advice is needed for physics -> quant transition.

[Lipschitz]

Hello,

I believe my question will look quite naive, sorry for that. Moreover, there are hundreds of posts like this "what are my odds to do 'phd in physics -> quant' transition", but each situation is different and as I understand, the situation in finances usually changes fast!

I am a theoretical physicist, finished my phd 3 years ago, had a practical postdoc in a national lab and now I work as a theoretical physicist in industry for 1 year. I am in my late twenties. My field is very practical (vs other branches of theoretical physics). Essentially, I solve ODEs and PDEs. Analyze analytically and solve numerically. I have broad experience writing and using different numerical methods: finite difference, finite volume, finite elements, spectral methods, spectral elements, etc. + different linear/nonlinear solvers, Krylov methods, multigrid, different iterative solvers, Newton methods, jacobian free methods etc. with different time marching strategies, i.e., explicit, implicit, symplectic, imex, etc. This is all in HPC settings, so I do it in parallel - MPI, openMP. I am not a mathematician who develops those methods (I did a bit), but I use them to solve my physics problems and I write code to use such methods. I do not blindly solve equations numerically, and I frequently do some analytical work as well, so I am proficient with math.

In terms of programming I think I am okayish. I mostly write C, C++, and julia. I can probably read another 10 languages including python. I played with lisp and haskell in the past. Without preparation, I probably can remember basic data structures and algorithms, like trees, stacks, pipes, quick sort, merge sort, state machine - that type of things. I know some OOP.

I will definitely need to refresh algorithms, some statistics, and brain teasers - this is just a crude overview of my background.

Main problem is that I do not know finances at all. I took some micro/macro economics classes in my undergrad, but I forgot all of it.

What am I looking for: I want an intellectually challenging job with lots of math and programming (I love programming and I want to become better in it). I do not care if the math I do describes physics or finances... Salary improvement is very desirable. This looks like a "quantitative researcher" position. I want to relocate to NY and I do not need a visa to work. I have other personal reasons which motivate me to change.

My question:
1) What are my odds to find a quantitative researcher position without internships or additional degrees? I am willing to spend time to prepare of course.
2) Does this timeline "3 months preparing + 3 months preparing and looking for a job" look too unreasonable?
3) What do you think is the starting salary I can hope for? What about a salary in 5 years?
4) How is the "finance sector" doing right now? what are the odds to just lose your job in ~1year. How difficult will it be to find a new job again?
5) There are lots of "quant" positions... what would be the most math heavy?
6) Anything crucial I should be aware of?
7) I know that this is a big shift, and I still evaluating if I should do it and I am learning about my options. So if you were/are in similar situation, can you please comment your thoughts and if you are happy with your decision.
8) Any other advice is very much appreciated!

Thank you very much!

 

[KillingField]

1) Why do you so wish to avoid internships? It'll still pay probably double a typical postdoc salary. That said, given the postdoc and that I infer you're currently employed on a permanent basis, you're good to apply for full time positions. See also my answer to 8).
2) No.
3) Depends. This is well documented, so check for the specific role you're going for.
4) Whenever there are major market moves there are winners and losers. At the end of the day, trading is a zero sum game. Many of the biggest players have posted record trading profits.
5) Depends. There's a lot of different kinds of math out there.
6) Probably, but as you said, everyone at the PhD level is a bit different, so it's difficult to know what you don't know or should know.
7) I've not come across many who regret their decision. I did a PhD in physics and am extremely happy having left academia for finance, for a multitude of reasons.
8) Given your background in solving physics problems numerically and a lack of finance knowledge (and I assume lack of knowledge of financial math), a quantitative developer role may be the easiest route to break into the industry. I'd expect you to be able to get interviews for entry level quant positions, too. The only way of knowing is by applying.

 

[Lipschitz]

Dear KillingField,

Thank you very much for your response. It is very helpful.
I have some follow up questions, I hope you do not mind.
1) You are right, I should consider internship. The main problem is that I cannot afford significant salary drop right now. And I will be looking for 150k starting salary, which is unrealistic for intern. I probably should adjust my expectations. Do you agree?
8) Thank you for this advice. Quantitate developer seems like a possible route. Will it be difficult to transfer to analyst position later? From what I wrote about myself it might look like I am more like a coding monkey, which might be what I did on my first postdoc (not on my phd and my current job). However, one of the reason I look into finances direction is that I realized that I enjoy math more than physics and numerical methods, so the idea that I might need to study new branches of math motivates me. My main weakness in math I believe --- stochastic calculus/stochastic differential equations (I know plenty about ODEs and PDEs, some probability and statistics, lots of linear algebra and calculus, some functional analysis, some abstract algebra, some differential geometry, some lie groups & symmetries theory, some perturbation theory, some optimization - I believe most of it is not relevant). In any rate, I will try to transfer to analyst position if I be able to break in finances world...

Thank you!

 

[Daniel Duffy]

You can easily learn SDE if you know ODE/PDE and good real analysis and a wee bit of Measure Theory (inflated importance).

 

[binomial-torrent]

Dear KillingField,

Thank you very much for your response. It is very helpful.
I have some follow up questions, I hope you do not mind.
1) You are right, I should consider internship. The main problem is that I cannot afford significant salary drop right now. And I will be looking for 150k starting salary, which is unrealistic for intern. I probably should adjust my expectations. Do you agree?
8) Thank you for this advice. Quantitate developer seems like a possible route. Will it be difficult to transfer to analyst position later? From what I wrote about myself it might look like I am more like a coding monkey, which might be what I did on my first postdoc (not on my phd and my current job). However, one of the reason I look into finances direction is that I realized that I enjoy math more than physics and numerical methods, so the idea that I might need to study new branches of math motivates me. My main weakness in math I believe --- stochastic calculus/stochastic differential equations (I know plenty about ODEs and PDEs, some probability and statistics, lots of linear algebra and calculus, some functional analysis, some abstract algebra, some differential geometry, some lie groups & symmetries theory, some perturbation theory, some optimization - I believe most of it is not relevant). In any rate, I will try to transfer to analyst position if I be able to break in finances world...

Thank you!

you could be a "quant analyst" on paper and still be a code monkey. or you could be a "quant developer" or "research engineer" working on numerical algorithms and bit manipulations. depends more on the shop and your team

 

[Lipschitz]

Thanks Daniel Duffy and binomial-torrent!

 

[Lipschitz]

Hi LilyBtown. You said "was". Did you finish your transfer? Do you like it? Any advises?

 

[Berenger]

... I cannot afford significant salary drop right now. And I will be looking for 150k starting salary, which is unrealistic for intern. I probably should adjust my expectations. Do you agree?...

If it's $/£/€, these are pretty wild expectations.
Take a look
glassdoor, new york, quant analyst + benefits not stated

Your background however sounds pretty solid in terms of knowledge, if you can translate it into interviewee skills you should be in a pretty good place.