• 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!

MSc. Theoretical Physics or Computer Science

Joined
10/14/13
Messages
37
Points
18
I am aware that an MFE programme is the obvious candidate to become a quant, however, I would prefer to teach myself most of the finance and focus my efforts on making myself a more rounded graduate student. I may even do an MFE after my MSc and/or PhD anyway if I feel it necessary.

Quant finance is one of the possible career paths I'm seriously considering and so I would like to hear thoughts on which programme would be more relevant or more highly thought of within the quant industry.

The courses I have applied for are:
MSc. Computer Science (Imperial)
MSc. Theoretical Physics (Edinburgh)
MSc. Theoretical Physics (Kings College London)

Just to say that I have read around this forum and I've picked up themes that the 90s and 00s were the era of the physicists and that 10s are the years of the computer scientists. In both Theoretical Physics programmes, however, I can choose optional modules in areas like like machine learning etc.
 
Are these courses good preparation in industry?

90% of stuff in all courses had little application. What is important:

1. Analytical thinking
2. Practical skills I (arithmetic)
3. Practical skills II (can you program, really?)
4. Soft skills.
 
I guess an MSc in Theoretical Physics smashes all those criteria then.

By 'smash' == 'satisfies'

What does such a course entail, these days? What kinds of topics?

IMO a good quant/engineer can take a model, design and code all by hisself.
 
Are these courses good preparation in industry?

90% of stuff in all courses had little application. What is important:

1. Analytical thinking
2. Practical skills I (arithmetic)
3. Practical skills II (can you program, really?)
4. Soft skills.

In the case of CS, a lot is devoid of any real math , so 1. is not necessarily a deliverable IMO.
 
This is the core course of KCL:

choose a minimum of 5 modules (max of 8) from below &
  • Advanced General Relativity
  • Advanced Quantum Field Theory
  • Foundations of Mathematical Physics
  • Lie Groups & Lie Algebras
  • Low-Dimensional Quantum Field Theory
  • Manifolds
  • Mathematical Methods for Theoretical Physics
  • Quantum Field Theory
  • Quantum Mechanics II
  • Spacetime Geometry & General Relativity
  • Standard Model Physics & Beyond
  • String Theory & Branes
  • Supersymmetry
The remaining modules can be drawn from a range of theoretical physics or pure mathematics MSc modules available in London, the Financial Mathematics MSc in King's. I also believe you can choose computing modules.

Edinburgh follows a similar structure as well.

The physics modules will probably not be hugely relevant for industry but will require quite difficult maths especially in linear algebra, calculus. Some of the modules would require high performance computing simulations ie. like you say, taking a model and getting on with the coding and design
 
This is the core course of KCL:

choose a minimum of 5 modules (max of 8) from below &
  • Advanced General Relativity
  • Advanced Quantum Field Theory
  • Foundations of Mathematical Physics
  • Lie Groups & Lie Algebras
  • Low-Dimensional Quantum Field Theory
  • Manifolds
  • Mathematical Methods for Theoretical Physics
  • Quantum Field Theory
  • Quantum Mechanics II
  • Spacetime Geometry & General Relativity
  • Standard Model Physics & Beyond
  • String Theory & Branes
  • Supersymmetry
The remaining modules can be drawn from a range of theoretical physics or pure mathematics MSc modules available in London, the Financial Mathematics MSc in King's. I also believe you can choose computing modules.

Edinburgh follows a similar structure as well.

The physics modules will probably not be hugely relevant for industry but will require quite difficult maths especially in linear algebra, calculus. Some of the modules would require high performance computing simulations ie. like you say, taking a model and getting on with the coding and design

I'm a bit sceptical to be honest. My 2 cents..

I did QM, SR and GM while at university at undergrad level. I cannot see why they would be useful.

Lie theory is too theoretical.

linear algebra, calculus
That's about 1st year maths.

Has this generation lost the plot? What about the Lagrange, Hamilton, .... PDE, FDM, numerical analysis are betrer.

high performance computing simulations
Big chance you will be doing MPI in Fortran, not exactly skills QF is looking for IMO.

Physics uses maths it is not 'hard maths' in the technical sense. But better than CS IMO.
 
