MFE Alternatives - For Quant Research

[Zaurald]

Hi everyone,

I am currently working as a Data Scientist at a top bank (mostly on the Operations side doing NLP). My inclination lies towards quant research (mostly utilizing ML/DL for investment research). Although I am sort of convinced about taking an MFE course, I want to reassess my decision. The major trigger is probably the curriculum which involves 1-2 ML/DL courses, and the rest of the curriculum is focused on math/finance. Also, as pointed out by a lot of folks here, and MFE won't lead directly to investment research. It involves a heavy coding background, as well as top-notch math skills to enter the quant research/trading jobs.

I am looking at the following options:
1. Operations Research - This looks like an ideal program involving a lot of math and computation. However, many have pointed out that this is a "mile wide but not a mile deep". I'm not completely sure how this could be a problem if one streamlines the curriculum according to the research interests.
2. Applied Mathematics - Too theoritical (less programming component), and I'm not sure if I'll even get in because my undergrad major is Telecommunications Engineering
3. MFE

Alternatively, if someone can provide some guidance on the basis of industry status-quo it'll be of great help. Thanks!

 

[Qui-Gon]

"Applied Mathematics - Too theoretical" ...

 

[noether-skolem]

I'm not entirely sure what your problem with MFE is, given that you think that AM is too theoretical and OR's focus is too wide.

 

[Zaurald]

I'm not entirely sure what your problem with MFE is, given that you think that AM is too theoretical and OR's focus is too wide.

I don't have a problem with MFE, I'm just unsure if I'll be the right fit. I don't wanna loose out on my ds skills, and going by the curriculum there are only a couple courses that involve that.

I really like the OR curriculum, but would it lead to quant research kind of jobs?

 

[noether-skolem]

I don't have a problem with MFE, I'm just unsure if I'll be the right fit. I don't wanna loose out on my ds skills, and going by the curriculum there are only a couple courses that involve that.

I really like the OR curriculum, but would it lead to quant research kind of jobs?

What makes you think that OR would be a better fit for you than FE if your goal is quant finance? FE programs are often part of the OR department which are tailored for finance jobs, more often than not you'll be able to take OR courses as a FE student, vice versa.

Why do you think you might not fit in an MFE program? I'm really confused here, what do you think OR has that FE doesn't have, which will help you get into quant research?

Maybe give a more detailed breakdown of what you like about OR, what you dislike about FE, what skills/experience/background you have and what you think are missing in order to get into quant research?

 

[Zaurald]

What makes you think that OR would be a better fit for you than FE if your goal is quant finance? FE programs are often part of the OR department which are tailored for finance jobs, more often than not you'll be able to take OR courses as a FE student, vice versa.

Why do you think you might not fit in an MFE program? I'm really confused here, what do you think OR has that FE doesn't have, which will help you get into quant research?

Maybe give a more detailed breakdown of what you like about OR, what you dislike about FE, what skills/experience/background you have and what you think are missing in order to get into quant research?

I think you're getting too defensive about the MFE. I never said I dislike MFE, or I'm against it. I was/am an MFE aspirant. However, I'm a little confused as to whether I should go for an MFE or would I be a right fit for other programs.

Let me provide a bit of my background. I'm currently working with the Operations and Risk Control division of a US bank. My work involves identifying potential risks and developing frameworks to mitigate risks and the associated impact. In order to achieve these objectives, I utilize a variety of techniques such as clustering, NLP, regression, etc. Prior to this job role, I was working as a Data Scientist for an airliner, with the divisions such as Operations & Strategy, Revenue Management, Corporate Strategy, etc. My job here was pretty similar but the problems were more complex and I had to do a lot of EDA, hypothesis testing, optimization, etc. I have been involved in problem spaces that are very similar to what an Operations Research program would offer (needless to say that I thoroughly enjoy working on these problems). On the contrary, when I look at the MFE curriculum, I don't find a lot of emphasis on ML or Big Data. I believe this is the reason why I think I may not be the right fit.

I am interested in finance (probably more on the FinTech side) which primarily employs ML/big data to make decisions, which is why quant research intrigues me. I know that programs are coming with a focus on data science but the existing curriculum has very few of that element (feel free to correct me if I'm wrong). OR has a good balance on math and ds which might be more suitable for me. Also going by the recent trend of how fintech/investment research firms hire, it is advisable to have a research-oriented mindset and the ability to experiment with novel techniques. So I'm contemplating if an MFE would actually help me develop that.

If you're a current/former student, I would like to know your experience with the program. Let me know your opinion on the points I made.

 

[noether-skolem]

Not trying to be defensive, sorry if I came across as such. I was just confused by your reasoning, most ML-related courses will be available to both FE and OR students and you seem to consider FE's focus on maths/finance as something you don't like, yet you agree that "It involves a heavy coding background, as well as top-notch math skills to enter the quant research/trading jobs".

Usually FE is a part of the OR dept focused on quant finance and courses are available for both programs so there's a huge overlap between what you can take as an FE student and as an OR student. There will be classes exclusive to certain programs so you should look closely into each program (look if university ABC's FE and OR programs offer what you want to learn, rather than looking at it broadly as FE programs vs OR programs, they're often similar enough).

