Where can I find which quantitative/technical skills are hot, and which ones aren't?

February 3, 2008 12:58 PM

Where can I find which quantitative/technical skills are hot, and which ones aren't?

I'm starting to plan for a career after college in a quantitative/technical field, such as quantitative finance or software. As I plan out what classes I will take for the rest of college, and what things I will learn in my spare time, it would be extremely useful to know what the most promising and valuable technical skills are. I have many different math, computer science, and engineering classes to choose from, and want to know what will be most useful and versatile when I finish college and want to get an industry job. (One indicator is that at my school, I have worked around some research groups who seem to be pessimistic about the future their field, and some who are extremely optimistic.) Where could I look for reviews/indices that discuss the outlook for different fields/skills?

Here are some examples of the concept I'm referring to:
-This ComputerWorld article (though it only discusses computer skills)
-The Bureau of Labor Statistics job profiles like this one.

I'm interested in knowing which disciplines/skills are up-and-coming, and also which ones are passe.

(Many obvious disclaimers apply: a field that's hot today might not be so tomorrow; and different skills are in demand in different places; and an ability to think critically and communicate effectively will never go out of style. I do understand all that; I just want to consider all sources of information when I make my decisions.)

posted by lunchbox to work & money (13 comments total)

If you want to work in quantitative finance, it seems the quickest way to get there is to get a master's in financial engineering/computational finance (what I'm doing) or get a PhD in math or physics. Are you an undergrad or grad student?
posted by pravit at 1:27 PM on February 3

Thanks pravit. I'm an undergrad.
posted by lunchbox at 1:42 PM on February 3

Well, here are my general tips:

If you haven't already, you should definitely learn C++. You don't need a class for this, just buy some programming books and mess around on your own. Knowing object-oriented programming is an invaluable asset, regardless if you'll end up programming in Java or C#. C++ has the advantage of being the language that most everybody knows and having a lot of things written in it.

Take classes in probability and statistics. These are good subjects to be familiar with regardless of what field you go into. They're a requirement if you want to do any work in quantitative finance, operations research, or actuarial work.

One thing that I still haven't done (but really should) is to become really skilled at Linux administration and that sort of thing. This is another thing that I imagine is more about tinkering around than actually taking any classes. I think there will always be demand for people skilled in Linux administration, PHP, mySQL, Perl, etc.

I don't think many quant finance jobs are open to undergrads, but you could always go the traditional IB analyst route (may not be quantitative enough for you though).
posted by pravit at 3:18 PM on February 3

This grad program is designed with that job market in mind. Why don't you poke through the curriculum and take back bearings--and then see what similar courses you can take now.
posted by Phred182 at 4:03 PM on February 3

Keep in mind that the master's in financial engineering is a professional degree and kind of limits you to a very specific role. It's not meant to be an intermediary step on the way to a PhD. The degree is very expensive (generally around $50K) and there usually isn't any financial aid available (the idea being that you get a high-paying job and pay off all the loans).

There are a lot of new MFE programs popping up, but in general you have to go somewhere well-known to secure a job: look at the programs offered by CMU, NYU, Columbia, UCB, Stanford, Michigan, and Toronto. The problem is that more and more people are hearing about this degree, and you'll be competing with hundreds of Ecole Polytechnique hotshots and the cream of the IIT Kharagpur crop for 1 of the 30 or so seats at each of these programs. I think 10-15% is the usual admit rate.

If you want to go this route, you need a high GPA (3.5 minimum) and a degree in a math-heavy discipline like mathematics, physics, engineering, or computer science (math and econ seems to be the optimal combination). In addition, you need to have taken (and done well in) multivariate calculus, linear algebra, probability, statistics, and preferrably partial differential equations, stochastic processes, and real analysis. I think most programs want people with 1-2 years work experience related to finance, although this req. can vary (I don't have any relevant work experience, but I managed to get in). Send me an e-mail if you need more info.
posted by pravit at 4:21 PM on February 3

Thanks so much Pravit (and others)! That info was extremely useful.

As for what you mentioned about Linux: yes, tinkering is the best way to do it. You can install a Linux distribution (such as Ubuntu) and learn to use the shell and so on.
posted by lunchbox at 8:03 PM on February 3

I am a grad student; Ms Robot is an actuary at a major insurance company.

