Black-Scholes versus GARCH Models: An Empirical Study of CSI 500 Index Option Pricing

This study empirically examined the performance of some popular option pricing models in the valuation of the CSI 500 ETF index options. The CSI 500 index tracks the mid-to-small cap equities in the Chinese A-share market. In particular, this study analyzed the market data from September 2023 to August 2024 to compare the performance of the Black-Scholes model and GARCH-variant models. Although specific characteristics of the CSI 500 index, including higher volatility and lower liquidity, theoretically may favor more sophisticated time-series models, this study finds the Black-Scholes model produces more accurate option valuation across different maturity ranges. This superior performance is most pronounced for options with shorter time to maturity and remains robust in out-of-sample data testing. All models produce better prediction accuracy for shorter-dated options. These results show that, presently market microstructure and market participant factors in the Chinese mid-cap equity options market may be dominating theoretical considerations in selecting the optimal option pricing methodology. This study serves as the first to comprehensively evaluate the performance of option pricing models in the Chinese mid-cap index options market, hence providing practical insights to market participants and regulators in various emerging derivatives markets.