TCS Daily

Sell Bin Laden, Buy Bush!

By Jan Arlid Snoen - June 3, 2004 12:00 AM

OSLO -- Last July a small scandal broke in Washington. Admiral John Poindexter, of Iran-Contra fame, had been responsible for planning a government-controlled online futures market, the Policy Analysis Market (PAM). The purpose was to collect information on political risks in the Middle East and worldwide. Commentators and politicians jumped on the idea that some of the potential contracts could involve terrorist attacks or assassinations, and dismissed the possibility that someone should be able to profit from correctly predicting such attacks as obscene.

Ironically, just a few days after the story broke, TradeSports put up a futures contract on Poindexter getting sacked before 30 August 2003. TradeSports is the leading Irish online futures market, also known as a betting exchange. Perhaps Poindexter would have been well advised to seek cooperation with TradeSports or the other main player, the British company Betfair, instead of establishing a new marketplace. After all, the pervasive political correctness that killed PAM seems to be a lesser problem in Britain and Ireland, where regulation of internet gambling also is much less restrictive than in the U.S. Instead of establishing their own marketplace at an initial cost of US$8 million, Poindexter could have paid the existing ones a trifle to set up betting lines of particular interest.

Over the last couple of years a number of such exchanges, where buyers and sellers for financial, political, legal and sports futures are matched, have sprung up. Political betting has been around for a long time, but the betting exchanges are far more sophisticated than the old bookmakers, even in their new online appearance. The commissions charged, typically 3-5 percent, are much more advantageous than the average bookmaker's payback-rate, usually in the 80-90 percent range. The exchanges offer constantly changing prices, instantly reacting to new market information, and are open 23 hours a day, every day. You also have the opportunity to bet against an outcome, and above all, to move in and out of positions.

The range of political bets is quite astonishing. Combining TradeSports and Betfair, you currently can bet on the U.S. presidential election, including VP nominee and winner in every single state, control of the Senate and House of Representatives, British, Canadian and Australian parliamentary elections, Tony Blair staying on as Labour party leader, bin Laden being captured or killed, the sacking of George Tenet or Donald Rumsfeld, U.S. Security Alert Level (perhaps a proxy for terrorist attacks) at the end of every single month for the rest of the year, the transfer of power in Iraq, the establishment of a Palestinian state and Israeli eviction of Arafat.

Besides the obvious entertainment value, the betting exchanges are information accumulators. Like all efficient future markets, they combine available information and sentiments to provide a snapshot of the likely future outcome. Whenever the U.S. troops have a particularly bad day in Iraq, Bush stumbles badly in a speech, the employment data are surprisingly good or new opinion poll data are released, the market price for Bush contracts moves. For those to busy following all the news, just checking the TradeSports homepage gives you an instant update on the state of the Bush campaign. This is the first thing Karl Rove should be doing every morning! It is interesting to note that The Economist has begun referring to the TradeSports markets at a regular basis when covering the U.S. election campaign.

For most of last year, the chance of a Bush victory fluctuated around 65 percent, peaking above 72 percent after the capture of Saddam. It has been falling almost constantly thereafter, and has recently been stabilizing above 55 percent. The Democratic nomination contest was a kind of a roller coaster ride, but the Kerry momentum was identified early on. Even after Kerry's nomination seemed secure, a possible "bimbo eruption" in February created a short-lived fall in the market and thus a profit opportunity for the cooler heads -- and a possibility to cash in most of the winnings for the fainthearted.

The Iowa Electronic Markets (IEM) has been around since 1988, mostly predicting the outcome of U.S. Presidential elections and other elections worldwide. This market has in general been considered more reliable than opinion polls and pundits. There is even research to prove that the IEM, as it stands close up to the elections, is better at predicting the outcome than the latest opinion polls. The new betting exchanges should be even better, given that liquidity is higher, and that real money is involved. The Iowa market is more of an academic exercise, where no punter is allowed to carry an account balance above US$500. It is reasonable to believe that many participants in such markets lack the discipline that comes from risking non-trivial stakes. Even so, IEM and TradeSports markets seem to follow each-other quite closely.

Officially the reasoning behind the DARPA (Defense Advanced Research Projects Agency) project headed by Poindexter was to generate useful information about the general political situation. A more troubling motive might have been to lure people with inside information about a particular incident, perhaps even terrorists themselves, to try to make a buck on this information, and then try to identify the punter. If we are to believe the official story, this was not the case, and the transaction limit of $100 seems to prove this, as such an amount would hardly be enough to motivate a terrorist. Anyway, using a futures market to collect information on such low probability occurrences as the timing and target of terrorist attacks or assassinations will probably not work. There is just too little relevant information around to form a basis for anything more than guesswork. This is also the verdict of the marketplace -- no such contracts have been offered.

Even so, there is nothing in principle that prevents a futures market to set up contracts that many would find morally troublesome. The most serious critique is that the existence of such a market, given sufficient liquidity and therefore potential profit, could create incentives to influence the outcome. This problem is well-known from sports betting, where match fixing is not uncommon. One strategy in preventing this is to exclude punters that might have a direct influence on the results, or as a minimum to be able identify them after the fact. TradeSports could not have accepted John Poindexter betting on his own departure, as he had some discretion in the precise timing of this outcome. Another measure is the one proposed by PAM and being the rule at IEM -- to put a low limit on the maximum bet, with the corresponding possible efficiency loss mentioned above.

Another safeguard would be to disallow betting on an extremely unwelcome outcome, such as a terrorist attack or the assassination of the American president. But as we have seen, people might differ on what such an unwelcome outcome is. TradeSports actually has put "prices on the heads" of both Saddam and bin Laden, and it easy to image an Arabic betting exchange putting up contracts on Ariel Sharon. The potential winnings on these contracts have of course been much smaller than the actual bounty the U.S administration placed on the capture of Saddam, and the bounty Al Qaeda conceivably has placed on the head of George Bush. To create incentives for people to really try to influence the outcome, the betting exchanges must increase potential winnings, perhaps by a factor of 100. Even then, the chance of terrorists actually playing these markets, given the high detection risk, must be considered miniscule. Better then for a terrorist to bet by proxy in the financial markets for stock or oil futures, or why not sell Continental Airlines short? After all, there have been unconfirmed rumours that Al Qaeda operatives shorted stocks on the New York Stock exchange before 9/11.

Although betting exchanges can be quite efficient as information accumulators even though liquidity is low, liquidity has to increase by several orders of magnitude if these markets were to fulfil another potential role, as risk management tools. If you are concerned that a Kerry victory in the fall will raise your taxes or cause a general downturn in the economy and hurt your business, you can insure against that by buying Kerry contracts. Total turnover on TradeSports presidential contracts are still less than US$4 million, and even if you add the unknown millions placed at old-fashioned bookmakers, this market is obviously still dominated by players that try to outguess other players, not by risk managers. It is conceivable, but highly unlikely, that these markets could grow so much that they could play a role in risk management. If that was to be the case, the need for safeguards against those who can influence the outcome would become acute.

And in case you wondered: The chance of getting bin Laden before the end of the year is about 30 percent, and the prospects for a Palestinian state at the end of 2005 looks grim, at 15 percent.


TCS Daily Archives