Advice needed. Thanks.

[BrianChan]

I am considering to pursue a master degree in Finance or Quantitative Finance (or Financial Engineering).

My Background:
- I graduated from BBA degree in Finance, and currently work as auditor in one of the Big4.
- I don't have strong maths background, just have basic knowledge in matrix, calculus, linear algebra.
- I have knowledge in VBA and R programming

My Career Goal:
- I am finding trading or research job (I want to study MFE because I want to gain some technical knowledge)
- I may find some equity research, PE fund or risk management after I get my CPA qualification
- I want to work with derivatives and technology.

My question is that if I studied a MFE program, could I still find jobs in finance industry other than quant?

It seems that MFE is more suitable for people with strong maths/programming background.
Should I study MFE or MSc in Finance to have a broader choice of career?

Please advise. Thanks a lot!

 

[Ken Abbott]

I recently asked a bunch of MFEs whether they ever used their knowledge of PDEs or stochastic processes in their daily work. No one said they did.

I believe that core math, stat, and finance knowledge (eg bond pricing) are more essential that a lot of the material currently being offered at these programs.

 

[Daniel Duffy]

I recently asked a bunch of MFEs whether they ever used their knowledge of PDEs or stochastic processes in their daily work. No one said they did.

I believe that core math, stat, and finance knowledge (eg bond pricing) are more essential that a lot of the material currently being offered at these programs.

AFAIK PDE methods form the core of many C++ quant libraries. Some models can only be done in this way.

PDE demands a high mathematical level, e.g. physics PhD. MFE education can never hope to do SDE or PDE in so much detail so that the graduates hit the ground running when they get out. So they use other methods I suppose.

It is a kind of specialised niche indeed.

 

[bigbadwolf]

PDE demands a high mathematical level, e.g. physics PhD. MFE education can never hope to do SDE or PDE in so much detail so that the graduates hit the ground running when they get out. So they use other methods I suppose.

Don't know about high level -- maybe if you're looking at Sobolev spaces and techniques from Hilbert space theory. But it does need time. I think typical MFE programs have one course on the Black-Scholes PDE, with a bare minimum of numerical methods to solve it. The same is true for stochastic analysis as well -- some topics presented hurriedly and out of context. Presented and received in a spirit of haste and cynicism.

 

[Daniel Duffy]

Don't know about high level -- maybe if you're looking at Sobolev spaces and techniques from Hilbert space theory. But it does need time. I think typical MFE programs have one course on the Black-Scholes PDE, with a bare minimum of numerical methods to solve it. The same is true for stochastic analysis as well -- some topics presented hurriedly and out of context. Presented and received in a spirit of haste and cynicism.

There are two issues (at least) IMO

1. A good understanding of the PDE in mathematical physics and their applications
2. Numerical applications of PDE

These are non-trivial pursuits if students want to get beyond the Euler method for the 1-factor heat equation. What is popular is the Crank-Nicolson method even though it fell out of favour in engineering some decades ago (see e.g. Strang/Fix book on FEM).

http://www.wilmott.com/pdfs/071203_duffy.pdf

 

[bigbadwolf]

There are two issues (at least) IMO

1. A good understanding of the PDE in mathematical physics and their applications
2. Numerical applications of PDE

These are non-trivial pursuits if students want to get beyond the Euler method for the 1-factor heat equation. What is popular is the Crank-Nicolson method even though it fell out of favour in engineering some decades ago (see e.g. Strang/Fix book on FEM)

I agree. 'Course this takes time. The slender exposure to the B-S PDE and the Crank-Nicolson method (and maybe other finite difference methods) is neither here nor there.

 

[BrianChan]

Thanks for the advice.

Would you think that a person without solid maths training (like me) can tackle the maths courses in MFE?
(with hardworking of course)

Also, what kind of characteristics are required for a quant/trader/MFE student?

Thanks again.

 

[ygmwy]

Why do you want to study MFE if you don't want to be a quant? Judging only based on what you said, an MFE is not for you. You might want to get an MBA. There are plenty of technical course offerings in an MBA program. An even better idea is to go for an MFin program like MIT's, where the student gets to choose how quantitative his or her coursework can be. There are also many specialized programs out there that do exactly what you want, but an MFE is definitely not going to give you what you are looking for.

Will hardwork get you through the coursework? I am inclined to say no, and the reason is that you are talking about one-year masters programs where a lot of intense math is packed into a few semesters of study. Students are widely expected to already have the basics down, and if you are stuck going over the basics of probability theory, you will never understand, nor pass the introductory quant classes. Even if you go to a super easy program and get straight A's, you will be asked difficult math questions in your quant interviews, and use math everyday in your work. I've been asked to solve stochastic differential equations over the phone! You're probably used to personality driven interviews, and more subjective evaluations. A quant trader? This doesn't sound like you.

That being said, you won't even get an interview for programs like Carnegie Mellon, Columbia, or NYU if you don't demonstrate in your application that you can handle multiple rough grad level math courses in rapid succession.

 

[BrianChan]

Why do you want to study MFE if you don't want to be a quant? Judging only based on what you said, an MFE is not for you. You might want to get an MBA. There are plenty of technical course offerings in an MBA program. An even better idea is to go for an MFin program like MIT's, where the student gets to choose how quantitative his or her coursework can be. There are also many specialized programs out there that do exactly what you want, but an MFE is definitely not going to give you what you are looking for.

Will hardwork get you through the coursework? I am inclined to say no, and the reason is that you are talking about one-year masters programs where a lot of intense math is packed into a few semesters of study. Students are widely expected to already have the basics down, and if you are stuck going over the basics of probability theory, you will never understand, nor pass the introductory quant classes. Even if you go to a super easy program and get straight A's, you will be asked difficult math questions in your quant interviews, and use math everyday in your work. I've been asked to solve stochastic differential equations over the phone! You're probably used to personality driven interviews, and more subjective evaluations. A quant trader? This doesn't sound like you.

That being said, you won't even get an interview for programs like Carnegie Mellon, Columbia, or NYU if you don't demonstrate in your application that you can handle multiple rough grad level math courses in rapid succession.

You have mentioned a very good point. I can't expect to improve a lot after a one-year program.

It makes me back to MFin or MSF. I want to study something more technical, instead of a master degree with lots of general knowledge.

MIT's one is great.
I will do some research on other MFin programs.. (I think Finance, Financial Economic, or Economic are similar or even the same, right?)

And I have to consider whether studying in UK/US or staying in Hong Kong.
(I come from Hong Kong =P)

 

[ygmwy]

The average student learns a lot in one year, just because they push you and teach you so much that you hardly have time to breath. But when it's all done and over with, you really have all the tools you need to go for a quant interview. That being said--you won't achieve this if you struggle with undergrad math classes. Especially probability theory, linear algebra, calculus, and computer science. If you don't have a strong background in this, then you will learn nothing and fail out of the fast-paced program anyway. Make sure you find the program that is right for you!

Good luck! My family is from HK too btw.