Tail‑Risk‑Augmented Asset Pricing : VaR and Expected Shortfall in the Swedish Equity Market, 2000–2019

Sammanfattning

This study examines whether systematic tail risk–measured by Value–at–Risk (VaR) and Expected Shortfall (ES)–is priced in the cross–section of Swedish stock returns over the period 2000–2019. Using a sample of 54 firms listed on the OMX Stockholm exchange, individual tail risk betas are estimated using an overlapping 60–month rolling window requiring a minimum of 30 valid monthly observations. Fama–MacBeth cross–sectional regressions are employed to evaluate whether tail risk exposure earns a statistically significant premium, controlling for firm size, downside beta, idiosyncratic volatility, and co–skewness, with standard errors adjusted using the Newey–West HAC correction. The primary monthly regressions indicate that standalone tail risk betas do not earn a statistically significant premium over the full sample or within isolated sub–periods. However, robustness checks spanning higher frequencies uncover a highly nuanced, frequency–dependent asset pricing structure. In weekly specifications, VaR–beta commands a significant positive premium, while ES–beta carries a significant negative coefficient. This opposing–sign phenomenon suggests that short–horizon investors differentiate between crash frequency (VaR) and crash severity (ES), treating severe tail losses as structural insurance hedges. Furthermore, sub–period analysis reveals that market–wide crash risk, rather than standalone firm tail risk, drives regime–dependent pricing. Co–skewness emerges as the most robust factor, undergoing a complete structural reversal from a highly significant negative premium pre–GFC to a significant positive premium post–GFC. This points to the 2008 Global Financial Crisis as a systemic learning event that permanently altered investor risk aversion regarding market–wide asymmetries.