Printer Friendly

Market reaction to bond downgradings followed by Chapter 11 filings.

What is the impact of a firm's bond downgrading on its stock price? When bond downgradings occur, can the stock market distinguish between those companies that later file for Chapter 11 and those that do not? If the stock market can indeed distinguish between the two cases, how early do the prices reflect it? Our paper will try to answer these interesting questions.

An investigation into the information value of rating companies' services is not new. Katz |10~, in one of the earliest works on bond rating changes, develops an event-oriented methodology for testing the efficiency of the bond market.(1) He concludes that no anticipation exists prior to a public announcement of a rating change. After the rating change, there is a lag of six to ten weeks before the yield-to-maturity fully adjusts to the new rating classification. Weinstein |14~ investigates whether bond rating changes contain new information by studying the bonds' prices during the time period surrounding the announcement of the change in ratings.(2) He concludes that the rating changes cause no significant price change during or after the announcement, and that adjustments in the market are made 18 to six months before the event. Thus, his study suggests that changes in ratings provide no new information. Pinches and Singleton |13~ study the effects of changes in bond ratings on the returns of stocks during the period from January 1950 to September 1972. For each stock, they derive its market return based on the stock's beta and measure the actual return against the expected return for a period of 30 months before to 12 months after a rating after a rating change. They conclude that the downgrading of bond, bonds provides no new information to the market. Finally, Griffin and Sanvincente |8~ use three different methodologies to study the effects of changes in ratings on common stock prices.(3) They find that although the upgradings of bonds have no effect on stock prices, downgradings do have a significant effect.

Gilson, John, and Lang |6~ study the market's ability to predict the success of financially distressed firms in restructuring their debt versus filing for bankruptcy. Using two-day mean market-model stock residuals from 1981-1985, they find that a negative average announcement return of -1.6% is associated with firms who successfully restructure their debt. A negative announcement return of -6.3, however, is associated with firms whose restructuring attempts ultimately fail. Their results indicate that the market, on average, correctly predicts firms' ability to restructure their debt.

Altman and Kao |1~ study the "drift" associated with rating changes for new bond issues. They find that there is a negative drift associated with an original issue that is subsequently downgraded, and a positive drift associated with an issue that is subsequently upgraded. These results are interesting in light of the empirical findings of our study, which show that the market, on average, is able to distinguish between firms experiencing identical downgradings where one subsequently files for bankruptcy and the other does not. Altman and Kao find that the "drift" of the downgrading event is much stronger than the "drift" of the upgrading event. Thus, rating agencies appear to be less responsive than the market when re-rating firms.

Hand, Holthausen, and Leftwich |9~ analyze the effect of bond rating agency announcements on bond and stock prices. They also analyze the daily excess returns associated with the announcement of additions to Standard & Poor's Credit Watch list. They find that there is no announcement effect for additions to the S&P Credit Watch list, but there is a significant excess return when the changes in the S&P Credit Watch list are classified as "unexpected."

Using the event-study methodology, we examine the announcement effect of the "first consistent downgradings" of bonds on stock prices. As later defined in more detail, the "first consistent downgradings" are the earliest in a series of downgradings, with no upgradings during the sample period. We also distinguish between firms whose bonds are downgraded and who later file for Chapter 11 and those firms with identical downgrades that do not file for Chapter 11.

We find that the market, on average, is able to distinguish between firms having identical downgradings when one subsequently files for Chapter 11 and the other does not. This implies that the market behaves as if it assigns an even lower grade to the bonds of the company that later files for Chapter 11 than was assigned during the down-grading downgrading by Standard & Poor's or Moody's. Because we find empirically that almost two years elapse between the first consistent downgrading and the Chapter 11 filing, our results point to a high degree of market insight into the future prospects of companies.

The methodology employed in this study is set forth in Section I. Section II provides the empirical results. Concluding remarks are given in Section III.

I. Research Design and Methodology

To study the impact of the downgrading of bonds on stock prices, an event-study methodology is used. We use the daily mean-adjusted excess return model as discussed by Fama |4~ and Masulis |11~. The event is defined as the "first consistent downgrading" of the bond.