The learning to think of proof-based math is kind of useful but got Ms Robot less traction in the job market than she hoped. In a physics BS I learned how to handle 800 kinds of equations; that was less useful that I hoped. Signal analysis (think EE) and advanced applied statistics (GLM/Zs, time series) make you worth bank. The underlying understanding that I've gotten from the theory classes has been instrumental in using those things correctly, but I can see that a more abridged version would have been possible. Basic understanding of these things got me an offer at Abbott that makes me angry that I didn't take it.
posted by a robot made out of meat at 10:31 AM on February 4

Oh, and while asking around at the quantitative math whatever masters at my institution for Ms Robot, they said that the purpose of it is a) remediation for people who got poor/no skills in undergrad b) foreigners who have no degree from inside the USA and think that they need one. An undergraduate degree in physics, CS, non-systems non-chem non-bio engineering, statistics, math, actuarial science, econ (with quant. concentration), etc etc should be perfectly well prepared. I can't recommend enough taking summers to do a coop and/or research to address this.
posted by a robot made out of meat at 10:42 AM on February 4

Oh, and while asking around at the quantitative math whatever masters at my institution for Ms Robot, they said that the purpose of it is a) remediation for people who got poor/no skills in undergrad b) foreigners who have no degree from inside the USA and think that they need one.

Err, I don't know which master's you're referring to, but if you mean financial engineering, it's impossible to get into one of those programs if you have "poor/no skills in undergrad." They're insanely competitive as long as we're talking about the ones I mentioned. Check out the resumes at the linked NYU program, for example.

An undergraduate degree in physics, CS, non-systems non-chem non-bio engineering, statistics, math, actuarial science, econ (with quant. concentration), etc etc should be perfectly well prepared.

Are we talking about being an actuary or quant finance here? I'm not very familiar with the actuarial industry but I've heard that undergrad is indeed enough to get started as an actuary as long as you pass two of the preliminary exams or whatever. If you want a job in quantitative finance, my understanding (and the general consensus at, say, wilmott.com) is that you need a PhD in math or physics or one of these new MFE degrees just to get an interview. I've heard DE Shaw does hire promising Ivy League math undergrads who have won a bajillion math olympiads, but I think they're the exception.
posted by pravit at 8:21 PM on February 4

The program was MS in "Financial Mathematics." Quant Finance is a bigger field than I think you're imagining. For example, a friend of mine went straight from undergrad to an i banking firm where he does decidedly quantitative financial analysis; any econ/commerce/math department will show you similar cases from the recent graduates. That was a number of years ago, and maybe it's changed. On that job board I saw plenty of non-PhD/MSc jobs. There's also the fact that job boards like that are biased away from the more entry-level and more easily qualified positions. If you demand a PhD in ThisExactBrachOfMathEcon then you have to advertise a lot more. Note that that job board has only a handful of employers.
posted by a robot made out of meat at 8:38 AM on February 5

Note that that job board has only a handful of employers.
It has a handful of headhunting agencies posting job ads, but they recruit for all sorts of employers. Note how the ads always say "Top Tier Bank", "Global Investment Bank", etc. And all of the ads for more quantitatively oriented positions ask for a math/physics PhD or MFE specifically.

As an EE from an accredited institution, all I can say is that my resume bounced for all the quant finance-related internships I applied for, and I've had plenty of math, high GPA, etc. I imagine it's much tougher to secure an actual full-time position as an undergrad. Most of the advice I've received suggests going for the MFE or PhD to get a more quantitative role, so that's what I'm doing. YMMV.
posted by pravit at 11:38 AM on February 5

You know, I think that it's quite possible that the Career Services person who I was talking to about it didn't mean that it was for people with really bad skills, just different than Ms Robot. I can see how going from EE to finance might be challenging depending on the focus of your degree, and that an intermediate degree would be useful there.

Also, my friend who went undergrad to ibanking did so in the good days of the job market. Of course a good application would have internships/research, and it's wrong of me to downplay getting the right kind of math for the job. I still would say that we're talking about different worlds, with "jobs in the finance industry which are quantitative" being less picky than "quantitative finance" jobs.
posted by a robot made out of meat at 4:25 PM on February 5

I still would say that we're talking about different worlds, with "jobs in the finance industry which are quantitative" being less picky than "quantitative finance" jobs.
I think you've nailed it here. :)
posted by pravit at 4:43 PM on February 5