Monday, February 15, 2010

So Mr. Finance Minister, I'm here for the bond auction but first, say hello to my little friend, I call him the Credit Default Swap

I am inclined to favour financial innovation. Since the financial crisis began an innovation that has come under attack is the credit default swap (CDS.) Usually I rise to the defence of the CDS and attack its detractors as people who do not understand how valuable the product is. Recently however, in the hands of the “Bond Vigilantes,” the traders in government securities who decide at what rates governments can borrow, the CDS is an extremely powerful weapon.

First of all what is a CDS? It is a side bet between two parties on the creditworthiness of a third party. Imagine that you think the markets are underestimating the chances of a Nakheel default, you and I could trade a CDS on Nakheel. Here’s how it would work. First we have to agree on how much we’ll bet and for how long. Let’s call it $1 million and let’s say we bet for five years. This means that if Nakheel defaults at any time in the next five years I have to pay you the amount of money that someone who lent $1 million to Nakheel would lose. Let’s say for example that Nakheel went bankrupt and then the bondholders sold off the Nakheel assets and had enough to pay the bondholders $0.03 on the dollar or $30,000 for every million lent. If that happened then I would owe you .97 times our million or $970,000.

Sadly there’s no such thing as a free lunch so I have to charge you for this. Let’s say I decide you have to pay me 5%. What this means is that every year for the next five years you have to pay me 5% of the million dollars we agreed or $50,000, but if Nakheel does indeed default you can stop making the payments. So, you have to pay me $50,000 a year for five years or until Nakheel defaults and if it does I have you pay you something around $900,000 depending on how much the creditors can get for the balance sheet. Sound like a good deal? It sure as hell is. Nakheel CDS would be a lot more than 5%, the Dubai government is trading at 6.5% now.

For the past 20 years the CDS market has been the fastest growing derivative product in the world. There are literally trillions dollars of these things outstanding. Why CDS are so popular? CDS’s are a useful tool for banks to manage their risks. Have you noticed how the bet is kind of like a loan? The person buying the credit protection is paying out coupons every year in return for his insurance. Let’s imagine that you are a Nakheel creditor and you lent Nakheel $1 million at 5% interest. If you traded this same credit default swap with me you now have no risk and all. Every year Nakheel pays you your $50,000 and you pay it to me. If Nakheel defaults, you recover what you can from Nakheel and I pay you the rest. Notice also that while you had to actually come up with the $1 million that you lent to Nakheel, in our CDS transaction there is no exchange of principal, there’s only a payout in event of default. This makes it a lot easier to trade CDS than to actually make loans, especially if the capital treatment of CDS is different from that of loans.

As a result, CDS have helped make credit much more accessible across the world for two reasons. First, a bank has made a loan and then hedged it with a CDS it frees up the balance sheet and then it make new loans. Second, they made it possible for a lot of people to provide credit who did not have the infrastructure of banks. As you can see from the example above if you sell a Credit Default Swap you have the same risk profile as if you had actually lent the money except that you don’t have to actually put up the money or go through the hassle of becoming a bank. If you buy a CDS you have the same risk profile as if you were the borrower. Thus people who are not in the banking business can use CDS to produce the same risks and returns as if they were.

Sadly it’s not all sweetness and light. There have been two big issues with the CDS market. The first issue was systemic. Because the entire financial system was using CDS to cross-insure its risks it became very difficult to understand what would happen in the event of the collapse of a major CDS issuer. They had the effect of blurring the balance sheets of all the financial institutions together. For example AIG insured a truly epic number of loans. So many that had it collapsed it would have constricted the balance sheets of almost all the banks in the world and caused another wave of Lehman style collapses. As a result the government was forced to rescue AIG.