Depending on the flow of relevant information to the rating agencies, the bond ratings of a company can either be upgraded, downgraded, or left unchanged. Over time, both upgrades and downgrades can be experienced for the same bond as the fortunes of the company change. In this study, our focus is on situations of financial distress and the ability of rating agencies to capture all relevant information known to the market. Therefore, we are interested in patterns of bond downgrades. To avoid contamination of the data, we restrict ourselves to downgradings which are free of the ambiguity that would be caused by an intervening bond upgrade. The "first consistent downgrading," then, is defined as the earliest downgrading after which there are no upgrades (although there can be additional downgrades) until the company either files for Chapter 11 bankruptcy or the study period ends. Prior to the date of the first consistent downgrading, either the bond rating was not changed from its original value or the previous change was an upgrade. Subsequent to the date of the first consistent downgrading, either the bond rating was not changed or it was downgraded further. The first consistent downgrading, then, is the initial warning from the rating agency that the company's financial condition is starting to deteriorate from previous expectations. Subsequent downgrading might reinforce the concerns about financial distress, but the first consistent downgrading is the initial signal. Hence, it is the appropriate event on which to concentrate.

Our research focuses on both financial distress and bankruptcy, which are long-term processes. Therefore, we select a two-year period (510 business days) ending 11 days before the event as the comparison period. Thus, our comparison period is 510 days to 11 days before the change in the bond's rating. The actual event, the change in the bond's rating, is taken as the 21-day period beginning ten days before and ending ten days after the downgrade announcement and the downgrade date itself.

Two samples are analyzed. The first sample contains companies listed on the New York or American Stock Exchange that experience a bond downgrade and subsequently file for Chapter 11. The other group constitutes a "matching sample" of firms experiencing an identical downgrading of their bonds but which do not later file for Chapter 11. We test the following three hypotheses:

(i) |Mathematical Expression Omitted~, where |Mathematical Expression Omitted~ denotes the mean excess return and the superscript "11" indicates the firm which files for Chapter 11 following the bond downgrade.

(ii) |Mathematical Expression Omitted~, where |Mathematical Expression Omitted~ denotes the mean excess return and the superscript "N" indicates the firms which do not file for Chapter 11 following the bond downgrade at the time when the firms in the Chapter 11 group do.

(iii) |Mathematical Expression Omitted~; that is, we test the equality of the means of the two groups.

To be more specific, if a firm's bond is downgraded from AA to B and it subsequently files for Chapter 11, we look at the monthly S&P Bond Guide for another firm (matching sample) which is downgraded from AA to B but which does not file for Chapter 11. If more than one firm experiencing the same downgrading is found, we select the one that is in the closest related industry. We hypothesize that the cumulative residuals of the bankrupt and matching groups are statistically different, implying a differential announcement effect for similar bond rating changes.

Finally, we separate the Chapter 11 group into two subgroups: subgroup 1 consists of companies that file for Chapter 11 and are later reorganized, and subgroup 2 consists of companies that file for Chapter 11 but are later liquidated. If, indeed, the market discriminates between firms filing for Chapter 11 followed by liquidation from those filing for Chapter 11 and overcoming the financial distress by reorganization, then the null hypothesis of no difference will be rejected. Because our sample of liquidated firms is small, we only indicate the hypothesis and some preliminary results. This issue certainly needs further research.

II. Data and Results

A. The Data

To study the stock's excess return as a company approaches bankruptcy, we select a sample of bonds based on two criteria. First, the bonds must be publicly traded and listed in the Standard & Poor's Bond Guide with a bond rating from either S&P's or Moody's. Second, the companies file for Chapter 11 between September 1977 and October 1988. Fifty corporate bonds had bond ratings and other available data from S&P's or Moody's adequate for our purposes. Exhibit 1 provides the number of months between the first downgrading by the two rating service companies and the filing for Chapter 11 protection.

