
February 9, 1994, pp. 1, 14-16.
Kaufman, George G., and Larry R. Mote,
“Glass-Steagall: repeal by regulatory and judi­
cial reinterpretation,” Banking Law Journal,
Vol. 107, September/October 1990, pp.
388-421.

FEDERAL RESERVE BANK OF CHICAGO




U.S. Department of Labor, Bureau of Labor
Statistics, Employment, Hours, and Earnings,
United States, 1909-90, Bulletin 2370, Vol. 2,
March 1991, and 1981-93, Bulletin 2429, Au­
gust 1993.
U.S. Treasury Department, Internal Reve­
nue Service, Statistics o f Income, Part 2,
Corporation Income Tax Returns, annual,
1938-89.

21

Funding small businesses
through the SBIC program

Elijah B re w e r III and Hesna G enay

“Ensuring the availability of
sufficient amounts of credit
for small- and medium-sized
businesses, at affordable inter­
est rates, is vital in any effort
to bring this nation out of recession, create new
jobs, and build a strong U.S. economy.” 1
These remarks reflect the growing concern
over the availability of funding to small busi­
nesses. Because small businesses are per­
ceived to be a major source of growth for the
U.S. economy, a number of policy initiatives
have been proposed recently in the Congress to
increase the availability of funds to these firms.
The debate about the availability of capital
to small business is not new. In 1958, the
Federal Reserve Board concluded that there
was a shortage of funds available to these
firms.2 In response, Congress authorized the
Small Business Administration (SBA) to char­
ter private small business investment compa­
nies (SBICs) to act as financial intermediaries
for small firms.
SBICs differ from other financial institu­
tions that fund small businesses. Traditional
financial intermediaries such as banks provide
short-term working capital financing to small
firms, while SBICs provide long-term funds,
not only through loans, but also through equity
investments.3 Furthermore, banking organiza­
tions are allowed to participate in the program;
hence, while banks are restricted from making
direct equity investments, they can do so indi­
rectly by establishing SBIC units. SBICs are
also unique in that they have access to govern­
ment subsidies and thus can leverage their

22




private capital with government funds, unlike
other venture capital firms.
These and other features raise a number of
interesting issues about the role of SBICs in
funding small businesses. In perfect capital
markets, firms can always raise funds for posi­
tive net present value projects. Capital market
imperfections that are caused by conflicts of
interest between outside investors and manag­
ers of firms and differences in the amount of
information available to them, however, can
impose costs on firms and inhibit the flow of
funds for profitable investment projects. It has
been argued that the characteristics of small
businesses exacerbate these problems.
A central issue in financing small firms is
the conflict between the types of investors and
financing that are most appropriate for these
firms. On one hand, because it is hard to eval­
uate and monitor small firms, and because they
Elijah Brewer III is associate professor of finance
at the University of Illinois at Urbana-Champaign
and a senior economist at the Federal Reserve
Bank of Chicago. Hesna Genay is an economist at
the Federal Reserve Bank of Chicago. We would
like to thank Herbert Baer fo r bringing to our
attention the distinctive features of the SBIC pro­
gram. We also thank the U.S. Small Business
Administration, in particular Gerry Dillon, Ned
Shepperson, and John W ilmeth, for providing us
with the data used in this article. We benefited
greatly from the comments of Douglas Evanoff,
Steven Strongin, and Paula W orthington.
Rosemary Berger, Veronica Woods, and Michael
York provided valuable research assistance.
The views expressed here are those of the
authors and do not necessarily reflect the views of
the Federal Reserve System or the U.S. Small
Business Administration.

ECONOMIC PERSPECTIVES

have few assets that can be collateralized, long­
term debt financing is likely to be costly for
them. On the other hand, equity financing
involves sizable fixed costs, and while banks
may have a comparative advantage in financ­
ing small firms, they are unable to provide
equity capital. Together, these facts restrict
the amount of equity financing available to
small firms. The SBIC program addresses
these issues by increasing the pool of long­
term debt and equity financing and by allowing
banking organizations to provide equity capital
to small firms.
If the SBIC program provides investment
opportunities that minimize the problems asso­
ciated with external finance, the type of financ­
ing provided by an SBIC should vary accord­
ing to the riskiness of the project and the iden­
tity of the SBIC. In particular, we expect
SBICs to provide debt financing primarily for
those activities that generate tangible assets
that can be pledged as collateral. By contrast,
we would expect equity financing to be domi­
nant in funding activities that generate relative­
ly few tangible assets. Moreover, if the SBIC
program affords banking organizations the
opportunity to utilize their comparative advan­
tage in evaluating and monitoring investments,
then we would expect bank-owned SBICs to
provide the majority of capital in the program
and to pursue a strategy of extensive equity
investments.

In this article, we explore these issues
using proprietary data obtained from the SBA.
To determine whether SBICs that are associat­
ed with banking organizations behave differ­
ently than other SBICs, we separated the
SBICs into two groups, bank-owned and nonbank-owned.4 The results indicate that bankowned SBICs do, in fact, pursue a strategy of
extensive equity financing, while non-bankowned SBICs appear to rely more on nonequi­
ty financing and on direct government subsi­
dies. Interestingly, bank-owned SBICs are
more profitable even though they rely far less
on government subsidies in the form of match­
ing funds to invest. This suggests that allow­
ing banks to participate in the SBIC program
provides some advantages over alternative
methods of financing small business, and that
direct government subsidies are not required
to enable investments in small businesses to
be profitable.

