Are Greek bonds pricing in a massive default?

By Felix Salmon
May 26, 2011
Martin Feldstein reckons that the market is pricing in a "massive" Greek default:

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

Martin Feldstein reckons that the market is pricing in a “massive” Greek default:

Even though the additional loans that Greece will soon receive from the European Union and the IMF carry low interest rates, the level of Greek debt will rise rapidly to unsustainable levels. That’s why market interest rates on privately held Greek bonds and prices for credit-default swaps indicate that a massive default is coming.

And a massive default, together with a very large sustained cut in the annual budget deficit, is, in fact, needed to restore Greek fiscal sustainability. More specifically, even if a default brings the country’s debt down to 60% of GDP, Greece would still have to reduce its annual budget deficit from the current 10% of GDP to about 3% if it is to prevent the debt ratio from rising again.

I don’t think this is true. The Greek yield curve is odd: 3-month T-bills yield about 6%, the 30-year bond yields about 10.8%, and the big spike is at 2 years out, where the yield is about 25%. Clearly the market isn’t pricing in any kind of massive default over the next three months. But let’s look at that benchmark 2-year bond, since if any instrument is pricing in a massive default it’s that one. The bond carries a coupon of 4.6%, and is trading at 71 cents on the dollar.

Now what would happen to that bond if there was a massive default? Let’s be conservative and say that Greece’s debt-to-GDP ratio will be about 150%, and let’s take Feldstein’s target ratio of 60% as where the country is going to end up. In that event, the face value of Greece’s debt would have to fall by at least 60%, and probably more, given the hit to GDP which normally accompanies a big default.

Clearly, traders pricing the 2-year note at 71 cents on the dollar are not pricing in any kind of event in which Greece will swap that note out for an instrument worth only 40 cents on the dollar.

In fact, any priced-in default looks decidedly modest to me. Let’s say that Greece defaults before the next coupon payment, which is due on May 20, 2012. And let’s say that traders in this risky asset want a return of at least 10% if there’s a default. Then someone buying the bond at 71 cents now is betting that it’s going to be worth at least 78 cents post-default. Which implies a pretty modest haircut of just 22% — something which would bring Greece’s debt-to-GDP ratio from 150% to a still-unsustainable 117%.

In fact, any priced-in haircut is even lower than that, since the post-default bonds aren’t going to trade at par.

My point here is that although yields on Greek debt are indeed high, they’re not anywhere near the really distressed levels that we’d expect to see if the market was expecting a massive and imminent default. If you buy a bond at 71 cents on the dollar, that’s cheap, to be sure, but there’s also an enormous amount of downside: if Greece does default in anything but the gentlest possible manner, then you’ll end up losing money.

The current price of Greek debt, then, says to me that traders are requiring hefty returns of 25% in order to pay them for taking the risk that Greece might default. They know that they’ll lose money if that happens, but they reckon that, more likely than not, Greece will muddle through — in which case they will make very good money.

The further you go out the curve, of course, the higher the default probability becomes. According to Thomson Reuters analytics, the CDS market is pricing in a 71% probability of a default in the next 5 years, and an 83% probability of a default in the next 10 years — both assuming a recovery rate of 41.5 cents on the dollar. But the Greek single-name CDS market is small and speculative; we have to be a bit careful about drawing too many conclusions from where it’s trading. If you want to protect yourself against default, you’re going to pay through the nose: this particular insurance market is very expensive. But if you’re buying Greek bonds at these levels, you’re not expecting a massive default. Quite the opposite.

More From Felix Salmon
Post Felix
The Piketty pessimist
The most expensive lottery ticket in the world
The problems of HFT, Joe Stiglitz edition
Private equity math, Nuveen edition
Five explanations for Greece’s bond yield
Comments
6 comments so far

This article and http://blogs.reuters.com/felix-salmon/20 11/05/25/why-clearxchange-is-great-for-p ayments/ don’t show up in your Facebook feeded. Technical problem or expected behavior?

Posted by S_2 | Report as abusive

Could you explain why 30-year yields are so low (relative to the 2-year), presumably if the chance of default is high then the 30-year note should have an even higher yield. Isn’t this the whole concept of a ‘normal’ yield curve?

