CDS demonization watch, Greece edition

By Felix Salmon
February 25, 2010
this one, headlined "Banks Bet Greece Defaults on Debt They Helped Hide". It's gaining a lot of traction: Ben Bernanke said today that he's looking into the issue of whether the CDS market is enabling some kind of run on the Greek government. I sincerely hope he was jut being polite to his Congressional overlords, rather than buying in to this theory.

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

My CDS Demonization Watch has been on the back burner for a while: I thought that the caravan had moved on. But right now, the most-read story in the NYT business section, getting a lot of attention in the Twittersphere, is this one, headlined “Banks Bet Greece Defaults on Debt They Helped Hide”. It’s gaining a lot of traction: Ben Bernanke said today that he’s looking into the issue of whether the CDS market is enabling some kind of run on the Greek government. I sincerely hope he was just being polite to his Congressional overlords, rather than buying in to this theory.

It’s worth looking at the NYT story in some detail, to see just how little sense it makes.

Bets by some of the same banks that helped Greece shroud its mounting debts may actually now be pushing the nation closer to the brink of financial ruin.

Echoing the kind of trades that nearly toppled the American International Group, the increasingly popular insurance against the risk of a Greek default is making it harder for Athens to raise the money it needs to pay its bills, according to traders and money managers.

The first thing worth noting here is that Greece is not anywhere near the brink of financial ruin. The CDS market is actually very good at showing when a borrower is near financial ruin: when that happens, spreads gap out past 1,000bp to something closer to 2,000bp or even 3,000bp. Greece’s CDS spreads peaked at about 400bp, which is high for an EU country, but is nowhere near distressed levels. Yes, every time the CDS spread rises it gets closer to distress, but that’s just as true — and just as unhelpful — if it goes from 30bp to 40bp.

The second point to note about these opening two paragraphs is the curious presence of AIG. AIG went bust because it wrote insurance; the NYT story is here implying that there’s some relevance to what’s happening with Greece, a reference credit that people are writing insurance on. AIG had to pay out billions of dollars to make good on CDS contracts; Greece has neither bought nor sold any CDS contracts at all. No sooner are the parallels made than they break down.

That doesn’t stop the NYT, however, which then proceeds to wheel out the cliché about CDS being “like buying fire insurance on your neighbor’s house”. The problem is that the analogy just doesn’t work in this case. Much later on in the article, after most people have stopped reading it, we’re told the truth of the matter:

European banks including the Swiss giants Credit Suisse and UBS, France’s Société Générale and BNP Paribas and Deutsche Bank of Germany have been among the heaviest buyers of swaps insurance, according to traders and bankers who asked for anonymity because they were not authorized to comment publicly.

That is because those countries are the most exposed. French banks hold $75.4 billion worth of Greek debt, followed by Swiss institutions, at $64 billion, according to the Bank for International Settlements. German banks’ exposure stands at $43.2 billion.

These banks aren’t buying insurance on someone else’s house, they’re buying insurance on their own house. As the old saying goes, if you owe $75,000 to the bank, you’ve got a problem. If you owe $75 billion to the bank, the bank has a problem. And in this case, the banks are doing their best to deal with that problem and manage their risk proactively.

The question here is whether their ability to do so in the CDS market is exacerbating matters for Greece. The mechanism here is complex, if it exists at all:

As banks and others rush into these swaps, the cost of insuring Greece’s debt rises. Alarmed by that bearish signal, bond investors then shun Greek bonds, making it harder for the country to borrow. That, in turn, adds to the anxiety — and the whole thing starts over again.

At the very least, this does a large disservice to bond investors. They’re not sheep who are happy to lend to any country unless or until its CDS spreads widen — in fact, at the margin, they’re more likely to lend to a country if they know that they’ll always be able to hedge that position in a liquid CDS market. Now, it’s true that as worries over Greece’s creditworthiness get more intense, Greece’s cost of funds goes up. But there’s a strong case to be made that absent the CDS market, Greece simply couldn’t borrow at all: the existence of the CDS market has made it easier (if more expensive) for Greece to borrow money, not harder.

