See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/223859920

The Impact of FDICIA and Prompt Corrective Action on Bank Capital and Risk:
Estimates Using a Simultaneous Equations Model

Article in Journal of Banking & Finance · June 2001

DOI: 10.1016/S0378-4266(00)00125-4

CITATIONS READS

371 669

2 authors, including:

Raj Aggarwal
Kent State University Foundation
278 PUBLICATIONS 6,133 CITATIONS

SEE PROFILE

All content following this page was uploaded by Raj Aggarwal on 18 February 2018.

The user has requested enhancement of the downloaded file.

 Journal of Banking & Finance 25 (2001) 1139±1160
www.elsevier.com/locate/econbase

The impact of FDICIA and prompt corrective

action on bank capital and risk: Estimates using
a simultaneous equations model q
Raj Aggarwal a, Kevin T. Jacques b,*

a
Firestone Chair in Finance, Kent State University, Kent, OH 44242, USA
b
Department of Economics and Finance, John Carroll University, University Heights, Cleveland,
OH 44118, USA
Received 25 September 1998; accepted 24 March 2000

Abstract

One of the requirements of the Federal Deposit Insurance Corporation Improvement

Act (FDICIA) was that bank regulators establish capital ratio zones that mandate
prompt corrective action (PCA) and early intervention in troubled banks. However,
prior research suggests that increases in regulatory capital standards can lead to o€-
setting increases in risk. This paper develops and estimates a 3SLS model to examine the
simultaneous impact of PCA on both bank capital and credit risk. The results document
that the FDICIA was e€ective in that, subsequent to its passage, US banks increased
their capital ratios without o€setting increases in credit risk. Ó 2001 Elsevier Science
B.V. All rights reserved.
JEL classi®cation: G2

Keywords: FDICIA; Bank regulation; Capital ratios

q
The views expressed are those of the authors and do not necessarily re¯ect those of John
Carroll University or of Kent State University.
*
Corresponding author. Tel.: +1-216-397-4655; fax: +1-216-397-1728.
E-mail address: [email protected] (K.T. Jacques).

0378-4266/01/$ - see front matter Ó 2001 Elsevier Science B.V. All rights reserved.
PII: S 0 3 7 8 - 4 2 6 6 ( 0 0 ) 0 0 1 2 5 - 4
1140 R. Aggarwal, K.T. Jacques / Journal of Banking & Finance 25 (2001) 1139±1160

1. Introduction

The nature of ecient and e€ective bank regulation continues to be of much

concern. In developing regulations, regulators face moral hazard, agency costs,
asymmetric information and many other obstacles, especially in the presence of
deposit insurance, and banks continue to fail throughout the world, often at
alarming rates. This paper examines and ®nds e€ective one attempt at regu-
latory reform, the Federal Deposit Insurance Corporation Improvement Act
(FDICIA) passed by the US Congress in 1991.
One of the major features of FDICIA was a new statutory framework for
bank supervision that detailed early intervention and corrective action by bank
regulators in dealing with troubled institutions. Speci®cally, Section 131 of
FDICIA, prompt corrective action (PCA), de®ned for banks, a series of capital
thresholds used to determine what supervisory actions would be taken by bank
regulators. As such, PCA classi®es banks into one of ®ve groups depending
upon their capital ratios. While banks above the highest threshold are con-
sidered well-capitalized, increasingly severe statutory restrictions and penalties
are required to be applied to banks that fall into the bottom three PCA cat-
egories.
FDICIA was a departure from other banking legislation in that it mandated
PCA, and thus, there is considerable interest in assessing its e€ectiveness. While
PCA was intended to supplement the existing supervisory authority of bank
regulators, FDICIA legislated mandatory intervention in undercapitalized in-
stitutions rather than regulatory discretion, in an e€ort to save undercapital-
ized banks from becoming insolvent. There is some evidence that these changes
in capital regulations made by regulators in the 1990±1992 period resulted in
increased pressure on bank holding companies (Wall and Peterson, 1995).
Indeed, informal observations indicate that FDICIA and PCA were associated
with increases in aggregate bank capital (Benston and Kaufman, 1997). 1
While the implementation of the PCA standards have focused attention on
bank capital ratios, insucient attention has been given to how PCA may have
impacted the level of credit risk in bank portfolios. Theory and some empirical
evidence suggests that increasing bank regulatory capital standards may have
the unintended e€ect of causing banks to engage in increasingly risky behavior
(e.g., Blum, 1999). But prior research does not speci®cally address how bank

1
In the two years following the passage of FDICIA, equity capital held by commercial banks in
the aggregate increased from $231.7 billion to $297.0 billion, the aggregate equity capital to asset
ratio increased from 6.75% to 8.01%, and the number of banks classi®ed under PCA as
undercapitalized, signi®cantly undercapitalized, or critically undercapitalized declined from 388 to
48. These data are taken from the Oce of the Comptroller of the Currency.
 R. Aggarwal, K.T. Jacques / Journal of Banking & Finance 25 (2001) 1139±1160 1141

capital ratios and risk levels responded simultaneously to the implementation

of PCA.
This study uses a simultaneous equations model and three-stage least
squares (3SLS) estimation to examine how the PCA standards simultaneously
impact bank capital ratios and credit risk. As compared to single equation
models of capital and risk that assume either bank capital ratios or risk are
exogenous to the bank, the simultaneous equations model and 3SLS technique
used here explicitly recognize the endogeneity of both capital ratios and credit
risk. The results presented here document that, after accounting for the end-
ogeneity of both capital ratios and credit risk, and the role of other plausible
variables, PCA had a signi®cant impact both in terms of raising capital ratios
and reducing credit risk for banks.

2. Bank capital, portfolio risk, and prompt corrective action

Given the belief that regulatory forbearance during the 1980s had exacer-
bated losses associated with bank and thrift failures, FDICIA contained two
key provisions designed to reduce the cost of failed banks. First, FDICIA
contained a provision for early closure of failing institutions, while they still
had a positive level of capital, as a solution to excessive losses to the deposit
insurance fund (Kane, 1983), and to the moral hazard problem created by
®xed-rate deposit insurance (Buser et al., 1981; Hovakimian and Kane, 2000).
The second key provision of FDICIA involved early intervention in problem
banks by bank regulators. It was suggested that PCA may lower resolution
costs for failed banks and reduce losses for deposit insurers by discouraging
healthy institutions from becoming undercapitalized, limiting the time to
failure of an undercapitalized bank, or by reducing the number of failures
among undercapitalized institutions.
The ®ve capital categories de®ned under PCA are: (1) well-capitalized; (2)
adequately capitalized; (3) undercapitalized; (4) signi®cantly undercapital-
ized; (5) critically undercapitalized. As shown in Table 1, bank classi®cation
into these categories depends on three di€erent capital ratios: (1) the total
risk-based capital ratio; (2) the Tier 1 risk-based capital ratio, and (3) the
Tier 1 leverage ratio. 2 For example, a well-capitalized bank must have a
total risk-based capital ratio greater than or equal to 10%, a Tier 1 risk-
based capital ratio greater than or equal to 6%, and a Tier 1 leverage ratio
greater than 5%, while the corresponding thresholds for adequately

2
FDICIA authorizes bank regulators to reclassify a bank in a lower capital category if, in the
opinion of the bank regulators, the bank is in operating in an unsafe or unsound manner.
1142 R. Aggarwal, K.T. Jacques / Journal of Banking & Finance 25 (2001) 1139±1160

