No Gambling With Trillions

No Gambling With Trillions

Ok, lets’s talk credit default swaps (CDSs): why they’re so dangerous and what can be done to fix them.

At the bottom all a CDS is is "you pay me money, and if X debt defaults, I’ll make it good." In other words, it is insurance, almost the same as homeowners insurance or life insurance. And since it’s insurance it should be regulated like insurance. But it isn’t.

The reasons the CDS swap market blew up include the following:

Bad Mathematical Modeling: Credit defaults are not independent variables. That is to say, they tend to happen in clusters and they aren’t evenly spread out over a curve. The math in most CDSs seems to have assumed that they were independent variables, that they didn’t cluster and that they were evenly spread out. So when a whole pile of defaults happened all at the same time, no one was prepared.

Taking Money Early: The risk of a credit default event (ie. having to pay out) doesn’t go down significantly over its term. If you’ve sold a 10 year swap, every year does not reduce the risk by 10%, for example, because defaults cluster. Unfortunately, what issuers of CDSs were doing was booking money every year as profit, usually at an even rate. So if it’s a 10 year CDS, take 10% profit each year. Since the risk hadn’t been reduced by 10% each year (indeed had hardly been reduced at all), this meant that they were booking profit they hadn’t earned.

Undercapitalized: Folks were selling CDSs without the necessary capital. You essentially didn’t have to have any money specifically committed to paying the CDS back. It would be like me saying "sure, if fred doesn’t pay you back 100K, I’ll make it good, if you pay me 10K now". Assume I could deal with that, mind you, and so could most people who issued CDSs. But let’s say I had hundreds of those outstanding. And suddenly a whole bunch of them come in all in the same year or two. Unless I’m very well capitalized, which these folks weren’t, I’m sunk. This is implicit in the math, events will cluster on occasion. Unfortunately, years go by without a lot of defaults, and so issuers tend to think "this is free money". They also think "the government won’t let these companies fail". They aren’t always right, though right now bet 2 is a pretty good one, because governments are paralyzed by fear at what will happen in the CDS market, among others, if another Lehman goes under.

How to Fix it

As noted above, credit default swaps are actually default insurance. To fix them you just regulate them like you regulate insurance.

Require use of standardized government approved actuarial tables and models: No, you don’t get to just make it up however your employee thinks it is. As with life insurance and most other kinds of insurance, the government and probably a professional association of quants will tell you what the risks are and what models you can use. You can use a more conservative model, but not a less conservative model. You will not be allowed to do the math wrong, and yes, your math will be audited. In fact, you will be required to submit your math for approval before using it in the market.

Require Strong Reserves You must have sufficient money to cover clustered events. You will be required to keep enough assets that you can meet the calls, and things like duration matching (making sure the time period of an asset matches the time period of a risk) will be required. The exact sort of securities which can be used as reserves will be specified, you will not be allowed to use assets which are not solid as reserves. One possibility is to simply require government insurance bonds, since the fact is that if you blow it, it’s the government that is going to have to clean up.

Do Not Allow Premature Booking of Profits: You don’t get to book the profits till the risk is gone. Period.

You get audited regularly We don’t assume your quants know what they’re doing, because even if they do, you’re probably telling them to cheat, and besides, they didn’t do it right when they were trusted to do it on their own, did they?

This is no longer a very high profit business: It’s insurance now, and insurance companies shouldn’t gamble. You will be nice and stable and profitable, but not immensely profitable. Your business will be boring and secure. That’s the way it should be, so you don’t cost us trillions of dollars to bail out again.

Concluding Remarks

There are very rarely new businesses under the sun. Credit Default Swaps were and are, just insurance. We know how to regulate insurance companies and all that is required is the political will to shut down the casino. The core problem here is the same as in every other "high" finance field—a lot of people are making a lot of money on the business and they don’t want it to stop. Ultimately the real reason they made money is because when they finally took a bunch of losses the government stepped in and covered much of them. Privatize the profits, socialize the losses. Or "heads I win, tails the taxpayer loses."

This is equally true of most types of swaps, many of which are essentially insurance contracts. For example, an interest rate swap is insurance against a rise or fall in interest rates. As a result this model, with some modifications, can be applied to the other swaps businesses. They will become a lot less profitable and a lot less sexy and once they cost as much as they should there probably won’t be nearly as many of them. But they also won’t be so dangerous that they threaten to swamp the entire economy.