Last edited:
From the stance of QF these courses are a monumental waste of time. I was going to say the same things Daniel Duffy is. What one interested in QF should be doing is mastering basic skills really well. Using sheaf cohomology isn't one of them. And just taking all these courses will merely addle your brain and make you useless for doing the simple things really well.
 
I just dumbed down the maths because I didn't know if you would know the various areas... Hamiltonian, Lagrangians, Complex number theory (contour integrations etc.), PDEs/ODEs, Stokes etc. these are all covered in standard BSc physics degrees.

@Daniel Duffy I've already done MPI/OpenMP and HPC in Fortran so I'd be doing it all C++ this time round

@bigbadwolf I can see where you're coming from but would an MSc in computer science be any better is the fundamental question?

I guess my point is that I want an eduction in something purer than what an MFE offers and I might do an MFE afterwards as a way of focusing in on finance. I am wondering which of these areas might tie in the best with an MFE
 
Last edited:
I just dumbed down the maths because I didn't know if you would know the various areas... Hamiltonian, Lagrangians, Complex number theory (contour integrations etc.), PDEs/ODEs, Stokes etc. these are all covered in standard BSc physics degrees.

@Daniel Duffy I've already done MPI/OpenMP and HPC in Fortran so I'd be doing it all C++ this time round

@bigbadwolf I can see where you're coming from but would an MSc in computer science be any better is the fundamental question?

I guess my point is that I want an eduction in something purer than what an MFE offers and I might do an MFE afterwards as a way of focusing in on finance. I am wondering which of these areas might tie in the best with an MFE
Fair enough. My conclusions remain unaltered.

The C code (I don't say C++) I've see in Physics is usually (very)weak and badly designed. In alll fairness, it's not core business.

Of course, Physics for its own sake is great fun. As BBW said, it's not numerate or practical enough.
 
I guess I wanted to avoid a Physics vs. MFE debate as the obvious winner is an MFE. However, I'm equally interested in doing an MSc in computer science or physics "for fun" I guess.
I'm using future prospects in QF as something to try and decide which I will do "for fun" before thinking about the possibility of becoming a quant and maybe doing an MFE
 
I can see where you're coming from but would an MSc in computer science be any better is the fundamental question?

If it's an MSc for those who already have a BSc in computing, then there will be courses in things like compiler design, computability theory, complexity theory, etc. -- none of which you really need in QF. If it's one of those conversion course MScs, then the programming courses might -- just might -- stand you in good stead. But my impression has been that they shy away from C/C++ and teach Java. The root of the problem, of course, is the disconnect between academia and the real world. Trust me, the MSc in theoretical physics leaves you neither here nor there. The courses serve merely as samples, as appetisers. Lie theory is an ocean unto itself. Likewise for QFT. Likewise for differential geometry and differential topology. It takes years to master any of these. What you will get is a smorgasbord of courses force-fed down your throat that will leave you reeling by year-end. And no real-world skills to speak of.
 
BTW the most famous quant with a PhD (Trinity College, Dublin) in Relativity is probably Phelim Boyle. But he also studied Accountancy and Maths as well.

PB is the originator of Monte Carlo for option pricing and the trinomial method.
 
While your reviewing his education can either of you comment on this combination of education:

University of Washington - Computational Finance & Risk Management Curriculum:
http://depts.washington.edu/compfin/sites/default/files/ctools/UW_Computational-Finance & Risk Management_Brochure_Final_080613.pdf

+

University of Washington - MS in Applied Math Curriculum:
-Vector Calculus and Complex Variables
-Scientific Computing
-Applied Linear Algebra & Introductory Numerical Methods
-Introduction to Dynamical Systems and Chaos
-Calculus of Variations
-Computational Methods for Data Analysis
-Numerical Analysis of Boundary Value Problems
-Dynamical Systems
-Methods for Partial Differential Equations
-High-Performance Scientific Computing
-Numerical Analysis of Time Dependent Problems

UW CFRM program is slightly low on Math compared to other MFE but if put together with their Applied Math program would it be a good combination of education?
 
