Or, losers in a Greek debt restructuring.
Some estimates courtesy of JPMorgan’s flows and liquidity team:
The ECB bought a large amount of Greek government bonds through its Securities Market Program. Our colleagues in Euro rates research estimate that the ECB bought around €40bn of Greek government bonds with €50bn of notional value, assuming an average purchase price of 80% to par. But the ECB has an even bigger exposure to Greece through its lending to Greek banks.
But JPM’s point about the difficulty in a Greek debt restructuring is a salient one given that the private financial sector has become so muddled with public liabilities. The Greek government has assumed a whole bunch of banking risk, which in turn has been pushed onto the ECB. Unfortunately, the Greek sovereign-bank model is one that’s also been infamously used in Ireland. Meanwhile Portugal has just announced it will up its own sovereign-bank loop by providing €35bn worth of government guarantees for bank bonds, to most likely be used as collateral at the ECB. The sovereign-bank loop is intensifying.
So we imagine that some of that top-secret European debate is not so much about the viability of a Greek debt restructuring per se — Greek debt sustainability is clearly on a downward spiral in a deflationary environment — but the possibility of triggering more restructurings, as suggested by JPM.
The good news is that there are those points at which a Greek debt restructuring becomes less onerous– such as after 2013, when the European Stability Mechanism (ESM) specifically set up to enable debt restructurings comes into effect. Before then, it looks very voluntary, or very painful.
Some estimates courtesy of JPMorgan’s flows and liquidity team:
The ECB bought a large amount of Greek government bonds through its Securities Market Program. Our colleagues in Euro rates research estimate that the ECB bought around €40bn of Greek government bonds with €50bn of notional value, assuming an average purchase price of 80% to par. But the ECB has an even bigger exposure to Greece through its lending to Greek banks.
Greek banks had borrowed €91bn from the ECB as of the end of February with collateral of €144bn. What does this collateral consist of? $48bn is Greek government bonds held by Greek banks on their balance sheet. €55bn consists of government-guaranteed bonds issued by Greek banks, €25bn of which was only issued at the end of last year for the Greek banks to meet new more punitive collateral requirements by the ECB. €8bn is zero-coupon bonds which the Greek government had lent to Greek banks in 2008. The remaining €33bn is likely to be Greek ABS/covered bond collateral. The Greek government agreed earlier this year to extend state-guarantees to Greek banks by another €30bn, but it appears that this new aid package has not been used by Greek banks. All this analysis suggests that 77% of the collateral that Greek banks posted with the ECB is government or government-guaranteed, which would be directly affected in the hypothetical scenario of a Greek debt restructuring. In addition, the remaining 23% of ABS/covered bank bond collateral would almost certainly be affected in the case of a Greek debt restructuring as the solvency of Greek banks would become an issue.
In total, the notional ECB exposure to Greece amounts to around €50bn + €144bn = €194bn. Against this notional exposure, the ECB has lent/invested €40bn + €91bn = €131bn or 68% of its notional exposure. These calculations imply that in a hypothetical case of a Greek debt restructuring, the ECB is protected for a haircut of up to 32%. Beyond that cushion, the ECB is exposed to losses. A hypothetical haircut of 50% would create losses of around €35bn for the ECB.
The Eurosystem has experienced losses on refinancing operations in the past during the Lehman crisis as 5 banks defaulted on their repo operations. The losses incurred by the Eurosystem are to be shared by all national central banks in proportion to their shares in the ECB’s capital. The Eurosystem has €81bn of capital and reserves currently, enough to withstand even a 50% Greek debt haircut. But it would be a lot more problematic for the ECB if other countries such as Ireland had to restructure. The exposure of the ECB to Ireland is similarly big but likely with a smaller cushion. The total exposure of the ECB to Ireland consists of around €20bn of bond purchases and €83bn of repos with domestic Irish banks. This excludes around €67bn of ELA lending which represents an exposure for the national central bank rather than the Eurosystem as a whole. But if domestic Irish banks had to replace their ELA borrowing with ECB borrowing over the coming months, the total exposure of the ECB to Ireland would rise to €170bn, well above of that of Greece.
Greek banks own €49bn of Greek bonds. Their equity amounts to €29bn. The market value of their equity is €12bn, suggesting that the market is already pricing in a loss of €17bn or 35%. A hypothetical haircut of 50% on Greek government debt would create losses of around €25bn, leaving only €4bn of equity (or 1% of assets) for the Greek banking system. But the losses for Greek banks would be much smaller if a Greek debt restructuring were to take place in mid 2013. The average maturity of their Greek government bond holdings is 5 years and roughly €10bn matures every year. By mid 2013, their Greek government bond holdings will drop to €25bn, i.e. half of their current holdings.
