Forums  > Careers  > Quant as fallback for Math grad student?  
     
Page 1 of 1
Display using:  

dmg_8


Total Posts: 16
Joined: Jul 2010
 
Posted: 2010-07-14 20:53
Hi,

I'm beginning graduate school in mathematics this fall. My interest is in pure math, but I'm not sold completely sold on an academic career yet. So I'm completing the MS and going to reevaluate things from there (getting funded, so why not).

First thing, does having an MS (as opposed to a PhD) even make quantitative finance jobs a realistic option?

Second, if I focus my research on something esoteric and pure, will that make me a lot less desirable than if I were to focus on something more computational and applied?

Finally, regarding programming: I'm fairly proficient in C (done a lot of my own projects, programmed for a mathematician for a year and a bit, worked with a lot of libraries, etc.). I like the low-level stuff, and play around with ASM as well. Is there any demand for quants to code in C and assembler (particularly the latter)? Beyond this, should I be adding anything else to my arsenal (please don't say Java!).

Thanks,
dmg_8

danko


Total Posts: 124
Joined: Mar 2007
 
Posted: 2010-07-15 00:33
Yo,

Some firms do require PhDs (BlueCrest for their quant researchers for example), and you would most likely be competing with them for similar roles (maybe at a bit lower pay) but generally the MS is fine.

As long as you are good at programming, I doubt it on the pure vs. applied question. Pure math guys obviously get a lot of respect intellectually. However, I suggest putting some time in to ensure that you're quick with your Strang-level Linear Algebra and Degroot-level probability and statistics as these sorts of things are considered fundamental and often come up in interviews.

"Is there any demand for quants to code in . . . assembler?" Haha, that is a good one. What do you think? You could tell them that for every month you were were on payroll, their code would run 0.001% faster with only modest deterioration in reliability. Due to the power of compounding, after a few years the improvement might almost be perceptible.

I bet you'd love this paper, (see the table on p. 5), which boastingly recounts a breakthough in the race to achieve big flops in the stupidest way possible.

You need to learn c++ or Java, probably the former because it is more common (and knowledge of it is a good sign you can become decent in the other two relatively quickly because it is the hardest), and you need to have a good understanding of Object Oriented programming. Expect to be aggressively grilled on c++ in interviews, so you need to know the stuff a lot better than is standard among academics.




dmg_8


Total Posts: 16
Joined: Jul 2010
 
Posted: 2010-07-15 00:58
My programming is probably not "good" relative to anyone whose smart and does it for a living, and my C++ is just C compiled with g++. So that'll have to be worked on over the next year or so. I don't have a good handle on exactly the sorts of things quants are coding, so just through ASM out there as a possibility.

If it won't make too much of a difference, I guess I'll just do my research in whatever I feel interested in, applications be damned. (Hello C*-algebras!)

P.S., thanks for the link. Lack a PDF reader right here, will have a look later.

danko


Total Posts: 124
Joined: Mar 2007
 
Posted: 2010-07-15 01:13
"I don't have a good handle on exactly the sorts of things quants are coding, so just through ASM out there as a possibility."

Recall that the objective is to maximize money/time.

Also, just to clarify, the primary purpose of the paper was humor. I wasn't suggesting you actually read the whole thing.


mj


Total Posts: 1049
Joined: Jun 2004
 
Posted: 2010-07-23 12:36
Do a PhD in stochastic calculus -- it's very hard and very pure. It's also a very easy side step into quant.


pj


Total Posts: 3613
Joined: Jun 2004
 
Posted: 2010-07-23 13:18
> Do a PhD in stochastic calculus -- it's very hard and
> very pure. It's also a very easy side step into quant.
So true...

вакансия "Программист Психологической службы" -але! у нас ошибко! не работает бля-бля-бля -вы хотите об этом поговорить?

Graeme


Total Posts: 1629
Joined: Jun 2004
 
Posted: 2010-07-23 13:57
and it's VERY good advice.

Graeme West

doctorwes


Total Posts: 577
Joined: May 2005
 
Posted: 2010-07-27 18:54
By the way, thinking of a career option as a "fallback" is not a recipe for success.

