Finance, as we know, has become an ultra high-technology business.

Gone are the days when people traded by the seat of their pants. Nearly gone, too, is plain old-fashioned forecasting. Instead, we have rocket science, or quantitative finance as it is usually known.Quantitative finance means using mathematics and computers to calculate the probability of something happening: a change in securities prices or in the volatility of certain derivatives, for example. Once you have worked out probability, you can calculate the risk and go from there to set a price for your deal.

''Quants,'' the people who use quantitative methods, are now the drivers at the sharp end of the international investment banking business. Or at least they were.

Last year's debacle over Long Term Capital Management, the large U.S. hedge fund that had to be rescued, has cast a shadow over quantitative methods - not least because two of the Nobel laureates who invented them, Myron Scholes and Robert Merton, sat on its board.

Is quant still OK? The short answer is yes, at least as far as the major investment banks are concerned. They continue to use it with gusto. But LTCM raised an important question: Are quantitative finance techniques sufficiently highly developed to provide reliable trading and risk management strategies for big banks? The answer is no.

As one senior investment banker recently put it to me: ''This is a remarkably immature field when you think of how much of the world's economy rests on it.''

So what are the investment banks doing about it? Distressingly little.

The logical thing would be for them to finance some urgent and fundamental research by the academic world to develop much more reliable quantitative tools. The theory of quantitative finance is so esoteric that only a few universities could do the job. But the banks are doing very little.

Our center recently carried out a survey of quantitative finance in London and New York. In the world's top two financial centers we found no surge of interest in theoretical work, no attempt to generate fresh research to reinforce the weaknesses exposed by LTCM, and no loosening of the purse strings. By our calculations, banks in London spend a mere 2 million pounds a year on academic research in quantitative finance - less than a hot-shot trader can earn on a single trade.

The banks defend this niggardliness by saying that they prefer to do all their research in-house, for competitive reasons. If they financed university research, everyone would benefit, so it would be money wasted.

This is understandable - to an extent. Investment banking is an intensely competitive business.

But the banks' attitude is extremely shortsighted. They should understand that their quant-based activities need to be underpinned by sound academic theory. Otherwise they will only be courting trouble.

This is a sad state of affairs. The relationship between academia and industry is seldom good in any area of business; they have such differing agendas. But the gap can nowhere be as wide as it is in finance. And the more banking becomes driven by science, the more dangerous this gap could become.

For the full story: Log In, Register for Free or Subscribe