Industry characteristics and the information content of debt downgrades.

Rajagopal, Sanjay ; Kohers, Theodor

INTRODUCTION

This paper begins by quantifying the effect of debt downgrades by credit rating agencies both on the re-rated firm's equity and that of its industry rivals. An earlier paper has considered a similar question, though it did not distinguish between industrial and non-industrial firms (Akhigbe et al., 1997). The present study adds to the existing body of knowledge by explaining the cross-sectional variation in abnormal returns across the industries within which the re-rating event occurs. The focus of the study remains on non-banking industries. In a broader context, by considering the industry-level characteristics that condition the impact of debt downgrades on incumbent firms, the study sheds light on the process of price formation in capital markets.

PREVIOUS STUDIES

Numerous studies over the last two decades have examined the security price effects of debt rating changes (see, for example, Katz, 1974; Weinstein, 1977; Pinches & Singleton, 1978; Holthausen & Leftwich, 1986; Zaima & McCarthy, 1988; Hand et al., 1992; Hsueh & Liu, 1992; and Schweitzer et al., 1992). Of these studies, the more recent ones lend strong support to the hypothesis that bond rating downgrades bring new information relevant to the pricing of the re-rated firm's equity.

The present work contends that such re-ratings convey information not only about the re-rated firm, but also about other firms in the same industry. In the process, it attempts to shed new light on the manner in which rating agencies contribute to the imposition of capital market discipline. An earlier study conducted an investigation of the industry-wide information effects of bond rating changes in the banking industry (Schweitzer et al., 1994). In contrast, the current study focuses on several non-banking industries. An examination of industrials, rather than just banks, provides a richer context in which to study the impact of rating changes, in part because it allows an analysis of how and to what extent industry-related factors condition the revaluation of equity caused by bond rating changes in an environment free of the regulatory conditions unique to banks.

A more recent study (Akhigbe, Madura & Whyte, 1997) has analyzed the industry-wide effects of bond rating changes for a sample of firms that includes non-financial industries. The present work, in addition to documenting the robustness of that study's results, contributes to the existing body of knowledge by analyzing how several industry-specific factors condition the cross-sectional variation in average abnormal returns across industries in which the re-rating event occurs.

METHODOLOGY AND DATA

This study employs standard event study methodology described in earlier studies of intra-industry effects of firm-specific events (see, for example, Mikkelson & Partch, 1985; Linn and McConnell, 1983; Slovin et al., 1995; and Zantout & Tsetsekos, 1994). The event day (Day 0) is defined as the day the announcement of a bond downgrade appeared in the Wall Street Journal. The market model parameters are estimated over the period -111 to -11 days relative to the event day and the CRSP Equally-Weighted Index is used as a proxy for the market return over day t. The present analysis uses discrete returns.

The study considers debt downgrade announcements by Moody's Investor Services in the six-year period 1990 to 1995, and considers bond downgrades announced by the two major rating agencies, Moody's Investor Services and Standard and Poor's Corporation. In order to be included in the study, these announcements had to have appeared in the Wall Street Journal (WSJ). The event day (Day 0) is defined as the date on which the story of the downgrade appeared in this newspaper. Multiple downgrades within a six-month period for the same company are eliminated. Also, only those downgrades are considered that are not accompanied by other relevant news about the firm within a three-day window surrounding the event.

The firms selected as rivals belong to the same Compustat-assigned 4-digit Standard Industrial Classification (SIC) code as the downgraded firm in the year of the downgrade. Also, in order for a firm to be included in the study, its daily stock returns had to be available in the Daily Returns File of the Center for Research in Security Prices (CRSP) for the period -111 to +10 days relative to the event date.

The WSJ Index is used to identify the date on which the announcement of the downgrade appeared in the Wall Street Journal. This Index also serves as the source of information on any confounding events during the three days surrounding the event date. Equity returns, along with the daily returns on the market index proxy, are retrieved from the CRSP Daily Returns and Calendar/ Index Files. Industry-specific characteristics are based on data provided by Compustat. The dates identifying periods of economic expansion and contraction are provided by the Survey of Current Business published by the U.S. Department of Commerce.

In all, the study includes 136 downgrades for 117 industries. The total number of rival firms is 2,195. The number of rivals range from 14 to 27, with an average of 19 rivals per industry.

