Home > Industry Analysis > Research & Analysis > FDIC Working Papers Series




FDIC Working Papers Series


Regulator Use of Market Data to Improve the Identification of Bank Financial Distress

December 2001
Working Paper 2001-01

Timothy J. Curry
Financial Economist
Federal Deposit Insurance Corporation
Washington, D.C. 20429
(202) 898-7372
tcurry@fdic.gov

Peter J. Elmer

Gary S. Fissel
Financial Economist
Federal Deposit Insurance Corporation
Washington, D.C. 20429
(202) 898-3949
gfissel@fdic.gov


1 Virtually all Call Report data are released to the public, typically about 75 days (10-11 weeks) after the end of each quarter to which they apply. Approximately the same data as those released to the public are made available internally to regulators about 2-3 weeks earlier. Although regulators receive the data a little earlier than the public, it is nevertheless possible that the market can process the information at a faster rate than regulators upon release.
Return to Document

2 It has been shown that there is a correlation between bank examinations and commercial bank write-off of assets and increased loan provisioning. This suggests that some institutions do hide "bad" news from the public in their financial statements until forced to make changes by the regulators. See Dahl, O'Keefe, and Hanweck (2000).
Return to Document

3 In the late 1990s, a sixth component was added to the CAMEL rating system, recognizing bank and thrift sensitivity to interest rate or market risk. But because the empirical portions of our analysis relate to ratings performed before the late 1990s, we reference the five-component rating system in effect at that time.
Return to Document

4 See Cole and Gunther (1998) for a discussion of off-site monitoring systems.
Return to Document

5 The Financial Institutions Reform, Recovery, and Enforcement Act of 1989 (FIRREA) mandated annual examinations for large banks and all banks with unsatisfactory supervisory ratings. Since then, the examination schedule for most banking organizations has been stretched out to approximately every 18 months.
Return to Document

6 For a detailed discussion of the bank examination process, see Curry et al. (1997), 463-475.
Return to Document

7 Informal enforcement actions are usually issued within three months after completion of the examination that leads to the downgrade and may require institutions to make changes, such as raising new equity capital, limiting the origination of certain types of loans, or increasing loan-loss reserves. Although regulators vary in their practices, the most common type of informal action accompanying a downgrade to 3 is a "memorandum of understanding" (MOU), which is written by bank supervisors and signed by bank officials and supervisors. MOUs specify activities that must be undertaken by the bank, time frames for implementing the new procedures, and special requirements for reporting to the bank's supervisor. A second type of agreement, known as a "board resolution," is drafted by the individual bank and signed by each member of the bank's board of directors; it commits the institution to a certain course of action. Since 1983, the FDIC has informed the banks it supervises of their composite CAMEL ratings as part of the FDIC's report to bank management on the results of the examination. The Comptroller of the Currency and the Federal Reserve began revealing their ratings in December 1988.
Return to Document

8 Supervisors do, however, provide individual component ratings on a bank's performance in the five categories of capital, assets, management, earnings, and liquidity. The overall, or composite, rating is based primarily on the ratings for each of the individual categories.
Return to Document

9 Formal enforcement actions are stringent legal decrees that are enforceable in courts and often carry heavy penalties for noncompliance. They are usually issued within three to nine months of the completion of the bank examination that resulted in reclassification to a rating of 4 or 5. After FIRREA, formal enforcement actions become part of the public record when issued. As noted by Curry et al. (1999), during the 1980-1994 period, 89 percent of all formal enforcement actions were imposed on banks with ratings of 4 or 5.
Return to Document

10 See Jordan, Peek, and Rosengren (2000).
Return to Document

11 French, Schwert, and Stambaugh (1987) lend empirical support to this view by documenting a positive relation between the volatility of market returns and market excess returns (market return minus T-bill yield).
Return to Document

12 Gallant, Rossi, and Tauchen (1992) find supportive empirical evidence that large daily price movements are followed by high trading volume.
Return to Document

13 In the case of thrifts, 1989 was a watershed year because of the passage of FIRREA, which provided the funds needed to resolve the thrift crisis and contained provisions (such as higher capital requirements) that improved the safety and soundness of financial institutions. See Gupta and Misra (1999) for an overview of changes made to the banking system throughout the 1980s and in the early 1990s. Although banks did not experience the same depth of problems as thrifts, the late 1980s nevertheless marked an important change in bank regulation because of a significant increase in the regularity of bank examinations as well as other new requirements. For example, FIRREA required annual examinations for banks with assets over $250 million or banks that had poor ratings; it also required that regulators take prompt corrective action for undercapitalized institutions. As noted, the examination requirement has since changed for most institutions: a periodic examination is now required approximately every 18 months.
Return to Document

