Bank contribution to financial sector systemic risk and expected returns: Evidence from large U.S. banks

Abstract

We estimate the contribution of large U.S. banks to the financial sector systemic risk by using value-at-risk (VaR), conditional value-at-risk (CoVaR), and two-stage least square (2SLS) methodology. Our sample is the monthly stock returns of 25 large U.S. banks from 1997 to 2021. We find that banks contributing more to the systemic risk have lower future returns on average. We also sort the portfolios’ future returns into five pentiles based on systemic risk contribution (SRC) and find that portfolios with high SRC earn lower future returns than those with low SRC. Our second contribution to the literature is the indication of the endogeneity problem in the SRC measures. We suggest an identification strategy for the estimation of SRC measures. Our results are contrary to some of the earlier studies, which concluded that the Dodd–Frank act of 2010 failed to eliminate the too-big-to-fail problem in banks. Such studies showed that anticipation of government subsidies has not been eliminated in the form of higher expected returns even for banks contributing more to systemic risk of the financial system. The results in our present study open a new research direction and are useful for investors and policymakers.