to cuchI would say quant trading is more econometrics rather than probability theory. [see eg euan sinclair volatility trading] but the point is that you just don't need so many people. each hedge fund has a handful of quant traders... IBs need(ed) 10-100 quant/quant devs developing and supporting new derivatives products.perhaps another difference is that quant traders are doing their analysis "off line" using R and other products, whereas the need for daily P&L and risk reports requires the maths runs within the trading system ( and consequent development of quant library). [DISCLAIMER I have no experience of quant trading!]

Perhaps we're witnessing quant market capitulation à la the one in the legal market?

Thanks for all of the great thoughtful posts. I have found the discussion very interesting and informative. I'm really glad I asked the question.I don't have an answer to the salary question, which is why I asked. I do think that there will always be a need for computational finance people.I've been struggling with some portfolio modeling software whose creator will go unnamed. The software is not easy to use and the documentation is some of the worst I've ever seen. In some cases the documentation is simply non-existent. The software comes in a risk analysis component and a portfolio modeling component (which give return, but not risk). The portfolio modeling component has some good features but its a black box. It is not very flexible and it is limited when it comes to model building. Struggling with the software I wondered why anyone would use it. I think that one reason is that this software is for people who can't build portfolio models in R or Python (or C++).So I speculate that there's a place for someone who knows how to build portfolio models and ask the right questions (what is the estimation error of the covariance matrix, how can it be constructed better, how can the model be made more predictive...) Of course this doesn't really say much about salaries.I will say that computational finance is one of the most demanding areas I know of and has some of the smartest people I've encountered. But then they are, at least in part, attracted by the money and if the money is not there, they may work elsewhere. There are those who say that this is a good thing, the view being the finance doesn't produce anything where as building something in Silicon Valley or Boston does. I don't fully subscribe to this idea, since Silicon Valley would not exist without robust markets.Perhaps one thing that will change is New York and London will not be the centers where quants work. The salaries have to be so much higher just to have the same life style you can have in Silicon Valley or Boston (which is not cheap). There is no reason that funds have to be in New York. So perhaps this will start to change.So one additional question: given the observations in this thread, have your career plans changed? Have you gotten "out" or are you thinking of getting "out"? Quant skills are relatively transferable to other areas, like engineering.

QuoteQuant skills are relatively transferable to other areas, like engineering. Depends on what you mean by 'engineering'. Chances are you will need to be retrained. Which skills are you considering?

Slightly off-topic: as per numbeo.com and expatistan.com, San Francisco is only 10-15% cheaper than New York. The people I know who moved from NYC metro area to SF bay area don't really consider the latter cheaper. As for salaries, there are no fundamental reasons that quants should make more money than other math/physics PhDs who can program. The only reason is being "close to the money", which is just a traditional "compensation inefficiency" which will probably disappear with more regulation and less risk-taking. On the other hand IT is probably in a bubble now which, too, can pop.

I'm already a software engineer/computer scientist. I'm in the UW Master's program as well (as a part time student).

QuoteAs for salaries, there are no fundamental reasons that quants should make more money than other math/physics PhDs who can program. The only reason is being "close to the money", which is just a traditional "compensation inefficiency" which will probably disappear with more regulation and less risk-taking. I'm not sure I totally agree with that. One argument why quants have high salaries is that as already identified, many of them really are very smart and well qualified. Being smart they have many interests. Furthermore, many have come from 'romantic' areas of endeavour, i.e. scientific research, where the culture is one of intellectual freedom. However these people are harnessed to a computer making financial calculations/software/'whatever', with high stress, long hours and often within a corporate culture with all that entails. For these people to do that there has to be additional compensation, otherwise they wouldn't choose to do it (for the most part). In contrast an equivalently qualified person in academia will tend to be paid less because they are compensated by the work itself and the more relaxed work culture.

QuoteOriginally posted by: CuchulainnQuotea) the demand for quants came from derivatives pricing.. that's why you had all these MSC in quant finance etc. teaching derivatives pricing (rather than on quant trading)Quant trading is less maths and more programming than traditional quant work?Yes, it is. I would even say 'smart programming.' Just coding a formula or an algorithm isn't enough.

QuoteOriginally posted by: neuroguyQuoteAs for salaries, there are no fundamental reasons that quants should make more money than other math/physics PhDs who can program. The only reason is being "close to the money", which is just a traditional "compensation inefficiency" which will probably disappear with more regulation and less risk-taking. I'm not sure I totally agree with that. One argument why quants have high salaries is that as already identified, many of them really are very smart and well qualified. Being smart they have many interests. Furthermore, many have come from 'romantic' areas of endeavour, i.e. scientific research, where the culture is one of intellectual freedom. However these people are harnessed to a computer making financial calculations/software/'whatever', with high stress, long hours and often within a corporate culture with all that entails. For these people to do that there has to be additional compensation, otherwise they wouldn't choose to do it (for the most part). In contrast an equivalently qualified person in academia will tend to be paid less because they are compensated by the work itself and the more relaxed work culture. I was comparing quants to other PhDs who work in industry, not in academia. Most of them are also "harnessed to a computer". An "equivalently qualified person" might not get a job in academia at all.The problem with your argument is that supply of people going into quant finance has been going up despite compensation decreasing. This means that there will be enough smart people to fill positions even with lower salaries.

Just a quick question: how wide is the compensation gap between a trader and a quant ? I suspect it to be pretty big right ?Because at the MFE of my school, the best tercile ends up as traders and not quants.

QuoteJust a quick question: how wide is the compensation gap between a trader and a quant ? I suspect it to be pretty big right ? Because at the MFE of my school, the best tercile ends up as traders and not quants.My perception is that trading, especially on the sell side, requires some different personality characteristics and talents than being a quant. Traders need to be able to think fast. They think about the current trade and the current position. They work with a lot of stress.Quants, especially on the buy side, tend to be more cerebral and less able (or willing) to handle stress and abuse. Building software systems and models is very different from intraday trading.I have only worked on the buy side, so my picture of the sell side is from books like Street Freak. People speak of "Wall Street" as a meritocracy. One meaning of this is that if you make a lot of money for your firm, you will get paid a proportionate amount. Traders tend to make more because they bring in more money.

QuoteOriginally posted by: iankPeople speak of "Wall Street" as a meritocracy. One meaning of this is that if you make a lot of money for your firm, you will get paid a proportionate amount. Traders tend to make more because they bring in more money.Paying people for being lucky does not make a meritocracy.

QuoteOriginally posted by: ChicagoGuyIsn't this just division of labor at work? Perhaps banks have found it easier to specialize in investing/trading, and have left the quant work to be done elsewhere.to be fair banks still have to have quants to constantly cope with the nonsensical regulatory requests

QuoteOriginally posted by: iankHave you gotten "out" or are you thinking of getting "out"? Quant skills are relatively transferable to other areas, like engineering.I will probably leave the quanty mumbo-jumbo and try to move to some management, non-technical sh!t. Like opening a bar.

QuoteOriginally posted by: katastrofaQuoteOriginally posted by: iankPeople speak of "Wall Street" as a meritocracy. One meaning of this is that if you make a lot of money for your firm, you will get paid a proportionate amount. Traders tend to make more because they bring in more money.Paying people for being lucky does not make a meritocracy.Yes. The quant business itself is quite meritocratic I think. Problem is that being a quant one almost necessarily hits the glass ceiling, and that makes the whole picture probably not so meritocratic.