Table 1
Capital ratios under prompt corrective actiona;b
Total risk-based Tier 1 risk-based Tier 1 leverage
capital (%) ratio (%) ratio (%)
Well-capitalized P10 P6 P5
Adequately capitalized P8 P4 P4c
Undercapitalized <8 <4 < 4c
Signi®cantly undercapitalized <6 <3 <3
Critically undercapitalized Tangible equity P 2
a
Source: Oce of the Comptroller of the Currency.
b
The tangible equity ratio equals Tier 1 capital plus cumulative preferred stock and related surplus
less intangibles except qualifying purchased mortgage servicing rights divided by the sum of total
assets less intangibles except qualifying purchased mortgage servicing rights.
c
As established by the Oce of the Comptroller of the Currency, the Tier 1 leverage ratio for
adequately capitalized and undercapitalized institutions equals 3 percent if the bank is rated a
CAMEL 1.

capitalized institutions are 8%, 4%, and 4%, respectively. 3 If a bank fails to
meet the minimum thresholds for adequate capital, it becomes undercapi-
talized, with mandatory restrictions being placed on its activities that be-
come increasingly severe as the bankÕs capital ratios deteriorate below
additional thresholds.
For example, as indicated in Appendix A, undercapitalized banks ± those
with total risk-based capital ratios less than 8%, Tier 1 risk-based ratios less
than 4%, and Tier 1 leverage ratios less than 4% ± are subject to a multitude of
restrictions that include the need to submit and implement a capital restoration
plan, limits on asset growth, and restrictions on expansion. In the extreme,
once a bankÕs tangible equity ratio falls to 2% or less, they are considered to be
critically undercapitalized and face not only more stringent restrictions on
activities than other undercapitalized banks, but also the appointment of a
conservator (receiver) within 90 days. 4
While PCA was intended to solve many of the problems associated with
regulatory discretion and forbearance, moral hazard, and deposit insurance
losses, PCA is not without criticisms and limitations. First, as Peek and Ro-
sengren (1996) note, providing e€ective early intervention is predicated on the
ability of ``troubled bank'' indicators to e€ectively identify problem institutions

3
It should be noted that a bank may not be classi®ed as well-capitalized if it is subject to either a
cease and desist order, a formal agreement with its regulator, a capital directive, or a PCA directive
to raise capital.
4
The tangible equity ratio equals the total of Tier 1 capital plus cumulative preferred stock and
related surplus less intangibles except qualifying purchased mortgage servicing rights divided by the
total of bank assets less intangible assets except qualifying purchased mortgage servicing rights.
 R. Aggarwal, K.T. Jacques / Journal of Banking & Finance 25 (2001) 1139±1160 1143

in a timely manner. Unfortunately, a number of studies have noted the inad-

equacy of such indicators in FDICIA. 5
A second problem with PCA is that it may prompt failing institutions to
increase risk. While the primary purpose of early closure as it applies to PCA is
to prevent banks from taking increasing levels of risk as they approach in-
solvency, Levonian (1991) demonstrates that early closure may fail to protect
the deposit insurance fund from losses because it creates incentives for banks to
take actions prior to closure that could increase the cost of the failure. This
occurs because an early closure policy increases the incentive for banks to hold
high-risk assets. Such a result is con®rmed by Davies and McManus (1991)
who ®nd that increasing the closure threshold for failing institutions may cause
some institutions to increase risk. They recommend an increasingly stringent
closure policy be implemented with increasing attention to the risks in bank
portfolios.
Finally, while early closure may lead banks to increase risk, some recent
evidence suggests that increasingly stringent bank capital regulations may have
the same e€ect (Kahane, 1977; Koehn and Santomero, 1980; Kim and
Santomero, 1988). Therefore, increases in minimum capital standards and the
imposition of other increasingly stringent restrictions by bank regulators may
cause banks to increase not just their capital ratios, but may also have the
unintended e€ect of causing them to increase their level of risk (Gennotte and
Pyle, 1991). Empirical work by Shrieves and Dahl (1992) indeed documents
that changes in bank capital ratios and asset risk are simultaneously and
positively related; increases in capital ratios lead to increasing levels of asset
risk and increases in asset risk lead banks to increase their capital ratios.
As this discussion shows, despite possible limitations, PCA may have had a
signi®cant impact on bank behavior. Indeed, if banks viewed the sanctions of
PCA as being costly, and regulators as being credible in carrying out the
sanctions, then banks may have altered their capital ratios and portfolios so as
to ensure compliance with the standards. However, as this discussion indicates,
in assessing the e€ectiveness of PCA, its impact on both bank capital and risk
must be assessed simultaneously. Furthermore, the model must account for the
confounding e€ects of other variables such as income, size, holding company
status, and location.

5
For example, Peek and Rosengren (1996, 1997) and Jones and King (1992, 1995), ®nd that the
capital ratio thresholds used in PCA are lagging indicators of a bankÕs ®nancial status. As a result,
Jones and King (1995) conclude that, had current PCA thresholds been applied in the late 1980s,
they would have failed to treat most failed banks as undercapitalized. Fortunately, as Peek and
Rosengren (1996) note, bank examiners use far more information than capital ratios in identifying
problem banks.
1144 R. Aggarwal, K.T. Jacques / Journal of Banking & Finance 25 (2001) 1139±1160

3. A simultaneous equations model for PCA and changes in bank capital and risk

To examine the possible impact of PCA on bank capital ratios and risk, the
simultaneous equations model used by Shrieves and Dahl (1992), and later
employed by Jacques and Nigro (1997) to study risk-based capital, is modi®ed
to incorporate the PCA zones. In this framework, observed changes in bank jÕs
capital ratio and risk in period t are modeled as the sum of two components, a
discretionary adjustment and a change caused by an exogenously determined
random shock. Thus:

DCAPj;t ˆ Dd CAPj;t ‡ Ej;t ; 1†

DRISKj;t ˆ Dd RISKj;t ‡ Uj;t ; 2†

where DCAPj;t and DRISKj;t are the observed changes in capital ratios and risk
levels, respectively, Dd CAPj;t and Dd RISKj;t represent the discretionary ad-
justments in capital ratios and risk, and Ej;t and Uj;t are exogenously deter-
mined random shocks. Here, capital is measured using the regulatory capital
ratios used in the PCA standards, and risk is measured using one of two
measures of credit risk: (1) the ratio of risk-weighted assets to total assets; or,
(2) the ratio of nonperforming loans to total assets. The discretionary adjust-
ment in a bankÕs capital ratio and risk level is modeled using the partial ad-
justment framework, thereby recognizing that banks may not be able to adjust
their desired capital ratio and risk level instantaneously. Thus:

Dd CAPj;t ˆ a CAPj;t CAPj;t 1 †; 3†

Dd RISKj;t ˆ b RISKj;t RISKj;t 1 †; 4†

where CAPj;t and RISKj;t are bank j's target capital ratio and risk level, re-
spectively. In the partial adjustment framework, discretionary adjustments in
the bankÕs capital ratio and risk level are proportional to the di€erence between
the target and its value in the previous period. Substituting Eqs. (3) and (4) into
Eqs. (1) and (2), respectively, yields:
DCAPj;t ˆ a CAPj;t CAPj;t 1 † ‡ Ej;t ; 5†

