Assessing the Risk Relevance of Accounting Variables in Diverse Economic conditions

This paper examines the association between accounting information and systematic (beta) risk. We extend previous research by using an updated data set, a range of risk measures that adjust for different market and time-varying conditions, and by examining whether the long-run association has changed over time. Results indicate a consistently strong association between accounting variables and systematic risk, with up to 67% predominantly explained by accounting variables that proxy for operational and growth risk. The strength of the association, however, varies according to industry and firm size. Results are robust to different market conditions, including mean reverting risk, time varying risk and financial leverage, but with a lower association in thin trading conditions. In an out-of-sample test we found that an accounting variable risk model is superior in predicting forward time-varying systematic risk proxied by an M-GARCH beta. Finally, we determined that the risk relevance of accounting information has not significantly declined over the past 30-years, contrary to other research that shows a decline in the price relevance of accounting information over the same period.