A while back a federal judge in New York decided that poker wasn't illegal gambling under federal law, and we discussed it here, and so I feel a certain sense of responsibility to all of the Dealbreaker readers who quit their jobs to start poker rooms. So I should tell you that that decision was reversed today on appeal, and poker is illegal again, so you'd better shut down your poker room and go back to trading derivatives ha ha ha.
Other than that I don't have much to say about the decision. The lower court's decision declaring poker legal was very scholarly about how poker is predominately a game of skill, not of chance, and so is not "gambling." But it was pretty loony on the law: poker is undisputedly illegal gambling under New York law, so the question of whether it was illegal under federal law turned on whether it's a "gambling business which is a violation of the law of" New York. The district court said, well, it's a violation of New York's gambling law, but it's not a "gambling business" under my reading of the word "gambling," because it's a game of skill, so you're free to go. The appeals court ignores all the chance vs. skill stuff and just says "it's gambling under New York law, so it's gambling under the statute, so you're hosed." That's sad but also probably right.
Here you can read Floyd Norris complaining that synthetic CDOs, and by extension most derivatives trades, are just gambling, the equivalent of a bank's "put[ting] together a betting book and taking out the bookie’s cut." This argument is not novel, nor have I ever found it that interesting, but it is true as far as it goes.1
It doesn't go that far, though; in particular, the analogy between Lawrence DiCristina's illegal Staten Island poker room and the multi-billion-dollar derivatives business doesn't make the former any more legal or the latter any less. The distinction is one of formalism and good lobbying, not of logic. Some games are illegal gambling because they're games of chance, and some games are illegal gambling even though they're games of skill. And some games - the game of speculating in the stock market not least among them - are fine even if, as seems likely,2 they are more driven by chance than poker is. If you try, as the district court did, to sort them out based on logic, you will end in a hopeless mess. The trick is to recognize that the lines are arbitrary, and to know how to stay on the right side of them, or failing that, how to move them. People who run derivatives books tend - mostly!3 - to know how to do those things. People who run Staten Island poker rooms tend not to.
U.S. v. DiCristina [2nd Cir.]
Earlier: Poker Is Not Gambling, Jury Still Out On Investing
1.Oh of course it's not true. Take Abacus, which is what Norris is talking about. The argument is:
But the securitization that Mr. Tourre helped to create was “synthetic.” It did not raise money that went, directly or indirectly, to homeowners or to those who had lent money to them. Instead, it picked a bunch of tranches from previous securitizations — tranches that no one involved in this deal had to own — and fashioned a new securitization in which one set of investors bet those securities would work out and another set bet they would not.
There are two ways of thinking about the gambling question. One is, were the parties to the trade involved out of gambling motives (entertainment, chance to make a big profit), or investing motives (have money that they need to transfer in time), or hedging motives (have risks that they need to offset with other trades), or what? There's a decent argument that Paulson, the protection buyer in Abacus, had gambling-ish motives; the long side is less clear and there's at least a case that the note buyer (though not ACA) had investment motives. But this is contingent; it's not a feature of the security itself. Some synthetic CDOs were created by banks to hedge their mortgage portfolios, and plenty of note buyers were just investors who wanted a few extra bps of yield versus non-synthetic trades. There's no requirement of gambling motives, unlike in, say, poker, where everyone's there to gamble. Err, to ply their skill.
The second question is whether Abacus "raise[d] money that went, directly or indirectly, to homeowners or to those who had lent money to them," and here you have sort of market-structure effects. Did the CDS liquidity provided by the existence of synthetic CDOs make it easier for lenders to provide mortgages? Maybe, I don't know, but certainly Norris doesn't either. Would that have been a bad thing, given that there was a mortgage bubble? Maybe? One theory I've heard is that synthetic CDOs actually provided the benefit of taming the mortgage bubble: if people couldn't get long mortgage risk via synthetic CDO they'd demand even more and worse actual mortgages. Synthetics are zero-sum, of course, making this theory sound a bit odd, but since they were pretty much the main way of shorting mortgages, and since Paulson and others were putting on large synthetic shorts, and since short selling aids price discovery, it's probably true. In any case that would imply that the long side of the synthetic trade, like the long side of a physical trade, was propping up the market and thus through no-arbitrage was making your mortgage cheaper. So the point is: Abacus probably did "raise money that went ... indirectly, to homeowners," for good or ill, you're welcome.
Is it gambling? Sure whatever. But life is complicated, markets are complicated, and there are limits to the social utility of liquidity provision but those limits are fairly obviously not "any secondary liquidity provision is bad."
2.Seriously the district court opinion is a good read, basically summarizing testimony that top poker players demonstrate persistent alpha, which is more than can be said for say mutual fund managers.