The 2012 Greek debt restructuring: how it happened and what it means Presentation at PEFM Seminar St.
Download ReportTranscript The 2012 Greek debt restructuring: how it happened and what it means Presentation at PEFM Seminar St.
The 2012 Greek debt restructuring: how it happened and what it means Presentation at PEFM Seminar St. Antony’s College, Oxford 11 November 2013 Jeromin Zettelmeyer* EBRD, PIIE and CEPR with Christoph Trebesch (LMU Munich and CESIfo) and Mitu Gulati (Duke University) *Strictly personal views Outline 1. A brief history of the Greek debt restructuring • 2. how did we go from denial to historic bond exchange to buyback of the new bonds? What were the distributional implications? • creditors vs. Greece; but also inter-creditor distribution 3. How was the collective action problem addressed? 4. Was Greece’s December 2012 buyback a “boondoggle”? 5. Assessment and implications • Was the restructuring necessary? • Could it have been handled better? • Could it serve as a template for future restructurings? Twists and turns of the Greek restructuring saga 1. May 2010 package: combination of large-scale financing and massive adjustment; no debt reduction 2. July 2011 package: increases both financing envelope (including through PSI) and required reforms/adjustment. • “Private sector involvement” w.r.t. financing (planned extension of maturities beyond 2020), but no debt relief. 3. October 2011 Eurogroup decision to carry out deep restructuring (50% face value reduction). 4. March-April 2012 debt exchange carrying out that decision (more or less); retires €200bn in bonds in exchange for €30bn quasi-cash (short term EFSF notes) and €56bn new bonds 5. December 2012: buyback of €32bn of new bonds using €11.3 in cash borrowed from the EFSF How did we get there?` 1. Notwithstanding significant initial adjustment structural reforms do not take off 2. The October 2010 “Deauville beachwalk” 3. Greater than expected output decline, unemployment, social and political stress. 4. Dwindling prospect of regaining market access 5. By May 2011, a sense that contagion could be contained 6. Unwillingness of Germany/IMF to provide extra support without some (July) or very deep PSI (October 2011) The end-result: debt reduction, but also a massive transformation of the debt structure T-Bills 15.0 Privately held Bonds: 205.6 Holdouts, 5.5 EU/EFSF 52.9 T-Bills; New 23.9 ECB/ NCBs 56.7 IMF, 20.1 Before (Feb 2012): • about €350 billion FV • Mostly owed to privately creditors Bonds 29.6 2012 restructurings IMF 22.1 ECB/ NCBs; 45.3 EU/ EFSF: 161.1 After (end-2012): • about €287 billion FV • 80 percent owed to official creditors Note: Charts show sum of bonds, T-bills and EU/IMF debt, some smaller sovereign debt items not shown. Outline 1. A brief history of the Greek debt restructuring • 2. how did we go from denial to historic bond exchange to buyback of the new bonds? What were the distributional implications? • creditors vs. Greece; but also inter-creditor distribution 3. How was the collective action problem addressed? 4. Was Greece’s December 2012 buyback a “boondoggle”? 5. Assessment and implications • Was the restructuring necessary? • Could it have been handled better? • Could it serve as a template for future restructurings? Distributional implications: private creditors vs. Greece 1. Present value “Haircut” (percentage loss of creditors): 59-65% • • Lower than headline figure of >75 percent claimed by creditors Difference due to methodology: need to compare PVs of new and old bonds, not PV of new with face value of old. 2. Debt relief to Greece: 51-55 percent of GDP • • This is net of bank recapitalisation costs. Much larger than Argentina (2005) and any other modern private sector debt restructuring. A historically large transfer from private creditors to Greece Computing “haircuts” (PSI-related creditor losses) (in percent of outstanding principal) Discounting approach 1/ Uniform discount rate 15.3 16.3 18.7 “Curve” Value of new securities received (PVnew) 23.1 22.5 21.2 22.8 Haircut in market definition (100-PVnew) 76.9 77.5 78.8 77.2 Value of old bonds (PVold) 65.3 63.3 59.0 56.5 Present value haircut 100*(1-PVnew/PVold) 64.6 64.4 64.0 59.6 1/ Used for discounting payment streams of both new and old Greek government bonds. For EFSF bill present value of 15 is assumed. GDP warrants valued at 0.3 percent of outstanding principal. How tough was haircut in historical comparison? 