• C++ Programming for Financial Engineering
    Highly recommended by thousands of MFE students. Covers essential C++ topics with applications to financial engineering. Learn more Join!
    Python for Finance with Intro to Data Science
    Gain practical understanding of Python to read, understand, and write professional Python code for your first day on the job. Learn more Join!
    An Intuition-Based Options Primer for FE
    Ideal for entry level positions interviews and graduate studies, specializing in options trading arbitrage and options valuation models. Learn more Join!

How to transition from postdoc in applied maths/physics to quant research?

I have a PhD in applied maths/astrophysics at a top UK university which isn't Oxford or Cambridge. I recently started working as a research fellow at the same university.

I want to work as a quant researcher. A recruiter for G-Research recently approached me. I was tempted to apply. However, after reading their practice interview exam, I realised my probability and statistics knowledge isn't good enough.

I am 25, and my background is in applied maths, physics and numerical analysis. I know a lot about solving differential equations and high-performance computing but very little about finance, probability and statistics.

I am unsure whether to teach myself the skills needed to be a quant researcher or apply for a masters degree.

I have been reading 'The concepts and practice of mathematical finance' by Mark S. Joshi and reading 1st and 2nd-year undergraduate lecture notes on probability and statistics. I'm worried that the book by Mark S. Joshi might be a bit out of date. For example, it does not mention machine learning which seems to be important these days. With my full-time job, I struggle to find the time/energy, so I am tempted to quit my job and dedicate my time to building my skillset.

Are there any books or courses you would recommend for me to get into quant research? Would you recommend a masters course? I have enough savings to pay a for masters.
 
Last edited:
I have basically no experience using it. The only programming languages I am comfortable with are Fortran, Python and a more obscure one called IDL.
 

Daniel Duffy

C++ author, trainer
Fortran is great, but very very niche. Python is popular. C++ is a skill.

background is in applied maths, physics and numerical analysis. I know a lot about solving differential equations and high-performance compute

I have a new PDE/FDM book (Wiley) coming out in a few months, so right up your alley?
 

Attachments

  • PDE2020.docx
    18.9 KB · Views: 56
Fortran is great, but very very niche. Python is popular. C++ is a skill.
I was thinking of reading Mark S. Joshi's 'C++ Design Patterns and Derivatives Pricing' after I finish with 'The concepts and practise'. Would you say that was a good idea? Or are there different books you would recommend?
 

Daniel Duffy

C++ author, trainer
I was thinking of reading Mark S. Joshi's 'C++ Design Patterns and Derivatives Pricing' after I finish with 'The concepts and practise'. Would you say that was a good idea? Or are there different books you would recommend?
This C++ book by the late Dr. Joshi is severely outdated.
The best way to learn C++ is to take Quantnet C++ online course.
One learns C++ by doing, not reading.
I have several books (lost count how many).

See also
 
This C++ book by the late Dr. Joshi is severely outdated.
The best way to learn C++ is to take Quantnet C++ online course.
One learns C++ by doing, not reading.
I have several books (lost count how many).

See also
would you have a cheaper alternative to suggest than quantnet ones for c++ ?
 

Daniel Duffy

C++ author, trainer
I have no real idea. Like 2nd hand cars, they can be cheap and cheerful, and pretty much useless.
QN C++ is great value for money.
 
I have a PhD in applied maths/astrophysics at a top UK university which isn't Oxford or Cambridge. I recently started working as a research fellow at the same university.

I want to work as a quant researcher. A recruiter for G-Research recently approached me. I was tempted to apply. However, after reading their practice interview exam, I realised my probability and statistics knowledge isn't good enough.

I am 25, and my background is in applied maths, physics and numerical analysis. I know a lot about solving differential equations and high-performance computing but very little about finance, probability and statistics.

I am unsure whether to teach myself the skills needed to be a quant researcher or apply for a masters degree.

I have been reading 'The concepts and practice of mathematical finance' by Mark S. Joshi and reading 1st and 2nd-year undergraduate lecture notes on probability and statistics. I'm worried that the book by Mark S. Joshi might be a bit out of date. For example, it does not mention machine learning which seems to be important these days. With my full-time job, I struggle to find the time/energy, so I am tempted to quit my job and dedicate my time to building my skillset.

Are there any books or courses you would recommend for me to get into quant research? Would you recommend a masters course? I have enough savings to pay a for masters.
Try to steer your research toward having some kind of stochastic process being involved in it, and get back to the recruiter in 2/3 years time. Maybe complete the CQF by Wilmott part-time in the meanwhile, so to acquire the basics in the field.

