The IMF revisits sovereign bankruptcy
Last weekend, at the IMF annual meetings, I moderated an official panel with the snooze-worthy title “Sovereign Debt Restructuring: Lessons from Recent Experience”. But the room was packed, and attention was rapt: everybody wanted to know what the panelists in general, and one in particular, thought about the subject at hand. All eyes were on first deputy managing director David Lipton, who kicked off proceedings with a dry but important speech in which he praised a recent Brookings report as “excellent”.
Lipton’s speech came in the wake of a 49-page paper the IMF released in May, which tried to lay out the issues as seen from Fund headquarters. The paper raised quite a few eyebrows, since it marked the first time in a decade that the IMF has talked in public about changing the international financial architecture around debt restructuring. Its last attempt to tackle the subject, known as the Sovereign Debt Restructuring Mechanism, or SDRM, died ignominiously, bereft of any US support.
What do I think about Lipton’s idea? If you scroll down to the end of this rather long and dense post, you’ll find out. It’s OK, there won’t be a quiz, and you don’t need a detailed understanding of everything I’m going to cover in order to understand my conclusions. But for the suprisingly large number of sovereign debt nerds out there, I’m first going to lay out the issues, as they were presented during the IMF meetings.
Lipton, in his speech, said that he was worried that “official resources, including from the Fund, would be used to pay out other creditors”. He also said that “in cases where the need for debt reduction may be unclear at the outset, in our view the key is to keep creditors on board while the debtor’s adjustment program is given a chance to work”.
This idea is very close to the “standstill” that was originally proposed as part of the SDRM; another name for it is “default”. And as veteran sovereign debt advisor Rafael Molina patiently explained later on in the panel, sovereign debt managers will, as a rule, do anything to avoid defaulting on their debt. As a result, tensions are naturally very high whenever this idea is brought up, despite the upbeat spin that the IMF puts on it in its paper:
The primary objective of creditor bail-in would be designed to ensure that creditors would not exit during the period while the Fund is providing financial assistance. This would also give more time for the Fund to determine whether the problem is one of liquidity or solvency. Accordingly, the measures would typically involve a rescheduling of debt, rather than the type of debt stock reduction that is normally required in circumstances where the debt is judged to be unsustainable. Providing the member with a more comfortable debt profile would also have the additional benefit of enhancing market confidence in the feasibility of the member’s adjustment efforts, thereby reducing the risk that the debt will, in fact, become unsustainable.
Translating into English, the IMF here is essentially saying this: “Sometimes we don’t know whether a country’s debt is too high. We need time to work that out. But if we’re lending, during that period, then while we’re deciding whether or not the country’s debt is sustainable, we’re going to force it to default on its private debt.”
There are two big worries at the IMF, according to Lipton. The first is that the IMF sometimes ends up lending money to sovereigns too late, after the country in question has already racked up an unsustainable quantity of debt. The second is that when the IMF does lend money to a troubled sovereign, that money just turns around and leaves the country entirely, in the form of coupon payments to foreign bondholders. In other words, the IMF doesn’t want to be the last chump lending, even as everybody else is using an IMF program as a path for smoothing their exit out of the country.
Hence the IMF’s latest idea, which is not (yet) a fully-fledged proposal, but which is intended to become one at some point. Basically, if you’re a country with a high and possibly unsustainable debt load, then new IMF loans are going to come with the mother of all strings attached: you’re going to have to stop paying back your existing private creditors in full and on time. Instead, you’re going to have to restructure those loans, somehow — term them out, force them to be rolled over — and in doing so you’re going to see all of the ratings agencies, and all of your CDS contracts, unambiguously declare that you have defaulted on your debts.
In the private sector, such financing is extremely rare: you don’t often see lenders to a company insisting that before they lend, the borrower has to stop paying back its current debts as they come due. There’s really only one situation where that ever happens, which is the case of debtor-in-possession financing after a company files for bankruptcy. So despite the fact that the IMF is adamant that it is not attempting — second time lucky — to create a sovereign bankruptcy regime, it does look as though the Fund is steering in that general direction.