DRISKj;t ˆ b RISKj;t RISKj;t 1 † ‡ Uj;t : 6†

Eqs. (5) and (6) state that observed changes in bank jÕs capital ratio and risk
level are a function of the target capital ratio and risk level in period t, the
capital ratio and level of portfolio risk in period t 1, and any random shocks.
The target capital ratio and risk level are not observable, but are assumed to
depend upon some set of observable variables. An example of an exogenously
 R. Aggarwal, K.T. Jacques / Journal of Banking & Finance 25 (2001) 1139±1160 1145

determined random shock to the bank that could in¯uence their capital ratio or
risk level is a change in the bankÕs macroeconomic environment.
Consistent with prior literature (Berger, 1995; Shrieves and Dahl, 1995), in
this study changes in a bankÕs capital ratio and risk are in¯uenced by a number
of explanatory variables including: the size of the bank (SIZE), whether the
bank is aliated with a multibank holding company (BHC), net income (INC),
holdings of government securities (SEC), liquidity (CASH), asset quality
(LLPROV), rural versus urban location (MSA), changes in risk (DRISKj;t ) and
capital ratios (DCAPj;t ), lagged capital ratios (CAPt 1 ) and risk levels
(RISKt 1 ), and the degree of regulatory pressure, as denoted by the PCA zone
of the bank. Given these variables for explaining changes in a bankÕs target
capital ratio and risk level, Eqs. (5) and (6) are:

DCAPj;t ˆ d0 ‡ d1 SIZEj;t ‡ d2 BHCj;t ‡ d3 INCj;t ‡ d4 DRISKj;t

‡ d5 SECj;t ‡ d6 CASHj;t ‡ d7 LLPROVj;t ‡ d8 MSAj;t
!
X X
‡ di PCA d11 ‡ di PCA CAPj;t 1 ‡ lj;t ; 7†
iˆ9;10 iˆ12;13

DRISKj;t ˆ k0 ‡ k1 SIZEj;t ‡ k2 BHCj;t ‡ k3 DCAPj;t k4 RISKj;t 1

‡ k5 SECj;t ‡ k6 CASHj;t ‡ k7 LLPROVj;t ‡ k8 MSAj;t

X
‡ kk PCA ‡ xj;t ; 8†
kˆ9;10

where lj;t and xj;t are disturbance terms. Here, SIZE is measured as the natural
log of bank j's total assets. As Shrieves and Dahl (1992) note, size may have an
impact on capital ratios and risk levels for a number of reasons including the
bankÕs investment opportunity set and its access to equity capital markets.
Thus large banks may be expected to hold less capital than smaller banks. BHC
is a dummy variable that equals unity for banks belonging to a multibank
holding company, while INC equals bank jÕs net income to asset ratio in period
t. If banks belonging to a multibank holding company have their capital ratios
and risk levels managed at the holding company level, they may have lower
target capital ratios and higher target risk levels than independent banks.
Following Jacques and Nigro (1997), the income to asset ratio (INC) is in-
cluded in the capital equation to account for the ability of pro®table banks to
increase their capital ratios by retaining earnings, and the fact that banks with
a negative return on assets are constrained in their ability to increase capital
through retained earnings.
The ratio of government securities to total assets (SEC) is included to ac-
count for the favorable interest rate environment of the early 1990s. A priori,
banks with signi®cant security holdings would be expected to have higher
1146 R. Aggarwal, K.T. Jacques / Journal of Banking & Finance 25 (2001) 1139±1160

capital ratios, as security sales in a falling interest rate environment would

increase earnings. Alternatively, if banks with large holdings of government
securities retained, rather than sold, these securities during a falling rate en-
vironment, then they may require lower capital. CASH is included in the si-
multaneous equation system to recognize that banks with greater cash
holdings, as a percentage of assets, have greater liquidity, less risk, and less
need for capital. Loan loss provisions as a percentage of assets (LLPROV) are
included as a proxy for asset quality, with higher LLPROV values being as-
sociated with greater credit risk and a resulting greater need for capital. Fi-
nally, MSA is a dummy variable to account for whether banks in metropolitan
statistical areas behave di€erently from rural banks.
Similar to Shrieves and Dahl (1992), DRISKj;t and DCAPj;t are also in-
cluded in the model to recognize the possible simultaneous relationship be-
tween changes in capital ratios and changes in risk. Shrieves and Dahl (1992)
®nd that changes in capital ratios and risk are positively related, a ®nding they
argue is consistent with a number of hypotheses including the unintended ef-
fects of minimum capital standards, regulatory costs, bankruptcy cost avoid-
ance, and managerial risk aversion. On the other hand, Jacques and Nigro
(1997) ®nd evidence that changes in capital ratios and risk are negatively re-
lated, a result they note may occur because of methodological ¯aws in the risk-
based capital standards. Alternatively, Shrieves and Dahl (1992) note that a
negative relationship may exist if banks seek to exploit the deposit insurance
subsidy.
Of particular interest in this study are the regulatory pressure variables.
Consistent with previous work, this study uses dummy variables to signify the
degree of regulatory pressure experienced by a bank. Speci®cally, the PCA
dummies are set such that:
PCAA ˆ 1 if the bank is classified as adequately capitalized; else ˆ 0;

PCAU ˆ 1 if the bank is classified in one of the three undercapitalized

zones undercapitalized; substantially undercapitalized;
or critically undercapitalized†; else ˆ 0:
These variables allow banks across di€erent PCA zones to respond di€erently,
both in capital ratios and in credit risk. A priori, banks classi®ed as under-
capitalized, PCAU, would be expected to have a stronger and faster response
than better capitalized institutions because PCA imposes penalties on their
activities. Furthermore, undercapitalized institutions may have had a stronger
response after 1992 than in 1992 because the PCA sanctions did not become
e€ective until 1993. Adequately capitalized banks (PCAA) may also increase
their capital ratios or reduce their credit risk if they perceive a cost from either
regulators or the market for not being classi®ed as well-capitalized. In addition,
 R. Aggarwal, K.T. Jacques / Journal of Banking & Finance 25 (2001) 1139±1160 1147

they may increase their capital ratios or reduce risk if they desire to maintain a
bu€er stock of capital above the regulatory minimum in order to protect
against shocks to income or equity as argued by Wall and Peterson (1987) and
Furlong (1992).
Furthermore, PCA-capital interactive terms (PCAA  CAPt 1 and
PCAU  CAPt 1 ) are included in the capital equation to allow banks in dif-
ferent PCA zones to have di€erent speeds of adjustment to their target capital
ratios. A priori, undercapitalized banks would be expected to adjust their
capital ratios at a faster rate than their more capitalized counterparts and,
because sanctions became e€ective in 1993, the speed of adjustment coecient
may have increased after 1993. Alternatively, if raising capital from external
sources is more costly for undercapitalized institutions, as argued by Baer and
McElravey (1993), then they may not be able to adjust their desired capital
ratios at a faster speed, regardless of the incentives created by the PCA sanc-
tions. Under these circumstances, the interactive terms would not be expected
to be statistically signi®cant. Given these de®nitions of regulatory pressure,
Eqs. (7) and (8) can be written:

