Forecasting Equity Index Volatility Spreads

Abstract

Equity market indices such as the S&P500, Dow Jones Industrial Average, NASDAQ Composite, and Russell 2000 have become some of the world’s most heavily traded financial products through their related ETFs, futures, and derivatives. While much work has been done on modelling relative returns and forecasting individual index volatilities, there has been little research into the relationship between index volatilities. We apply the established methodologies of volatility forecasting such as univariate and multivariate GARCH and ARFIMA modelling in order to generate forecasts for equity index volatility spreads. While this approach emphasizes the dynamics of individual volatility time series such as persistence in volatility shocks and asymmetrical reaction to return shocks, we find that these models produce forecasts that are less or about equal in accuracy to simpler moving average models.