Does the distress factor hypothesis explain the size and value effects in equity returns?

Abstract

The distress factor hypothesis says that value stocks and small stocks are distressed and therefore higher returns on such stocks are merely a compensation for higher risk. I test this hypothesis using z-scores, a cleaner proxy for bankruptcy risk than other proxies used in the literature such as dividend reductions or omissions. I find that unconditionally, distressed stocks earn significantly lower returns than non-distressed stocks and much underperformance is uninfluenced by size and B/M factors. I also find that z-score, size and B/M effects are stronger in different months suggesting little common variation between the three factors. The results show that size and B/M effects are unrelated to bankruptcy risk on an unconditional basis. Of crucial importance is a consideration of the time varying behaviour of bankruptcy risk premia and I consider explicitly the impact of changes in GDP growth rate and the impact of stock market movements on the pricing of distressed firms. I find that risk of bankruptcy is a systematic risk with distressed stocks registering strong underperformance during ‘bad’ states of the world. As with unconditional analysis, the results show there is no link between distress factor and size and B/M effects. Size and B/M effects are stronger in non-distressed stocks. To ensure that the empirical results are robust across different methodologies, I significantly expand on the work of Dichev (1998) by employing two different portfolio formation methods and individual securities in my analysis. My main results on z-scores are robust though size and B/M effects are sensitive to alternative trading rules. I also test the Fama & French (1993) three-factor model for the UK and find that it is unable to explain returns on negative z-score portfolios. A four-factor model that includes a factor mimicking the z-score effect is better specified. The primary contribution of this study is the direct evidence it provides on the distress factor hypothesis of higher returns on value stocks and small stocks and the four-factor model for stock returns. This research has important implications both for extant asset pricing theories and for practitioners especially in evaluation of portfolio performance and computation of abnormal returns.