Abstract
This paper sits at the intersection of derivative pricing and systemic-risk measurement by studying how system-wide tail dependence and crisis amplification can be incorporated into the valuation of European options. The research is motivated by the robust measurement of conditional downside vulnerability, and the adjustment of option values based on each asset's individual contribution to systemic distress. We subsequently compare a conventional econometric approach against a more flexible machine-learning alternative for estimating this tail-risk component. Empirically, we find substantial heterogeneity in downside tail exposure across broad asset classes, implying that conditioning on systemic distress provides information beyond marginal (stand-alone) risk. We also show that the learning-based tail estimator produces systematically more conservative tail-risk assessments than the econometric baseline, suggesting that adaptive methods may better capture extreme, time-varying distress dynamics. Incorporating systemic tail risk into option pricing yields economically consistent effects-lower call values and higher put values-reflecting compensation for downside externalities that intensify under crisis regimes. Finally, we show that endogenizing the strength of the systemic adjustment generates plausible cross-asset variation, with pricing impacts primarily explained by differences in the individual assets' contribution to systemic-risk spillovers.