# Blog

Sovereign QE and national central banks

Inflation continues to fall. The first outright deflation numbers since 2009 have been recorded and inflation expectations are clearly dis-anchored. Inflation and inflation expectations are now clearly below the 1.7 to 1.9 percent range and inflation expectations are 1.5 percent for the EMU 5y5y forward inflation swap. More worryingly, wage growth in Germany is starting to fall as a result of lower inflation expectations.

It is clear that the European Central Bank will have to act. It is true that the ECB cannot solve all of the euro area’s problems: governments have a clear obligation to move ahead more quickly with structural reforms that address the deep divergences in the euro area and with more public investment to trigger growth. However, slow government progress cannot be an excuse for the ECB not to act. It looks as if the decision has already been taken to buy sovereign bonds. The policy debate is now about how to design a sovereign-bond purchase programme.

An option reportedly seriously under consideration is to leave national central banks to take on the risk of default on sovereign bonds, while the market risk will remain with the Eurosystem as a whole (see for example Reuters). This would be at best ineffective, at worst dangerous. In fact, the attempt to make national central banks shoulder hard default risks is bound to fail in the Eurosystem. The following technical points need to be considered to properly assess this matter.

  1. Signalling: Buying sovereign bonds but leaving national central banks to take on the risk of default would be a strong signal that the ECB is no longer a “joint and several” institution. It would effectively be a declaration that the ECB cannot act and purchase government bonds as a euro-area institution in the interest of, and on behalf of, the entire euro area. This could severely undermine the ECB’s credibility not just in the sovereign purchase programme but also more broadly as an institution.
  2. Pari-passu: ECB executive board member Benoit Coeure argued recently in an interview with France24 (1) that it is illegal according to the treaty to reschedule or restructure the Greek debt that the ECB holds1)Ce portefeuille-là ne peut pas être structuré. Ce serait illégal et contraire aux traités de rééchelonner une créance de la Banque centrale sur un Etat. Je crois que les traités européens sont très clairs là-dessus. Read the full interview here. He argued that the ECB bought this debt for monetary policy purposes and that any restructuring of such a portfolio would be against the treaties. (2) It seems to be clear that the ECB views voluntary participation in a restructuring as monetary financing and therefore in violation with Article 1232)Contraintes…l’une d’entre elles – mais pas la seule – étant qu’un portefeuille de politique monétaire détenu par la BCE ne peut pas être rééchelonné. Ça, ça serait clairement contraire aux traités. Ibid.

A different question is how to consider losses that the ECB would make on a forced restructuring. What would happen if a debtor country unilaterally declares that it will default on its debt, or that alternatively a majority of creditors with collective action clauses declare that they accept a restructuring of the bond? In my interpretation, the ECB could do nothing against such a restructuring and would have to accept its consequences. As this would not be voluntary and it would happen after the monetary policy decision has been taken, it would not amount to monetary financing and would therefore be legal. If this interpretation is wrong, then, de facto, the ECB would never be able to buy pari-passu, including through the OMT programme.

Overall, it is currently debated if the ECB could actually assume the risk of default of sovereign bonds it would buy. I would argue that it has that possibility according to the treaty if the restructuring is forced on the ECB, but the ECB cannot agree to a voluntary exchange leading to a loss. But the legal debate on this is ongoing. The question of whether the ECB can buy at pari-passu is one of the central points that Advocate General Pedro Cruz Villalon of the EU Court of Justice (ECJ) will discuss in his opinion on the German Constitutional Court’s challenge of the OMT, to be published tomorrow.3)The issue of seniority of the ECB is already alluded to in sentence 132 of the ECJ’s Pringle judgement when comparing article 123 with article 125 of the TFEU.

This uncertainty is the main reason why the ECB Governing Council is currently discussing whether default risk should remain with the national central banks. The fear is that the ECB might violate the European treaties by accepting the risk of default of individual member states that could be distributed throughout the Eurosystem. The hope of those proposing a solution that would see the national central banks buying the sovereign bonds of their country is that it would be a way to avoid the redistributive effects of monetary policy across the common currency area. It is seen as a way of avoiding potential cross-border fiscal implications of monetary policy.

  1. So what would happen if the national central banks were to buy the sovereign bonds of their countries only? The following discussion will show that this cannot solve the pari-passu issue in a satisfactory way. On the contrary, the construction would be very complicated and would also lead to a de-facto super-seniority of the national central bank over the private creditors. Ultimately, isolating default risk in one national central bank is impossible as long as there remains a Eurosystem. Here is why:

Take the example of a country with 25 percent of its debt in the hands of its national central bank. What would happen if the national government had to decide to impose a haircut on all of its debt in order to reduce the burden on its taxpayers? There are essentially two options in such a case:

  • The national treasury could decides to exempt the national central bank from participating in the loss. This would mean that the remaining 75 percent of the debt would bear a greater part of the burden. The private creditors would be junior. Knowing about this possibility, they would ask ex-ante for a higher risk premium, which would render the purchase much less effective in the first place.
  • Alternatively, the national treasury could include the national central bank in the haircut. The national central bank would incur a loss, its equity would fall or even become negative. Normally, a central bank would then go to its treasury, pass on the losses and ask for a recapitalisation. This would essentially mean that the treasury would not benefit from defaulting on this part of the debt and again, the other creditors would have to bear a greater part of the burden. They would be junior and again ask for a risk premium ex ante.

So the purely national purchase of national sovereign debt would either leave the private creditors as junior creditors, or the national central bank has to accept negative equity. What would negative equity mean for a central bank? De facto it would mean that the national central bank, that has created euros to buy government debt, would have lost the claim on the government. It would still owe the euros it has created to the rest of the Eurosystem4)A similar discussion took place when national central banks gave emergency liquidity assistance (ELA). ELA is provided at the risk of the national central bank, but if the national central bank has negative equity, it can no longer honour the claim of the Eurosystem, and therefore the loss is passed on to the Eurosystem.. The Eurosystem could now either ask the national central bank to return that liability, which it is unable to do without a recapitalisation of its government. Or, the Eurosystem could decide to leave the claim standing relative to the national central bank. In that case, the loss made on the sovereign debt would de facto have been transferred to the Eurosystem. In other words, the attempt to leave default risk with the national central bank will have failed.

To sum up, either government bond purchases made by national central banks are super-senior or the potential default risk on the government bonds will be passed on to the Eurosystem as a whole. In the former case, the QE bond purchase would be rather ineffective. In the latter case, the only way to avoid losses for the Eurosystem would be to use other national central bank assets, such as gold or potentially future seignorage.

Overall, this discussion shows that monetary policy in the monetary union reaches the limits of feasibility if the principle of joint and several liability at the level of the Eurosystem is given up. Surely it was not the treaty’s intention to create a central bank that is ineffective when the zero lower bound is reached and the only way to increase the money supply is bond purchases. I would therefore argue that the ECB should be able to buy sovereign bonds according to the capital key in order to fulfil its treaty-based monetary policy mandate. It could leave some constructive ambiguity about what would happen in the case of a default, which in any case is an extreme scenario. Ideally, there should be some implicit understanding that ECB would share in losses in case debt is restructured by majority decision of the creditors.

A programme for national central banks to purchase national debt would send a negative signal to markets, and it would also not fully and effectively address the concerns of all those that worry about the potential cross-border fiscal effects of monetary policy. For the ECB to be effective in current circumstances, policymakers will have to accept the potential fiscal implications of monetary policy.

Published by Bruegel