Market Implied Conformal Volatility Intervals

Volatility is a fundamental input for pricing and risk management of financial instruments. In the following work we propose an algorithm to estimate the market implied uncertainty of future realized volatility. Our method interprets the market implied volatility as a point prediction of future realized volatility and applies online conformal prediction to estimate the uncertainty of this prediction. We analyze rolling coverage and width of several nonconformity scores over 15 years of daily data. The results suggest that conformal prediction can be used to infer market implied prediction intervals for realized volatility.