This paper analyses whether realized generalized autoregressive conditional heteroscedasticity (GARCH) models are useful for pricing Nikkei 225 options. This model enables us to estimate simultaneously the dynamics of stock returns using both realized volatility (RV) and daily return data. The analysis also examines whether realized GARCH models using realized kernels (RK) and realized ranges (RR) improve the option-pricing performance. Comparing the empirical results, for call options, EGARCH models perform better; however, for put options, realized GARCH models with RK without nontrading hour returns perform better than those with RV or RR.
一橋大学 Official Web Site ≫
