Online exchanges let political junkies bet on candidates they think will win

Tuesday, October 26, 2004

For Jonathan Sargent, any talk about a presidential candidate's political stock rising or falling is no mere figure of speech. The 21-year-old University of Iowa senior has bet $75 that President Bush will win re-election, purchasing stock in the incumbent in an online stock market that lets investors invest their money in shares of the candidate they think will win.

		Graphic: Bush vs. Kerry stock performance

Such online stock market schemes in which presidential futures trade like pork bellies do on the Chicago Board of Trade or shares of General Motors do on the New York Stock Exchange may be just for fun for people such as Bush-backer Sargent -- and a way to pick up a few bucks. If Bush wins, Sargent stands to double his investment. But to a cadre of academics and economists who believe markets are likely more telling than polls, the way people are betting with their money could be a tell-tale sign of who the likely winner will be come Nov. 2 .

"We're about twice as accurate as polls," said George Neumann, director of the Iowa Electronic Markets, or IEM, where Sargent invested in Bush futures. Run by the University of Iowa's Tippie College of Business, the IEM is regarded as the grandfather of political betting sites. Yesterday, it recorded a 4 p.m. EDT closing price of 61 cents per share for Bush's stock and a 39-cent close for Sen. John F. Kerry's in a "winner take all" contest in which traders will receive $1 for each of their holdings if their candidate prevails.

In other words, the markets are giving Bush a 61-to-39 percent edge over Kerry, a fatter margin than currently reflected in most polls, which show the outcome as a toss-up. However, another set of IEM stocks that reflect investors' bets on the share of the vote each candidate will get have the candidates neck-and-neck. The data indicate the markets are showing some conviction about a Bush win, but also betting the race will end in a photo finish.

Neumann, an economist, was one of three professors who founded IEM in 1988 after seeing pundits flummoxed that year when Michigan's Democratic presidential primary contradicted polls and crowned Jesse Jackson its victor. The Iowa academics wanted to find out whether markets -- as they've been proven to be in a number of economic matters -- would be a good if not better predictor of political futures than polls.

The Bush and Kerry campaigns have expressed little interest in the site, and a spokesman for the Gallup Organization, one of the nation's best-known pollsters, said the firm is fixated on making sure its methods are accurate as possible and does not pay attention to online exchanges and betting sites. But in academic and economic circles, the market data has a keen following.

To Sargent, the University of Iowa senior, the IEM is "getting students who might not usually pay attention to political campaigns to focus more on what the real issues are and to follow the campaigns more closely." The IEM exchange on the Bush-Kerry race is open to anyone but investments are limited to $500 and the exchange takes no commissions. The IEM has more than 3,000 active traders, the majority of whom are neither students nor faculty.

Some other political exchanges and wagering sites don't limit investments and some do take commissions. Some are run like casinos, with the house setting odds. Others are run like financial markets, letting traders establish prices. Some use play money instead of the real thing and some, such as TradeSports.com and InTrade.com boast tens of thousands of traders and even more followers. Most such sites are run offshore so as not to run afoul of state and federal gambling laws.

While lots of people trade or check online political stocks purely for fun, some do so because the outcome of presidential elections can have huge impact on the fortunes of different companies and on the overall economy.

"There are lots of people sitting in Wall Street firms looking at them because people really do believe what party is in charge will matter to business interests," said Robin Hanson, an economist and markets researcher at Virginia's George Mason University.

Few of the sites outside of IEM have been around long enough to suggest they are better gauges of public opinion than polls, but that doesn't stop others from asserting that market data generally is more accurate.

One is Mike Knesevitch, a former hedge fund manager and trader who now serves as marketing director for InTrade.com. He mocked the scant post-debate samples some pollsters used. "If you wanted to find out the probability [of a particular candidate winning], are you going to look at a public opinion poll with 24 likely voters or would you look at an exchange with 40,000 individuals who put their money where their mouth is?"

Virtual or otherwise, betting on political futures has a long history, some of which played out on or near the nation's traditional financial exchanges.

In an academic paper, economist and market researcher Koleman Strumpf at the University of North Carolina said that a large, active and highly public market for betting on presidential elections existed throughout much of U.S. history before World War II. Among other places, traders placed their bets in the lobby of the New York Stock Exchange. Gambling laws and polling that put an end to the public political exchanges, he said.

The reappearance of formal markets to forecast political outcomes doesn't surprise Strumpf in light of their track record between 1868 and 1940, the period he studied. "There were no polls. News traveled by telegraph, so it was very slow to reach people and yet, the markets worked," he said.

Despite the wider lead markets such as IEM and TradeSports give Bush than polls do, the race still is close, Stumpf said. "The markets say Bush has about a 60 percent chance of winning. They're definitely picking Bush, but there's some uncertainty." He noted that at the same point in the 1996 race between President Bill Clinton and Bob Dole, Clinton's stock was selling for 94 cents, meaning the market gave him a 94 percent chance of victory. "It was clear the market knew he was going to win," Strumpf said.

Strumpf also said online political markets thus far seem immune to manipulation by party operatives, which was not the case with their historical forerunners. But just as in the past, there appear to be attempts.

For example, although Bush has been maintaining steady lead over Kerry for weeks of virtual trading, a large mid-month bet crushed the incumbent's stock, to 10 cents on TradeSports.com, after it had hovered around 60 cents for days before. Strumpf and other academics suspect the dive was driven by vested interests, but because "it only lasted five minutes," it suggested to analysts that the markets aren't easily fooled, he said.

That episode got to the heart of the reason market theorists generally believe that market data has a good chance of outperforming polls. Polls typically ask likely voters which candidate they'll back. Markets ask speculators or investors, who must gather information to enhance their odds, which candidate will win. Indeed, IEM research suggests that traders often cast their ballots for different candidates than the one on which they stake their bets.