Forecasting security's volatility using low-frequency and high-frequency historical data and option-implied information

Low-frequency historical data, high-frequency historical data and option data are three major sources of forecasting the underlying security’s volatility. In this paper, we propose a unified GARCH-Ito-OI model to integrate... [ view full abstract ]

Low-frequency historical data, high-frequency historical data and option data are three major sources of forecasting the underlying security’s volatility. In this paper, we propose a unified GARCH-Ito-OI model to integrate three information sources. Instead of using options’ price data directly, we extract the option-implied information, such as implied volatility, from the option data and consider it as an exogenous variable. We provide the quasi-maximum likelihood estimators for the parameters and establish the asymptotic theory for the estimators. The empirical analysis shows that the proposed GARCH-Ito-OI model has better out-of-sample forecasting performances than the models, which rely on two information sources.