Can Greece be forced out of the euro? The role of the ECB in restricting funding avenues to Greece - will Target2 be next?

Ioannis Glinavos (@iGlinavos), Senior Lecturer, University of Westminster https://iglinavos.wordpress.com/

The European Central Bank (ECB) has come under harsh criticism for its support (or lack thereof) of Greece since the election of Syriza. The following comment charts the progressive tightening of funding conditions for Greece against the background of the ECB rules, and reflects on options in case an agreement is not reached to address Greece’s immediate funding needs in June 2015.

ECB stops accepting Greek government bonds as collateral

Since Greece accepted the first bailout in 2010, it has largely not been able to raise money in the markets (apart from domestic T-Bill issues). Continued support from the Troika (disbursement of bailout funds) is dependent on the successful completion of periodic reviews. As the last review was never completed successfully, Greece has not received a bailout fund payment since the summer of 2014.

The avenue through which the Greek government continued to finance its deficits in the absence of bailout disbursements was by borrowing more from its commercial banks. Indeed, although the Greek government is unable to raise long-term funding on the bond markets, it increased its borrowing by means of short-term treasury bills. The Greek government was able to borrow from its banks because those banks can borrow from the Bank of Greece (BoG) and, in turn, the BoG can borrow from the ECB so long as Greece remains in the euro. The banks themselves would be in no position to object to taking on more government debt, for political reasons and especially because sovereign default would mean that their existing holdings of government debt are written down leading ultimately to larger state ownership (a catch 22 situation for the domestic banking system). The ending of EU/IMF lending to Greece has not therefore been a binding constraint on its government budget or its foreign borrowing. It would be as if Greece had obtained ‘bailout’ lending from the loan facility or EFSF after all, causing a faster rise in Eurosystem debt, instead.

Before February 2015 and while this final assessment was being argued over, Greece did continue to finance itself via the ECB by selling bonds to its commercial banks, which then deposited those bonds as collateral with the national central banks (NCBs) in order to gain the funds (through the ECB) needed to pay for the bonds. Correspondent account balances (NCB-ECB) only pay the ECB discount rate as interest, so this is a cheap form of financing. In practice the ECB had tried to persuade NCBs to stop abuse of these accounts. The ECB had pressured Greece, Ireland, and Portugal at the beginning of the crisis to seek bilateral rescue loans and EFSF/ESM funds rather than use their banks and ECB credits to finance their deficits and rollovers. For this to work of course, state paper needs to be accepted as collateral by the ECB. Prior to the 2008 crisis, only A-rated paper was acceptable collateral. This was reduced to BBB- in October 2008 to allow for the large expansion of ESCB credit. As Greece was being threatened with a credit rating below investment grade, the ECB dropped this minimum rating requirement for Greek government in May 2010.

This ‘allowance’ for Greece ended on 4 February 2015 when the ECB’s Governing Council lifted the waiver of minimum credit rating requirements for marketable instruments issued or guaranteed by the Hellenic Republic. This suspension was in line with existing Eurosystem rules, since it were not possible to assume a successful conclusion of the programme review.

ECB rations ELA

The loss of direct access to the ECB credit line meant that the BoG had to extend its use of Emergency Liquidity Assistance (ELA) which is not subject to ECB collateral rules. Although ELA is supposed to be for short periods, there is the precedent of the Irish central bank that used it extensively. The ECB Council could order the BoG to cease ELA, but this seems unlikely given the Irish precedent. The way ELA works (the rules determining its use are extremely limited, a mere 2 page document) is by the NCB requesting it, and the ECB supplying it, unless a 2/3 majority of the governing council objects. ELA can only be provided to ‘solvent’ financial institutions and cannot be used to directly finance a state, as this would violate the No-Bailout clause in the Treaty. Further, the ECB has made it clear that the so-called Securities Market Programme portfolio of Greek bonds bought by the ECB cannot be restructured because that would be equivalent to granting an overdraft to the country and that would be contrary to Article 123 of the Treaty on the Functioning of the European Union. The ECB has continued nonetheless to support the Greek banking system via allowing incremental increases to the ELA, plugging the hole that is opening as deposits fly out in the slow motion bank run that has been in progress since elections were called at the end of 2014.

