There can be no banking system without a lender of last resort. Some countries have various mechanisms that provide lending in first resort when a bank fails. This can be a deposit insurance agency, a rainy day fund fed by bank dues, or an understanding that banks rescue each other when the need arises. But large banks or a collection of small banks typically have balance sheets that far exceed available resources. This is why central banks must be lenders of last resort, even when their role is clouded in constructive ambiguity to mitigate moral hazard.
Bank crises do not just involve lenders in first and last resort. Sizeable sums of money cannot be injected into a black hole. The authorities must have complete and up-to-date knowledge of each bank so that they can detect whether and how much money is needed in the case of failure. Because this is taxpayers’ money, the authorities must be able to determine the viability of a failing bank and, if needed, how to proceed to shut it down with minimal impact on depositors and other creditors, including deciding which creditors need protection and which should bear losses alongside shareholders. Because bank crises are often explosive and sometimes threaten to spread throughout the banking and financial sector, the decisions must sometimes be made in a matter of hours. The authorities must always be ready and perfectly informed. This is the role of supervisors and of the resolution authority that may be called upon to restructure or close the banks down.
While the details vary, all developed countries and many emerging market countries have developed such institutions. This is also the case in the Eurozone countries, and this is a major problem. As noted by Begg et al. (1998), with one central bank, Eurozone countries need one regulator, one supervisor, and one resolution authority, not N of them, where N is the number of member countries.
Like many other central banks, the ECB has always insisted that it is not a lender of last resort. The usual reason is that central banks wish to avoid giving commercial banks an implicit bailout insurance that would encourage risky behaviour. It is understood, however, that the central bank would step in if a serious bank crisis were to occur. But this is not the case of the ECB, for two good reasons at least.
First, the ECB has limited real-time knowledge of the situation of banks. Information can be requested but it will filter through national supervision authorities whose first allegiance is to their governments. High potential costs of a bank failure, closeness with banks, and protectionist feelings are likely to limit the free flow of information to the ECB. In addition, the ECB has no authority to close insolvent banks or even to design restructuring plans. Intervening as lender of last resort, the ECB would provide money without any control.
Second, lending in last resort can be costly; the ECB could suffer losses. In one country, whether the losses occur at the central bank or at the treasury makes little difference since the taxpayers are always the residual burden bearers. Within the Eurozone, ECB losses are borne by taxpayers from all member countries. The European Treaty stipulates that the ECB would intervene on behalf of the authorities of the failing bank, who would be responsible for any losses. The arrangement can work in the case of one bank with no cross-border activities. However, cross-border activities – encouraged to promote a single financial market – would open up litigations. In addition, the relevant government may be unable to live up to its legal obligation if it is already under pressure because of its own indebtedness.
The result is particularly disturbing, with many features of a Nash outcome remindful of the tragedy of the commons. It is in the interest of every member government with fragile banks to ‘share the burden’ with the other members. Given the size of the amounts potentially involved, national authorities have a strong interest to deny that any national bank is in difficulty for as long as possible, until the costs are so large that a central bank intervention becomes necessary. After the rescue, the authorities have a further interest in protecting its bank from being broken down as part of a resolution whose costs, if any, could be shared. Knowing this, the ECB has every reason to seek to avoid getting involved. Yet, when the costs exceed resources available to governments, the ECB cannot stand aside as a national banking system unravels, with a high probability of contagion to other countries.
This is why, in spite of the constructive ambiguity that some want to retain, the ECB must be recognised as the lender of last resort to Eurozone banks. At the very least, it must be accepted that there may exist crisis situations when the ECB will have to intervene, and consequences ought to be drawn. These consequences are that the Eurozone needs a single regulator, a single supervisor, a single resolution authority and, most likely, a common deposit insurance mechanism. This is what defines a banking union.
A banking union is politically difficult to fathom. It involves a transfer of competence from national to Eurozone authorities. It entails apparent income redistribution among countries. It requires the setting up of new institutions. Its need is only apparent at crisis time, even though its existence is bound to change incentives of both banks and governments. The consequences of these changing incentives are unpalatable to banks in quiet times inasmuch as they result in less risk-taking and less profits.
The clash between the required coherence of what has to be done and the predictable political opposition is particularly worrisome. At the current stage of the crisis, most Eurozone banks are not seen as close to failing. The Irish and Spanish wake-up calls have brought home the need for a banking union of some sort, but the threat of a systemic banking crisis remains remote. It matters little that good policymaking requires planning for the worst, resistance is bound to be strong as long as most countries feel safe. The risk is to adopt some but not all the measures that are part of the definition of a banking union.
Indeed, the proposal that the European Commission has put forth in early September 2012 only includes regulation and supervision. Resolution and a deposit insurance mechanism are postponed. Even worse, a number of governments want to strictly limit the number of banks that would fall under the European supervisor’s authority and to delay the project. The very same reasons and actors that blocked any discussion of these issues at the time when the Maastricht Treaty was under negotiation are back in action.
It is essential to understand that a partial banking union is no better than no banking union at all, and possibly worse. Imagine that we only have a single regulator and that the common supervisor only looks at large banks. Imagine that a series of governments, small and large, restructure their public debts, an occurrence that many regard as unavoidable. Because banks typically hold large amounts of national bonds, a public-debt restructuring is likely to cause deep losses in small and large banks. The 2007-8 scenario has amply shown how mutual suspicion promptly steps in and brings the interbank market to a halt. The ECB must then provide liquidity to individual banks, small and large. In addition, some banks may fail, bringing along others. The ECB is facing its role as lending in last resort. If it only supervises large banks, it cannot provide liquidity and, if need be, emergency assistance to the smaller banks. Financially hard-pressed governments will need to provide resources that they do not possess and cannot even borrow. Healthier governments may start down the road followed by Ireland and Spain and find themselves losing market access as well. Large banks too will be engulfed. With no resolution authority, the ECB will not feel able to inject amounts that could reach several trillions of euros. Apocalypse now.
This is a scenario, of course. It is not just plausible, it is reasonably probable.
Begg, D, P de Grauwe, F Giavazzi, H Uhlig and C Wyplosz (1998), "The ECB: Safe at Any Speed?", Monitoring the European Central Bank 1, CEPR.