In general ML/DL won't be the focus of an MFE program but it should be one of the available specialization tracks (at least at Columbia we have intro ML, some data mining/analytics courses, RL, DL, and probably others I can't remember - most of them are taken by FE, OR and other students in the dept). The curriculum still has to cover enough of the maths as a baseline (intro math finance, stats, time series, optimization, stochastic, Monte Carlo) because these are things you'd be expected to know if you go into quant research. If you've already covered these and think it'd be a waste of time going through them again then yeah FE/OR may not be for you. You can probably look into CS or data science programs to strengthen your ML/programming background in that case.

 

[Zaurald]

Makes sense. I guess I need to research a bit more on both the programs and probably assess on what skills I'm currently lacking that can be fulfilled by either of these programs.

Another question - Would I need to complete certifications like CFA/FRM/CQF to get into an MFE program? I see most of the applicants have these credentials, or else they are already working in a quant role.

 

[Andy Zhang]

Not trying to be defensive, sorry if I came across as such. I was just confused by your reasoning, most ML-related courses will be available to both FE and OR students and you seem to consider FE's focus on maths/finance as something you don't like, yet you agree that "It involves a heavy coding background, as well as top-notch math skills to enter the quant research/trading jobs".

Usually FE is a part of the OR dept focused on quant finance and courses are available for both programs so there's a huge overlap between what you can take as an FE student and as an OR student. There will be classes exclusive to certain programs so you should look closely into each program (look if university ABC's FE and OR programs offer what you want to learn, rather than looking at it broadly as FE programs vs OR programs, they're often similar enough).

In general ML/DL won't be the focus of an MFE program but it should be one of the available specialization tracks (at least at Columbia we have intro ML, some data mining/analytics courses, RL, DL, and probably others I can't remember - most of them are taken by FE, OR and other students in the dept). The curriculum still has to cover enough of the maths as a baseline (intro math finance, stats, time series, optimization, stochastic, Monte Carlo) because these are things you'd be expected to know if you go into quant research. If you've already covered these and think it'd be a waste of time going through them again then yeah FE/OR may not be for you. You can probably look into CS or data science programs to strengthen your ML/programming background in that case.

Just to clarify, lots of the baseline maths that @noether-skolem mentioned are required for the FE program and not the OR program. In the OR program (at Columbia), the core courses are prob and stats, optimization, stochastics, and (Monte Carlo) simulation. Of these, only prob and stats can be waived, but you must pass a waiver exam, which takes place during orientation, to do so. For optimization and stochastics, you can also take waivers for those, but then you'll have to do their PhD equivalents--which are much, much harder. Since there's no times series course requirement for OR, I've had friends in the OR program who took a time series course from the stats department instead.

@Zaurald Also, from my interviews, I've found that the baseline maths are actually quite useful. You will be tested on some of those (check the green book by Xinfeng Zhou or even the pocket book by Dr. Dan Stefanica and his colleagues and you'll see what I mean).

Makes sense. I guess I need to research a bit more on both the programs and probably assess on what skills I'm currently lacking that can be fulfilled by either of these programs.

Another question - Would I need to complete certifications like CFA/FRM/CQF to get into an MFE program? I see most of the applicants have these credentials, or else they are already working in a quant role.

Hope I'm not intruding here. No, you don't need them. CFA is generally done by people who either don't have a finance background or want to do something in investment banking. I've also seen FRM/CQF, but they are not a necessity to get into any top FE program. I think you might see these because lots of applicants don't come from a finance background.

 

[noether-skolem]

Just to clarify, lots of the baseline maths that @noether-skolem mentioned are required for the FE program and not the OR program. In the OR program (at Columbia), the core courses are prob and stats, optimization, stochastics, and (Monte Carlo) simulation.

The requirements are not exactly the same but still quite similar it seems (with overlap on stats, optimization, stochastic, Monte Carlo).

Hope I'm not intruding here. No, you don't need them. CFA is generally done by people who either don't have a finance background or want to do something in investment banking. I've also seen FRM/CQF, but they are not a necessity to get into any top FE program. I think you might see these because lots of applicants don't come from a finance background.

Thanks for answering, I completely agree with this. Also worth to note many of those who have/had CFA took the exams under employer sponsorship. If you're working at a financial firm and can get them to pay for the exams, I'd say go for it. It still won't be very relevant for quant roles but it wouldn't hurt you.

 

[Zaurald]

Thanks for your detailed replies @Andy Zhang @noether-skolem. Appreciate it!

Will definitely look forward to your suggestions.

 

[LeviDyett]

Research is always interesting, it is developmental and it unlocks your potential. I understand people who do not succeed in college or university assignments on their own due to a number of reasons that I, as a student, also have, so I often recommend to all colleagues and acquaintances use Custom Research Papers From Professional Writers - Paperell.com where at any time you can make an order for custom research papers, and the experts will do everything for you quickly and efficiently, try it.

 

[binomial-torrent]

just get a MS in computer science/machine learning