HYPOTHESES

This study segregates rival firms by industry and analyzes the effect of specific industry-level characteristics on the nature of industry-wide information conveyed by bond downgrades. Previous studies suggest, for instance, that the financial leverage and the degree of competition within an industry may be important factors in explaining the stock price reactions of rival firms (Lang and Stulz [15]). To explore this possibility, we model the average abnormal returns to all rival firms within an industry as follows:

[CAR.sub.oi] = f ([AR.sub.o]; INDLEV; HI; DMAT; CYCLE),

Where:

[CAR.sub.oi] = Average abnormal returns over Days 0 and 1 for all rival firms;

[AR.sub.o] = The abnormal returns to the downgraded firms on the event day;

INDLEV = Average total debt to total asset ratio, which measures financial leverage;

HI =H erfindahl Index, which measures industry concentration;

DMAT = Average of short-term to long-term debt ratio, which measures debt maturity; and

CYCLE = 0 if the downgrade occurred during an economic contraction, and 1 otherwise.

The coefficients of [AR.sub.o], INDLEV, and CYCLE in a linear regression are expected to be positive, negative, positive and negative, respectively. That is, we expect the (negative) effect on average rival equity to be greater if the re-rated firm is affected more strongly (negatively) by the rating change, to be greater in a more levered industry, and to be greater during an expansion, when a downgrade would be a greater surprise. There are no a priori expectations with regard to the specific direction of influence of HI, and DMAT.

RESULTS

Table 1 shows the average abnormal returns to downgraded firms over a period of 21 days surrounding the event date. The results confirm the finding of previous studies that debt downgrades have a negative influence on the equity value of the re-rated firm. Thus, this action of rating agencies brings new information to the capital markets.

The effect of debt downgrades on industry rivals can be seen in Table 2. These results provide substantial evidence of a predominant "contagion effect" stemming from bond downgrades. As in the case of downgraded firms, the group of rival firms experiences a highly significant and negative average abnormal return on the announcement day. No statistically significant average abnormal return is observed for Day -1, but a negative abnormal return significant at the 10% level is seen on Day +1. On the announcement day, these rivals as a group experienced an average abnormal return of -0.325% (Z-value:-3.831) which is significant at the 1% level. It should be noted that the typical event study focuses on a narrow window over which the influence of a maximum of confounding events can be eliminated. Even though there are a few significant average abnormal returns prior to Day -1 and after Day +1, these may stem from other events that were not relevant to the screening process because they lay outside the event window. Within this window, however, the abnormal returns to rivals were negative and large.

To put these results in perspective, the -0.325% average abnormal return to rival firms on the day of the announcement is, in absolute value, almost twice the size (1.9 times) of any other abnormal return outside the event window in the 21-day period reported. The next largest abnormal return is 60% of that observed on the event day, and occurs on Day +1. This too is understandable, since news pertaining directly to a firm (the one downgraded) may be expected to be fully reflected in its price more quickly than in those of its rivals. Thus, no significant stock price adjustment recurs on Day +1 for the group of downgraded firms, while some is seen for the rivals.

The results presented in Table 2 suggest the presence of a contagion effect from bond downgrades, wherein the news of a downward rating revision at one firm is received by investors as adverse news for the average firm in the same industry. It also appears that the reaction to the announcement is immediate, and that most of the new information contained in the re-rating is impounded in rival stock prices on the same day. In addition, some downward revision in stock price, albeit of a smaller magnitude, seems to occur on the day following the announcement.

Next, we turn to the discussion of industry-specific characteristics that might condition any intra-industry effects arising from bond downgrades. In addition, the role of changes in economic conditions is considered. Table 3 shows the estimated relationship between the average abnormal returns to industry rivals as a group and the abnormal return to the downgraded firm. The model also includes measures of leverage, concentration, and debt maturity at the industry level. Finally, a dummy variable is included to distinguish a short period of contraction from the rest of the sample period.

Of the 136 instances of downgrades included in this study, data was available on all five explanatory variables for 128 cases. The multicollinearity diagnostics did not indicate the presence of collinearity between any regressors. The overall regression was significant at the 1.17% level, and explained approximately 8% of the cross-sectional variation in the average cumulative abnormal returns to rivals over Days 0 and 1.

The results indicate that the abnormal returns to downgraded firms on the event day play a significant role in explaining the differences in abnormal returns in the cross-section of industries. Although not reported here, the estimated coefficient of ARo is significant at the 2.5% level. The positive sign to the coefficient points to an intra-industry contagion effect arising from debt downgrades. In terms of absolute magnitude, every 1% increase in abnormal return to the downgraded firm elicits approximately a 0.203% increase in the cumulative abnormal return to the corresponding portfolio of industry rivals.

It was expected that any contagion effect would be stronger in industries with higher leverage because of the higher elasticity of equity value to changes in expected cash flows. Thus a significantly negative sign was predicted for the estimated coefficient of INDLEV. Table 3 shows, however, that in the present sample of downgrades, industrial leverage was not significant in explaining the cross-sectional differences in the cumulative average abnormal returns to rivals. It should be noted that leverage was measured as the ratio of total debt (long-term and short-term) to total assets. The results pertaining to another explanatory variable, DMAT, which are discussed below, seem to indicate that the inclusion of short-term debt in the leverage measure might be causing industrial leverage not to be a significant determinant of the size of contagion.

