In this paper weekly volatility forecasts are considered with applications to risk management; in particular hedge ratios and VaR calculations, with the aim of identifying the most appropriate model for risk management practice.
The study considers a variety of models, including those typically employed within the risk management industry, such as averaging and smoothing techniques, as well as those favored in academic circles, such as the GARCH genre of models, and a more recent realized volatility approach which incorporates both the simplicity in construction favored by the finance industry and the flexibility and theoretical underpinnings recommended by academics.
The results support the view that this realized volatility measure provides not only superior volatility forecasts per se, but also allows for improved hedge ratio and VaR calculations.
The research findings carry practical implications for the conduct of risk management, namely that volatility forecasts are best obtained using the realized volatility approach.
It is therefore proposed that a future direction for risk management practice may be to utilize such measures, while more generally it is hoped that such approaches may improve the cross‐fertilization of ideas and practice between the academic and practitioner communities.
