Monday, December 8, 2008

"It’s about one of the nice side benefits of CDS: the habit they have of pointing out who is going to get into trouble next. "

I want to keep this post and comments for my own reference, so I'm going to enter it here. It's about the use of CDSs to signal trouble. It should not come as any surprise that an insurance policy, meant to cover default or foreclosure, would be sensitive to the likelihood of its having to be paid out. If it were tradeable and saleable, the price should be based on how likely it is that the insurance policy will have to be paid out. It should be more expensive to purchase as it becomes more likely to be paid out. The buying and selling of these insurance policies is predicated on being able to assess the likelihood of its kicking into effect. That's its reason for being, as it were.

Here's James Kwak on Baseline Scenario
:

"No, this isn’t another article about how credit default swaps (CDS) have ruined or are going to ruin the economy. It’s about one of the nice side benefits of CDS: the habit they have of pointing out who is going to get into trouble next. And it has pretty Bloomberg charts!

As everyone probably knows by know, a CDS is insurance against default on a bond or bond-like security. If you think about it for a while, you will realize that this means the price of the CDS reflects the market expectation that the issuer will default."

That's right.

"The price of a credit default swap is referred to as its “spread,” and is denominated in basis points (bp), or one-hundredths of a percentage point. For example, right now a Citigroup CDS has a spread of 255.5 bp, or 2.555%. That means that, to insure $100 of Citigroup debt, you have to pay $2.555 per year."

Price=Spread on CDSs.

"CDS exist for various durations and on many different kinds of debt. If someone doesn’t specify the duration or the type of debt, he is usually referring to a 5-year CDS on senior debt. That means that the contract will be open for 5 years, during which one party (the insured) pays premiums and the other (the insurer) promises to pay off if Citigroup defaults. If there is no default within 5 years, the insurer gets to keep the premiums."

Just like home insurance.

"Look at it from the standpoint of the insurer. If Citi doesn’t default, I get $2.555 x 5 = $12.775. If Citi defaults immediately, I have to pay $100. That implies that I think there is about a 12.8% chance that Citi will default (ignoring the time value of money). Actually, my expectation of a default is actually somewhat higher, for a couple of reasons. First, if Citi defaults 4-1/2 years from now, I have to pay $100, but I’ve collected the $12.775 in the meantime (assume premiums are paid at the beginning of each year for simplicity), so my loss is only $87.225. Second, in any case I don’t have to pay the full $100; I only have to pay $100 minus the value of the security, which is unlikely to be zero even in the case of a bankruptcy. For example, Lehman bonds were only worth 9 cents on the dollar (so insurers had to pay out 91 cents), but Washington Mutual bonds were worth 57 cents. So my net loss will be lower, which means that my expectation of a default is higher. (The expectation is the money I expect to gain if there is no default, divided by the net amount I expect to lose if there is a default.)"

The basic points have already been made. However one does it, you're trying to assess the likelihood of default.

"Luckily, Bloomberg can calculate all of this for you, and right now they say the chance of a Citigroup default in the next 5 years is 16.2%. (That’s using a recovery rate of 40 cents on the dollar, but you can type in whatever rate you want.) You can see the valuation on the right side of the screen below.

Citigroup valuation

OK, that’s interesting, but why call credit default swaps heralds of doom? Because CDS have shown the ability to identify what financial institutions (or countries) are going to get into trouble next. When the market starts getting nervous about a company and thinks it is more likely to default, insurance on that company’s debt starts getting more expensive. And this tends to happen before you start reading about that company in the newspaper."

So, you can use CDSs to assess the financial health of a company, or its likelihood to default. This post implies that this modus of assessing a company's finacial health is quite vaulauble because it predates a lot of other information for public viewing.

"Here are a few examples, in which I compare CDS prices to my home-grown “mainstream media” indicator, which is when the first article appeared in the New York Times saying a company was in danger of failure (as opposed to just taking writedowns along with every other bank). This is not scientific, because really you would want to compare the company’s CDS curve to an index of other companies in the industry to separate out sector-wide trends, but you get the point.

This is the chart of Bear Stearns’s CDS. Note that the price started climbing steeply in late February.The first Times article about Bear Stearns’s troubles was published on March 11, referring to the plunge in the stock price the previous day.

