Evaluating the time-varying impact of economic data on the accuracy of stock market volatility forecasts

I assess the time-variation in predictive ability arising from the inclusion of macroeconomic and financial data in a GARCH-MIDAS model for stock market volatility. I consider whether the relative forecasting performance is affected by the state of the business cycle or the market environment. Results suggest predictive ability varies significantly over time, especially over long horizons. A central result is that models including macroeconomic data are useful for forecasting in low volatility periods. On the other hand, financial data performs overall surprisingly poorly. No single forecasting model or combination scheme is superior on all horizons and in all time periods, and while the term spread improves forecasting performance in particular over long horizons, forecast combinations perform
well over the medium term.