We show that increases in oil prices, rather than changes in oil prices, can predict stock returns. The revealed stock return predictability is both statistically and economically significant. The forecasting performance of oil price increases is not affected by changes in the choice of subsample, a considerable advantage over other popular predictors. We obtain greater forecasting gains by adding oil price increases as an additional predictor to univariate macro models. This forecasting improvement is also present when using multivariate information methods. The success of oil-macro models in forecasting stock returns is robust to a large battery of robustness tests. Oil price increases predict stock returns by affecting future industrial production and discount rates.

In this study, we forecast the realized volatility of the S&P 500 index using the heterogeneous autoregressive model for realized volatility (HAR-RV) and its various extensions. Our models take into account the time-varying property of the models’ parameters and the volatility of realized volatility. A dynamic model averaging (DMA) approach is used to combine the forecasts of the individual models. Our empirical results suggest that DMA can generate more accurate forecasts than individual model in both statistical and economic senses. Models that use time-varying parameters have greater forecasting accuracy than models that use the constant coefficients. The superiority of time-varying parameter models is also found in volatility density forecasting.