The other criticism of CDS trades is that they are used as a weapon by speculators seeking to attack otherwise solvent companies. The theory is that buying insurance against default on loans you haven’t made is like buying insurance on a house you don’t own and then lighting it on fire to collect the insurance money. This argument was made with regard to Lehman Brothers. I disagreed with that line of argument for the same reason that the truth is a valid defence in a libel case. The truth is that the Lehman bondholders got $.06 on the dollar meaning there was a hole in the balance sheet in the hundreds of billions. People buying Lehman CDS did not cause the balance sheet hole, Lehman’s management did. So this is where the fire insurance analogy breaks down, it wasn’t possible for the CDS buyers to create the carnage at Lehman, only Lehman’s management could do that. I’d say the analogy is hurricane insurance. Even if you buy hurricane insurance on someone else’s house, you can’t then cause a hurricane unless you’re God.

Recently however events in Dubai and to a much greater extent in Greece have been making me question the CDS as a product. As mentioned, Greece has to roll $75 billion of debt in 2010. This will be done in a series of auctions of new bonds the proceeds of which will go to repay old bonds and to finance the large deficit that Greece is currently running. What usually happens is the world’s bond investors decide what interest rate they need to be paid in order to justify the risk of default and then vote with their dollars. If they stand back from the auction or demand higher rates the government has to pay higher rates. If they stand back altogether the country is forced into default. You could think of it as there is a certain amount of demand in the world for Greek debt at each level of interest rates and at every auction the Greeks and the bond holders find out what level of interest rates match the demand of investors with the amount of debt the Greeks need to issue.

Now enter the CDS. Remember as we said above that a CDS enables anyone to replicate the returns of a loan. This is true for sovereign loans as well. So not only can people who are bearish on Greece refuse to participate in the auction they can front run the Greeks themselves and soak up the demand for the Greek bonds by buying CDS protection. If you sell me a CDS I pay you the coupons that you would receive from Greece and you only lose money if Greece defaults. This is exactly the same position you would be in if you actually lent the money to Greece. Except in our trade the Greeks don’t get any money at all! In essence I have disguised myself as Greece and absorbed some of the markets’ willingness to lend to Greece.

How important is this? To be sure the real issues are the fundamentals of Greece, can they fix their deficit problem? Will the EU bail them out? But the CDS can become a significant factor in its own right. It really depends on how many people in the world are bearish Greece. Imagine that there are people willing to bet $75 billion that Greece will default. That would have the effect on the market of creating another Greece or effectively doubling the supply of Greek debt and thereby pushing up the price at which the Greeks borrow. By buying CDS on Greece the bond vigilantes absorb some of the interest in lending to Greece and thereby actually push up the cost borrowing for Greece. If the Greeks have to pay more to borrow, it actually widens the Greek deficit and helps to bring about the outcome on which they are betting. So they can cause the hurricane by buying hurricane insurance, or as they see it, they can play God.


Anonymous said...

fantastic analogy... and as always, very well explained...

Anonymous said...

Here is another analogy.

Town A and Town D both are on banks of a big river and have built huge dams on the river

Town A is upstream and Town D is downstream on the river.

There are a bunch of guys running around in Town D saying that a flood is coming. This naturally causes headaches for Town D officials as they have no way of knowing if there is actually a flood coming or those rumor mongers are investment bankers in disguise trying to create a panic and sell more boats and CDSs.

Town D officials royally and rightfully ignore the rumors and build more racing tracks, buy some trinkets and get busy in building the next biggest 'whatever'. In any case they have the best consultants to advise them on risk management!!

Town A officials hear the rumor and think, "If Town D is going to get flooded then chances are that Town A is also going to get flooded". Why take a risk, It would be good 'risk management' practice to release a bit of water from the big dam they control.

So they release the water and hey Town D actually gets flooded.

Who is the Abi Daddy now!!

Anonymous said...

