Washingtonpost.com July 31, 2003 Thursday 02:00 PM SECTION: LIVEONLINE LENGTH: 3553 words HEADLINE: Editorial Focus: Value in Terrorism Futures? SOURCE: washingtonpost.com BYLINE: Assistant Professor of Economics, Stanford Graduate School of Economics HIGHLIGHT: Eric Zitzewitz, assistant professor of Economics at Stanford University, discussed piece on the questionable value of terrorism futures. BODY: On July 29, 2003, the Pentagon discarded a program that would have enabled market assessments of terrorist strikes in the Middle East. What were the reasons behind the Pentagon's decision to terminate the plan? Should there have been a future for the terrorism futures market? Eric Zitzewitz, Assistant Professor of Economics at Stanford University, was online Thursday, July 31 at 2:00 ET to discuss his column ( http://www.washingtonpost.com/wp-dyn/articles/A5696-2003Jul30.html The Furor Over Terrorism Futures) on the questionable value of terrorism futures. Editor's Note: washingtonpost.com moderators retain editorial control over Live Online discussions and choose the most relevant questions for guests and hosts; guests and hosts can decline to answer questions. ________________________________________________ washingtonpost.com: Welcome to Editorial's Live Online with Eric Zitzewitz. ________________________________________________ Washington, D.C.: In paragraph five of your editorial, you state that "we don't have the same doubts about financial markets." If by "we" you mean yourself and Mr Wolfers, fair enough. But if by "we" you mean the American people, you may want to read any of thousands of editorials over the past year bewailing the fact that Americans have no faith, no confidence in financial markets, after seeing them exploited and abused by corporate criminals. Why should a tool so deeply flawed that it produced major crashes in 1928, 1987, and 2001 (to name but a few) be so good at predicting the future in the Middle East? In paragraph one, you state that saving lives mitigates moral questions. Is this not the rationale that was used by the Federal Government in justifying the racist Tuskeegee syphilis experiments and testing nuclear weapons on American soldiers in the 1950s? And, for that matter, used by the Nazis to justify freezing Jews to death in order to study hypothermia? Could you expand on your rationale, please? Eric Zitzewitz: I think part of the disconnect between economists and non-economists on these issues in due to the bubble, as you suggest. In 2000, many economists were tearing their hair out over the media and Congress's apparently blind faith that, if the Nasdaq is 5000, who am I to argue? Now opinion has swung in the other direction. But once every 30 year bubbles are the exception rather than the rule. I think your the analogies you draw are a little out of proportion. We clearly were not saying anything like "the ends always justify the means." ________________________________________________ Portland, Ore.: In general I would say that the current futures market, derived from experiences with agriculture, do work. However I agree with our political leaders that the idea of "terrorism futures" is purely absurd. Haven't economists "debunked" the efficient market theories? Robert Samuelson wrote in his July 24th column a kind tribute to Charles Kindleberger, who documented cases where the market was wildly inefficient.And with the recent stock market bubble, I don't see many dissenting views. Why wouldn't "terrorism futures" be subject to such bubbles? And wouldn't that defeat the whole point of it? Eric Zitzewitz: One thing that's gotten lost is that "terrorism futures" is a bit of a mislabeling of what these guys were up to. Almost all of the suggested securities I saw on their Web site (and they probably made matters worse by pulling it down) were about political events. Sadly, in that region, removal from office often comes via terrorism, and so I think some of the securities may have asked about assassinations. But for the most part, they seemed to be simply extending the Iowa election markets to another context. I've got no idea if it would have worked, but it's worked pretty well on U.S. politics. See above on the bubble. Also, part of why the bubble didn't correct is that there was so much money flowing in from Motley Fool types that value investors were overwhelmed. This seems unlikely to happen in this market -- the amounts are stake are just too small. ________________________________________________ Somewhere, USA: This could have been just an off-the-wall proposal, but did it (notwithstanding who knew about it beforehand) reflect a mind-set perceived as permeating the Administration? The attitude that terrorism against the U.S. (as opposed to terrorism against other countries, that's been going on for years), excuses any activity, including imprisoning people for months or years without charges because it's the most effect way to elicit evidence? I see more callous cynicism among some-- not all-- members of this Administration than I've seen since the Nixon White House. It scares me. Overreacting? Eric Zitzewitz: I have the same feeling sometimes about other administration efforts. But I think this project is more a reflection of genuine efforts to come up with a creative, innovative solution. Even if this one doesn't fly politically, calling