VoxEU Column Financial Markets International Finance

Systemic liquidity risk: A European approach

How should financial regulators address problems stemming from liquidity risk? This column argues that the liquidity coverage and net funding ratios proposed for Basel III are economically and politically impractical. It recommends using those ratios as long-term targets while imposing ‘prudential risk surcharges’ on deviations from the targets.

The repeated bursts of financial distress in Europe in 2010-11 reflect vulnerabilities built up in the previous decade and are germane to the roots of the credit crisis.

Abundant global liquidity relaxed funding constraints for banks and their borrowers, whether governments, firms, or consumers. Private and public debt grew faster than domestic savings as they were funded externally, by wholesale funding. Such funding is cheap because it is short-term, uninsured, and uninformed, and therefore prone to runs. This classic problem of ‘hot money’ for developing countries has now reached developed economies, since they have become large net borrowers.

Credit grew fastest in the Eurozone’s periphery, where the stability induced by the euro eased historical concerns about private productivity or fiscal laxness. Banks abandoned organic growth on local business credit, and escalated lending to unsustainable real estate booms and excess public consumption. As this balance-sheet expansion was built on a very unstable funding structure, Eurozone banks are now visibly over-reliant on jittery wholesale credit flows.

A radical new architecture is needed to restore proper credit incentives and strengthen resilience, moving banks away from a failed business model. Central to this transformation is to steer a desirable structure of bank funding. A banking system based on more stable funding (retail deposits and informed, long-term investors) would also promote a focus on local business credit opportunities, moving away from the oversized carry trades, investing in risky global assets and funding it with unstable global liquidity.

Prudential control of liquidity risk under Basel III

In the ongoing regulatory reform, attention has focused on raising capital ratios. Under Basel II, capital charges were inadequate, relying excessively on credit ratings and industry modelling. The new proposals are a large improvement, and a broad agreement has now been reached.

Basel III also seeks to address liquidity risk, recognized as the biggest gap in Basel II. Understanding bank equity is, unfortunately, much easier than understanding liquidity risk. After all, even a bank with a 10% adjusted-capital ratio usually has a 95/96% debt-to-assets ratio, after removing capital discounts for safe assets (such as all Eurozone sovereign debt!). So the key question is – how can we control refinancing risk for 95% of bank funding, given that banks are tempted to raise cheap, short-term funding and rely on central bank rescues should a run occur?

Academic opinion agrees that unstable funding imposes a negative risk externality. Both risk charges and mandatory ratios have been proposed to contain liquidity risk buildup (Perotti and Suarez 2009; Acharya, Khrishnamurti and Perotti 2011). However, the Basel III proposals have been cast solely in terms of ratios.

Two standards have been introduced:

  • Liquidity coverage ratios: buffers of liquid assets as a fraction of less stable funding.
  • Net funding ratios: quantitative limits to short-term funding.

While very tight levels of such ratios can ensure stability, this approach is too narrow.

Ratios are:

  • Too rigid. As they are fixed, they need to be set tight enough to be adequate all the time. As a result they are easily characterised as very expensive, and create massive resistance.
  • Not countercyclical. In fact, buffers are clearly procyclical (Perotti and Suarez 2010). The reason is that buffers discourage aggregate net liquidity risk only if they are costly. But in good times, the wholesale funding spread for banks is minimal (it was basically zero in 2004-07), while it jumps in a crisis. So unstable funding exposure is not discouraged, and net exposure will be the same as without buffers. Even ex post, buffers are clearly insufficient to contain systemic runs.
  • Very distortionary (relative to charges). They penalise the more efficient lenders, which will be rigidly constrained. This is analogous to the reason why quotas are usually less efficient than tariffs.

On the positive side, quantity limits on unstable funding are a robust solution when banks are very undercapitalised. In that case, many banks are too prone to gambling to rely on charges alone. Yet in practice, when banks’ capital ratios are weak, authorities are forced to offer extensive liquidity support, and to abandon tight standards. So ratios would work because they are very constraining, but will not be used precisely because they are.

It is extraordinary that contrary to a broad academic consensus, central banks’ lists of macroprudential tools do not consider alternatives to ratios, even though the approach is at serious risk of derailment.

