Trading the election

Since everybody has already been talking nonstop about the future elections of 2008, I consulted a different kind of poll: the 2008 election futures markets at Intrade, the Dublin-based online futures exchange. Intrade commoditizes future events by marrying market structures to individual betting on highly speculative – but not unpredictable – events. Put another way, Intrade offers a platform on which individuals from around the world auction event outcomes, with the event’s estimated probability as the event’s price. It does not function as “the house,” but simply facilitates bets made between individuals for a small transaction fee.

Intrade has had mixed success as a predictor. It was widely hailed in 2004 as a superior predictor of state outcomes by the night before the election – before election-day exit polls (which later turned out to be wrong) jinxed the market’s overall performance. In the 2006 midterms, it called the House correctly, and also called virtually all individual Senate races correctly – but it also predicted that the Democrats would not win as many races as they did; for example, it gave the Democrats only a 30 percent chance at winning the Senate, even though it called virtually all individual races correctly. (Essentially, the markets underestimated the degree of intercorrelation between congressional races.)

The consensus among political pundits similarly held that the Democrats would win the House, but not the Senate – leaving Intrade vulnerable to the charge that it merely distilled the opinions of talking heads, as opposed to bringing anything fundamentally new to the table.

However, with the markets so young and liquidity so low, evaluation is difficult. If history is any guide, however, prediction markets have huge growth potential. From 1880 to 1932, a massive betting market thrived on Wall Street, and it acquired an almost mystical status in American media for “never being wrong.” According to an academic study by Profs. Koleman Strumpf and Paul Rhode, the Wall Street prediction markets peaked in 1916 with an astounding $156 million of volume - in 1916 dollars (!!). The prediction market died as collateral damage from FDR’s New Deal jihad against the securities industry, and betting markets did not make a comeback until the mid-1990’s.

They are also probably the target of manipulation by rival campaigns. Manipulation charges were rampant with regards to the earlier Wall Street market, and in 2004 there were blatant attempts to drive down the price of the Bush reelection derivative to zero. (They failed ignominiously as the markets bounced back immediately, probably costing the manipulator six figures’ worth of losses.)

 True to historical precedent, charges of manipulation are already flying. Ron Paul’s online army seems to think that Paul has a 2 percent-plus chance of winning the GOP nomination, and are zealously bidding him up. The 2008.PRES.CLINTON(H) contract in particular seems to have been manipulated for approximately one month. For an excellent summary of the arguments that a Clinton supporter has been propping up the 2008.PRES.CLINTON(H) contract, I’d recommend Stanford Graduate Business School Prof. Eric Zitzewitz’s writeup. Or, let the picture speak the thousand words instead:

(Note the gigantic spike from a post-announcement average of 27 to a mid-May high of 44, for no discernible reason. Also know that, considering the amount of publicly available polling data, the value of political “inside information” is actually quite minimal, except in very specific instances, i.e., if Al Gore has actually decided to run or not.)

Intrade also features contracts on whether the United States will “neutralize” Osama bin Laden by certain dates, whether the United States will launch an overt air strike on Iran by certain dates, how many hurricanes there will be in 2007, and whether an immigration bill will pass the House and the Senate by the end of 2007, among many others. They remain an opportunity for current-events junkies to alchemize information addiction into gold, and for outsiders to watch traders speculate on outcomes of highly, but not completely unpredictable events.

Thanks to America’s terrible regulatory environment, exchange managers have little reason to base their operations in the United States; investment usually requires a bank wire. If history is any indication, the market has orders of magnitude left to grow to reach its 1916 peak, and provides plenty of volatility and insight in the meantime.