How to Restructure Greek Debt Lee C. Buchheit G. Mitu Gulati

Electronic copy available at: http://ssrn.com/abstract=1603304
Draft
5/7/10
How to Restructure Greek Debt
Lee C. Buchheit
G. Mitu Gulati
Abstract
Plan A for addressing the Greek debt crisis has taken the form
of a €110 billion financial support package for Greece announced by the
European Union and the International Monetary Fund on May 2, 2010. A
significant part of that €110 billion, if and when it is disbursed, will be used to
repay maturing Greek debt obligations, in full and on time. The success of
Plan A is not inevitable; among other things, it will require the Greeks to
accept -- and to stick to -- a harsh fiscal adjustment program for several
years.
If Plan A does not prosper, what are the alternatives? And how
quickly could a Plan B be mobilized and executed?
This paper outlines the elements of one possible Plan B, a
restructuring of Greece’s roughly €300 billion of government debt. Prior
sovereign debt restructurings provide considerable guidance for how such a
restructuring might be shaped. But several key features of the Greek debt
stock could make this operation significantly different from any previous
sovereign debt workouts.
To be sure, a restructuring of Greek debt will not relieve the
country from the painful prospect of significant fiscal adjustment, nor will it
displace the need for financial support from the official sector. But it may
change how some of those funds are spent (for example, backstopping the
domestic banking system as opposed to paying off maturing debt in full).
This paper does not speculate about whether a restructuring of
Greek debt will in fact become necessary or politically feasible. It focuses
only on the how, not the whether or the when, of such a debt restructuring.
* * * *
Electronic copy available at: http://ssrn.com/abstract=1603304
Draft
5/7/10
How to Restructure Greek Debt
Lee C. Buchheit*
G. Mitu Gulati**
This paper offers no opinion on whether Greece’s debt should
be restructured and, if so, when or on what financial terms. Nor does it
speculate on what consequences, intended or unintended, might flow from
such a restructuring. We will limit ourselves solely to the issue of how such a
restructuring might efficiently be undertaken -- drawing on the lessons of past
sovereign debt restructurings (successful and otherwise) -- if a decision is
made to proceed with a debt restructuring at some point in the future.
Financial Aspects of the Debt Stock
Greece’s total debt as of end-April 2010 was approximately
€319 billion. Of that figure, the vast majority -- approximately €294 billion --
was in the form of bonds, with another €8.6 issued as Treasury bills.
Virtually all of this debt stock was denominated in Euros. Small
amounts (in aggregate, less than 2% of the total) are outstanding in U.S.
dollars, Japanese yen and Swiss francs.
Information about the holders of Greek bonds is anecdotal.
From press reports, however, it appears that French and German banks
have the heaviest exposures,1 but mutual funds, pension funds, hedge funds
and other categories of investors also own Greek bonds. Significantly, the
extent of retail (non-institutional) ownership appears to be small.
* Cleary Gottlieb Steen & Hamilton, LLP
** Duke University Law School
1 James Wilson et al., “Worries persist on exposure to Greece,” Fin. Times, April 30, 2010.
2
A significant amount of Greek bonds have been discounted by
European commercial banks with the European Central Bank (“ECB”). On
May 3, 2010, the ECB announced that those bonds would continue to be
eligible for discounting, notwithstanding the downgrading of Greek bonds
below investment grade that occurred the prior week.2
Legal Aspects of the Debt Stock
Governing law. From the legal standpoint, the salient feature of
Greece’s bond debt is that approximately 90% of the total is governed by
Greek law. Only about €25 billion of the bond debt was issued under the law
of another jurisdiction, and most of that under English law.
Collective action clauses. It does not appear from our survey
of Greek bond documentation that the instruments issued under local law
contain provisions permitting the holders to amend the terms of the bonds
after issuance (other than to correct obvious errors or technical matters).
Prior to 2004, Greek bonds issued under English law contained collective
action clauses (“CACs”) that appear3 to permit holders of 66 percent of an
issue to modify payment terms in a manner that would bind all other holders.
After 2004, Greece altered this clause in its English law bonds.4 This new
version of the CAC permits amendments to payment terms of a bond, as well
as certain other key features of the instrument, with the consent of holders of
75% or more of an issue.
Negative Pledge. Greece does not appear to have included a
negative pledge clause in its bonds issued under local law.
