Greek lessons for central banks acting as lenders of last resort
30.07.2015
Lender of last resort
Greek lessons for central banks
Greece has returned to the news headlines over the past few weeks, following a period when fears of a crisis in the country appeared to fade due to reports of positive growth and a primary budget surplus. But the new Syriza government, elected in January, wanted to renegotiate Greece’s debt and change the conditions of financial support program which ultimately resulted in a decline in confidence in its ability to repay its debt. As is often the case in these situations, the first victims were banks, as depositors once more started to withdraw their funds due to concerns about the ongoing viability of the financial system.
Side head: ‘LoLR in action’
Best practice dictates that after commercial bankers have used up their liquid assets
to meet the outflow of deposits, they turn to their central banks in their capacity as ‘lenders of last resort’ (LoLR) to seek additional funding. As long as the problem is related to the liquidity of banks rather than their solvency, loans provided by the central bank via its LoLR function are safe, because the nature of the LoLR lender must not be fiscal in nature and the credit extended should be repaid in due course.
There are at least two good reasons for central banks not to engage in fiscal activity. First, it should be the role of government and not the central bank to increase the debt level of a country, and so placing a burden on future public budgets (and ultimately to tax payers). Secondly, LoLR activities should not raise issues related to compliance with anti-competition rules that are relevant in most developed economies. Any government support for failing companies should be the subject of an anti-competition assessment. To ensure LoLR activities remain outside these areas, the liquidity provision must be clearly based on credit and not be hidden capital injections. As a result, most central banks can provide LoLR services only against strong collateral.
In practice, from time to time, losses from central bank credit lines (LoLR mainly) do occur, for various reasons. Sometimes a bank is actually in worse shape than the central bank expected. On other occasions the prevailing environment facing institutions may have deteriorated rapidly, due to external conditions such as the state of the economy or the markets involved. However, even in these situations, the losses from LoLR operations tend not to be too detrimental to the credibility of the central bank. So, despite the ‘quasi-fiscal’ nature of such loans, there was no demand by central banks to make the system of LoLR less risky by, for example, asking for guarantees from their respective governments.
LoLR, Greece and the eurozone
The situation is even more complicated in the case of Greece, as it is a member of the eurozone and the Euro system. As a result, the LoLR funds – called emergency liquidity assistance (ELA) – while provided by the local central bank, the Bank of Greece, in effect sit on the accounts of the European Central Bank (ECB). Eligible collateral for ELA includes Greek government paper, as well as other domestic financial assets. As a result in current situation, the future quality of this collateral depends fully on the result of government negotiations with creditors, that will preserve ability of Greek government to service its debt.
For many months, the Greek government, led by prime minister Alexis Tsipras, was unwilling to accept some of the structural reforms and fiscal deficit cuts that were the conditions of providing badly needed financial resources – as Greece lacks access to financial markets. Without the financial support of new loans from eurozone countries, the Greek government could not service its debt and pay government bills, such as state employee wages and pensions.
Negotiations were deadlocked for weeks, with the Greek government refusing to accept the conditions placed on loans and its creditors unwilling to provide money without adherence to a strict programme of economic reform. During this period, the ECB provided tens of billions of euros in ELA, to some extent substituting loans that could have been provided by creditors – the ECB also allowed banks to buy government bills, which were accepted as collateral for ECB loans.
The rationale for the ECB’s decision is not difficult to understand:
- The banks were solvent before the crisis and were suffering a liquidity crunch;
- The government is set to negotiate new loans that will restore its solvency, so collateral is of good quality, assuming the talks are successful; and
- If the ECB stopped providing ELA, thus preventing the Bank of Greece from offering liquidity support, Greek banks would collapse, making the negotiation of new loans irrelevant under current terms, as the economy would change dramatically with the fall of the banking sector. In effect, the ECB could be deciding Greece’s economic future.
The ELA does not come without consequences, however. The ECB’s exposure to Greek credit has increased substantially (nearly 90 bil. Euros of ELA loans were provided). So, while eurozone governments were negotiating their credit lines with the Greek government and pretending that the eurozone exposure to Greece would not rise until an agreement was reached, the central bank had to continue to increase its Greek risk. While governments were hesitating on how best to deal with ‘the Greek problem’, the ECB was forced to replace their actions by offering ‘invisible’ loans, which kept its economy afloat, while pushing the ECB’s total exposure to Greece closer towards the magnitude of loans provided by the eurozone (loans provided vie different programs by ESM/EFSM reached 130,9 bil. Euros, while total exposure of ECB was 117,9 bil. at the same time).