for their heads, as a couple Senators are intent on doing, seems counterproductive. ________________________________________________ Bethesda, Md.: It astounds me that this administration would come up with such a thing. Why would you make decisions based on what a trader on K Street or Michigan Avenue thinks, or really, is guessing? Shouldn't intellegence be based on something substantive, instead of some greedy person's opinion? It would make more sense to get intelligence from a kabob vendor in Tehran than some day trader in D.C.! Eric Zitzewitz: If this market worked (and I don't know that it would have, but thought it was worth a try), it would be because kebob vendors, or people who talk to them, like journalists or U.S. or European relative, were participating. ________________________________________________ Washington, D.C.: Eric, A couple of questions from Jerry Knight, business columnist for The Washington Post: Your op-ed repeats the oft-heard comment that the futures market is better at predicting the weather in Florida than the weather service. Can you tell me who did that research and where it was published? Second it's a truism in the futures market that the vast majority of traders lose money. Which means they're wrong. How does that square with assertions about the predictive ability of the markets? Eric Zitzewitz: Jerry, The reference is: Orange Juice and Weather, Richard Roll, American Economic Review v74, n5 (December 1984): 861-80. I believe that most retail futures trader lose money primarily due to commissions and bid-ask spreads. So this is pretty consistent with market efficiency -- these retail traders think they can spot trends, but they end up breaking even on the actual trades, and lose money on the transaction costs. Now U.S. commodity markets are already so efficient that its hard for amateurs to beat them. The Saddam security market was also quite efficient, although sometimes information was incorporated with a short lag. So there were profit opportunities for the quick, but they closed fast. How efficient these markets would have become is an open question. ________________________________________________ Kennesaw, Ga.: I have no idea whether the Policy Analysis Markets idea was a good one or not. My question is how long were the Senators who denounced it aware DARPA was asking for it to be funded? The tone of their remarks strong suggested they hadn't been aware of it for very long. Knowing how much politicians hate being surprised, I have to think PAM advocates within the administration screwed up. Also, I wonder if the association of this idea with Adm. Poindexter damaged its prospects. Your thoughts? Eric Zitzewitz: I don't know the background of how Senators were told of it, but agree they seemed angry about being surprised. Fortune did do an article on it in March or April. Probably right about Poindexter. A lot of people may have been surprised to hear he was back in government. ________________________________________________ New York, N.Y.: We heard a story back in the early 90s that Saddam (or other higher Iraqi officials) made a ton of money from oil futures because they made a big bet on oil price rises before invading Kuwait. Is this true or a mere urban legend? If true, doesn't it point to the possibility of market manipulation by terrorists or the like? Eric Zitzewitz: Yes, and there was the story about Osama shorting Boeing and United before 9/11. The point we make is that these markets don't create new opportunities in this area, since you can make far more money in existing markets. What they might do is make the market signals easier to interpret. ________________________________________________ New York, N.Y.: I don't understand how this makes economic sense. I've heard the wonderful abilities of futures markets to predict better than analysts or pollers, but isn't this success contingent on a marketplace where information is disseminated broadly and efficiently? Aren't futures market with only a few participants, with a monopoly of information, easily manipulated? And when these few participants, won't there be an even greater incentive to manipulate? Eric Zitzewitz: I actually think the terrorism markets part of this project is less interesting than the general political markets part. If Arafat is losing the support of the Palestinians, there will be lots of little micro-signals than journalists, etc. will be picking up on. If the right people play in these markets, it could pick up those signals. That said, the journalists would need to embrace the basic idea. With terrorism, you're right. Most of us know nothing, and a few people have inside info. All you could hope for is that someone involved in the project uses the market to make a few bucks, and that tips you off. Seems like a long shot, but who knows? ________________________________________________ Phoenix, Ariz.: It seems clear that the main problem with the proposed market was its appearance -- it was too easy for politicians to take outrage at, like Proxmire used to do with his "Golden Fleece" award. Is there any way, now that this has been "exposed" and trashed, that the government can still make use of this kind of technology for intelligence purposes? Eric Zitzewitz: Its quite possible that, given all the publicity, the private sector will pick up on this and run some