La Ristrutturazione del Debito Greco Decisa a Bruxelles

Riproduco sotto una esaustiva spiegazione delle decisioni circa la ristrutturazione del debito greco prese ieri a Bruxelles. L’autore è David Nowakowski, Fixed Income and Credit Strategist at RGE (Roubini Global Economics)
For over a year here at RGE, we have argued that a solution to the Greek debt insolvency crisis should include an orderly and market-oriented but coercive debt exchange, with a par bond for hold to maturity investors (banks, insurance companies, pension funds) and a discount bond for mark-to-market investors (hedge funds, etc.). As we reiterated earlier in a note on July 13:

“Since a restructuring of the Greek public debt is the only feasible option, the question is how to achieve an orderly restructuring without causing a banking crisis in Greece or the creditor countries. The solution that we have proposed for quite a while is the same that has been used in a number of sovereign debt restructurings in emerging market economies (for example, Uruguay, Pakistan and Ukraine): an exchange of the entire stock of public debt with new debt with the same face value, a significant maturity extension (20 to 30 years) and a coupon that is not only much lower than the current unsustainable market rates but also a little lower than the original coupon. A maturity extension with no face-value reduction would provide massive debt relief to the Greek sovereign: A euro due in 30 years is worth less than a euro due in a year on a net present value (NPV) basis, and such a par bond is equivalent—on an NPV basis—to a discount bond with lower face value and a higher coupon. And an exchange offer should provide a menu that includes—in addition to a par bond for hold-to-maturity investors such as banks—a discount bond that would be appealing to mark-to-market investors such as hedge funds.”

And indeed, today the eurozone (EZ) has proposed, and the private sector—as represented by the Institute of International Finance (IIF) that includes most of the large global banks —has accepted, an orderly debt exchange of the entire public debt of Greece coming to maturity between 2012 and 2020: The agreement includes two par bond options and two discount bond options. The maturity extension is 30 years for three of the options and 15 years for the fourth discount bond option. Private-sector involvement is estimated at EUR50 billion in 2011-2014, to reach EUR106 billion through 2019. So, as we recommended, the maturity extension will include all of the debt maturing in the next decade, as opposed to—as suggested by some—the debt maturing only in 2012-2014; thus, the amount of debt relief for Greece is now much more significant than a partial debt re-profiling of only some maturities.

Also the coupon on the new bonds is relatively low and—as we recommended—close to the original coupon: The coupon on the par bond will be 4.0% during the first five years, 4.5% during the next five years, and 5.0% for years 2011-2030; the coupon on the discount bond will be obviously higher: 6.0%, 6.5% and 6.8%, respectively. These are rates that are relatively low and, as in the Brady plan, show a step-up feature for the coupon.

We had recommended that credit enhancement should not be given in this debt exchange as they would be too expensive and unwarranted but we did not rule out their use:

“Talk of credit enhancements—like in the Brady plan and the ill-conceived French initiative for a voluntary rollover—should be forgotten as such sweeteners are extremely expensive for the sovereign and not justified. They involve creditors not providing new money or debtor-in-possession (DIP) financing but just rolling over their existing claims. Thus, the private creditors of Greece do not deserve such enhancements and will accept an exchange offer without such sweeteners, just as creditors did for Pakistan, Ukraine, Uruguay, etc.”

Unfortunately, credit enhancements have been added—as in the original Brady Plan—to this debt exchange: Three of the four new bonds have a full principal collateral guarantee (collateralized by 30-year zero coupon AAA Bonds), while the fourth bond has a partial collateral. The principal is partially collateralized through funds held in an escrow account. This suggests that the implied interest rate on the new debt is higher than the official coupon by about 100 basis points. Thus, the overall cost for Greece will be higher than the original coupons, but much lower than the current market yields. However, if Greece is successful in servicing these new bonds, it may choose—as did countries that participated in the original Brady plan—to retire the new bonds early and thus recapture the AAA bonds that it has to purchase to provide the collateral guarantee. So the additional cost to Greece that comes from the principal guarantee may end up to be ex-post lower than its ex-ante value.

Overall, the IIF debt exchange proposal is quite generous to banks, and does not provide enough enhanced debt sustainability to Greece, which goes from paying a 4.5% coupon to paying roughly a 4.5% coupon (a bit of deferred interest, made up by a higher coupon later). In addition, Greece is asked to provide a principal guarantee. This means that the country needs to purchase a 30y zero-coupon strip; a France OAT 0 Apr-2041 has a 4.265% yield and trades at 29. So for every 100 that is rolled over, Greece will have to pay 4.5% to the bondholder that extends maturity, and borrow 29 cents from the EFSF (which lends at 3.5%) to put a French or EFSF principal strip in escrow somewhere. Greece won’t actually lose money on this, as eventually the bond will be paid off and the monies in escrow released (as long as it does not default on the new bonds), but in the meantime it adds the 102bps in interest cost (0.29 x 3.5%) to the debt service. So the debt service to bondholders and the EFSF on the existing stock of bond debt goes up to around 5.5% and eventually 6.5% (bonds maturing in 5-years would be guaranteed by a 35-year strip, so the cost would be a bit lower, etc).

