View original document

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

Economic Review
Federal Reserve Bank of Dallas
March 1985

1

A Comparative Analysis
of Mature Hispanic-Owned Banks
Robert T. Clair

Evidence presented in this article shows that the
financial performance of Hispanic-owned banks has
been on a par with that of nonminority-owned
banks. Minority-owned banks, as a whole, have
suffered from high loan losses and expensive
deposits; consequently, their financial performance
has been below that of nonminority-owned banks. By
charging higher interest rates on loans and imposing
higher service charges on deposits, Hispanic-owned
banks have been able to offset their higher costs .

12

FDIC Settlement Practices
and the Size of Failed Banks
Eugenie D. Short

Examination of data on failed banks from 1921 to
1984 indicates that the relative size of failures
increased significantly after 1972. Before that
year, most bank failures were small. FDIC practices
for settling bank failures give preferential treatment
to the depositors of large banks . Incentives provided
by deposit guarantees induce banks to increase their
exposure to risk . It is possible that stronger implicit
guarantees to the uninsured depositors of large banks
led these institutions to incur even greater risk. Such
preferential treatment may have altered the size
distribution of bank failures in this country.

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

A Comparative Analysis
of Mature Hispanic-Owned Banks
Robert T. Clair
Economist
Federal Reserve Bank of Dallas

The structure and performance of minority-owned
commercial banks have been analyzed in a sizable
amount of published research. A bank is classified
as minority-owned if over half of the equity is held
by individuals in minority groups. This research has
shown that minority-owned banks are not as profitable as nonminority-owned banks. Highly volatile
deposit bases and scarce profitable lending opportunities are cited as the structural causes of their
relatively poor financial performance.'
Despite the sizable amount of published research
concerning minority-owned banks, none of this
research has directly analyzed Hispanic-owned
banks. Initial studies were devoted to black-owned
banks, and later research examined all minorityowned banks as a single group. The primary purpose
of this article is to conduct a comparative analysis
of the structure and performance of Hispanic-owned
commercial banks of the type previously conducted
for black- and minority-owned banks.
I n contrast to previous minority-owned bank

The views expressed are those of the author and do not necessarily
reflect the positions of the Federal Reserve Bank of Dallas or the
Federal Reserve System.
Economic Review I March 1985

studies, the analysis of Hispanic-owned banks shows
that their profitability is on a par with that of
nonminority-owned banks. There is evidence that
the structure of Hispanic-owned banks is significantly different from the structure of nonminorityowned banks. Furthermore, these structural differences are similar to those previously established
between minority-owned banks and nonminorityowned banks. Hispanic-owned banks have dealt
with the structural differences by pricing loans to
compensate the bank for exposure to credit risk
and by setting service charges higher to offset, at
least partially, higher noninterest expenses. As a
result, the rates of return on assets and equity
are not significantly different from those of their
nonminority-owned competitors. To the extent that
Hispanic-owned banks have been characterized as
poor-performing minority-owned banks, they have
been misrepresented.

1. "Structural" is used here to describe the characteristics of the
markets in which the banks operate, and these characteristics
are beyond the control of bank management. Structural
characteristics would include the income profile of depositors,
the variability of borrowers' incomes (an important cause of
loan defaults), labor costs, and other operating costs.

Summary of literature on minority-owned banks
The early studies of minority-owned banks dealt
with black-owned banks. 2 These studies focused on
the long-run viability of these institutions and on
their economic and sociological contribution to the
minority community. Subsequent research was
similar in approach, but better techniques for sample selection were used to control for problems involving bank size, economies of scale, locational
differences, and bank age. The research was also expanded to include the study of all minority-owned
banks as well as black-owned banks. 3
The evidence presented in these studies indicated
that minority-owned banks were less profitable than
their nonminority-owned counterparts. Two structural causes have consistently been put forward as
explanations for the poor financial performances.
Minority-owned banks generally have a deposit base
that is a relatively more expensive source of
loanable funds, and they face limited profitable
lending opportunities. In addition, some studies
have also cited less-experienced management as a
cause of poor financial performance at minorityowned banks. 4

2. See Andrew F. Brimmer, "The Black Banks: An Assessment of
Perform<ince and> Prospects," Journal of Finance 26 (May 1971):
379-405; and Edward D. Irons, "Black Banking- Problems and
Prospects," Journal of Finance 26 (May 1971): 407-25.

3. See Timothy Bates and William Bradford, "An Analysis of the
Portfolio Behavior of Black-Owned Commercial Banks," Journal of Finance 35 (June 1980): 753-68; Harold Black, "An
Analysis of Minority Banks," Research Paper no. 77-6, Division
of Economic Research and Analysis, Office of the Comptroller
of the Currency (Washington, D.C., 1977); John T. Boorman,
"The Prospects for Minority-Owned Commercial Banks: A
Comparative Performance Analysis," Journal of Bank Research
4 (Winter 1974): 263-79; John T. Boorman and Myron L. Kwast,
"The Start-Up Experience of Minority-Owned Commercial
Banks: A Comparative Analysis," Journal of Finance 29
(September 1974): 1123-41; Myron L. Kwast, "New MinorityOwned Commercial Banks: A Statistical Analysis," Journal of
Bank Research 12 (Spring 1981): 37-45; and Myron L. Kwast
and Harold Black, "An Analysis of the Behavior of Mature
Black-Owned Commercial Banks," Journal of Economics and
Business 35, no. 1 (1983): 41-54.
4. The quality of bank management is not used in this research
as an explanatory variable of bank financial performance.
Only structural differences will be examined. See Mona J.
Gardner, "Minority Owned Banks: A Managerial and Performance Analysis," Journal of Bank Research 15 (Spring 1984):

26-34.
2

The higher cost of the deposit base has been attributed to three sources. First, minority-owned
banks typically serve low-income communities. Consequently, the average balance in deposit accounts
is lower at minority-owned banks. To the extent
that there are fixed costs per deposit account, a
minority-owned bank will incur greater costs to
maintain the same dollar amount of deposits as a
nonminority-owned bank.
Moreover, some minority-owned bank depositors,
most likely those with lower incomes, have used
savings accounts as transaction accounts in place of
traditional checking accounts. Apparently, the opportunity cost of the time spent by the depositor in
conducting transactions with a savings account is
less than the fees for checking services. As a result,
there is more frequent use of the teller windows for
transactions. The frequent withdrawals lead to
unusually high teller expenses; hence, the average
variable cost of servicing the deposits is higher at
minority-owned banks. 5
The third cause of the expensive deposit base is
the volatil ity of deposits. Private deposits have been
shown to be more volatile at minority-owned banks
than at nonminority-owned banks, and a large portion of total deposits consists of relatively volatile
government deposits. 6 Beginning in 1971, the
Federal Government has used the Minority Bank
Deposit Program to place deposits at minorityowned banks. Generally, government deposits are
quite volatile and account for a significantly larger
share of total deposits at minority-owned banks than
at nonm inority-owned banks. 7
The above attributes of the deposit bases of
minority-owned banks result in three characteristic
financial ratios. First, the volatility of deposits implies a greater need for liquidity to meet deposit
outflows. Minority-owned banks hold a greater proportion of assets in very liquid forms, such as cash
and federal funds sold. Second, the cost of servicing
their deposit base is higher per dollar deposited. As
a result, the ratio of salaries and employee benefits

5. See Boorman, "The Prospects for Minority-Owned Commercial
Banks."
6. See Bates and Bradford, "An Analysis of the Portfolio Behavior
of Black-Owned Commercial Banks."
7. For one black-owned bank, government deposits were 66 percent of total deposits at the end of 1983.