The contagion effect should be stronger in more concentrated industries if rivals in such industries, with their large market shares, could more readily be recognized by investors as competing in the same resource and product markets. This would cause a stronger extrapolation of any negative news inferred from a bond downgrade to these prominent rivals. As Lang & Stulz, 1992, suggest, however, the relationship between contagion and industry concentration is an empirical question. Table 3 shows that the estimated coefficient of the measure of industry concentration, HI, is not different from zero at any reasonable level of significance. Thus, any contagion effect resulting from the present sample of bond downgrades appears to be unrelated to the degree of competition within the industry.

To the extent that creditors become reluctant to roll-over debt upon the arrival of adverse information, the contagion effect would tend to be stronger in industries with a higher proportion of debt in shorter maturities. However, for a given amount of debt, the pure leverage effect would be greater for a higher proportion of long term debt. Thus, the net impact of debt maturity on the magnitude of the contagion effect depends on which of these two effects dominates. The results in Table 3 indicate that the estimated coefficient of DMAT is positive and significant at the 10% level under a two-tailed test. Since DMAT expresses the ratio of short-term to long-term debt, this suggests that, for a given leverage ratio, industries with higher proportions of long-term to short-term debt experience more negative abnormal returns. Thus, it appears that the pure leverage effect arising from long-term debt dominates any contagion arising from excessive liquidations by short-term creditors. Further, these results suggest that the poor explanatory power of INDLEV might derive from the fact that it is calculated using both long-term and short-term debt.

Finally, a categorical variable is included in the analysis to distinguish between periods of expansion and contraction from 1990 to 1995. One contraction can be identified for these six years, which lasted from August 1990 to March 1991. Of the 136 downgrades included in this study, 21 occurred during this period. It was expected that rating downgrades during periods of expansion would possess greater information content for the investors and result in a greater contagion effect. The results in Table 3 reveal, however, that the estimated coefficient for CYCLE is not significantly different from zero. Thus, downgrades during expansions do not appear to have a differential impact on the magnitude of contagion. The 21 downgrades during the contractionary period represent slightly over 16% of the sample of 128 announcements included in the regression. Such a low variation in CYCLE might have contributed to its lack of significance as an explanatory variable.

CONCLUSIONS

The present paper shows that debt downgrades by rating agencies have an impact on equity valuation beyond the re-rated firm. In particular, the announcement of a downgrade has a negative effect on the equity value of rival firms. The present study extends previous research by examining specific industry-level characteristics for their role in conditioning the average industry response to the debt downgrade announcement. The study finds that, the strength of the "signal", namely, the size of the abnormal return to the re-rated firm on the event date, has a significant effect on the average industry response to the rating change. A second factor which possesses some explanatory power is the average proportion of long-term debt in the industry. The results suggest that the pure leverage effect arising from long-term debt dominates any contagion arising from excessive liquidations by short-term creditors.

REFERENCES

Akhigbe, A., J. Madura & A.M. Whyte (1997). Intra-Industry Effects of Bond Rating Adustments. Journal of Financial Research, 20(4), 545-561.

Hand, J. R., R. W. Holthausen & R. W. Leftwich (1992). The Effect of Bond Rating Agency Announcements on Bond and Stock Prices. Journal of Finance, 47 (2), 773-752.

Holthausen, R. W. & R. W. Leftwich (1986). The Effect of Bond Rating Changes on Common Stock Prices. Journal of Financial Economics, 17, 57-89.

Hsueh, L. P. & Y. A. Liu (1992). Market Anticipation and the Effect of Bond Rating Changes on Common Stock Prices. Journal of Business Research, 24, 225-239.

Katz, S (1974). The Price Adjustment Process of Bonds to Rating Reclassifications: A Test of Bond Market Efficiency. Journal of Finance, 29, 551-559.

Lang, L. H. & R. M. Stulz (1992). Contagion and Competitive Intra-Industry Effects of Bankruptcy Announcements: An Empirical Analysis. Journal of Financial Economics, 32, 45-60.

Linn, S. C. & J. J. McConnell (1983). An Empirical Investigation of the Impact of Anti-Takeover Amendments of Common Stock Prices. Journal of Financial Economics, 11, 361-399.

Mikkelson, W. H. & M. M. Partch (1985). Stock Price Effects and Costs of Secondary Distributions. Journal of Financial Economics, 14, 165-194.

Pinches, G. E. & J. C. Singleton (1978). The Adjustment of Stock Prices to Bond Rating Changes. Journal of Finance, 33, 29-44.

