How do you restructure a contingent liability?

By Felix Salmon
July 12, 2013
Lee Buchheit and Mitu Gulati have another great paper out, on the subject of how on earth a sovereign's meant to restructure its contingent liabilities.

" data-share-img="" data-share="twitter,facebook,linkedin,reddit,google" data-share-count="true">

Lee Buchheit and Mitu Gulati have another great paper out on the subject of how on earth a sovereign is meant to restructure its contingent liabilities.

Here’s the problem, in a single chart:


What you’re looking at here is the way in which European countries especially have turned to sovereign guarantees as a way of masking the true size of their national debt. Many of these guarantees, once upon a time, would have simply been sovereign loans: the nation would have borrowed the money, and then lent it on, at a modest profit, to the borrower. But European countries want to do everything in their power not to borrow money right now. So instead of lending and borrowing, they simply guarantee a borrower’s debt instead. That brings down the price of debt for the borrower, but it doesn’t show up in any national debt-to-GDP statistics.

As you can see from the chart, this technique is increasingly popular — which means, in turn, that it’s going to have to be addressed in future debt restructurings. When sovereign guarantees were de minimis, they could be — and were — ignored. But at this point, they’re too big to be ignored. For one thing, sovereign bondholders won’t allow it. Let’s say Ruritania has $3 billion in bonds and $2 billion in guarantees. The bonds then get restructured, so they’re worth only $1 billion: a 66% haircut. But then, let’s say that half of the borrowers with sovereign guarantees go bust. Will Ruritania really pay out the $1 billion in full, with no haircut at all? It wouldn’t be fair to the original bondholders, that’s for sure.

As Buchheit and Gulati explain:

The restructurer’s dilemma is that contingent liabilities, if they are of any material size, cannot safely be left out of a sovereign debt restructuring, nor can they easily be included in a sovereign debt restructuring. This problem wasn’t a problem for so long as contingent liabilities represented only a small part of the debt stocks of affected countries. But for many countries, that period ended with the commencement of the financial crisis in 2008. The problem will therefore be unavoidable in at least some of the sovereign debt restructurings yet to come.

They do come up with one possible solution. If the guarantees are issued under domestic law, then the problem can be solved with legislation, along these lines:

All guarantees issued by the Republic of Ruritania in respect of debt obligations of third parties that are eligible to participate in the [Ruritanian restructuring] shall, if called by the beneficiary at any time after the closing of the [Ruritanian restructuring], be satisfied and discharged in full by delivery to the creditor of consideration equivalent to that offered in the [Ruritanian restructuring].

But the bond markets might have seen this one coming. After seeing the awesome power of domestic legislation in the Greek context, look what the market is now demanding in terms of sovereign guarantees:

While domestic-law guarantees were commonplace before 2010, that’s changing; at this point, foreign-law guarantees are catching up, and will almost certainly soon become an outright majority of the total.*

All of which leaves us with the paper’s conclusion — that contingent liabilities are going to have to be addressed somehow, in future restructurings, but no one knows how. If you thought that sovereign debt restructuring has been difficult until now, you ain’t seen nothing yet.

Update: I’ve removed a chart showing foreign-law guarantees making a huge proportion of recent bond issuance. That turns out not to be the case, necessarily. But I wouldn’t be surprised if we get there sooner rather than later.


We welcome comments that advance the story through relevant opinion, anecdotes, links and data. If you see a comment that you believe is irrelevant or inappropriate, you can flag it to our editors by using the report abuse links. Views expressed in the comments do not represent those of Reuters. For more information on our comment policy, see

I’d like to see more asset backed sovereign lending. Want to offer a general obligation T-bill… maybe the market will demand 4% for the 10-year term. Want a cheaper rate pledge something of value. Could be wireless spectrum, could be offshore drilling rights, fishing rights, water rights, you name it.

Today I (and a few thousand of my fellow drivers) wasted 45 minutes to drive around a municipal tractor which was mowing grass in the highway median. Some how the huge back tire had come off the rim and they were trying to jack it up in the break down lane. Two state police troopers blocked the slow lane leaving only one open travel lane. The value of time lost by the citizens driving north on interstate 295 (backed up 20 solid miles) exceeded the annual wages of the dullard responsible for driving that tractor. The obvious solution would have been to leave the tractor in the ditch, have the driver picked up, and then retrieve the disabled tractor during off peak driving time. The grass to be mowed will still be there tomorrow. After that glaring illustration of the total lack of understanding or respect for the value of the general public’s time, I’m ready to turn over the road to a private public partnership and relive the state of Maine of it’s burden.

Posted by y2kurtus | Report as abusive

Interesting idea, y2kurtus. Except I would suggest that the greatest risk in lending to the US is the value of the currency, not default. Backing a debt with assets wouldn’t change that much.

Posted by TFF | Report as abusive