Volatility Components: Evidence from VIX Futures Market

Abstract:     
In this paper we analyze CBOE VIX futures price time series data from Mar. 2004 to Feb. 2008. We derive a new pricing framework for VIX futures that is convenient to study variance term structure dynamics. Our main contribution to existing literature is the identification of the number of factors implicit in VIX futures term structure. We find that three-factor model is ideal to characterize the variance term structure. We further construct and estimate structured two- and three-factor models to identify the components and find similar results.

Keywords: VIX futures, loglinear model, Kalman filter, Principal Component Analysis (PCA), variance term structure

JEL Classifications: C52, G12, G13

Working Paper Series