Posted by chappy8 | Report as abusive

In 2001 Argentina, markets consistently underpredicted the risk and magnitude of default until the bitter end.

Some stories from back in the day sound eerily familiar.

June 13, 2001: “… The Argentine government is struggling to balance its books and there are fears that it could soon default on its $130 billion (£92 billion) debt. Earlier this week it only just managed to raise $800m. Yesterday, Domingo Cavallo, economic minister, launched a series of emergency measures aimed at stabilising the situation. The measures include an 8pc to 10pc pay cut for all government ministers and civil servants, as well as a clamp-down on those who fail to pay their taxes… Standard & Poors, the credit rating agency, cut Argentina’s sovereign debt rating to B-minus from single B and this caused international funds to sell the government’s bonds. This drove market interest rates up to 14.2pc, but they fell back to 13.6pc.”

August 29, 2001: “Argentina’s bonds rose after two days of declines on expectations U.S. Treasury officials may provide details about the country’s planned debt swap in a meeting today with some holders of Argentine debt. Argentina’s benchmark bond due 2005 rose 1.5 to an offer price of 76.81, according to J.P. Morgan Securities. That cut the yield to 22.5 percent.”

November 3, 2001: “The G7 group of major industrialised nations said it was “pleased” with the plan to lower the cost of debt. The International Monetary Fund has twice come to the aid of Argentina since December in a bid to stop the financial woes spreading across the region. A spokesman said a further bailout from the IMF was “not on the cards”. Argentina tied its currency to the dollar 10 years ago and insists it will not renege on this.”

December 11, 2001: “.. the country’s risk premium on bonds jumped amid growing concerns about a possible default on debt payments. The yield on Argentine bonds is now 42.06pc above that of comparable US Treasury bonds…. The Argentine government has little room to manoeuvre as it tries to implement austerity measures so that it can receive the latest tranche of IMF aid that would avert a default on its $132 billion debt later this month.”

By December 20 there were riots in the streets of Buenos Aires. In January, Argentina defaulted and dropped the dollar peg. Peso fell 70%.

Posted by Nameless | Report as abusive

chappy8: 10.8% on a 30-year is not low. Consider that, if you buy €1,000 of German 30-year bonds (for simplicity, treat them as zero coupon) at the current yield of 3.52%, in 30 years you will get €2,800. But if you buy €1,000 of Greek 30-year bonds at 10.8% and Greece never defaults on them, you will get €21,700. In other words, the market assumes that, at some point down the road, Greek 30-year will get a haircut that makes it lose almost 90% of its value.

Posted by Nameless | Report as abusive

Actually I take that back. You can’t treat them as zero coupon, results are badly off. Instead let’s do it right.

Assume that the German 30-year is perfectly safe. Therefore we can use 3.52% as the discount rate to calculate its value. Net present value of future cash flows from €1,000 worth of German 30-year bonds is exactly €1,000 (a tautology).

At the same discount rate, net present value of future cash flows from €1,000 worth of Greek 30-year bonds (assuming no default) comes out to €2,335.

It follows that, in agreement with Thomson Reuters, the market prices something like the near-certainty of a default/restructuring on the scale of 40 cents on the dollar.

Posted by Nameless | Report as abusive

Math, math, math.

Take the 30-year Greek bond. Compound it out 30 years. Take the 30-year bund. Do the same. Divide the expected Greek total return by the expected German total return. Then divide by 30.

You will see that the market is discounting a 30% annual loss likelihood over the next 30 years. That is, if Greek-like debt loses money 30% of the time, a rational investor is indifferent between the two. Note: this is a LOSS probability. If you assume a 50% recovery, then the probability of default is 60%.

Through the magic of compounding, a lower 30-year yield is required to discount this same likelihood than is necessary with a 2-year bond. But the probability is similar.

Posted by Publius | Report as abusive
Post Your Comment

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 http://blogs.reuters.com/fulldisclosure/2010/09/27/toward-a-more-thoughtful-conversation-on-stories/