Scheiber is worried that people like Morgan Stanley’s Howie Hubler — who lost $9 billion at what was essentially an in-house hedge fund — will simply now repeat their failures at standalone funds, if banks are barred from taking those kind of bets. But that misses the point. Hubler could put on those bets only because there was no prime broker breathing heavily over his shoulder, and because he had the full faith and credit of all of Morgan Stanley backing him up. The same is true of CDO losses at places like Merrill Lynch and Citigroup. And it’s also true of the pair of Bear Stearns hedge funds whose implosion marked the beginning of the crisis — when their prime broker sensibly withdrew, Bear Stearns itself stepped in to take losses on them. ...
But in general there’s one thing that the hedge fund system does well, and that’s confine hedge fund losses to the investors in those funds. Hedge funds will blow up occasionally, and that’s fine; the investors in those funds will lose money, and people betting against those funds will make money, and there will be few if any systemic repercussions. Even if the losses exceed the amount invested in the fund, those excess losses will be borne with few systemic implications by the fund’s prime broker.
Don't get me wrong--I'm all in favor of institutions that can fail without posing systemic risks. All things equal, I'd like to see lots of small institutions rather than a few big ones. I just think Salmon (and Mallaby) way overstate the benefits of what they propose. Here's why: Mallaby says the social benefit of hedge funds is their willingness to place smart, contrarian bets, which would nudge asset prices back toward their fundamental values when irrational behavior--bubbles or panics--starts driving them in the other direction.
The problem is that there are only so many supple-minded contrarians out there. Simply working at a hedge fund doesn't magically transform you into such a being, though it may confer advantages if you happen to be one already. If we follow Mallaby's recommendation and encourage hedge-fund proliferation, my guess is we're pretty quickly going to outrun the supply of worthy managers and end up with a bunch of mediocre ones--people no less prone to bubble or panic psychology (or just plain stupidity and carelessness) than when they worked at banks. The failure of a single hedge fund wouldn't be a problem, as Salmon points out. Neither would the failure of a handful or even a few dozen. But the failure of hundreds or thousands simultaneously probably would be a problem. And I think you could see that sort of thing if you took all the people who were making proprietary trades at banks and other big institutions and set them up at hedge funds.
Yes, the hedge funds create somewhat better incentives for traders (as I concede in my piece). And, yes, there are insititutional features, like their relationships with prime brokers, that impose healthy constraints. The downside is that they're far less regulated and transparent overall. So while I don't think shifting a bunch of risk to hedge funds would worsen the pre-crisis status quo, and I'm prepared to believe it may even improve things at the margins, I don't think it would improve them very much. You'd still have a small minority of independent thinkers and a big mass of crowd-chasers.
In fact, both Salmon and Mallaby concede that crowding is a real problem already. When you have a bunch of funds making the same bet with a lot of leverage, any event that triggers selling is going to vaporize a lot of them, because everyone has to sell at once and no one will be buying, sending prices through the floor. As Salmon notes, "given their extreme secrecy, [hedge funds] simply don’t know when they’re entering a crowded trade — as many of them discovered painfully during the quant meltdown of 2007." Wouldn't this problem get substantially worse if you dramatically increased the number of hedge funds--especially since the new funds are likely to dilute the talent pool?
P.S. Salmon also complains that I'm a bit unfair to Howie Hubler, the Morgan Stanley trader who blew $9 billion, and whom I accuse of being susceptible to market fads. He writes: "Hubler wasn’t jumping on to a market bandwagon: in fact, he thought he was betting against subprime bonds. He just funded that bet in a very unfortunate manner." It's true that Hubler bet against the riskiest subprime mortgage securities. But, as I recall (and I don't have the Michael Lewis book in front of me), he was long triple-A rated subprime mortgage securities (basically the higher-quality portion of a bundle of subprime mortgages that have been sliced up a few times), believing them to be virtually riskless because the ratings agencies had blessed them as such. This doesn't sound like exculpatory evidence. A number of people who analyzed these bets on their own found them to be pretty dodgy, the ratings notwithstanding. In fact, Hubler's bet was similar to AIG's, except it came a few years later, by which point it was even less excusable.