The events in Cyprus during the past few weeks have put in the spotlight what, in a European context, has almost been a taboo: can a major bank be allowed to fail? Looking back at recent financial history, it is difficult to find examples of fully-fledged bank failures where creditors have been forced to take losses. Bank rescues have led to the “too big and too interconnected to fail” proposition. This subverts the incentives for bank owners and, not least, for creditors, since the infrequency of lenders being faced with losses has entrenched moral hazard in the system.
Why is it that banks have not been allowed to fail? One factor is precedent: by not letting banks fail, expectations of a similar response in the future have been built into the system. This has triggered a self-fulfilling circle. Time inconsistency also plays a role: the rewards from letting banks fail – reducing moral hazard and the risk of the situation repeating itself – are long-term, whereas the costs – in the form of financial destabilisation and voter outrage – are likely to be felt imminently.
There is no concept more fundamental in economics and finance than that of expectations. When unexpected events occur, disorder is likely. This is why the plans to impose haircuts on depositors in Cyprus revealed on 16 March caused such disquiet. Not only were deposits in general to be hit, but those below the insured level of €100,000 would also be subject to bail-in, i.e. write-downs and/or conversion of their claims into equity. Cyprus therefore indeed seemed like a harbinger of change – and an unexpected change – in terms of how bank failures are dealt with in Europe.
As the European Parliament's lead negotiator on the Bank Recovery and Resolution Directive, which seeks to establish a European framework of measures to address these very issues, I have been following the Cypriot case closely. Without commenting on the Cypriot example in itself, a number of points that I think are fundamental deserve highlighting.
Firstly, management of bank failures must be as predictable as possible. One cannot alter the rules of the game without adequate prior notice: if that happens, even the most well-considered and justified line of action will be detrimental to market and depositor confidence. Secondly, the bail-in tool is complex to use. It has many benefits, not least in terms of its pre-emptive effects and the risk awareness it will create. However, risks of contagion could cause worse problems than those which initially triggered the action, especially in banks whose only creditors are depositors.
And thirdly, bail-in is unlikely to be sufficient in crises of systemic proportions. When the very infrastructure of our economies, such as the system of payment, is in peril, the government has a responsibility to act: assuming ownership provides taxpayers with the best chance of getting their money back.
By not allowing for government intervention as a last resort, future bank rescues will be more or less solely dependent on bail-in. In a systemic crisis, this might cause more problems than it solves and thus reduce the possibility of bail-in being used at all. If markets know governments may intervene, bail-in will be more credible. Thus rescuing banks can produce both a reduction of moral hazard in the banking system and an enhancement of its systemic stability.
Gunnar Hökmark is an MEP, vice-president of the EPP Group and rapporteur for the Bank Recovery and Resolution Directive.