Schweitzer, R., S. H. Szewczyk & R. Varma (1992). Bond Rating Agencies and Their Role in Bank Market Discipline. Journal of Financial Services Research, 6, 249-263.

Schweitzer, R., S. H. Szewczyk, & R. Varma (1994). The Intra-Industry Effects of Bank Bond Rating Changes. Proceedings of the Thirtieth Annual Conference on Bank Structure and Competition, Federal Reserve Bank of Chicago.

Slovin, M. B., M. E. Sushka & S. R. Ferraro (1995). A Comparison of the Information Conveyed by Equity Carve-Outs, Spin-Offs, and Asset Sell-Offs. Journal of Financial Economics, 37, 89-104.

Weinstein, M. I. (1977). The Effect of a Rating Change Announcement on Bond Price. Journal of Financial Economics, 5, 329-350.

Zaima, J. K. & J. McCarthy (1988). The Impact of Bond Rating Changes on Common Stocks and Bonds: Tests of the Wealth Redistribution Hypothesis. Financial Review, 23(4), 483-498.

Zantout, Z. Z. & G. P. Tsetsekos (1994). The Wealth Effects of Announcements of R&D Expenditure Increases. Journal of Financial Research, 17, 205-216.

Sanjay Rajagopal, Montreat College

Theodor Kohers, Mississippi State University

Table 1: Average Abnormal Returns to Downgraded Firms Between
Day -10 and Day +10 Relative to the Event Date

 Day Relative Average Z-value # Observations
to Event Date Abnormal
 Return

 -10 -0.137 -0.692 136
 -9 0.042 0.895 136
 -8 0.042 -0.250 136
 -7 0.036 0.082 136

 -6 -0.059 0.900 136
 -5 -0.343 -0.593 136
 -4 -0.048 -1.444 136
 -3 -0.049 -0.188 136
 -2 0.584 2.120 * 136
 -1 -0.121 -0.972 136
 0 -1.072 -3.651 *** 136
 +1 0.262 0.924 136
 +2 -0.393 -0.658 136
 +3 0.433 1.477 136
 +4 -0.260 0.381 136
 +5 0.649 2.303 ** 136
 +6 0.010 0.473 136
 +7 0.065 -0.931 136
 +8 0.178 1.484 136
 +9 -0.337 -1.234 136
 +10 0.068 0.232 136

*** Significant at the 1% level under a two-tailed test.

** Significant at the 5% level under a two-tailed test.

* Significant at the 10% level under a two-tailed test.

Table 2: Average Abnormal Returns to Rival Firms Between
Day -10 and Day +10 Relative to the Event Date

Event Date Average Z value # Observations
 Abnormal
 Return

 -10 -0.015 0.447 2195
 -9 0.036 1.060 2195
 -8 0.103 0.787 2195
 -7 0.034 -0.401 2195
 -6 0.105 2.574 ** 2195
 -5 0.089 1.998 ** 2195
 -4 -0.114 -0.456 2195
 -3 0.061 1.138 2195
 -2 -0.128 -2.176 2195
 -1 0.151 0.543 2195
 0 -0.325 -3.831 *** 2195
 +1 -0.197 -1.774 * 2195
 +2 0.136 2.230 ** 2194
 +3 -0.145 -2.067 2193
 +4 -0.090 -1.671 2193
 +5 -0.050 2.438 ** 2175
 +6 -0.170 -2.137 * 2175
 +7 0.040 1.227 2175
 +8 -0.115 -2.858 *** 2175
 +9 0.081 -0.654 2175
 +10 0.087 0.412 2175

*** Significant at the 1% level under a two-tailed test.

** Significant at the 5% level under a two-tailed test.

* Significant at the 10% level under a two-tailed test.

Table 3: Regression Results Explaining Average Industry Abnormal
Returns Associated With Bond Downgrade Announcements Over the
Period 1990 - 1995

CA[R.sub.0,1] = [alpha] + [[beta].sub.1] (A[R.sub.0]) +
[[beta].sub.2] (INDLEV) + [[beta].sub.3] (HI) + [[beta].sub.4]
(DMAT) + [[beta].sub.5] (CYCLE) + [epsilon]

Regression Variables Coefficient t-value

Intercept -0.0003 -0.024
AR to Downgraded Firm 0.2029 2.316 **
Industry Leverage 0.0100 0.419
Concentration -0.0068 -1.008
Debt Maturity 0.0002 1.696 *
Cycle -0.0056 -0.854
F = 3.09 **
Adj. R-square 8%
N 128

*** Significant at the 1% level under a two-tailed test.

** Significant at the 5% level under a two-tailed test.

* Significant at the 10% level under a two-tailed test.

Note: The t-values reported above have been corrected for
heteroscedasticity using White's consistent estimates of
standard errors for the coefficients.