Information Discreteness and Stock Market Returns: Evidence from China

The frog-in-the-pan (FIP) hypothesis posits that investor attention is limited and less responsive to persistent small-scale information. We construct information discreteness (ID) as a proxy variable grounded in the measure of information continuity and investigate its relationship with aggregate stock returns. Empirical evidence indicates that ID can be a powerful negative indicator for predicting excess stock returns in the Chinese market, with its superior in- and out-of-sample predictive performance relative to several commonly used predictors. Moreover, integrating ID into asset allocation within a mean-variance framework yields considerable economic gains for investors. ID’s predictability mainly arises from the discount rate channel and its strong association with price jumps and investor sentiment. Additionally, ID influences stock market returns by affecting market trading activity. Importantly, the forecasting strength of ID becomes markedly stronger in phases characterized by economic recessions and low market sentiment.