Note that the original French proposal for a voluntary “Vienna Initiative” that required 30-cents cash payout was much more onerous for Greece, with implied yields on the new exchanged debt north of 10%; luckily, that unsustainable offer for Greece was scrapped. Now the principal guarantee limits losses for creditors to 20-25%. At the assumed exit yield of 9%, the new 30-year bonds would indeed trade at about 75 (46 annuity at 9% + 29 principal strip at 4.27%). If instead Greece’s exit yield is 13%, the new bonds would be worth about 62, a bit higher than most maturities trade today, so enough to induce even non-IIF members—that are mostly mark to market investors—to go into the exchange.

Still, the cost for Greece to fund the purchase of this 30-year zero coupon bond will be much higher than that of the Latin American economies that followed the Brady plan: As bond yields on long-dated AAA bonds are much higher today than they were in the early 1990s, today’s value of a 30-year zero coupon bond is much higher than two decades ago. So, unless long rates fall further in the next few years the benefits to Greece to retire early such Brady-like collateral will be very modest. The EU pledged to back Greek collateral for an amount of EUR35 billion, what it will approximately cost to provide collateral (via 30-year zero coupon AAA bonds) for a face value of debt to be exchanged of about 109 billion euros. But it is Greece that will pay for this collateral.

The IIF expect large participation rate—about 90%—in this debt exchange; this expected participation rate is likely and would be similar to that in previous orderly sovereign debt restructurings (Pakistan, Ukraine, Uruguay). The holdout risk would have been very low even without the credit sweeteners, as there were other coercive ways to bind in potential holdouts; but the credit sweetener ensures a very high participation rate for all investors still holding Greek debt, both hold to maturity ones and mark to market ones.

As we have pointed out in numerous notes, the “credit event” issue and the downgrade of Greece to Selective Default by rating agencies will be a red herring issue as, in any successful debt exchange, the country will be in Selective Default for only a few days or weeks. As we pointed out for a while:

“Would rating agencies consider such an exchange offer a credit event? Yes, they would, as they did in Uruguay and other episodes of sovereign debt exchange as these exchanges occur under the implicit threat of default. But this should be of little concern in ensuring an orderly debt restructuring. In the case of Uruguay, the downgrade of the sovereign debt to SD occurred not after the announcement of the intention to do an exchange offer but rather for a short period of two weeks between the launch of the exchange offer and its successful completion. After the conclusion of the offer, the country was upgraded back from SD to a junk rating as the offer significantly increased the solvency of the country and its sovereign. Greece and the ECB could comfortably live with a similar temporary downgrade to SD—for a few weeks, at least.”

Will this credit event—as defined by the rating agencies—trigger the CDS? The answer, as we argued before, is most likely not:

“Would this par bond exchange offer be viewed as a credit event? Not in the sense that it would trigger CDS contracts. Indeed, ISDA rules suggest that a debt exchange is considered a ‘restructuring’ or credit event only if either collective action clauses (CACs) or domestic legislation is used to change the financial terms of the debt. In the case of Greece, the debt does not include CACs, and domestic legislation would not be necessary to ensure a successful debt exchange with a small number of holdouts; other tools could be used to deal with potential holdouts. Rather paradoxically, a debt restructuring that does not trigger CDSs could damage this asset class: Those who bought CDS insurance may find that the expected protection is not delivered if a debt restructuring turns out not to be a credit event for the purposes of triggering CDSs. Moreover, even if domestic legislation were used and CDSs were triggered the net outstanding amounts are so small in the case of Greece—about €5 billion—that the triggering of CDSs would have little systemic effect.”

Moreover, the fact that private creditors—specifically the IIF that represents all the major global banks—are fully on board for this debt exchange increases the probability that ISDA will not consider this exchange offer as a credit event; thus, it is very likely the CDS will not be triggered even if this exchange offer is effectively a credit event. It is voluntary but successful only under the implicit threat of default.