Normally when you buy a ton of puts on a company's equity, you can drive the stock down quite a bit, but eventually it will recover when the dealers take their delta hedges off. You buy a boatload of CDS protection against a company with a funding gap, even a year or two out, and you can put them out of business as their trade creditors get spooked. It's like the really bad market practice of when PIPE buyers hedge their deals before they're even priced, knowing the company is trying to raise equity. Large CDS players can be the functional equivalent of insiders without needing to sign documents or disclose anything to anyone. I think the sunlight of position disclosure (making CDS buyers vulnerable to similar sharks when they over-position) is the best way to deal with this.

sigismondo said...

Dear Ken,

From a systemic risk perspective, is it not enough to regulate that only banks can issue CDS and only against a provision on the principal (the 1 million) for the duration of the contract on their balance sheet?

Best regards,

Ken said...


That is a popular theory but I think it would actually destroy CDS as a useful product. The suggestion you have, that only banks would be allowed to issue CDS and only against owned liabilites of the companies in question would have two effects.

First it would massively constrict the supply of CDS in the world and as a result it would tighten credit more generally. Second it would make the CDS market highly illiquid. Only banks could sell them, and only people with extant risk could buy them and would have to sell them once the risks were gone. This would by necessity make the CDS market smaller than the actual debts outstanding because only a small fraction of lenders would seek to hedge at any given moment.

By limiting both the supply and demand the bid-offer spread would necessarily increase because any supplier would in essence be committed to the CDS for the duration of the buyers holding period of the reference asset. This means that they could not effectively control their risk and would therefore have to charge a higher premium for it.

One of the major benefits of the CDS is that they trade so much more freely than bonds. Think of it this way: the CDS is really an aggregated bond market for the reference asset. Imagine a company has issued $10 million 1 year, $10, million 3 year, $10 million 5 year, and $10 million 10 year bonds. A large fraction of bond buyers are hold to mataurity investors. This means that only a small proportion, say $2 million of each issue actually trade. The illiquidity premium people charge for this is substantial.

Now, enter the CDS, in the event of default any of the 4 issues would be deliverable against the CDS so it has the effect of aggregating all the issued debt of the company. So now even if you only count the trading proportion of the bond issues, the avaliable notional is $8 million or 400% the original, this substantially diminishes the liquidity premium. But also anyone can trade them so Perhaps the real supply of CDS might be more like $10-$12 million.

This has imporant follow on effects as people begin to question the credit worthiness of the borrower. It is a lot easier for people who are supportive of the company to sell protection via CDS than to scare up a seller of the actual bond because the market is centralized and aggregated. This has the effect of lowering the ultimate borrowing costs by lowering the transaction costs associated with making credit avaliable.

If you impose the limits on both the supply and demand of CDS you eliminate the relief they bring to the liquidity premium people charge in the bond market. It's throwing the baby out with the bathwater.


sigismondo said...


first of all thanks for taking the time to respond.

Secondly, I am ignorant and headstrong, a bad combination :)

Thirdly, every time that I discuss CDS i look out of the window to check if the city is still standing. A nice city, btw: I live next to grosvenor house, if that tells you anything :)

So, being ignorant, my main points are very simple:

1- financial institutions are required to provide backing on their balance sheets for making loans or providing insurances. I don't know the exact form or shape of these provisions, and over the years I have lost a lot of faith on the competence and relevance of the government regulators, but notwithstanding these requirements are there for some VERY GOOD reasons, right? Please please please say 'yes' :) I do not like a product that essentially allows me, the incompetent and irresponsible sigismondo, to issue you, the competent but equally irresponsible Ken, an insurance without backing. Is like giving an atomic bomb to

2- what scared me of the crisis was not that people lost money and jobs and houses or that banks went bankrupt or even that they were bailed out with taxpayers money. What scares me to death is that nobody knows the exposure of anybody else anymore. Since two (2) years. This is the dissolution of any system of trust. Armageddon. Halas. If to correct this I only need to kill one particular financial product, I will do it in a split second.
Finally: Do you see a way to throw the water without throwing the baby?

Best regards