100 Fourth largest haircut among since 1975 (except for HIPCs) YEM BOL TGO NIC ETH SEN MOZ STPSLEMRT BIH GUY GUY ZMB UGA GIN NER ALB IRQ COG TZA CMRCMR HND ARG BOL HND CRI SRB ECU PERCIV POL NIC 0 50 MEX BGR JOR MDG VNM RUS RUS DOM DOMMOZ CUB GMBCOD POL NIC RUS NER TGO KEN CUB JAM CUB PHL PHL ECU NIC NGA POL CRIMDG MARNGA ECU CODNERCODPOL MWI VEN LBR SEN CRICOD PAN MKD ROU ARG ROU SEN JAM CHLJAM MEX CODARG NGA COD BRA PRY SEN MWI POL COD URY BRA RUS NIC GIN PHL URY POL MAR DZA TUR ZAF ARG MAR URY SRB TUR MDG NGA BRA BRA MEX UKR TUR JAM CHL GAB TTO ECU PHL JAM PAN UKRPAK SRB MDG CHL ZAF ROU PAN POL PAK HRV UKR VEN DZA TUR ZAF CHL GAB SRB PER ECU ECU MEX VEN JAM SVN JAM MEX NGA BRA NGA URY CHL MEX NGA PER SRB UKR BRA POL SDN MDA GRC SYC CIV MDA GRD BLZ URY DOM DOM 1975m1 1980m1 1985m1 1990m1 1995m1 2000m1 2005m1 2010m1 Short maturities hit much harder than long end • Consequence of one-size-fits-all offer (no menu) and low coupon rate on old bonds compared to exit yields Bond-by-bond haircuts using uniform 15.8% discount rate 90 80 70 60 50 40 30 20 10 0 0 2 4 6 8 10 12 14 remaining duration (years) 16 18 20 22 Computing debt relief Old liability Assumed discount rate (percent) 3.5 5.0 8.0 15.3 Value of €199.2 bn in old bonds 217.2 199.5 171.9 130.1 Value of €61.4 new bonds 61.9 49.8 33.6 15.7 Value of €29.7 bn EFSF PSI sweetener 31.4 25.3 17.2 8.2 Value of €25 bn EFSF bank recap loan Present value debt relief net of recap costs € billion per cent of GDP 25.7 21.5 15.3 7.6 99.2 51.2 103.7 53.5 106.4 54.9 98.9 51.1 New liabilities Historically high debt relief Combination of high debt relief in percentage of PV of old debt and large outstanding debt volume as a share of GDP Restructuring Episode Investor Debt relief Volume GDP Debt relief (% GDP) Haircut (percent) (€mn) (€mn) Russia, 2000 52.6 33.2 21759 196302 3.7 Ecuador, 2000 28.6 24.8 4826 12050 9.9 Argentina Phase 1, 2001 40.5 30.8 31537 203097 4.8 Argentina, 2005 75.0 70.9 47090 138468 24.1 Uruguay - External, 2003 13.4 -5.3 2294 17251 -0.7 Uruguay - Domestic 2003 22.3 0.0 1209 17251 0.0 Greece, 2012 64.6 52.0 199210 193750 53.5 Source: Sturzenegger and Zettelmeyer, 2007, except Greece; authors' calculations (Greece); WEO, Eurostat Outline 1. A brief history of the Greek debt restructuring • 2. how did we go from denial to historic bond exchange to buyback of the new bonds? What were the distributional implications? • creditors vs. Greece; but also inter-creditor distribution 3. How was the collective action problem addressed? 4. Was Greece’s December 2012 buyback a “boondoggle”? 5. Assessment • Was the restructuring necessary? • Could it have been handled better? • Could it serve as a template for future restructurings? How the free rider problem was addressed Quite differently from the typical bond exchanges of the 1990s and early 2000s (e.g. Ecuador, Argentina): 1. More reliance on formalised negotiations with a creditor committee consisting mostly of large institutions susceptible to official influence and peer pressure 2. Much less reliance on threats of non-payment • Avoided almost entirely (until the last minute) 3. More reliance on CACs • “retrofitted” onto domestic law debt via act of parliament, 66% majority threshold across all domestic bonds 4. Much more reliance on “sweeteners”: cash, and a seniority upgrade. “Carrots”: make new instruments as safe and liquid as possible 1. 15% of old principal in short term EFSF bonds • Quasi-cash, made up 2/3 of the new value received! • The largest cash sweetener ever offered in a restructuring 2. New bonds issued under English law with usual creditor protections. 