The full text on this page is automatically extracted from the file linked above and may contain errors and inconsistencies.
B a n k F ra g ility : P e r c e p t io n a n d
H is to ric a l E v id e n c e
George G. Kaufman
Working Papers Series
Issues in Financial Regulation
Research Department
Federal Reserve Bank of Chicago
September 1996 (WP-96-18)
FEDERAL RESERVE BANK
OF CHICAGO
Draft
Septembers 1996
B a n k F ra g ility : P e r c e p t i o n a n d H is to r ic a l E v id e n c e
George G. Kaufman*
(Loyola University Chicago and Federal Reserve Bank of Chicago)
The fragility of banks and the financial system has been a persistent source of public
concern throughout modem history. Anna Schwartz (1995, p. 445) defines financial fragility
as a state in which "the ability of the financial system to withstand economic shocks is
weak." It follows that it is feared that fragility will result in failure. Thus, if the financial
system were more fragile than other sectors of the economy, we should expect the failure
rate of financial institutions and markets to be higher than that in other sectors. Commercial
banks are generally viewed as both the most important and the most fragile of all financial
institutions. Moreover, their failure is viewed as having the most adverse effect on the
economy (Minsky, 1995). As a result, much public policy in banking in all countries is
directed at coping with the fragility of banks in order to reduce the probability of their failure
(Benston and Kaufman, 1995). But the recent history of widespread bank failures in
'Earlier drafts of this paper were presented at the Federal Reserve Bank of Chicago and the Western
Economic Association meetings. I am indebted to the participants at these sessions, particularly Ed Stevens
(Federal Reserve Bank of Cleveland), for their helpful comments. Kristine DeCastro provided research
assistance. The views expressed in this paper are those of the author. They do not necessarily represent the
position of the Federal Reserve Bank of Chicago or the Federal Reserve System.
2
almost all countries-differing widely in economic, social, cultural, and political structurecasts doubt about the efficacy of these prudential policies.1 This paper examines:
- The degree of bank fragility relative to the perception of their fragility,
- Why banks are viewed as more fragile than other firms,
- The reasons that the greater fragility has generally not resulted in greater failure,
- Changes in the indicators of bank fragility through time,
- The relationship of bank fragility and the bank safety net, and
• The implications of bank fragility for bank failures in the future.
Because of their availability, the data analyzed in this paper are only for the United States.
The primary purpose of this paper is to challenge a number of widely held
conventional wisdoms about the fragility of banks and the need for and effectiveness of
much of the extant prudential bank regulatory policies that appear to be both costly and
even counterproductive. As such, the paper should be viewed as a "bugle call" to motivate
and encourage additional research in this important public policy area.
1.
Why B anks A re V iewed A s More F ragile T han Other F irms
A review of the historical and contemporary literature suggests that banks are
perceived to be relatively more fragile than other firms primarily because of three stylized
facts. They have:
1. Low capital-to-assets ratios (high leverage),
1A recent survey found that 131 of the 181 IMF m em ber countries have reported serious banking problems
and failures in recen t y e a rs (Lindgren, Garcia, and Saal, forthcoming 1996).
- *
*
* t
3
2. Low cash-to-assets ratios (fractional reserve banking), and
3. High demand deposits (debt)-to-total deposits (debt) ratios (high potential for run).
The first factor reduces the ability of a bank to absorb losses and avoid insolvency. A given
adverse shock, say, a credit loss (default) or an interest rate mismatch loss, is thus more
likely to drive a bank into insolvency than it is to drive a nonbank firm that has a higher
capital-to-assets ratio into insolvency. Small holdings of cash increase a bank's need to sell
earning assets in order to meet deposit outflows. To the extent the bank's earning assets
are not very liquid, fire-sale losses may be generated when they are sold. This reason
takes on greater significance the more illiquid or opaque are the bank's assets. A high
proportion of demand debt intensifies the effects of the low cash ratio as a large
unexpected withdrawal of deposits would force a fast sale of assets and potentially
increase the bank's fire-sale losses. Thus, depositor runs on banks, which almost by
definition occur unannounced, are likely to ignite liquidity problems, which, in turn, are
widely considered to lead to solvency problems because of the banks' low capital ratios.