FEDERAL RESERVE



BANK OF CHICAGO

The article is organized in four sections.
The first section discusses the economic impli­
cations of the SBIC program. The second
section examines the types of investments
SBICs make, the cross-sectional differences in
characteristics and investment strategies, and
the differences between bank-owned and other
SBICs. The third section presents evidence
concerning the impact of SBICs’ asset-mix
decisions on profitability. The final section
contains concluding remarks.
T h e e c o n o m ic s o f fin a n c in g sm a ll
b u sin e sse s and th e SB IC p ro g ra m

One of the central questions in the debate
over small business financing is how the char­
acteristics of these firms affect their funding.
It is often argued that in small firms, the infor­
mation gap between outside investors and
managers of firms is greater and the conflicts
of interest among different stakeholders are
more severe.
Small businesses tend to be newer, private
companies without established public track
records.5 Moreover, most small firms are in
trade and service industries, which tend to have
high ratios of intangible assets that cannot be
pledged as collateral for loans. Small busi­
nesses also tend to have high failure rates and
are concentrated in highly volatile industries.6
Although the probability of failure is higher for
small firms, comparisons of small and large
surviving firms indicate that small firms grow
faster. In other words, while young firms are
likely to have very little cash flow in the short
run, their future growth opportunities tend to
be high.
These features of small firms tend to exac­
erbate the problems associated with capital
market imperfections that raise the cost of
external financing and inhibit the flow of funds
to them. Evidence suggests that collateral,
restrictive covenants, mixed equity and debt
financing, and long-term relationships with
investors mitigate some of these problems.7
But such solutions involve fixed costs that are
burdensome to firms that need to raise only
small amounts of funds. The usual response
to these problems has been either to provide
government subsidies to defray the fixed costs,
or to relax regulations on financial institutions
to encourage the flow of funds to small busi­
nesses.8 The SBIC program offers both of
these features.

23

Under the program, a company may be
chartered to operate as an SBIC if it satisfies
minimum private capital requirements. SBICs
provide equity capital or long-term loans to
firms having net worth less than $6 million or
average net income less than $2 million in the
preceding two years.9 In addition, SBICs may
receive government-guaranteed funds through
issuances of debentures and other obligations
which can be purchased directly or guaranteed
by the SBA. At present, SBICs must have a
minimum of $2.5 million in private capital and
may receive up to $3 in SBA funds for every
$1 of private capital.
SBICs are also subject to restrictions on
the types and forms of their investments, sum­
marized in box 1. Because the SBIC program

was designed to encourage the flow of long­
term capital to small firms, the regulations
specify a minimum maturity for loans and a
maximum rate of interest that can be charged.
Although regulations allow SBICs to invest in
the equity of small businesses, they are not
permitted to gain control of a small business
without prior SBA approval or a plan of dives­
titure. SBICs may invest only in qualifying
small businesses, or, if an SBIC has temporari­
ly idle funds, in certain short-term investments.
In addition to providing subsidized funds
through the SBA, the SBIC program allows
banking organizations to provide equity financ­
ing to small firms. If, as has been argued,
banks have a comparative advantage in evalu­
ating and monitoring small firms, then bank

BOX 1

C urrent SBIC regulations: a summ ary
S o u rc e s o f S B IC fu n d s

■ Minimum private capital requirement is $2.5
million in capital and paid-in surplus.
■ SBICs can obtain up to $3 in SBA funds for
every $1 of private capital.
■ SBA funds can be obtained either through sales
of debentures to the SBA or through issues of
SBA-guaranteed debentures. The majority of the
outstanding SBA-guaranteed debentures issued
by SBICs are ten-year debentures. Currently, the
SBA is restructuring the regulations of the SBIC
program. Once the restructuring is completed,
SBICs will also be able to obtain SBA funds
through issues of preferred securities. In addi­
tion, the maximum amount of SBA funds that
any one SBIC can obtain is to be raised from $35
million to $90 million.
■ The interest rate on SBA-guaranteed debentures
is the interest rate on Treasury securities of
comparable maturity. In addition, the SBA
charges a premium averaging 60 to 100 basis
points over the interest rate of comparable Trea­
sury securities.
Uses o f SBIC fu n d s

■ SBICs may invest only in qualifying small busi­
ness concerns or, if the SBIC has temporarily
idle funds, in certain short-term investments.
SBICs may not invest in other SBICs, investment
or finance companies, finance-type leasing com­
panies, unproved real estate, companies with
less than one-half of their investments in the

24




U.S., or companies not engaged in regular and
continuous business.
■ SBICs may not acquire a controlling interest in
a small business unless a plan of divestiture
is filed with the SBA. SBICs may not invest
more than 25 percent of their capital in any one
small business.
■ The minimum maturity of SBIC loans is 5 years.
The maximum interest rate that can be charged
on these loans (the “maximum cost of money”) is
determined by the SBA. If the current rate on
ten-year debentures sold by the SBA is less than
8 percent, then the maximum cost of money is 15
percent on loans and 14 percent on debt securi­
ties. If the debenture rate is more than 8 percent,
then the maximum cost of money is the deben­
ture rate plus 800 basis points on loans, or the
debenture rate plus 700 basis points on debt
securities.
O ve rsig h t

■ Each SBIC must be audited by an independent
accredited auditor to determine whether the
SBIC’s financial statements conform to generally
accepted accounting rules and to SBA regula­
tions. In addition, SBICs are subject to annual
SBA examinations.
Note: The information in this table is not exhaustive but
only highlights the principal regulations o f the SBIC
program. The formal text o f the full SBIC regulations is
given in section 13 CFR 107 o f the SBA regulations.