What if Greece fails?
The weakness of this approach has become clear. In the event that loans to the Greek government were not provided, the ECB would face a very different situation, namely:
- The risk of the Greek government defaulting on its euro-denominated collateral has increased, and after it failed to pay an instalment of IMF loan of 1,6 bil. Euros at the end of June, Greek bonds can hardly be considered good-quality collateral;
- The quality of assets held by banks has deteriorated, directly due to the weakness of government credit and indirectly due to the expected decline of the economy in the event of government default. The capital of commercial banks is suffering, so they have a lower solvency level compared with a few weeks ago;
- The ECB, due to decline of quality of collateral and other consequences of Greek non honouring debt payments, cannot provide more ELA, but without this, Greek banks will not have sufficient liquidity to serve clients (which resulted in their ‘temporary closure’ in the form of an extended bank holiday, and even after they reopened there were limits had to be placed on withdrawals. Even so, the banks will run out of money soon ); and
- As the new supervisor (directly or indirectly) of European banks, the ECB could declare a bank insolvent, but since it provides liquidity it is in a conflicted position (insolvency of banks would likely cause large loss for ECB, while extension of ESM program can support capital of Greek banks).
While it was clear that even in circumstances where there was a high probability of another breakdown in negotiations between Greece and creditors, it was in the best interest of Europe for the ECB to ‘stay calm’ and wait for a solution. But this is inconsistent with the ‘technocratic’ working of the LoLR function, as the ECB faces substantial risks in the event that a long-standing agreement cannot be reached, which would be very bad for financial interest of the ECB.
- The Greek banks will become illiquid and eventually insolvent;
- The ECB’s exposure to the Bank of Greece would no longer be considered ‘safe’ and it would have to report a loss at least in magnitude to the portion of ELA provided;
- This loss would offset the future profits of the ECB for many years or trigger a call for shareholders to increase its capital;
- Even non-eurozone members would share this loss (in the case of capital increase), despite not being involved in the Greek loan negotiations; and
- This might prompt some people to consider the ECB’s provision of ELA to Greek banks insufficiently prudent.
LoLR and the future
A number of major issues are raised by the ECB’s extension of ELA to Greece.
Firstly, acting as a LoLR is a risky business. In many cases, the risk is so high that a central bank must implicitly accept a some probability of loss. For a country with its own currency and central bank, the consequences are ‘national’, and therefore lie between the local central bank and local politicians. But in a monetary union, where the central bank is detached from nation states and local politicians, the situation is riskier for its credibility, as can be seen in the case of Greece.
Secondly, in economies with a developed banking sector, there is a strong link between banks and fiscal policy. As seen in Greece for some time, the LoLR can act as a substitute for the access of governments to the private debt market. This is not a good situation for central banks. Necessary, timely and economically viable solutions by governments can be postponed due to the actions of the central bank, which in turn builds a large and risky credit exposure. Greek risk sitting on the ECB’s books illustrates this point.
The ‘clean’ solution is that when a government is involved in LoLR credit, it takes ex ante responsibility for any arising losses. In practice, it is hard to foresee that a government will vote down a proposal to extend LoLR facilities to a bank that is considered by the central bank to be solvent. The risk of this encouraging government to take over supervision from central banks is not, in my view, a real risk, but it is debatable whether supervision within a central bank is good or bad for its operational activities and credibility.
After difficult negotiations in recent weeks, there is a high chance of agreement between Greece and the eurozone on a new debt programme (and we can ask questions, to which extend freezing of new ELA loans played role in turnaround of Greek government and acceptance of conditions, that we refused before). So the issue of the ECB’s exposure to the country could be eventually to diminish, as the New program can strengthen the capital of the banks too. Nevertheless, the issues raised are general in nature and central banks should consider whether the current policy of keeping the LoLR function locked within the walls of the central bank is the best strategy.
Ludek Niedermayer is the former deputy governor of the Czech National Bank