Although each of these three factors by themselves are perceived to make banks
fragile, the interaction among the three produces an even higher degree of fragility that
gives rise to heightened concern. However, the interrelationship of these three factors
permits substitutions among them so that various combinations of the three can yield the
same degree of overall fragility. Empirically, banks, on average, have substantially lower
capital-asset ratios and higher (short-term) demand-to-total-debt ratios than other firms,
but they have substantially higher cash-to-asset ratios. The values of the cash-to-asset and
4
capital-to-asset, ratios for banks and nonfinancial firms and demand to total deposits for
banks for selected periods from 1870 through 1992 are shown in Table 1.
2.
A ctual vs . Perceived B ank F ragility
The fragility of banks is viewed as a public policy concern both because it is feared
that it may cause a larger number of failures and because of the substantial adverse
consequences that bank failures are widely perceived to cause in other sectors of the
economy. And, at least for the United States, the annual rate of bank failures has been
higher than the rate of nonbank failures, on average, since the end of the Civil War. As can
be seen from Table 2, the annual average bank failure rate from 1870 through 1994 was
1.03 percent, while that for nonbanks was 0.78 percent.2 The annual variability in bank
failures is also considerably greater than that of nonbank failures. As a result however,
in most years, the bank failure rate was substantially below that of nonbanks. Indeed, if
one excludes the unusually large number of bank failures during the Great Depression
from 1930 to 1933, the average bank failure rate for the remaining 120 year period would
be lower than that of nonbanks (0.65% vs. 0.76%). Likewise, in the period from 1870 to
1913, before the introduction of the first federal government safety net under banks in the
^ e s e figures are likely to overstate the bank failure rate somewhat as the number of banks reported
before 1896, but not the number of failures, appears to substantially understate the actual number (Benston,
et al, 1986). This causes the reported failure rate before 1896 to be overstated. On the other hand, temporary
bank suspensions are not included as failures.
5
form of the Federal Reserve's lender of last resort facilities in 1914, the failure rate for
banks was also lower than that for nonbanks (0.91% to 1.01%).3,4
Moreover, the lower bank failure rate in the pre-safety net period and the only
moderately higher overall bank failure rate occurred despite the fact that until very recently
banks in the U.S. have been prohibited from reducing their risk of failure as much as they
may have liked by government legislation and regulation that restricted them from
diversifying fully across either geographic (unit banking) or product lines. Indeed,
particularly in the earlier years, a stranger to the U.S. may rightfully have believed that the
U.S. banking system was almost designed to maximize failure.
Thus, there appears to be an inconsistency between the widely held perception of
substantial bank fragility based on the values of three financial fragility ratios and the
historical evidence of only a slightly, if any in many periods, higher failure rate than for
nonbanks. The historical bank failure rate suggests little actual fragility relative to other
firms. The next section considers why the perceived high fragility did not translate into an
unusually high failure rate.
’ Using 1914 for the introduction of the initial bank safety net and as the appropriate dividing year between
the pre- and post-safety net periods is supported by an analysis by Federal Reserve Chairman Alan
Greenspan, who argued that
Overtime, societies concluded that [bank] leverage and intermediation were essential to economic
performance, but also that some bank failures could have unacceptable economic costs. In response,
central banks were created and were accorded new responsibilities, and what we now call prudential
regulation evolved. In the United States, these initiatives took the shape of the creation of the Federal
Reserve in 1913 after several financial panics in the late 19th and early 20th centuries, and of federal
deposit insurance and a broadened role for bank supervisors in the 1930's....This expanded role of
governments, central banks, and bank supervisors implies a complex approach to managing and even sharing
the risks of failure between governments and privately owned banks (Greenspan, 1996, p. 5).
"Because of the high variability in the bank failure rate, none of the differences between the bank and
nonbank failure rates for any of these periods are statistically significant at the 5 percent level.
6
3.