The central Bank of Greece held directly €7bn of Greek government bonds as of the end of February. A hypothetical haircut of 50% on these bond holdings would wipe out its entire capital and reserves of €3bn.
Greek social security and other public entities hold around €30bn notional of Greek government bonds. They have already applied a loss of 30% in these holdings. A hypothetical haircut of 50% would create additional €6bn of losses vs. current financial assets of €31bn.
European banks hold €50bn of Greek government bonds according to Q3 2010 BIS data. Even a 50% hypothetical haircut would be manageable. But it becomes more problematic when ones looks at the total exposure of European banks to Greece, including private sector loans, repos, guarantees and credit commitments. These private sector claims are also likely to suffer in the case of a Greek debt restructuring. According to BIS, European banks’ total exposure to Greece was €165bn at the end of Q3 2010, driven by French banks (€68bn) and German banks (€50bn). The potential losses for European banks would be more threatening if other countries such as Ireland were to restructure. According to BIS, European banks’ total exposure to Ireland (both public and private sector exposure) was €450bn at the end of Q3 2010, driven by British banks (€165bn), German banks (€150bn) and French banks (€57bn).Now, there’s nothing surprising about a bank resisting the idea of bond haircuts.
But JPM’s point about the difficulty in a Greek debt restructuring is a salient one given that the private financial sector has become so muddled with public liabilities. The Greek government has assumed a whole bunch of banking risk, which in turn has been pushed onto the ECB. Unfortunately, the Greek sovereign-bank model is one that’s also been infamously used in Ireland. Meanwhile Portugal has just announced it will up its own sovereign-bank loop by providing €35bn worth of government guarantees for bank bonds, to most likely be used as collateral at the ECB. The sovereign-bank loop is intensifying.
So we imagine that some of that top-secret European debate is not so much about the viability of a Greek debt restructuring per se — Greek debt sustainability is clearly on a downward spiral in a deflationary environment — but the possibility of triggering more restructurings, as suggested by JPM.
The good news is that there are those points at which a Greek debt restructuring becomes less onerous– such as after 2013, when the European Stability Mechanism (ESM) specifically set up to enable debt restructurings comes into effect. Before then, it looks very voluntary, or very painful.
Argentina: the default scenario for Greece
Posted by Joseph Cotterill on May 12 10:30.
We did warn that financial markets would have to come to vivid terms with the comparison one day…
This is not a flattering chart for the eurozone:
It’s a two-thirds haircut on Greek debt, in 2012, before the supposed 2013 threshold for a ‘real’ restructuring, and less friendlier to Greece holders too. Importantly, the haircut comes in terms of net present value. A NPV cut could be done via maturity and coupon alterations, not necessarily face value.
It’s a familiar taboo if you’ve followed the travails of Greek banks’ exposure to government bonds within their banking books and the capital risks involved. Nevertheless, a 67 per cent haircut explodes all notions of ‘soft’ versus ‘hard’ restructuring and it’s really worth exploring what this number means.
Above all, there’s the shadow of Argentina’s torturous 2005 debt exchange after its 2001 default.
That is a eurozone reality check.
___________________________
Fiscal adjustment <–> haircuts
Why 2012?, you must be asking.
The chart comes via Barclays Capital analysts. BarCap now assume that eurozone officials will wait until Greece is at least on the way to sustainable primary balances after a sovereign restructuring, which is in the first half of 2012. But, with the balance liable to come in at -2.5 per cent this year — it’s implicit here that a 67 per cent haircut must be made on Greek debt in 2012 to push Greece towards sustainability.
Here’s another chart:
Interesting to note that BarCap assume restructuring of Greece’s official debt would be heavy as part of the haircut, with maturities pushed out by 10 years, but far from enough:
Recovery rates, again
Where it gets really interesting however is how BarCap have modelled the recovery rates on post-restructuring Greek debt — again, a subject close to FT Alphaville’s heart.
Modelling all of Greece’s debt as one bond that pays out the average interest rate for Greece (5.15 per cent), and which possesses Greek debt’s average maturity (eight years) BarCap apply the 67 per cent haircut to derive this chart:
OK — time to recycle this recent Moody’s chart of historical rates of recovery in defaulted sovereign bonds:
Note Russia’s and Argentina’s recoveries especially.
FT Alphaville had already been expecting low recovery rates owing to the sheer amount of debt stock that Greece is burdened with. However…
Twenty-five cents in the euro would be a landmark in the history of sovereign debt restructuring. Not only that, it’s really still not being priced either in Greek debt or CDS. In the latter market, recoveries of 40-45 have been more common.
___________________________
Greek banks as the caveat to a 2012 haircut
But what about Greek banks’ capital?, you might still be asking.