DCAPj;t ˆ d0 ‡ d1 SIZEj;t ‡ d2 BHCj;t ‡ d3 INCj;t ‡ d4 DRISKj;t

‡ d5 SECj;t ‡ d6 CASHj;t ‡ d7 LLPROVj;t ‡ d8 MSAj;t
‡ d9 PCAA ‡ d10 PCAU d11 CAPj;t 1
 
d12 PCAA CAPj;t 1 d13 PCAU CAPj;t 1 ‡ lj;t ; 9†

DRISKj;t ˆ k0 ‡ k1 SIZEj;t ‡ k2 BHCj;t ‡ k3 DCAPj;t k4 RISKj;t 1

‡ k5 SECj;t ‡ k6 CASHj;t ‡ k7 LLPROVj;t ‡ k8 MSAj;t

‡ k9 PCAA ‡ k10 PCAU ‡ xj;t : 10†

Estimation of Eqs. (9) and (10) requires measures of both bank capital ratios
and credit risk. As noted earlier, given the regulatory capital requirements
associated with PCA, capital ratios are measured according to the three cor-
responding regulatory standards: the Tier 1 leverage ratio (T1LEV), the Tier 1
risk-based capital ratio (T1RBC), and the total risk-based capital ratio
(RBCR). Following Avery and Berger (1991), Shrieves and Dahl (1992), Berger
and Udell (1993), and Berger (1995), credit risk is measured using either: (1) the
ratio of total risk-weighted assets to total assets (RWATA); or, (2) the ratio of
nonperforming loans to total assets (NPL). Avery and Berger (1991) have
shown that RWATA correlates with risky behavior. Because nonperforming
loans re¯ect the ex-post outcome of lending decisions, consistent with Shrieves
and Dahl (1992), credit risk in a given year is measured using NPL in the
following year.
1148 R. Aggarwal, K.T. Jacques / Journal of Banking & Finance 25 (2001) 1139±1160

4. Data and statistical methodology

As noted earlier, the number of undercapitalized institutions declined the

year after FDICIA was passed, and during the years after it became e€ective.
This result is not surprising because, a priori, mandatory restrictions would
begin to be placed on the activities of undercapitalized banks beginning in
December 1992. And if banks adjusted to their target capital ratios and risk
levels slowly over time, then they may have begun adjusting their portfolios in
1992 to account for PCA. 6
This study covers the period after the passage of FDICIA, but before im-
plementation of PCA (1992), and the ®rst four years after the PCA standards
became e€ective (1993±1996). In addition, because risk-based capital data are
unavailable prior to the end of 1990, 1991 is used as a control period. 7 This
study examines 1685 FDIC-insured commercial banks with assets of $100
million or more in existence from year-end 1990 through year-end 1997, using
Call Report data. 8 The simultaneous equations model is estimated using the
3SLS procedure which recognizes the endogeneity of both bank capital ratios
and risk levels in a simultaneous equations framework. This study utilizes 3SLS
because it provides consistent estimates of the parameters, whereas ordinary
least squares estimates are both biased and inecient. In addition, because
3SLS is a full-information estimator, it produces parameter estimates that are
asymptotically more ecient than 2SLS. 9

6
In studying the implementation of the risk-based capital standards, Haubrich and Watchel
(1993) note that because the composition of bank portfolios can be changed quickly, and because
banks appeared to have experienced a period of learning, the impact of the risk-based capital
standards appeared more clearly after the implementation date. The same argument may be true
for PCA, although learning by banks should be less signi®cant with PCA because all of the capital
ratios that de®ne the thresholds had been measured since at least December 1990.
7
While our equations include some terms to isolate and assess the impact of PCA, a word of
caution is necessary. Any analysis of PCA is complicated by other factors present during this time
period, such as the implementation of the risk-based capital standards, the 1990±1991 credit
crunch, and other provisions of FDICIA, which may make it dicult to de®nitively assess the
impact of the PCA standards.
8
One year of data are lost because, as noted earlier, the nonperforming loan to total asset ratio
in a given year is used to measure credit risk in the previous year. Thus, 1997 data on
nonperforming loans are used to estimate the risk variable in 1996.
9
As noted by Intriligator (1978), 3SLS can be sensitive to speci®cation or measurement error
and, under these conditions, 2SLS may be preferable. Estimation of Eqs. (9) and (10) using 2SLS
does not signi®cantly alter the ®ndings of this study. Therefore, we use 3SLS because its estimators
are asymptotically more ecient than those produced using 2SLS. In addition, the use of the partial
adjustment model may introduce serial correlation into our error terms. 3SLS eliminates this
problem because, as Intriligator notes, the 3SLS technique can be interpreted as an extension of
generalized least squares (GLS) to a simultaneous equation system.
 R. Aggarwal, K.T. Jacques / Journal of Banking & Finance 25 (2001) 1139±1160 1149

5. Results

Table 2 shows the summary statistics for banks in our sample. An informal
observation of bank capital and risk data in the early 1990s indicate that not
only did aggregate capital levels increase, but that credit risk levels decreased as
well. The data also indicate that, as expected, banks classi®ed as undercapi-
talized, substantially undercapitalized, or critically undercapitalized (hereafter
referred to as ``undercapitalized'') increased their capital levels and reduced
their credit risk levels more than did either adequately capitalized or well
capitalized banks.
However, this informal assessment does not clarify if the increases in capital
ratios and decreases in risk were indeed due to PCA or due to the simultaneous
movement in some other variable such as bank income. Given the nature of the
yield curve and the benign environment for banking in the years following the
passage of FDICIA, it can be contended that the observed changes in bank
capital ratios and risks may be due to factors other than PCA. The more
comprehensive simultaneous equations model employed herein addresses these
concerns as it includes terms that account for these confounding factors in-
cluding bank income, size, holding company status, asset quality, liquidity, and
holdings of government securities.

5.1. 3SLS model estimates

The results of estimating the simultaneous system of Eqs. (9) and (10) are
presented in Tables 3 and 4. Table 3 provides the results using the Tier 1 le-
verage ratio as the measure of capital, while Table 4 measures capital using the
risk-based capital ratio. In both cases, risk is measured using RWATA. Tables
5 and 6, using the two capital ratios with NPL as the credit risk measure are
shown in Appendix B. 10 An examination of the results reveals that all of the
variables included to explain variations in capital ratios and risk levels are
statistically signi®cant in at least some of the equations.
Bank size (SIZE) had a negative e€ect on capital ratios and a positive
e€ect on risk levels. One possible interpretation of this ®nding is that larger
banks, because of their greater access to capital markets did not increase
capital ratios as much as smaller banks until the PCA standards became
e€ective. As hypothesized, multibank holding company status (BHC) ap-
pears to reduce capital ratios and increase risk levels. The income to asset
ratio (INC) was found to have a positive e€ect on bank capital ratios,

10
Estimation of the equation system using the Tier 1 risk-based capital ratio (the third of the
three capital requirements under PCA) is excluded for the sake of brevity. Results are similar to
those reported in Tables 3±6 and are available from the authors.
 1150