R easons Why A pparent F ragility May Not T ranslate Into G reater F ailure
Fragility per se does not imply breakage or, in the case of firms, failure. Rather, it
implies "handle with care." Thus, the breakage rate for fragile glass, porcelain, and even
economists' egos need not be unusually high if these objects are handled correctly. Only
the potential for breakage is automatically greater. Similar for banks. The numerical values
of the three fragility ratios signal that banks need to be handled with care in order to avoid
failure. And that is what the market appears to have done. In the period before the
establishment of the safety net under banks in the United States, first in 1914 with the
establishment of the Federal Reserve lender of last resort facilities and then reinforced by
the introduction of deposit insurance by the FDIC in 1934, there was relatively little
government prudential regulation of banks and much market regulation.5-6As noted earlier,
the average annual failure rate for banks before 1914 was slightly lower than that for
nonbanks. Moreover, both the average annual bank failure rate and the annual standard
deviation were lower than afterwards (0.91% vs. 1.09% and 1.00% vs. 3.32%), when
5Th e third part of the modem safety net in the U.S. is finality by the Federal Reserve in payments
settlements over Fedwire. It became important in the 1960s, when technology permitted real-time large value
fund transfers during the day rather than only dayend transfers. Because the Fed requires settlement only at
dayend, intraday or daylight overdrafts became possible. Because the Federal Reserve guarantees all fund
transfers at the time they are made (provides finality), it assumes the risk of default if settlement cannot be
made by the paying bank at dayend. T o stimulate use of Fedwire, the Federal Reserve did not restrict or price
such overdrafts until the late 1980s. Even since then, the charges have been well below what private market
would have charged for such credit
6Nonprudential regulations were more common in this period. For example, restrictions on the number and
location of banking offices have always been specified in federal and state legislation and restrictions on
product powers are specified in federal and state charters. These restrictions were imposed primarily to
reduce competition and increase revenues to the government rather
than for prudential purposes (Wallis, Sylla, and Legler, 1994 and Sylla, Wallis, and Legler, 1995). Restrictions
on geographic and product powers of bank holding companies, however, did not originate until 1956. For an
analysis of the justification for bank regulation, see Benston and Kaufman, 1996.
7
government prudential regulation increased in importance as a source of discipline on the
banks and market regulation declined.7
Before the establishment of the safety net, the market not only permitted the observed
numerical values of the three fragility ratios that were perceived to make banks fragile to
continue, but actually determined their values through depositor and other customer
discipline, presumably after taking account of the risks of failure. Thus, the market did not
appear to view these values as particularly fragile or dangerous relative to other firms. Why
may the market have permitted these "fragile" values before 1914?
The market appears to have established the lower private capital ratios for banks than
nonbanks before 1914 for three reasons. (Data on balance sheet accounts and capital
ratios for nonfinancial firms as a whole are difficult to obtain before 1926. But capital ratios
for selected nonfinancial industries for the earlier period indicate capital ratios well in
excess of 50 percent. Capital ratios for nonfinancial firms for selected years from 1902 to
1992 are shown in Appendix Tables A1 and A2.) One, the reported capital ratios
understate the actual capital ratios. To pay losses to depositors at failed institutions,
shareholders of all national banks and of state banks in some states were subject to
double liability in which the shareholders were liable in case of failure beyond the value of
their stock holdings in an amount up to the initial par value of the securities. Payments
equal to about one-half of the stated liability were, on average, collected from shareholders
at failed national banks to repay losses to depositors (Macey and Miller, 1992). But even
7A s is discussed later, although the increase in the annual mean bank failure rate is not statistically
significant at the 5 percent level, the decline in the nonbank failure rate between these two periods is
statistically significant
8
after adding this sum to the banks' capital, their capital ratios were still below that of
nonbanks. Moreover, banks tend to have greater off-balance sheet assets than nonbanks.
This causes their reported capital to on balance sheet assets to overstate their overall
capital ratios.
Two, on average, depositor creditors suffered substantially smaller relative losses at
failed banks than did creditors at failed nonbanks firms (Kaufman, 1994). The smaller
losses appear to reflect a quicker failure resolution process, in part attributable to quick
suspensions of banks by regulators when depositor runs produced liquidity problems and
sufficiently large fire-sales losses to bring about possible insolvency. Regulators
suspended the bank's operations until they could determine whether the institution had
sufficient capital to reopen or could raise the appropriate amount quickly. This gave banks
less opportunity to operate while insolvent and from ongoing operations as well to continue
to generate losses to "gamble for resurrection." Bankruptcy courts in the U.S., to which
banks are not subject, apparently gave nonbank firms greater opportunity to do both. Thus,,
depositors and other creditors were willing to permit banks to operate on a smaller capital
base.
Three, the losses from fire-sales of assets may not have been perceived as very large
and well within the loss absorbing capabilities of the low capital ratios maintained by the
banks. Even when such losses were greater than a bank's capital they were, as noted
above, smaller than those experienced by creditors of insolvent nonbanks. This suggests
that the market believed that, at least, some portion of the banks' earning assets were not
as illiquid or opaque as is sometimes suggested, even in the period before banks were
9
operated much beyond local markets and technology increased the marketability of loans
through better information and means like securitization. The lower capital ratios before the
introduction of the safety-net suggest that the market viewed banks to be less rather than
more fragile than other firms and/or subject to smaller creditor losses when they did foil.