You don't need a lot of background in Probability really, none in Statistics, to read Concepts and Practice by M. Joshi, what is required is: knowledge of Normal distribution (he explains Lognormal directly in the book), how to take expected value and variance, Central Limit theorem, and basics on covariance/correlated matrices.
What you really need is the time and energy to learn a good 360 pages, out of 428 overall, as if you were preparing for an exam.
If you hold a postdoc position and have tight deadlines, trying to scrape together research proposals etc, you most likely cannot afford the effort.
Maybe a simpler, more compact, less verbose book may help accomplish the goal (I went through Concepts, found it useful, but perhaps I could have been better served by the more succinct Baxter-Rennie book, seen as it receives good press, and both books are steered more towards solving for prices through martingales-Montecarlo, rather than PDE-finite differences as in the Wilmott book).

C++ could be learned through the C++ Primer by Lippmann, but that's another 1237 pages (new edition out Oct 2021) ...

You definitely have the background and potential to learn the ropes in this field... It doesn't need a lot of Probability knowledge, at least sell-side, but you will have to set aside years of self-study if you feel a financial engineering degree is too costly and a bit of a step backwards for a postDoc, and a bit of a gamble.
 
Try to steer your research toward having some kind of stochastic process being involved in it, and get back to the recruiter in 2/3 years time. Maybe complete the CQF by Wilmott part-time in the meanwhile, so to acquire the basics in the field.

You don't need a lot of background in Probability really, none in Statistics, to read Concepts and Practice by M. Joshi, what is required is: knowledge of Normal distribution (he explains Lognormal directly in the book), how to take expected value and variance, Central Limit theorem, and basics on covariance/correlated matrices.
What you really need is the time and energy to learn a good 360 pages, out of 428 overall, as if you were preparing for an exam.
If you hold a postdoc position and have tight deadlines, trying to scrape together research proposals etc, you most likely cannot afford the effort.
Maybe a simpler, more compact, less verbose book may help accomplish the goal (I went through Concepts, found it useful, but perhaps I could have been better served by the more succinct Baxter-Rennie book, seen as it receives good press, and both books are steered more towards solving for prices through martingales-Montecarlo, rather than PDE-finite differences as in the Wilmott book).

C++ could be learned through the C++ Primer by Lippmann, but that's another 1237 pages (new edition out Oct 2021) ...

You definitely have the background and potential to learn the ropes in this field... It doesn't need a lot of Probability knowledge, at least sell-side, but you will have to set aside years of self-study if you feel a financial engineering degree is too costly and a bit of a step backwards for a postDoc, and a bit of a gamble.
Would you have other advice regarding learning c++ ? , I'm in a different situation than peanutlex but would like to learn c++ for a similar goal without having to pay a price as high as 1000 $
 
Would you have other advice regarding learning c++ ? , I'm in a different situation than peanutlex but would like to learn c++ for a similar goal without having to pay a price as high as 1000 $
There are tons of videos on youtube that teach you C++. I personally like Tech with Tim (though I watched him for Python). If you want to do it for certificate/admissions purposes then you might have to bite the bullet. But I know you already got into a bunch of top programs so if you're doing it a prep then I think youtube will serve you fine. Just remember to code often so you don't forget!
 
Would you have other advice regarding learning c++ ? , I'm in a different situation than peanutlex but would like to learn c++ for a similar goal without having to pay a price as high as 1000 $
Either $1000 or 1000 pages, I am convinced there is no alternative, quicker way. I don't think videolessons or web tutorials will quite cut it.
There is Accelerated C++ by Koenig at 300 pages and is highly praised, but omits to explain the modern features of C++ as it is from year 2000.
This is probably the best online free resource on C++. Perhaps you could read Accelerated C++, complete all exercises, and integrate the C++11/14 new features by studying them on learncpp.com. You should be able to identify by yourself what these new features are.
Even this solution will take months of self-study.

Then there is the issue of Object Oriented Design Patterns, that they will invariably require you to know for Quant Developer roles that are not internships or graduate career-start roles.
Maybe Joshi's C++ book, + the Quantlib implementation book by Ballabio can help in clarifying which are the most used patterns in quant finance. On design patterns there is also Daniel Duffy's latest book which should be up to date, but at about 1000 pages can only serve as reference. The author prowls around this forum, and will certainly not miss this opportunity to expand upon the virtues of his own book.
 
Last edited:
Try to steer your research toward having some kind of stochastic process being involved in it, and get back to the recruiter in 2/3 years time. Maybe complete the CQF by Wilmott part-time in the meanwhile, so to acquire the basics in the field.

