In 1992 seven postgraduate students started their PhDs in the high-energy physics group at Imperial College, London. Six years later, three of them now work in international finance in the City of London. Although neither the research councils nor the financial sector keep records of how many physicists enter the City, it is clear that more and more are doing so. "[PhD physicists] are more in demand than they were five years ago, " says Ivan Collins, a senior consultant at Michelangelo Associates, a company that specializes in international financial recruitment.

Theoretical physicists have long been prized in the City for their skills in modelling and programming. But it is no longer just theoretical physicists who become rocket scientists - experimentalists with experience of data analysis are also in demand. "We look for people with the quantitative skills to build models and to understand risk, " says Collins.

Most physicists who work in the City are quantitative analysists or "quants". They model the past behaviour of the stock market in order to predict the range of values that a stock or share might have in the future. Quants deal in derivatives - financial products such as futures and options with a price derived from something else. In a future, you agree to buy something at a given time in the future for a price agreed now. In an option, you pay for the right - rather than the obligation - to buy something at a given price in the future. By modelling the stock market, quants can calculate a fair price for a given derivative (Physics World April 1995 p8).

"It is a common misconception that we predict what is going to happen on the market - that is not what we do, " says Han Lee, who has a PhD in theoretical condensed matter physics and now works as a quant and derivatives trader for Toyko-Mitsubishi International. Lee's work involves calculating a fair price and the associated risk for complex financial products based on interest rate models.

Lee finished his PhD at Cambridge University in 1992, then spent two years as a post-doc before entering the City. "Relatively speaking, it is difficult to get a permanent job in theoretical physics, " says Lee. "Matching the real world with maths is a transferable skill, " he says. "And the financial rewards are much greater."

Rocket scientists are indeed handsomely rewarded for their work. PhD quants with no experience can expect a starting salary of £35 000-40 000, according to Collins. After one year, this typically rises to £150 000; five years into the job, the quants will be earning between £250 000 and £500 000 a year.

Many of the rocket scientists who command these salaries are top-notch physicists who could have made a career in academic research. John Sleath, a monetary analysist at the Bank of England, shared the Royal Astronomical Society's astronomy prize last year for his "outstanding doctoral thesis" on the computer simulation of the formation and evolution of galaxies. After completing a PhD at Cambridge University, Sleath spent two years as a post-doc at the University of Wales at Cardiff.

But a new career beckoned. "I got the feeling that astrophysics does not matter, " says Sleath. He now models the time structure of interest rates; the research is long-term but is ultimately used by the bank's monetary policy committee to set interest rates. "I now have the feeling that I am contributing to something worthwhile, " he says.

Sleath also had complaints about other aspects of academic work. "I never enjoyed teaching, but it was a necessary part of career progression, " he adds. "Now, I still do all the bits [of research] that I enjoy but with more purpose. I can do research without the teaching and for more money."

Working in finance gives physicists the opportunity to show off skills that could be neglected in universities. "Being a good organizer, a good communicator and empathetic are not necessary to be a good physicist...I wanted to display a few more of my talents, " says Jessica James, vice president in the strategy risk management advisory group at the First National Bank of Chicago. James, who did her PhD in theoretical atomic physics at Oxford University, is obviously well organized - she handed in her thesis, got married, moved house and started her new job all in the same week.

"The whole of finance is to do with probabilities - that is very physics-y, " she says. "For example, we might need a Monte Carlo simulation to look at the possible value of a deal at the end of a time period. Monte Carlo simulations and programming are things that physicists do."

James designs trading strategies for the bank and advises external clients. "A client might have a portfolio of 12 currencies, whose value fluctuates considerably, " she says. "We can show the client the possible range of values that his investment will have in the future. If that range is too great to be acceptable, we can hedge the risk."

Entry to the City was straightforward for James. "I had a small lectureship at Oxford but was looking around for something else, " she says. "My supervisor came in with a letter from the First National Bank of Chicago, and he said 'Jess, you are looking for a job, aren't you, ' and I thought 'Oh, well it won't do me any harm'. The e-mail address on the letter had been mis-typed but I could work out what it should have read and so I applied. I was the only one who passed that particular intelligence test."

So where do rocket scientists end up? Although a job in the City is lucrative, it is not necessarily secure. "The paternal instincts of institutions are becoming less and less evident, " says Collins. "A person who has been in the City for five years may well have had three jobs, " he adds. Quants tend to remain in the City until they are in their late thirties or early forties. Some then move into senior management roles; others have made enough money to retire.

"Ultimately, I would like to work and write, " says James, who is currently writing a book about interest rate modelling with Nick Webber of the University of Warwick. "And I want to write good, bad science fiction - but that is another story."