Information and stock returns: Evidence from high-frequency data in China's ETF options market

China's options market has experienced rapid development. This study moves beyond the predominant focus on implied volatility to examine implied higher moments—specifically, skewness and kurtosis. These moments are crucial for characterizing the full distribution of expected returns: implied skewness captures the market's expectation of asymmetric tail risk, while implied kurtosis gauges the anticipated probability of extreme returns, reflecting market fears of disruptive volatility events. Unlike existing literature, we examine CSI 300 and CSI 500 ETF options alongside the SSE 50ETF. Utilizing high-frequency data from September 2022 to December 2024, we first confirm the informational superiority of model-free implied volatility (MFIV) in predicting realized volatility. We then systematically investigate the asymmetric relationship between returns and these implied higher moments using a nonparametric quantile regression model. Our findings indicate a pronounced asymmetry: in high-volatility regimes, negative returns disproportionately increase implied volatility. Conversely, when skewness and kurtosis are elevated, positive returns exert a stronger influence, suggesting that in extreme market states, positive shocks amplify concerns about potential bubbles or destabilizing rallies. This asymmetry persists under external shocks. Our research provides a basis for formulating condition-specific hedging strategies and offers new empirical evidence on pricing behaviors in China's multi-tiered options market.