Table 2
Variable means
Variable 1991 1992 1993±1996
Well-cap- Adequately All under- Well-cap- Adequately All under- Well-cap- Adequately All under-
italized capitalized capitalized italized capitalized capitalized italized captitalized capitalized
RBCR 17.39% 10.16% 8.64% 17.16% 11.15% 9.77% 16.58% 10.55% 9.36%
T1LEV 8.52% 6.25% 5.28% 8.64% 6.70% 5.47% 9.01% 6.37% 5.56%
T1RBC 16.18% 8.75% 6.90% 15.92% 9.36% 8.03% 15.26% 9.02% 8.00%
RWATA 59.81 73.27 77.35 59.06 73.90 70.75 63.30 73.36 72.39
NPL 4.06 6.32 8.85 3.43 5.26 8.54 2.89 5.66 11.73
DRBCR 2.11% 0.76% 1.48% 0.14% 1.87% 2.91% )0.09% 1.03% 3.27%
DT1LEV 0.08% 0.61% 0.93% 0.22% 0.79% 1.39% 0.12% 0.48% 2.04%
DT1RBC 2.03% 0.77% 1.44% 0.12% 1.59% 2.58% )0.001% 1.00% 3.18%
DRWATA )6.57 )3.71 )3.39 )1.28 )3.48 )4.00 0.92 )1.96 )2.63
DNPL )0.10 0.28 )0.20 )0.79 )0.86 )2.79 )0.18 )0.83 )2.53
SIZE 12.36 13.31 14.06 12.47 13.82 13.49 12.88 13.21 12.34
BHC 0.36 0.56 0.57 0.38 0.58 0.32 0.45 0.29 0.11
INC 0.010 0.007 0.005 0.011 0.010 )0.001 0.012 0.003 )0.008
SEC 0.10 0.05 0.05 0.11 0.06 0.07 0.08 0.06 0.06
CASH 0.06 0.07 0.08 0.06 0.07 0.07 0.05 0.06 0.05
LLPROV 0.004 0.009 0.015 0.004 0.010 0.01 0.003 0.009 0.009
R. Aggarwal, K.T. Jacques / Journal of Banking & Finance 25 (2001) 1139±1160
 R. Aggarwal, K.T. Jacques / Journal of Banking & Finance 25 (2001) 1139±1160 1151

Table 3
Three-stage least square estimates of PCA on bank capital ratios (TILEV) and risk (RWATA)
Variable 1991 1992 1993±1996
DT1LEV DRWATA DT1LEV DRWATA DT1LEV DRWATA
Intercept 0.016 )0.030 0.012 0.070 0.013 0.061
(4.68) ()1.39) (3.69) (4.33) (7.62) (9.26)
SIZE )0.001 0.013 )0.001 0.003 )0.001 0.002
()1.89) (7.34) ()2.79) (2.22) ()5.36) (4.14)
BHC )0.002 0.004 )0.002 0.009 )0.003 0.008
()4.47) (1.18) ()2.94) (3.41) ()11.25) (6.40)
INC 0.585 ± 0.651 ± 0.438 ±
(21.38) (23.75) (23.43)
SEC 0.000 )0.231 0.001 )0.156 0.011 )0.081
(0.04) ()12.21) (0.38) ()11.39) (7.78) ()12.40)
CASH )0.007 )0.062 0.026 )0.085 0.020 )0.052
()1.38) ()1.83) (4.93) ()3.17) (5.24) ()3.57)
MSA 0.000 0.000 )0.000 )0.000 0.000 0.000
(0.58) (5.45) ()0.10) ()1.62) (2.33) (0.44)
LLPROV 0.072 1.393 0.189 0.313 )0.073 0.326
(2.09) (6.24) (5.64) (1.83) ()3.01) (3.33)
CAPt 1 )0.143 ± )0.112 ± )0.105 ±
()13.81) ()11.15) ()18.33)
RISKt 1 ± )0.271 ± )0.164 ± )0.121
()17.38) ()13.84) ()22.87)
DCAPITAL ± 0.772 ± )0.387 ± 0.849
(2.93) ()1.84) (6.29)
DRISK 0.026 ± 0.027 ± 0.035 ±
(3.02) (1.86) (3.15)
PCAA 0.001 0.022 0.015 )0.004 0.016 )0.021
(0.33) (4.64) (2.74) ()0.91) (2.83) ()4.99)
PCAU 0.024 0.012 0.057 )0.008 0.042 )0.037
(4.66) (1.22) (7.27) ()0.95) (6.24) ()4.61)
PCAACAPt 1 )0.023 ± )0.117 ± )0.122 ±
()0.57) ()2.04) ()1.99)
PCAUCAPt 1 )0.270 ± )0.618 ± )0.330 ±
()3.85) ()5.54) ()3.15)
System weighted 0.266 0.236 0.117
R2
*
Signi®cant at the 5% level.
**
Signi®cant at the 10% level.

suggesting that the increased pro®tability of commercial banks in the early

1990s played an important role in increasing bank capital ratios. Not sur-
prisingly, banks with greater holdings of government securities (SEC) also
had more capital in the post-1992 equations, and similar to the ®ndings for
CASH, banks with greater holdings had lower credit risk levels. Further-
more, banks with lower asset quality (higher LLPROV) had greater risk.
Finally, banks in standard metropolitian statistical areas (MSA) were sel-
1152 R. Aggarwal, K.T. Jacques / Journal of Banking & Finance 25 (2001) 1139±1160

Table 4
Three-stage least squares estimates of PCA on bank capital ratios (RBCR) and risk (RWATA)
Variable 1991 1992 1993±1996
DRBCR DRWATA DRBCR DRWATA DRBCR DRWATA
Intercept )0.011 )0.024 0.156 0.068 0.051 0.066
()0.54) ()1.06) (4.47) (4.31) (5.05) (9.82)
SIZE )0.001 0.013 )0.006 0.003 )0.002 0.002
()0.55) (7.06) ()2.21) (2.16) ()2.10) (3.88)
BHC )0.003 0.004 )0.025 0.009 )0.013 0.006
()0.87) (1.25) ()4.15) (3.43) ()7.48) (5.10)
INC 0.617 ± 2.185 ± 1.074 ±
(4.05) (8.71) (10.98)
SEC 0.020 )0.256 0.314 )0.156 0.139 )0.076
(1.26) ()12.43) (10.34) ()11.58) (14.26) ()11.44)
CASH )0.86 )0.074 0.200 )0.086 0.053 )0.043
()2.96) ()2.12) ()3.28) ()3.23) (2.30) ()2.89)
MSA 0.000 0.000 0.000 )0.000 )0.000 0.000
()0.19) (5.58) (1.39) ()1.57) ()0.11) (0.89)
LLPROV )0.212 1.368 0.690 0.302 0.102 0.185
()1.06) (5.89) (1.78) (1.79) (0.68) (1.90)
CAPt 1 0.215 ± )0.743 ± )0.335 ±
(29.47) ()65.57) ()41.17)
RISKt 1 ± )0.278 ± )0.160 ± )0.125
()17.42) ()13.71) ()22.94)
DCAPITAL ± 0.076 ± )0.013 ± 0.026
(1.96) ()2.05) (1.66)
DRISK )0.127 ± 1.161 ± 0.477 ±
()2.57) (6.66) (6.93)
PCAA 0.049 0.023 0.067 )0.007 0.039 )0.017
(1.95) (4.71) (1.31) ()1.57) (1.20) ()4.18)
PCAU 0.063 0.019 0.091 )0.012 0.045 )0.021
(2.13) (1.95) (1.24) ()1.58) (1.21) ()2.89)
PCAACAPt 1 )0.434 ± )0.641 ± )0.268 ±
()1.66) ()1.21) ()0.80)
PCAUCAPt 1 )0.536 ± )1.035 ± )0.188 ±
()1.34) ()0.99) ()0.33)
System 0.338 0.659 0.172
weighted R2
*
Signi®cant at the 5% level.
**
Signi®cant at the 10% level.

dom found to di€er in their changes in either capital or risk from banks in
rural areas.
The parameter estimates on lagged capital ratios and risk levels were gen-
erally negative and signi®cant, with estimates ranging from )0.105 to )0.743.
In general, these values imply slow adjustment of bank capital ratios and risk
to desired levels. In addition, Tables 3 and 4 show a predominantly positive
relationship between changes in capital and changes in credit risk, this being
 R. Aggarwal, K.T. Jacques / Journal of Banking & Finance 25 (2001) 1139±1160 1153

consistent with a number of hypotheses discussed earlier including the unin-

tended e€ect of PCA on bank risk levels.

