The CDS market and Greece’s default

By Felix Salmon
July 22, 2011

ISDA has made the right decision: the Greek bond default does not and should not count as a “credit event” for the purposes of whether Greek credit default swaps will get triggered.

This is the right decision for two reasons. Firstly, the swap is voluntary. If you don’t want to suffer a haircut, or see your six-month maturity suddenly become a 30-year maturity, then all you have to do is nothing. If the CDS paid out, that would be tantamount to giving free money to anybody who wanted it: just buy short-dated Greek debt and also credit protection on that debt. The bonds will pay out in full, and the CDS would pay out as well.*

Secondly, this should be a large nail in the coffin of the CDS market generally. Credit default swaps were designed primarily for banks: it took many years before they became widely-traded speculative instruments in their own right. The idea behind them was that banks could keep loans on their balance sheets while at the same time hedging the risk that they would default. That was easier and cheaper than selling the loans outright, and also helped banks maintain good relations with their borrowers.

In the case of Greece, however, banks are going to take a 21% haircut on their loans to the country — and if they hedged themselves in the CDS market, too bad. Whatever they paid for their CDS protection — and if they bought it recently, it could be quite a lot of money — will not help them one bit. If buying CDS doesn’t help you in the event of a default, then there’s really no point in buying CDS.

So even though there isn’t going to be a formal credit event in Greece, have no doubt that this is a default — or, at least, that it will be a default if the banks are correct in anticipating a participation rate of 90%. But I have to say I’m surprised they’re so bullish, and I will be extremely impressed if Greece wakes up one morning to find that 90% of its private-sector debt has been tendered into the exchange.

For one thing, the exchange is structured in a pretty unsophisticated way. No matter whether you have a bond maturing tomorrow or a bond maturing in a decade, you’re swapping it into exactly the same 30-year instrument. Banks with short-dated Greek debt aren’t going to be happy about this, and will have a strong incentive to quietly sell that debt on the secondary market to someone willing to just hold it to maturity.

And more generally, it’s almost impossible to see why anybody who isn’t a bank would tender into this exchange. If you’re an individual holding Greek bonds, or a big bond investor like Pimco, the obvious thing to do is to hold on to your debt for the time being, since the exchange includes no carrots and no sticks. The only incentive for you to tender into the exchange is that the new bonds will be partially collateralized with zero-coupon 30-year bonds. Triple-A-rated 30-year bonds in euros currently yield 4%, which means that the collateral in the new Greek instruments will be worth less than 31 cents on the dollar. And so far, there’s been zero indication that holdouts will get anything less than all their money back, in full and on time; they might have a bit less liquidity, but bond investors are used to illiquid instruments.

Is it really the case that over 90% of Greek bonds are held by banks which will tender all of their Greek debt into the exchange? Maybe so; non-bank investors looking to go long Greek credit might well have found it easier and cheaper to do so by writing credit protection in the CDS market rather than buying cash bonds. In which case they’re smiling broadly today.

If that’s the case, then maybe this deal is one of those rare occasions where the CDS market was genuinely useful. Back in the 1980s, when sovereign debt was held overwhelmingly by banks, negotiations about restructuring that debt could be held with the London Club of bank creditors. Non-bank investors didn’t really matter. But then the loan market became a bond market, and investors in sovereign debt were mainly non-banks; no longer were negotiations even really possible.

Now, however, we seem to have come full circle: banks are able to put together bond-restructuring deals on their own, without worrying much about non-bank bond investors. And one reason would seem to be that the non-bank investors have largely moved from the bond market to the CDS market.

Is this scheme going to work? There’s a big collective-action problem at the banks, and an even bigger problem with the non-banks. So the 90% target is ambitious. But so far I haven’t seen much doubt. I will say this: if the 90% target is achieved, then the Institute for International Finance will deserve a lot of credit. I’m no fan of the organization — I’ve been very rude about it for many years. But if it manages to pull this off, it will finally and genuinely have justified its existence.

*Update: Kid Dynamite asks a good question in the comments: if the bonds paid out in full, wouldn’t the CDS auction clear at par, with the CDS paying out nothing? No. To a first approximation, the CDS auction clears at the price of the cheapest-to-deliver Greek bond, and Greek bonds are trading at a substantial discount. The cheapest-to-deliver bond will probably have a very long maturity, and if there was a credit event for CDS purposes, then holders of credit default swaps would get a nice check in the mail.


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


you wrote “The bonds will pay out in full, and the CDS would pay out as well.”

correct me if I’m wrong, but if the bonds pay out in full, then don’t the CDS pay out nothing? (oversimplified, of course)

Posted by KidDynamite | Report as abusive

“If the CDS paid out, that would be tantamount to giving free money to anybody who wanted it: just buy short-dated Greek debt and also credit protection on that debt.”