ECONOMIC PERSPECTIVES

participation in such programs as the SBIC
program should increase the amount of funds
available to small firms.1 Until 1976, banks
0
were prohibited from owning more than 50
percent of any one SBIC, and no bank could
invest more than 2 percent of its capital and
surplus in SBICs. Now, the only constraint on
bank ownership is that no bank or bank hold­
ing company may invest more than 5 percent
of its capital and surplus in SBICs. Further­
more, directors, officers, and employees of a
bank may also serve as officers, directors, or
employees of an SBIC.
The increase in the pool of equity capital
available to small firms should offer several
advantages. Because residual claimants can
share in the potential benefits of the invest­
ments and share the risk with fixed claimants,
the program may lower the cost of capital to
small firms. Additional capital also improves
the balance sheets of these firms, making it
easier for them to obtain funds from other
sources. Moreover, if SBICs are better able
to process information about small firms, then
an investment by an SBIC would signal to
other investors that the firm offers profit
opportunities.
According to SBA statistics, 1,320 compa­
nies became licensed as SBICs between 1959
and 1992.1 At the end of fiscal year 1992,
1
there were 204 active SBICs with $3 billion in
capital resources. Over two-thirds of this capi­
tal was obtained from private sources; the
remainder was supplied by the SBA either
through guarantees of debentures issued by the
SBICs or through purchases of such deben­
tures. The majority of SBA leverage is provid­
ed through guarantees of debentures, which

require direct outlay of SBA funds only in the
event of a default by an SBIC.
S B IC s' fin a n c ia l c h a ra c te ris tic s
and in v e stm e n ts

The SBA has an extensive database on all
SBICs. For this article, we examined its files
on SBICs’ history, reports of condition, and
investments in order to determine whether
SBICs offer different types of financing to
different types of small businesses, and to
examine the relationship between SBICs’ prof­
itability and their financial characteristics.
The reports of condition cover each year from
1986 to 1991, while the investment data cover
each year from 1983 to 1992. The sample
changes each year because many institutions
were liquidated, merged, or voluntarily surren­
dered their licenses.1
2
Table 1 reports some of the developments
in the SBIC program from 1986 to 1991. Dur­
ing this period, the total assets and capital of
SBICs increased by more than 28 percent and
50 percent, respectively. By the end of fiscal
year 1991, the total assets of the companies in
the program were over $4 billion and capital
resources had reached almost $3 billion. As
table 2 shows, these total dollar figures repre­
sent an average of $24.1 million in total assets
and almost $17 million in total capital per firm
in 1991.1
3
The higher growth rate of total capital
relative to total assets indicates that SBICs
leveraged less of their assets in 1991 than in
1986. In fact, SBICs’ total amount of SBA
financing outstanding actually declined over
that period. This decline is indicative of two
general trends within the SBIC program. First,
the number of active SBICs declined signifi-

TABLE 1

Development of the SBIC program
A ll S B IC s
1986

B a n k - o w n e d S B IC s
1991

1986

1991

TA

$3.30 b illio n

$4.24 b illio n

$1.89 b illio n

$3.08 b illio n

TOTCAP

$1.99 b illio n

$2.99 b illio n

$1.35 b illio n

$2.46 b illio n

PRIVCAP

$1.28 b illio n

$2.16 b illion

$0.83 b illio n

$1.75 b illio n

SBAFUND
N

$878 m illio n
292

$575 m illio n
176

$246 m illio n

100

$129 m illio n

68

N o te : V a ria b le s are d e fin e d in b o x 2.

FEDERAL RESERVE



BANK OF CHICAGO

25

BO X 2

Definitions of variables
DEBT
EQUITY
EQUITY and DEBT
LOANS
LOSS
N
PDEBT
PEQUITY
PLOANS
PRIVCAP

SBIC disbursements as purchases of debt instruments with equity features,
such as convertible bonds
SBIC disbursements as purchases of equity
SBIC disbursements as simultaneous purchases of equity and debt instruments
SBIC disbursements as loans
the ratio of provision for losses on accounts receivables to gross expenses
number of observations
the ratio of the stock of debt securities with equity features to total portfolio of
investments, with all assets measured by their market value
the ratio of the stock of equity securities to total portfolio of investments, with all
assets measured by their market value
the ratio of the stock of loans to total portfolio of investments, with all assets
measured by their market value
private capital defined as capital plus paid-in surplus

ROE-BV

the three-year average ratio of net income to book value of equity, 1989-91

ROE-MV

the three-year average ratio of net income to total capital (market value), 1989-91

SBAFUND
SBALEV
TA
TOTCAP

total amount of funds owed to the SB A
SBAFUND/PRIVCAP
market value of total assets, including unrealized gains or losses on portfolio
securities
market value of total capital, including unrealized gains or losses on securities held

cantly during those years. While a few SBICs
were formed during the period, a substantial
number either surrendered their license or went
into liquidation. At time of liquidation, those
firms held about $467 million in outstanding
SBA loans, which accounts for part of the
decline in the SBA leverage. Second, the
groups of SBICs that experienced the largest
growth in assets and capital—bank-owned
SBICs—used less SBA leverage on average.
During the 1986-91 period, of all SBICs,
bank-owned companies had the highest growth
rates in total assets and capital. In fact, over
the same period, the total assets of non-bankowned SBICs actually declined. Bank-owned
SBICs typically financed their growth through
private capital and relied less on SBA funds.
As table 2 shows, in 1991 bank-owned SBICs
had approximately $0.21 in SBA funds for
every $1 o f private capital, which was signifi­
cantly lower than the comparable figure for
non-bank-owned SBICs. Bank-owned SBICs
also tended to be larger and to have more total

capital relative to assets than non-bank-owned
SBICs. The higher capital ratios at bankowned SBICs suggest that those SBICs had a
greater cushion against unanticipated losses on
investments. The differences between bankowned and other SBICs are also evident in the
composition of their portfolios. In 1991, nonbank-owned SBICs had, on average, 41 percent
of their portfolios in loans and the remaining
59 percent in securities with equity features,
such as straight equity and convertible debt
securities. Among bank-owned SBICs, loans
represented only 11 percent o f their portfolios.
The differences in the portfolio composi­
tions of bank-owned versus other SBICs may
also explain the differences in their capital
structures. Until 1992, prepayment of SBA
financings entailed prohibitive costs. As a
result, SBICs that received SBA financing
when interest rates were high could not refi­
nance their debt when interest rates started to
fall, as they did in 1986. In other words, the
ex post costs of SBA funds were relatively

