Determination of optimal sampling frequency for conditional variance models in the Indian context

Abstract

This paper evaluates the performance of conditional variance models using high frequency data of National Stock Index (S&P CNX NIFTY) and determines the optimal sampling frequency for the best daily volatility forecast. A linear combination of the realized volatilities calculated at two different frequencies is used as benchmark to evaluate the volatility forecasting ability of the conditional variance models (GARCH (1,1)) at different sampling frequencies. From the analysis, it is determined that the sampling at 30 minutes gives the best forecast for the daily volatility. But the forecasting ability of these models is deteriorated by the non-normal property of mean adjusted returns which is an assumption in conditional variance models. Nevertheless, the optimum frequency should remain same even in the case of appropriate distribution assumption should be 30 minutes, as the same distribution is assumed for all the models at different sampling frequencies. Key words: Realized Volatility, conditional variance, daily volatility forecasting