@KidDynamite and Felix: Buying a bond plus CDS should be risk free, modulus some counterparty risk. I would say that if the CDS paid out in this case, then the CDS would be pointless.

Posted by guanix | Report as abusive

Perhaps I am confused, but when exercising the CDS, doesn’t the purchaser of the CDS have to surrender the bond to the CDS seller?

To confuse matters more, it is unclear who would get paid. Would not the swap be required to trigger the CDS?
If the original bond is swapped for a new discount bond, would not the original bond be then returned to the CDS seller. But the owner of the original bond is now Greece (or the EFSF??). Does it not follow that Greece would get the CDS payment in this swap?

Posted by Kosta0101 | Report as abusive

If it is voluntary, and those who choose not to participate will be paid in full, why will any bank participate? Shouldn’t the management of any bank who voluntarily takes an unnecessary loss like this deserve to be fired?

Posted by MattJ | Report as abusive

@KidDynamite, KostaD101, yes, the most common style of CDS settles by exchanging one of a number of deliverable bonds for a fixed amount of cash. (So there is an embedded cheapest-to-deliver option.) But this is far from universal; digital (aka recovery) CDS just pay out a fixed amount of cash.

Posted by Greycap | Report as abusive

@Greycap – thanks – do we have any way of knowing what kind of CDS was written on Greece? Digital CDS sound insane… ie, a “default” event could be triggered, resulting in a binary payout on the CDS, and the bonds still get full recovery, right?

Posted by KidDynamite | Report as abusive

I’m sorry, I still don’t understand. If a bond holder doesn’t tender but has CDS coverage, then either the bond is paid in full or CDS makes up the loss. Isn’t that right? Why, then, are the CDS sellers off the hook?

Posted by RussAbbott | Report as abusive

This won’t mean that the CDS market will go away for sovereign debt. It just means that the restructuring language will change to include events like this, should enough of the market care about this.

Posted by DavidMerkel | Report as abusive

Felix, you wrote:

“banks are going to take a 21% haircut on their loans to the country — and if they hedged themselves in the CDS market, too bad. Whatever they paid for their CDS protection — and if they bought it recently, it could be quite a lot of money — will not help them one bit.”

This is not correct. Those who hedged themselves in the CDS market can just decline to participate, and keep holding on to their bonds. Either these will pay out in full or there will be a credit event in the future, which will trigger CDS payouts. In fact, I see no reason why any bondholder who bought protection will participate voluntarily.

Those who did not hedge face default risk if they decline to participate in the exchange. This risk could be significant, depending on the overall rate of participation.

Posted by rajivsethi | Report as abusive

I agree with rajivsethi. If you hold Greek bonds and are hedged with CDS, there is absolutely no reason for you to participate. You would be taking the haircut for no reason. Either the bonds pay off or they do not (in which case the CDS do). There is no third option whereby the bonds pay off for those who participate and don’t for those that do not, which is the scenario some seem to be envisioning for some inexplicable reason.

Posted by RueTheDay | Report as abusive

It’s shocking that such a large percentage of outstanding bonds are expected to take a haircut without being forced into it. Even more so considering that this round of help still won’t be enough to right the ship.

After this new money is spent and another round of budget cuts the Greek budget will still be out of balance for the next budget cycle. If they want to borrow money in less than 2 years time than they can’t be in default at all or they won’t have access to the debt markets.

I can see centeral banks offering to take a haircut but it’s astounding that large publicly trade stock banks who are already short on capital would agree to this deal. I can think of some U.S. states that would love to get these terms from the Federal goverment!

Posted by y2kurtus | Report as abusive

If the bond hasn’t experienced a credit event then the CDS won’t pay out. Life would of course be a lot simpler if a CDS paid out both in the event of a credit event or in the absence of one.

RueTheDay, depends who your counterparty is…. If they were writing alot of protection on Greek debt i wouldn’t party till the cheque clears.

Posted by Danny_Black | Report as abusive

ok maybe i misinterpreted “The bonds will pay out in full, and the CDS would pay out as well”. Are you suggesting that if ISDA had declared this a default then you could buy short dated debt and protection, immediately collect on the CDS and hopefully get a payment in a couple of weeks? This would of course depend on you finding someone stupid enough to write a protection on an event that has already happened.

Posted by Danny_Black | Report as abusive

“banks are going to take a 21% haircut on their loans to the country — and if they hedged themselves in the CDS market, too bad”
CDS may (and shall) not be triggered, but I don’t think banks are writing down hedged bonds — unless they voluntarily participate of course.
Too early to say, though.

Posted by vastunghia | Report as abusive