Because after all, at €29bn, it’s at real risk of being wiped out wherever the haircut is more than 50 per cent — this is partly why there’s been interest in forestalling this cut to 2013, when their exposures would have likely halved, rather than doing it in 2012 and therefore requiring heavy recapitalisation. There’s no easy answer here although under a NPV haircut, we’ve already pointed out that accounting and regulatory capital treatments could limit losses.
In addition, BarCap (like Roubini before) mention the possibility of presenting creditors with a menu of different bonds tailored to capital treatments in an exchange of debt — for example, zero coupon bonds which protect face values, or discount bonds which work vice versa. Add in exotic accouterments like GDP warrants and we’re getting closer to debt being able to be traded away.
And guess what accompanied Argentina’s 2005 exchange?
___________________________
But — Greek banks may yet prevent a strong haircut before 2013, leaving the eurozone with some combination of maturity swapping and extra bailout loans.
Of course, BarCap’s haircut is a model that aims to identify the optimal debt reduction for Greece given its expected fiscal path throughout the next four years. It’s still very likely indeed that eurozone states and banks will try to nibble the bullet, rather than bite it.
(Equally, the model is necessarily colourless on some tail risk in Greek debt values, such as the country leaving the euro)
The point is though — if a 67 per cent haircut is needed in 2012, what’s going to be required in 2013?
This is not a flattering chart for the eurozone:
It’s a familiar taboo if you’ve followed the travails of Greek banks’ exposure to government bonds within their banking books and the capital risks involved. Nevertheless, a 67 per cent haircut explodes all notions of ‘soft’ versus ‘hard’ restructuring and it’s really worth exploring what this number means.
Above all, there’s the shadow of Argentina’s torturous 2005 debt exchange after its 2001 default.
That is a eurozone reality check.
___________________________
Fiscal adjustment <–> haircuts
Why 2012?, you must be asking.
The chart comes via Barclays Capital analysts. BarCap now assume that eurozone officials will wait until Greece is at least on the way to sustainable primary balances after a sovereign restructuring, which is in the first half of 2012. But, with the balance liable to come in at -2.5 per cent this year — it’s implicit here that a 67 per cent haircut must be made on Greek debt in 2012 to push Greece towards sustainability.
Here’s another chart:
Interesting to note that BarCap assume restructuring of Greece’s official debt would be heavy as part of the haircut, with maturities pushed out by 10 years, but far from enough:
Given the implicit debt reduction obtained by the official sector (c. 9% of GDP), bonded debt needs to contribute an additional 83% of GDP. Given that this represents c.120% of GDP by end-2011, it is equivalent to a haircut of 67%…___________________________
Recovery rates, again
Where it gets really interesting however is how BarCap have modelled the recovery rates on post-restructuring Greek debt — again, a subject close to FT Alphaville’s heart.
Modelling all of Greece’s debt as one bond that pays out the average interest rate for Greece (5.15 per cent), and which possesses Greek debt’s average maturity (eight years) BarCap apply the 67 per cent haircut to derive this chart:
OK — time to recycle this recent Moody’s chart of historical rates of recovery in defaulted sovereign bonds:
Note Russia’s and Argentina’s recoveries especially.
FT Alphaville had already been expecting low recovery rates owing to the sheer amount of debt stock that Greece is burdened with. However…
Twenty-five cents in the euro would be a landmark in the history of sovereign debt restructuring. Not only that, it’s really still not being priced either in Greek debt or CDS. In the latter market, recoveries of 40-45 have been more common.
___________________________
Greek banks as the caveat to a 2012 haircut
But what about Greek banks’ capital?, you might still be asking.
Because after all, at €29bn, it’s at real risk of being wiped out wherever the haircut is more than 50 per cent — this is partly why there’s been interest in forestalling this cut to 2013, when their exposures would have likely halved, rather than doing it in 2012 and therefore requiring heavy recapitalisation. There’s no easy answer here although under a NPV haircut, we’ve already pointed out that accounting and regulatory capital treatments could limit losses.
In addition, BarCap (like Roubini before) mention the possibility of presenting creditors with a menu of different bonds tailored to capital treatments in an exchange of debt — for example, zero coupon bonds which protect face values, or discount bonds which work vice versa. Add in exotic accouterments like GDP warrants and we’re getting closer to debt being able to be traded away.
And guess what accompanied Argentina’s 2005 exchange?
___________________________
But — Greek banks may yet prevent a strong haircut before 2013, leaving the eurozone with some combination of maturity swapping and extra bailout loans.
Of course, BarCap’s haircut is a model that aims to identify the optimal debt reduction for Greece given its expected fiscal path throughout the next four years. It’s still very likely indeed that eurozone states and banks will try to nibble the bullet, rather than bite it.
(Equally, the model is necessarily colourless on some tail risk in Greek debt values, such as the country leaving the euro)
The point is though — if a 67 per cent haircut is needed in 2012, what’s going to be required in 2013?