Federal Reserve Bank of Dallas

to total deposits is significantly higher at minorityowned banks. On the other hand, because their
deposit base has a larger share of savings accounts,
minority-owned banks have lower interest expenses
per dollar deposited than do their nonminorityowned counterparts that raise funds through
relatively more expensive time deposits. The lower
interest expenses will offset, to some extent, the
higher salary and benefit costs.
The second structural cause of poor financial
performance at minority-owned banks, as cited in
previous studies, is a lack of profitable lending opportunities. Finding creditworthy borrowers appears
to be a problem at minority-owned banks. The empirical results show that mi-nority-owned banks consistently have higher loan-loss provisions relative to
total loans than do nonminority-owned banks. Furthermore, their loan-to-asset ratio is significantly
lower, partly reflecting the greater need for liquidity
mentioned earlier. There is also evidence that attempts to enforce tighter credit standards, in an effort to reduce loan losses, have resulted in further
declines in the loan-asset ratio. 8
Analysis of Hispanic-owned banks
The empirical analysis presented here addresses
two basic questions. First, do Hispanic-owned banks
face structural characterisitics common to other
minority-owned banks? Second, do these structural
characteristics result in lower financial performance
at Hispanic-owned banks relative to nonminorityowned banks, as they have at other minority-owned
banks? Before these questions are addressed, the
data and statistical technique used are described.
The data. The data were obtained from the
Report of Condition and the Report of Income. A
statistical sample was constructed to control for
bank age, asset size, multibank holding company affiliation, and location. Annual data were collected
for 1980 through 1983. The resulting data sample
had 82 observations for Hispanic-owned banks and
3,226 observations for nonminority-owned banks.
This data sample was used to calculate average
financial ratios for Hispanic- and nonminority-owned
banks. A t test was employed to determine whether
the ratios for Hispanic- and nonminority-owned

banks were significantly different. An F test was
conducted to determine whether the variances
could be treated as identical, and the t tests were
adjusted appropriately if the variances were significantly different.
The decision to compare mature Hispanic-owned
banks with mature nonminority-owned banks was
made to eliminate any distortions that might be
associated with the start-up experiences of H ispanicowned banks. Examination of mature banks would
suggest that any differences between Hispanic- and
nonminority-owned banks would be of a long-term
nature. Furthermore, this approach would permit a
comparison of empirical results with results obtained for mature black-owned banks. 9 Hence, the
sample was restricted to banks that had been in
operation at least four years. Previous research has
shown that the significant fixed cost associated with
establishment of a new bank is usually eliminated
by the fourth year of operation.'o
The sample was further restricted to banks with
assets of at least $10 million but not more than
$150 million. This restriction was imposed to reduce
any problems or differences attributable to
economies of scale or to large-bank access to different financial markets. As a result, two Hispanicowned banks had to be dropped from the sample.
Notably, the average total assets of these two banks
were 10 times average assets of the remaining
Hispanic-owned banks.
Banks held by multibank holding companies were
also dropped from the sample. Previous research
has shown that affiliated banks are significantly different from independent banks in terms of asset
structure, loan portfolio composition, pricing policy,
and capital ization." Properly identifying the causes
of differences between Hispanic- and nonminorityowned banks would be difficult, if not impossible, if
the factor of holding company structure were not
controlled.
Finally, the nonminority-owned banks were required to be located in the same markets served by

9. See Kwast and Black, "An Analysis of the Behavior of Mature
Black-Owned Commercial Banks."

10. See Kwast, "New Minority-Owned Commercial Banks."
11. See Duane B. Graddy and Reuben Kyle, III, "Affiliated Bank
8. See Boorman, "The Prospects for Minority-Owned Cqmmercial
Banks."

Economic Review I March 1985

Performance and the Simultaneity of Financial DecisionMaking," Journal of Finance 35 (September 1980): 951-57.

3

Table 1

SELECTED CHARACTERISTICS OF DEPOSIT
BASES FOR MATURE HISPANICAND NONMINORITY-OWNED BANKS
Hispanicowned
banks
Ratio

Nonminorityowned
banks

Percent, 1980-83

.91 *

.57

Salaries and employee benefits/total deposits

2.52*

2.12

Interest paid on deposits/total deposits.

5.81 *

6.24

Service charges on deposits/total deposits

* Significant at the .05 level.
SOURCE OF PRIMARY DATA: Board of Governors, Federal Reserve System.

Hispanic-owned banks. "Market" is defined here as
a standard metropolitan statistical area (SMSA), if
possible, or a county, for rural banks. This is the
standard definition of a market in the minorityowned bank literature, and the limitations of this
definition have been discussed in the literature. 12
These limitations become important in interpreting
the empirical results.
Asset and liability structure. The first issue examined was whether Hispanic-owned banks have
deposit bases that are significantly different from
those of nonminority-owned banks and similar in
nature to deposit bases of all minority-owned banks
as a group. Three financial ratios are cited in the
I iterature as characteristic of deposit bases of
minority-owned banks. The higher level of account
activity per deposit suggests that the ratio of service
charges on deposits to total deposits at minorityowned banks will be higher. Second, since the
average deposit account at minority-owned banks
tends to be smaller and more volatile than at
nonminority-owned banks, the ratio of employee
expenses to total deposits is likely to be higher.
Finally, the ratio of interest cost of deposits to total
deposits is likely to be lower at institutions with a
large proportion of savings accounts-a common
characteristic of minority-owned banks.
These ratios are presented in Table 1 for both
Hispanic- and nonminority-owned banks, The ratios
are statistically different in all three cases and sup-

12. See Black, "An Analysis of Minority Banks."
4

port the hypothesis that Hispanic-owned banks have
a deposit structure that is different from the structure for nonminority-owned banks. Furthermore,
the differences are similar to those established in
previous studies of minority-owned banks.
Hispanic-owned banks also were examined to
determine whether the structural characteristics of
their asset portfolios are significantly different from
those of nonminority-owned banks. Asset characteristics have typically been compared by examining
the composition of assets and the loan-loss experience. These two measures, however, are not independent of each other. If a minority-owned bank
faces a lack of profitable lending opportunities, one
option for the bank is to accept an asset portfolio
that is more heavily invested in securities. In this
case, the loan-asset ratio is usually significantly less
than at comparable nonminority-owned banks. On
the other hand, the minority-owned bank could
make efforts to increase lending, but the loans are
more likely to be made to less creditworthy borrowers. As a result, net loan losses rise relative to
total loans.
Evidence indicates that Hispanic-owned banks
face asset markets with characteristics different
from those of nonminority-owned competitors and
that these differences are similar to those established for minority-owned banks as a group. The
ratio of net loan losses to total loans at Hispanicowned banks, reported in Table 2, is nearly double
that of nonminority-owned banks, a difference that
is statistically significant at the 1-percent level. The
loan-asset ratio is essentially the same at HispanicFederal Reserve Bank of Dallas

Table 2

SELECTED CHARACTERISTICS OF ASSETS AT MATURE
HISPANIC- AND NONMINORITY-OWNED BANKS
Hispanicowned
banks

Nonminorityowned
banks

Percent, 1980-83

Ratio

Net loan,losses/total loans

1.01 *

.58

Total loans/total assets

53.15

52.07

Cash plus net federal funds sold/total deposits.

26.14*

17.61

Government securities/total deposits.