2 See Press Release, ECB Announces Change in Eligibility of Debt Instruments Issued or
Guaranteed by the Greek Government (May 3, 2010) available at
http://www.ecb.int/press/pr/date/2010/html/pr100503.en.html).
3 We say “appear” because the descriptions of the modification clauses in the Offering
Circulars for these bonds routinely conflate the notion of a quorum (that is, the number of
bonds required to activate a meeting of bondholders) with the percentage of bonds required
to approve an Extraordinary Resolution.
4 In 2002, a working group of the G-10 issued a report recommending reform of the standard
sovereign debt contract. See Report of the G-10 Working Group on Contractual Clauses
(September 2002) (available at http://www.bis.org/publ/gten08.htm). The Report contained
a recommended version of a collective action clause for sovereign bonds as well as other
model clauses. The new Greek CAC derives from this source.
3
The foreign law Greek bonds we have examined do contain a
negative pledge clause, but in a very unusual form. This clause requires
Greece equally and ratably to secure each of these bond issues if ever it
creates or permits to subsist any form of security interest over its revenues,
properties or assets to secure any “External Indebtedness.”
In the case of a typical emerging market sovereign bond,
external indebtedness is defined to cover borrowings denominated in a
currency other than the currency of the issuing State. And so it was with
Greek bonds issued prior to January 1, 2000 when the Euro was adopted as
the common currency of the European Union. After that date, however, the
Euro became the “lawful currency” of Greece. But Greece’s form of negative
pledge clause did not change for more than four years after the adoption of
the Euro5; the “lawful currency” just quietly ceased being the Drachma and
became instead the Euro. In practical terms, this means that Greece’s
negative pledge clause in its foreign law bonds issued prior to 2004 would
only be triggered by the creation of a lien to secure a non Euro-denominated
Greek debt.
Events of Default. Most of the Greek bonds we have examined
incorporate a standard set of Events of Default (variations can be found in a
few bonds). These include:
• failure to pay interest (with a 30-day grace period);
• other covenant defaults (with a 30-day grace period after
written notice to the issuer);
• “External Indebtedness” (above a de minimis threshold) is
accelerated, or a payment is missed under such External
5 Starting in 2004, the definition of “External Indebtedness” in Greece’s foreign law-governed
bonds changed. Under the new version of the definition, indebtedness for borrowed money
is “External” if either (i) it is denominated in a currency other than Euros or (ii) borrowed from
foreigners under a foreign law-governed contract. This change in Greece’s standard bond
documentation occurred at roughly the same time as the country began using a CAC
modeled on the G-10 recommended clause, and adopted a 25% voting threshold for
acceleration (also recommended by the G-10).
4
Indebtedness and continues beyond the specified grace
period;6
• a general moratorium is declared on “External
Indebtedness”; and
• any government order or decree prevents Greece from
performing its obligations under the bonds.
The important point to note here is that the cross-acceleration,
cross-default and moratorium event of default clauses in these bonds apply
only to External Indebtedness; a term that in Greek bonds issued prior to
2004 excludes Euro-denominated obligations. Thus, these Events of Default
in pre-2004 bonds could only be triggered by an event affecting the small
amount of Greek debt that is denominated in currencies other than Euros.
Stated differently, a payment default on a Greek Euro-denominated bond, or
the acceleration of such an instrument, would not have cross-default
consequences across much of the debt stock.7
Bondholder remedies. Prior to 2004 (when Greece adopted a
new form of CAC), individual bondholders could accelerate their bonds
following the occurrence of an Event of Default. Bonds issued after that time
incorporate a requirement that holders of 25% of the bonds vote in favor of
an acceleration -- one of the provisions recommended by the G-10 Working
Group on Contractual Clauses in 2002. (Individual rights of acceleration are
now rarely found in emerging market sovereign bonds.)
Greece in a Restructuring Context
Greece would enjoy some distinct advantages in the
restructuring of its debt, as well as a few disadvantages, in comparison with
other countries that have been forced to undergo the process.
Advantages
• Greece’s debt is overwhelming in the form of bonds, a
characteristic shared by countries like Argentina (2001-05),
6 These cross-acceleration and cross-default clauses appear in some, but not all, of
Greece’s bonds issued under local law.
