The European Commission (2012) has now presented its legislative proposal for a banking union whose key element is a ‘Single Supervisory Mechanism’ to be headed by the ECB, but it says nothing about deposit insurance at the national level. Is that viable?
Setting the incentives
Economists use a ‘backward’ approach when looking at the link between supervision and deposit insurance and resolution. The endgame of deposit insurance and resolution drives the incentives for ex ante supervision (Schoenmaker 2013).
A key point here is that the purpose of good supervision is not to prevent banks from taking any risks and thus make sure that no bank ever fails. The purpose of good supervision is to equilibrate the relationship between risk and reward for the private sector, especially for bank managers and the owners of banks. Growth and innovation would be stifled if banks were not allowed to take any risk. With good supervision bankers and their investors would, however, be forced to accept the consequences if any risky investment goes awry – possibly up to the point that the bank has to be closed or restructured. This implies that there will still be bank failures even if the ECB were to do the best possible job as the supervisor of the EZ’s banking system.
Dewatripont and Tirole (1994) stress the point that as depositors are guaranteed, they will no longer have an incentive to monitor the bank. Normally the supervisor then takes over the monitoring role representing the depositors. This is naturally the case at the national level, where both the supervisor and the deposit insurance system are part of the same government. But this would not be Europe if only supervision were centralised and national authorities remained responsible for deposit guarantee and restructuring. The ECB would have an incentive to offload the fiscal cost of any problem to these national authorities.
As long as deposit insurance and resolution were to remain at the national level serious conflicts would necessarily arise if the ECB thinks that any given bank needs to be restructured or closed down. The ECB would do this on the basis of its assessment of the viability of the bank and any danger it might represent to systemic stability at the EZ level. By contrast the national deposit guarantee systems, and more generally the national authority responsible for bank restructuring (i.e. in practice today’s supervisors and finance ministries), would have a tendency to minimise their own costs by keeping the bank alive through support from the ECB. National authorities would have naturally a tendency to blame an ‘unfair’ ECB for not recognising the strength of ‘their’ bank which should not be closed, but saved. This type of conflict is likely to be especially prevalent at the start of the new system when the ECB has to discover all the skeletons in the closet hitherto hidden by national supervisors.
Over time other conflicts will arise. For example, if the ECB has made a mistake and let a bank take too much risk. National authorities would then have a point in complaining if they had to pay up for the costs.
The best way to avoid these potential conflicts and provide the new EZ supervisor with proper incentives is to gradually move deposit insurance and resolution to the EZ level as well, thus ensuring eventually the needed align of responsibilities. A gradual introduction would ensure that both national and EU-level authorities have ‘a skin in the game’ during the transition. See Schoenmaker and Gros (2012) on how to gradually introduce a European Deposit Insurance and Resolution Authority (EDIRA).
In the US, a two-tier system emerged with banks chartered at the state or federal level. In the end the smaller, state-chartered banks became less and less important. This analogy suggests a compromise: groups of smaller banks with their own mutual guarantee system (e.g. savings or mutual banks) might not fall under the direct supervision of the ECB, with national supervisors, who have a better local knowledge, remaining responsible for the day-to-day supervision of these banks.
However, a large group of small banks with a similar business model can also represent a danger for systemic stability. This implies that, the national supervisors would remain responsible towards the ECB for the systemic stability of these groups of banks and that the ECB should be given the right to require all the information required to assess the stability of these groups.
Systemic crisis: National versus EZ
A standard objection against a common deposit insurance fund based on the FDIC model is that it would be too small to deal with systemic crisis. This is undoubtedly true for a systemic crisis at the EU or EZ level. However, one should also ask the counter question: would the system be more stable with the existing patchwork of national schemes, only some of which are actually pre-funded? The answer must obviously be no. A common, prefunded, deposit insurance and resolution system could only improve the situation.
Moreover, in Europe real estate booms and other shocks will remain for some time concentrated at the national level. Systemic crisis are thus more likely to arise at the national level than at EZ level (Spain and Ireland provide vivid examples).
Recent IMF and World Bank studies find that the 'typical' fiscal cost of a crisis seems to be about 5% of GDP. If one accepts this figure for the 'typical' fiscal cost of a crisis it follows that an EDIRA endowed with a fund of €50 billion should be sufficient to handle a 'typical' crisis for all smaller member countries (like Greece, Portugal or Ireland today) and possibly even for Spain with a GDP of €1,000 billion. Moreover, in case of a very large crisis the funding base of the EDIRA (the EZ banking system) would be much stronger and more credible should the costs go above this figure.
