TCS Daily

Futures Shock

By James Pethokoukis - July 30, 2003 12:00 AM

You would think that if any member of the Senate Armed Services Committee understood a thing or two about futures markets, it would be the Cattle Queen of Arkansas, Hillary Rodham Clinton. Not only did Clinton display Soros-esque market timing during her days as a futures speculator, but she's a native of Chicago, home of the Chicago Board of Trade and the Chicago Mercantile Exchange. (It was the Merc where she famously turned $10,000 into $100,000 on cattle futures.) But the junior senator from New York was derisively dramatic when she described the Pentagon's proposed futures market for predicting terrorist strikes as a "futures market in death" as if it were a financial version of the death match in Stephen King's Running Man.


But Clinton was hardly alone on the panel in her apoplectic criticism. Senate Democratic Leader Tom Daschle denounced the trial program sponsored by the Defense Advanced Projects Agency as "an incentive to actually commit acts of terrorism." Clinton and Daschle's Senate counterparts in the GOP -- the party with the reputation of having greater faith the wisdom of markets -- saw a political Perfect Storm brewing and quickly put the kabosh on the Policy Analysis Market project which was supposed to start signing up potential traders on August 1 and be up and running on October 1.


This is just the sort of nightmare that George Mason University professor Robin Hanson fretted about when he first proposed the idea for PAM to the DARPA. "My suggestion was to go for the big time, but my wariness with these sorts of things is that the bigger thing you go for, the more you risk stepping on someone's toes or offending someone," Hanson told me.


And PAM was a pretty ambitious thing. As planned, the site would have allowed traders to purchase futures contracts based on various economic and political scenarios in the Middle East -- including U.S. recognition of a Palestinian state, GDP growth rates and political unrest -- and, more controversially, specific events such as the assassination of national leaders and the commitment of terrorist acts. From the trading patterns, the Pentagon hoped to assign probabilities on all of the above geopolitical happenings to better inform policy decisions. And think those senators are raging now? They probably would have been even madder had they realized PAM would have even allowed bets on the number of potential deaths from terrorist acts, as Hanson told me.


Clearly the politics of the project were a bit dicey, as Hanson realized. The only thing it lacked as a total public-relations disaster was a creepy logo, like that pyramid/eye combo from the now-defunct Total, I mean, Terrorism Information Awareness program. (Interestingly, retired Adm. John Poindexter of Iran-Contra fame, runs the DARPA department in charge of both programs.)


In a DARPA press release announcing the termination of the program, the agency said that "in light of the recent concerns surrounding the program, it became clear that it simply did not make sense to continue our participation in this effort." But at the same time, the statement seemed to imply that idea behind the PAM market was a sound one. "Futures markets have proven themselves to be good at predicting such things as elections results; they are often better than expert opinions," it stated. Indeed, if you perform a market-based analysis of the program -- evaluating its potential risks and potential reward -- the "terror market" really does make a lot of sense. First, the two major criticisms:


It's unseemly or in bad taste. The "yuck factor" seemed to be the crux of Clinton's critique, that bets on life and death are inherently and unforgivably repulsive. Yet the entire life insurance industry is based on carefully researched bets on the likelihood of policyholders living to a certain age based on various statistical criteria. And what about the "post-life" or funeral services industry dominated by publicly traded companies like Service Corp. and Stewart. If I am investing in these stocks, I am implicitly betting in favor of busy funeral homes and betting against any miracle cures or life-extension technologies occurring any time soon. Heck, the entire stock market for the past two years has involved in a massive bet about the chances of more terrorists strikes and their possible impact on the economy, with traders betting right after 9-11 that the worst was far from over.


Terrorists could profit from it. This is certainly true. A terrorist group could bet on a terrorist act happening at a particular time period and then go out and make it so. Of course, they can already do this through the existing oil futures market (by going long and then attacking the Saudi royal family, for instance) or even though the stock market. A group could buy put options on the S&P 500 and then attack the New York Stock Exchange or short Disney and attack one of its theme parks. Let your -- or Tom Clancy's -- imagination run wild. And recall, there were rumors that terrorists shorted airlines stocks before 9-11. So the PAM doesn't really present any opportunity that is radically different from what's already out there -- plus the existing financial markets provides a lot more financial opportunity than the $1 a contract payoffs in the PAM market.


Sort of the flipside of this argument is that the terrorists could bet that an attack wouldn't occur, thus pushing prices down and sending a misleading signal to the Pentagon. I asked Charles Polk, CEO of Net Exchange, one of the private companies involved in setting up the program, about this very issue. Polk pointed out that to really push a price, the terrorists would have to bet in volume, which would mean drawing attention to themselves and creating Internet and financial footprints that would provide clues to their whereabouts. In a way, the possibility of attracting actual terrorists to the site would be a good thing; easier to find them in an Internet café than in a cave or that now infamous border region between Pakistan and Afghanistan.


So what about the potential upside of PAM? As Hanson says, "If it has a large potential payoff and it doesn't cost that much to try, why not try?" Indeed, who cares about a "yuck factor" or terrorists pocketing a few grand if thousands of lives could be saved? The real question, then, is whether information markets like PAM really provide useful information? It seems they do. As Hanson notes in his research of information markets, racetrack market odds, for instance, have been found to improve on the prediction of racetrack experts. Similarly, orange juice commodity futures improve upon government weather forecasts. And political betting markets like the Iowa Electronic Markets consistently beat major national opinion polls in predicting U.S. presidential election results. Betting markets even beat Hewlett Packard official forecasts six times out of eight at predicting HP printer sales.


Then there's the recent Stanford study of "Saddam Securities," future contracts traded via an online betting exchange called Tradesport, based in Ireland. Before Gulf War 2.0 started, participants traded in Saddam Securities by buying and selling a contract that paid $10 if "Saddam Hussein is not President/Leader of Iraq by [Date]," and zippo otherwise. The researchers found that the security provided "a plausible estimate of the probability of war."    


As Hanson concludes in his analysis of these markets: "Betting markets, and speculative markets ... seem to do very well at aggregating information. To have a say in a speculative market, you have to 'put your money where your mouth is.' Those who know they are not relevant experts shut up, and those who do not know this eventually lose their money, and then shut up. Speculative markets in essence offer to pay anyone who sees a bias in current market prices to come and correct that bias."


But these markets seems to do pretty well even if only fake money is at stake. The Foresight Exchange, around since the mid 1990s, allows traders to make bets on terrorists attacks -- and pretty much anything else -- with pretend money. And in a study of its predictive prowess, Douglas Hubbard, a risk consultant in the IT industry, found that when the Foresight Exchange markets said an event had a 30 percent or 50 percent or 70 percent chances of happening, the outocme pretty much fit those forecasts. Ken Kittlitz, co-founder of the Foresight Exchange, told me that "even though we only use play money, people try to bet rationally because they feel inside that they have their reputations on the line."


And in that lies the possible salvation of a futures market that could help analyze the probabilities of future terrorism. Use pretend money. Although it might seem that it would take real dough to achieve a high level of accuracy, Hubbard told me that in his analysis of such markets, having actual money at stake only marginally improves the accuracy. Of course, while that would derail the "terrorist can profit" argument, some critics would still see the concept as unsettling. Yet one look at the news headlines -- such as reports today that terrorists may be planning more suicide hijackings -- reminds us that there are problems far more unsettling in the world, problems worth risking some visceral discomfort in order to stop or prevent.


James Pethokoukis is a senior writer at U.S. News & World Report where he also writes the Next News science and technology column for the U.S. News web site.


TCS Daily Archives