7 In Greek bonds issued after 2004, these Events of Default would be triggered by an event
affecting Republic debt that was either non Euro-denominated or borrowed from foreigners
under a foreign-law governed agreement.
5
Ecuador (2000) and Uruguay (2003). Greece will therefore
be able to avoid the complexities and intercreditor rivalries
that can be occasioned by a diverse creditor universe. Iraq
(2005-08), for example, faced a commercial creditor class
composed of banks, suppliers, construction companies,
various kinds of trade creditors and even individuals. In the
end, Iraq had to settle more than 13,000 individual claims
dating back to the Saddam era.
• Greece’s debt records are fresh and up to date. When
sovereign debts have been in default for prolonged periods
before they are restructured (for example, Liberia (2009) --
more than 25 years in arrears; Iraq (2005-08) -- more than
15 years in arrears), the task is much harder.
• Greece enjoys considerable financial support from
multilateral and bilateral sources. This opens up the
possibility of being able to “credit enhance” any new Greek
debt instrument that might be issued as part of a debt
restructuring. Greece is therefore in a position similar to the
countries that issued collateralized Brady Bonds starting in
1990.
• Related to the last point, the odd nature of Greece’s
negative pledge clauses noted above (the clauses in bonds
issued prior to 2004 would not be triggered by the creation
of liens to secure future Euro-denominated indebtedness of
Greece) means that some form of collateralized Brady Bond
approach might be legally feasible for Greece without
having to obtain waivers of negative pledge restrictions in
foreign law-governed instruments issued prior to 2004.
• The fact that so few Greek bonds seem to have fallen into
the hands of retail (non-institutional) investors may also be
a blessing. Those countries (for example, Argentina (2001-
05)), that have had to deal with thousands of retail
bondholders in a debt restructuring have found this a messy
and thankless task.
• By far, however, the greatest advantage that Greece would
enjoy in a restructuring of its debt derives from the fact that
so much of the debt stock is expressly governed by Greek
law (90% or more, if our figures are correct). This raises the
6
possibility, discussed in more detail below, that the
restructuring could be facilitated in some way by a change
to Greek law.
Several other countries have restructured local lawgoverned
debts. Russia restructured Russian law-governed
instruments known as GKOs and OFZs in 1998 in parallel
with the restructuring of its English law commercial bank
debt. Uruguay restructured its local law bonds (on the
same terms as its foreign law bonds) in 2003. In each of
these prior cases, however, the local law bonds were also
denominated in local currency and formed only part of the
overall stock of the debt being restructured. While the Euro
is certainly now the local currency of Greece, it is a good
deal more besides that.
No other debtor country in modern history has been in a
position significantly to affect the outcome of a sovereign
debt restructuring by changing some feature of the law by
which the vast majority of the instruments are governed.
Disadvantages
Several features of the Greek situation may complicate any
restructuring of its debt.
• A significant percentage (perhaps more than 30 percent) of
the bonds are believed to be owned by Greek institutional
holders. A restructuring that dramatically impairs the value
of that paper could therefore place further strains on the
Greek domestic financial sector. Jamaica, which
restructured its local currency debt earlier this year, had to
walk very gingerly in designing its transaction because so
much (perhaps as much as 75%) of the affected paper was
held by Jamaican financial institutions.
• European banks are the largest holders of Greek bonds.
The stability of those institutions will be therefore very much
on the minds of Greece’s multilateral and bilateral
supporters should a debt restructuring prove unavoidable.
The sovereign debts that triggered the global debt crisis of
the 1980s were nearly all owed to international commercial
7
banks. When the debt crisis broke in 1982, the official
sector overseers of the restructuring process therefore had
to keep one eye on the debtor countries and the other on
the stability of the banking systems in the industrialized
creditor countries. It remained an uncomfortable position
for nearly seven years until the banks had time to build up
the loan loss reserves that permitted them to accept losses
in Brady Bond exchanges without alarming regulators,
stockholders or their own creditors.
In contrast, sovereign debt crises of the last 10 years or so
have affected mostly non-bank creditors -- hedge funds,
pension funds, other institutional holders of emerging
market sovereign debt, sometimes even individuals. Those
crises did not threaten the stability of the banking sectors in
creditor countries.
