Instead of the traditional volatility forecasting framework, this paper introduces an innovative two-step process for identifying implied volatility (IV)’s mispricing at a large cross section. The two-step process includes identifying the contribution of historical volatility and remaining other firm features, leaving the residual as mispricing. Trading 10-1 long-short portfolio on 1-month at-the-money (ATM) straddle based on the residual yields a high information ratio (IR) of 335.4711%, surpassing other trading signals in academic. This method not only offers a superior predictive indicator for option returns but also proves robust against changes in liquidity and transaction costs, maintaining its performance advantage over other signals, as confirmed by the double-sorting analysis.