19.07*

27.47

* Significant at the .05 level.
SOURCE OF PRIMARY DATA: Board of Governors, Federal Reserve System.

and nonminority-owned banks. This may suggest
that management at Hispanic-owned banks is willing
to incur higher loan losses as a cost of maintaining
an average proportion of assets in loans.
Table 2 also shows an interesting difference in the
composition of investments. Financial assets other
than loans have been split into two broad categories- primary and secondary sources of liquidity.
Primary sources include cash and near-cash items,
such as federal funds sold. Secondary sources are
basically investment securities. I n past research, it
has been reported that a high primary liquidity ratio
is evidence of a lack of profitable lending opportunities.13 When this ratio is examined with the
secondary liquidity ratio and the loan-asset ratio,
the evidence for Hispanic-owned banks suggests
that higher levels of primary sources of liquidity are
exactly offset by lower levels of secondary sources.
This structure is more likely to result from the
volatile deposit base than from any lack of profitable lending opportunities.
Financial performance. The composition of total
operating expenses and income demonstrates some
effects of the different structural characteristics of
Hispanic-owned banks. Several measures of performance are presented in Table 3. The data indicate
that total operating expenses relative to total assets
are significantly higher at Hispanic-owned banks,
primarily because of significantly higher labor costs

13. See Brimmer, "The Black Banks."

Economic Review I March 1985

and loan losses. The higher expenses, however, are
offset by higher income. The higher income can be
attributed to a higher gross return on loans and
higher service charges per deposit.
Because the higher income offsets higher expenses, the financial performance of Hispanicowned banks is on a par with that of nonminorityowned banks. The after-tax returns on assets and on
equity are not significantly different from those of
nonminority-owned banks. 14 This result is contrary
to nearly all the published research that compares
minority- and nonminority-owned banks. (Interpretation and analysis of study results are presented in
the next section.)
Mature Hispanic- and black-owned commercial
banks are compared with their nonminority-owned
competitors in the Appendix. The evidence there indicates a high degree of similarity in the liability
structure of these two types of minority-owned
banks. Furthermore, it appears that both Hispanicand black-owned banks have relatively high rates of
net loan loss. Compared with their nonminorityowned competitors, Hispanic-owned banks generate
significantly more operating income relative to
assets while black-owned banks generate significantly less. At Hispanic-owned banks the gross

14. The difference of 3.3 percentage points in the return on
equity may appear large, but the variances of such returns for
both minority- and nonminority-owned banks are also large.
To state definitively that the ratios are unequal would be subject to a probability of error in excess of 30 percent.

5

Table 3
FINANCIAL PERFORMANCE MEASURES FOR MATURE
HISPANIC- AND NONMINORITY-OWNED BANKS
H ispanicowned
banks

Nonminorityowned
banks

Percent, 1980-83

Ratio

Total operating expenses/total assets.

10.39*

9.56

Total operating income/total assets.

11.62*

10.87

Net income after taxes/total assets.

.78

.97

Net income after taxes/total equity

6.78

10.08

* Significant at the .05 level.
SOURCE OF PRIMARY DATA: Board of Governors, Federal Reserve System.

return on loans is significantly higher than at
nonminority-owned banks; at black-owned banks the
gross return on loans is less than at nonminorityowned banks. Similarly, the ratio of service charges
on deposits to total private deposits is significantly
higher at Hispanic-owned banks than at nonminority-owned banks, but there is little difference
in this ratio between black- and nonminority-owned
banks. As a result, there is no significant difference
between the return on assets at Hispanic- and
nonminority-owned banks, but the return on assets
at black-owned banks is significantly lower than the
return at nonminority-owned banks.

Interpretation of the empirical results
The most likely explanation for the comparable performance of Hispanic- and nonminority-owned
banks, despite the structural differences, is that
Hispanic-owned banks have priced their loans and
services higher to compensate, at least partially, for
the higher costs incurred while operating in structurally different asset and liability markets. If service charges on deposits are priced to cover the
higher handling cost attributable to the difference
in deposit bases, then the higher cost of servicing
deposits will be offset by higher income from service charges. In this case, the higher costs resulting
from the difference in liability structure would not
lead to lower performance. Similarly, if interest
rates on loans include risk premiums to compensate
the bank for exposure to credit risk, the net return
on loans to borrowers with various risk profiles
6

should be the same. If this were the case, the structurally different asset market should have no effect
on the return on assets or the return on equity.
As has already been established, the loan portfolios of Hispanic-owned banks are more exposed to
credit risk. The gross return on loans at Hispanicowned banks is 83 basis points higher than at
nonminority-owned banks, a statistically significant
difference. After adjustment for the higher loan
losses, however, the net return on loans at Hispanicowned banks is only 31 basis points above that of
nonminority-owned banks, and the difference is no
longer significant. 1s The implication is that Hispanicowned banks have dealt with the problem of serving
a loan market that is characterized by higher
default rates by raising interest rates to cover loan
losses.
Two solutions to the problems of minority-owned
banks have been offered by previous researchers.
The problem of an expensive deposit base should be
addressed by raising service charges to a level that
will cover the expenses of handling the deposits, Apparently, the experience of Hispanic-owned banks
may be an example of this successful strategy. It
has been suggested that the problem of high loan
losses be addressed by enforcing higher credit standards. The evidence for Hispanic-owned banks offers no support of this proposed solution; to the
15. The net return on loans was calculated as interest and fees
earned on loans less net chargeoffs (chargeoffs less
recoveries) divided by total loans.
Federal Reserve Bank of Dallas

contrary, it may point out that this solution is unnecessarily restrictive. Interestingly, the previous
researchers have focused exclusively on a nonprice
solution for loans, but several have recommended a
price solution for volatile deposits.
The previously proposed solution of tightening
credit standards was tied closely to the assumption
that minority- and nonminority-owned banks operate
in the same market. An implicit assumption that
minority-owned banks had to match the loan interest rates of the nonminority-owned banks was
also made. Emphasis was placed on lowering loan
losses to raise net income. The ability to increase interest and fee income from loans was not viewed as
a viable option.
Empirical evidence suggests that minority- and
nonminority-owned banks do not operate in the
same market. In studies analyzing these banks, the
market is usually defined as the SMSA or county in
which the minority-owned bank is located. This
geographical definition of a market has not dealt explicitly with the notion that a market is a collection
of economic agents transacting with each other. The
point has been made that banks operating in the
same SMSA may be serving primary market areas
that have sharply different economic characteristics. ' • In support of this view, the empirical work
presented in these studies tends to suggest important differences in the primary markets. The volatility of deposits, for example, indicates a difference
in the characteristics of depositors.
Under the alternative hypothesis that minorityowned banks are serving different markets, they
would be expected to have different structural
characteristics. These characteristics might include a
volatile deposit base or a loan portfolio with a
higher loan default rate. A lower return on assets,
however, would not necessarily be expected. Competition should exert pressure on bank management
to place funds where the bank would receive a competitive return, adjusted for risk. This alternative
hypothesis provides a better explanation of the empirical evidence obtained by examining Hispanicowned banks.
Conclusion

The evidence for mature Hispanic-owned banks

16. See Black, "An Analysis of Minority Banks."

Economic Review I March 1985

shows that these banks face the same structural
characteristics as minority-owned banks in generalspecifically, volatile deposits and high loan losses.
The Hispanic-owned banks, however, perform as
well as nonminority-owned banks. This equal financial performance is atypical in the literature on
minority-owned banks.
The experience of Hispanic-owned banks shows
that relatively high loan losses are not in themselves
a cause of poor financial performance. High loan
losses should not result in lower profits unless the
bank is not being compensated for its exposure to
credit risk. The approach of charging higher interest
rates in order to cover the expense of higher loan
losses diverges sharply from the prescriptions of
previous researchers. These researchers have consistently recommended that the problem of relatively high loan losses would best be solved by enforcing higher credit standards. This recommendation is
questionable on the basis of its implicit assumption
that minority- and nonminority-owned banks operate
in the same market. This assumption has been maintained despite evidence that minority-owned banks
operate in structurally different asset and liability
markets.
It is likely that many minority-owned banks could
profitably employ the same methods as Hispanicowned banks of pricing services and loans to attempt to cover handling expenses and loan losses. '7
In addition, it may be an improvement in the provision of credit to borrowers for minority-owned
banks to be willing to extend credit to riskier borrowers, albeit at higher interest rates. These banks
might establish multiple credit standards, and interest rates on loans would increase with the
riskiness of the borrower. This approach would be
an improvement over enforcing a single credit standard and refusing to lend to riskier borrowers
regardless of the interest rate that could be charged
on the loans. It is possible, however, that the transaction and information costs of determining the

17. Precise measurement of the cost of deposits at Hispanicowned banks is not possible. I nterest expenses are less at
these banks but noninterest expenses are higher, and it is not
possible to determine accurately what portion of non interest
expenses should be attributed to deposits. There is evidence
that deposits at Hispanic-owned banks are expensive to service but the banks' net income is not adversely affected;
hence, the implication is that deposit costs are being covered.