Indeed, it looks like the Fund is aiming for a kind of muddled middle — one which is the worst of both possible worlds. On the one hand, it’s going to force countries to default on their debts; on the other hand, after coming out of default, the countries in question will still have debts which are far too high. The good thing about bankruptcy proceedings is that companies emerge from them in viable form, even if doing so means that creditors are forced to take substantial losses. But if all the IMF wants is “more time for the Fund to determine whether the problem is one of liquidity or solvency” — and if it then makes the determination that the country is actually insolvent after all — then the mild restructuring will have to be followed with a much harsher one. And if defaulting once is bad, defaulting twice in short succession is much, much worse.
The official sector understands this fact about markets, although it has never much liked it. Official-sector creditors, including the IMF and its shareholders, never mark to market: when they make loans, they always keep those loans on their book at par, until such a day as the loans are officially written off. The result is the classic series of Paris Club restructurings, where loans get rescheduled, coupons get reduced, and the present value of the debt continually declines — but not where it matters, on the books of the creditors. If you’re a sovereign creditor, rescheduling debts costs you almost nothing, and so it is indeed a great way of buying time for the debtor.
But the Paris Club, just like the IMF, doesn’t like the idea that its generosity will be taken to the bank by private-sector creditors. As a result, it insists on something called “comparability of treatment”: if a country is getting a good deal from the Paris Club, then it needs to negotiate something similar with its private-sector creditors at the same time. Which might be possible when the private-sector creditors are banks, which can sit down across a negotiating table and hammer something out — especially if the banks in question are able to continue to hold the loans on their books at par, on the grounds that they’re being held to maturity.
When the private-sector creditors are bondholders, however, everything changes. The cost of restructuring bonds — a very public default — is much higher than the cost of restructuring loans. And so there’s a constant tension between the official sector and the private sector when it comes to restructurings. The official sector likes to drag things out, in a series of deals; the private sector much prefers to do a single big deal — what’s known in the trade as “one and done”.
In other words, the noises coming out of the IMF aren’t particularly new: they’re pretty standard official-sector whines. What’s new is that it’s the IMF making them, after a decade of studious silence on the subject.
The Fund is treading much more cautiously this time around than it did ten years ago. It’s listening to just about everybody with an interest in the subject, including the wide range of views represented by its own board. (While creditor countries can see where the Fund is coming from, debtor countries are generally pretty solidly opposed, since they fear that any moves in this direction will only serve to drive up their borrowing costs.) Before my panel, for instance, the Fund hosted a discussion between 20-odd experts from the private sector and academia, with a wide range of views.
I was impressed by the certainty with which the various participants expressed their views, because I have no particular dog in this fight, and generally just find myself agreeing with whoever the last person to speak was. Virtually everybody there had an extremely strong case, despite the fact that many of them were deeply opposed to each other. I’ll try to give a flavor of the full range of views here, since it was hard to do that in the public panel. Obviously, many of these views are mutually inconsistent, but they’re all held by some important constituency somewhere — which should give you an idea of how hard the IMF’s job is, if it wants to achieve any kind of consensus on this issue.
The best place to start, though, if only because the views have names attached to them, is the Brookings report. The lead authors include Lee Buchheit, the dean of sovereign debt restructuring, and lawyer who has represented debtor nations around the world. Also on the lead-author list is Jeromin Zettelmeyer of the EBRD, although he’s at pains to point out that his involvement in the report was personal, and that the report does not necessarily reflect the views of the EBRD. More interestingly still, if you look down the list of other committee members involved in putting this report together, you’ll find the name of the biggest buy-sider of them all, Mohamed El-Erian. From what I can glean, he was not an active participant in the process. But the fact that his name is here at all is noteworthy, since there’s a lot in the report to antagonize the buy side in general.
The report unapologetically uses the b-word in its title, “Revisiting Sovereign Bankruptcy”, and is a pretty radical document, if by “radical” you mean the opposite of “market-friendly”. It says that there’s now “a much stronger case for an orderly sovereign bankruptcy regime today than ten years ago”, and lays out what is referred to by critics as the “Greece and Griesa” argument. Basically, the lesson of Greece — and of the euro crisis more generally — is that sovereign overindebtedness is a much bigger problem than anybody thought it was in 2003. As such, if some kind of sovereign bankruptcy regime helped to constrain the amount that countries could borrow on the private market, well, that might be a good thing, rather than — as people thought in 2003 — a self-evidently bad thing. On top of that, Thomas Griesa, a federal judge in New York, has created a huge precedent with his rulings about the way that the pari passu clause can and should be interpreted in bond documentation. In the wake of those rulings, bondholders are more likely to hold out, rather than tendering their bonds into an exchange. As a result, bond exchanges are going to be more and more difficult to execute successfully, and therefore there needs to be some kind of alternative way to restructure a nation’s bonded debt.
