Convert-to-Surrender Bonds: A Proposal of How to Reduce Risk-Taking Incentives in the Banking System

We argue that contingent convertible capital (CoCo-Bonds) might have perverse risk-taking incentives for banks (asset substitution problem) and discourage them from investing in positive NPV projects and issuing new equity in times of crisis (debt overhang problem). Whenever the conversion price is set too high - as in the case of the Lloyds CoCo-Bond issuance in November 2009 - a wealth transfer will take place at the time of conversion. This will exacerbate both the asset substitution problem and the debt overhang problem. We propose a new type of contingent convertible capital for banks - which we label Convert-to-Surrender Bonds (CoSu-Bonds) - which eliminates both the asset substitution and the debt overhang problem. CoSu-Bond convert into equity once the equity ratio falls below a certain threshold and CoSu-Bond holders take over the bank while equity holders are totally wiped out. This instrument makes equity holders naturally risk averse and gives incentives to equity holders to raise new equity and to invest also in slightly negative NPV projects in times of financial distress. Our instrument has a unique price if it is augmented by a simple option for old equity holders to issue equity if the trigger is hit, thereby circumventing the problem of multiple equilibria. We develop a tractable model to analyze risk-taking incentives in the banking sector and present a detailled analysis of the Lloyds CoCo-Bond issuance. We finally discuss several policy issues associated with this new instrument including robustness and unique equilibria.