Supporting the Greek banks, and supporting Syriza through them are two different things however and the ECB has been trying to ban Greek commercial banks from buying any more government T-bills. It was reported in March 2015 that the ECB instructed Greece’s biggest banks to refrain from adding (short term) Greek government exposure. More specifically, the ECB included their recent warnings on capping Greek T-bill holdings at Greek banks in its legal framework. In March 2015, Greek banks held around €11B of T-bills, while the Greek government has a Troika-induced limit of €15B T-bill issuance (total amount outstanding). The new legal framework by the ECB would thus imply that Greek banks can’t cover this possible €4B shortfall if foreign investors don’t re-invest their maturing T-bills. The ECB already had an official cap on the amount of T-bills Greek banks can use for funding through ELA (€3.5B as of March).

This came on top of some more subtle changes, restricting the ability of Greek banks to suck liquidity out of the Eurosystem. In March the ECB also changed the rules for state-guaranteed bonds. This is another kettle of fish than the sovereign bonds (discussed above) that the ECB no longer accepts as collateral for Greece. While the ECB had prevented commercial banks from depositing sovereign bonds as collateral to borrow direct from the ECB, it continued to directly accept commercial bank bonds guaranteed by the Greek state. This is no more. The Governing Council of the ECB adopted Decision ECB/2013/6, which prevents, as of 1 March 2015, the use as collateral in Eurosystem monetary policy operations of uncovered government-guaranteed bank bonds that have been issued by the counterparty itself or an entity closely linked to that counterparty. This Decision, which aimed to ensure the equal treatment of counterparties in Eurosystem monetary policy operations (supposedly!) and simplify the relevant legal provisions, following the measures implemented on July 2012, which limited counterparties’ use of uncovered government-guaranteed bank bonds that they themselves have issued.

This little known practice (now unavailable for Greek banks) worked as follows. A commercial bank would lend money to itself by issuing a bond which it did not intend to sell. Such phantom bond was issued in order to hand it over to the European Central Bank as collateral in exchange for a cash loan. Normally, of course, the ECB would never accept such a phantom bond as collateral, as it would amount to a total circular reason for financing. It would be an assault on the meaning of collateral and a gross violation of the ECB’s rulebook. This is why the bank would take its phantom bond first to the Greek government and had it guarantee it. With the government’s guarantee stamped on it, the ECB then accepted the bank’s phantom bond and handed over the cash as the Greek taxpayer had, in the meantime, unknowingly provided the collateral for the bank’s loan.

Some European governments (Greece included) had launched schemes guaranteeing bonds issued by credit institutions shortly after the outbreak of the financial crisis in order to support their banking systems. Nevertheless, this market development suggests that the introduction of the eligibility of own-use government-guaranteed bonds accompanying the suspension of the minimum credit rating has also allowed a substantial fraction of these increasingly issued bonds to find their way into reverse transactions for refinancing credits with the ECB. Government guarantees are of importance because of two reasons. Firstly, government guarantees for risky assets pose a risk for taxpayers in case of bank default. Secondly, government guarantees can influence the valuation of the collateral as well as its credit rating, and thereby its refinancing conditions. In February 2009, the ECB extended the acceptance of own-use assets to all those guaranteed by governments. In principle, this made it possible to securitize assets into bonds, which are retained, thus never assessed by the market or a rating agency, and can still be used as collateral for refinancing credits due to the government guarantee. Moreover, the conditions in terms of valuation haircuts would be appealing if the rating of the guaranteeing government is higher than that of the issuer. As explained above, this facility is no longer available.

Could conflict with ECB end in expulsion from Target2?

The current conflict scenario may lead to Greece missing the bundled IMF payment at the end of June. If this is treated as a default event (this is doubtful, but possible) it may further impair the position of Greek banks. Even if the ECB does not label the Greek banking system insolvent (thus not eligible for ELA support), it will most certainly increase the haircut on GGBs, making it even more difficult for Greek banks to pledge collateral to benefit from ELA. A further deterioration in relations which leads to comprehensive default on sovereign debt (and/or Grexit) will put the ECB in a position where it will need to stop supporting the Greek banking system, and by extension the defaulting Greek government. It is difficult though to see the BoG cooperating with the ECB in bringing about the destruction of the Greek banking system.