A restructuring of Greek debt will, in this respect, rekindle
fretful memories of the global debt crisis of the 1980s in the
minds of official sector observers.
• Greece’s debt is denominated in Euros, a currency shared
by other members of the European Union. When a Mexico
or a Philippines restructured debts denominated in U.S.
dollars in the 1980s, no one -- for that reason alone -- lost
confidence in the U.S. dollar. This same assurance cannot
be given about the restructuring of Euro-denominated debts
owed by an EU Member State.
How Might It Be Done?
Which brings us to the main event. If one were to attempt to
glean the lessons of the 50 or 60 sovereign debt restructurings of the modern
era, what would they teach about how a Greek debt restructuring should be
managed in order to achieve a prompt, orderly and fair outcome? There is
one lesson from this history that is inescapable. A sovereign debt crisis can
be a painful experience for both the debtor and its creditors; a mismanaged
sovereign debt crisis can be a catastrophically painful experience.
Transaction structure. The most likely structure for such a
transaction would be an exchange offer -- new bonds of the Hellenic
Republic would be offered in exchange for the Republic’s existing bonds.
The terms of the new bonds would determine the nature and extent of the
8
debt relief that the transaction would provide to Greece. Exchange offers
have been the norm for sovereign debt restructurings of middle income
countries since the first Brady Bond transactions in 1990.
Eligible debt. The debt eligible to participate in such an
exchange would presumably be all outstanding Greek bonds, excluding
perhaps the Treasury bills. Short-term Treasury bills have been excluded
from a number of recent sovereign debt restructurings in order to keep that
market sweet for the government’s emergency financing needs.
An interesting question will be whether some way can be found
to exclude from the restructuring the bonds in retail hands, or at least to
moderate the terms of any restructuring of those bonds. This will depend on
two things. First, the total amount of bonds in the hands of natural persons
would have to be relatively small. Institutional holders may recognize the
public relations benefit of not having widows and orphans complaining on the
evening news, but only up to a point. If the exclusion of retail holders
appreciably increases the severity of the financial sacrifice that must be
borne by institutional creditors, these sympathies will quickly fade. Second,
some mechanism must be found that will permit the government to identify
which bonds are in retail hands when the restructuring is announced.
Otherwise, bonds will tend to migrate temporarily into the hands of
individuals until the restructuring storm passes over.
New instruments. The terms of the new instrument or
instruments that would be offered in such an exchange will be a function of
the nature and extent of the debt relief the transaction is designed to
achieve. At the soft end of the spectrum would be a simple “reprofiling” of
existing bonds (or some discrete portion of them such as bonds maturing
over the next three to five years) involving a deferral of the maturity date of
each affected bond. Uruguay (2003) stretched out the maturity date of each
of its bonds by five years, while leaving the coupons untouched.
At the sharper end of the spectrum would be a transaction
designed to achieve a significant net present value (“NPV”) reduction in the
stock of debt. If Brady Bonds were chosen as the model for the transaction,
this might entail allowing holders to elect to exchange their existing credits
for either a Par Bond (a new bond exchanged at par for existing instruments,
having a long maturity and a low coupon), or a Discount Bond (a new bond
exchanged for existing instruments at a discount from the face amount of
those instruments, but typically carrying a higher coupon and perhaps a
shorter maturity than the Par Bond). The precise financial terms of the Par
9
Bond and the Discount Bond would be calibrated to achieve an equivalent
NPV reduction.
CACs. Some version of a collective action clause appears in
most of Greece’s foreign law-governed bonds. There is no reason not to use
these clauses to minimize the number of non-participating creditors.
It would work as follows: each tender of an existing bond
containing a CAC would contain a power of attorney from the owner of that
bond in favor of the government (or its exchange agent) to vote that bond at
a bondholders’ meeting (or in a written action by bondholders) in favor of a
resolution that, if approved by the requisite supermajority of holders of that
instrument, would either cause the totality of that bond to be tendered in the
exchange or cause the payment terms of the bond to be amended so as to
match the terms of one of the new instruments being offered in the
exchange. Such a resolution, if approved by the requisite supermajority of
holders (66% or 75% in Greek bonds governed by English law) would
automatically bind all holders.
