(FORTUNE Magazine) – AS ANY LONDONER WITH A LICK OF sense will tell you, house prices in the British capital--up 200% over the past decade--are overdue for a correction. And thanks to the miracle of modern derivatives markets, London's real estate pessimists can, unlike their sadly disadvantaged counterparts in New York City or Los Angeles, put their money where their fears are by hedging their houses with warrants that will pay off if British housing prices drop.

Only ... Londoners aren't buying. Trading in the put warrants on the London Stock Exchange is scant, and prices have dropped steadily since they were issued last year. "People in London are all saying the bubble has peaked and it's going down," says Yale economist Robert Shiller, who plans to introduce similar securities in the U.S. "But they don't do anything about it."

Can you really blame them? Pessimists have been warning for years that house prices can't keep rising at their current torrid pace--15.1% a year in the U.S., according to the National Association of Realtors. Someday they will be proved right. But ever since the Dutch went nuts over tulip bulbs in the 1630s, figuring out when a bubble will burst has proved a task too difficult for even the finest minds (and finest computers). The end may be inevitable, but there's no telling when or how it will occur.

This is the strange reality of financial bubbles. When one really gets going, the people who--correctly--warn that it won't go on can be wrong year after year after year. Meanwhile, the overenthusiastic, ill-informed sorts who cast such worries to the wind can get seriously rich. Conventional economic theory holds that market prices are set by prudent and rational buyers and sellers. When all the profits flow to the imprudent and the irrational, though, they set the prices--and economic laws temporarily cease to apply.

Yale's Shiller, the world's leading expert on asset prices gone wild, argues that this is happening in real estate markets right now. The second edition of his bestseller, Irrational Exuberance, includes a new chapter making the case that we are in a housing bubble. But there's still that pesky question of when it all will end. At least part of Shiller's stellar reputation these days can be attributed to the fortunate coincidence that the first edition of his book hit stores just as stock prices peaked in March 2000. "When I came out with the second edition, I was wondering if I would be as lucky in timing the real estate market," Shiller says. "I don't think I will be. Real estate in the U.S. has enough upward momentum to keep going for a while."

At a conference on behavioral finance put on in May by Yale's School of Management, Shiller delivered a keynote speech on the parallels between the stock bubble of the 1990s and the real estate craziness of today. Tellingly, though, the rest of the day's discussions barely touched on bubbles at all. It's not that the assembled Yale professors, Wall Street traders, and hedge fund managers disagreed with Shiller's arguments. It's just that they have come to despair of finding a reliable way to make money off the periodic insanities of the markets.

In the late 1990s, for example, there was no shortage of smart Wall Streeters who thought the stock market was overvalued. But those who bet against stocks too early were wiped out, or at least lost clients by the boatload. As a result, markets came to be dominated not by smart Wall Streeters but by the nitwits who thought prices would go up forever. Until, as was inevitable, prices stopped rising.

Why was this inevitable? Because an asset derives its true value from the income that it will throw off in the future. With a stock that means, strictly speaking, the dividends it pays. With an owner-occupied house, it's what economists call "imputed rent"--what you would have to pay to rent an equivalent house.

This means that growth in asset prices cannot sustainably outstrip economic growth, because profits and rents tend over time to rise in step with--or often a couple of steps behind--the economy as a whole. Assets can sometimes be revalued upward when investors reassess the riskiness of those income flows, as seems to have happened with stocks during the second half of the 20th century. But that's a one-time jump. Asset prices simply cannot keep growing faster than the economy ad infinitum.

To bring the argument back to houses: In many U.S. markets today, the monthly cost of buying a house is significantly higher than the monthly rent on an equivalent house. So by the "imputed rent" test, it makes no sense to buy. People buy anyway because they assume that house prices will keep rising, meaning they can sell out later for a profit. That of course requires that another buyer will come along who also assumes that prices will keep rising, which will require yet another such optimistic buyer down the road, and so on, and so on. This happens to be the definition of a Ponzi scheme.

In reality, the long-run rate of housing price appreciation in the U.S. and elsewhere dramatically trails economic growth. That's because, over time, the main determinant of house prices has been building costs, and as the productivity of the construction industry has risen, inflation-adjusted costs have dropped. In the most expensive real estate markets today it is land scarcity that is driving prices up. But that can't go on forever: At a certain point the high price of land will drive away the very economic activity that made the land valuable.

In other words, this real estate market craziness cannot continue. But to make a bet about when it's going to end--now that would be really crazy.

FEEDBACK first@fortunemail.com

FEEDBACK jfox@fortunemail.com