5.2. Impact of PCA on bank capital ratios

The results in Tables 3 and 4 provide some rather interesting insights re-
garding the impact of the PCA provisions on changes in capital ratios and
credit risk. In the capital equations, the impact of the regulatory pressure
variables are captured by both an intercept term (PCAA or PCAU) and a
speed of adjustment term (PCAA  CAPt 1 or PCAU  CAPt 1 ). For ade-
quately capitalized banks, regulatory pressure brought about by the PCA
standards had a positive and signi®cant impact on capital ratios in both 1992
and after 1993 if capital is measured using the T1LEV, and no impact on
capital in either period if capital is measured using the RBCR. Furthermore,
the parameter estimate on PCAA in 1991, the control period, was insigni®cant
in when capital was measured using T1LEV, but signi®cant and positive when
capital was measured using RBCR. This suggests that adequately capitalized
banks were increasing their leverage ratio, but not necessarily their risk-based
capital ratio, in response to the PCA standards. One possible explanation is
that prior to FDICIA, adequately capital banks were still adjusting to the risk-
based capital ratio, but following the passage of FDICIA in 1991, these banks
began increasing their leverage ratio so as to become even better capitalized. In
addition, the magnitude of the parameter estimates on PCAA in Table 3 for the
1992 and 1993±1996 are similar, suggesting that the response by adequately
capitalized banks was similar for the period following the announcement of
PCA (1992) and the period when PCA was in e€ect.
Undercapitalized banks seem to behave similarly. Examining the coecients
on PCAU in Tables 3 and 4, undercapitalized banks showed a strong response
to PCA in their leverage ratios, but not in their risk-based capital ratios. The
parameter estimates on PCAU for 1991 in Table 3 equals 0.024 and in Table 4
equals 0.063, a result which suggests undercapitalized banks were still in-
creasing their capital ratios in response to the risk-based standards. In the 1992
and 1993±1996 periods, the parameter estimate on PCAU in Table 3 increases
dramatically to 0.057 and 0.042, respectively, while they are not signi®cant in
Table 4. These ®ndings suggest that independent of the adjustment to risk-
based capital, PCA was e€ective in getting undercapitalized institutions to
increase their leverage ratios. The signi®cance of PCAU during the 1992 and
1993±1996 periods is not surprising, but rather suggests that when PCA
standards were announced in December 1991, undercapitalized banks contin-
ued increasing their leverage ratios, recognizing that failure to meet adequately
capitalized standards by the beginning of 1993 could result in regulatory
sanctions.
1154 R. Aggarwal, K.T. Jacques / Journal of Banking & Finance 25 (2001) 1139±1160

A priori, undercapitalized banks would be expected to have a stronger

capital response to the PCA standards than adequately capitalized banks be-
cause more stringent regulatory sanctions could be placed on undercapitalized
banks. A comparison of the parameter estimates on PCAA and PCAU in
Table 3 shows that indeed in 1992 and 1993±1996 undercapitalized banks re-
sponse in their leverage ratios was at a rate roughly two to four times greater
than at adequately capitalized banks.

5.3. Speed of adjustment in the capital equations

In addition to a direct response to the regulatory pressure brought about by

the PCA, the standards may have also caused banks to increase the speed with
which they adjusted their capital ratios. Table 3 provides evidence that both
adequately capitalized and undercapitalized institutions exhibited signi®cantly
faster speeds of adjustment of their leverage ratios than did well-capitalized
banks, and that speed of adjustment increased markedly in 1992 and 1993±
1996. For adequately capitalized banks, the speed of adjustment coecient
(PCAA  CAPt 1 ) was not signi®cant in 1991, but increased to )0.117 in 1992
and )0.122 in 1993±1996. In addition, undercapitalized institutions exhibited
signi®cantly faster speeds of adjustment (PCAU  CAPt 1 ) than did well-cap-
italized banks, and their speed of adjustment increased dramatically after 1991.
Here the magnitude of the speed of adjustment coecients is more than twice
as large in 1992 as in 1991, and over twenty percent larger in 1993±1996 than in
1991. These ®ndings suggest that both undercapitalized and adequately capi-
talized banks increased the speed of adjustment of their leverage ratios to the
desired level after the announcement and implementation of PCA. This result
is to be expected for undercapitalized banks in particular because at the end of
1991 these banks had approximately one year to improve their capital position
or face possible regulatory sanctions as part of PCA.
Comparing the speed of adjustment coecients in Table 3 between ade-
quately capitalized and undercapitalized banks also provides some evidence of
di€erent adjustment speeds. Speci®cally, in 1992, undercapitalized banks
showed faster adjustment speeds of their leverage ratios ()0.618 vs. )0.117)
than their adequately capitalized counterparts. A similar result is true for the
1993±1996 period, but here the coecient on the interactive term equals
)0.330 for undercapitalized banks vs. only )0.122 for adequately capitalized
banks.

5.4. Impact of PCA on bank portfolio risk

With regard to the impact of the PCA standards on risk, the parameter
estimates for PCAA are similar regardless of the capital ratio used. In 1991, the
 R. Aggarwal, K.T. Jacques / Journal of Banking & Finance 25 (2001) 1139±1160 1155

parameter estimates on PCAA in the risk equations are positive and signi®cant
using both the Tier 1 leverage ratio and the total risk-based capital ratio. This
®nding suggests that adequately capitalized banks were increasing credit risk in
1991 prior to FDICIA. In 1992, the parameter estimates on PCAA are in-
signi®cant, while the estimates in 1993±1996 are negative and signi®cant
equaling )0.021 when capital is measured using the leverage ratio and )0.017
when capital is measured using the total risk-based capital ratio. Such a result
is not surprising because mandatory sanctions for banks being undercapital-
ized could be employed by regulators beginning in 1993, and these banks may
have decreased their risk levels to both improve their capital ratios and to
reduce their susceptibility to ¯uctuations in capital ratios due to exogenous
shocks.
With respect to the credit risk of undercapitalized institutions (PCAU),
the results in Tables 3 and 4 suggest that regulatory pressure brought about
by PCA also led them to signi®cantly decrease their level of risk. Similar to
adequately capitalized banks, the parameter estimates on PCAU are nega-
tive and signi®cant in 1993±1996, but not signi®cant in 1992. Thus, PCA
was e€ective in reducing credit risk only after 1993 when PCA became
e€ective.
It is also interesting to note that while both adequately capitalized
and undercapitalized banks had an incentive to reduce the credit risk in
their portfolio of assets, undercapitalized banks appear to have reduced
risk to a greater degree. In Table 3, the coecient on PCAU in the
1993±1996 risk equation equals )0.037 vs. )0.021 for PCAA, while in
Table 4 the coecients on PCAU and PCAA in the post-1993 risk
equations equal )0.021 and )0.017, respectively. Both sets of results
suggest that while adequately capitalized and undercapitalized banks re-
duced their level of risk, the response of undercapitalized banks appears
stronger.
In summary, the results in Tables 3 and 4 provide evidence that imple-
mentation of the PCA standards, complete with mandatory restrictions on the
activities of undercapitalized institutions, brought about signi®cant reductions
in the risk levels of both adequately capitalized and undercapitalized banks,
particularly once the standards went into e€ect in 1993. Given federal deposit
insurance and banks e€orts to attempt to circumvent new regulations, bank
regulatory policy must account for the tendency of banks to o€set new capital
requirements by making o€setting changes in risk. In this sense, the results in
this section suggest that the PCA provisions of FDICIA have been a success.
While the results do not necessarily imply that bank capital is adequate from a
public policy perspective, they do suggest that banks responded to PCA by
increasing their leverage ratios and reducing their credit risk levels, both
changes being associated with improvements in the safety and soundness of the
banking system.
1156 R. Aggarwal, K.T. Jacques / Journal of Banking & Finance 25 (2001) 1139±1160