26

ECONOMIC PERSPECTIVES




TABLE 2

Financial characteristics of SBICs, 1991
______________ A l l S B IC s ___________
V a r ia b le

M ean

_________ B a n k - o w n e d S B IC s

S t. d e v ia t io n

TA

$12.28 m illio n

$45.33 m illio n 3

98.54

49.83

$16.97 m illio n

PRIVCAP

S t. d e v ia tio n

$36.21 m illio n 3

75.94

32.24

$24.12 m illio n

TOTCAP

M ean

$25.71 m illio n 3

48.55

64.72

SBALEV

$0.82

0.97

$0.21b

0.42

PLOANS

0.29

0.39

0.24

PDEBT

0.52

0.39

0 .11b
0.68

0.34

PEQUITY

0.15

0.26

0.183

0.27

ROE-MV

-0.06

0.29

-0.03

0.29

ROE-BV

0.02

0.58

-0.02

0.23

N o te : A ll fig u re s are fo r th e fis c a l y e a r 1991 e x c e p t R O E-M V an d ROE-BV, w h ic h are th e th re e -y e a r
a v e ra g e s fo r th e p e rio d 1989-91. V a ria b le s are d e fin e d in b o x 2.
S ig n ific a n tly h ig h e r th a n th e c o m p a ra b le n u m b e r fo r n o n -b a n k -o w n e d SBICs; p < .05.
S ig n if ic a n t ly lo w e r th a n th e c o m p a ra b le n u m b e r fo r n o n -b a n k -o w n e d SBICs; p < .05.

high for these firms. A General Accounting
Office report indicates that the costs of SBA
funds were particularly high for SBICs that
specialized in equity investments.1 Firms that
4
had a large fraction of their portfolio in equity
investments did not have regular cash flows
from their investments and frequently experi­
enced difficulties in meeting their obligations.
Bank-owned SBICs, however, were less likely
to be subject to these forces. Although a large

fraction of their portfolios consisted of equity
investments, they had more equity capital and
less SBA leverage than other SBICs.
The differences in the growth rates of total
assets of bank-owned and other SBICs are also
reflected in their total disbursements. As table

3 shows, between 1983 and 1992, SBICs in­
vested almost $4.7 billion in over 18,900 trans­
actions. Bank-owned SBICs provided about
$2.8 billion in over 5,300 of these transactions.
O f the $4.7 billion invested by all SBICs in the
1983-92 period, $1.3 billion was in loans; the
remaining $3.4 billion was divided among
equity-related investments.
On a year-by-year basis, investments by
all SBICs increased between 1983 and 1988;
thereafter, they declined. This suggests that if
a small firm was unable to obtain funding from
banks during the years 1990 to 1993, it was
unlikely to obtain funding from an SBIC. Fur­
thermore, the decline in SBIC investments
between 1989 and 1992 was not confined to

TABLE 3

Types of investments made by SBICs
A ll S B IC s
In v e s tm e n t ty p e

B a n k - o w n e d S B IC s

A v e r a g e s iz e

T o ta l a m o u n t

(d o lla rs )

(m illio n d o lla rs )

(d o lla rs )

$1,279.99

LOANS

T o ta l a m o u n t

(m illio n d o lla rs )

$127,782

$311.08

$209,341

DEBT

723.68

237,117

357.19

358,899

1,798.22

366,610

1,423.62

577,533

859.96

895,792

704.30

1,623,811

4 ,6 6 1 .8 5

EQUITY

2 4 6 ,2 1 7

2 ,7 9 6 .1 9

5 1 9 ,7 3 8

EQUITY and DEBT
T o ta l

A v e r a g e s iz e

N o te : N u m b e rs are th e d o lla r a m o u n ts o f th e flo w o f in v e s tm e n ts m a d e in th e p e rio d 1983-92.
V a ria b le s are d e fin e d in b o x 2.