7

creditworthiness of some borrowers might outweigh
the potential gains.
Future research in comparative studies of
minority-owned banks should attempt to control the
data sample, if possible, so that both the minorityand the nonminority-owned banks serve the same
primary market area. The present market definitions

of county or SMSA are likely to be far too broad to
ensure that primary market areas are similar. If
such control is not possible, the assumption that
the minority- and nonminority-owned banks are
operating in the same market should be viewed with
at least some skepticism.

Appendix
A Comparison of Mature
Black- and Hispanic-Owned Banks
The empirical work in this study was formulated to
allow a comparison of the results for mature Hispanicowned banks with the results obtained by Kwast and
Black in their study of mature black-owned banks.' A
direct statistical comparison of Hispanic- and blackowned banks is not possible because there are too few
banks operating in the same markets to obtain a sample of adequate size. A comparison can be made,
however, by calculating the differences between the
ratios of Hispanic-owned banks and their nonminorityowned competitors and comparing them with the differences in the same ratios for black-owned banks and
their nonminority-owned competitors. The data
samples were controlled for bank size and age.

Asset and liability composition
The asset composition, reported in Table A, shows
some interesting similarities and some striking differences. First, the difference in the loan-asset ratio is
negative for black-owned banks compared with nonminority banks; for Hispanic-owned banks there is no
significant difference for the same ratio. Furthermore,
investment in Federal Government securities is quite
different for the two types of minority banks. Blackowned banks held a larger proportion of these
securities than their nonminority counterparts, while

1. Myron L. Kwast and Harold Black. "An Analysis of the Behavior of
Mature Black-Owned Commercial Banks," Journal of Economics
and Business 35, no. 1 (1983): 41-54.

8

Hispanic-owned banks held less. I n contrast, the difference in the primary liquidity ratio-cash and
federal funds sold as a percentage of total assets - is
nearly identical for black- and Hispanic-owned banks,
but Hispanic-owned banks split their liquid assets
more evenly between cash and federal funds sold
while black-owned banks invest primarily in federal
funds sold.
The composition of the loan portfolios is strikingly
different. Black-owned banks are heavily concentrated in real estate lending. The effect of this concentration appears to be reduced consumer lending.
Hispanic-owned banks, on the other hand, hold much
less of their loan portfolio in real estate loans than do
their nonminority competitors. Instead, Hispanicowned banks lend more to both businesses and
consumers.
The composition of liabilities at Hispanic- and
black-owned banks shows similar types of differences
from their respective nonminority competitors, though
the magnitudes differ in some cases. The most notable
difference in magnitude is in government deposits, the
sum of Federal Government deposits and state and
local government deposits. Both black- and Hispanicowned banks have significantly higher ratios for these
deposits, but the difference is far larger at blackowned banks. These deposits are usually required to
be collateralized with government securities. This may
be an explanation for the large holdings of Federal
Government securities at black-owned banks. Correspondingly, demand deposits and time deposits acFederal Reserve Bank of Dallas

Table A
DIFFERENCES IN ASSET AND LIABILITY COMPOSITION
OF MATURE HISPANIC- AND BLACK-OWNED BANKS
RELATIVE TO THEIR NONMINORITY COMPETITORS
Hispanicowned
banks

Blackowned
banks

Difference from
nonminority-owned
competitors
(Percent)

Item

Total capital as percent of total assets.

.1

Asset components as percent of total assets
Cash and due.
Federal Government securities
State and local government securities
Federal funds sold and repurchase agreements.
Gross loans
Real estate loans
Commercial and industrial loans.
Consumer loans
Other assets.
Loan components as percent of gross loans
Real estate loans.
Commercial and industrial loans.
Consumer loans
Other loans
Liability components as percent of total liabilities
Demand deposits of individuals, partnerships, and corporations ..
Time deposits of individuals, partnerships, and corporations.
U.S. Government deposits.
State and local government deposits
Federal funds purchased and repurchase agreements
Other deposits and liabilities.

.5*

3.7*
-2.9*
-5.5*
4.8*
1.1
-5.2*
2.3
4.5*
-1.2*

.9*
9.3*
-6.3*
7.3*
-11.3*'
.3
-3.6*
-8.1 *
.0

-10.9*
5.1 *
7.5*
-1.7

7.9*
1.2
-10.6*
1.4*

-.3
-7.6*
2.2*
5.0*
-.9*
1.7*

-8.3*
-10.5*
15.9*
3.9*
-1.0*
.1

* Significant at the .05 level.
NOTE: Data for Hispanic-owned banks are for 1980 through 1983.
Data for black-owned banks are for June 1976 through June 1979.
SOURCES OF PRIMARY DATA:
Board of Governors, Federal Reserve System.
Kwast and Black, "An Analysis of the Behavior of Mature Black-Owned Commercial Banks."

count for smaller shares of total liabilities at both
Hispanic- and black-owned banks.
Sources and uses of income
The differences in the sources of income, reported in
Table B, at Hispanic- and black-owned banks are, for
the most part, directly related to their differences in
asset composition. Black-owned banks earn significantly less income from loans than their nonminority
counterparts, while Hispanic-owned banks show no
difference. This is directly related to the com para-

Economic Review I March 1985

tively low loan-asset ratio at black-owned banks. Conversely, black-owned banks earn relatively more income from Federal Government securities because
their asset portfolios are more heavily invested in
those securities.
The final marked difference in sources of income is
evident in service charges on deposit accounts.
Hispanic-owned banks earn relatively more from this
category than their nonminority competitors. In contrast, black-owned banks earn about the same share of
income from this category as do their nonminority
9

Table B
DIFFERENCES IN SOURCES AND USES OF INCOME
FOR MATURE HISPANIC- AND BLACK-OWNED BANKS
RELATIVE TO THEIR NONMINORITY COMPETITORS
Hispanicowned
banks

Blackowned
banks

Difference from
nonm inority-owned
competitors
(Percent)

Item

Operating income components as percent of total operating income
Interest and fees on loans
Interest on Federal Government securities.
I nterest on state and local government securities
I nterest on federal funds sold and repurchase agreements.
Service charges on deposit accounts
Other service charges and fees
Other income

1.1
-2.3*
3.4*
3.4*
2.3*
.6*
-1.1

-12.4*
8.7*
-3.9*
7.5*
.3
.6
-1.0*

Operating expense components as percent of total operating expenses
Salaries and employee benefits.
Interest on deposits.
. ......... .
Interest on subordinated notes and debentures.
Occupancy expenses ..
Provision for loan loss.
Other operating expenses .................... .

2.6*
-7.9*
-.2*
-.6*
2.1 *
4.0*

3.5*
-7.7*
.8*
.0
.3
3.4*

* Significant at the .05 level.
NOTE: Data for Hispanic-owned banks are for 1980 through 1983.
Data for black-owned banks are for June 1976 through June 1979.
SOURCES OF PRIMARY DATA:
Board of Governors, Federal Reserve System.
Kwast and Black, "An Analysis of the Behavior of Mature Black-Owned Commercial Banks."

competitors. This is important evidence that Hispanicowned banks are using service charges to help cover
the handling expenses of their deposits.
Among the uses of income, there is a high degree of
similarity, and only two categories warrant comment- interest on subordinated notes and debentures
and provision for loan loss. Most Hispanic-owned
banks do not use subordinated debt; consequently, a
comparison here may be misleading. The relatively
high provision for loan loss at Hispanic-owned banks
seems appropriate, considering their higher ratio of net
loan losses to total loans. It is surprising that blackowned banks, which as a group are known for problems with high loan losses, set aside the same proportion of total expenses for these losses as do their
nonminority competitors.
Rates of return

The overall financial performance, reported in Table
C, of Hispanic- and black-owned banks relative to their
respective nonminority competitors is the basis for the
10

conclusions of this research. At Hispanic-owned
banks, higher expenses are offset by higher revenues,
and overall return on assets is insignificantly different
from the nonminority banks. In contrast, the financial
performance of the black-owned banks is significantly
less than their nonminority competitors. The basic
premise that Hispanic-owned banks price loans and
services to cover their higher expenses is supported by
an examination of the more narrowly defined returns.
The gross return on loans is significantly higher'at
Hispanic-owned banks than at nonminority banks,
while tliere is little difference at black-owned banks.
This higher return on loans at Hispanic-owned banks
helps offset the higher rate of net loan loss that is
common to both Hispanic- and black-owned banks.'
Similarly, the return from service charges on total
deposits is also higher at Hispanic-owned banks