Last edited:
While your reviewing his education can either of you comment on this combination of education:

University of Washington - Computational Finance & Risk Management Curriculum:
http://depts.washington.edu/compfin/sites/default/files/ctools/UW_Computational-Finance & Risk Management_Brochure_Final_080613.pdf

+

University of Washington - MS in Applied Math Curriculum:
-Vector Calculus and Complex Variables
-Scientific Computing
-Applied Linear Algebra & Introductory Numerical Methods
-Introduction to Dynamical Systems and Chaos
-Calculus of Variations
-Computational Methods for Data Analysis
-Numerical Analysis of Boundary Value Problems
-Dynamical Systems
-Methods for Partial Differential Equations
-High-Performance Scientific Computing
-Numerical Analysis of Time Dependent Problems

UW CFRM program is slightly low on Math compared to other MFE but if put together with their Applied Math program would it be a good combination of education?
At a glance, MS seems to be doing practical topics.

Just my 2 cents of course.
 
From the stance of QF these courses are a monumental waste of time. I was going to say the same things Daniel Duffy is. What one interested in QF should be doing is mastering basic skills really well. Using sheaf cohomology isn't one of them. And just taking all these courses will merely addle your brain and make you useless for doing the simple things really well.

Agreed. None are relevant to quant finance, and doing a course like this will not help you.

In the UK, the most important factor is brand name. Imperial has the best brand name in your list, but doing a masters in CS is a monumental waste of time.

Finally, what do you want to do with your life? If you want to be a quant, all of those courses are detrimental. If you want to be a physicist, then why have you applied to a CS course?

You sound utterly confused.

Make no mistake about it - if you want a quantitative career, the single most important thing is to be a good programmer. You can't learn that in a year, so any taught course is largely a waste of time, unless you're going to Harvard.
 
BTW the most famous quant with a PhD (Trinity College, Dublin) in Relativity is probably Phelim Boyle. But he also studied Accountancy and Maths as well.

PB is the originator of Monte Carlo for option pricing and the trinomial method.

Who? I wish people would stop looking at their alma mater with such rose-tinted specs. Nobody cares about TCD, especially not in finance.
 
If it's an MSc for those who already have a BSc in computing, then there will be courses in things like compiler design, computability theory, complexity theory, etc. -- none of which you really need in QF. If it's one of those conversion course MScs, then the programming courses might -- just might -- stand you in good stead. But my impression has been that they shy away from C/C++ and teach Java. The root of the problem, of course, is the disconnect between academia and the real world. Trust me, the MSc in theoretical physics leaves you neither here nor there. The courses serve merely as samples, as appetisers. Lie theory is an ocean unto itself. Likewise for QFT. Likewise for differential geometry and differential topology. It takes years to master any of these. What you will get is a smorgasbord of courses force-fed down your throat that will leave you reeling by year-end. And no real-world skills to speak of.

Absolutely correct, and my exact experience after studying for a masters in theoretical physics.

If you want to be a physicist, do a PhD.

If you want to work in finance, do a PhD.

One year masters courses are really a waste of time - avoid if at all possible. If you get to spend one year doing a project then that's certainly useful.
 
I guess I wanted to avoid a Physics vs. MFE debate as the obvious winner is an MFE. However, I'm equally interested in doing an MSc in computer science or physics "for fun" I guess.
I'm using future prospects in QF as something to try and decide which I will do "for fun" before thinking about the possibility of becoming a quant and maybe doing an MFE


At present, you have no hope of getting into QF. Gone are the days where you can come out of Uni with an MSc or MSci and stand a chance of getting a reasonably quantitative job. What people in QF look for is skills. What skills do you have? You know basic algebra, and you know something about time-dilation, Brillouin zones and geo-stationary orbits. Only the algebra is useful. What is very useful, on the other hand, is being able to program. So if you want to work in QF, become a programmer first. That does not mean work through some C++ for finance book, like Daniel Duffy's, though that's a start, it means getting to the stage where you're answering people's questions on StackOverflow and starting and contributing to open-source software projects.
 
I guess I wanted to avoid a Physics vs. MFE debate as the obvious winner is an MFE. However, I'm equally interested in doing an MSc in computer science or physics "for fun" I guess.
I'm using future prospects in QF as something to try and decide which I will do "for fun" before thinking about the possibility of becoming a quant and maybe doing an MFE


Your last sentence is incomprehensible. Tip #2, learn how to write. It's a life skill and it's especially useful for writing impressive cover letters to pedantic hiring managers.
 
Back
Top