3. “Co-financing agreement” governing new bonds: • Enshrines pro-rata repayments to EFSF (w.r.t. PSI related repayments) and new bondholders • EFSF agreement needed for changes in payment, new bond issuance beyond a ceiling of 90 billion Summary: how Greece got to 97% participation 1. Official + peer pressure on large regulated institutions: →perhaps 50-60% of domestic law debt 2. Plus: combination of “carrots” and “sticks” (non-payment threats) →83% of domestic law debt (>> 66% CAC threshold) 3. Plus: domestic aggregate CAC triggered: →100 percent of domestic law debt 4. Foreign law instruments: 61% participation rate before CACs (so must reflect pressure, carrots, sticks), 71% after CACs. Upper axis: per cent of domestic law debt 60 0 Official and peer pressure (P) 0 83 CACs (foreign law) 100 “Carrots” (C) and Domestic P&C&S “Sticks” (S), dom. CAC - foreign Lower axis: per cent of total eligible debt 86 97 Outline 1. A brief history of the Greek debt restructuring • 2. how did we go from denial to historic bond exchange to buyback of the new bonds? What were the distributional implications? • creditors vs. Greece; but also inter-creditor distribution 3. How was the collective action problem addressed? 4. Was Greece’s December 2012 buyback a “boondoggle”? 5. Assessment and implications • Was the restructuring necessary? • Could it have been handled better? • Could it serve as a template for future restructurings? Boondoggle alert: sharp rise in Greek debt price ahead of buyback Price of Greek government bonds Buyback 50 45 Asmussen statement Buyback announced 40 35 30 25 20 15 10 5 2023 bond price 2042 bond price 0 • Caveat: price rise may also reflect anticipation of resumption in programme lending, improved official lending terms Numbers suggest no boondoggle, but additional debt relief was small 1. Buyback passes “boondoggle” test: reduction in market value of debt exceeds PV increase in EFSF debt. 2. Debt relief: 6-11 per cent of 2012 GDP (difference between PV of retired and additional EFSF debt, using 3.5-8 percent discount rate) 3. But good use of taxpayer money? Probably not. For example, debt relief could have been higher if buyback had been conducted at negotiated pre-buyback prices. Outline 1. A brief history of the Greek debt restructuring • 2. how did we go from denial to historic bond exchange to buyback of the new bonds? What were the distributional implications? • creditors vs. Greece; but also inter-creditor distribution 3. How was the collective action problem addressed? 4. Was Greece’s December 2012 buyback a “boondoggle”? 5. Assessment and implications • Was the restructuring necessary? • Could it have been handled better? • Could it serve as a template for future restructurings? Was the restructuring necessary? – Yes The argument against: decision to restructure in principle (July 2011) contributed to spread of Eurozone crisis to Italy and Spain. However, deep restructuring still was the right decision. 1. Greek debt unsustainable -- clear to IMF staff since beginning of programme; official position of IMF as of October 2011. 2. Only alternative to deep debt restructurings would have been a large official transfer (not just a loan), in excess of €100 billion. 3. Eurozone had instruments to undertake the restructuring while limiting the contagion (although it was initially reluctant to use them) 4. In the end restructuring was smooth and EZ did not fall apart. Could restructuring have been handled better? - Yes. 1. Left money on the table for tax payers (Greek and European): • One size fits all approach (PV cost: €21 billion) • Bank creditors went scot free (PV cost: €22 billion) • Equal treatment of foreign and domestic law bondholders • Came too late (by at least 6 months; cost >€10 billion); • Exclusion of ECB (initially even ECB profits) Result: Will require additional debt relief (in one go, or drip-drip-drip) 2. Exceptionally large cash component, financed by EFSF • implies that almost all residual risk is now borne by the European taxpayer. 