For a systemic crisis at the EZ level it would anyway be a question of the ECB intervening, because at that point the fiscal costs be too large for any mere deposit insurance scheme. Moreover the ESM, which in our mind would provide the ultimate fiscal backstop, would be able to deal with a ‘typical’ crisis even at the EZ level (5% of about €10,000 billion GDP for the EZ is about equal to the €500 billion lending capacity of the ESM).
Moreover, an EDIRA would be able to considerably diminish the threat of deposit flight due to convertibility risk in any country subject to a financial crisis. Depositors in any country subject to a systemic crisis (at the national level) will ask themselves whether their deposits are safe given that the national deposit guarantee system will be too small to deal with the problem, and given that the government itself might experience financing difficulties. This means that national deposit insurance can do little to prevent national bank runs, much as in the US of the 1930s state guarantee systems could not prevent local-level bank runs.
The debate about Banking Union is running into the typical chicken-and-egg problem: Most academic observers agree that deposit guarantee and resolution should be organised at the same level as supervision. But at present only the creation of a ‘Single Supervisory Mechanism’ (SSM) to be headed by the ECB is being discussed; with deposit insurance and resolution to be considered only later when this SSM has shown its effectiveness.
We argue that the SSM is unlikely to be working well unless a European Deposit Insurance and Resolution Agency is introduced gradually at the same time.
Dewatripont, M and J Tirole (1994), Prudential Regulation of Banks, Cambridge (MA): MIT Press.
European Commission (2012), A Roadmap towards a Banking Union, COM(2012) 510 final, Brussels.
Golembe, C (1960), “The Deposit Insurance Legislation of 1933”, Political Science Quarterly, 76:181-195.
Pisani-Ferry, J and G Wolff (2012), “The Fiscal Implications of a Banking Union”, Bruegel Policy Brief, Issue 2012/02.
Schoenmaker, D (2013), Governance of International Banking: The Financial Trilemma, Oxford University Press, forthcoming.
Schoenmaker, D and D Gros (2012), “A European Deposit Insurance and Resolution Fund: An Update”, Duisenberg School of Finance Policy Paper, No. 26, and CEPS Policy Brief No. 283.
Thies, C and D Gerlowski (1989), “Deposit Insurance: A History of Failure”, Cato Journal, 8:677-693.
Appendix: A lesson from US history
A federal deposit insurance system (FDIC) was introduced in 1933 as part of the wide ranging reforms implemented by the incoming Roosevelt administration which had to deal with a wholesale collapse in the banking system. Interestingly, the US Congress took the initiative for federal deposit insurance. Henry Steagall, chairman of the House Committee on Banking and Currency, introduced federal deposit insurance in the Banking Act of 1933. The US Treasury Secretary was against, but lost the case. The creation of the FDIC did succeed in re-establishing almost immediately confidence in the banking system which had been shaken because deposit guarantee had up to then be a responsibility of the state level and most of the deposit schemes at the state level had become insolvent (Golembe 1960).
There were two obvious reasons for the weaknesses of the state deposit insurance systems (Thies and Gerlowski 1989):
- Small undiversified banks exposed to local real estate bubbles and agricultural difficulties;
- Excessive risk taking due to insurance (moral hazard).
The first reason is very true today. The Spanish and Irish deposit insurance funds cannot handle the multiple failures of small undiversified banks resulting from a local boom and bust. Federal insurance would diversify this risk of local shocks.
Of course, good supervision is needed to prevent excessive risk taking and limit the exposure of the insurance fund. That brings us back to Dewatripont and Tirole (1994), who argue that the supervisor should monitor banks on behalf of the deposit insurance fund.
This is of course not to say that the FDIC solved all problems besetting the US banking system at the time or today. However, the FDIC was an important factor in establishing confidence and the system has by now worked successfully for eight decades, resolving hundreds of banks at minimal cost to the taxpayer. Even during the banking crisis the FDIC was able to take over a large number of banks, pay out depositors and restructure the rest, thus contributing to the stability of the system as a whole. But it is not yet clear how the FDIC would handle large failures. Only after the crisis, the Dodd-Frank Act also gave resolution powers for the large banks to the FDIC. Of course, federal insurance is helpful, but not a foolproof system. The savings and loans debacle, during which many small savings and loans were exposed to the interest rate shock of Volcker, shows that a federal fund can also be insufficient if the problem is not local but widespread.