Risk Markets And Politics

Monday, February 06, 2006

New York Summit Recap, Part 1

The turn-out was excellent, but there were no big announcements or public tip-offs. This was more of a networking event than an informational seminar. The bankers and venture capitalists far outnumbered the university professors. Other industries with representatives in attendance included pharmaceuticals, autos, energy and accounting. Regulation is a significant concern and there is a sense that controversies over the legality of prediction markets may suddenly escalate, and that "the industry" should be prepared for this — or proactively press for changes in gambling laws. Here are some speaker highlights in no particular order. Watch for Part 2 later this week.

Trade Exchange Network CEO John Delaney expects the prediction market industry to experience 500% year-over-year growth in 2006, but played devil's advocate and examined various risks to this success. What if prediction markets are just a fad? Will the markets be able to improve liquidity? What if there is a major prediction market failure? Will the industry receive mainstream validation and how will this affect regulation? He seemed more concerned about the latter two questions, and proposed an industry association to help navigate these issues. For instance, the association could maintain a historical database of market results that would facilitate a general objective evaluation of prediction markets if there should be a notable specific failure. Bo Cowgill of Google thinks that an industry association could lobby for changes in gambling laws in the United States, but cautioned that it should also get non-corporate interests involved.

Delaney said almost nothing about the Intrade EBOT. When asked about it, he noted that Intrade is still plodding along in its quest to become a CFTC Designated Contract Market. It is unclear whether something is imminently in the works with the EBOT, or if Intrade has moved the venture to the backburner, looking forward to the less limiting DCM status. Delaney did report that Intrade is maintaining an internal prediction market on the date of their CFTC approval, but declined to give any prices. He also announced a "Prediction Market Experiment" and handed out username/password combinations to the audience which will allow them to participate in various play-money markets concerning the future of the industry and the concerns that he had previously outlined. (Some quotes will be included in Part 2.) Delaney also mentioned that rising interest rates are beneficial to Intrade, but acknowledged that higher rates will also distort the probalistic interpretations of their long-dated contracts since the interest isn't passed back to most depositors.

Charles Polk, CEO of Common Knowledge Markets, described the Turkish Avian Flu Outbreak Market, which was subsidized through the Gerson Lehrman Group. This was essentially a real-money market, but it was fully-funded. In other words, the traders competed over a pool of funds to which they did not contribute. Polk described it as an alternate way to compensate consultants (i.e., the traders). This is an interesting example with respect to gambling laws.

Chris Hibbert of CommerceNet is among the most well-versed in the legality of prediction markets and argues that such markets will improve their chances of favorable legislation if they can demonstrate that they are providing socially useful information. To this end, he seems constructive on real-money markets subsidized by information-seeking patrons, as in Charles Polk's case.

On an interesting technical note, while discussing Zocalo, Chris Hibbert described a way for exchanges to alter the liquidity of "N-way claim" markets by posting orders that can trigger the simultaneous buying or selling of the full set of possibilities. For example, consider this $10-per-contract election market whose width is $4 (13 to buy the full distribution minus 9 to sell it). The exchange can lower the Liberal offer to 5, tightening the market width to 3:
Conservative Party24522452
Liberal Party24622452
All Others21222122

The exchange is not exposed to the market because if someone buys their offer at 5, it immediately sells the Conservatives at 4 and the Field at 1, thus selling all possible outcomes for $10, or the full distribution at 100%:
Conservative Party24521452
Liberal Party24512451
All Others21221122

Presumably, all other trades are locked-out of the processing queue while the exchange flattens its exposure in this manner. It is not always possible for the exchange to find such a combination of trades. Even if there is a suitable combination, having it executed may actually increase the market width, as would be the case if the last example were iterated:
Conservative Party14522352
Liberal Party24512462
All Others1122--22

Most importantly, this method always decreases the market depth. In these examples, 1 contract of depth was added while 2 were removed. Unfortunately, it doesn't work in reverse (adding 2 to remove 1) because the 2 added contracts would have to be traded by others simultaneously. Even if market depth is always removed, it might be beneficial for an exchange to make these sorts of trades in order to capture commissions. This would be the case with Tradesports and its policy of charging price-takers commission, but not price-makers.

Marco Ottaviani discussed some subtleties of market and price formation. He stressed the no-trade concept. He described how market participants exaggerate the importance of the information and analysis on which they base their trades - their "secret signals" - and this among several other factors causes the favorite-longshot bias.

Koleman Strumpf summarised his paper Manipulating Political Stock Markets, which detailed the political gambling markets of the late 19th and early 20th centuries in the United States. He does not interpret the 1919 Black Sox scandal as being important in the fall of the original election markets, although he is working on a paper involving that scandal. He finds it curious that political parties are so reluctant to hedge their fortunes by selling themselves short in political markets, especially since it might often be desirable to appear weaker in order to make the opposition complacent and/or galvanize one's own voters.

Again, check back later in the week for Part 2 of the recap, which will include Russell Andersson of Hedgestreet, Emile Servan-Schreiber of NewsFutures, and James Surowiecki.


  • Regarding Chris Hibbert's liquidity method, it "always decreases the market depth" if the number of (exclusive and exhaustive) claims is greater than 2. If N > 2, it can only add depth indirectly by emboldening other price-makers, but if there are a preponderance of takers, watch out.

    By Blogger Jason Ruspini, at 8:59 PM  

Post a Comment

<< Home