Wednesday, 4 April 2012

No political events derivatives for the US [updated]

The CFTC has barred real money political event derivatives in the US.

Among the reasons given for the CFTC not allowing such contracts under their jurisdiction is this howler:
WHEREAS, the Political Event Contracts can potentially be used in ways that would have an adverse effect on the integrity of  elections, for example by creating monetary incentives to vote for particular candidates even when such a vote may be contrary to the voter's political views of such candidates; 
Yup. And allowing lotteries might have people choose favourite numbers that are contrary to their actual views about the merits of numbers.

The actual chances of your vote changing an election are infinitesimal. Voting to make your contract more likely to pay out makes as much sense, in the limit, as thinking you can improve your odds when you've bet against your favourite team by yelling at the television.

The CFTC highlights the importance of an economic purpose test in determining whether the contracts are contrary to the public interest. What determines lack of economic purpose? Nothing more than assertion.
WHEREAS, the unpredictability of  the specific economic consequences of  an election
means that the Political Event Contracts cannot reasonably be expected to be used for hedging purposes;
That's an interesting conjecture, but the evidence points pretty strongly against it. Here's Brian Knight in the Journal of Public Economics.
This paper tests for the capitalization of policy platforms into equity prices using a sample of 70 firms favored under Bush or Gore platforms during the 2000 U.S. Presidential Election. Two sources of daily data during the six months leading up to the election are incorporated: firm-specific equity returns and the probability of a Bush victory as implied by prices from the Iowa Electronic Market. For this group of politically sensitive firms, the daily baseline estimates demonstrate that platforms are capitalized into equity prices: under a Bush administration, relative to a counterfactual Gore administration, Bush-favored firms are worth 3% more and Gore-favored firms are worth 6% less, implying a statistically significant differential return of 9%. Estimates based on weekly returns are even stronger, suggesting a differential return of 16%. The most sensitive sectors include tobacco, worth 13% more under a favorable Bush administration, Microsoft competitors, worth 15% less under an unfavorable Bush administration, and alternative energy companies, worth 16% less under an unfavorable Bush administration. A corresponding analysis of campaign contributions, which allows for heterogeneity in the importance of policy platforms to the firms, supports the baseline estimates.
And see too Snowberg, Wolfers and Zitzewitz.

So we have decent evidence that political events markets tell us something important about price movements for politically sensitive stocks. But the CFTC says:
WHEREAS, there is no situation in which the Political Event Contracts' prices could form the basis for the pricing of a commercial transaction involving a physical commodity, financial asset or service, which demonstrates that the Political Event Contracts have no price basing utility
Really? No situation? That's pretty harsh. I can imagine some. The Knight article reasonably suggests a pile of commercial transactions whose pricing could depend on prices in political event contracts.

The ruling is a pretty big disappointment. It would have been one thing if they'd had a reasonably argued case against allowing political event markets under CFTC jurisdiction. Instead they're using assertions of no utility as justification for a ban, when the evidence isn't really on their side.

HT: Chris Masse

[Update: wording clarifications, economic purpose test noted]

No comments:

Post a Comment