(FORTUNE Magazine) – Your servant has not mentioned the Iowa Political Stock Market (IPSM) since November 1988, when he noted -- in a column that went to press just before election day -- that this remarkable institution was forecasting a solid Bush victory. To be sure, most folks with frontal lobes were doing the same. But the IPSM proved far more accurate than the polls in 1988, and has continued to outdo them ever since. The IPSM was concocted by economists at the University of Iowa. They are somewhat committed to a ''rational expectations'' vision of how the world works, meaning that they believe markets to be superior forecasters of events in which market participants have a financial stake. How do you give people a stake in elections? Brilliant solution: Set up a market in which people trade shares in different candidates. In the original market, set up in the spring of 1988, trading was limited to students and faculty members at the University of Iowa. Packages of shares sold for $2.50 apiece, and the basic rule was that each politician's share would be redeemed for $2.50 multiplied by the pol's proportion of the popular vote. If, for example, you thought that Dukakis would get 52% of the vote, you would presumably have been willing to pay up to $1.30 ($2.50 x .52) for his shares. On election eve, when Gallup showed Bush ahead by 11.5 percentage points, and Harris had him up by only 4.7 points, IPSM prices implied he was ahead by 7.6 points. Bush won with a margin of 7.7. More recently, IPSM came within one percentage point of predicting the 1990 Iowa Senate race (won by incumbent Tom Harkin). It also did well in recent elections in Germany and Denmark, where the traders were local college students replicating the Iowa system. In each case, the student investors did at least as well as heavily financed national polling organizations. How do the students do it? Robert Forsythe, chairman of the University of Iowa Economics Department and a ringleader in IPSM, has studied the process intensively but still comes away feeling that ''there's an element of magic in it.'' All the political markets have included many participants who let emotion govern their pricing judgments and are clearly not too well informed. But the markets seem also to attract a certain number of ''sharks'' (Forsythe's term) who invest heavily, research carefully, and manage to exploit other investors' prejudices -- a process presumably not too different from what you see every day in Wall Street's efficient markets. Returns to sharks have run about 35% to 40%. Forsythe and his colleagues got into the political stock market mainly as a means of testing some academic hypotheses about market efficiency. Now that they are in it, however, they are struck by the potential utility of such markets to business -- which could use them to hedge against political risks. Right now, the futures markets offer hedging opportunities against all sorts of catastrophes: disastrous weather, interest-rate spikes, even turmoil in the Middle East (which you might hedge against by going long on crude oil contracts). So why not a large, well-financed futures market in which you could buy shares in Mario Cuomo, to pick one liberal menace not quite at random, and thereby recoup some of the losses you would expect to suffer upon Mario's coming to power? Why should such catastrophes be hedgeable only by a few academics in Iowa? Hey, what about us poor New Yorkers?