Michael Lewis has a grand theory about the CDS market: It was all one big mechanism for paying traders to take a whole bunch of tail risk which would eventually blow up all of Wall Street.
You know, I have yet to have a financial person persuade me that there's a really useful reason for a credit default swap. I know why they exist and I know why they're used. They're mostly used as speculative instruments. And the people who are selling the insurance are mostly selling it because they don't pay a price for it until everything goes bad.
This doesn't even make internal sense. Yes, there are some people who speculated with credit default swaps: bought them when they thought they were going to go higher, and sold them when they thought they were going to go lower. Those people, by definition, were marking to market, and would profit by closing out their positions when the market moved in their favor.
On the other hand, there were the long-term investors -- pension funds, insurance companies, that sort of thing -- who would sell five-year credit protection with every intention of simply insuring that credit for the full five years, and pocketing the insurance premiums. Those players can't be considered to be speculators -- but it was they, most notably at AIG Financial Products, who refused to mark their positions to market and therefore didn't "pay a price" until it was far too late.
In any case, you don't need to look for logical inconsistencies to see that Lewis doesn't understand credit default swaps as well as he thinks he does. You just need to see him getting the basics of the market upside-down:
Let's take a bond, let's say a General Electric bond. A General Electric bond trades at some spread over treasuries. So let's say you get, I dunno, in normal times, 75 basis points over treasuries, or 100 basis points over Treasuries, over the equivalent maturity in Treasury bonds. So you get paid more investing in GE. And what does that represent? You get paid more because you're taking the risk that GE is going welsh on its debts. That the GE bond is going to default. So the bond market is already pricing the risk of owning General Electric bonds. So then these credit default swaps come along. Someone will sell you a credit default swap -- what enables the market is that it's cheaper than that 75 basis point spread -- and he's saying that in doing this he knows GE is less likely default than the bond market believes.
What Lewis is saying here is that the driving force behind the CDS market -- "what enables" it -- is the fact that the basis between CDS spreads and bond spreads is negative: CDS spreads were lower than bond spreads on the same credit.
Except, they weren't. In reality, the CDS-bond basis has nearly always been positive, certainly in the days when the CDS market was growing fastest. Far from implying that GE was less likely to default than the bond market believed, the prices on CDS implied that it was more likely to default than the spreads on GE's bonds implied. Just to take the most obvious example, the credit default swap on the US government is trading at about 60bp, which is 60bp wide of the risk-free rate as measured by the Treasury market: it's impossible to have a negative basis on US government CDS.
Now there are lots of technical and fundamental factors behind the CDS-bond basis, which aren't worth getting into here. But if Lewis really believes that a negative basis was the "enabler" of the market, he really should start talking to more financial people.
And yes, there is "a really useful reason for a credit default swap" -- it's pretty much the same as the really useful reason for the existence of equity markets. In the stock market, there are lots of sellers, and lots of buyers, and the public visibility of the market-clearing price creates a lot of value. The bond market, by contrast, has always been much more illiquid: Bond investors tend to be buy-and-hold types (remember those pension funds and insurance companies) who buy up bonds at issue and then hold them to maturity. As a result, it can be very hard to short any given bond, or to get a useful bead on exactly what the market-clearing price on any given company's debt might be. Unless you have a liquid CDS market, which is very useful indeed when it comes to price discovery.
I'm pretty sure that Lewis understands the importance of short-sellers to the stock market; it's weird he doesn't understand that the existence of short-sellers in the credit market can serve an equally useful purpose. But if demonizing short-sellers is a popular and easy sport these days, demonizing the CDS market is easier still. More's the pity.