Robert Merton makes a great point on this subject. In complete markets, any derivative can be replicated by some trading strategy. Conversely, any trading strategy can be represented by a derivative. Any individual with known risk preferences can identify an optimal trading strategy over the course of her life. So any individual should just simply buy one derivative and then get on with their lives, leaving the seller of the derivative to execute the trading strategy. The two fabulous consequences of this are (1) those that want them can have jobs for life executing hedge trades and (2) the rest of us can go to the beach having made a pile of money selling these products to retail punters! The quant finance revolution isn't over until the day comes when everyone has bought their derivative!

In the time frame of sciences quant finance is a baby. An inexact science it may be, but junior all the same. If you look at when risk was seriously considered quantitatively in markets compared to history of finance there is a big gap. The bubble being discussed is really a catch up period to the technology and speed at which markets have grown, there's likely to be some over-compensation. I'm not drawing a direct comparison but the Dot Com bubble is similar. There are still plenty of people making money out of the internet, even if it's not as high profile, and the GOOD internet children are still in high demand. As long as one firm keeps quant finance a priority all the others will too (prisoners dilemma).I have read a number of senior members saying it's the ever changing nature of trading that makes it so interesting and unique. If things are so dynamic, you can't relax and rest on yesterday's model.

QuoteNot only energy companies benefit from weather derivatives, but other sectors too. I don't believe it has been mentioned yet in this thread....but I believe there was discussion of it some time ago on this site.....Insurance.........consider the hubbub in Europe a few months ago as a good example of a demand driver...

QuoteOriginally posted by: asproAs long as one firm keeps quant finance a priority all the others will too (prisoners dilemma).Very nicely put, I can only hope I am that prisoner and let ITO 33 become other's people dilemma.

Let me answer this question with a story. The year is 1969 and Seattle is undergoing a major recession as Boeing is suffering from the reduction in military spending and other factors. A friend of my family was among the world experts in materials for wings for supersonic flight, having spent over 20 years, starting in graduate school before there was supersonic flight, on the subject. He was fired from a very high-paying job at Boeing and got a job driving a cab (literally). There were no aerospace jobs, the professional organization took out ads in campus newspapers telling students not to go into the field, as there were already enough people trained for the next 30 years."Never specialize in hardware," my cab-driving friend told me, "hardware changes." So he taught himself computer programming, founded a company with a few friends in similar circumstances, and sold out for about $10 million eight years later (when $10 million was a lot of money).Also, of course, aerospace kept hiring, it just hired cheap newcomers trained in the latest hardware and kept firing expensive people with lots of experience in old hardware. The only people who prospered consistently were those whose skill did not depend on specifics.In finance today, the market is tough for the physics-trained quant expert in C++ mathematical programming of Monte Carlo and numerical solution of stochastic partial differential equations. That will get worse, not better. There are lots of quantitative jobs in finance, but they are going to cheap newcomers who have different skills. You might not recognize a lot of these people as quants, because they are better at SQL than C++, and know more about search algorithms than differential equations.The places of employment are changing as well. Quants are moving off desks and trading floors, and into middle and back office. The largest organizations are shedding quants, while ever-smaller organizations are hiring.People who really understand the software: good mathematicians who know real finance, will always do well, even if they have to drive a cab for a couple years before starting their own companies. Anyone who specializes in a specific technique or environment, will have a hard time.This is true anywhere, anytime, any industry. It's not a bubble, it's life.

Thanks for all your thoughts -- some very interesting ideas, especially the 'personalised derivatives'! I reckon I'd be happy to give someone a few bucks to hedge me against the big basket of things that could go wrong beyond my control (though isn't this just 'insurance' rather than sexy exciting derivatives?)Aaron -- interesting to hear your 'stay general' strategy, as it seems there is much pressure to specialise in one niche field and get mega rich off the latest CDS multidimensional lattice pricing techniche or whatever.Also interesting to hear your description of Mathsy IT people as quants - this is kind of what I do (degree in CompSci,MSc, work for a hedge fund, mining equity tick data) but I never thought of myself as a 'proper' quant as I don't do PDEs or price derivatives.It still scares me that one day everything will be programmed on Bloomberg by a handful of the best quants, and everyone else will get sacked though....

QuoteOriginally posted by: foxxIt still scares me that one day everything will be programmed on Bloomberg by a handful of the best quants, and everyone else will get sacked though....Don't be scared.By ITO 33's lemma:(Something will not be programmed on Bloomberg) OR (Something will, not by the best quants).

QuoteOriginally posted by: foxxIt still scares me that one day everything will be programmed on Bloomberg by a handful of the best quants, and everyone else will get sacked though....Perhaps your words should be taken with a bit of caution. In fact, Bloomberg had employed one of the shrewdest quants of THIS forum...