of these markets. Perhaps that's the way it should have been done in the first place -- it avoids all the bad associations with the Pentagon for some people. ________________________________________________ Washington, D.C.: Professor Zitzewitz, Great column today. This is a very interesting topic. It seems as though this is more a political dilemma than anything else. In your closing sentence, we can sense (and understand) your frustration with the politicians. Maybe it all comes down to symbolism and representation. The intent and idea sound solid. Do you think there's a way to frame this idea so that it can be perceived as morally acceptable, and so the politicians can't use the issue as political ammo? Instead of blatantly investing/betting/gambling on futures, could some variation of the fantasy sports leagues be a viable model? Instead of players, you could draft targets. And their value could fluctuate according to the so-called market. Maybe this could be more politically acceptable? Unfortunately, that seems to be key at this stage of the game. I know next to nothing about economics, so all the accuracy in the original futures idea could be lost in a fantasy league model. Anyway, if we can develop some new benign tool for improving our terrorism predicting power, we can't let politics kill it. Very interested to get your thoughts on alternatives to the Pentagon betting model. Thanks Eric Zitzewitz: A fantasy team of assassination targets? OK, even I think that's a little over the line. One way to reframe the idea might be to move away from assassination and terrorism and limit it to less emotionally loaded political events. ________________________________________________ Washington, D.C.: The fact that the Pentagon entertained the idea of this program seems to be a backhanded statement about our intelligence. Could you comment on the status of our intelligence and the need for programs like the futures market? Eric Zitzewitz: Let me make sure we were clear. I think people in our intelligence agencies work extremely hard for well less than they could make in other occupations. *We* were certainly not making any backhanded statements about their abilities or efforts. Our main point is that their job is a tough one, and we thought there was a chance that this project would help. ________________________________________________ New York, N.Y.: It seems like the private sector can step in and fill the void left by the DoD for a futures market on terrorist and political markets. Is there any reason to believe that this would be a lesser solution to the one proposed by DARPA? Wouldn't the DoD get its market without the crying of Congress? Eric Zitzewitz: That's not a bad idea. It's possible though that there are some markets that are worth running from a policy perspective that a private company couldn't make enough money from running. Clearly, there was enough demand for a market on Saddam or the Iraq war that the private sector provided. But a coup in Jordan? ________________________________________________ Corvallis, Ore.: The examples of futures markets working that you cited all have one thing in common: highly uniform information is available to the public. With election predictions, we all have access to the same newspaper articles, tv stories, and polls. With printer sales forecasts, all participants have access to the same market data from places like gartner, etc. However, with terror futures, the information is highly non-uniform. There are people who are experts that know a lot about particular political situations. Most people have very little knowledge about potential terrorist attacks, and their estimates of the probability of various events would be purely speculative. Whereas, just a few people with inside information could profit handsomely without significantly moving the market price. How is the terrorism futures market supposed to work with the absence of uniform information? Eric Zitzewitz: Good points about info. Part of why I mentioned earlier that the policy markets (i.e., is Arafat losing influence) may work better than the terrorism ones (i.e., is there a plot to kill him). There lots of people have a little bit of information. ________________________________________________ Washington, D.C.: One of the fundamental features of efficient markets is that its impossible to obtain better-than-average risk-adjusted returns. Also, in stock markets, past market behavior is not an accurate predictor of future market behavior. So what the market will be tomorrow is like taking a random walk. If the "terror market" is efficient, how can the US government use it to beat it? In other words, suppose the terror market says there is a 60 percent chance that Al Qaeda will strike by the end of December. How can the U.S. government beat the collective wisdom of the market and change those odds? If it could, the market would no longer be efficient because it wouldn't be able to factor in the knowledge of the U.S. actions. If it's not efficient, then what type of market is it, other than a fancy system for coaxing people who may or may not have actual knowledge of potential terror events to reveal that knowledge? Eric Zitzewitz: That's an important question. The minute you start taking actions based on market prices, you affect the prices themselves if your actions are anticipated. If