Of course the ECB had to let go its request that a credit event should be avoided at any cost. And ECB President Jean-Claude Trichet, in the post-summit press conference, did not rule out that the possibility that the new exchanged debt will be acceptable as collateral for the ECB in spite of the likelihood that rating agencies will downgrade the debt to SD. The fig leaf for the ECB is twofold: Firstly, as we pointed out, the debt will be in SD rating only for a very short period of time. Thus, as we argued before:

“The ECB has made the irrational argument that it could not accept repo debt that is downgraded to SD as collateral. That position is illogical: Since the ECB has substantially reduced its criteria for what is acceptable collateral, it could live with a short period in which the Greek debt is downgraded to SD. During that period, the ECB may decide to not accept further Greek debt as collateral for repos, but it would be forced to maintain its existing exposure to the debt that was already offered as collateral. And after the upgrade from SD has occurred it would continue its current practice of accepting junk debt as collateral for repo at the appropriate haircut.”

Secondly, the most important fig leaf for the ECB is that the new exchanged debt will have a principal guarantee: even leaving aside the fact that the ECB repos operations are vastly over-collateralized in the first place, the additional principal guarantee via the 30-year zero coupon AAA bonds fully reassures the ECB of the safety of such collateral.

These four instruments “will be priced to produce a 21% Net Present Value (NPV) loss based on an assumed discount rate of 9%.” This is a loss that is lower than the one priced in the current market value of the bonds and thus reduces the debt relief benefit to Greece below what is priced by the markets today; what reduces the NPV loss for creditors is of course the very sweet and rich principal value collateral guarantee. Thus, the degree of NPV debt relief that Greece obtains is lower than the one priced by the markets and lower than our estimates of the required debt relief (30 to 50% NPV debt reduction) to make the debt sustainable.

The IIF Financing Offer does not mention whether the new bonds will include collective action clauses (CACs); most likely they will include such CACs along the lines of previous orderly sovereign debt exchange, such as in Uruguay. As we have argued before, the introduction of CACs will make it easier to do a second round of debt reduction in a few years if—as is possible—the current debt exchange is not sufficient to restore debt sustainability in Greece.

The other elements of the Greek rescue plan are quite standard and along the lines of what we suggested and recommended as sensible and likely (little reliance on buyback; long maturities and low rates on official financing; use of EFSF to recapitalize Greek banks; more flexible rules on use of the EFSF):
As we had argued, a debt buyback option was one of the least efficient ways to provide debt relief to Greece, because most of the benefits of a buyback would go to the creditors as the residual price of the debt goes higher as the buybacks occur. Thus, in the final plan, as a fig leaf for those supporting buybacks, this option is kept; but such buybacks are expected to cut Greek debt by EUR12.6 billion, a figure that implies a token contribution of buybacks to the private-sector involvement in Greece; most of the PSI—as we recommended—will occur through the debt exchange.
Official aid (excluding PSI) will be about 109 billion euros. The official loans to Greece will have a rate of about 3.5% with maturities of at least 15 years and up to 30 years; there will also be a grace period of 10 years.
The EU will also lend another 20 billion euros to Greece—possibly via the now more flexible European Financial Stability Facility (EFSF)—to recapitalize the Greek banks.
The 440-billion-euro EFSF will now be allowed to buy debt in the secondary market to do buybacks; it will also be used to recapitalize weak EZ banks and to intervene to reduce market pressures against the bonds of other EZ periphery economies (such as Italy and Spain).

The EU officials, Trichet and the IMF also reiterated that this coercive debt restructuring of Greece’s debt is an “exceptional case” that will apply only to Greece. That is clearly—as we have repeatedly argued—wishful thinking and cheap talk: Both Portugal and Ireland are likely to be as insolvent as Greece; and in a year’s time, when Plan A for Portugal and Ireland fails—as it did for Greece—the need for an orderly debt restructuring of their public debt will become clear and unavoidable. The only way that Ireland will likely avoid a sovereign debt restructuring would be to reverse its seriously flawed decision not to treat all of the senior—guaranteed and unguaranteed—debt of its insolvent banks. Hopefully, in due time, Irish policy makers will regain their senses and do the right thing; otherwise the likelihood of a coercive debt restructuring—a la Greece—becomes very high.

So, all in all this is an OK solution for Greece along the lines of what we had strongly recommended; the main caveat is that the private creditors accepted PSI in exchange for a very sweet credit enhancement that significantly increases (relative to risk-free rates) the implied interest rate on the new bonds, even if the overall yields will be lower than the current high and unsustainable market rates; thus, the credit sweetener significantly reduces the NPV debt reduction that Greece will obtain to only 21%, a level that, over time, may not be sufficient to restore debt sustainability that—based on RGE calculations—requires a 30 to 50% NPV reduction of the debt burden.

David Nowakowski, Fixed Income and Credit Strategist at RGE, contributed to this note.