2. Net loan losses are loans charged off as uncollectible less
recoveries on loans that were previously charged off.

Federal Reserve Bank of Dallas

Table C
DIFFERENCES IN RATES OF RETURN AND COST
AT MATURE HISPANIC- AND BLACK-OWNED BANKS
RELATIVE TO THEIR NONMINORITY COMPETITORS
Hispanicowned
banks

Blackowned
banks

Difference from
nonm inority-owned
competitors
(Percent)

Item

Income and expenses as percent of total assets
Total operating income .................... .
Net operating income before taxes .............. .
Net income after taxes ..
Total operating expenses.
Average rates of return
I nterest and fees on loans/gross loans ......... .
Interest on Federal Government securities/
total Federal Government securities
Interest on state and local government securities/
total state and local government securities.
Service charges on deposits/total deposits
of individuals, partnerships, and corporations
Average rates of cost
Salaries and employee benefits/total assets
Interest on deposits/time deposits of individuals,
partnerships, and corporations
Net loan loss/total loans
....... .

.8*
-.1
-.2
.8*

-.2*
-.3*
-.2*
.1 *

.8*

-.1t

-1.2*

-.1t

-.3

-.7*

.5*

.1t

.4*

.2*

1.2
.4*

.3*
.5*

* Significant at the .05 level.
t t test not computed.
NOTE: Data for Hispanic-owned banks are for 1980 through 1983.
Data for black-owned banks are for June 1976 through June 1979.
SOURCES OF PRIMARY DATA:
Board of Governors, Federal Reserve System.
Kwast and Black, "An Analysis of the Behavior of Mature Black-Owned Commercial Banks."

relative to nonminority banks, and this return is about
the same for black-owned banks and their nonminority
competitors. The higher return from service charges at
Hispanic-owned banks helps offset the cost of salaries

Economic Review I March 1985

and employee benefits per dollar of assets. This rate of
cost is significantly higher at both black- and H ispanicowned banks.

11

FDIC Settlement Practices
and the Size of Failed Banks
Eugenie D. Short
Assistant Vice President and Senior Economist
Federal Reserve Bank of Dallas

Throughout most of the post-World War II period,
the annual number and the size of failed banks
have been small. In the recent past, both factors
have changed. Failed banks in 1984 totaled 79, the
largest number of failures since the Great Depression. Moreover, one of the nation's largest banks
required a major Federal Deposit Insurance Corporation rescue package to continue operations.
During the past three years the average failure
rate for banks increased to 0.37 percent, compared
with 0.07 percent for the entire 1946-84 period. As
shown in Chart 1, the failure rate for banks is now
similar to the rate in the early years following the
introduction of FDIC insurance in 1934.' Although
the bank failure rate is still exceptionally low compared with the failure rate for other industries, the
magnitude of the increase in bank failures has
heightened concerns about the strength of the U.S.
banking system.

The views expressed are those of the author and do not necessarily
reflect the positions of the Federal Reserve Bank of Dallas or the
Federal Reserve System. Special recognition is given to W. Michael
Cox for helpful comments on this paper and to Franklin D. Berger,
Keith R. Phillips, and Phyllis C. Katsigris for research and
programming assistance.

12

A number of different reasons have been given
for the recent rise in the number and size of failed
banks. The three most notable are the severity of
the last recession, the sharp adjustment from a highinflation environment to a low-inflation environ-

1. I n this article, banks closed because of financial difficulty and
those requiring major rescue packages to continue operations
are treated as failures. The failure rate in Chart 1 was
calculated by dividing the annual number of failed banks by
the total number of banks operating in the same year. Data for
1934 to 1983 on commercial banks and mutual savings banks
closed because of financial difficulty were obtained from the
1983 Annual Report of the Federal Deposit Insurance Corporation, Table 122, page 53. Data on banks closed in 1984
because of financial difficulty were obtained directly from the

FDIC.
Major rescue packages were arranged by the FDIC for First
Pennsylvania Bank, Philadelphia, in 1980 and for Continental
Illinois National Bank and Trust Company of Chicago in 1984.
Severe financial difficulty at Seattle-First National Bank in
1983 induced the Federal Reserve Bank of San Francisco to
facilitate an out-of-state merger arrangement between
BankAmerica Corporation and Seafirst Corporation. First Pennsylvania Bank, Continental Illinois Bank and Trust, and SeattleFirst National Bank are all treated as failures in this article.
Difficulties resulting from excessive exposure to interest rate
risk have also required special arrangements for some mutual

Federal Reserve Bank of Dallas

Chart 1

Bank Failure Rate
FAILED BANKS AS PERCENT OF ALL BANKS
.55 .---~--------------------------------------------------------~
.50
.45

