Impacts of lagged returns on the risk-return relationship of Chinese aggregate stock market: Evidence from different data frequencies

Abstract

In this paper we investigate if lagged returns have impacts on the relationship between the conditional mean and variance of Chinese aggregate stock market. We consider different frequent data and the influence of lagged returns on both expected returns and risk-return tradeoff simultaneously. We investigate both Shanghai Stock Exchange Index and Shenzhen Stock Exchange Index and conduct both in-sample analysis and out-of-sample analysis. The empirical findings suggest that the risk-return tradeoff is time-varying and affected by lagged returns, but the results are sensitive to different data frequencies. At daily data frequency, we found statistically significant negative impacts of negative lagged returns on future risk-aversion level but significant positive impacts on future expected returns. The out-of-sample return forecasting test suggest that the GARCH-in-Mean model with lagged returns can improve the forecasting accuracy compared to the GARCH-in-Mean model without lagged returns at all data frequencies. Lastly, applying the GARCH-in-Mean model with lagged returns to 10-day data can bring about evident utility improvement for risk-averse investors even when the trading cost is non-zero and risk-free rate is zero.