Do we need a separate resolution fund?

Dirk Schoenmaker 14 January 2010



The 2007-2009 financial crisis was resolved through massive government support across the world. Calls are now being made to establish a separate resolution fund to deal with future financial crises (HM Treasury, 2009). Such a programme would be funded by premia levied on the financial sector. The question is whether such a separate resolution fund should be established. I would argue not. Crisis resolution should be considered a core government task run from the general budget. Even if no separate fund would be established, the government could decide to levy a separate tax on the financial sector.

A separate resolution fund?

Maintaining the stability of the financial system is an important objective of government. The rationale for a separate resolution fund1 to maintain financial stability is twofold. The first aim is to have sufficient funds available when a crisis hits. However, the incidence and severity of crises are difficult to assess on an actuarial basis. Before the current crisis, the incidence and severity would have been calculated on a low-medium base, reflecting occasional crises over the last 30 years (since the breakdown of Bretton Woods). After the current crisis, the incidence and severity would probably be calculated on a high base, reflecting the current crisis and the Great Depression. This is the classical problem of disaster myopia (Guttentag and Herring, 1984). Moreover, macro-events such as systemic crises are inherently difficult to predict. So, the size of the required funds cannot be estimated in any meaningful way.

A second aim is to internalise the costs of financial crises. This can be done by a Pigouvian tax, as set out by Perotti and Suarez (2009). The levy, charge, or premium on the financial sector is basically a tax, which should be set at a level to counter (i.e. neutralise) the negative externality caused by financial institutions. By its nature, this levy is not necessarily related to the required actuarial premium (insofar as such an actuarial premium could be calculated). A properly set systemic levy also addresses moral hazard as the levy is set in such a way to internalise the negative externality. In that way, the levy neutralises any incentive for risky behaviour.

Related arguments

Another argument against an ex ante resolution fund is its pro-cyclicality. Countercyclical premia suggest that levies should be higher in good times (i.e. economic upturn) and lower in bad times (i.e. economic downturn). The running of an insurance fund is typically pro-cyclical (Niehaus and Terry, 1993). In good times, premia tend to be lower as future risks appear benign. Even worse, premium holidays can be granted when the fund reaches its target size. This is pro-cyclicality in its extreme. Prior to a crisis, levies are reduced to zero, inducing further expansion of the financial sector and thus increasing the likelihood of a crisis. An ex post resolution fund to recover the costs of the current crisis is even worse, as premiums are charged in the current economic downturn (further weakening the banking sector) and do not address moral hazard.

Micro-arguments of fiscal discipline could support the establishment of a separate fund. To increase fiscal discipline and prevent governments from spending “free” resources separate funds could be established, such as basic state pension funds, export insurance funds, etc. However, other, more effective mechanisms are available to achieve fiscal discipline.

Take Sweden, which proposes to set up a fund of 2.5% of GDP. Whereas there could be some advantages in building up a masse de manoeuvre in advance, there are strong political arguments against that, since such ex ante contributions would raise the measured fiscal deficit. So, Sweden would be better advised to reduce its fiscal deficit by 2.5%.


Financial stability is a public good (De Haan et al. 2009), as the producer cannot exclude anybody from consuming the good (non-excludable) and consumption by one does not affect consumption by others (non-rival). I, therefore, consider public resolution of financial crisis as a core government task. Reviewing the arguments, I believe that the establishment of a separate resolution fund, financed ex ante or ex post, is not desirable.

Although a separate fund can be helpful as a first line of defence, it can never be foolproof. Sweden, for example, has set a target of 2.5% of GDP. The current public resolution costs have been well above this size in many countries. It may even give a false sense of safety. Moreover, the setting of a target size suggests that premia can go to zero when the target is reached. This is clearly pro-cyclical and counterproductive. So, a buffer-based approach (i.e. the fund acts as a buffer) has major drawbacks.

The general budget of the government is meant to bear macroeconomic risks, such as financial crises. At the same time, the government may wish to levy Pigouvian taxes on the financial system to counter negative externalities. This is a flow-based approach – cash outflows in times of crises and cash inflows in stable times. As I have argued above, the inflows and outflows need not necessarily match. As with any government task, there is no need to balance each task separately. The only constraint on governments is to achieve an overall balance of inflows and outflows over time.


1 The set up of a resolution fund to recapitalise banks in difficulties is different from a deposit insurance scheme. A deposit insurance scheme is privately funded by banks. A separate ex ante funded deposit insurance scheme makes sense, as it allows charging risk-based premia reducing moral hazard and prevents an extra charge on banks in an economic downturn when payouts are usually needed. To be credible, a deposit insurance scheme typically has a government backstop guarantee. In line with my arguments against a resolution fund, there is no need to pre-fund this guarantee.


De Haan, J., S. Oosterloo and D. Schoenmaker (2009), European Financial Markets and Institutions, Cambridge University Press.

Guttentag, J. and R. Herring (1984), “Credit Rationing and Financial Disorder”, Journal of Finance, 39, 1359-1382.

HM Treasury (2009), “Risk, Reward and Responsibility: The Financial Sector and Society”, Discussion Document, London.

Niehaus, G., and A. Terry (1993), “Evidence on the Time Series Properties of Insurance Premiums and Causes of the Underwriting Cycle”, Journal of Risk and Insurance, 60, 466-479.

Perotti, E., and J. Suarez (2009), “Liquidity Risk Charges as a Macro-Prudential Tool”, CEPR Policy Insight, No. 40, London.



Topics:  Macroeconomic policy

Tags:  financial crisis, Pigouvian tax, resolution fund

Professor of Banking and Finance at the Rotterdam School of Management, Erasmus University Rotterdam; Senior Fellow, Bruegel; and Research Fellow, CEPR