The cash held by banks relative to their assets is considerably greater in more recent
years than that held by nonfinancial firms. It is unlikely that this was reversed in earlier
years. In large part, the higher cash ratios of banks are likely to reflect reserve
requirements imposed on banks by law or regulation. It is unlikely that ceteris paribus, the
cash ratio contributed greatly, if at all, to the perceived fragility of banks.*
The implications of the higher ratio of demand to total debt for perceived fragility may
be interpreted in two opposite ways. On the one hand, these deposits may be withdrawn
quickly and without notice and force the hurried sale of earning assets, possibly at a loss.
This would increase perceived fragility. On the other hand, particularly before government
deposit insurance, the ability to remove their funds quickly increases the effectiveness of
monitoring by informed bank depositors by permitting them to exert more effective market
discipline (Flannery, 1994; Calomiris and Kahn, 1991). That is, the threat of a run, as
opposed to a run itself, is a powerful source of market discipline that operates to stabilize
banking and reduce its fragility (Kaufman, 1988). Thus, one cannot say without additional
evidence whether a high demand-to-total deposit ratio contributes to bank fragility or
stability.
*The higher cash-to-asset ratios may explain part of the lower capital and higher demand deposit ratios.
10
4.
C hanges
in
Bank F ragility C onditions T hrough T ime
and the
E ffects
of the
S afety Net
Changes in the economic environment through time may be expected to change the
values of all three indicators of bank fragility. Two major economic developments that may
be expected to have affected the values of the indicators are 1) the introduction of the bank
safety net and 2) advances in communications and computer technology. Table 1 shows
that the values of all three of the banks' fragility measures-capital-to-asset, cash-to-asset,
and demand-to-total deposits-have declined through time.
With infrequent exception, these ratios and changes are consistent with what theory
would predict. (These hypothesized effects are summarized in Table 3.) The introduction
of the government safety net, which, until the enactment of FDICIA in 1991, tended to be
poorly designed and underpriced, particularly after the introduction of deposit insurance
with its design flaws of nonrisk related premiums and weak regulatory intervention and
closure rules, ceteris paribus, should lower the reported capital ratios of banks by
permitting them to substitute public capital for private capital and encouraging moral hazard
behavior on the part of shareholders (Kane, 1989a and b; Benston and Kaufman, 1988;
and Kaufman, 1995).' Indeed, the 1994 Annual Report should also lower the banks' cash
of the FDIC drive this point home when it states that “the FDIC remains today the symbol
9ln addition to underpriced deposit insurance premiums, the Fed frequently extended credit to insolvent
and near-insolvent banks at lower interest rates than the market would have charged and did not charge for
or limit daylight overdrafts in clearing settlement until the late 1980s (U .S . House of Representatives, 1991).
Thus, Fed practices likely encouraged both moral hazard and agency problems.
Th e current emphasis on public rather than private capital as a measure of bank financial strength is
reflected in a recent statement by the FD IC that "the FD IC remains today the symbol of banking confidence”
(Federal Deposit Insurance Corporation, 1995, p. 35).
11
of banking confidence” (FDIC, 1995, p. 35). The introduction of the underpriced safety net
should also lower the banks’ cash ratios to the extent permitted by reserve requirements
both by encouraging moral hazard behavior and by providing for an alternative source of
liquidity through the Fed's discount window.
The direction of the impact of the safety net on the demand to total deposit ratio is
less certain. On the one hand, to the extent the safety net reduces the need for depositor
monitoring, it should reduce the ratio. But, on the other hand, to the extent the safety net
permits greater fragility to exist without a widespread perception or concern over increased
failure or losses to depositors, it would encourage an increase in the ratio.
Advances in communications and computer technology, ceteris paribus, should
increase the liquidity of bank earning assets through better information and securitization
and permit banks to reduce their risk exposures through additional product-type and
geographic diversification. In addition, these should advances permit quicker, cheaper, and
more effective bank monitoring by depositors. These forces should in turn reduce the need
for banks to hold both private capital and cash, subject to any reserve requirement
constraint On the other hand, however, to the extent that these advances have permitted
banks to engage in a greater volume of accounting "off-balance sheet" activities relative
to accounting "on-balance sheet" activities, the reported capital to only on balance sheet
assets may increase, even though the capital to overall economic balance sheet asset ratio
is constant or even declining slightly. To the extent that the advances in technology have
also permitted depositors to develop closer substitutes for demand deposits, e.g., money
12
market funds, the demand for bank demand deposits should be reduced and the demand
to total deposit ratio decline.