.40

.35
.30
.25

COMMERCIAL BANKS AND MUTUAL SAVINGS BANKS

.20
.15

.10
.05

o

~~~~L-~~~~~~~L-~~~-L~~~

'36

'40

'44

'48

'52

'56

'60

'64

__L-~~~-L-L~~LJ

'68

'72

'76

'80

'84

SOURCES OF PRIMARY DATA: Board of Governo". Federail{e,erve System
Federal Deposit Insurance Corporation
U S Bureau of the Census

ment, and financial deregulation. 2 I ncreased attention is also being given to the impact of federal
deposit insurance on bank failures.
The FDIC charges a fixed premium for deposit insurance without regard to the riskiness of bank portfolios-an action that insulates banks from the full
cost of incurring risk. By law, the FDIC protects
depositors up to $100,000 at insured institutions.
Deposits in excess of $100,000 are uninsured funds.
But the manner in which the FDIC has settled failed
banks has provided de facto 100-percent coverage
to both insured and uninsured depositors. These full
deposit guarantees have encouraged banks to increase their exposure to risk, thereby increasing the

savings banks to continue operations. Because of the constraints on obtaining information about these arrangements for
individual institutions, mutual savings banks that received
special assistance to continue operations were not included as
failures.
2. For additional discussion of the causes of recent financial
stress, see Eugenie D. Short and Gerald P. O'Driscoll, Jr.,
"Deposit Insurance and Financial Stability," Business Forum 8
(Summer 1983): 10-13.

Economic Review I March 1985

probability of failure. Moreover, the perception that
the holders of uninsured deposits at large banks
receive greater protection than the uninsured
depositors of small banks has provided stronger incentives to large banking institutions to increase
their exposure to risk.
This article focuses on the effect of FDIC policies
on the relative size of failed banks. It is argued that
preferential treatment given to depositors of large
banks has altered the size distribution of bank
failures in this country to a higher proportion of
large-bank failures.

FDIC insurance and bank risk decisions
Federal deposit insurance was authorized by the
Banking Act of 1933 to restore confidence in the
U.S. banking system. But the FDIC was created as
one component of financial legislation, most of
which imposed restrictions on bank activity in an effort to constrain risk taking. Banks were prohibited,
among other things, from underwriting corporate
securities, paying interest on demand deposits, and
paying interest on savings and time deposits in excess of allowed limits. Asset and liability constraints
13

and limits to geographic expansion were intended to
ensure safe banking by reducing competition. As a
result, incentives provided by deposit insurance to
undertake excessive risk were partially offset.
By the middle to late 1960s, however, financial
innovation and technological change initiated a
period of gradual or de facto deregulation. As
regulations constraining risk taking were removed or
circumvented, deposit guarantees provided by the
FDIC became increasingly important in insulating
insured banks from the full cost of incurring additional risk.
Literature on the impact of FDIC policies on bank
risk decisions is extensive. 3 Briefly, the fixed-rate
premium on FDIC insurance provides incentives for
banks to increase their exposure to risk. The higher
yields on riskier investments are not offset by higher
insurance premiums. Similarly, extensive reliance on
purchase and assumption (P&A) transactions to settle failed banks also provides incentives for banks
to increase their exposure to risk.
With a P&A transaction, all nonsubordinated
liabilities, including uninsured deposits, are transferred to an assuming bank. The main benefit of the
P&A settlement is that it avoids interruption in the
availability of funds to all depositors. But this
benefit has also generated a negative side effect- a
sharp reduction in incentives for depositors to
monitor the risk exposure of the banks in which
they place funds.
The availability of federal deposit insurance
reduced the probabil ity of deposit runs at banks,
but it also reduced constraints against risk taking
that would normally be imposed by the holders of
uninsured deposits. Extensive reliance on P&A transactions to settle failed banks plus fixed-rate pricing
of deposit insurance established incentives for
banks to increase their exposure to risk. I n addition,

3. Some examples are John H. Kareken and Neil Wallace,
"Deposit Insurance and Bank Regulation: A Partial-Equilibrium
Exposition," Journal of Business 51 (July 1978): 413-38; John H.
Kareken, "The First Step in Bank Deregulation: What About
the FDIC?" American Economic Review 73 (May 1983, Papers
and Proceedings, 1982): 198-203; Mark J. Flannery, "Deposit
Insurance Creates a Need for Bank Regulation," Business
Review, Federal Reserve Bank of Philadelphia, January/
February 1982, 17-27; and Eugenie D. Short and Gerald P.
O'Driscoll, Jr., "Deregulation and Deposit Insurance,"
Economic Review, Federal Reserve Bank of Dallas, September

1983,11-22.
14

FDIC settlement practices provided the holders of
uninsured deposits at larger banks with a greater
degree of protection than was provided in the case
of smaller banks. The increase in the relative size of
failed banks, including some of the nation's largest
banks, may reflect this policy bias.

Bank failures since FDIC insurance
The primary objective of federal deposit insurance
was to prevent bank runs and consequent failures. If
a failure occurred, deposit insurance would quickly
restore any circulating medium of exchange
destroyed or made unavailable to the public as a
result of failure. A related objective was to provide
financial protection to the small bank creditor.
The increased confidence in the banking system
because of FDIC insurance reduced deposit runs at
U.S. banks and thereby reduced the number of bank
failures. During the 1920s, bank failures had averaged 635 per year. Additional failures during the
three banking crises between 1930 and 1933 brought
the average number of failures in the four-year
period to 2,277 banks per year. Since 1934, however,
the annual number and rate of bank failures have
been relatively low. In the 51-year period, only 890
banks failed (893 if the three large commercial
banks that required major loan-assistance packages
to continue operations are treated as failures). But
nearly 75 percent of these failures occurred in the
first nine years-from 1934 to 1942-and in the last
three years. The annual number of bank failures in
those two periods averaged 54.4 and 57.0, respectively. During the interim period from 1943 to 1981,
the average number of failures was only 5.9 banks
per year.
The rise in bank failures during the past three
years has renewed concerns about the impact of
government deposit guarantees on bank decisions to
incur risk. Increased exposure to risk has been a
significant determinant of bank failures. 4 In addition, the relative size of failed banks is now

4. We recently completed an empirical examination of the impact on bank failures of bank decisions to incur risk. See
Eugenie D. Short, Gerald P. O'Driscoll, Jr., and Franklin D.
Berger, "Recent Bank Failures: Determinants and Consequences" (Paper presented at the Annual Meetings of the
Allied Social Science Associations, Dallas, Texas, 29 December
1984). The hypothesis tested in that paper is that bank failure
is directly related to portfolio decisions made by bank
Federal Reserve Bank of Dallas

Chart 2

Failure Rate, by Bank Size
FAILED BANKS AS PERCENT OF ALL BANKS IN CATEGORY
.60 r-----------------------------------------------------------~

.55
.50

.45
COMMERCIAL BANKS AND MUTUAL SAVINGS BANKS

.40

DEPOSITS GREATER THAN $100 MILLION
---- DEPOSITS LESS THAN $100 MILLION

.35
.30
.25
.20

.15
.10

.05

o
'60

'62

'64

'66

'68

'70

'72

'74

'76

'78

'80

'82

'84

SOLIRCES OF PRIMARY DATA: Board of Governors, Federal Reserve System
Federal Deposit Insurance Corporation.

significantly larger than in any other period on
record. This change in the size distribution of
failures may reflect the unintended impact of
federal deposit guarantees.

Relative size of bank failures
Most bank failures in this country have been small.
This tendency primarily reflects the size distribution
of the U.S. banking system, Of the 15,388 banks
operating in June 1984, roughly 85 percent had less
than $100 million in deposits. Thus, a larger number
of failures at small institutions seems likely. During
the 42 years from 1931 to 1972, no banks with

managers to accept risk. To identify those portfolio decisions,
financial ratios of failed and nonfailed banks are examined.
Probit analysis is used to determine differences in these ratios
in 1964, 1975, and 1982-83. The results of the study indicate
that there are statistically significant differences between
failed and nonfailed banks for several of the financial ratios.
The critical ratios appear to be loans to assets, capital to
assets, core deposits to liabilities, and purchased funds to
liabilities, The results suggest that managerial decisions to accept more risk have played an important role in bank failures,

Economic Review I March 1985

deposits in excess of $100 million failed,S In 8 of the
past 12 years, however, the failure rate for banks
with deposits in excess of $100 million-a size
category often used to differentiate between large
and small institutions-has been greater than the
failure rate for smaller banks (Chart 2).
Using a $100 million size definition for large
banks introduces an inflation bias into this size
comparison. Annual growth in the nation's money
supply raises the dollar level of deposits at all
banks. Hence, the number of banks with deposits in
excess of $100 million tends to increase over time,
The number of banks with deposits in excess of
$100 million increased from an average of 442
banks during 1959-72 to an average of 1,516 banks
in the post-1972 period. Similarly, the proportion of
banks with deposits in excess of $100 million rose
from an average of 3 percent of all banks from 1959

5. In December 1930 the Bank of the United States, with over
$200 million in deposits, failed. At the time, that was the
largest commercial bank failure in U.S. history.