A quick Google search turns up several connections between C*-algebras and stochastic calculus (e.g. this thesis, and reference [69] therein, where C*-algebras are used to construct Gaussian fields and Poisson processes - I haven't read these) so perhaps you could work on something interesting and quant-relevant at the same time?



filthy


Total Posts: 1265
Joined: Jun 2004
 
Posted: 2010-07-27 19:13
how many quant jobs really require stochastic calculus? i know it is often used as an interview filter but aren't most jobs much more reliant on numerical methods now?

"Game's the same, just got more fierce"

dmg_8


Total Posts: 16
Joined: Jul 2010
 
Posted: 2010-07-27 19:24
Actually I've looked into that since I originally made that post. I talked to a few professors about this and some have suggested some applications of stuff I'm interested in (nowhere near as specific as C* algebras, those links sound interesting) to finance. My knowledge of dynamical systems/probability theory needs to be sharpened before I can really get a grip on the stuff (the prof sent me a paper), but it looks interesting on the surface.

I've just seen probability theory from a rigorous point of view for the first time (using Folland's book 'Real Analysis') and I like it a lot. I found the basic stats courses I've taken sort of dull, but looking at all the measure theory, harmonic analysis, and distribution theory behind Folland's exposition it seems there'd be no shortage of deep, interesting math with possible applications to quant finance.

dmg_8


Total Posts: 16
Joined: Jul 2010
 
Posted: 2010-07-27 19:24
oops, double post

filthy


Total Posts: 1265
Joined: Jun 2004
 
Posted: 2010-07-27 20:13

i think you are close to making the mistake of trying to find something both interesting enough for you to study for 3-4 years and it also generally useful in finance. most of the commonly used stuff in finance isn't very deep, certainly not by pure math standards and not even really by physics standards. remember that most papers are much more rigorous (ie full of crap) than the stuff actually done in banks and funds.

if you are going to be doing a pure math phd you are clearly smart. why don't you chose something you love and then do the best thesis possible? if it doesn't work out you can read hull in a week and become a trader or a quant.

but if you chose something that you don't love (at least at the start because you will hate the entire world at the end...) you will do a crappy job, you won't get decent academic offers and you may still find it hard to get into quant.

practical trading math is trivial. worry about it later.


"Game's the same, just got more fierce"

ClopeMan


Total Posts: 328
Joined: Jan 2007
 
Posted: 2010-07-27 20:57
totally agree with filty; follow your natural interests, worry about the rest later on, there will be plenty of time (and you might like academic research/teaching as well)

my best advice: get yourself a 4 ish months summer internship in a bank (or whatever trading shop) BEFORE you complete your phd thesis; that's by far easier to get than a full time job and makes your cv more competitive, on top of giving you a taste of the type of work you might end up with before you make a final move... (same advice hold on teaching a course before you complete your thesis)



twisted objective mind

dmg_8


Total Posts: 16
Joined: Jul 2010
 
Posted: 2010-07-28 01:43
Heh, maybe using the word "fallback" sent the wrong impression - finance is of interest to me (more generally, money, politics, power, etc.) in its own right. I'm only embarking on a masters right now, continuing for a PhD afterwards is a possibility. I love math, but I don't know if it will be fulfilling in the longterm. Spending obscene amounts of time doing nothing but math (as I've done for 2 years now) will eventually take a toll (social isolation, dulling of personality, etc.). I read Alexandre Grothendieck's description of this and disturbingly it felt familiar.

In any case, I want to at some point become a "money-man". Programming, applied math, leadership, networking, interpersonal relationships, salesmanship, and all the other things associated with the business world seem more sustainable (in terms of leading a balanced life) in the longterm. Not sure when, but eventually I want to explore all of that. So for the next year I'll continue building my up my math and programming to lay a foundation for that while continuing to dig deeper into abstract theory. If I go on for the PhD (in something pure) I'll be able sleep a little better at night knowing my programming/computational skills (which can generate money more readily than abstract theory) are sharp.

edit to add: thanks for the advice though, will look into internships for next summer.