14 This correspondence is important because the public equity of banks held by holding companies is typically issued at the holding-company level, whereas detailed Call Report financial data are reported at the bank level. Banks are also distinguished from their holding companies in bankruptcy, because individual banks are taken over by the Federal Deposit Insurance Corporation whereas their holding companies fall under the purview of standard bankruptcy law.
Return to Document

15 Analysis of multibank-holding-company stocks carries disadvantages as well as advantages compared with analysis of single-bank holding companies and nonholding companies. For example, multibank holding companies tend to be large institutions that are widely traded and rated by nationally recognized rating agencies. Although single-bank holding companies and banks not affiliated with holding companies tend to have the opposite characteristics, their Call Report data nevertheless correspond directly to the institution that is publicly traded, and their financial data are far more extensive than financial data released at the holding company level. Moreover, the many activities of holding-company subsidiaries cannot be separated from the aggregated data reported at the holding company level, and this lack of separability obscures the extensive information released by individual banks. Market signals at the holding-company level may or may not correspond to the performance of the bank subsidiary. The potential disconnect between the performance of individual banks and the market signals of their holding companies may widen as holding companies diversify into additional nonbank activities subsequent to the passage of the Gramm-Leach-Bliley Act of 1999.
Return to Document

16 Examinations that lead to rating downgrades typically take one to two weeks to complete, and they conclude with a notification to management that the institution's rating has been downgraded. Thus, the zero quarter can be regarded as approximately contemporaneous with the rating change.
Return to Document

17 The industry value-weighted index was created from approximately 2,200 banking institutions that could be identified on the CRSP tapes and tied back to their specific charters. Separate value-weighted indexes were created for banks and for thrifts using the CRSP utility for creating value-weighted indexes (DSXPORT). At the beginning of each year, the sample of banks or thrifts was established; then the index was calculated for that year. The final index combined the yearly indexes into a continuous long-term series.
Return to Document

18 The Pettway study examined only six large banking organizations.
Return to Document

19 As institutions approach failure, they may also drop out of the sample because their stock prices are dropped from CRSP, given the de-listing rules of the various exchanges. In our sample, the most common reasons for de-listing were insufficient number of market makers and insufficient capital.
Return to Document

20 Logistic regression has been used extensively in this type of analysis, especially by Sinkey (1975), Elmer and Borowski (1988), Gajewski (1989), and Cole and Gunther (1995 and 1998).
Return to Document

21 EQ_AS and NC_RES are independent variables that have high correlation values. These high correlations range between -.82530 and -.85646 in specifications 1 to 3 of Table 8.
Return to Document

22 See Pindyck and Rubinfeld (1991), 240, for a discussion of the likelihood ratio test. See Greene (2000), 306, for a discussion of the AIC.
Return to Document

23 It should be noted that the decline in dividend payments to shareholders of downgraded banks and thrifts as reflected in all specifications in Table 8 could be the result of a combination of both management decisions to reduce dividends because of financial problems as well as regulatory orders imposed by bank supervisors.
Return to Document

24 There is an inverse relationship between the level of the AIC value and the effectiveness of the model. The lower the value, the more effective the model.
Return to Document

25 The logit models are tested for their accuracy of classifications on both in-sample and out-of-sample data. The in-sample data refer to the data set for the periods used to construct the model from the - 4 quarter before the downgrades. The estimated logit model for the in-sample data was run on 94 observations for the pre-3-rated group and 114 for the pre-4/5-rated group to test the effectiveness of the model in accurately classifying the observations. The out-of-sample tests were run on the remaining 20 percent of the sample-specifically, 32 observations for the pre-3-rated group and 41 for the pre-4/5-rated group-to determine the accuracy of the forecasts or classifications. Each of the observations for the pre-period was matched against a highly rated institution in the regression models. For the post-period, the in-sample tests amounted to 91 observations for the 3-rated group and 108 for the 4/5-rated group. The out-of sample tests were conducted on 31 observations for the post-3-rated group and 36 for the post-4/5-rated groups.
Return to Document

26 These critical probabilities are derived from the ratio of the regression sample that experienced further financial distress to the total sample that was used in the regression.
Return to Document

Last Updated 03/01/2002 insurance-research@fdic.gov