Measuring and Testing Systemic Risk from the Cross Section of Stock Returns

This study quantifies the presence of financial distress in the cross section of stock returns. Systemic risk is defined as the occurrence of simultaneous tail events for a large fraction of firms. A tail event is interpreted as evidence of downside risk in the tail distribution of financial returns. This definition is made operational through a bootstrap test of first order stochastic dominance that corrects for estimation of the conditional volatility model parameters. The outcome of these individual tests is aggregated to construct two indicators of systemic risk. One indicator measures the fraction of firms that suffer simultaneous tail events. The second indicator aggregates the value of the test statistic over the cross section of assets. Large values of these indicators signal the potential of systemic events. This intuition is formalized by introducing two cross-sectional asymptotically Gaussian tests to assess statistically for the presence of systemic risk. The application of these methods to the cross section of FTSE-100 stock returns uncovers empirical evidence of financial instability during the period 2003 to 2006 and after the outbreak of the COVID-19 pandemic but not during the period 2007-2009 suggesting that the turbulences in the latter period were captured by suitable parametric risk models