Thus, the observed declines in the capital and cash ratios are generally consistent
with both the introduction of the safety net and advances technology explanations. The
decline in the reported capital ratio reflects the greater combined impact of the substitution
of public for private capital and a greater decrease in perceived risk of both on- and offbalance sheet activities than the effect of an increase in off-balance sheet activities. The
observed decline in the demand deposit ratio is consistent with both the technology
explanation and the hypothesis that the safety net protects greater fragility and thus
reduces the need for depositor discipline, but inconsistent with the reduced depositor
concern hypothesis. However, to the extent that either of the first two effects alone or in
combination with each other are strong, they can outweigh any effect of the reduced need
for depositor concern explanation of the impact of the safety net and need not disprove the
existence of that effect. The decline in the cash ratio is also likely to reflect the prolonged
decline in reserve requirements, particularly as, except for the 1930s, excess reserves are
typically quite small. We attempt to differentiate between these two explanations in the next
section.
Lastly, it is interesting to consider whether the observed changes are consistent with
concern over bank fragility by the market. The reductions in both the capital and cash ratios
would suggest no. However, it may also reflect the fact that many of the costs of bank
failure are now absorbed by the deposit insurance agency or taxpayer rather than by
13
individual depositors. On the other hand, the decrease in the demand deposit ratio is
consistent with this hypothesis.
5.
T he Implications of Bank F ragility for Bank F ailures in the F uture
The decrease in the capital and cash-to-asset ratios through time suggests that banks
may have become more fragile. As discussed in the previous section, the increase in
fragility is consistent with both the impact that the introduction of a poorly designed bank
safety net and the impact that advances in technology may be expected to have. It is,
however, possible to differentiate between these two hypotheses. To the extent that the
safety net makes any contribution to an increase in fragility, so that fragility is greater than
in its absence, the increase in fragility would translate into an increase in bank failures, if
advances in technology were the only force increasing fragility, in the absence of a
decrease in market discipline or an increase in macroeconomic instability, there should be
no increase in bank failures. As noted earlier, the average bank failure rate increased
somewhat, or, at least, did not decrease between the pre- and post-safety net periods. At
the same time, however, the average annual nonbank failure rate declined substantially
from 1.01 percent to 0.65 percent, a rate only 60 percent as great as that of banks10. As
nonbanks are not insured and subject only to market discipline, the decline in their failure
rate suggests both that macroeconomic or regional instability is unlikely to have increased
greatly relative to the extant degree of market discipline and that the intensity of market
10The decline in the nonbank failure rate is statistically significant at the 5 percent level.
14
discipline on noninsured firms is unlikely to have declined greatly. In the absence of the
introduction of the safety net, there is no reason not to expect the bank failure rate to also
have declined significantly.
Thus, it appears likely that the observed increases in the bank fragility indicators may
at least partially be attributed to the introduction and subsequent broadening of the safety
net under banking and that the safety net was poorly designed in terms of not sufficiently
curtailing the moral hazard and principal-agent problems that were introduced concurrently:
In particular, the decline in capital ratios left many banks with insufficient capital to
withstand losses from the instability in the macro or regional economies that occurred in
the 1930s and 1980s and that they were better able to absorb before. That is, the
introduction of the safety net increased their perceived and actual fragility.
6.
