Should the ECB be lender of last resort?

A guiding principle for decision-making in crisis management is “he who pays the piper calls the tune” (Goodhart and Schoenmaker 2009). So long as resolution is organised on a national basis, the national governments will normally want to oversee and undertake the function of supervision. That has been the current setup for financial supervision and crisis management, which have been nationally organised.

Following this principle, the ECB clarified that the national central banks should remain the lender of last resort (LOLR) for individual banks with problems at the time when EMU started (Padoa-Schioppa 1999). This is the individual LOLR function. By contrast, the ECB had assumed the more general aggregate liquidity support function from the start of EMU, as this general role is intrinsically linked to monetary policy operations.

Lender of last resort in the banking union

Supervision and resolution will shortly be moved to the European level in the banking union. The question then arises: what should happen with the individual LOLR function? As LOLR is typically not codified in legislation, this is currently a decision for the national central banks in the Eurosystem. Recent speeches (e.g. Lautenschläger 2014) seem to suggest that the ECB is planning to keep this LOLR function at the national level.

However, the national central banks cannot call the tune anymore in the banking union. The ECB will supervise the significant banks (these are the 120 largest banks in the banking union; the national supervisory authorities remain responsible for the local banks). We argue that the individual LOLR function for those significant banks should then move – in tandem with supervision – to the ECB.

What government involvement?

While traditionally central banks assumed the LOLR function on their own account and risk, the national government, as ultimate risk-bearer, will usually insist on having a say.

Treasury involvement can be at different levels. The first option is that the central bank has the freedom to provide emergency liquidity assistance to illiquid but solvent banks (although we would argue that the notion of an illiquid but insolvent bank is a misnomer; the need for liquidity assistance often arises because banks cannot get funding in the market due to doubts about their solvency). An example is the Netherlands, where the Netherlands Bank can provide special credit in exceptional cases (Dutch MoU 2007). There is no need for government approval.

The second option is prior Treasury approval. Because of the risk to public funds, the Memorandum of Understanding on Financial Crisis Management between HM Treasury and the Bank of England requires Treasury approval for any emergency liquidity assistance provided by the Bank (UK MoU 2012).

The third option is a full government guarantee (and thus also approval). The ECB is, for example, only prepared to consider buying the riskier tranches in the new asset buying programmes if these tranches have a government guarantee (Constâncio 2014). In that way, the ECB can protect its balance sheet.

Although central banks are not for profit (Goodhart 1988), the LOLR function is part of their banking functions. Central banks should therefore take the necessary precautionary measures, such as collateral, whenever possible to foster good risk management. We thus consider the third option as less appropriate. The first option is easy to implement. For the second alternative, the ECB needs to find the ‘European minister of finance’ for approval. The president of the Eurogroup could perform that function. That is consistent with his role as chair of the Board of Governors of the European Stability Mechanism, the fiscal backstop of the participating member countries.

Political tensions

The stated aim of the banking union is to break the diabolic loop between banks and governments in the Eurozone. The Single Supervisory Mechanism and the Single Resolution Mechanism therefore move supervision and resolution from the national government level to the Eurozone level. We argue that the LOLR role for the significant banks should likewise be moved to this level.

Nevertheless, this new banking union setup can lead to political tensions between the centre – the ECB (as supervisor and LOLR) and the Single Resolution Board – and the national authorities.

A first tension is that the ECB may be too tough. In its supervisory capacity, the ECB can decide to hand over a bank to the Single Resolution Board. It may push a bank into resolution without providing LOLR assistance. We can, for example, imagine that the ECB may refuse emergency liquidity assistance to a bank well before the national government of the country where the bank is headquartered is prepared to accept that. While most of us would probably accept that such earlier closure might be the right thing to do, it may set off a political row, with the national government claiming that the ECB’s actions are premature, ultra vires, etc.

Such political rows can disturb the cohesion of the Eurozone. It is basically a conflict of differing interests. The ECB and the Single Resolution Board have a remit for the soundness and stability of the banking union as a whole, while national governments worry about their national banking systems. With the move to a banking union, European legislators have decided that the Eurozone interest should prevail. The Eurogroup, with a mediator role for its President, is likely to be the decision centre where such a political row needs to be settled, though the Eurogroup cannot give orders to the ECB or the Single Resolution Board. This was already the case in the pre-banking union era, when the President of the Eurogroup, for example, played a leading role in the rescue of the Cypriot banking system.

It should be noted that the ECB with its new supervisory remit from 4 November has the power to hand over a significant bank to the Single Resolution Board (by withdrawing that bank’s licence) if it deems such a bank insolvent. The ECB can, and will, exercise that power regardless of taking up the LOLR role.

Another tension is that the ECB may be too lenient and fail to prevent failures. If the ECB were to be the LOLR, it would face the consequences of (the lack of) its supervisory actions. That would make the system incentive compatible. Moreover, the President of the Eurogroup may restrain the ECB from overly generous LOLR support for ailing banks. Failing to assume the LOLR role could cause a political backlash against the ECB if a national central bank had to provide LOLR as a result of (perceived) supervisory failings of the ECB. The national central bank would have to pay up, even though it was not primarily responsible for the failure.

In conclusion, we argue that the ECB should also take on the LOLR role for the significant banks when it starts supervision. That would make the governance system incentive compatible (i.e. provide the right incentives to the ECB for conducting good supervision). By the same token, the national central banks should keep the LOLR role for the local national banks, as the primary responsibility for supervising these banks would also remain at the national level.

Published in collaboration with VoxEU

Authors: Charles Goodhart was the Norman Sosnow Professor of Banking and Finance at the London School of Economics until 2002; he is now an Emeritus Professor in the Financial Markets Group. Dirk Schoenmaker is Dean of the Duisenberg School of Finance and Professor of Finance, Banking and Insurance at the VU University Amsterdam.

Image: An illuminated euro sign is seen in front of the headquarters of the European Central Bank (ECB) in the late evening in Frankfurt January 8, 2013. REUTERS/Kai Pfaffenbach