You don't need a lot of background in Probability really, none in Statistics, to read Concepts and Practice by M. Joshi, what is required is: knowledge of Normal distribution (he explains Lognormal directly in the book), how to take expected value and variance, Central Limit theorem, and basics on covariance/correlated matrices.
What you really need is the time and energy to learn a good 360 pages, out of 428 overall, as if you were preparing for an exam.
If you hold a postdoc position and have tight deadlines, trying to scrape together research proposals etc, you most likely cannot afford the effort.
Maybe a simpler, more compact, less verbose book may help accomplish the goal (I went through Concepts, found it useful, but perhaps I could have been better served by the more succinct Baxter-Rennie book, seen as it receives good press, and both books are steered more towards solving for prices through martingales-Montecarlo, rather than PDE-finite differences as in the Wilmott book).

C++ could be learned through the C++ Primer by Lippmann, but that's another 1237 pages (new edition out Oct 2021) ...

You definitely have the background and potential to learn the ropes in this field... It doesn't need a lot of Probability knowledge, at least sell-side, but you will have to set aside years of self-study if you feel a financial engineering degree is too costly and a bit of a step backwards for a postDoc, and a bit of a gamble.
Thank you for the advice. It's a bit of a shame that I will have to set aside years of self-study. I was hoping to get there within about a year if I were to quit my job and learn the skills full-time.

It will be pretty tricky to steer my research as my funding is for a specific project and my line manager is very autocratic.

I am open to doing a masters. I am someone who learns better in a structured course as opposed to reading a textbook by myself. However, I don't know if I can justify the cost. Also, is there a danger I could be seen as someone incapable of leaving a university? I wouldn't say I am bothered about taking a step back. The main thing I am after is a fulfilling job and where I feel like my efforts are adequately rewarded.
 
Last edited:
Thank you for the advice. It's a bit of a shame that I will have to set aside years of self-study. I was hoping to get there within about a year if I were to quit my job and learn the skills full-time.

It will be pretty tricky to steer my research as my funding is for a specific project and my line manager is very autocratic.

I am open to doing a masters. I am someone who learns better in a structured course as opposed to reading a textbook by myself. However, I don't know if I can justify the cost. Also, is there a danger I could be seen as someone incapable of leaving a university?
Perhaps you could stay where you are, and do as you are autocratically told, until you have saved the amount of money required for a master.
In the meanwhile on your own time you should learn C++ though, that's the most critical skill to possess. You could integrate some C++ in your research as a way.
Having attained the PhD should dispell any notion of you "being incapable of this or that", the important thing then is to do well in interviews.
 
Perhaps you could stay where you are, and do as you are autocratically told, until you have saved the amount of money required for a master.
In the meanwhile on your own time you should learn C++ though, that's the most critical skill to possess. You could integrate some C++ in your research as a way.
Having attained the PhD should dispell any notion of you "being incapable of this or that", the important thing then is to do well in interviews.
I have enough savings to do a masters now. I guess the pros of doing one would be:
  • The structured course means I will learn the work more efficiently.
  • Build up a network (although during COVID this might not be true).
  • A certificate that proves I have done the work, which is harder to prove in self-study.
The cons would be:
  • Costs about £30-£40k.
  • If I go to a university in London the living costs will be much more expensive than where I am currently living.
  • I may decide midway through that this is not for me.
Do you think I missed anything important in this pros and cons list?
 
Do you think I missed anything important in this pros and cons list?
No. I agree it is a bit of a plunge, but ultimately you'll have to decide by yourself if you want to stay in academia or not, and if not whether to pursue a career in quantitative finance, or in some other technical field, if there is any that attracts you more.
 
No. I agree it is a bit of a plunge, but ultimately you'll have to decide by yourself if you want to stay in academia or not, and if not whether to pursue a career in quantitative finance, or in some other technical field, if there is any that attracts you more.
What would a masters for £37.5k at, say Imperial College London give me that the CQF (for £14k) wouldn't?
 
What would a masters for £37.5k at, say Imperial College London give me that the CQF (for £14k) wouldn't?
I cannot really say for certain as I have frequented neither.
CQF seems more oriented towards people currently employed in finance who want to know more about quantitative finance, many are sent there by their own banks/employers who pay all fees.
A regular Master of Science will entail all the rigours of university learning.
 
I am an undergrad in math and stats with not too much coding experience or quant work experience.. i do not know what kind of quant role i want
I am going to start my MS applied math degree in fall 2021
When is a good time to start applying for summer 2022 internships? and when does the application period begin and end ?
How can i make my application strong for internships? I am already doing coding courses in python and revising math and stats topics from my undergrad curriculum
 
Top