If the ECB did prohibit ELA, depriving the BoG of any approved means of lending to its banks, the BoG would have no option (if the Government does not wish to issue its own currency) other than to defy the ECB and continue to lend anyway, given the consequence of not doing do: the closure of its banks for the want of liquidity. What could the ECB do to prevent this? The only way for the ECB to stop this indirect Eurosystem lending to the Greek government would be by ordering other NCBs to refuse further credit to the BoG, shutting the BoG out of the Target2 system. This scenario is reminiscent of the breakup of the post-USSR ruble zone. Such action however would prevent clearance of cross-border payments out of Greece and amount to the expulsion of Greece from the euro. The free flow of credit between Eurozone NCBs is an essential feature of monetary union. It is what keeps a euro in a Greek bank equal to a euro in banks elsewhere. As long as Greece remains in the euro, it cannot be excluded from Eurosystem credit, so Germany and any other euro countries that still have sound finances will keep lending, whether or not the Greek government defaults. If this is not done via an official loan facility, it will go through the Eurosystem (ECB), and it will increase (as it clearly has) if uncertainty about Greece remaining in the euro accelerates the flight of capital. The ECB cannot avoid continued lending to Greece or any other troubled country that remains in the euro. The ECB (or, more accurately, its owners, the NCBs that constitute the Eurosystem) is the lender of last resort whether it likes it or not. This creates a paradox. The ECB cannot legally expel Greece from the Eurozone, yet by shutting it out of Target2 it will de-facto create a Greek euro that will float against the old-euro creating valuation differentials. In any event, the legality of expelling Greece from Target2 would surely be challenged by Greece in the CJEU.

Some questions for Mr Draghi

The Greek government has complained that the ECB has placed a noose around Greece’s neck. It would be more accurate to say that the noose is around the government’s neck, but there are some serious questions now facing the ECB as the crisis evolves. I would like to ask Mr Draghi the following:

·         How will the ECB treat a default on IMF loans?

·         Will the ECB allow ELA to continue if Greece is rated as in default by the agencies?

·     Will ELA support be dependent on the introduction of capital controls in case of sovereign default?

·         How will the ECB react to ‘non-aligned’ actions by the BoG in case of default?

Further reading:

Ruparel, Even If Deal Is Reached With Greece, The Drama Is Just Beginning

Garber, The Mechanics of Intra Euro Capital Flight,  

Buiter, The implications of intra-euro area imbalances in credit flows

Varoufakis, How the Greek Banks Secured an Additional, Hidden €41 billion Bailout from European taxpayers/

Whittaker, Eurosystem debts, Greece, and the role of banknotes

Barnard & Peers: chapter 19

Art credit: www.rollingalpha.com

Read More

Clash of the Judicial Titans: Will the Euro survive?

Steve Peers

The German Federal Constitutional Court (BVerfG) has finally sent a reference to the Court of Justice of the European Union (CJEU). Moreover, it has chosen a crucially important issue to ask questions about. This issue is the validity of the Outright Monetary Transactions (OMT) policy of the European Central Bank (ECB) – the very policy which is credited with keeping the EU’s single currency alive, perhaps single-handedly. If either of these courts rules that this policy is invalid, the very existence of the EU’s single currency could be called into question.

 In a nutshell, the OMT constitutes a promise made by the ECB, back in the summer of 2012, that it would if necessary purchase the government bonds of troubled eurozone Member States on the secondary market (ie, from banks and other financial institutions which own those bonds). The purpose of the policy was to shore up confidence as regards the specific economies in question, and therefore in the single currency in general. In return, among other things, the countries concerned would have to sign up to austerity programmes. While the OMT has never actually been triggered, the mere possibility of its use appears to have calmed financial markets’ doubts about the survival of the single currency (and about the continued use of the euro by all the EU Member States which currently use it) considerably. The BVerfG’s decision to send any question to the CJEU is itself historic. It comes soon after the first references to the Court of Justice from the Spanish Constitutional Court (Melloni) and the French Constitutional Court (Case C-168/13 F, which concerned Jeremy Forrest, the British schoolteacher who ran off with one of his pupils). So all the big beasts among Europe’s constitutional courts have now engaged with the CJEU.

 However, it is clear that the BVerfG at least still does not really regard the CJEU as the king of the jungle. It asserts its intention to find that the OMT programme breaches the German constitution, depending on what the CJEU has to say. On the other hand, the CJEU has always asserted that it is the sole judge of whether an EU act is valid (going back to the judgment in Foto-Frost). Can the single currency survive this conflict between constitutional principles?

The legal issues 

The BVerfG (along with many others, particularly in Germany) doubts that the OMT programme is legal, because it constitutes economic policy (rather than monetary policy) and because it amounts to the ECB buying up eurozone governments’ debt, which is not permitted under the Treaties which the EU is founded up.

Let’s have a closer look at these arguments. First of all, is the OMT programme an economic policy? While the EU has established an economic and monetary union (EMU) among eurozone Member States, and economic and monetary policies are obviously closely related, there is nonetheless a sharp legal difference between the two policies as a matter of EU law. Monetary policy is an exclusive competence of the EU, as far as the eurozone Member States are concerned. Within the EU, the ECB is in charge of that policy.