6. Conclusions

In view of the continuing interest in bank regulation, this study documents

the e€ectiveness of a regulatory approach, the prompt corrective action
provision of FDICIA passed by the US Congress in 1991. Speci®cally, this
paper documents the bene®cial impacts of the prompt corrective action pro-
visions on bank capital ratios and credit risk levels. As prior research indicated
that banks may increase risk in response to changes in regulatory capital
standards, this study used a three stage least squares estimation of a simulta-
neous equation model to account for this simultaneity bias and the impact of
other confounding variables.
The results presented here show that both adequately capitalized and
undercapitalized banks increased their leverage ratios in response to PCA,
during both the announcement period, 1992, and the years after the standards
went into e€ect, 1993±1996. In addition, the ®ndings of this study provide
evidence that banks, both adequately capitalized and undercapitalized, in-
creased the speed with which they adjusted their leverage ratios to the desired
levels in 1992 and in 1993±1996. Finally, with respect to credit risk, the re-
sults for the 1993±1996 period suggest that banks also signi®cantly reduced
their level of credit risk in response to PCA. This is not surprising because
undercapitalized banks were faced with the possibility of signi®cant regula-
tory sanctions if they did not meet certain capital thresholds, while ade-
quately capitalized banks may have lessened risk to reduce the probability of
a signi®cant reduction in their capital ratios resulting from an exogenous
shock.
Taken as a whole, the results suggest that subsequent to the passage of
FDICIA in 1991, US banks increased their capital ratios without o€setting
increases in credit risk, thus suggesting that the PCA provisions of FDICIA
were e€ective. While these results do not guarantee that bank capital levels are
adequate, they suggest that PCA has been successful in increasing bank capital
ratios without the unintended e€ect of increasing risk as has been noted in
prior research.

Acknowledgements

The authors thank Richard Anderson, Drew Dahl, Stephen Hiemstra,

Stephen Kane, Thomas Lutton, David Nebhut, Peter Nigro, Swamy Parava-
stu, Tara Rice, Gordon Sick, Larry Wall, Ying Yan and two anonymous
referees for their helpful comments. A special thanks to James Barth for dis-
cussions that led to the idea behind this paper. David Roderer provided in-
valuable research assistance. Part of this paper was done while Mr. Jacques was
aliated with the Oce of the Comptroller of the Currency and while Raj
 R. Aggarwal, K.T. Jacques / Journal of Banking & Finance 25 (2001) 1139±1160 1157

Aggarwal was at John Carroll University. The authors alone are responsible
for the contents.

Appendix A. Restrictions on undercapitalized, substantially undercapitalized, and

critically undercapitalized banks

Restrictions on undercapitalized banks

 close monitoring by the regulatory agency;
 submission of a capital restoration plan;
 asset growth restrictions;
 restrictions on both expansion plans and new lines of business.
Restrictions on substantially undercapitalized banks
 required capital restoration by either selling stock, requiring the bank to be
acquired by a holding company, or merging the bank into another bank;
 restrictions on transactions with aliates and aliated banks;
 restrictions on interest rates paid on deposits;
 asset growth restrictions;
 restriction on bank activities that may pose excessive risk;
 changes in bank management;
 a prohibition on deposits from correspondent banks;
 prior approval required for capital distributions by the holding company;
 divestiture of an bank operating subsidiary or divestiture at the holding
company level;
 regulator may impose restrictions that the FDIC places on critically
undercapitalized banks if necessary.
Restrictions on critically undercapitalized banks
Without approval of the FDIC, the bank may not:
 enter into a material transaction;
 extend credit for a highly leveraged transaction;
 amend the bankÕs charter or bylaws, unless required to do so;
 make signi®cant changes in accounting methods;
 engage in certain covered transactions;
 pay excessive compensation or bonuses;
 pay excessive interest rates.

In addition, under Section 131 of FDICIA, all banks are prohibited from
paying dividends that would leave the bank undercapitalized and from paying
management fees to any person in control of the bank that would leave the
bank undercapitalized.
Source: Oce of the Comptroller of the Currency.
1158 R. Aggarwal, K.T. Jacques / Journal of Banking & Finance 25 (2001) 1139±1160

Appendix B

See Tables 5 and 6.

Table 5
Three-stage least squares estimates of PCA on bank capital ratios (T1LEV) and risk (NPL)
Variable 1991 1992 1993±1996
DT1LEV DNPL DT1LEV DNPL DT1LEV DNPL
Intercept 0.011 )0.006 0.014 0.018 0.015 0.012
(3.27) ()1.04) (4.24) (2.72) (8.32) (5.64)
SIZE )0.001 0.001 )0.001 )0.001 )0.001 )0.000
()0.86) (1.92) ()3.41) ()1.87) ()5.55) ()0.10)
BHC )0.002 )0.002 )0.002 )0.003 )0.003 )0.001
()4.42) ()1.53) ()3.10) ()2.22) ()10.59) ()1.47)
INC 0.608 ± 0.665 ± 0.462 ±
(21.47) (22.31) (24.45)
SEC )0.001 )0.011 )0.000 )0.007 0.011 )0.012
()0.50) ()2.21) ()0.02) ()1.40) (7.46) ()5.99)
CASH )0.004 )0.014 0.026 )0.005 0.019 )0.019
()0.88) ()1.41) (4.51) ()1.41) (4.84) ()3.85)
MSA 0.000 0.000 )0.000 0.000 0.000 )0.000
(1.45) (2.26) ()0.31) (1.11) (1.59) ()1.97)
LLPROV 0.071 0.275 0.134 0.254 )0.134 0.285
(1.97) (3.64) (3.60) (3.11) ()5.29) (7.93)
CAPt 1 )0.154 ± )0.121 ± )0.113 ±
()15.13) ()12.20) ()19.66)
RISKt 1 ± )0.313 ± )0.256 ± )0.364
()19.65) ()13.42) ()46.02)
DCAPITAL ± )0.097 ± )0.131 ± )0.069
()1.22) ()1.47) ()1.53)
DRISK )0.063 ± )0.212 ± )0.120 ±
()2.39) ()5.84) ()7.00)
PCAA )0.001 0.000 0.012 0.003 0.012 )0.001
()0.28) (0.05) (2.25) (1.65) (2.03) ()0.79)
PCAU )0.237 )0.005 0.049 0.003 0.035 )0.007
()3.35) ()1.81) (6.31) (0.67) (5.19) ()2.73)
PCAACAPt 1 0.004 ± )0.088 ± )0.093 ±
(0.09) ()1.59) ()1.49)
PCAUCAPt 1 )0.237 ± )0.538 ± )0.307 ±
()3.35) ()4.93) ()2.92)

