Yahoo! Research - Yahoo! - More Y!

Dr. David M. Pennock Email: PENNOCKD   -a@tt-   YAHOO-INC   d.o.t   COM

Y!IM:

Home · Summary & Contact · Blog · Publications · New and Noteworthy · Curriculum Vitae Questions · What is Yahoo! Research? · What do you do there?

Commentary: The Good Side of the 'Terror Futures' Idea (Yes, There is One) Note: The opinions expressed here are my own, and do not necessarily reflect the opinions or views of my company or any other person or organization. Given the uproar over the so-called terror futures market, you might be wondering what those mad scientists at DARPA were thinking. Believe it or not, there really is another side to this story, though the initial media firestorm did not uncover it, as both Democrats and Republicans with very cursory knowledge of the plan tried to distance themselves as quickly as possible. In the aftermath, several excellent commentaries and articles from well-known academics and journalists have been written defending the market, or at least the core ideas behind the market. Here are some voices on the other side of terror futures issue, from... The New York Times The Washington Post Fortune Newsweek Slate Wired (I am quoted here) Wired again (I am quoted here too) CNN (and a 2nd, more humorous, article from CNN) The BBC The Washington Post again The New York Times again The Mercury News The San Francisco Chronicle Federal Computer Week Economic Principles The Lighthouse Op Ed in The Lighthouse Several bloggers have also joined the chorus Here is a great compiliation of information by Robin Hanson, one of the primary scientists behind the market proposal (If you're only going to read one thing, stop now and read the first NYTimes article above, by Hal Varian.) My two cents First note that futures based on terrorist activities were to play a very small role in the market, or perhaps none at all. The main thrust of the market was to predict various quantities relating the the economic health and stability of countries in the middle east. Yes, there are legitimate concerns about a completely uncontrolled terror futures market. But with strict investment limits, careful record keeping, proper audit trails, carefully crafted financial instruments, and possibly even restricted access to security-cleared personnel, there would have been little risk, and great potential benefits. Yes, it may have failed. This was a research project -- an experiment -- and research sometimes fails. But it may also have succeeded in delivering superior information and superior forecasts about extremely important world events, something most can admit appear lacking using current methodologies. The Policy Analysis Market (PAM), dubbed "Terror Futures" by congressional leaders and the media, was based on decades of research in economics and finance that markets provide accurate predictions: more accurate than most other forecasting tools we have. Markets offer a way to tap into the collective wisdom of a population. Markets don't care who is the loudest, who is the most aggressive, who is the most powerful, or who is the most persuasive; markets only care about accuracy. Those who have better information and can better predict the future win. Those with poor information or poor abilities to process information lose. In this way, the good information filters to the top, yielding forecasts that reflect about as good a vision of the future as can be found by any other means. This kind of cold indifference to personalities is often exactly what is needed to uncover the truth. The ability of markets to provide unbiased forecasts is not just speculation: it is borne out by countless scientific investigations of real markets, including options markets, futures markets, sports betting markets, and even market games where no real money changes hands. PAM also could have provided a means for people and organizations to insure or hedge against major world events. Oil companies, for example, are very much exposed to the risk of instability in the middle east. A market like PAM would give them the opportunity to hedge those risks. A major objection to PAM was that terrorists or their sympathizers could manipulate the market to divert attention from their true targets. This is possible, but the evidence from existing markets is that manipulation is very hard, because markets punish misinformation. Those attempting to artificially inflate some investment stand to lose significant sums of money, and may not even effect the market price if one or two smart other traders realize what is going on. In fact, there is evidence from the Iowa Electronic Market, a futures market in political elections, that a coordinated attempt to bid up the price of Pat Buchanan shares failed miserably, barely effecting prices, and causing all the recruited Buchanan supporters to lose their collective shirts. A much more serious objection to PAM was that terrorists themselves might benefit by enacting the very thing they bet on, thus funding their next terrorist act. This threat was overblown, since the market would have limited investment to $100 or so. Terrorists looking for ways to profit on their actions would do much better to short sell airline stocks or oil stocks, or otherwise exploit existing markets that have effectively no limits on the amount that can be invested. There are plenty of ways to manipulate markets and commit fraud, including, say, perpetrating arson to collect on insurance. But because there are legitimate reasons for stocks and for insurance, we don't abolish them just because of the potential for fraud. We instead monitor for fraud and punish it severely when discovered. It is an unfortunate truth that we sometime resort to paying large sums of money to unsavory characters in order to capture other unsavory characters. It's true that PAM might have benefited some undesirable elements, and perhaps this is enough to shut it down in its original form. But their benefit would have been limited, and our benefit could have been great. Note that, contrary to reports, the government would not have directly paid anyone: any one trader's winnings would have come directly from some other trader's losses. Then there