Does the market believe in loss-absorbing bank debt?

We propose a simple model to estimate the risk-neutral loss distribution from the credit spreads of long-term debt instruments with different seniorities. We apply our model to a sample of global systemically important banks that have issued bail-in debt in order to meet the total loss-absorbing capacity (TLAC) requirements established after the global financial crisis. Bail-in debt is a new debt category that absorbs losses in a gone-concern situation and that ranks between subordinated debt and non-eligible senior debt. With a structural model for these three debt layers, we calibrate the tail of the risk-neutral loss distribution such that it is consistent with the observed market prices. Based on this loss distribution, we find that the expected loss in a gone-concern situation exceeds TLAC for most banks and that the risk-neutral probability that TLAC will not be sufficient to cover the losses in such a situation is approximately 50%. The large expected losses that we find with our model are a consequence of the similar pricing of bail-in debt relative to other senior debt. We argue that regulators should promote further clarity about the subordination and the conversion mechanism of bail-in debt to achieve a more differentiated pricing that is more in line with regulatory expectations.