15

to 1972 to 10 percent in the post-1972 period. 6 Thus,
a larger number of failures in the size category over
$100 million was likely in the latter period.
To eliminate this inflation bias, figures on the
relative size of failed banks were calculated by
dividing the average size of failed banks each year
by the average size of all banks in the same year. 7
This relative size variable, 5, is plotted in Chart 3 for
1921 through 1984. Statistical analysis of these data
indicates that the size distribution of failed banks
changed markedly in the mid-1970s. The procedure
used was to separate the 64-year period into two
subperiods of size nand (64 - n), where n = 2, ... , 62.
Within each subperiod the mean and standard
deviation of the relative size variable were
calculated. An F test was applied to determine
whether the standard deviations of the two
subperiods could be treated as identical. The appropriate t statistic for each subperiod was
calculated to examine differences in the means of
the two subperiods. The root mean square error
over the entire sample was minimized by breaking
the period at n = 52, which corresponds to 1972. The
value of the t statistic for the difference-in-means
test at that breaking point was 3.64. A critical t
value of 2.91 is required for significance at the
99-percent confidence level; hence, the null
hypothesis that there is no difference in the means
in these two periods was rejected. In addition,
results of F tests for this sample indicated that the
standard deviation of 5 is significantly higher in the
1973-84 period than in the 1921-72 period.
This article offers an explanation for the general
increase in the relative size of failed banks over the

6. Data on deposits of individual U.S. banks are available back
to 1959 from the FDIC's Consolidated Report of Condition
(bank Call Report data).

7. The average size of failed banks was calculated by dividing
annual data on total deposits of failed banks by the total
number of failures during the year. The average size of all
banks was calculated by dividing annual data on total deposits
of all banks by the total number of banks. Annual data on the
number of failures and the total deposits of failed banks from
1921 to 1933 were obtained from U.S. Bureau of the Census,
Historical Statistics of the United States, Colonial Times to
1970, Bicentennial Edition, pt. 2 (Washington, D.C: Government Printing Office, 1975). Comparable data for 1934 to 1983
were obtained from the 1983 Annual Report of the FDIC Data
for 1984 were obtained directly from the FDIC The three large
commercial banks that required rescue packages were in16

extended 1921-84 period. It does not provide
guidance about when a break in the sample should
occur, nor does it attempt to explain the obvious
sharp jumps in the relative size of failed banks in
1973, 1974, and 1980 (Chart 3). I n those three years
the relative size of failed banks increased to 3.24,
7.22, and 5.97, respectively. The increase in each
year reflected the failure of one of the nation's
largest banks. In 1973 the United States National
Bank, San Diego, California, with $932 million in
deposits, was closed. In 1974, Franklin National
Bank, New York City, with $1.4 billion in deposits,
was closed. In 1980 the First Pennsylvania Bank,
Philadelphia, with $5.3 billion in deposits, was a
problem bank.
Two of those outlier years-1974 and 1980-were
during recessions. Banks, like other firms, are more
likely to fail during, or in the aftermath of, periods
of econom ic recession. To control for the impact of
recessions on the relative size of failed banks, the
difference-in-means test was applied to a smaller
sample that only included the relative size of failed
banks in nonrecession years.8 The adjusted sample
included 46 observations. It was examined to determine whether a statistically significant increase
occurred in the relative size of failed banks in the
post-1972 period even if failures during recession
years were excluded from the sample.
The same procedure was used. The 46-year period
was separated into two subperiods of size nand
(46 - n), where n = 2, ... , 44. The root mean square
error was minimized by breaking the period at
n = 38, which corresponds to 1972, the same breaking point as in the larger sample. The value of the t
statistic for the difference-in-means test at the 1972

eluded as failed banks.
Annual data on the number of banks and the total deposits
of banks from 1921 to 1970 were obtained from U.S. Bureau of
the Census, Historical Statistics, Colonial Times to 1970. Comparable data for 1971 to 1984 were obtained from the Reports
of Condition for banks.

8. Recession years were determined by using the dates of peaks
and troughs of U.S. business cycles for 1854 through 1982 from
CITIBASE: Citibank Economic Database, app. A (New York:
Citibank). If six months or more in a single year were in a
period of economic downturn, the year was considered a recession year. With that definition, the following were excluded
from the sample: 1921, 1923, 1924, 1927, 1930, 1931, 1932,
1937,1938,1945,1949,1953,1960,1970,1974,1980,1981, and
1982.

Federal Reserve Bank of Dallas

Chart 3

Relative Size of Failed Banks
SIZE VARIABLE, S

B.O

r---------------------------------------------------------------,

1'.0

6.0

S =

:\VERAGE DEPOSITS OF FAILED BANKS
AVERAGE DEPOSITS OF ALL BANKS

5.0
4.0
3.0
2.0

COMMERCIAL BANKS AND MUTUAL SAVINGS BANKS

1.0

'20

'24

'28

'32

'36

'40

'44

'48

'52

'56

'60

'64

'68

'72

'76

'80

'84

SOURCES OF PRIMARY DATA: Board of Covernor" federal Reservp System
Federal Deposit

InSUfdnCf'

Corpofdtlon

US Bureau of the Census

breaking point was 3.51, again significant at the
95-percent confidence level. These results also indicate that the relative size of failed banks was
significantly larger in the post-1972 period than in
earl ier years.
The change in the size distribution of failed banks
has, among other things, renewed concerns about
the impact of federal deposit guarantees on bank
behavior. The FDIC acknowledges that extensive use
of purchase and assumption transactions has altered
incentives for bank depositors to monitor the risk
exposure of the banks in which they hold funds. The
agency has also identified the negative consequences resulting from this practice. In its published
manuscript about the need for deposit insurance
reform, the agency notes that the preferential
treatment given to larger banks has encouraged
depositors to place funds, at virtually no risk, in
these larger institutions. 9 This preferential treatment
not only requires smaller institutions to pay proportionally more for uninsured deposits but also has
removed incentives for the holders of uninsured
deposits at large banks to evaluate the financial
Economic Review I March 1985

condition of these banks.
As the use of assisted deposit assumptions has
become more common and increased numbers of
depositors and investors continue to be shielded
from losses in large banking organizations, the
public's perception of the relative safety of funds
appears to have become altered, Many believe
that no large American bank will be paid off even
if it were allowed to fail, and have acted accordingly. In addition to driving large depositors from
smaller to larger banks, this growing perception of
almost absolute safety of funds in large institutions is having the effect of removing the consideration of bank risk from business decisions.
(FDIC, Deposit Insurance in a Changing Environment, chap. 3, p. 1)

9. For a thorough discussion of the views of the FDIC on the impact of extensive reliance on P&A transactions to settle bank
failures, see Deposit Insurance in a Changing Environment: A
Study of the Current System of Deposit Insurance Pursuant to
Section 712 of the Garn-St Germain Depository Institutions Act
of 1982, Submitted to the United States Congress by the Federal
Deposit Insurance Corporation (Washington, D.C.: Federal
Deposit I nsurance Corporation, 1983).

17

FDIC settlement practices: the large-bank bias
Since 1934 the FDIC has settled 407 bank failures
with P&A transactions and 340 failures with some
form of deposit payoff.'o Virtually all large banks,
however have been settled with P&A transactions.
Four f~ctors are usually cited as important determinants of the FDIC's bias in favor of merger settlements for large banks." First, the premium price
paid to acquire the failed bank's charter is ordinarily large enough to reduce the estimated cost of
a merger below that of a deposit payoff. This factor
is also generally important for small banks as well.
Second, larger banks have a greater proportion of
their liabilities in uninsured deposits; hence, a
deposit payoff tends to be more disruptive at larger
institutions. Third, a deposit payoff is more difficult
with larger banks because it can involve a substantial immediate cash outlay by the FDIC. Although
the FDIC may recoup a significant portion of this
outlay after the failed bank's assets have been liquidated, concerns that the public's confidence in
the banking system would be shaken if the agency
were required to reduce its reserve base by a large
initial payoff tend to bias the decision to favor a
merger solution. Finally, the agency is concerned
about negative spillover effects for sound banks
from the closing of unsound banks, especially large
ones.
Before 1982, all large banks-those with deposits
in excess of $100 million-that were closed because
of financial difficulty were settled with P&A transactions. In 1982 the FDIC's settlement of Penn Square
Bank of Oklahoma City generated some uncertainty
about the agency's policy for settling large failures.
At the time of failure, the bank had $470 million in
10. Data on the number of banks settled as deposit assumptions
and deposit payoffs from 1934 to 1983 were obtained from
the 1983 Annual Report of the Federal Deposit Insurance Corporation, Table 125, page 56. Unpublished data for 1984 were
obtained directly from the FDIC. The 11 banks settled with
modified payouts in 1984 were treated as deposit payoffs.
(The banks are listed in the table accompanying this article.)

11. For a thorough discussion of the FDIC's preference for using
P&A transactions to settle failed banks, see Paul M. Horvitz,
