The aftermath of the 2008 financial crisis has triggered a lively debate on bank regulation. Tougher capital adequacy requirements and more restrictions on bankers’ pay have been two issues that have played a prominent role in this debate.
In a column on Vox Research, Hans Gersbach has suggested a new regulatory instrument called Crisis Contracts. The idea of this regulation is as follows: previous earnings of bank managers are subjected to a retroactive tax whenever a banking crisis compels governments to bailout banks. This retroactive tax is not imposed on those banks which need to be bailed out, but on all managers in the entire banking sector. In this way, the regulator imposes a kind of collective liability on bank managers.
In our joint research project, we have developed a modeling framework in which the idea of Crisis Contracts is analyzed. The results of this research have been reported in our working paper entitled “On the Economics of Crisis Contracts.”
The question of interest in our paper is whether Crisis Contracts can be effective in deterring risk-taking, reducing the probability of a banking crisis, and enhancing social welfare. In our research paper, we construct what we believe to be the simplest model economy that can be used to address this issue. In our model, bankers choose between safe or risky investment strategies. If more than some specified proportion of bankers choose a risky strategy, a systemic crisis may be triggered. In the event of such a crisis, part of the losses is absorbed by the taxpayer through government bailouts. Therefore, bankers have incentives to take excessive risks. We develop a game theoretic model in which the aforementioned strategic interaction among bankers takes place in two subsequent periods. In the absence of Crisis Contracts, all bankers choose a risky investment strategy in both periods, leading to a maximal probability of a crisis.
Next, we introduce Crisis Contracts into our model. That is, bankers’ pay from the first period becomes subject to retroactive taxation if and when a crisis occurs in the second period. We establish conditions on basic model parameters under which the aggregate level of risk, as well as the probability of banking crises, are reduced by the introduction of Crisis Contracts. We show that the relevant parameter conditions are suitably interpreted as equivalent to sufficient capital adequacy requirements. Moreover, there is a subtle relationship between both kinds of regulations in our model economy; for Crisis Contracts to be effective, some minimal capital adequacy requirements must be met. In the presence of such minimal capital adequacy requirements, it is, to some extent, possible to use Crisis Contracts and capital adequacy requirements as substitutes. A somewhat higher retroactive tax on bankers’ pay can make up for a somewhat lower capital adequacy requirement, and vice versa.
In sum, our first pass analysis suggests that Crisis Contracts may be an effective and valuable complement to capital adequacy requirements, and may enhance their effectiveness. While similar results could be obtained by imposing explicit caps on bankers’ pay, we argue that Crisis Contracts have a number of advantages over such pay caps. For instance, pay caps require the regulator to gather more information about payment practices. Moreover, contrary to pay caps, Crisis Contracts do not curtail the freedom to contract, and they only become effective once a crisis has actually occurred.
As it has been pointed out, our research goal has been to analyze the effects of Crisis Contracts in the framework of a formally sound game theoretic model of bankers’ strategic (investment) behavior. In order to conduct such an analysis, it is vital to think of the economic environment in a highly stylized way, abstracting away from many aspects which play a role in reality. For instance, one limitation of our model is that the risk of a systemic crisis in one time period depends only on the investment decisions made in that same time period. Consequently, our analysis does not account for systemic risks which build up gradually over time.