C onclusion
This paper has provided evidence that, at least in the U.S., the perceived fragility of
banks arising from their low capital-to-asset and cash-to-asset ratios and their high
demand to total deposit (debt) ratio has not always been translated into actual fragility as
reflected in a high failure rate. Indeed, although for the entire 1870 to 1994 period analyzed
the bank failure rate in the U.S. was moderately higher than that of nonbank firms, before
the introduction of the first safety net under banks in 1914, the bank failure rate was slightly
lower than that of nonbanks. Both the perceived fragility of banks, as measured by two of
the three indicator ratios, and the annual average failure rate of banks increased after the
introduction of the safety net and the accompanying partial replacement of market
15
discipline by regulatory discipline. Indeed, in the period since 1914, the average annual
failure rate for banks exceeded that for nonbanks by nearly 70 percent
Although the increase in the fragility measures is consistent with the introduction of
the safety net, it is also generally consistent with the advances in technology and
communications. However, in the absence of a concurrent decrease in market discipline
brought about for reasons other than either the introduction of the safety net or an increase
in macroeconomic instability, the increase in the bank failure rate is inconsistent with
greater fragility from advances in technology and communications alone. Moreover, the
statistically significant decline in the average failure rate of noninsured nonbanks between
the pre- and post-1914 periods suggests that macroeconomic and regional instability did
not increase relative to the degree or market discipline and, ceteris paribus, contribute to
the increased (or unchanged) bank failure rate. Rather, it appears that the introduction of
the federal government safety net, which was intended in large measure to reduce the
likelihood of systemic risk, has unintentionally increased the fragility of banks by reducing,
in particular, their capital ratios to levels inconsistent with the prevailing instability in the
macro and regional economies and contributed significantly to an increase in the instability
of both individual banks and the banking system (Kaufman, 1996). Because banks also
operated with lower capital ratios than nonbanks before 1914, the market did not appear
to perceive them as unusually fragile. The capital ratios were sufficiently high to restrict
both the bank failure rate and creditor (depositor) losses to below that of nonbanks.
Thus, the public perception of banking as fragile and susceptible to failure is currently
largely correct, but ironically primarily because of government policies intended to
16
enhance the safety of banks. Before the introduction of such policies in 1914, regulation
by the marketplace apparently was effective in preventing the perceived fragility of banks
from being translated into actual fragility. That is, the market appears to have handled
"fragile" banks with care and did not permit them to fail at a greater rate or with greater cost
than perceived less fragile nonbanks. The increase in bank failures since the introduction
of the safety net thus reflects primarily regulatory (government) failure than market failure.
The safety net was introduced in 1914 and expanded in 1934 in response to periodic sharp
increases in the rate of bank failures, particularly in 1907-1908 and 1929-1933. Rather
than repair defects in the market mechanism that had produced relatively low average
rates of bank failures, the government chose more dramatic and visible policies.11 In
banking, at least, the perfect became the enemy of the good.
To the extent that these policies have reduced or eliminated systemic risk, which may
or may not have existed before 1914 and which may have existed in the 1930s, they
appear to have done so only at a very high cost to society. Because the evidence
presented in this paper is primarily circumstantial, the conclusions are only suggestive
rather than conclusive. Nevertheless, they challenge much conventional wisdom about
banking and suggest that the usefulness and effectiveness of particularly the pre-FDICIA
"This outcome, of course, is not unique to public policy applied to banking. Charles Schultze has
recently observed that
[N]oneconomists have an almost universal desire to deal with market failures through carefully specified
regulation rather than a change in incentive structures. Such specification is the natural function of
lawyers....When government intervenes in the marketplace, our political leaders typically rule out the
manipulation of economic incentives to deter undesirable actions because reliance on market
responses injects an uncertain, partially random, and therefore “unfair* set of forces into the picture
(Schultze, 1986, p. 27).
17
prudential regulations on banks are worthy of more rigorous analysis and reevaluation.12
Moreover, because as noted earlier, serious and costly banking problems have plagued
many other countries in recent years despite the existence of similar prudential regulatory
policies, particularly the safety net, the evidence developed for the U.S. also suggests that
a reexamination of the costs and benefits of these policies is warranted worldwide.
12A study of historical stock and bond market volatility in the U.S. came to a similar conclusion. The
authors found that
In contrast with impressions one receives from financial and other historians, the National Banking
era before 1914 exhibits less, not more, stock and bond market volatility than the period from 1914
to the present This raises questions about the effectiveness of financial and regulatory reforms that
were put in place in 1914 and later years. (Wilson, Sylla, and Jones, 1990, p. 112).
TABLE 1
Fragility Measures for Banks And Nonfinancial Firms
1870-1992
a All banks, Historical Statistics of the llnitecLSlates, Colonial Timesto 1970 (U.S. Department of Commerce, Bureau of the Census).
b All insured commercial banks, EDIC Statistics onBanking - Historical 1934^1994.
c Nonfinancial firms include corporations other than banks, security dealers, insurance companies and agents, real estate firms, and investment
companies. Historical Statistics of the United States, Colonial Times to 19Z0; Statistical,AbstraCLof theiinited States (U.S. Department of
Commerce, Bureau of the Census), (various years); and Statistics of.Income: CorporatiorLlncome Tax Return (U.S. Internal Revenue Service),
(various years).
d Balance Sheets for the US. Economy 1945-1994 (Board of Governors of the Federal Reserve System).
e All commercial banks, Historical Statistics of the United States, Colonial Times to !97D.