FEDERAL RESERVE



BANK OF CHICAGO

27

bank-owned SBICs; in fact, invest­
FIGURE 1
ments by non-bank-owned SBICs
The top industries in which SBICs invested
declined more than those by bankl 983-92
owned SBICs.
Communications
A comparison of the flows of
investments by bank- and non­
Electronic & other
bank-owned SBICs indicates that
electrical equipment
their investment patterns are con­
sistent with the composition of
Business services
their portfolios. During the period
1983-92, more than one-half of the
Industrial machinery
$2.8 billion invested by bankand equipment
owned SBICs was in the form of
straight equity investments. More­
Holding companies
over, bank-owned SBICs account­
Wholesale durable goods
ed for about three-fourths of all
Health services
investments with equity features.
Printing and
|
Instruments
publishing
_
,.
Figure l shows the ten indus­
oho/
General building contractors
3.3 %
tries in which SBICs invested the
largest amounts over that period.
Investments in these top ten indus­
degrees of diversification across industries.1
5
tries accounted for more than one-half of total
As table 4 shows, the top three and top ten
investments. The largest amounts of invest­
industries in which bank-owned SBICs made
ments were made in communications, electron­
ic equipment, and in business services. While
investments accounted for approximately 23
percent and 57 percent of the portfolio of these
investments of all SBICs appear to be concen­
institutions, respectively. In contrast, the
trated mostly in service and high-technology
industries, there are significant cross-sectional
shares of the three largest industries in the
portfolios of other SBICs were 26 percent and
differences in the industries invested in and the
degree of diversification. SBICs owned by
62 percent, respectively.
We also examined the investments of
banks and other financial institutions invested
mostly in firms in the semiconductor and com­
SBICs according to the purpose for which
puter equipment industries. In contrast, SBICs
financing was obtained. Figure 2 shows the
owned by nonfmancial firms made a little
main reasons for which small businesses ob­
tained SBIC financing. O f the $4.7 billion
less than one-half of their investments in gro­
cery stores.
invested by all SBICs, about one-half was used
There are also differences between bankfor operating capital, one-third to acquire exist­
owned and other SBICs in terms of their
ing businesses, and the remainder to consoli­
date debts, fund research and development
(R&D) and marketing activities, and acquire
TABLE 4
or construct plants, buildings, machinery,
Portfolio shares of the top three
and land.
industries in which SBICs invested
Figure 3 shows the percentage of funds
1983-92
that were provided as loans, as well as the
percentage of funds provided by bank-owned
P e rc e n ta g e o f
In d u s tr y
< ll i n v e s t m e n t s
a
SBICs for each type of activity. When SBIC
funds were provided for activities generating
B a n k - o w n e d S B IC s
little collateral (such as R&D, marketing, and
Top three industries
22.6
acquisition of existing businesses), a large
Top ten industries
57.0
fraction of the funds was provided through
equity investments and by bank-owned SBICs.
N o n - b a n k - o w n e d S B IC s
For example, bank-owned SBICs supplied
Top three industries
26.1
more than three-fourths of the funds for R&D
Top ten industries
61.5
activities, primarily through equity participa-

28



ECONOMIC PERSPECTIVES

tion. Research and development, marketing,
and acquisition of existing businesses are risky
activities that are difficult to monitor and that
allow managers a great deal of discretion over
the disbursement of funds. As a result, the
agency costs of debt are likely to be high, and
funds are more likely to be supplied through
equity participations. On the other hand, when
funds financed such activities as new building
and plant construction, more than 85 percent
was provided through loans, and banks provid­
ed only 22 percent. This type of activity gen­
erates tangible assets and allows little manage­
ment discretion. Consequently, the agency

FEDERAL RESERVE BANK OF CHICAGO




costs of debt are likely to be low­
er; lenders can monitor managers
easily, minimizing the ability
of managers to shift funds to
riskier projects.
Similarly, firms in hightechnology industries tend to
invest in risky projects that gener­
ate very small or negative cash
flows in the short term, yet the
future profit opportunities of
these firms are relatively high.
As a result, when these firms
borrow funds, their probability of
bankruptcy is high. Furthermore,
investors that lend to these firms
cannot share in the surplus of
high-growth opportunities. In
contrast, when SBICs invest in
the equity of these firms as resid­
ual claimants, they share in the surplus. The
fact that bank-owned SBICs, which tend to
specialize in equity investments, invest in hightechnology firms suggests that agency costs of
debt financing are significant for these firms.
T h e p ro fita b ility o f SB IC s

As in any other business, an SBIC’s asset
quality, financial leverage, and investment mix
are likely to affect its profitability. Return on
equity (ROE), as measured by the ratio of net
earnings to equity, is perhaps the most com­
monly used measure of profitability. From the
standpoint of financial theory, ROE provides a
proxy for the returns available to
shareholders. An SBIC with low
earnings as a percentage of share­
holder claims is likely to experi­
ence falling share prices and
therefore increased costs of exter­
nal capital. In such a case, the
company’s growth potential is
likely to be lowered commensurately.
Examination of the mean
values of ROE in table 2 reveals
that bank-owned SBICs were
more profitable than non-bankowned SBICs during the years
1989 to 1991. Although it ap­
pears that all SBICs had negative
or very low average ROEs in that
period, there were significant
cross-sectional differences.

29

Some of these differences are related to SBA
leverage and the mix of SBICs’ investment
portfolios.
An SBIC’s investment portfolio consists
of loans, debt securities with equity features,
and equity interest. Because SBICs assume
credit risk exposure on these investments, asset
quality is particularly important for them. If
an SBIC is highly leveraged, large loan or
security losses can bring insolvency. The
quality of assets will be affected both by man­
agement’s control over its credit review func­
tion and by economic conditions. A decline in
credit quality can lead to write-offs and re­
duced earnings on the investments.
Loans are likely to be the least risky of
these types of investments. While higher risk
investments should be positively associated
with higher ROE, imprudent use of asset pow­
ers and inadequate risk management practices
will produce lower or negative ROE. Thus,
changes in investment mix can either increase
or decrease ROE. We calculated investment
mix (PLOANS) by dividing loans by total
portfolio of investments.
A more direct measure of the riskiness
of the investment portfolio is the loss experi­
ence (LOSS), measured by the provision for
losses on accounts receivable divided by gross
expenses. Other things being equal, a higher
loss provision reflects a higher degree of
expected loss in the investment portfolio.
Therefore, this ratio should be negatively relat­
ed to ROE.
Another variable that can influence ROE is
the amount o f SBA leverage (SBALEV). We
calculated this variable by dividing the dollar
value of debt that an SBIC owes to the SBA by
the sum of the private paid-in capital and paidin surplus of the SBIC. We expect that the
higher the leverage, the more likely it is that an
SBIC will have trouble repaying its obliga­
tions. On the other hand, greater leverage may
enable some SBICs to earn higher returns.
Thus, across SBICs, high SBA leverage may or
may not be indicative of lower ROEs.
The return on equity may also be related
to asset size (TA) because firm size may serve
as a proxy for SBIC asset diversification.
Large SBICs are more likely to have better
diversified investment portfolios than small
SBICs. Moreover, larger SBICs are more
likely to have professional managers with

considerable expertise and thus should show
better performance.
The following equation provides a simple
econometric specification o f the relationship
between ROE and the above-mentioned
variable,