Creditor consultations. As part of the restructuring process,
Greece would have to consult with significant holders of its paper, or with
committees or other ad hoc groups representing those holders. Such
consultations would be necessary to garner widespread support for the
restructuring. In addition, the IMF’s so-called “Lending Into Arrears” policy
requires a member country facing a debt restructuring to make “a good faith
effort to reach a collaborative agreement with its creditors”.
Credit enhancement. If Greece’s multilateral and bilateral
supporters were prepared to allow some portion of their emergency financing
to be used for this purpose, Greece might be able to enhance to
attractiveness of the new bonds it would offer in the exchange. In the typical
Brady Bond context, for example, the issuing countries borrowed money
from the IMF and the World Bank and used those funds to purchase U.S.
Government zero coupon obligations that were then pledged to secure the
principal payment due on the Brady Bonds at maturity. In many cases, a
pool of cash equal to 12-24 months of interest accruals on the Brady Bonds
was also set aside and pledged as partial security for the interest due on the
Brady Bonds. So the Brady Bonds represented U.S. Government risk for
principal due at maturity, and issuing country risk for interest during the life of
the Bond (apart from any amount pledged as rolling interest collateral).
The negative pledge clauses in the debt instruments of the
Brady countries did not pose an insuperable obstacle to these collateralized
10
transactions. The commercial bank creditors waived the application of the
negative pledge restrictions in their agreements, as did the World Bank and
other multilateral development banks. In the small number of cases where a
Brady country also had outstanding international bonds, arrangements were
put in place to post “equal and ratable security”.
For the reasons noted above, Greece’s negative pledge
clauses contained in bonds issued prior to 2004 would not protect holders of
those bonds in the event that Greece were to collateralize a future Eurodenominated
issue of securities. In the case of existing bonds with negative
pledge clauses that would be triggered by the issuance of a new
collateralized security, Greece would presumably seek a waiver of the
negative pledge restrictions when it invited tenders of those issues in the
exchange offer. Failing receipt of such a waiver, Greece would be obliged to
post “equal and ratable” collateral for that issue of bonds.
Credit enhancement need not take the form of collateral
security. Indeed, over time, the market grew less fond of Brady Bonds and
most countries have retired their Brady Bonds early. Obtaining a partial
guarantee of the new instruments from a creditworthy party is another option.
(The Seychelles obtained a partial guarantee from the African Development
Bank of the new debt instrument issued by the Seychelles in connection with
its debt restructuring earlier this year.) Another alternative would be to offer
holders of the new instruments a continuing “put” of those instruments to a
creditworthy party, presumably at some discount from face value. This
would allow an institutional holder of the paper to ensure that the instrument
will always have a minimum floor value, no matter what happens to the
trading price in the secondary market.
The Tactical Implications of Local Law Bonds
International investors are often leery of buying debt securities
of emerging market sovereign issuers that are governed by the law of the
issuing state. Why? Because investors fear that the sovereign might
someday be tempted to change its own law in a way that would impair the
value or the enforceability of those securities. Such changes in local law
would normally be respected by American and English courts if the debt
instruments are expressly -- or otherwise found to be -- governed by that
local law.
International capital market borrowings by industrialized
countries sometimes follow a different model. Many of these countries have
found that foreign investors are prepared to purchase local law-governed
11
debt securities of the sovereign. These investment decisions are
presumably based on a belief that industrialized countries are less likely than
some of their emerging market brethren to risk eroding future investor
confidence by opportunistically changing their own law in order to reduce
government debt service burdens.
No country in Greece’s position would lightly consider a change
of local law as an easy method of dealing with a sovereign debt crisis. The
following factors, among others, counsel extreme caution before embarking
on such a remedy.
• If done once, future investors will fear that it could be
done again. The debtor country may therefore be
compelled in future borrowings (in which international
investor participation is sought) to specify a foreign law
as the governing law of its debt instruments.
• A dramatic change in local law by one country might
allow a worm of doubt to slip into the heads of capital
market investors in other similarly-situated countries,
driving up borrowing costs around the board.
• The official sector supporters of the debtor country will
presumably balk at any action of this kind that could
unleash the forces of contagion and instability upon
other countries whose debt stocks also contain
predominantly local law-governed instruments.
• The more dramatic or confiscatory the effect of the
change of law, the higher the likelihood that it would be
subject to a successful legal challenge. More on this
below.