the price of Event A going to 2 percent was enough for the government to prevent it with 100 percent certainty, then the price would never go to 2 percent. But suppose that even if the government tried, it could only prevent events 50 percent of the time. If I thought an event was 2 percent likely with no government action, I'd still be willing to pay 1 percent for the security, even taking their preventative action into account. And if I was right, 1 percent would be the correct probability. So the prices would still have informational value, and, assuming the market was right about the efficacy of government action, they'd still be the right probabilities. Make sense? ________________________________________________ Edgewater, Md.: There seems to be an underlying assumption in all this that it's important that the exchange be based on real money. However, many articles, including one published in SCIENCE in 2001, have pointed out the fact that even "virtual money" exchanges like the Hollywood Stock Exchange and the Foresight Exchange are remarkably accurate in their predictions. My own experience as a long-time operator of NewsFutures.com (more than 26,000 markets created over 3 years, on all kinds of subjects) is that virtual money markets such as ours are indeed very accurate, and definitely more accurate than classic polls. Beyond material incentives, the essential factor in getting good informational value out of a market seems to be the knowledge that the traders bring to the negotiation table. Are you aware of any data that would back-up the idea that real-money actually makes a difference, or is it just a hunch? Eric Zitzewitz: OK, I've been fielding some tough questions, so here's one in return. I was on newsfutures yesterday, and your market has Bush's reelection at about 50 percent. Tradesports.com, a real money market, and most bookmakers have it at 66 percent or so. If your market were a real money market, there would be an arbitrage opportunity (and from what I know about Tradesports, plenty of people ready to jump on it for sport as well as for profit). So your price would be driven to whatever the prevailing level was. But that isn't happening. 50% vs. 66 percent is a big difference, and my guess is that most people would think that 66 percent was closer to the truth. So doesn't that one datapoint suggest that having real money at stake helps you get better prices? ________________________________________________ Ithaca, N.Y.: Do you happen to know if the NWS (or the Farmer's Almanac for that matter) have taken to studying OJ futures as part of their predictive medicine? It would seem to me that human psychology would inhibit the use of these types of data in an effective way. Just curious if any study has taken place on this premise, that "experts" would be unwilling to accept conclusions that differ from their "conventional wisdom" or "common sense". Eric Zitzewitz: We now have weather futures (on heating and cooling degree days in major cities), so you don't have to infer the weather from OJ prices anymore. My understanding is that utilities find them useful in planning for future electricity demand. ________________________________________________ Madison, Wis.: Aren't we gambling with matters of principle and policy? Aren't these fundamentally in the political and moral domain? Again, this fails the means-ends test. Eric Zitzewitz: The ends don't automatically justify the means, but nor are they irrelevant. My guess is that the ethical questions are a close enough call though that if people thought the markets would really prove valuable, that might tip the balance. The fact that some people thought that they were a hoax suggests that many people are skeptical of their value. We work with these type of markets more closely than many people, and thought it was worth weighing in with our view that some of them could be far more valuable than most people realize. ________________________________________________ Alexandria, Va.: Eric: In response to the question about letting the private sector fill in, you suggested that this wouldn't be a bad idea, but I spoke with one of the economists behind this project a month ago, and he told me that government sponsorship was a way around laws against online betting. Is there another way (especially now that such a stink has been raised about the ethics of these markets)? Eric Zitzewitz: The private sector markets I know of are run offshore. Making it easier for them run legally onshore would be a positive step, in my personal opinion. ________________________________________________ Cambridge, Mass.: Sorry for the off-topic question but your recent paper on arb opportunities in some classes of mutual funds was very interesting. Are you personally less inclined to invest in an international fund despite the diversification benefit it would provide a portfolio? Eric Zitzewitz: The average international fund is losing a percent of assets a year to this arbitrage. So yes, I'd avoid blindly choosing an intl. fund until they resolve this. That said, there are some funds that are better than others. Vanguard, Managers, Delaware, TIAA-CREF are examples. ________________________________________________ washingtonpost.com: Thank you for joining us today and submitting your questions. Professor Zitzewitz, thank you for your answers.