« Ask the Editors: What's Wrong with Deflation? | Main | Chrysler is Doomed »
Mar 30 2009, 8:47 am
Why Credit Default Swaps are Dangerous
Felix Salmon says that credit default swaps are just like bonds. Charles Davi says they are just like futures and forwards (actually, they are most like options), which are derivatives that provide liquidity. But CDS are different from either of these.
The difference between credit default swaps and typical financial derivatives is the long-term nature of the commitment with CDS. On organized futures and options markets, it is essentially impossible to obtain a long-dated option. That is, if I want to buy an option that expires three years from now, I have no real chance of doing so. In most markets, such options are not traded at all. Even when there are option contracts that exist for more than six months ahead, trading volumes are slim, so that it is quite difficult to take a position of any size.
Long-dated options are only sold over-the-counter, with CDS being the main example. I believe that some currency swaps and interest-rate swaps sold over-the-counter also are long-dated. I suspect that those contracts are dangerous as well. If interest rates on Treasuries rise to double-digit rates in the next few years, then my guess is that some sellers of interest-rate swaps will blow up.
If you sell a long-dated option, then you need much more in terms of capital and loss reserves to cover your bet than is the case with options that expire over the next few months. Unless, of course, you are willing to take the risk of blowing up in a few years in order to get some nice profits in the near term. My guess is that if the writers of all forms of long-dated options were required to put up sufficient capital and loss reserves, the markets in those options would shrink considerably.
Speaking of capital, you need it in order to invest in bonds. But when you create a synthetic position in bonds by writing credit default swaps, you can do so without putting up capital. With bonds, investors who put debt or equity in your firm have a clear picture of what their risk and return profiles look like. With credit default swaps, you hide your leveraged bond position from regulators and investor.
There is nothing intrinsically opaque about credit default swaps. Regulators and accountants could require firms that are net sellers of credit default swaps to translate those positions into bond holdings and put these synthetic bonds on their balance sheets. My guess is that had such a policy been in place in 2000, the CDS market would not have taken off.
The difference between credit default swaps and typical financial derivatives is the long-term nature of the commitment with CDS. On organized futures and options markets, it is essentially impossible to obtain a long-dated option. That is, if I want to buy an option that expires three years from now, I have no real chance of doing so. In most markets, such options are not traded at all. Even when there are option contracts that exist for more than six months ahead, trading volumes are slim, so that it is quite difficult to take a position of any size.
Long-dated options are only sold over-the-counter, with CDS being the main example. I believe that some currency swaps and interest-rate swaps sold over-the-counter also are long-dated. I suspect that those contracts are dangerous as well. If interest rates on Treasuries rise to double-digit rates in the next few years, then my guess is that some sellers of interest-rate swaps will blow up.
If you sell a long-dated option, then you need much more in terms of capital and loss reserves to cover your bet than is the case with options that expire over the next few months. Unless, of course, you are willing to take the risk of blowing up in a few years in order to get some nice profits in the near term. My guess is that if the writers of all forms of long-dated options were required to put up sufficient capital and loss reserves, the markets in those options would shrink considerably.
Speaking of capital, you need it in order to invest in bonds. But when you create a synthetic position in bonds by writing credit default swaps, you can do so without putting up capital. With bonds, investors who put debt or equity in your firm have a clear picture of what their risk and return profiles look like. With credit default swaps, you hide your leveraged bond position from regulators and investor.
There is nothing intrinsically opaque about credit default swaps. Regulators and accountants could require firms that are net sellers of credit default swaps to translate those positions into bond holdings and put these synthetic bonds on their balance sheets. My guess is that had such a policy been in place in 2000, the CDS market would not have taken off.










The people who were selling CDS on a large scale (led (I hope there is no one who did more) by AIG) got away with it for a while because Moody's and their fellows were very, very, slow in revising the sellers' credit ratings. Long term put options based upon nothing much more than a rating were and always will be likely to prove disastrous to the seller, worthless to the buyer, or both.
As Kling notes, CDS or similar options based upon large quantities of capital are possible; but unlikley to be profitable. However, there is a familiar occupation in which people are and have been getting rich doing this sort of business with limited capital: bookmaking.
Bookmakers quite frequently lay bets that would bankrupt them if they had to pay out in full. They do this because they know they can lay off part of the potential loss if paying out begins to seem likely. Secondly, bookmakers rapidly change in their favour the odds they offer to new punters whenever the initial odds they are offering attract quantities of money. My guess is that there is viable space in the financial world for firms to offer CDS and other puts on the same basis, long as well as short. What is more, it would be quick and easy for rating agencies, and parties offering these options, to judge the soundness of the party they were dealing with: if they act like a sound bookie, they are probably good for the money; if they don't, don't bet with them.
Part of AIG Financial Products simply acted like a very stupid, or very drunk, bookie. They accepted masses of potentially bankrupting bets without knowing where and how much they would be able to lay off if necessary. When a queue of punters formed at the door begging AIG to take their money, they listened to the punters pretty tales about why they wanted to bet rather than assuming the punters knew something they did not. If the rating agencies had been on the watch for non-bookie-like behaviour, AIG's AAA rating would have disapppeared well before the main mass of its crippling CDS were sold.
The banking elite need to exercise extreme caution here. It's true, the American people are very apathetic and lazy for the most part. However, they are beginning to become very angry with the likes of Chase Bank, Citi and a few others. If the financial situation doesn't at least start to show signs of relieving pressure on the average American, there may be a real risk of widespread civil unrest, and possibly worse.
It's time to get this house of cards in order gentlemen.
Arnold,
The collateral aka margin for a CDS is two ways, i.e., either party could end up posting, which means it is not like an option, but a forward contract. That said, the income of a protection seller is similar to that of a bond holder. So it's not an either or situation. The collateral operates like a forward contract but the economic exposure to the protection seller is very similar owning a bond.
The long-dated option issue you cite is largely dealt with through the use of collateral. CDS is a bilateral traded instrument, traded between sophisticated parties that want to get paid. To say they are dangerous is to say that at least one of those parties either doesn't want to get paid or doesn't know how to get paid.