In conducting the event study of the effect of the bond downgrading on the daily rates of return on stocks, we first determine the exact press release dates of the bond rating changes.(4) Because we could not find relevant data for two continuous years from the CRSP Daily Return Tape for all 50 companies in the matched sample (i.e., firms which did not file for Chapter 11), some of the observations in our sample are omitted. This leaves us with 22 pairs of bankrupt and nonbankrupt companies. Thus, for each firm in the bankrupt group we have a nonbankrupt firm in the matching sample which has an equivalent downgrading during the same time period.

B. The Event-Study Results

Exhibits 2 through 4 give the statistical results of our study. The null hypothesis asserts that the average residual at the event day (the bond's downgrading date), |E.sub.1~(0), is zero for the Chapter 11 group. As shown in Exhibit 2, the t-statistic for testing the null hypothesis is -21.81. Thus, TABULAR DATA OMITTED the hypothesis is strongly rejected. For the matching sample, the t-statistic is only -1.48. This indicates that the negative abnormal return is much less pronounced for the firms experiencing bond downgrades which are not followed by a Chapter 11 filing. The negative sign on both t-statistics suggests that bond downgradings negatively affect shareholder wealth, which is consistent with our expectations.

The t-statistic for testing the equality of the mean excess return in the Chapter 11 and matching samples is -15.54. It is significant at the 0.5% level, so the null hypothesis that the two means are equal is rejected. This implies that the daily excess return of companies in the Chapter 11 sample is significantly lower than the corresponding residual return of the non-Chapter 11 group with the same bond downgrading.

Because the firms in the Chapter 11 and matching samples are selected such that both have identical rating changes, the fact that these two samples reveal different residual returns around the downgrading suggests that while the rating agencies provide new information to the equity market, this information is not complete. Even though the bonds are downgraded to the same degree, the market's stronger reaction to the downgrading of bonds for companies in the Chapter 11 sample suggests that the market knows in advance that these companies are in worse financial shape than the companies in the matching sample. Therefore, because of this stronger market reaction and because of the fact that those companies later file for bankruptcy, it appears that the rating agencies should have downgraded those companies' bonds even further to include the bankruptcy risk that the market itself predicts. Since they do not downgrade these bonds more than they do the bonds of companies which do not later file for bankruptcy, our study suggests that the rating agencies were not as insightful as the mature investors. Their ability to distinguish between the two groups is particularly interesting in light of the fact that, on average, the date of the Chapter 11 filing is almost two years after the bond's first consistent downgrading date.

When we examine the residual return on firms which are reorganized (subgroup 1) and those that are liquidated (subgroup 2), we find that the mean residual return is -3.29% for companies which are later liquidated, and -2.92% for companies which are reorganized. The difference between the two means is not significant (with a t-value of -0.32). A larger sample of liquidated firms is needed to be able to analyze whether the market can distinguish between these two groups.


Thus, Exhibit 2 indicates that even though the rating companies assign the same downgrading to two firms, the market is able to distinguish those with severe financial distress which ultimately file for Chapter 11 from those firms which do not file. These results are particularly strong since the event is defined as the first consistent bond downgrading, which is, on average, 23.06 months before the date of the Chapter 11 filing. Thus, the market "refines," so to speak, the rating service companies downgrading a long time before the severe financial distress (Chapter 11) is revealed.(5)

Exhibit 3 provides the daily residual return of stocks and cumulative residual return in the 21 days surrounding the downgrading announcement for the bankrupt group and for the matching sample, respectively. It is obvious from Exhibits 3, 4 and 5, that the market is able to differentiate between the bankrupt firms and the matching sample. While the bankrupt group is characterized by many significant negative t-values, the matching sample shows only a few negative values. Exhibit 5 shows diagrammatically the cumulative residual for both the Chapter 11 and the matching samples. It shows that the downgrades have a definite negative impact on the stock's daily rate of return. More importantly, the relation between the two curves drawn in Exhibit 5 indicates that the cumulative impact of the first bond downgrading is much more TABULAR DATA OMITTED severe for companies that later file for Chapter 11 than for those that do not.