3. Soft approach vis-a-vis holdouts: large cash sweetener ex ante; repaid in full ex post. Does Greece provide a template for other countries wanting to restructure? – Not really. Many European countries have large outstanding local law debt stocks – hence could use retrofit CAC, seniority upgrades as a substitute for confrontation. But: no template. 1. Concentration of sovereign debt in the hands of domestic banks limits maximum debt relief. 2. Less domestic law debt in several countries (Cyprus) 3. Foreign law bonds will be much harder to restructure: • Large-scale officially financed cash offers unlikely to be repeated (and not desirable). • Holdouts emboldened by successes in Greece, recent court cases (NML vs Argentina, Assenagon). How an (unlikely) new sovereign restructuring would need to look like in Europe 1. As in Greece, use domestic law to restructure domestic law bond holders. 2. Would likely need to include some classes of bank liabilities in restructuring (junior debt; possibly senior debt) 3. A much more coercive approach towards foreign bond holders (drop the pretense of voluntariness) Note: does not necessarily imply a higher haircut vis a vis foreign law bondholders – depends on required debt relief; quantity of “bailin-able” bank debt. Conclusion: a much riskier – financially and legally – operation than in Greece. Could be done, but unlikely to happen. Thank you Backup slides Debtor coerciveness: historical comparison Date Total coerciveness (max. 10) PostDefault Full Moratorium? Forced Exchange Explicit Threats Mexico (Brady deal) 4.2.90 3 Yes No No Yes Nigeria (Brady deal) 1.12.91 8 Yes No Yes Yes Brazil (Brady deal) 15.4.94 7 Yes Yes No Yes Bulgaria (Brady deal) 29.6.94 6 Yes Yes No Yes 27.10.94 5 Yes Yes No No 1.3.97 9 Yes Yes No Yes 13.12.99 3 No No No Yes 7.4.00 2 No No No Yes Ecuador 23.8.00 6 Yes No Yes No Russia (PRINs & IANs) 25.8.00 6 Yes Yes No No Uruguay 29.5.03 1 No No No No 1.4.05 9 Yes Yes Yes Yes Belize 20.2.07 2 Yes No No No Greece 9.3.12 2 No No No Yes Poland (Brady deal) Peru (Brady deal) Pakistan (Bond debt) Ukraine (Global Exch.) Argentina (Ext. debt) Source for all countries except Greece: Enderlein, Trebesch and von Daniels, 2012. Yield curve after the exchange (all new bonds long) Exit yields of new bonds on first day of trading 20 Yield 15 10 5 0 0 2 4 6 8 10 12 14 duration (years) 16 18 20 22 24 Can attempt to “impute” complete yield curve Peak default density: 0.5 years; SD 3 months Peak default density: 2 years; SD 6 months Peak default density: 0.5 years; SD 6 months Peak default density: 2 years; SD 12 months Volatility of yields after issuance Yields of shortest dated and longest new bonds, March-June 2012 35 25 20 15 10 2023 bond yield 5 2042 bond yield 25/6/12 18/6/12 11/6/12 4/6/12 28/5/12 21/5/12 14/5/12 7/5/12 30/4/12 23/4/12 16/4/12 9/4/12 2/4/12 26/3/12 19/3/12 0 12/3/12 Yield of new bonds 30 Computing PSI-induced creditor losses • Compare present value (PV) of new bonds with PV of old bonds, both evaluated at same, post-exchange market yield. Differs from market practitioners who compare PV of new bonds to outstanding face value of old bonds • Concept measures the loss incurred by a participating creditor if the counterfactual is keeping the old bond and being repaid with the same risk as participating creditors In contrast, counterfactual in practitioner concept is full and immediate repayment – no such right in Greece • Hence our concept measures the temptation to free ride at the time of exchange, conditional on repayment risk as it emerges after the exchange. Computing debt relief Creditor losses and debtor gains (debt relief) may not be the same for three reasons: 1. Exit yield may not be appropriate discount rate: if debtor intends to repay, and expects to borrow again in the future, should use expected borrowing rate in normal times • Not appropriate if assume debtor is chronically credit constraint – need to use higher discount rate (Dias et al, 2012). 2. From Greece’s perspective, the value of borrowed PSI sweetener should be lower than it’s face value (€29.7 bn) 3. Greece borrowed €22 bn from EFSF to compensate its banks for PSI-related losses