"Failures of Large Banks: Implications for Banking Supervision and Deposit Insurance," Journal of Financial and Quantitative Analysis 10 (November 1975): 589-601; and Barbara A.
Bennett, "Bank Regulation and Deposit Insurance: Controlling
the FDIC's Losses," Economic Review, Federal Reserve Bank
of San Francisco, Spring 1984, 16-30.

18

deposits. Despite its relatively large size, a deposit
payoff was used to settle Penn Square. The holders
of uninsured deposits did incur financial loss.
Penn Square involved litigation problems that
precluded assumption by another bank. But t~e
treatment given to depositors of large banks since
the Penn Square settlement has strengthened the
perception that the holders of uninsured deposits at
large banks will be treated like fully insured
depositors. In 1983 the FDIC arranged a P&A. transaction to settle the First National Bank of Midland,
which had $575 million in deposits. More recently,
the inability to arrange an independent merger for
Continental Illinois led the FDIC to arrange a major
rescue package for the nation's seventh largest
bank.'2
The magnitude of potential losses to the holders
of uninsured deposits at Continental, coupled with
concerns about the impact on other financial institutions, induced the FDIC to offer full protection to
all general creditors before c1osure.'3 Subse~uently,
during congressional testimony on the Continental
settlement, C. Todd Conover, Comptroller of the
Currency, testified that the Federal Government
would not allow the nation's 11 largest banks to
fail.'4 That statement provided a verbal guarantee
of 100-percent coverage to all depositors and
general creditors of the largest banks. But the verbal
guarantee merely made explicit a perception th~t
previous policy practices had already led de~osltors
of large banks to have; that is, holders of uninsured
deposits at large banks would be treated like fully
insured depositors.
The explicit guarantee provided to the nation's

12. According to the definition of bank failure used in this article, Continental Illinois is considered a failed bank. Legally,
however, the bank was not closed.

13. 'The FDIC's response to concerns about the problems at Con:tinental Illinois represented a departure from the norm in
dealing with severe financial difficulty at large banks. Unable
to arrange a merger, the FDIC implemented a rescue plan
under which the agency became the major stockholder of the
bank holding company. The FDIC accepted the equivalent of
80 percent of Continental Illinois Corporation's equity, and it
assumed as much as $4.5 billion in problem loans at the lead
bank, Continental Illinois National Bank and Trust Company.

14. For additional discussion of Conover's testimony, see Tim
Carrington, "U.S. Won't Let 11 Biggest Banks in Nation Fail,"
Wall Street Journal, 20 September 1984.

Federal Reserve Bank of Dallas

FAILED BANKS SETTLED IN 1983 AND 1984
BY "MODIFIED DEPOSIT PAYOFF" APPROACH
Deposits
(Millions
of dollars)

1983

Union National Bank of Chicago, Chicago, Illinois.
Atkinson Trust and Savings Bank, Atkinson, Illinois ..

$24.5
18.9

1984

Seminole State National Bank, Seminole, Texas.

41.3
153.3

Heritage Bank, Anaheim, California.
Security National Bank of Lubbock, Lubbock, Texas.

40.6

Gamaliel Bank, Gamaliel, Kentucky

21.6

United of America Bank, Chicago, Illinois.

29.0

West Coast Bank, Los Angeles, California.

154.8

First National Bank, Snyder, Texas.

15.2

The National Bank of Carmel, Carmel-By-The-Sea, California.

70.8

First Continental Bank & Trust Company of Del City, Del City, Oklahoma.

92.3

Stewardship Bank of Oregon, Portland, Oregon.
The Dayton Bank & Trust Company, Dayton, Tennessee.

5.4
47.5

SOURCE: Federal Deposit Insurance Corporation.

largest banks was in sharp contrast to the treatment
of several smaller banks that were settled with
modified payoffs in 1983 and 1984. (See the table.)
The uninsured depositors at those banks did incur
financial loss, with initial payouts ranging from 35
to 75 percent of the dollar value of their holdings.
The FDIC's purpose in using modified payoffs at
those 13 banks was to test procedures that might
reestablish incentives for holders of uninsured
deposits to monitor the risk exposure of their banks,
incentives that had been eliminated by full insurance guarantees. But the FDIC did not apply the
practice uniformly at large and small banks.
Although two banks with deposits just over $150
million were settled with modified payoffs,
depositors at the largest institutions, including First
National Bank of Midland and Continental Illinois,
received full plrOtection.
Criticisms against the preferential treatment given
to larger banks are likely to constrain the FDIC from
using modified payouts to settle failed banks. Further attempts to reintroduce pricing constraints on
risk taking by providing less than full protection to
Economic Review I March 1985

the holders of uninsured deposits will probably be
curtailed, at least in the near future.
FDIC insurance, deposit pricing,
and information about bank risk
The incentive mechanism established by full deposit
insurance guarantees has inhibited the flow of information that would normally be produced by the
frequent adjustments of depositors to changes in
the risk exposure of individual banks. With full insurance guarantees, bank depositors have little
incentive to adjust to incremental changes in the
risk exposure of individual banks. Hence, bank
depositors generally do not require large interest
rate premiums to keep funds at riskier institutions.
The full (implicit) guarantee provided to banks
has altered both the timing and the magnitude of
price and quantity adjustments in the bank deposit
market. By sharply reducing the probability that
bank depositors would incur any financial loss from
bank failures, FDIC policies removed incentives for
uninsured depositors to make continuous price and
quantity adjustments on the basis of new informa19

tion on the risk exposure of individual banks. It is
only after problems of significant proportions arise
that large and potentially unstable price and
quantity adjustments develop in the bank deposit
market. The pattern of information flow is quite
different from that which is normally produced by
continuous price and quantity adjustments.
The 100-percent insurance protection provided to
depositors is de facto, not de jure. To eliminate any
chance of incurring financial loss-the magnitude of
which can be large-uninsured depositors tend to
remove funds from troubled institutions before failure. The structure of interest rates in the market for
uninsured bank deposits can then become sharply
graduated, or "tiered," with risk premiums reflecting
the low but positive probability that uninsured
depositors will incur financial loss. As the financial
condition of a troubled bank worsens, large
outflows of uninsured funds can and do occur. But
these large deposit outflows only develop after
serious problems have been identified. At this point,
the bank may indeed undertake actions to reduce
its exposure to risk. If taken earlier, these actions
might have prevented the very problems the bank
faces. By the time corrective actions are taken, however, the poor quality of the troubled bank's asset
portfolio limits the bank's ability to reduce its exposure to risk. In this way, removal of constraints
against risk taking that would normally be imposed
by the holders of uninsured deposits may have increased the likelihood of failure at large insured
banks.

Conclusion
Full insurance coverage of bank deposits has encouraged banks to increase their exposure to risk
and thereby has increased the probability of bank
failures. Risk premiums on bank deposit rates no
longer provide an effective constraint against risk

20

taking. Unless a bank, or group of banks, is in
serious danger of failing, the holders of uninsured
deposits do not require large interest rate premiums
to keep their funds at the riskier institutions. If a
bank experiences serious earnings difficulties,
outflows of funds from the troubled bank induce it
to adjust deposit rates upward in accordance with
revised assessments of the bank's risk exposure. But
these quantity and pricing adjustments take place
after sizable problems have been identified - problems sufficiently large to raise questions about the
bank's solvency.
In sharp contrast to the way creditors normally
impose constraints on risk takers, bank depositors
no longer provide effective restraints against risk
taking. When banks fail, stockholders and subordinated creditors incur financial loss, but losses to
depositors have been infrequent and small. As a
result, depositors-the principal creditors of
banks - have I ittle reason to be concerned about the
financial condition of their banks.
The de facto 100-percent protection furnished
bank depositors appears to have reduced differentials on deposit rates offered by banks with different
risk profiles to small fractions of those that would
normally be required to compensate for risk differences. Thus, banks generally, and large banks in
particular, operate in a unique environment. Their
principal creditors-depositors- impose little or no
constraint on risk decisions.
The elimination of risk premiums on bank
deposits commensurate with risk differences among
banks increases the probability that sizable errors
will be made in the actual risk exposure of banks.
Similarly, the bias in favor of large banks increases
the probability of large-bank failures. The rise in the
bank failure rate and in the relative size of failed
banks in the post-1972 period indicates that both
predictions have already occurred.

Federal Reserve Bank of Dallas

FEDERAL RESERVE BANK OF DALLAS
STATION K, DALLAS, TEXAS 75222
ADDRESS CORRECTION REQUESTED

BULK RATE
U.S. POSTAGE

PAID
PERMIT NO. 151

The Economic Review is published by the Federal Reserve Bank of Dallas and
will be issued six times in 1985 (January, March, May, July, September, and
Nqveltlber). This publication is sent to the mailing list without charge, and additional copies of most issues are available from the Public Affairs Department,
Federal Reserve Bank of Dallas, Station K, Dallas, Texas 75222. Articles may be
reprinted on the condition that the source is credited.