Beyond the S&P 500: examining the role of external volatilities in market forecasting

Abstract

This study examines the impact of various exogenous volatilities on S&P 500 volatility using HAR-RV-type models. In this regard, a diverse range of exogenous volatilities-including assets, commodities, oil, gold, bonds, interest rates, the CBOE Volatility Index (VIX), and the Economic Policy Uncertainty (EPU) index-are incorporated into the Heterogenous Autoregressive (HAR) model. The exogenous volatilities are classified into three groups: European Market Information, U.S. Market Information, and U.S. Financial and Policy-Related Information. The study aims to determine which group of information contributes most effectively to forecasting S&P 500 stock market volatility. The results reveal several noteworthy points. First, the predictive accuracy of the baseline HAR-RV model is improved by incorporating exogenous explanatory variables. Second, the combination and the kitchen-sink approaches outperform models with single exogenous regressors, particularly for the U.S. Market Information group. Third, the U.S. Market Information group demonstrates superior predictive power compared to the other two groups, highlighting the interconnected nature of U.S. financial markets.