On the other hand, economic policy is primarily a matter for Member States; the Union (primarily the Council) only coordinates such policies.While this division may not make much economic sense, any significant shift of powers over economic policy to the EU would have been impossible to agree politically, and have raised great(er) doubts about the EU’s legitimacy.

In its Pringle judgment of 2012, the CJEU ruled that the treaty establishing the European Stability Mechanism (ESM), which establishes a system for eurozone Member States to lend financial support to each other, was an act of economic policy, not monetary policy. But the CJEU did not define what ‘monetary policy’ consisted of.

The BVerfG doubts that the OMT programme is valid because it constitutes an independent economic policy of the ECB. But as the BVerfG itself notes, the Treaties provide (in Art. 127(1) TFEU) that the European System of Central Banks (which the ECB forms a key part of) ‘shall support the general economic policies in the Union with a view to contributing to the achievement of the objectives of the Union’ set out in Art. 3 TEU. Arguably the OMT programme is exactly that: support for the economic policies of Member States (including the ESM) as regards the EU’s objective of establishing and maintaining a single currency. As long as the economic conditionality linked to the OMT programme is no different from, or at least consistent with, the economic conditionality linked to the ESM and the EU rules on economic governance (ie, control of excessive deficits), then the OMT programme cannot be said to constitute a separate economic policy of the ECB.

Secondly, does the OMT programme circumvent the ban on buying government debt? Article 123 TFEU specifies that the ECB and national central banks cannot ‘purchase directly’ the ‘debt instruments’ of eurozone Member States’ governments. Obviously, the word ‘directly’ is significant; if the authors of the Treaties had wanted to ban the ECB from ever owning a government bond issued by a eurozone Member State, they would have left that word out. So it can hardly be doubted that the ECB can purchase such bonds from financial institutions, at least on a modest and non-systematic basis. 

But the BVerfG is concerned about the underlying purpose of the ban on direct purchases of government bonds: to prevent governments from being ‘propped up’ by the central bank. If all government bonds issued by Eurozone Member States were systematically snapped up by the ECB from the financial institutions which had initially bought them, the ECB would arguably be circumventing the ban on direct purchases.

The best approach to this objection is to interpret the ECB’s powers in light of its obligation to contribute to achieving the EU’s objectives, in particular the development of the single currency. In the ordinary course of events, purchasing significant numbers of eurozone government bonds directly on the secondary market might not have a strong link to the existence of the single currency. But in the current circumstances, it does. So this justifies a flexible approach to the limits which might otherwise apply to the ECB’s actions – provided that such purchases are on secondary markets, and are necessary to ensure the single currency’s survival.

The judicial politics 

The BVerfG states clearly what it expects the CJEU to do: to interpret the OMT programme in accordance with its specified constraints, otherwise it will rule that the programme is in breach of the German constitution. But the BVerfG has warned the CJEU before, and then not gone through with its threats. It has come to resemble an angry parent issuing increasingly dire threats to a naughty child – and then not following through on them. The naughty child soon realises that the threats won’t be carried out, and adapts her (mis)behaviour accordingly. 

And there is another factor at play here. When the BVerfG was asked if Germany could ratify (for instance) the Lisbon Treaty, there was a clear route to ensure that its judgment was carried out. Very simply, it could have ordered the German government not to ratify that Treaty. Similarly, in cases concerning the EU’s banana market legislation, or the interpretation of EU age discrimination rules, it could have ordered the German administration and courts not to apply the rules concerned. But it is less clear exactly what it can do to stop the actions of the ECB, which presumably would not consider itself bound by a BVerfG decision. The BVerfG refers to requiring the German government to act (unless there is a retroactive amendment of the German constitution), but exactly how that will stop the application of the OMT programme is unclear. So our naughty child in Luxembourg – and her naughty brother in Frankfurt – must know that their ostensibly strict parent probably can’t carry out this particular threat.

In that case, how will they behave? It might be expected that many of the concerns raised by the BVerfG can be addressed, but not all. In particular, it is hard to see how the OMT programme could achieve its objectives if the ECB does not have the power to buy significant number of government bonds on secondary markets. Possibly the CJEU will allow such purchases to continue, subject to certain conditions – which the BVerfG may decide are strict enough to meet its concerns. Judging from the overall tone of the latter court’s ruling, it is looking to find a way to uphold the validity of the OMT programme despite its fundamental objections.


Barnard & Peers: chapter 19

Read More