Predicting stock market returns with average correlation and average variance: Decomposition approach

Does the observed stock market return variance predict future stock market return? I demonstrate that decomposing individual stock returns into systematic and idiosyncratic parts and using these components separately in constructing the average correlation and the average variance of individual stock returns is crucial for predicting future stock market return. I find that only the average correlation and the average variance for the systematic part of the individual stock returns predict future stock market returns. The results contribute to the literature on the riskreturn tradeoff for the stock market return and offer a new approach in reflecting aggregate wealth risk.