The Brookings paper essentially comes up with two separate sovereign bankruptcy regimes: one for the Eurozone, and one for everybody else. This makes sense to me: a one-size-fits-all solution is never going to be optimal with respect to addressing the very specific and idiosyncratic nature of European sovereign debt problems. In both cases, however, the Brookings paper proposes major new legislative changes: an amendment of the IMF articles, as well as an amendment of the ESM treaty in Europe. And in both cases the amended treaties would set an ex ante level for debt-to-GDP ratios, over which no new official-sector money would be lent without a debt restructuring taking place.
There are obvious problems with using a blunt instrument like the debt-to-GDP ratio to determine whether a debt restructuring is necessary. In fact, the Brookings paper goes in the exact opposite direction to the IMF, which is trying to make its debt sustainability analyses increasingly sophisticated. (For instance, if you look at the debt sustainability analysis on pages 43-45 of the Italy Article IV report, you’ll see a range of scenarios, with associated probabilities, rather than just a simple projection.) Obviously, the simple debt-to-GDP ratio excludes a huge number of very important factors: Japan, with its stratospheric debt-to-GDP ratio, is still vastly more creditworthy than Argentina, whose debt-to-GDP ratio is very low. And more generally, flows tend to matter more than stocks: ceteris paribus, the market will always prefer a country with a 100% debt-to-GDP ratio which is in primary balance, on the one hand, to a country with a 50% debt-to-GDP ratio and a primary deficit of 15% of GDP, on the other.
The Brookings paper argues that trying to put together a more complex threshhold, or even coming up with a bespoke number for every country on a case-by-case basis, is an approach which is certain to be gamed, and ultimately to fail. The point of this proposal is to tie the official sector’s hands — since one of the articles of faith underlying both the current IMF project and the Brookings paper is that the official sector has been too weak when it comes to saying no to countries needing funds. The official sector faces “a simple choice,” says the Brookings paper: “to adopt simple ex-ante rules that constrain discretion when structuring rescues, or to adopt no rules. Given the biases that currently exist against any form of debt restructuring, pure discretion would lead to bigger errors than simple rules.”
This is a highly contentious statement: Much of the buy side, for one, is genuinely convinced that the system isn’t broken, and shouldn’t be fixed. Putting in place a system where official-sector funding requires debt default, they say, plausibly enough, will only make matters worse: any time that a country even thinks about approaching the IMF or the ESM for help, markets will plunge, and bondholders will rush to be the first ones out the door.
Under this view, there are many countries — Brazil in 2002, Portugal and Italy more recently — which are clear success stories thanks to official-sector support, and which would have turned into debt-default failures had this kind of a scheme been in place. Official-sector types, on the other hand, are not entirely convinced that Portugal and Italy really are successes, and they reckon that even Brazil got very lucky with booming commodity prices and extraordinarily low global interest rates.
On the other hand, default isn’t what it used to be. Bond defaults used to be rare, just because sovereign bonds were rare: until the Brady plan in the 1990s, most sovereigns with any propensity to default would simply borrow money from banks rather than attempt to tap the bond market. Once bonds replaced loans, then bond defaults started. At first, they were shocking, but each new default makes them seem less so: statistically speaking, if you have a lot of sovereigns with a lot of debt, then some of them are going to default.
As a result, there’s a significant constituency — including quite a lot of academics looking at things like market access and bond spreads — that has a rather more nuanced view of default than the black-and-white idea that it’s always and everywhere a really bad idea, to be avoided at all costs. Under this view, if you look at successful bond restructurings in the past, then countries do indeed get punished quite harshly in terms of market access and bond spreads if they impose harsh and punitive haircuts on their creditors. On the other hand, if the restructuring is market-friendly and is really more of a reprofiling than an outright haircut, then the country often suffers quite minimal adverse consequences.
