Guest post today from Gabriel Weil, who many know from the Intrade forums as GAW38. I think he's one of the smartest guys on the forum, and I hope he posts here more often.
The controversy over US.GOVT.HEALTHPLAN.DEC09, the contract on the public option, illustrates the limits of prediction markets. It's really easy to write a contract for election results and be confident that it will eventually be clear what the expiry prices should be. For legislative outcomes, and other events that involve a wide array of qualitatively different potential results, this task becomes much more difficult. As a result, I usually stay away from such markets. In this case, I took advantage of the situation a little and shorted the public option contract after one of the clarifications, and was able to take a decent profit after the price plunged to single digits.
Interestingly, even after the clarifications, the intrade price continued to be cited as a indicator of the likelihood that the final bill would include a public option. See this post: http://www.marginalrevolution.com/marginalrevolution/2009/10/the-resurrection-of-the-public-plan.html on Marginal Revolution, a blog run by two George Mason economists, Tyler Cowan and Alex Taborrok. Their colleague, Robin Hanson, is a strong proponent of using prediction markets to guide public policy. I agree that is a project worth exploring, but one of the problems it would run into is developing a mechanism for incorporating the many nuances of public policy with clear and well-defined contract rules. Sure, intrade could be doing a lot better with writing clear rules than it has to date, but it also has the luxury of using broad generalities. Any effort to use prediction markets in policy formation must drill down to a much greater extent to the details on policy debates.
This is one of many challenges that prediction markets like intrade must confront if they are to advance from a curiosity to something of truly great social value and importance. It's also difficult to tell how serious an obstacle this is just yet, because intrade has not made a great effort to overcome it.
About Me
- Jesse Livermore
- A neuroscientist by day, and a crusher of Intrade by night. Starting in 2008, I've turned a single $50 deposit into more than $50,000. Follow me on Twitter here. Contact me at wiserthanthecrowd@gmail.com
Blog Archive
-
►
2010
(22)
-
►
January
(19)
- Now this is Hubris
- Efficient markets what?
- Obama hires McCain's campaign manager
- Test your Intrade IQ
- Fallout from Brown
- I should have seen this coming
- Same old song
- These are ALWAYS funny
- And... Intrade is down
- One more word before the results come in
- Election Day in MA
- What happens to health care?
- Even Obama thinks Coakley will lose
- More Massachusetts Polls
- Bad Sign
- Health Insurers Stock
- What can Brown do for you?
- Hack job on Monsanto
- New Year Update
-
►
January
(19)
2 comments:
*GAW838
I'm glad I didn't go anywhere near this contract. I was surprised that they would rule that an 'opt-out' plan is not a public option. Does anybody have a reasonable solution to the problem of clarifying contracts?
First, I agree with the social value aspect. I'm getting really tired of people sitting in judgment of "socially useless" trading. At their best, prediction markets are socially useful trading that people can understand.
But yeah, designing legislative contracts is tough, and just imagine if the markets were significant enough that legislators reacted to them. (Whoever tries to make a real go of this in the US will have to maintain a good, constructive relationship with Washington.)
In terms of contract design, first ask if there is a comparable non-binary contract that might be traded as actively. For binaries, if you take the hedging approach, the challenge is to make the contract general enough to avoid false negatives but specific enough to avoid false positives. False negatives are probably the more common danger, e.g. you have a contract that pays if income tax rate is above a certain level, but, oh, now there's a surtax and you still pay higher taxes but the contract expires at zero. This possibility can be mitigated with an elastic clause in the contract rules that gives latitude to the exchange so that it can best satisfy the hedging intention of the contracts when the settlement source is rendered unusable or incomplete. Not perfect, but if traders don't like that clause, they don't have to trade the contract. This is essentially what I did with Intrade for the income tax contracts, and it did come into play for the surtax case.
http://www.intrade.com/jsp/intrade/common/c_cd.jsp?conDetailID=575720&z=1257964679348
The surtax treatment should also make the contracts more active since they are affected by additional newsflow, and that is what they desperately need to see active trading.
Another option, which might be more attractive if there is real money involved, is to somehow spin-off or split a contract if a problem arises.
Post a Comment