There is a connection between the CDS market and the cash bond market, thanks to the concept of delta hedging — people who sell credit protection on Greece will often end up selling Greek bonds at some point in order to manage their exposure. But that connection is much more tenuous than the alternative, which is that of banks looking to reduce their Greece exposure all dumping their Greek bonds onto the market at the same time. The result of that kind of operation would be spreads much wider than 400bp.

The weirdest bit of all in the NYT article is the way it places the blame not on people trading Greek CDS, but rather on people trading a more general sovereign CDS index:

Last September, the company, the Markit Group of London, introduced the iTraxx SovX Western Europe index, which is based on such swaps and let traders gamble on Greece shortly before the crisis. Such derivatives have assumed an outsize role in Europe’s debt crisis, as traders focus on their daily gyrations.

If you want to gamble on Greece, you can gamble on Greece: gambling on a broader Western Europe index makes little sense. What’s more, insofar as people are trading the iTraxx index, they are very unlikely to delta-hedge in the bond market — they’re just taking positions for a few hours or days, trading in and out. Blaming the iTraxx index for Greece’s problems makes no more sense than blaming the ABX index for the subprime crisis: it’s a symptom, not a cause.

But by far the worst part of the NYT piece is its headline: at no point in the article does it come close to making the case that banks in general, or Goldman Sachs in particular, are betting on a Greek default. At worst, banks are hedging their large exposure to Greece and other PIGS nations using an index they helped to create. But the fact is that Greece’s financing burden — the title of the NYT webpage is “Trades in Greek Debt Add to Country’s Financing Burden” — is entirely its own making. Blaming the banks here makes no sense at all.


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

More generally, anyone who can write a CDS who was otherwise inclined to lend to Greece will need a higher interest rate to be convinced to lend. There are limits to arbitrage here, but they’re not huge right now, in my (poorly informed) estimation.

I’d like to highlight the distinction between
- the existence of a CDS market is hurting Greece
- the widening of Greece’s spreads in the CDS market is hurting Greece;
in particular I’d compare this to
- international trade hurts the US
- a rise in the price of imports on international markets hurts the US.
In both cases, the latter statement is probably true, while the former is probably false; the existence of the market in each case does the most good for the buyer when the prices are low, and what you’re looking at is not a change to where the market is imposing a net cost so much as a reduction in the amount of benefit from what it was able to provide under other circumstances.

Posted by dWj | Report as abusive

Greek 1-year CDS recently hit +700. you are looking at the 5-year, which is not as useful for indicating investors’ perception of the likelihood of an imminent default.

you might want to consider who -writes- PIIGS sovereign CDS protection. considerable risk there which no one is talking about.

you also might want to consider that in the event of a Greek default GS is likely to receive payouts on both its Greek CDS and the “balloon payment” on its currency swap with Greece. it would also get to stop making payments to the CDS protection writer. forget whether GS is “betting” on a Greek default (although I would not be surprised if they are); what would GS prefer to see happen?

CDS create perverse incentives. this is a problem.

Posted by N.Mycroft | Report as abusive

You have graph of cds spreads here in america, between 2007 and nowish?

Posted by Uncle_Billy | Report as abusive

nice read. can we have more commentary on articles like this. great to have a different perspective.

Posted by savo | Report as abusive

I swear, I wrote a nearly identical post last night (which I obviously never got around to finishing today). We clearly read the NYT article the same way, because I had made almost all the same points that you make — e.g., how tired this particular CDS conspiracy is; the fact that it assumes that bond investors are incredibly stupid/naïve; that there’s no actual mechanism by which rising CDS spreads cause bond yields to rise; etc. It’s really uncanny. Great minds think alike!