I can identify two main drivers for the future of the profession.1- A relentless increase in computing power. This makes complex algorithms possible which were only available in theory a few years ago. This is true for the calibration of PDE, for Monte-Carlo simulations, for risk management problems... As more computing power becomes available in the coming years (and computer scientists are still predicting incredible steps forwards), new algotithms and a new quantitative finance will emerge.2- The second point is probably related to the first. The industry is structuring itself very fast today. Few years ago, research was conducted by big banks alone. They developed everything in-house, with large propietary research projects. There was no space for research boutiques. Hedge funds have introduced a different business model. They outsource large chuncks of their research and deveolpment effort and they seek to purchase the best quant products available. This has created a market for niche players of very high quality. In the future, the best traders will be in hedge funds, talking to the best quants in specialized research outfits. A brave new world is emerging. Who is saying that Quant finance has no future? It is always when markets are tough that talents are revealed. Who needs a good quant when you make 40% a year investing in the first stupid dot com at hand? Who needs risk management when stocks only go up and default is an abstract word? Who needs quants when the banking industry lives with the impression that a nice static VaR calculation will insure their positions (thank you Bale II)? My final thoughts:- Markets are tough- They will become a lot tougher- The serious Quants will emerge from the chaosCheersNewNumberTwo

QuoteOriginally posted by: DavidIn fact, Bloomberg had employed one of the shrewdest quants of THIS forum...Like I said, no reason to be scared.It certainly is a small world, but THIS forum is certainly smaller...

ITO 33's lemma (for the positive minded):(We have the best quants) OR (We have programmed something other quants will never program)

Other people's di-lemma:(Have the best quants) XOR (Program everything)

QuoteOriginally posted by: NewNumberTwoA relentless increase in computing power. This makes complex algorithms possible which were only available in theory a few years ago. This is true for the calibration of PDE, for Monte-Carlo simulations, for risk management problems... As more computing power becomes available in the coming years (and computer scientists a still predicting incredible steps forwards), new algotithms and a new quantitative finance will emerge.A substance that was fairly unprecedented only six months ago gradually gaining some confidence, at last. Actually, advanced quantum computers and complex algorithms are already a fact in the computerized science laboratories, but not in the financial laboratories yet. Is this to say that Binary algorithms have been squeezed to the point of near exhaustion? It could be the difficult and unconceivable obstacle, though. Beyond the computing power of the processors, quantum algorithms may be difficult to deal with in terms of randomness. In the Binary language for example, the internal parameters are merely string of zeros and ones, while the quantum algorithms contain both zeros and ones and every possible number between them. This of course product, in probabilistic terms, powerful simulations compared to current (pseudo-random) simulations running by Monte-Carlo engine. The moral here is that one cannot rely upon a classical computing device to properly executes a probabilistic algorithm, hence a possible dilemma in the future might be the ability of a Quant to manipulate quantum events and thus product string of truly random processes. However, quantum events apparently contradicted the tendency toward stability and uniformity in quantitative finance. Then again, in tough environment perhaps that the best Quants will be able to adopt imaginative and innovative approaches in a way to master the perplexing events to come. Finally, I do agree; markets will be much tougher however. Regards,David

Increase in computing power cuts two ways. A lot of problems have exact, simple-minded solutions if you have enough computing power, all the clever stuff people did was to tease out approximate solutions with one-billionth of the computing power. I spent a lot of time in graduate school learning how to integrate, search and handle large matrices on a computer; none of those skills are useful to me today.If the increase in computing power leads to new types of products and strategies, it will support a new generation of quants. But if it leads to simple, off-the-shelf algorithms for all currently-traded products, it will compete with quants.As computers have evolved, I have seen both of these things happen many times in many fields.What I do know is not to wed "quant" with "computer." It may be that the future of quantitative finance is not in computers, that all those functions will be transferred to IT, with one old holdover doing model validation and signing off on the reports.

QuoteOriginally posted by: AaronIf the increase in computing power leads to new types of products and strategies, it will support a new generation of quants. But if it leads to simple, off-the-shelf algorithms for all currently-traded products, it will compete with quants.My view of the place where IT, quants, and computers should be moving in the future:- Computing power and increased sophistication of the theoretical models (e.g. incomplete markets) will allow "specialized research outfits" to actually deliver solutions that work, as NN2 puts it. (I think he has in mind an outfit like his own).- The best hedge funds will no longer hire quants to program the above models but to calibrate them, eventually to simplify the problem and reduce the number of parameters (as Aaron sees it).- The best traders in the best hedge funds (or in the best banks, when banks follow into this new trend) will just have to learn to fly these complex machines.