  • The ratios have been successfully portrayed as very tight (as perhaps they need to be since they are set once and for all), and unaffordable in the current climate. The US regulators are currently under heavy lobbying pressure, and given the political climate, they are unlikely to adopt very tight recommendations. Inexplicably, the European Commission has neglected any reference in its CRD4 report to the most critical ratio, the net funding ratio. This omission has caused shock and concern, leading the ECB board to demand an explanation of the Commission’s intentions.
  • The introduction of these standards has been considerably delayed, with the liquidity coverage ratios not scheduled for some years, and the net funding ratios postponed for much longer. It is extraordinary that no European-level prudential measure will be in place for so many years.
  • Even the definition of these ratios has not yet been agreed. The definition of the net funding ratio standards is particularly controversial, and likely to be seriously weakened.
  • The buffer measures (liquidity coverage ratios) would require the creation of massive buffers given the current highly mismatched bank funding. These are seen as very costly, and there is an issue of insufficient forms of safe liquid assets to invest in anyway. Under current rules any Eurozone sovereign debt would qualify as a buffer, a curious prudential solution in the midst of the Eurozone sovereign debt crisis.
A key strategic choice

How, then, can regulators introduce prudential measures on liquidity risk that will be effective but not too onerous? The measures must also be introduced earlier than 2019 without disruption to be politically feasible.

A concrete solution, based on broad academic consensus, would be as follows.

Central banks have, during two years of Basel III negotiation, defined desirable liquidity positions in terms of standard ratios. These may be introduced as long-term targets, next to less demanding standards to be implemented immediately.

Banks may choose not to comply with the (higher) desirable standards because of individual circumstances or business model choices. In that case, they would be charged ‘prudential risk surcharges’ on the difference between the desirable and the actual ratios.

Risk charges may start quite low, certainly in a confidence crisis.

These fees would not reflect a direct insurance promise, but reflect the risk externality caused by individual bank strategies on systemic liquidity risk. As such, they represent a non-fiscal form of ‘bank taxation’ which targets risk creation, rather than transaction volumes.

The banks which would be more affected are those with the lowest retail deposits, which have expanded their balance sheets by relying on wholesale funding. This would rebalance the current bias where non-deposit funding is de facto insured but evades insurance charges.

The critical feature of ratios is that they may be adjusted countercyclically to stem excess growth of unstable funding. Raising charges would be much easier than adjusting ratios, as they have lower adjustment and disruption costs than quantity adjustments (which as a result are usually delayed for years).

Charges would be less rigid than absolute ratios, enabling individual banks to optimise their adjustment over time. Public disclosure would be limited to aggregate volumes.

The presence of the charges would ensure that:

  • Supervisors would be able to monitor on a constant basis the liquidity risk buildup at the individual and system level.
  • All banks will be induced to monitor the difference between the desired and current liquidity standing. Up to 2008, most large banks did not have a central tracking of their liquidity exposure.
  • Risky strategies could be discouraged in good times without raising interest rates.
  • At present, the authority for imposing such charges would lie with national central banks. For some, this step would require legislation.
  • Ideally, such a step should be coordinated at the level of the European Union. This would justify an EU directive proposal, ensuring a critical role for the approval process in Brussels for the CRD proposal.
  • International coordination of such charges would be desirable though not indispensable. There is no violation of the principle of level playing field, of course, if banks with different risk contributions were charged different amounts. An analogy would be with insurance premia, which may well differ across risk profiles. To illustrate this point further, it is clear that it would have been desirable in 2005-07 for Spain and Ireland to have higher charges on wholesale bank borrowings than, say, Germany, where there was no credit-fuelled real estate boom.
  • The natural locus of coordination to set rates would be the ESRB, which would gain an (indirect) macroprudential tool and thus a role. This is consistent with the European Parliament’s stated preferences for a more concrete empowerment of the ESRB as a macroprudential regulator.
  • Charges would accumulate in reserve funds for general financial-stability purposes. 



Acharya, Viral, Arvind Krishnamurthy, and Enrico Perotti (2011), “A consensus view on liquidity risk”, VoxEU.org, 14 September.

Brunnermeier, Markus, Gary Gorton, and Arvind Krishnamurthy (2011), “Risk Topography”, NBER Macroannual 2011.

Perotti, Enrico (2010), “Systemic liquidity risk and bankruptcy exceptions”, VoxEU.org, 13 October.

Perotti, Enrico and Javier Suarez (2011), “The Simple Analytics of Systemic Liquidity Risk Regulation”, VoxEU.org, 16 March.

Perotti, Enrico and Javier Suarez (forthcoming), “A Pigovian Approach to Liquidity Regulation”, International Journal of Central Banking.

Shin, Hyun Song (2010), “Macroprudential Policies Beyond Basel III”, Policy memo. 


1,470 Reads