Leverage Ratios, Industry Norms, and Stock Price Reaction: An Empirical Investigation of Stock-for-Debt Transactions - Statistical Data Included

Financial Management (Financial Management Association), Summer, 1999 by Robert M. Hull

The event study research of security offerings has largely failed to explore how stock prices react when a firm changes its debt-to-equity ratio (DE) in relationship to its industry DE norm.(1) This lack of serious regard by event studies for the role of an industry DE benchmark is puzzling given the insight of financial leverage ratio research. This line of research suggests that an industry DE benchmark should prove useful in predicting the direction and magnitude of stock returns that accompany pure leverage-change announcements.

In this paper, I am motivated by the notion that an industry DE norm (e.g., median or mean) is a useful benchmark when investors evaluate a stock's true worth. The research hypothesis is that firms moving "closer to" industry DE benchmarks should have a market response that is positive when compared to firms moving "away from" industry DE benchmarks. To test this hypothesis, I obtain a working sample of 338 observations where firms announce public common stock offerings. This sample is distinctive in two respects that are important for achieving the research aim.

First, it is distinctive in that the sole purpose of each offering is for debt reduction. Not only are the productive assets not directly altered, but simultaneously changing both equity and debt produces a large alteration in a firm's DE. Given the notion that firms operate within target DE ranges, large movements in DEs may be required to detect if the market reacts consistently with the view that an industry DE benchmark is a wealth-maximizing target.

Second, the sample is distinctive in its size (n = 338). Despite the support by leverage ratio research for the importance of an industry DE benchmark, one can argue that an industry benchmark DE may not always be a good estimate of what is perceived as its wealth-maximizing DE. Consequently, a large sample offers the possibility of overcoming estimation problems if one can assume that errors in estimating become less of a concern as the sample size increases.

In support of the research hypothesis, I find that the three-day mean cumulative abnormal return (CAR) of -1.91% for "closer to" firms is less negative than the -3.41% CAR found for "away from" firms.(2) The difference of 1.50% between the two CARs is statistically significant at the 0.01 level. Even if industry DE norms are not precise approximations of optimal DEs, the significant CAR difference indicates that industry DEs are perceived by the market as desirable DEs. In further support of the research hypothesis, I conduct other tests (in particular, regression analysis) that show the robustness of the findings concerning the importance of an industry DE norm.

I organize the remainder of the paper as follows. Section I reviews the literature, while Section II describes the sample, "change in DE" measurement, summary statistics, methodology, and statistical tests. I report empirical results in Section III and summarize the findings in Section IV.

I. Literature Review

In this section, I review capital structure theory and prior empirical research.(3) I find that a firm's industry DE norm is usable in empirical tests as a benchmark to generate stock price predictions.

A. Capital Structure Models

The capital structure model of Modigliani and Miller (1963) posits that a firm's value increases as its DE increases due to a corporate tax shield effect. Extensions (e.g., Kraus and Litzenberger, 1972; Jensen and Meckling, 1976; DeAngelo and Masulis, 1980; Kim, 1982; Ross, 1985; and Leland, 1994) argue that an increasing DE leads to ever rising leverage-related costs such that firm value will eventually stop increasing. Dynamic optimal models, such as Fischer, Heinkel, and Zechner (1989) and Mauer and Triantis (1994), do not advocate a static optimal capital structure. Nonetheless, the optimal dynamic financing policy is still characterized by a tradeoff between the corporate tax shield advantage of debt and the leverage-related costs of debt.

Asymmetric information signaling models posit different levels of information between insiders and outsiders such that insider behavior conveys information about firm value to outsiders. These models predict that a change in a firm's mix of debt and equity contains news about stock value. For example, Leland and Pyle (1977) argue that security offering announcements will lead to stock returns that are positively related to the expected change in insider ownership proportions. Ross (1977) claims that leverage increases convey positive news concerning the firm's capacity to service a larger amount of debt. Similarly, it can be argued that leverage decreases signal negative news. Fama (1985) asserts that firms announcing bank debt agreements signal positive news. This is because bankers are privy to inside information and would not approve the loan if negative news was gotten in the lending process. Similarly, firms that announce bank debt reductions convey unfavorable inside information via banker actions. Adverse selection signaling theorists (Myers and Majluf, 1984; and Lucas and McDonald, 1990) posit that the market suspects stock overvaluation when managers announce a stock offering. The negative signaling perceived by outsiders can be lessened if there is a reduced information advantage to insiders.