30

ECONOMIC PERSPECTIVES




(1) ROE = a 0 + a,PLOANS + a 2 LOSS +
a,SBALEV + a , TA + e,
3
4
where e is an error term. We estimated equa­
tion 1 using time-series cross-sectional data
over the period 1986-91. To determine wheth­
er the portfolio decisions of bank-owned
SBICs have a different impact on ROE than
those of other SBICs, we estimated separate
coefficients for the two types of institutions.
SBICs must report each investment using
historical cost (book value) and historical cost
plus any unrealized gains or losses embedded
in the security (market value). We used book
values in the estimation of equation 1 to check
the reasonableness of our results using market
values. Finally, we transformed each of the
independent variables to examine how a one
standard deviation change in that variable
translates into changes in ROE. We calculated
the transformed variables by taking each vari­
able, subtracting its mean value over the sam­
ple period, and dividing by its standard devia­
tion. Assuming that each variable is a normal
random variable, one can show that the trans­
formed variable is its standard normal variate.
The results of estimating equation 1 ap­
pear in table 5. The first two columns present
the results using the non-transformed variables,
and the last two columns present the results for
the transformed variables. The market value
results in column one show that SBA leverage
is negatively correlated with ROE for both
bank-owned and other SBICs. Greater use of
subordinated debt and debentures provided by
the SBA tends to reduce profitability. Losses
on accounts receivable (LOSS) are negatively
correlated with ROE for both types of SBICs,
but they have a significant impact only for
non-bank-owned SBICs. Since non-bankowned SBICs tend to hold relatively more
loans than equity compared to bank-owned
SBICs, it is not surprising that the ROEs of
non-bank-owned SBICs are more sensitive to
changes in loss experience.

TABLE 5

The relationship between ROE and portfolio decision variables
T ra n s fo rm e d

T ra n s fo rm e d

M a rk e t ROE

Book ROE

m a rk e t R O E

book ROE

INTERCEPT

-0.0353
(-2.015)*

-0.0447
(-2.185)*

-0.0296
(-2.015)*

-0.0342
(-2.185)

BLOSSa

-0.1491
(-1.554)

-0.1457
(-1.524)

-0.0104
(-1.554)

-0.0102
(-1.524)

BSBALEV3

-0.0679
(-2.613)*

-0.0668
(-2.554) *

-0.0275
(-2.613) *

-0.0271
(-2.554)

BPLOANS3

0.1310
(3.249)*

0.1421
(3.311)*

0.0237
(3.249) *

0.0258
(3.311)

OLOSSb

-0.4167
(-4.670) *

-0.4537
(-4.398) *

-0.0413
(-4.670) *

-0.0449
(-4.398)

OSBALEVb

-0.0213
(-2.483)*

-0.0310
(-2.542) *

-0.0224
(-2.483) *

-0.0271
(2.542)

OPLOANSb

0.0813
(3.734)*

0.1042
(3.835) *

0.0327
(3.734)*

0.0427
(3.835)

TA

0.8229
(5.573)*

1.1230
(6.191)*

0.0360
(5.573)*

0.0426
(6.191)

BDUM C

0.0025
(0.116)

0.0057
(0.234)

0.0025
(0.116)

0.0057
(0.234)

0.051

0.056

0.051

0.056

10.393

11.356

10.393

11.356

1,398

1,398

1,398

1,398

V a r ia b le s

R2
F -statistic
N

N o te : T h e in d e p e n d e n t v a ria b le s ROE, P LO A N S, an d T A in th e "B o o k R O E" c o lu m n are m e a s u re d as
m a rk e t v a lu e less th e u n re a liz e d g a in s o r losses on s e c u ritie s . A n e s tim a tio n o f th e re s id u a ls fro m th e
o rd in a r y le a s t s q u a re s re g re s s io n e q u a tio n in d ic a te d th e pre se n ce o f h e te ro s c e d a s tic ity in th e e rro r te rm .
A s a re s u lt, w e use W h ite 's (1980) h e te ro s c e d a s tic -c o n s is te n t e s tim a te o f th e c o e ffic ie n t s ta n d a rd e rro rs to
c o m p u te th e t-s ta tis tic s (in p a re n th e s e s ).
“T h e le tte r B b e fo re a v a ria b le re fe rs to a b a n k -o w n e d SBIC v a ria b le .
bT h e le tte r 0 b e fo re a v a ria b le re fe rs to a n o n -b a n k -o w n e d SBIC v a ria b le .

C
BDUM is a dummy variable taking on a value of one for bank-owned SBICs, zero otherwise.
1 p < .10.

Larger SBICs tend to have higher ROE.
This suggests that large SBICs can diversify
their investment portfolio so as to achieve
superior performance. The variable measuring
investment composition is positively correlated
with ROE. A shift in the investment portfolio
from equity to loans tends to raise ROEs for
both bank-owned and other SBICs. This is an
important result because much of the discus­
sion about banking organizations’ involvement
with SBICs has to do with their using SBICs
to hold equity securities. Banks claim they are
losing market share in their traditional areas
of lending and deposit-taking and therefore
need, among other things, to be able to invest

FEDERAL RESERVE BAMC OF CHICAGO




directly in business enterprises. Regulators
worry, however, that these direct investments
may increase the riskiness of banking organiza­
tions and lower their profitability. We find
that bank-owned SBICs with above-average
investment in loans tend to have above-average
ROEs. This implies that a shift in the invest­
ment mix from loans to equity is likely to re­
duce profitability. However, to assess the
effect of equity investments on the riskiness of
banking organizations, it is not enough to show
that SBICs specializing in equity investments
have below-average ROEs; one must also eval­
uate whether they have higher or lower vari­
ability of ROE. When we used book value

