This paper offers a novel approach to identify the relationship between extensive option panels and market returns using functional predictive regression. Employing our approach on the options and realized returns of the S&P 500, we achieve a remarkable performance in predicting S&P 500 monthly returns, yielding a 4.720% (6.198%) in-sample (out-of-sample) R2. The performance of our approach is superior to that of other well-known predictors and equilibrium models. The out-of-sample performance delivers substantial utility gains over historical averages. We find that both the use of option panels and the adoption of functional regression are indispensable for the outperformance.

JEL classification codes: G12, G17

Keywords: functional predictive regression, return predictability, risk-neutral measure, option market, market risk premium