Does the weather affect the Chinese stock markets? Evidence from the analysis of DCCA cross-correlation coefficient

Recent studies confirm that weather affects the Chinese stock markets, based on a linear model. This paper revisits this topic using DCCA cross-correlation coefficient (ρDCCA(n)), which is a nonlinear method, to determine if weather variables (i.e., temperature, humidity, wind and sunshine duration) affect the returns/volatilities of the Shanghai and Shenzhen stock markets. We propose an asymmetric ρDCCA(n) by improving the traditional ρDCCA(n) to determine if different cross-correlated properties exist when one time series trending is either positive or negative. Further, we improve a statistical test for the asymmetric ρDCCA(n). We find that cross-correlation exists between weather variables and the stock markets on certain time scales and that the cross-correlation is asymmetric. We also analyze the cross-correlation at different intervals; that is, the relationship between weather variables and the stock markets at different intervals is not always the same as the relationship on the whole.