Modelling the volatility of TOCOM energy futures: A regime switching realised volatility approach

This paper combines the Heterogeneous Autoregressive Realised Volatility (HAR-RV) model and Markov Regime Switching (MRS) approach to estimate forecast volatility of energy futures contracts traded at Tokyo Commodity Exchange (TOCOM). The proposed MRS-HAR-RV allows dynamics realised change as market conditions change. dataset consists intraday prices for gasoline, kerosene crude oil futures. Estimation results suggest that can capture price better than alternative models. However, out-of-sample evaluation show only produce forecasts more liquid contracts. Moreover, seems less over-predict under-predict compared HAR-RV, HAR-RV-CJ, GARCH, MRS-GARCH