Check out the Chart of the Day on Bloomberg too, which shows that the European sovereign CDS market is tiny compared to the overall European sovereign bond market (i.e., less than 1% the size). The idea that the CDS market is somehow driving yields on Greek debt is really beyond ridiculous.

Posted by EconOfContempt | Report as abusive

There are basically 3 channels through which the angst of investors about the creditworthiness of the country concerned expressed: (1) The credit quality notes of Rating agencies for countries, (2) the yield differentials (spreads) in the bond market, and (3) the credit default swaps.

We should therefore not be surprised if all these indicators show in the same direction. But one big difference is that in the case of CDS we have “off-market” values. And the CDS has no disclosure obligations. Meaning that derivatives are used extensively to circumvent investment restrictions, tax legislations etc, as Satyajit Das remarked recently in a newspaper essay. Therefore the derivatives market is now larger than the base market. The speculative market contributes to increase risk premia.

Posted by CEZMI-DISPINAR | Report as abusive

If as a credit default swap demonizer I join such illustrious company as George Soros, Charlie Munger and Paul Volcker, all of whom have strongly ‘demonized’ credit default swaps, then I am deeply honored by the title. Soros and Munger both called for an outright ban.

You too should be a CDS demonizer, Felix. It is the moral high ground.

The ultimate problem with credit default swaps, to use your own words, Felix, is that they can be end-of-world insurance. For the issuer, profits come now, and in the unlikely event that default actually occurs, well who cares, that is the end of the world anyway.

Writing credit default insurance without the ability to pay is fraud, pure and simple. It happened on a massive scale at AIG and there have been no convictions.

End-of-world insurance in the forms of CDSs on governments and government-backstopped institutions is a license to print money.

We are in the biggest debt bubble in history, particularly with respect to government debt. We are not getting proper market signals regarding the scale of the problem in the CDS market or the bond market. Yes they move in tandem, probably due to arbitrage, but neither properly accounts for default risk. If issuers were required to reserve appropriately and actually account for the end-of-world you speak of, things would reprice dramatically and become a market rather than a fraud.

Instead, under the cover of default insurance, sovereign debt towers to greater heights, putting off and worsening the ultimate reckoning.

Posted by DanHess | Report as abusive

“You too should be a CDS demonizer, Felix.”

“to use your own words, Felix.”

Yes, this is probably all very true, but you’re coming off like an axe-murderer staring into Felix’s eyes.

[insert brilliant analysis of the Libyan bond market here]

Posted by Uncle_Billy | Report as abusive

DanHess does not come off as an axe murderer to me. He’s making a lot of sense. We all saw how AIG was bailed out and paid its counterparties with government funds, funds it did not have and in an honest market would never have been able to insure. Can you answer that instead of calling people axe murderers?

Posted by rdubeau | Report as abusive

They did not have the funds because of the collateral arrangements in the CDS contracts. In terms of likely defaults, they probably did have the funds.

Posted by JCH1952 | Report as abusive

JCH1952, what an odd thing to say!

“They did not have the funds because of the collateral arrangements in the CDS contracts. In terms of likely defaults, they probably did have the funds.”

JCH1952, AIG still owes the United States a bit more than $100 billion dollars in bailouts, including $70 billion in loans and interest, and $35 billion in relation to collateral the government accepted that proved to have a far lower market value.

If they have money somewhere, why don’t they pay it back? The answer of course is that they are broke. They are selling off their crown jewels and still can only pay a fraction of what they owe.

You are right that the collateral arrangements created an immediate liquidity crisis, but we have seen that many of those mortgage-backed securities really were worthless. Goldman and others wanted to be paid immediately because they saw or suspected that behind the facade AIG was insolvent, and they were right. They wanted to jump to the head of the line, before bankruptcy reduced their claim.

Indeed the default rate on subprime adjustable MBS has been running at close to 40% and rising! So these weren’t just paper losses! These MBSs have really been defaulting massively. Even prime fixed mortgages are defaulting massively.

Posted by DanHess | Report as abusive