Implied volatility curve: an empirical study on Chinese SSE 50 ETF options

This paper studies the determinants shaping the implied volatility curve of Chinese SSE 50 ETF options and the predictability of its patterns. Initially, the research examines the economic variables influencing the curve's curvature. Our analysis reveals a pronounced disparity in how investor sentiment and risk appetite impact the curvature of call versus put options' volatility curves. Subsequently, the study distinguishes four patterns of the volatility curve, exploring the causes of their formation. We observe that shifts in market conditions precipitate alterations in the curve's patterns, with pessimistic investor sentiment and diminished risk appetite correlating with a more pronounced skewness in the curve. In the final phase, the research employs machine learning techniques to prognosticate the patterns of the volatility curve. The chosen model demonstrates robust proficiency in forecasting Smile and Smirk patterns, yet it notably struggles to predict the Frown pattern, which typically emerges alongside major market emergencies.