The forecasting abilities of implied and econometric variance–covariance models across financial measures

Abstract

This study evaluates the forecasting ability of implied and econometric forecasting models under statistical and financial evaluation measures. The econometric models outperform the implied model on all criteria. Overall, the exponentially weighted moving average model is the best forecasting model since it produces hedged portfolios with the lowest variance. It also has the ability to time market fluctuations while maintaining minimal capital allocation, and thus enhances profitability when employed as a tool for trading strategies. Our empirical results suggest that firms should explore the use of dynamic statistical forecasting models rather than relying on the implied model.