Is there a future for futures trading?

Betting on terrorism is bad public relations, but is it an important tool in shaping anti-terrorism policy? Some experts say that a Defense Advanced Research Projects Agency futures trading market similar to the system that works for orange juice or petroleum markets could have gone a long way toward identifying likely future terrorist activities — it if had not abruptly gone down in flames last summer.

"With the safety of the American people at risk, it would be irresponsible of government officials to not actively explore a promising idea like this one," said Thomas Malone, a professor of management at the Massachusetts Institute of Technology. "Failure to explore it puts us at greater risk."

Known as the Policy Analysis Market, the electronic forum would have allowed people inside and outside the government to express opinions about events unfolding in the Middle East that relate to military activity, political stability, economic activity or U.S. policies. But the fledgling research project was canceled last August after a firestorm of criticism from two senators who charged that traders could profit on others' misfortune.

If the idea of a futures market is to overcome such criticism, military and homeland security officials must learn some important lessons. Topping the list is understanding that even for projects that have a large information technology component at their cores, the biggest challenges often relate to communications and managing people, not technology.

The potential is clear: Futures markets offer an efficient way to collect reliable information, proponents say.

"You may have a bunch of people in a government bureaucracy who may be smart and know a lot, but they also know there are certain things their bosses want to hear," said Robin Hanson, assistant professor of economics at George Mason University and a key contributor to DARPA's effort. "A futures trading system provides more incentive to be honest because of the monetary rewards. When you know something, you say what you think. When don't know something, you don't say anything."

DARPA's program was similar to the University of Iowa's Iowa Electronic Markets, which has tried to predict the outcome of U.S. elections for the past 16 years. School officials say the market has been more accurate than election polls 76 percent of the time.

Policy Analysis Market officials sought to look at a much wider range of contingencies, Hanson said. "Some of the questions we wanted to ask were very conditional," he said. "If we move our troops out of Saudi Arabia, how does that change political stability there and in other countries? Once you start asking these kinds of questions, the number of additional questions increases exponentially."

Hanson addressed the problem with what he calls combinatorial trading technology, which would have allowed traders to manage hundreds of thousands of interrelated parameters by browsing a DARPA Web site for prevailing estimates in each category.

"Traders would browse different countries, different quarters, different estimates," he said. "When you find a number you think is wrong, you tell [the system] one you think is better, and the system tells you what kind of bet goes along with that opinion. If it turns out your estimate is closer to the truth, you make money. If it's farther from the truth, you lose."

Bets would not have exceeded $100 and trades would not predict individual terrorist attacks, he added.

Although the number of parameters is large, the technology to handle trading is relatively simple. "It's not a matter of developing some complex [artificial intelligence] application," Malone said. "Several companies have developed such systems."

In hindsight

If experts believe futures trading is viable in the anti-terrorism context and the technology to build the market is readily accessible, why did the Policy Analysis Market come to such an abrupt end?

One reason was its openness to traders outside the intelligence community, said Forrest Nelson, professor of economics at the University of Iowa and a founder of the Iowa Electronic Markets. Nelson worked on the Policy Analysis Market as part of a small firm university officials created for that purpose.

"We wanted to set up a market on terrorist attacks, but we wanted it done within the confines of the security agencies," Nelson said. "We wanted traders from the CIA, FBI and maybe Israeli intelligence — a relatively small number of traders with private information. The only people who knew about it would be people who observed or traded in this market."

Such secrecy would have avoided the image problems that led to the program's demise, he said. The DARPA project's biggest challenge was effectively communicating the connection between futures markets and terrorism. Public relations problems surfaced because a Web site describing the effort provided eyebrow-raising hypothetical trade examples, such as the assassination of Palestinian leader Yasser Arafat and a missile launch by North Korea. Although project organizers said those examples did not reflect the true nature of the idea, they created the political firestorm that killed the effort.

"You have to be very careful about what you present about a government project because you don't get any chance to defend yourself," Hanson said. "An accusation can be made, and people will simply respond to that accusation with a rush to judgment."

Another program participant, who requested anonymity because of the controversy, blamed political naiveté for the program's demise.

"The way it was portrayed in some fashion merited the reaction," the participant said. "But it was not portrayed correctly. What was canceled was not what was being built. The idea became part of a larger set of activities that go on in Washington [D.C.] that are not connected with main intent."

The lesson? "If you want to save something, don't put great big targets on it," the source said, "not if you want something to have any shot in hell of succeeding."

Joch is a business and technology writer based in New England. He can be reached at ajoch@monad.net.

***

Policy Analysis Market

Project: Policy Analysis Market, sponsored by the Defense Advanced Research Projects Agency.

Purpose: A Web-based futures trading forum to gather expert opinions about possible future terrorist activity.

Status: Canceled August 2003.