Are we living in an illusion? A fresh look at the importance of bank capital in the quest for stability

Using a sample of 2350 listed banks from 51 countries in 1990–2018, we find that changes in capital or even the amount of capital itself is not significantly related to concurrent changes in the banks’ probability of default. We combine four methods to deal with endogeneity issues (effect of unobserved confounders and reverse causality): linear models with fixed effects, novel instrumental variables in 2SLS analyses, generalised method of moments, and structural equation modelling. Hence, our conclusions can be seen as evidence of the lack of causal impact from capital to bank stability. This could be explained by the possibility of higher risk taken to cover the cost of capital offsetting the loss absorption benefits of capital and by different loss perceptions of shareholders that may reduce their incentives to monitor bank managers. Our results are corroborated by several robustness tests involving different capital and stability measures and alternative model specifications. In an additional step, we also show that promoting changes in other aspects normally considered by regulators and supervisors (asset quality, management quality, earnings, liquidity risk, and sensitivity to market risk) does not lead to higher stability either. In sum, we contribute to the literature by applying original approaches for causal inference in the context of bank capital and stability. For regulators, we leave a message indicating that the prevalent regulatory framework is likely based on an illusion given that making banks increase their capital does not necessarily make them more resilient.