At the end of 2014, the ECB’s holdings of Greek government bonds amounted to approximately €20 billion (Table 1). These bonds had an average remaining maturity of 3.5 years. On 20 July 2015, €3.5 billion of Greek government bonds are coming to maturity. The Greek government is supposed to repay these bonds to the ECB. As more of these bonds will come to maturity during the next three years, the Greek government will have to make repayments to the ECB until the Greek bonds disappear from the ECB’s balance sheet.
Does it make sense to insist that Greece repays the bonds held by the ECB now?
Table 1: Eurosystem SMP holdings (31 Dec 2014)
|Issuer country||Nominal (€ billions|
In order to answer this question, we have to go to the basics of central banking so as to understand what the implications are of government bond purchases and sales by the central bank. A good starting point is Pâris and Wyplosz (2014) and De Grauwe and Ji (2015).
Let us first consider the case of a standalone country. When a central bank buys government bonds, it substitutes one type of liability of the public sector with another one. Government bonds that promise a fixed interest rate are replaced by a monetary liability without interest but carrying an inflation risk. At the moment of the purchase, the government bonds cease to exist from an economic point of view. All that is left of the bonds is a monetary liability of the central bank (which is one branch of the public sector).
- It is important to stress that there are fiscal implications of this bond buying by the central bank.
These arise at the moment of the purchase of the bonds by the central bank.
Prior to the purchase, the government paid interest to private investors. After the purchase, the government pays interest to the central bank. But the central bank is part of the public sector and returns the interest to the government.
- As long as the bonds are on the balance sheet of the central bank, a circular flow of interest payments is organised from the government to the central bank and back to the government.
This effectively means the government is relieved from paying interest on the bonds held by the central bank. The interest relief, therefore, allows the government to reduce taxation or increase spending.
Another way to put this is that when the government bonds are on the balance sheet of the central bank, there is effectively debt relief, i.e. interest payments are waived. This is in fact the seigniorage that the central bank creates by issuing money. This is distributed every year to the government as long as the central bank holds these bonds on its balance sheet.
The price the taxpayers pay is that there might be more inflation. But if ‘monetary dominance’ prevails – i.e. if the central bank targets an inflation rate at, say, 2% and creates each year an amount of money base that is consistent with this target – there is no risk of inflation.
It is important to understand that the value of the bonds purchased under monetary dominance and kept on the central bank’s balance sheet has no bearing for the taxpayer. These bonds could be given a value of zero, or any other value, without any taxpayer suffering or gaining from this.
Suppose for example that the central bank were to write off these bonds (give it a value of 0). This would mean that the circular flow of interest payments would stop. But that would lead to the same economic outcome for the central bank and for the government. The central bank would not record a profit (interest inflow) and would not record a payment to the government (interest outflow).
All that counts is that the central bank that promises an inflation rate of 2% per year keeps its liabilities (money base) under control. What the central bank has on the asset side of its balance sheet is of no economic significance. In the limit, these assets could be written down to zero. Following standard accounting convention, this would then show up as negative equity. But for a central bank equity is of no importance. A modern central bank can easily live with negative equity as it cannot default, i.e. its liabilities (money base) are not a claim on its assets. For example, a holder of a bank note has no claim on any of the central bank’s assets. A banknote is a claim on goods and services in the market place, no more no less.
Bond buying in a monetary union
Does this analysis also hold in a monetary union like the Eurozone? The answer is yes, provided we structure the bond-buying programme carefully.
Suppose the ECB buys an amount of government bonds of Greece. This leads to interest payments from the Greek government to the ECB, which then distributes the profit it makes to the member countries according to the equity share of each country in the ECB (Germany 27%, France 20%, etc.).
Such a bond purchase therefore implies fiscal transfers from Greece to the other member countries. However, there has been an agreement within the ECB to return the Greek interest payments to the Greek government. Thus we reach the situation obtained in a standalone country discussed in the previous paragraphs. That is, the Greek bonds held by the ECB lead to a circular flow from Greece to the ECB and back to Greece. The ECB could easily stop this circular flow by writing off the bonds. This would have no effect on the other member countries.