Scotty


Total Posts: 727
Joined: Jun 2004
 
Posted: 2010-07-28 04:29

No, I think "fallback" sums it up alright.  You've surmised that maths is not likely to be a lucrative career option, so you figured you can make money in Finance.  You don't seem to be interested in Finance per se as in the money, right?

For the last decade or more we have had a lot of guys like that come into the industry for these reasons. They have had a certain technical proficiency (maths or physics, usually) but a very weak understanding of finance and finance theory.  It has not be very constructive.


“Whatever you do, or dream you can, begin it. Boldness has genius and power and magic in it.”

jslade


Total Posts: 1246
Joined: Feb 2007
 
Posted: 2010-07-28 08:40
Threadjack: do you have that Grothendieck reference handy?

FWIIW, the standard math escape hatch I know about is actuarial science. It's not as sexy as being a quant nerd (as if that were even possible), but it seems to be a straightforward series of hoops to jump through, and it provides a decent, low-key lifestyle. Quite a number of my mathematical pals have done this; none of them are unhappy with their decision. You don't even have to be good at computer science for that sort of thing.

"Learning, n. The kind of ignorance distinguishing the studious."

frstwrldprblm


Total Posts: 271
Joined: Apr 2008
 
Posted: 2010-07-28 15:39
@ dmg_8
one thing I will say as a lower rung poster on here (regarding experience and academic level):

a MS will definitely help you stand out, and opportunities will come your way conditional on strong internships. if you are doing math I would definitely pick up Hull, and then go read (www.markjoshi.com/downloads/advice.pdf) mj's advice paper, DC's quant job guide and just try and get a feel for things. don't expect hull to be a very challenging read, that is not its intention. treat hull like a bench guide for products. product knowledge is what you will not have exposure to in your current masters, but it is not a tough thing to get a high level understanding of.

i heavily agree with the internship points people have raised. i completed one strong internship, and one that i made strong by my own efforts (doing more than asked just to challenge myself). this alone will get you attention.

i would say learning coding look at C++ (C++, Java or C# is a usual requirement). Some places are flexible (not all), but i would recommend C++ (if you know C++, Java will seem easier (converse does not hold), and if you know C++ and Java then a language like Python will seem easier). wrt difficulty, Java is easier than C++ but if you have been able to do C, C++ should not be an issue.

just try hard at the internships (pick those carefully) and there is no reason that what you want will not become within reach.

diogenes


Total Posts: 78
Joined: Apr 2006
 
Posted: 2010-08-10 05:12
Threadjack: do you have that Grothendieck reference handy?

+1 on the reference request.

If you want power, then I would stay away from being a quant.

dmg_8


Total Posts: 16
Joined: Jul 2010
 
Posted: 2010-08-19 16:54
jslade, diogenes:

Pardon the slow reply. If you can read french, try and track down a copy of "Recoltes et semailles" ("Reaping and Sowing"), that is AG's autobiography. Previously it was posted here:

http://www.grothendieckcircle.org/

but apparantely AG has recently demanded that (among other things he's authored) be removed from the public domain. Nonetheless, there's plenty of interesting (English) biographical info there.

Regarding what I mentioned earlier, try these two links:

http://www.ams.org/notices/200409/fea-grothendieck-part1.pdf

http://www.scribd.com/doc/35435936/As-If-Summoned-from-the-Void-The-Life-of-Alexandre-Grothendieck-Part2

Specifically, look at the section under the heading "A Spirit in Stagnation" in the second one. But I'd recommend reading the whole thing, very interesting.

It's interesting that it appears other mathematicians have gone through similar withdrawals (for lack of a better word) as AG (e.g., Perelman, Godel, ). And probably many other non-famous examples we'll never hear about (hence my concerns giving your entire self to math could be harmful).


dmg_8


Total Posts: 16
Joined: Jul 2010
 
Posted: 2010-08-19 16:59
Scotty,

Well you could argue anything is a "fallback" if you want to go far enough. I like math/science/programming. Ideally, I'd have no job and just explore those at my leisure. But I like to eat, thus that is not a lucrative option. So I'll be "falling back" in some sense no matter what.
Previous Thread :: Next Thread 
Page 1 of 1