Those that see default in shades of gray like to draw distinctions and invent new words: rather than “default”, they say, can’t we use the word “treatment” instead? And in any case, don’t most bondholders, in reality, tend to agree on what needs to be done? After all, the elegance of bond markets is that when a country gets into trouble, the original holders of the debt are likely to sell it at a loss to people who can make a profit, even if (especially if) the country goes through a successful restructuring. Argentina is a particularly contentious outlier: in pretty much every other restructuring, holdouts really haven’t been an issue. And even Argentina can be seen as a success story in one respect: because its default happened so slowly, and was telegraphed so clearly, it had very little in the way of spill-over effects. It’s the sudden defaults, like Russia’s, which are the most dangerous: a clearly-foretold default allows the bonds to be held in large part by speculative investors who might actually make money in the event of a restructuring. As a result, a mild default like the one proposed by the IMF could do wonders in terms of minimizing contagion risks.
This argument doesn’t persuade the more hard-minded market types, who say that if holdouts haven’t been an issue up until now, that proves that the current system is working. What’s more, if the official sector starts bigfooting into the markets and telling countries to default, then at that point you can be sure that holdouts will be an issue — especially since the hedge funds who buy sovereign debt at low prices are exactly the hedge funds which are most likely to hold out. The official sector always seems to want the private sector to take a very large haircut, and restructurings have been successful to date only because the debtor and the markets have had the ability to push back against the IMF. (See, for example, Uruguay.)
Some participants go even further, and say that the current system is already too biased against bondholders. Sure, they’ve done OK for themselves — so far — in Portugal and Italy. But look at cases like Greece, Argentina, and Russia: in those countries, bondholders had to take enormous haircuts, in part because there was so much official-sector financing which insisted on preferred-creditor status. If a country has to pay off its preferred creditors in full, then any given necessary debt reduction is going to fall all the more heavily on the private sector. And if the IMF is going to come up with some kind of system for imposing losses on the private sector, that’s a clear sign that the Fund thinks that the private sector’s losses right now are too small.
Lipton sounds at best overoptimistic and at worst downright disingenuous when he says that “providing the member with a more comfortable debt profile would also have the additional benefit of enhancing market confidence in the feasibility of the member’s adjustment efforts, thereby reducing the risk that the debt will, in fact, become unsustainable.” No bondholders are ever going to actively welcome an IMF program which includes a bond default, even if that default in practice doesn’t involve much in the way of present-value losses. Lipton seems to believe that such a program could cause bond yields to go down; that seems like wishful thinking to me.
So, where do I stand on all this? I’d make three main points.
Firstly, this whole thing looks to me a bit too much like a solution in search of a problem. Hard cases make bad law, and when your entire reason for taking action can be boiled down to Greece-and-Griesa, two cases which are highly unique and idiosyncratic, then you’re always going to run the risk of overreach. The big story in emerging markets over the past decade is that their local debt markets have become much deeper and more liquid, meaning that they’re increasingly capable of funding themselves domestically. And domestic debt is not an issue: any time a government wants to default on its domestic debt it can, or alternatively it can just print money to cover the obligations.
The main exception to this rule is the eurozone, where governments do not have control over their own currency. Which is another way of saying that at heart this is really a European problem, rather than a global one, and that it should be solved at a European level, rather than at the level of the IMF. What makes sense for Ireland is unlikely to make sense for Brazil.
What’s more, in Europe, even more than in the rest of the world, debt restructuring is a very bad way of saving money overall. Sure, you can impose losses on sovereign bondholders — but those sovereign bondholders are very likely to be domestic banks. Which means that for every euro you save in government debt, you’re going to end up spending roughly one euro in bank recapitalization.
So rather than concentrate on bond exchanges, I think the Fund — along with Europe’s technocrats — should think much more about alternative ways of retiring outstanding debt at below face value. Ecuador did this very effectively in its most recent default: it simply used cash to buy up its own debt at a deep discount. And if you look at the amount of money that the official sector ploughed into the Greek exchange, it wouldn’t have taken much more to simply buy up most of the outstanding private debt on the public markets, and get rid of it that way.
I do understand that Greece has worried the IMF — and that Griesa has worried them even more. But it’s crazy to let a single vulture fund — Elliott Associates — effectively set the agenda for the design of the entire international financial architecture. Ten years ago, Anne Krueger was prompted to propose SDRM by Elliott’s shenanigans in Peru; today, Lipton is similarly motivated by Elliott’s successes against Argentina. (Not that Elliott has actually gotten paid, yet.) But if the IMF ends up moving in the direction it’s thinking about, the result could end up being counterproductive for everybody. So let’s think seriously about Europe, first. Only then, and only if a European solution proves obviously successful, should we start considering extending something similar to the rest of the world.