Bear Stearns

This is AIG. It looks like an instantaneous spike in mid-September, but the price had been climbing steadily, from double digits in May to 300 bp in mid-August to 430 bp on September 4. The first Times article appeared on September 12, again describing events on September 11. By September 10, however, CDS spreads were already up to 517 bp.

sg20081128756051

And this is Iceland. In the middle of 2007, Iceland’s CDS were priced below 10 bp. They spent most of July and August this year in the high 200s, passed 300 in mid-September, and reached 395 bp on Friday, September 26. Iceland only reached the attention of the mainstream media on Monday, September 29 (Times article the next day, in which Iceland barely got a mention).

Iceland

So whose CDS spreads are climbing now? That will have to wait for another, or several other, posts."

So, using CDSs as an assessment tool about the financial health of a company can be quite useful, especially if you are considering buying its stock or even its bonds.

Here are my questions, with some answers:

"James,

You make a good point about CDSs being useful. But couldn’t the same information of trouble ahead be reflected in the:
1) Price of the stock
2) Price of the Bond, or even Mortgage
3) Credit Ratings
4) Analysts Ratings
5) Company filings
Is there any reason to believe that buyers and sellers of CDSs have access to any more info than other traders in these other investments? Or that CDSs are more finely tuned to express movements in company prospects? All the CDS does is reflect the same, common, info differently, or am I wrong?"

“E.g., Goldman Sachs bought AIG CDSs in early August when AIG failed to meet collateral calls made by GS against AIG.’

I’m assuming that GS had some AIG bonds or product, against which they’re buying protection in case of default. Or, are you saying that GS went out and purchased some existing AIG CDSs, betting one way or another on a default. I’m still trying to see counterparties or anyone can see that a diligent investor can’t.

“CDS prices of financial companies often reveal problems that are evident to trading their counterparties but that are not yet apparent to equity and debt investors of financial companies.”

This sounds like counterparties are purchasing CDSs in case AIG defaults.When this happens, everyone knows there’s a problem, but not before. That makes sense, assuming that wired investors or others in the loop of this company’s investors don’t find out as well.

On the other hand, it certainly can’t help the fortunes of the company they’re buying insurance against. It looks less like hedging than pushing. I’m assuming that’s what you mean by shorting.

Also, you have to remember, there’s a buyer, and a seller. I can see one having the info, but surely the other end of the deal must see something else."

From James Kwak:

"The Goldman-AIG case is unusual but interesting. Goldman had some CDS contracts with AIG on other (non-AIG) bonds. As those contracts moved against AIG, Goldman demanded additional collateral (to protect itself against AIG defaulting and being unable to honor the contracts). AIG refused to give Goldman all the collateral they wanted. So Goldman went out and bought CDS protection on AIG.

Put another way: They knew AIG was in trouble because of their existing trading relationship with AIG, and that’s why they bought CDS protection on AIG. That is something that the general public would not have been able to know. How often this type of scenario occurs I do not know."

Here's me again:

“I’m assuming that GS had some AIG bonds or product, against which they’re buying protection in case of default.”

Your answer:

“Goldman had some CDS contracts with AIG on other (non-AIG) bonds. As those contracts moved against AIG, Goldman demanded additional collateral (to protect itself against AIG defaulting and being unable to honor the contracts). AIG refused to give Goldman all the collateral they wanted. So Goldman went out and bought CDS protection on AIG.

Put another way: They knew AIG was in trouble because of their existing trading relationship with AIG, and that’s why they bought CDS protection on AIG. That is something that the general public would not have been able to know. How often this type of scenario occurs I do not know.”

That’s what I wanted to know. However, would the price or anything about the CDSs become known. Who’s trading the ones on the market? Or is it that, GS can keep the CDS private, and the info will only get out if they try and sell them?

What you say makes sense, in that there is a chance, as I said, that if the info about these GS buys gets out, it can further harm AIG. However, if there is a CDS for any company, it does seem that the only reason to purchase a CDS is if you feel that there is a chance of default, and someone else who is willing to take your position on. CDSs don’t seem like Canaries to me, they’ve already passed on. CDSs seem like a reaction to Canaries, and validation of their deaths."

And that's it. A few other commenters were very helpful as well.

No comments: