Do industry returns predict the stock market? A reprise using the random forest

The prior work reports conflicting evidence on the information content of industry returns for the market index return. We reexamine the out of sample predictive ability of industry returns by considering several relatively advanced methods from the statistical learning literature. We show that when the random forest method, which accounts for both linear and nonlinear dynamics, is used for regression, industry returns indeed contain significant out of sample forecasting power for the market index return. Moreover, our analysis also presents evidence for lead-lag relations among individual industry returns. The reported findings are consistent with the implications of the gradual diffusion of information hypothesis.