31

measures, reported in the second column of table
5, the results are qualitatively the same as the
market value results in the first column. This
suggests that differences in accounting practices
apparently have very little effect on the estimat­
ed relationships between profitability and the
portfolio decision variables.
The results also suggest that SBICs with
above-average investments in loans and belowaverage SBA leverage will have above-average
ROEs. Furthermore, as the third and fourth
columns of table 5 show, the implied differences
in ROE are not trivial. For instance, the market
value results in column three indicate that for
bank-owned SBICs, a one standard deviation
increase in loans as a percentage of investments
would yield a 237 basis point increase in ROE.
A one standard deviation decrease in SBA lever­
age causes ROE to rise by 275 basis points. The
sensitivity of non-bank-owned SBICs’ ROE to
change in the above two variables is not differ­
ent from that of bank-owned SBICs. The book
value results in column four yield similar results
in these cases.
Overall, the results seem to indicate that
SBICs receiving above-average SBA leverage
perform more poorly than other investment
companies. SBICs that specialize in equity
investments are less profitable, on average, than
other firms. Nevertheless, the results suggest
that banking organizations, like other firms, tend
to perform better when allowed to provide
mixed loan-equity financing.
Our preliminary examination of the sources
of these relationships between profitability and
characteristics of SBICs suggest that the results
in table 5 are particularly strong for those insti­
tutions that did not survive our sample period.1
6
Furthermore, even though bank-owned and
other SBICs had similar parameter estimates,
test results indicate that the two groups had
significantly different regression equations. In
other words, the relationship between profitabil­
ity and firm characteristics is different for bankowned and non-bank-owned SBICs.
C o n c lu s io n s

The SBIC program appears to go a long
way toward resolving the conflict between the
types of institutions that are appropriate for
financing small businesses and the types of
financing they need. If, as has been argued,
banks have a comparative advantage in evaluat­
ing and monitoring small firms, allowing banks

32




to participate in such programs as the SBIC
program may offer significant advantages in
small business financing.
The empirical results in this article support
this argument. SBICs associated with banking
organizations play a significant role in the
program. On average, bank-owned SBICs
were significantly larger, had more capital,
obtained less SBA leverage, and invested a
greater portion of their portfolio in equity in­
vestments than non-bank-owned SBICs. Fur­
thermore, while the total assets and capital of
non-bank-owned SBICs declined over the
period from 1983 to 1992, the total assets and
capital of bank-owned SBICs grew.
These results suggest that bank-owned
SBICs were an essential part of the program
and that they took advantage of their expanded
powers by pursuing an extensive strategy of
equity investments. The evidence also sug­
gests that such equity investments were partic­
ularly important in funding activities and in­
dustries that are perceived to have high costs of
debt financing. Specifically, equity financing
and financing by bank-owned SBICs were
prominent for activities and industries that
generate few tangible assets and give greater
management discretion in the use of funds.
The empirical results on the relationship
between SBIC profitability and portfolio deci­
sions indicate that profitability is positively
related to size, the measure of asset quality,
and the ratio of loans to total investments. On
the other hand, profitability is negatively relat­
ed to SBA leverage. In addition, bank-owned
SBICs, which typically relied less on SBA
leverage, had higher returns on equity than
other SBICs. These results suggest that offer­
ing SBA subsidies was relatively less effective
in encouraging the flow of funds to small firms
in the long term than was allowing banking
organizations to participate in the program.
Our analysis in this article and our prelim­
inary results on the percentage of disburse­
ments that were repeat financings raise some
interesting questions.1 Do the investment
7
patterns of SBICs change over the course of
their relationship with small firms? In other
words, do SBICs learn more about small firms
as their relationships with them develop, and
is this reflected in their investment patterns?
Does the type and amount of investment in
first-time financings differ from those in subse­
quent financings? Are SBICs more likely

ECONOMIC PERSPECTIVES

to provide management services at the begin­
ning of their relationship with firms, or in
subsequent financings? Do small businesses
tend to obtain funds from more than one
SBIC? How do the SBIC units of banking
organizations contribute to the overall perfor­

mance and riskiness of banks? We plan to
address these questions in our future research.
We also plan to examine in more detail the
relationship between the profitability of SBICs
and their characteristics.

FOOTNOTES
'See Kanjorski (1993).

average net income less than $6 million. At present, these
revisions are under review.

2Board o f Governors (1958).
3See U.S. Small Business Administration (1992) for a
discussion o f a recent survey on small business financing.
4An SBIC is classified as bank-owned i f at least 10 percent
o f its equity is controlled by a banking organization.
5Petersen and Rajan (1994) report that nearly 75 percent o f
the firms in their sample, which consists o f 3,404 small
firms, are less than 10 years old. Furthermore, the majority
o f firms in the sample are partnerships, sole proprietorships,
and Chapter S corporations.
6Evidence on the industries that are dominated by small
businesses and the failure rates o f these firms is reported in
White (1982); Brown, Hamilton, and M edoff (1990); and
U.S. Small Business Administration (1992).
7Berger and Udell (1990, 1994) report that two-thirds o f
commercial bank loans and over 50 percent o f lines o f
credit to small firms are secured by collateral. Bank lend­
ing to small firms also appears to be positively correlated
with the amount o f assets that can be pledged as collateral
(Hooks and Opler 1993). Furthermore, according to Dia­
mond (1991) and others, asymmetric information problems
decrease as lenders learn more about firms through deposit­
taking and previous lending arrangements. Empirical
evidence in Petersen and Rajan (1994) and Berger and
Udell (1994) supports this argument.
8For example, the SB A offers guarantees on bank loans to
small businesses and the Small Business Incentive Act,
recently introduced by Senator Christopher Dodd, would
make it easier for investors to finance small businesses by
amending the Securities Act o f 1933 and the Investment
Company Act o f 1940.
9In the last year, the SBA has proposed to increase the
coverage o f the program by redefining small firms as those
that have net worth less than $ 18 m illion or two-year