One legislative measure that might be perceived as balanced
and proportional in these circumstances, however, would be to enact what
amounts to a statutory collective action clause. It could operate in this way:
local law would be changed to say that if the overall exchange offer is
supported by a supermajority of affected debtholders (say, 75%, to use the
conventional CAC threshold), then the terms of any untendered local law
bonds would automatically be amended so that their payment terms (maturity
profile and interest rate) match those of one of the new instruments being
issued in the exchange.
12
Such a law, let’s call it a “Mopping-Up Law”, would thus operate
in the manner of a contractual collective action clause in a syndicated debt
instrument. Once the supermajority of creditors is persuaded to support an
amendment to the payment terms of the instrument, their decision
automatically binds any dissident minority.
Viewed another way, the Mopping-Up Law would merely
replicate at the level of the sovereign borrower the same protection enjoyed
by corporate borrowers in many countries, including Greece. For example,
we understand that in corporate reorganization proceedings under Greek
bankruptcy law, if a plan of reorganization is accepted by two thirds of the
affected creditors (including at least 40 percent of “privileged claims” such as
secured or senior claims), it will -- with court approval -- bind all creditors. A
Mopping-Up Law would achieve a similar result but at the level of a
sovereign borrower in need of a debt reorganization.
Facilitating a sovereign debt restructuring through some form of
Mopping-Up Law would be consistent with the fundamental principle that a
sovereign debtor bears the burden of persuading its creditors that a debt
restructuring is essential, that the terms of the restructuring are proportional
to the debtor’s needs, and that the sovereign is implementing economic
policies designed to restore financial health. The only question is whether
the sovereign must persuade every last debtholder of these elements, or just
a specified supermajority of affected creditors. The trend in recent years, as
evidenced by the rapidity with which CACs have been introduced into New
York-law sovereign bonds, is in favor of the supermajority threshold.
Even the relatively mild step of facilitating a debt restructuring
through the passage of a Mopping-Up Law of some kind, however, could
draw a legal challenge. In the case of Greece, such a challenge could come
from three possible sources. The first is Article 17 of the Greek Constitution.
That Article declares that no one shall be deprived of property “except for
public benefit” and conditional upon payment of full compensation
corresponding to the value of the expropriated property. The question, it
seems to us (non-Greek lawyers that we are), is whether a mandatory
alteration of the payment terms of a local law Greek bond in the context of a
generalized debt restructuring could be said to impair the value of that bond;
an instrument that, in the absence of a successful restructuring, would have
in any event been highly impaired in value. Also of possible relevance may
be Article 106 of the Greek Constitution which gives the State broad powers
to “consolidate social peace and protect the general interest.”
13
A second source of possible legal concern might lie in the
European Convention on Human Rights and its Protocols. Article 1 of
Protocol No. 1 protects the right to the “peaceful enjoyment of possessions”.
This right may be restricted only in the public interest and only through
measures that do not impose an individual and excessive burden on the
private party. That said, Article 15 of the Convention permits measures,
otherwise inconsistent with the Convention, to deal with a “public emergency
threatening the life of the nation”.
Finally, foreign holders of local law-governed Greek bonds
subject to the Mopping-Up Law might look to Greece’s Bilateral Investment
Treaties for redress. BITs protect against expropriation without
compensation, as well as unfair and inequitable treatment. It appears that
Greece has signed more than 40 BITs with bilateral partners.
Assuming some version of a Mopping-Up Law could survive
any legal challenge, however, it could have significant tactical implications for
a Greek debt restructuring. More than 90% of Greek bonds are governed by
local law. If, to use our example, holders of 75% of all eligible bonds (local
law and foreign law) were to support a restructuring, our version of a
Mopping-Up Law should operate to ensure that more than 90% of the debt
stock will be covered by the restructuring. The Mopping-Up Law would not
affect holders of foreign law bonds. Participation by those holders would
need to be encouraged by moral suasion and the use of contractual
collective action clauses in the relevant bonds.
How Long Would a Restructuring Take?
Our guess? If done efficiently, five to six months, less if
necessary.
One month or so would be needed for preparation; one to two
months for creditor consultations; one month during which the exchange
offer would be in the market and road shows would take place; another four
to six weeks to convene bondholder meetings for those bonds containing
CACs, and two to four weeks to prepare for a closing.
If done efficiently.
* * * *