Price Efficiency, Bubbles, Crashes and Crash Risk: Evidence from Chinese Stock Market

Abstract

When there is bad news hoarding from managers, returns of stocks are no longer efficient. We hypothesize that a proxy for efficient returns predicts stock price bubbles, crashes and crash risk. We find evidence in support of our hypotheses. Lagged price efficiency significantly predicts bubbles, crashes and crash risk in multivariate linear regressions and logit regressions, as predicted by our hypotheses. We also find that the lagged probability of bubbles is only correlated with future returns. In contrast, the lagged probability of crashes is correlated with both future returns and fundamental values of stocks. This result validates our explanation for the formation of bubbles and crashes. Finally, the out-of-sample accuracy ratio of our bubble and the crash prediction model is higher than in previous studies. Our results provide alternative explanations of the mechanics of stock price bubbles and crashes and are helpful for academicians, investors and policymakers.