Iowa Alumni Magazine | October 2008 | Features

Predicting the Future

By Emily Grosvenor
Who knows which presidential candidate will end up in the White House after next month's election? A UI economics research tool just might.

After Barack Obama won the Democratic nomination, Ian Hoffman, 95JD, made bank. Not from an office pool, but from a high-tech economic teaching tool that for 20 years has enabled the University of Iowa to out-predict many traditional political polls.

In the run-up to the Iowa caucuses almost a year ago, Hoffman, an attorney with the U.S. Department of Justice in Cleveland, was taking in speeches by the Democratic contenders at small town events across central Iowa. No candidate impressed him more than Obama, whom he listened to in a middle school cafeteria in Nevada.

So, like any good citizen, Hoffman started learning more about the man. And, like any good capitalist, he bought in—picking up shares of Obama in the Iowa Electronic Markets' (IEM) Democratic primary contest when the senator was trading at about 30 cents.

Hoffman is just one of a cadre of young, ambitious, politically engaged UI alumni who trade frequently on the IEM, Internet-based prediction markets run by a group of UI business school professors that allow students and the public to trade "shares" of political candidates for real-money profit. Of the 1,000 active IEM participants, a few dozen are students using the markets in courses, while a few hundred are UI alumni.

Only a small fraction consists of frequent traders like Hoffman: smart, savvy guys who buy and sell to take advantage of the markets' fluid response to the news. While the stakes aren't especially high, the personal reward is immeasurable. "It's the idea that you believe you're smarter than everybody else and you can make a little money at the same time," Hoffman explains.

So far, Hoffman's gut feeling at the caucuses has proven prophetic in the IEM's general election contest as well. At press time, Obama is trading at 57 cents to McCain's 43 cents, meaning that IEM investors believe Obama has a 57 percent chance of winning the popular presidential election in November. Hoffman remains confident in his pick—although the race is far from over, and the markets respond constantly as conditions change.

Since their emergence in the early 1990s, prediction markets—for everything from movie blockbusters to Super Bowl champions—have been the darlings of pop economists, who see them as a more accurate way to anticipate the outcome of events than traditional opinion polls. They are glorified gambles—the office pool made dynamic.

No such market has been more watched, disputed, and praised than the U.S. Presidential Election Markets, developed by professors in the Henry B. Tippie College of Business. The markets work like this: traders buy and sell future "contracts" based on how they think a candidate will do in a given scenario, from the nomination process to the popular election. The minimum investment is $5 (the amount most students begin with), and the maximum is $500. The price investors pay for the contracts reflects the markets' best guess of a candidate's chance of victory.

"We can't tell you who's going to win," says Thomas Rietz, 88PhD, associate professor of finance and an IEM market manager. "We can only tell what the market is saying is the probability of a certain outcome."

More often than not, though, the IEM gets it right. Its success offers a homegrown example of a counterintuitive idea: groups of people—under certain conditions—are far from the chaotic, madding crowds many imagine them to be. New Yorker business reporter James Surowiecki outlined those conditions in his best-selling book The Wisdom of Crowds: Why the Many Are Smarter Than the Few and How Collective Wisdom Shapes Business, Economies, Societies and Nations. The book explores situations in which diverse groups of people acting independently, in their own financial interest, and regardless of their level of intelligence, can come to remarkably wise collective decisions.

So it seems to work with the IEM, which aggregates a diverse range of opinions and predictions. Put another way, the wisdom of crowds is what makes the IEM the country's most accurate political crystal ball.

These days, when the IEM sends out press releases of its investors' latest predictions, people take notice. But the markets were not always so well regarded. An early New York Times story called it "a bizarre experiment." Economics professor Forrest Nelson vividly remembers the moment he and his colleagues envisioned that experiment. In 1988, while discussing Jesse Jackson's unexpected win at the Michigan primary, the UI professors wondered, What if we gave people in the know an incentive to reveal their information in a market setting?

The UI wasn't the first institution to run prediction markets to illustrate economic concepts—but it was the first to institutionalize it and run it as a research project using real money. By mid-July of 1988, Nelson and his colleagues had persuaded 35 traders to invest. Back then, the professors ran trades on a single mainframe computer in the UI Main Library. In need of more traders to participate, they went to the economics department and asked to create a class based on the market. The result was a one-credit, dual economics/political science evening course. The instructors showed up the first night and found more than 100 people crowded into a room built for 30. They promptly cancelled class and regrouped to accommodate a bigger crowd.