System weighted R2 0.262 0.231 0.224

*
Signi®cant at the 5% level.
**
Signi®cant at the 10% level.
 R. Aggarwal, K.T. Jacques / Journal of Banking & Finance 25 (2001) 1139±1160 1159

Table 6
Three-stage least squares estimates of PCA on bank capital ratios (RBCR) and risk (NPL)
Variable 1991 1992 1993±1996
DRBCR DNPL DRBCR DNPL DRBCR DNPL
Intercept 0.013 )0.007 0.158 0.017* 0.058 0.012
(0.67) ()1.12) (5.60) (2.59) (6.34) (5.55)
SIZE )0.002 0.001 )0.007 )0.001 )0.002 )0.000
()1.16) (1.94) ()2.98) ()1.77) ()2.52) ()0.02)
BHC )0.003 )0.002 )0.018 )0.003 )0.012 )0.001
()0.82) ()1.49) ()3.68) ()2.28) ()7.16) ()1.15)
INC 0.489 ± 2.195 ± 0.966 ±
(2.96) (8.91) (9.63)
SEC 0.031 )0.010 0.230 )0.007 0.132 )0.012
(1.99) ()1.90) (10.34) ()1.32) (14.93) ()6.38)
CASH )0.099 )0.012 0.169 )0.007 0.041 )0.020
()3.32) ()1.23) (3.45) ()0.56) (1.93) ()4.07)
MSA )0.000 0.000 )0.000 0.000 )0.000 )0.000
()0.43) (2.17) ()0.40) (1.20) ()0.64) ()2.08)
LLPROV )0.290 0.280 0.231 0.257 )0.241 0.302
()1.37) (3.71) (0.73) (3.15) ()1.76) (8.77)
CAPt 1 0.211 ± )0.718 ± )0.324 ±
(28.68) ()82.77) ()44.75)
RISKt 1 ± )0.311 ± )0.256 ± )0.366
()19.63) ()13.46) ()46.34)
DCAPITAL ± 0.001 ± )0.001 ± 0.003
(0.05) ()0.02) (0.53)
DRISK 0.059 ± )0.253 ± )0.210 ±
(0.37) ()0.81) ()2.24)
PCAA 0.042 )0.001 )0.023 0.002 )0.008 )0.001
(1.56) ()0.04) ()0.45) (1.29) ()0.24) ()0.98)
PCAU 0.081 )0.006 0.040 0.001 0.026 )0.009
(2.59) ()2.23) (0.54) (0.27) (0.70) ()3.32)
PCAACAPt 1 )0.374 ± 0.077 ± 0.075 ±
()1.34) (0.14) (0.22)
PCAUCAPt 1 )0.794 ± )0.710 ± )0.269 ±
()1.88) ()0.68) ()0.46)
System weighted R2 0.318 0.689 0.261
*
Signi®cant at the 5% level.
**
Signi®cant at the 10% level.

References

Avery, R.B., Berger, A.N., 1991. Risk-based capital and deposit insurance reform. Journal of
Banking and Finance 15, 847±874.
Baer, H.L., McElravey, J.N., 1993. Capital shocks and bank growth: 1973 to 1991. Economic
Perspectives, Federal Reserve Bank of Chicago 17, 1±21.
 1160 R. Aggarwal, K.T. Jacques / Journal of Banking & Finance 25 (2001) 1139±1160

Benston, G.J., Kaufman, G.G., 1997. FDICIA after ®ve years. Journal of Economic Perspectives
11, 139±158.
Berger, A.N., 1995. The relationship between capital and earnings in banking. Journal of Money,
Credit, and Banking 27, 432±456.
Berger, A.N., Udell, G.F., 1993. Securitization, risk and the liquidity problem in banking. In:
Klausner, M., White, L.J. (Eds.), Structural Changes in Banking. Homewood, IL, pp. 227±291.
Blum, J., 1999. Do capital requirements reduce risks in banking? Journal of Banking and Finance
23, 755±771.
Buser, S.A., Chen, A.H., Kane, E.J, 1981. Federal deposit insurance, regulatory policy, and
optimal bank capital. Journal of Finance 35, 51±60.
Davies, S.M., McManus, D.A., 1991. The e€ects of closure policies on bank risk-taking. Journal of
Banking and Finance 15, 917±938.
Furlong, F.T., 1992. Capital regulation and bank lending. Economic Review, Federal Reserve
Bank of San Francisco 3, 23±33.
Gennotte, G., Pyle, D., 1991. Capital controls and bank risk. Journal of Banking and Finance 15,
805±824.
Haubrich, J.G., Watchel, P., 1993. Capital requirements and shifts in commercial bank portfolios.
Economic Review, Federal Reserve Bank of Cleveland 3, 2±15.
Hovakimian, A., Kane, E.J., 2000. E€ectiveness of capital regulation at US commercial Banks,
1985 to 1994. Journal of Finance 55, 451±468.
Intriligator, M.D., 1978. Econometric Models Techniques and Applications. Prentice-Hall,
Englewood Cli€s, NJ.
Jacques, K., Nigro, P., 1997. Risk-based capital, portfolio risk, and bank capital: A simultaneous
equations approach. Journal of Economics and Business 49, 533±547.
Jones, D.S., King, K.K., 1992. The implementation of prompt corrective action. In: Proceedings of
the Conference on Bank Structure and Competition, Federal Reserve Bank of Chicago, pp. 68-
100.
Jones, D., King, K.K., 1995. The implementation of prompt corrective action: An assessment.
Journal of Banking and Finance 19, 491±510.
Kahane, Y., 1977. Capital adequacy and the regulation of ®nancial intermediaries. Journal of
Banking and Finance 1, 207±217.
Kane, E., 1983. A six-point program for deposit-insurance reform. Housing Finance Review 2,
269±278.
Kim, D., Santomero, A.M., 1988. Risk in banking and capital regulation. Journal of Finance 43,
1219±1233.
Koehn, M., Santomero, A.M., 1980. Regulation of bank capital and portfolio risk. Journal of
Finance 35, 1235±1250.
Levonian, M.E., 1991. What happens if banks are closed ÔearlyÕ? In: Proceedings of the Conference
on Bank Structure and Competition, Federal Reserve Bank of Chicago, pp. 273±295.
Peek, J., Rosengren, E.S., 1996. The use of capital ratios to trigger intervention in problem banks: Too
little, too late. New England Economic Review, Federal Reserve Bank of Boston, pp. 49-58.
Peek, J., Rosengren, E.S., 1997. Will legislated early intervention prevent the next banking crisis?.
Southern Economic Journal 64, 268±280.
Shrieves, R.E., Dahl, D., 1992. The relationship between risk and capital in commercial banks.
Journal of Banking and Finance 16, 439±457.
Shrieves, R.E., Dahl, D., 1995. Regulation, recession, and bank lending behavior: The 1990 credit
crunch. Journal of Financial Services Research 9, 5±30.
Wall, L.D., Peterson, D.R., 1987. The e€ect of capital adequacy guidelines on large bank holding
companies. Journal of Banking and Finance 11, 581±600.
Wall, L.D., Peterson, D.R., 1995. Bank holding company capital targets in the early 1990s: The
regulators versus the market. Journal of Banking and Finance 19, 563±574.

View publication stats