Tax banks to discourage systemic-risk creation, not to fund bailouts

Enrico Perotti 07 February 2010



The burning issue of funding the bailout has finally led to the first policy action on financial taxation. The good news is that it is not a Tobin tax on all financial transactions, which would be a very crude and distortionary solution. Financial intermediaries have indeed grown too large, but discouraging all financial transactions suppresses activity and fails to target problematic practices.

The Obama proposal for bank taxation has simple flat rates on uninsured bank liabilities. This is a better target than total liabilities since deposits were already insured, and the intervention bailed out wholesale funding.

But is such a flat tax designed to control risk creation? John Kay (2010) argues against it. Meanwhile Viral Acharya and Mathew Richardson (2010) argue that the bailouts have generated more moral hazard and suggest a fee discouraging all activity that creates systemic risk – not just leverage – and moreover that banks should be paying more in the good times when risk taking is more attractive.

In recent research, Javier Suarez and I (2009a, b) suggested a more subtle policy than President Obama’s – a Pigouvian tax based on banks’ individual contribution to systemic-risk creation, measured by their exposure to uninsured short-term funding. As in the Obama tax, this approach exempts insured deposits and targets the risk of sudden withdrawals of wholesale funding, which was the engine of the last crisis. Critically, our tax is sharper for shorter-term funding and decreases to zero for medium-term liabilities that do bear risk. In other words, it targets the externality caused by funding fragility and offers strong incentive effects in good times.

Liquidity charges are complementary to countercyclical capital requirements. Higher capital ratios will control asset risk and improve risk absorption, but they would not stop systemic propagation during a panic. When losses lead to rapid withdrawals of massive amounts of uninsured funding, they spread to other markets by forcing fire sales, which in turn trigger more frantic deleveraging.

Liquidity charges contain risk without relying exclusively on restrictions on admissible investments. It would discourage banks from running large proprietary trading with cheap short-term funding –  and in particular from playing a simple carry trade that adds little value to economic activity. Importantly, it would charge intermediaries in good times, raising the cost of opportunistic risk creation when intermediaries grow quickly with unstable short-term funding for investing in risky assets.

Unlike capital requirements, liquidity charges raise revenues. A fraction should go to general tax revenues, because bank instability hurts the whole economy, a part may flow into a bank stability fund.

Finally, liquidity charges (in excess of a basic tax) are a natural macro-prudential tool. Because central banks and supervisors serve the two goals of monetary and financial stability, an instrument distinct from interest rate policy is needed for financial stability. Raising the interest rate is too blunt, as it hurts the whole economy and is therefore used too sparingly. Under our proposal, macro-prudential authorities would be able to adjust tax rates (or surcharges) to slow down rapid credit creation and risk accumulation in a timely fashion. The best discipline for timely intervention is the attribution of a distinct tool and associated responsibility to a coordinating entity such as the European Systemic Risk Board.

The effect of liquidity charges would be to induce the financial system to rely less on unstable short-term funding and create an opportunity cost for simple carry trade strategy of funding high yield risky position with cheap but fleeting borrowing. Proprietary trading that adds no value would be discouraged (not just for banks) without forcing a generalised prohibition.

There is really no reason why so much of the financial intermediation is funded with near-demandable debt that does not correspond to retail depositor needs. In the crisis, short-term investors which were unwisely supporting risky strategies were able to escape in time. We need to do less ex post insurance and more ex ante discouraging. Yet it is important to do using well targeted taxes, not just draconian market segmentation and quantity regulation.


Acharya, Viral and Mathew Richardson (2010), “Making sense of Obama’s bank reform plans”,, 24 January.

Kay, John (2010), “Why ‘too big to fail’ insurance is the worst of all worlds”, Financial Times, 2 February.           

Perotti, Enrico, and Javier Suarez (2009a), “Liquidity Insurance for Systemic Crises”, CEPR Policy Insight 31, February.

Perotti, Enrico, and Javier Suarez (2009b), “Liquidity Risk Charges as a Macro-prudential Tool”, CEPR Policy Insight 40, November.




Topics:  Financial markets

Tags:  bank regulation, Pigouvian tax, Obama's bank reforms

Professor of International Finance, University of Amsterdam; Research Fellow, CEPR


CEPR Policy Research