Trading on trends: How the ordering of historical volume predicts Chinese stock returns?

Abstract

In examining return prediction strategies in China’s stock market, we find that the chronological return ordering is ineffective within a one-month window. To overcome this limitation, we introduce a more robust measure, named chronological turnover ordering (${\text{CTO}}_3$), calculated using turnover in the past three months. As anticipated, ${\text{CTO}}_3$ demonstrates statistically significant predictability for returns, indicating a tendency among investors to overvalue stocks with high recent and low distant turnover. Bivariate portfolio analysis reveals that ${\text{CTO}}_3$ performs more effectively during high-sentiment periods and on stocks with high investor attention. This research contributes significantly to understanding investor behavior and market dynamics in China.