18
19
TABLE 2
Annual Failure Rates for Banks and Nonbank Firms
1870-1994
Banks*
Nonbanksb
(Percent)
1870-1994
1.03
(2.74)c
0.78
(0.35)
1870-1913
0.91
(1.00)
1.01
(0.21)
1914-1994
1.09
(3.32)
0.65
(0.35)
1870-1929, 1934-1994
0.65
(0.87)
0.76
(0.34)
* Includes chartered commercial and savings banks.
b Excludes finance, real estate, and railroad firms, professionals, and farmers.
c Standard deviations in parenthesis.
Sources: Annual Report (Board of Governors of the Federal Reserve System), (various years);
Annual Statistical Digest 1970-1979 (Board of Governors of the Federal Reserve System); Annual
Reports (Federal Deposit Insurance Corporation), (1990 and 1994) ; Historical Statistics of the
United States, Colonial Times to 1970 (U.S. Department of Commerce, Bureau of the Census);
and Statistical Abstract of the United States (U.S. Department of Commerce, Bureau of the
Census), (various years).
TABLE 3
Changes in Indicators of Bank Fragility Through Time
Stylized Facts About Fragility
_______ Measures
Consistent With Introduction
of Safety Net
Consistent with Advances in
Technology
K/A: Low and declining
Yes, banks willing to gamble
more
C/A: Declining
Yes, alternate source of liquidity8 Yes, permits greater
diversification8
DD/TD: High and declining
Yes and no:
Yes-requires less monitoring
No-increases fragility
K
A
C
DD
TD
8
=
=
=
=
=
=
Net worth
Total on-balance sheet assets
Cash on hand or balances at other banks
Demand and other transaction deposits
Total deposits
Subject to reserve requirement constraint.
Yes and no:
Yes-Permits greater
diversification of assets and
improved monitoring
No-reported "on-balance sheet"
ratio rises to maintain constant
total balance sheet ratio.
Yes, permits more close
substitutes for transaction
purposes
Consistent With
Fragility Concern
No
No
Yes and no:
Yes-declining
No-high level
T A B L E A1
CAPITAL-TO-ASSET RATIOS AT CORPORATIONS BY INDUSTRY SELECT YEARS, 1926-1992
INDUSTRY{
Year
1926
1930
1940
1950
1960
1970
1980
1986
1990
1992
Total
45.5
48.3
43.2
37.4
33.9
28.5
25.5
26.1
26.1
28.5
Excluding
Finance
60.3
62.3
61.0
61.2
56.1
44.9
39.3
35.5
32.5
32.3
Finance
21.3
28.9
27.1
13.4
14.9
14.2
13.2
18.8
21.0
25.7
Construction
40.5
48.3
49.7
43.8
34.6
28.6
24.7
24.1
26.2
31.0
S o u r c e : S ta tis tic s o f In c o m e : C o rp o ra tio n In c o m e T a x R e tu r n s
Treasury), various years.
Mining
68.6
69.8
70.9
67.0
63.0
57.1
42.6
47.8
50.0
48.6
Manufacturing
71.5
75.3
72.9
68.5
64.5
51.2
43.8
38.4
35.5
34.3
Public
Utilities
46.3
50.7
48.7
51.4
48.5
42.3
37.3
37.7
35.0
34.1
Trade
63.0
63.6
59.8
58.2
50.4
41.1
34.4
28.0
24.5
25.6
Services
52.5
57.1
48.2
53.3
38.6
32.8
29.7
26.2
24.0
29.7
(Washington, D.C.: U.S. Internal Revenue Service, Department of the
21
TABLE A2
'
' ' ' '
Year
Telegraph
CAPITAL-TO-ASSET RATIOS FOR SELECTED INDUSTRIES 1902-1970
• ' .: *
'
INDUSTRY
Street and Electric
Railroads
Telephone
SElectric Light and Power
Mining
(percent)
1902
69
62
52
1907
67
54
49
1912
60
50
46
52
1917
56
58
42
48
1922
54
58
37
44
1927
34
47
1932
31
50
1937
48
1940
49*
1950
45
1960
42
1970
40
’Break is series.
