The VIX Index: Forecasting Power and Perfomance in a Risk Management Framework

This paper explores the information content and forecasting power of the VIX Index computed by CBOE (Chicago Board Options Exchange) on two different levels. The analysis is organised around two research questions. The first question is aimed at understanding whether the Volatility Index (VIX), due to its forward-looking nature, forecasts the future realised volatility better than other estimation techniques that are based on historical data. This part of the analysis is in line with a rich stream of literature on the topic, and our contribution intends to test whether the empirical results obtained by other researchers hold true in the years of hightened volatility following the Lehman Brothers collapse. The second research question aims to evaluate the perfomance of the VIX within a risk management framework, exploring an aspect that has been scarcely analysed in the literature and that has produced relatively contradictory results. In particular, we use the VIX alongside other volality measures to compute the Value-at-Risk (VaR) metric for a hypothetical portfolio replicating the Standard & Poor's 500 Index. The various measures of maximum potential loss are then backtested against actual returns and compared in order to understand which one is more effective. Results show that the VIX index possesses a strong information content, even if it is an upward biased forecast of realised performance. When used to compute VaR however, the measures based on VIX are less effective than others using different volatility estimations, especially during periods of higher turbulence.