is the moral issue: isn't it simply wrong to place bets on death and destruction? Well, if so, nearly everyone in the civilized world is guilty: we all bet on misfortune every day. When you purchase life, health, or property insurance, you are essentially betting that something bad will happen. Your insurance company is taking the other side of the bet. If nothing happens, the insurance company keeps your payment. If something terrible does happen -- you wreck your car, or require major surgery, for example -- you "win" the bet and the insurance company pays you your due. Of course you don't actually wish for tragedy to strike, but if it does, you want the peace of mind to know that you will receive monetary compensation. In fact, large building owners are now taking out terrorism insurance: in doing so, they are in effect betting on terror, albeit in a way that almost no one would object to. I can understand the shock and outrage people felt when they first heard about PAM. Had I not known anything about the project, I likely would have reacted the same way. But knowing something of the other side, and seeing the incredible firestorm burn out of control so fast, with nearly all coverage and commentary coming in exceedingly one sided, was both frustrating and frightening to behold. It certainly makes me think twice about all those seemingly outrageous scandals I hear about that I otherwise know little about. It's a shame that PAM -- and all the research programs surrounding it, most having nothing to do with terrorism -- were dismissed out of hand in a single day with little discussion and little investigation into their merits. The bottom line is that PAM, in some form, might have worked. I certainly hope that other academic and commercial projects investigating the predictive power of markets will proceed. Related Links These lists are by no means exhaustive -- consider this section "under construction" Real markets: Real forecasts The proof of the forecasting power of markets is in the pudding. Beyond our standard financial derivatives markets (e.g., options and futures markets), here are some interesting real-money markets with a proven track record of predicting the outcome of a huge number of events spanning a wide variety of topics. The Iowa Electronic Market (IEM). A not-for-profit market operated by the University of Iowa Tippie College of Business. Known for its political election markets, IEM also offers a markets on a few other topics, for example the Federal Reserve's monetary policy. TradeSports. Offers markets on a wide range of topics, including sports, politics, and news. Cantor Index. Cantor Index is a UK-based subsidiary of Cantor Fitzgerald. Cantor Index now owns the Hollywood Stock Exchange (HSX), and runs a real-money version of HSX's play-money market game. Many other so-called spread betting markets are run in the UK and elsewhere, allowing traders to bet real money on where various market indices are headed, or how sporting events will unfold. The World Sports Exchange. Mainly limited to sporting events, but also offers markets in elections (e.g., the 2004 Democratic Nominee), the Oscars, American Idol, etc. The site is also a sportsbook. To get to the markets, from the above link click on "Interactives" or "Futures Markets" on the left navigation panel. The novelty of the sports markets on TradeSports, World Sports Exchanges, Cantor Index, other spread-betting sites, and other so-called "in-game" or "in-running" or "peer-to-peer" (P2P) sports betting sites, is that trades/bets can continue even after the game starts, right up until the final moment. So as information bearing on the outcome of the game is revealed -- points are scored, players are injured, penalties are called, etc. -- traders can react in real time. Prices reflect the traders' current collective sentiment on the likely outcome of the bet at any given moment. Play-money market games: These traders aren't kidding around A variety of play-money market games are cropping up on the Web. They operate like real markets, but no real money changes hands, thus side-stepping thorny legal and regulatory issues. Research shows that even market games can provide decent forecasts, sometimes outperforming experts or polls. It seems players care dearly about their fictitious portfolios, and the lure of funny money -- fueling community stature within the game, or sometimes convertible into very real prizes -- is enough to drive a form of informational efficiency and forecast accuracy in these games, if not as strong and well-tested as with their real-money brethren. The Hollywood Stock Exchange (HSX). Probably the most popular and well-known market game. Offers markets in movie box office results, actor market value, Oscars, Emmys, and more. NewsFutures. A well-designed and fast growing market game covering anything that is newsworthy, including politics, current events, entertainment, and sports. There are also French and Hungarian versions of the site. The Foresight Exchange (FX). The granddaddy of market games. With an old-school interface, this site, inspired by Robin Hanson's Idea Futures concept, is focused mainly (though certainly not exclusively) on events of scientific and technological import. FX is the only site with a mechanism for users to craft and contribute their own market contracts. (A few) People, Organizations, and Companies Joyce Berg, Thomas Gruca George Neumann, Robert Forsythe, Forrest Nelson, and Thomas Rietz -- IEM Board of Directors, University of Iowa Tippie College of Business Leonard Brody -- Ipreo Alexander Costakis -- Hollywood Stock Exchange Ely Dahan -- UCLA John Delaney -- TradeSports Lewis Findlay -- Cantor Index Leslie Fine -- HP Labs Michael Foster -- DARPA Carol Gebert -- Incentive Markets Robin Hanson, Vernon Smith -- George Mason University Ken Kittlitz -- Foresight Exchange Anthony Kwasnica, Lee Giles, Tracy Mullen -- Penn State John Ledyard, John Lin, Charles Plott -- Caltech Jim Neiers -- Neoteric Technologies Charles Polk -- NetExchange Carsten Schmidt -- Max Planck Institute for Research into Economic Systems Emile Servan-Schreiber -- NewsFutures Leighton Vaughan Williams -- The Nottingham Trent University