l0The special role o f banking organizations in the finan­
cial system is examined in Diamond (1984), James
(1987), and Haubrich (1989).
"U.S. Small Business Administration (1993).
12Although the data comprise the S B A ’s entire computer
database on SBICs, there are a few m issing observations.
According to our calculations, there are 94 companies for
which there are m issing financial statements in the 198691 period and 14 firms that have no data for investments.
Since these represent a small fraction o f the database, we
do not expect our qualitative results to be affected signifi­
cantly by the m issing observations.
"Despite the healthy gains in the 1986-91 period, SBIC
funds represent a small fraction o f the total funds in
venture capital. According to statistics reported in Deger
(1993), venture capital firms managed $32.87 billion in
total capital in 1991, representing a 36 percent increase
from 1986.
I4U.S. General Accounting Office (1993).
"Diversification across industries was calculated using
the flow o f investments in the 1983-92 period. Therefore,
this is a measure o f diversification for new investments
during this period and does not, necessarily, reflect the
degree o f diversification for the entire portfolio. Never­
theless, the period examined is sufficiently long for the
diversification o f new investments to be a good measure
o f diversification o f the entire portfolio.
"These are institutions that either surrendered their licens­
es, went into liquidation, or merged during 1986-91.
"More than one-half o f all transactions in our sample
were repeat financings. Bank-owned SBICs accounted
for more o f the repeat transactions than did other SBICs,
and equity-related investments were more likely to be
repeat financings than were loans.

REFERENCES

Berger, Allen N., and Gregory F. Udell,
“Collateral, loan quality, and bank risk,” Jour­
nal o f Monetary Economics, Vol. 25, January
1990, pp. 21-42.

FEDERAL RESERVE



BANK OF CHICAGO

____________ , “Lines of credit and relation­
ship lending in small firm finance,” working
paper, Board of Governors of the Federal Re­
serve System, February 1994.

33

Board of Governors of the Federal Reserve
System, Report to the Committees on Small
Business of the U.S. Congress, Financing
Small Business, Washington, DC: U.S. Gov­
ernment Printing Office, 1958.

the House of Representatives Subcommittee on
Economic Growth of the Banking Committee,
103rd U.S. Congress, First Session, Washing­
ton: Government Printing Office, February 16,
1993, pp. 50-51.

Brown, Charles, James Hamilton, and
James Medoff, Employers Large and Small,
Cambridge: Harvard University Press, 1990.

Petersen, Mitchell A., and Raghuram G.
Rajan, “The benefits of lending relationships:
evidence from small business data,” Journal
o f Finance, Vol. 49, N o.l, March 1994,
pp. 3-37.

Deger, Renee, “Capital levels decline again,”
Venture Capital Journal, November 1993,
pp. 34-37.
Diamond, Douglas W., “Financial intermedia­
tion and delegated monitoring,” Review o f
Economic Studies, Vol. 51, 1984, pp. 393-414.
______________ , “Monitoring and reputation:
the choice between bank loans and directly
placed debt,” Journal o f Political Economy,
Vol. 99, 1991, pp. 688-721.

U.S. General Accounting Office, Report to the
Chairman of the Committee on Small Business
of the U.S. Senate, Small Business: Financial
Health o f Small Business Investment Compa­
nies, Washington, DC: General Accounting
Office, May 1993.
U.S. Small Business Administration, The
State o f Small Business: A Report o f the Presi­
dent, Washington, DC: U.S. Government Print­
ing Office, 1992.

Haubrich, Joseph, “Financial intermediation,
delegated monitoring, and long-term relation­
ships,” Journal o f Banking and Finance, Vol.
13, 1989, pp. 9-20.

____________ , Investment Division, SBIC
Program: Statistical Package, Washington, DC:
Small Business Administration, March 1993.

Hooks, Linda, and Tim C. Opler, “The de­
terminants of corporate bank borrowing,”
working paper, Southern Methodist University,
May 1993.

White, Halbert, “A heteroskedasticity-consistent covariance matrix estimator and a direct
test for heteroskedasticity,” Econometrics, Vol.
48, 1980, pp. 817-838.

James, Christopher, “Some evidence on the
uniqueness of bank loans,” Journal o f Finan­
cial Economics, Vol. 19, 1987, pp. 217-235.

White, Lawrence, “The determinants of the
relative importance o f small business,” Review
o f Economics and Statistics, Vol. 64, No. 1,
February 1982, pp. 42-49.

Kanjorski, Paul, “Opening statement of the
Honorable Paul E. Kanjorski,” Hearing before

34



ECONOMIC PERSPECTIVES

ECONOMIC PERSPECTIVES
BULK RATE
P u blic Inform ation C en ter

Federal Reserve Bank of Chicago
P.O. Box 834
Chicago, Illinois 60690-0834

U.S. POSTAGE
PAID
CHICAGO, ILLINOIS
PERMIT NO. 1942

Do N o t F o rw a rd
A d d ress C o rre c tio n R equested
R e tu rn P o stag e G u a ra n te e d

Mailing label corrections or deletions:
M a k e c h a n g e s o r m ark

Delete from mailing list

o n th e la b e l an d fa x th e la b e l to 3 1 2 - 3 2 2 - 2 3 4 1 ,
o r s e n d to: M a il S e r v ic e s ,
F e d e r a l R e s e r v e B a n k o f C h ic a g o ,
P .O . B o x 8 3 4 , C h ic a g o , I llin o is 6 0 6 9 0 - 0 8 3 4

FEDERAL RESERVE BANK
OF CHICAGO