Looking only at the lower curve in Exhibit 5, one may conclude that the rating services do provide important new information since the residuals are negative around the event date. The upper curve, though, reveals just the opposite -- namely, the downgrading does not provide any new information to the market. Because the market seems to implicitly rate the bonds more accurately than Moody's or Standard & Poor's, we conclude that, overall, the rating agencies provide only limited new information to investors, and that, nevertheless, the information they provide is important.

These results are provocative. If, in fact, the market is able to assess the higher bankruptcy potential for the Chapter 11 sample, why is it not already incorporated into the stock price? In other words, why does the bond downgrading TABULAR DATA OMITTED itself stimulate a strong negative reaction if the superior knowledge already exists in the market? We suggest that one possible explanation is due to "cue redundancy" (see Einhorn, Kleinmuntz, and Kleinmuntz |3~).

When decisions have to be made in the absence of all needed information (that is, under uncertainty), managers or investors usually base their decisions on available data that are known to be associated with or indicative of the value of the unknown but vital information. A series of consistent confirming data points -- redundant cues--finally reach some threshold so that the manager or investor is willing to act as if the missing information were known. This is the theory of cue redundancy.

In the present study, the market may have information pointing to a concern about the continuing viability of a company, but not quite enough to inspire action. The bond downgrading, then, may be the independent confirming data point required to cross the threshold, or to reinforce the conclusion of looming financial distress, and thus the market reacts accordingly. For companies in the non-Chapter 11 sample, evidence of future financial difficulties may not be as strong (less consistency or redundancy of cues), so the bond downgrading by itself is not sufficient to stimulate action.

III. Concluding Remarks

The response of the financial markets to a firm's bankruptcy is investigated. We examine how bond downgradings impact the market as measured by the daily residuals of a company's stock rate of return. We distinguish between the downgrading of bonds followed by Chapter 11 filings and similar bond downgradings which are not followed by Chapter 11. The main conclusion of this paper is: An event-study of the stock market reveals that for firms subsequently filing for bankruptcy, a negative abnormal return exists in the event period and, in particular, on the event date (t = 0). Therefore, the downgrading of the bonds, on average, conveys new information to the market resulting in a negative excess return. In this respect, the agencies provide important information to the market. However, when we take a matching sample with identical bond downgradings which are not followed by Chapter 11, we find that for the matching sample the excess return is almost zero (in comparison with about -3% for the Chapter 11 group). Thus, the stock market differentiates between two identical downgradings. This implies that the agency rating services do not provide sufficiently refined ratings, or are unable to distinguish between the two evolutionary patterns of financial distress.

When the cumulative excess return is measured, we find that in the entire event period (-10 to +10 days), it is -8.97% for the Chapter 11 group and approximately zero for the matching sample. Even though we have identical downgradings for the two samples, the market "assigns" a much sharper downgrading to firms which, two years later, indeed, file for Chapter 11. This difference can be explained by the theory of cue redundancy. In summary, while the rating agencies convey important information, it seems that the errors and omissions involved with the bond ratings are not negligible.

1 His data consists of electric utilities bonds from 1966 to 1972. He looks for "unusual behavior" in a bond's yield-to-maturity 12 months prior to and five months after a rating change.

2 His sample consists of utilities and industrial bonds from July 1962 to July 1974. Weinstein starts with portfolios which, for every month, contain every bond in his sample that had a given rating in that month. He then constructs a series of risk-adjusted returns for each bond by subtracting the return on the appropriate rating class portfolio from the return on the given bond. He selects the bonds that have a rating change and examines whether those bonds have abnormal returns during time periods around rating changes.

3 Their study contains 180 rating changes from 1960 to 1975. First, they use a portfolio method similar to that of Weinstein |14~. Then they employ a one-factor model and a two-factor model, basing their expected stock prices on beta, as employed by Pinches and Singleton |13~.