"It served as a great teaching laboratory," recalls Nelson. Participating in the markets taught the economics majors much about politics. The political scientists, in turn, were astounded by how much they discovered about markets.

The economics department has been integrating prediction markets into its courses ever since. One class runs a prediction market to anticipate box office returns of a seasonal movie blockbuster, while a finance class runs a market to predict whether the Federal Reserve will lower interest rates.

Ever since the 1940s, economists have known that prices contain information about future developments within marketplaces. A political candidate's price on the IEM refers to how much confidence investors hold in that person, just as stock prices reflect investors' confidence in a product or a business.

In its inaugural effort, the IEM predicted the breakdown of the presidential election results to within 0.25 percent, easily trumping the polls. Since then, its U.S. Presidential Markets have provided students and traders with insights about how information is absorbed into the marketplace. In the market's early years, investors often responded to minor political gaffes and stump speeches. This year, political imbroglios such as Obama's Reverend Wright controversy—which dominated the news media for weeks this past May—have had only a minor effect on the IEM market standings.

As with any market, the trick is to buy low and sell high. Since the presidential markets include winner-takes-all, most traders buy contracts early in election season and just let their money sit until the end, at which point the investors who had money on the winning candidate receive a check in the mail from the university. The high-frequency traders, though, will buy and sell on a daily basis.

"I check it every day when there's a lot going on," says Kent Oliven, 92BS, 94MA, a frequent IEM trader and self-proclaimed "political news junkie" who works as a government finance director for the Village of Palos Park in suburban Chicago. "This is a hobby I have time for."

Successful traders learn to anticipate how the markets respond to events. They scour headlines to see which groups candidates are speaking to, and they guess at the content of candidates' speeches. They educate themselves about past policy decisions and the organizational structures that support most successful political campaigns. In some cases, they are insiders in the political arena who have the best information of all—the facts that aren't public.

The average active IEM trader is young (20s to 30s), white, and male with a connection to the UI business school, although the participants are becoming more diverse. All agree that the smart way to play is to put your money on who you anticipate will win based on the best information you have, not who you want to win. That factor helps make the markets more accurate than traditional polling.

For Oliven, the IEM's greatest value is how it gives traders an incentive to engage with their world. "You're never going to recoup the amount of time you spend playing on the markets," he says. "What I've learned about the world by reading newspapers has been as important as what I learned in any classroom."

Sometimes, though, the markets are a crap shoot. Even the most well-read traders experience the thrill of profit and the agony of loss. Hoffman won big when the UI ran a market for the Congressional race of 1994. He bought shares of future U.S. Congresswoman Karen Shepherd of Utah when they were trading at a nickel (such a trade, if the maximum of $500 were invested, would have created a windfall of $10,000). His worst loss occurred in the 2000 election, when he vacillated between Gore and Bush, weighing his personal preferences for the incumbent vice president over a newcomer he had an inkling might win. He put his money on Bush.

"I figured, if Bush wins, then I win," he says. "And if Bush loses, the country wins. Win-win, whatever happens."

That year, Bush and Gore were both trading at 50 cents right up until election night, meaning investors expected a toss-up. The equation wasn't as simple as Hoffman had predicted, though. Gore won the popular vote, but Bush won the election. The traders who had their money on Gore took the prize at the IEM.

"So, I lost, and the country lost," Hoffman says with a laugh.

For the next generation of markets, UI researchers are working on models that will focus not only on who will win an election but how that outcome will affect policy decisions. Other markets already take prediction to that next level of usefulness.

At the UI, an influenza prediction market open only to invited doctors, public health experts, and other health care providers aims to pinpoint which strains of flu will hit at certain times. Ideally, that early-warning information can help health officials and drug companies prepare an effective strategy against the outbreak.

The business school has also worked with private companies to develop ways to use prediction markets to foresee productivity and sales levels. In fact, such markets can be used in any business situation where a certain level of risk is involved and where employees might have an incentive to share good information.

As for the future of this unusual business tool, Nelson makes a calculated—and confident—guess: "I predict more and more use of prediction markets within industry. It's nice to see people put their money where their mouth is."