S ou rce: U.S. D ep artm en t o f C om m erce, Historical Statistics, p , 939.
22
23
REFEREN CES
Benston, George J., Robert A. Eisenbeis, Paul M. Horvitz, Edward J. Kane, and George
G. Kaufman, Perspectives on Safe and Sound Banking. Cambridge, MA.: MIT Press,
1986.
Benston, George J. and George G. Kaufman, RisKand Solvency Regulation of Depository
Institutions: Past Policies and Current Options. New York: Salomon Brothers Center,
Graduate School of Business, New York University, 1989.
Benston, George J. and George G. Kaufman, "Is the Banking and Payments System
Fragile," Journal of Financial Services Research. December 1995, pp. 209-240.
Benston, George J. and George G. Kaufman, 'The Appropriate Role of Bank Regulation,”
Economic Journal, May 1996. pp. 688-697.
Calomiris, Charles W. and Charles M. Kahn, "The Role of Demandable Debt in
Restructuring Optimal Banking Arrangements," American Economic Review, June
1991, pp. 497-513.
Federal Deposit Insurance Corporation, 1994 Annual Report, Washington, D.C., 1995.
Flannery, Mark J., "Debt Maturity and the Deadweight Cost of Leverage: Optimally
Financing Banking Firms," American Economic Review, March 1994, pp. 320-331.
Greenspan, Alan, "Remarks at the International Conference of Bank Supervisors in
Stockholm, Sweden," Washington, D.C.: Board of Governors of the Federal Reserve
System, June 13,1996.
Kane, Edward J., The S&L Mess, Washington, D.C.: Urban Institute, 1989a.
Kane, Edward J., "Changing Incentives Facing Financial-Services Regulators," Journal of
Financial Services Research, September 1989b, pp. 265-274.
Kaufman, George G., "Bank Runs: Causes, Benefits, and Costs," Cato Journal, Winter
1988, pp. 559-587.
Kaufman, George G., "Capital in Banking: Past, Present and Future," Journal of Financial
Services Research, April 1992, pp. 385-402.
Kaufman, George G., "Bank Contagion: A Review of the Theory and Evidence,” Journal
of Financial Services Research, April 1994, pp. 123-150.
Kaufman, George G., ’The U.S. Banking Debacle of the 1980s: An Overview and
Lessons," Financier, May 1995, pp. 9-26.
Kaufman, George G., "Bank Failures, Systemic Risk, and Bank Regulation," Working
Paper Series (96-1), Federal Reserve Bank of Chicago, January 1996.
24
Lindgren, Carl-Johan, Gillian Garcia and Matthew Saal, Bank Soundness and
Macroeconomic Policy, Washington, D.C.: International Monetary Fund (1996,
forthcoming).
Macey, Jonathan R. and Geoffrey P. Miller, "Double Liability of Bank Shareholders: History
and Implications," Wake Forest Law Review 27TNo. 1,1992, pp. 31-62.
Minsky, Hyman P., "Financial Factors in the Economics of Capitalism," Journal of Financial
Services Research, December 1995, pp. 197-208
Schultze, Charles L., "The CEA: An Inside Voice for Mainstream Economics,” JoumaLof
Economic Perspectives, Summer 1996, pp. 23-39.
Schwartz, Anna J., "Coping With Financial Fragility: A Global Perspective,” Journal of
Financial Services Research, December 1995, pp. 445-451.
Sylla, Richard, John J. Wallis, and John B. Legler, "Historical Economics: U.S. State and
Local Government," NBER Reporter, April 1995, pp. 14-16.
U.S. House of Representatives, Committee on Banking, Financial and Urban Affairs, "An
Analysis of Federal Reserve Discount Window Loans to Failed Institutions," Staff
Report, June 11,1991.
Wallis, John J., Richard E. Sylla, and John B. Legler, 'The Interaction of Taxation and
Regulation in Nineteenth Century U.S. Banking," In Claudia Goldin and Gary Lidecap,
eds., The Regulated Economy, Chicago: University of Chicago Press, 1994, pp. 121144.
Wilson, Jack W., Richard E. Sylla, and Charles P. Jones, "Financial Market Panics and
Volatility in the Long-Run, 1830-1988" in Eugene N. White, ed., Crashes and Panics:
The Lessons From History, Homewood, II.: Dow Jones-lrwin, 1990, pp. 85-125.
@FedFRASER