4 Our two criteria for selecting companies for the Chapter 11 and matching samples for the event study were: (i) the companies had to be listed either on the New York or the American Stock Exchange; and (ii) there must be two years of continuous data available for the companies.

5 In cases where there are further downgradings before Chapter 11 filing, the market's interpretation may be ahead of the rating agency's only up to the second downgrading, because the rating agency may reveal additional negative information about the firm in its second and further downgradings before the firm files for Chapter 11. Thus, the market may be ahead of the rating agency only during the period between the first consistent downgrading and the next (second) downgrading.


1. E.I. Altman and D.L. Kao, "The Implication of Corporate Bond Rating Drift," Financial Analysts Journal (May-June 1992), pp. 64-75.

2. E.I. Altman, R.G. Haldeman, and P. Narayanan, "ZETA Analysis: A New Model to Identify Bankruptcy Risk of Corporations," Journal of Banking and Finance (June 1977), pp. 29-54.

3. H.J. Einhorn, D.N. Kleinmuntz, and B. Kleinmuntz, "Linear Regression and Process-Tracing Model of Judgment," Psychological Review (September 1979), pp. 456-485.

4. E.F. Fama, Foundations of Finance, New York, Basic Books, Inc., 1976.

5. E.F. Fama, L. Fisher, M.C. Jensen, and R. Roll, "The Adjustment of Stock Prices to New Information," International Economic Review (February 1969), pp. 1-21.

6. S.C. Gilson, K. John, and H.P. Lang, "Troubled Debt Restructurings -- An Empirical Study of Private Reorganization of Firms in Default," Journal of Financial Economics (October 1990), pp. 315-353.

7. P. Grier and S. Katz, "The Differential Effects of Bond Rating Changes Among Industrial and Public Utility Bonds by Maturity," Journal of Business (April 1976), pp. 226-239.

8. P. Griffin and A. Sanvincente, "Common Stock Returns and Rating Changes: A Methodological Comparison," Journal of Finance (March 1982), pp. 103-119.

9. R.M. Hand, R.W. Holthausen, and R.W. Leftwich, "The Effect of Bond Rating Agency Announcements on Bond and Stock Prices," Journal of Finance (June 1992), pp. 733-752.

10. S. Katz, "The Price Adjustment Process of Bonds to Rating Reclassifications: A Test of Bond Market Efficiency," Journal of Finance (May 1974), pp. 551-559.

11. R.W. Masulis, "The Effects of Capital Structure Change on Security Prices: A Study of Exchange Offers," Journal of Financial Economics (June 1980), pp. 139-177.

12. D.C. Montgomery, Design and Analysis of Experiments, New York, John Wiley & Sons Inc., 2nd edition, 1984.

13. G.E. Pinches and J.C. Singleton, "The Adjustment of Stock Prices to Bond Rating Changes," Journal of Finance (March 1978), pp. 29-44.

14. M.I. Weinstein, "The Effect of a Rating Change Announcement on Bond Price," Journal of Financial Economics (December 1977), pp. 329-350.

Keqian Bi is an Associate Professor of Finance at the McLaren School of Business, University of San Francisco, San Francisco, California. Haim Levy is a Professor of Finance at The Hebrew University, Jerusalem, Israel, and at the University of Florida, Gainesville, Florida.
COPYRIGHT 1993 Financial Management Association
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 1993 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Title Annotation:Financial Distress Special Issue; filing under bankruptcy law
Author:Bi, Keqian; Levy, Haim
Publication:Financial Management
Date:Sep 22, 1993
Previous Article:Recognizing financial distress patterns using a neural network tool.
Next Article:Troubled savings and loan institutions: turnaround strategies under insolvency.

Related Articles
Bankruptcy planning.
Absolute priority rule violations and risk incentives for financially distressed firms.
The emerging market for trade claims of bankrupt firms.
An empirical examination of prepackaged bankruptcy.
Bankruptcy as a risk management tool: economic and social implications.

Terms of use | Copyright © 2017 Farlex, Inc. | Feedback | For webmasters