Note that this is also the rule under the QE programme. The government bonds (80%) bought by the ECB are put on the national banks’ balance sheets. As a result, a circular movement of interest payments is organised from the government to the ECB and back to the same government.
Should the Greek government repay bonds held by the ECB?
Let us now use these insights to evaluate the question of whether the ECB should insist that the Greek government repays the Greek bonds held by the ECB.
The Greek government bonds were bought by the ECB in the context of the SMP programme, which intended to stabilise the financial system in 2011. It is often forgotten, but the Treaty explicitly states that the ECB shall contribute to financial stability (Article 3 of Protocol to the Treaty).
What are the implications of a Greek repayment? When Greece repays the bonds held by the ECB two things happen.
- First, Greece loses the seigniorage gain (in the form of the waiver on interest payments) it enjoyed while the bonds were on the ECB’s balance sheet.
Greece will have to start paying interest again. Thus, its effective debt burden increases.
- Second, at the moment Greece repays the bonds the money base in the Eurozone declines.
Put differently, the repayment is equivalent to a reverse QE.
One could call this ‘quantitative shrinking’ (QS). The ECB, however, is pursuing QE. In the framework of this QE programme, it has set a target path for the money base.
- The ECB will have to compensate the decline in the money base produced by the Greek repayment by additional purchases of government bonds of other member countries.
As a result, the governments of the other member states will enjoy extra seigniorage gains (in the form of a waiver of interest payments on the bonds purchased by the ECB).
In other words, the other countries will enjoy a debt relief that is the counterpart of the increase in the debt burden of Greece
This is where the absurdity of the repayment by Greece comes in. Such a repayment implies that the seigniorage enjoyed by Greece is taken away and redistributed to the other member countries. Or, put differently, the Greek effective debt burden is increased, while the debt burden of the other member countries is reduced. One can ask the question whether this is the right thing to do when so many today agree that the Greek debt burden has become unsustainable.
What should the ECB do?
It would be easy for the ECB to solve this problem. The easiest way would be to write off the Geek bonds. As we have argued earlier such a write-off has no fiscal implications. It means that the circular flow of interest from Greece to the ECB and back is discontinued. In contrast, the repayment of the bonds by Greece has fiscal implications and leads to transfers from Greece to the other member countries. This is quite a perverse transfer given the sorry state of the Greek public finances.
The easy solution is hard
There is a lot of resistance against the easy solution.
- First there is resistance in the ECB.
Like many central banks, it is concerned to show profits and a positive equity. These concerns, however, are misplaced and these profits and losses only have a life in the mind of accountants. Similarly, the equity position of the ECB has no real consequences and should not influence the ECB’s decision. The latter should be guided by concerns about price stability and financial stability.
- Second, there is also resistance against a write-off outside the ECB because such a write-off is associated with a loss that the taxpayers, especially the hard-working German taxpayer, will have to bear.
As we have shown, this loss is purely of an accounting nature. It does not affect the other member countries. It does not imply that German taxpayers will bear a higher tax bill.
Fears, when held strongly, become a reality. One has to take these into account. Therefore it is worth considering another solution. This would consist of the ECB extending the maturity of the Greek bonds. This would lead to exactly the same economic effects (although the accounting would be different). The Greek government would continue to pay interest to the ECB, which would then dutifully return these to the Greek government. Other countries would not be affected.
Both these solutions would prevent the perverse redistribution of seigniorage from Greece to the other member states.
Buiter, W. (2008), “Can Central Banks Go Broke?”, CEPR Policy Insight 24, 16 May.
Corsetti, G. and L. Dedola (2013), “Is the euro a foreign currency to member states?”, VoxEU.org, 5 June.
De Grauwe, P. and Y. Ji (2015), “Quantitative Easing in the Eurozone: It is possible without fiscal transfers”, VoxEU, 15 January.
Pâris, P. C. and Wyplosz (2014), “The PADRE Plan: Politcally Acceptable Debt Restructuring in the Eurozone”, VoxEU, 28 January.
 See Buiter (2008) and Corsetti and Dedola (2013).