By Sarah Lovenheim
washingtonpost.com Staff Writer
Thursday, October 2, 2008 2:23 PM
Last September, Bethan Brome Lilja, a Massachusetts commercial photographer, made a bet of sorts that seemed absurd at the time: that Sen. John McCain (R-Ariz.) would win the Republican presidential nomination.
Although McCain's campaign organization had imploded that summer and his polling numbers were near rock-bottom, Lilja purchased about $75,000 worth of contracts on an unusual futures market called Intrade speculating that McCain would pull out the nomination. She also wagered that former New York City Mayor Rudolph Giuliani would lose his campaign for the GOP nomination, despite polls showing him very much in the running.
Today, Lilja's high-stakes political speculation looks prescient, and she has made more than $300,000 off the market -- roughly four times the amount she initially invested.
Intrade, a nine-year-old financial exchange based in Dublin, Ireland, invites investors worldwide to trade contracts tied to political outcomes. The exchange operates much like a financial futures market, only profits are contingent on the result of a political race or an electoral map shakeup, rather than a company's growth. The price of a contract indicates how likely investors consider an outcome.
For example, a contract tied to the prediction that McCain will win the election selling for 47 points means that investors believe McCain has a 47 percent chance of winning the election. On Intrade, a point represents 10 cents, so one share of stock in McCain is worth $4.70. Confident investors place tens of thousands of dollars or more on one contract.
Prediction markets have been around in one form or another for decades. Throughout the 2008 presidential campaign, Intrade has generated nearly $50 million worth of futures contracts. Betfair, a market based in the United Kingdom also attracts millions of dollars in bets. Its Web site predicts nearly $70 million will be invested on the U.S. presidential election.
U.S. law prevents predictive trading nationwide, with the exception of the Iowa Electronic Markets which began as an academic research tool and today caps investments at $500.
The markets at times have worked as an uncanny bellwether of political trends -- in some cases proving to be a more accurate indicator of political trends than the pundits or the polls.
A few days before Democratic presidential candidate Barack Obama (Ill.) announced his selection of Sen. Joseph Biden (D-Del.) as his running mate on Aug. 23, a contract linked to the prospects of Biden being picked jumped 300 percent on Intrade. Shortly before Obama picked Biden, Sen. Evan Bayh (Ind.) and Gov. Tim Kaine (Va.) were on pundits' short lists.
And as soon as a blog leaked news of a jet landing in Anchorage to pick up Alaska Gov. Sarah Palin the day before McCain announced his selection, a contract tied to Palin being picked as McCain's running mate soared in market value. Market expert Justin Wolfers, an associate professor at the University of Pennsylvania, noted "That was picked up in seconds on Intrade." The mainstream news media didn't pick up the speculation on Palin until several hours later.
Investors drawn to this form of speculation "have to be a little savvy," said economist Dean Karlan of Yale University. "At the end of the day it's a security," he said.
And to do well in this form of speculation, experts say, investors have to check their political leanings at the door and invest money strictly based on who they think will win or lose. Lilja, a Democrat, will only bet on Republican outcomes. "I don't want my own personal bias to get in the way," she said.
Sometimes she confers with financial analyst Alex Forshaw, who has also bet on Intrade, before hedging a bet. Forshaw believes investors should make cautious predictions. "When I was in college, I didn't do much besides bet on political outcomes. It worked very well, but I got overconfident, and my model 'blinked' for about two weeks out of 10 months, which was enough to cost me everything I'd gained," he said.
Still, Forshaw tracks market trends because he believes most investors are wiser than that. "Anybody who thinks his opinion is materially different from the herd's, and meaningfully more predictive, should invest in the futures markets," he said. "The market is a democracy in the short run and a meritocracy in the long run. It punishes stupidity of academics and political 'experts' just as ruthlessly as it punishes the stupidity of overconfident college students."
Intrade's Vice President of Business Development Chad Rigetti believes the markets often are more insightful than polls. "Polls do not respond to news in real time. Market prices do," he said.
Yet while investors respond quickly to some events, they react cautiously to others.
While McCain's standing in the polls noticeably improved immediately after the Republican National Convention and his choice of Palin as his running mate in early September, the price of shares tied to McCain on Intrade barely changed.
From his perch at the Wharton School of the University of Pennsylvania, Professor of Business and Public Policy Justin Wolfers watched the market daily. "The two weeks after the convention were the two most stable weeks in the markets," he said. "Markets were quite unimpressed by the choreography of the conventions and correctly responded."
Most investors, he explained, realize voters ultimately vote for the top of the ticket, rather than the bottom. "They were not going to place more or less money on a prediction regardless of public sentiment reflected in polls," he said.
With convention bounces all but faded, national polling -- which currently shows Obama with a four-point lead over McCain, according to an aggregation of polls by Real Clear Politics -- is more in line with Intrade's market.
Karlan, the Yale economics professor, said that when he is around a group of economists, "people will talk about the Intrade odds much more than poll data." Karlan believes the data's worth watching because "You're not going to invest thousands of dollars unless you know something beyond the average voter."
Prediction markets predate polls and were considered a reliable source to track political trends in the early 1900s. Polling emerged as a source of political intelligence in the 1930s. "If you go back to the 1912 election, large-scale prediction markets were the basis of [forecasting] the race," said Wolfers.
These markets subsequently declined in importance, but made a comeback after accurate political forecasts in 2004, according to Professor Keith Chen of Yale. "Intrade investors who bet on the electoral college map, accurately predicted that the presidential candidate to win Iowa would win the election," said Chen. They based their predictions off of mathematical calculations.
These speculators predicted the number of states President Bush and Democratic challenger John Kerry were likely to carry and ranked the likelihood of each. The candidate who could win Iowa, they determined, would win all of the states needed to enter the White House. On Election Day, Bush won Iowa, and all of the other states calculated in his favor. Polls, however, speculated otherwise. Election day exit polls suggested Kerry would win the popular vote, based on a widespread belief that he would win the electoral-rich state of Ohio. That night, Kerry lost Ohio and lost the election.
During this election cycle, Chen says economists applying the probability formula to investments say Nevada will be the pivotal state. While some polls suggest that Ohio is again the battle-ground state likely to determine the election, economists like Chen will watch Nevada. Investors believe the winner of Nevada will win the necessary electoral votes from other swing states to claim victory.
And what happens once the election's over? What will economists and politically savvy investors like Lilja bet on the markets? Contracts tied to potential policies of the next Congress and the state of our economy -- already growing in value -- could become the next jackpot.