• 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!

Avoid this U of Toronto MMF program: SCAM

What Carl said in the thread is consistent with what I know about this MMF program at UT.

I did my undergrad in the UT for stats, and also got the offer from MMF last year, and turned it down in the end for many reasons.

Firstly, if you guys really go through the selection process, you surely find it very random & cheap and not quantitative at all. In contrast with its competitor, the selection standard of MQF program at University of Waterloo is extremely high with very hard written test (it makes sense as it has the largest MATH Faculty in the world) and tech-based interviews.

Secondly, MMF program is essentially NOT part of UT, and the CLASSROOM is not on Campus at all. Most instructors are NOT from academia so what they teach is just like baby questions and not quantitative (yes, Jaimungal is only one from Stats department). Can you guys imagine finishing Stochastic Calculus in only 13 lectures? 13 lectures cannot even cover chapter 4,5,6 in Shreve's book (vol 2).

Thirdly, full-time placement is not real. I have a few friends in MMF and some of them were still seeking a position after 4 months of graduation. Admittedly, the internship placement for MMF is good. However, this is a result of forcing students to take the first-come offer. You don't have any choice here.
 
Hello Carl, you are acting as a sort of journalist in this thread, and are presenting some very impactful information to applicants of the program (whose deadline is in a few days). I appreciate that you are helping others avoid a potential mistake.

Since you are acting as a journalist could you provide some sort of evidence that you have attended the program and had these experiences? I noticed that you joined this website yesterday so adding some support to your claims would clarify your motives. Maybe a picture of a student ID with personal information blacked out or something. This is the kind of thing a real reporter would do to support his/her claims.
You right I will post scan of admit letter and tcard, need to stay anonymous though.
Edit: i also post pics of classroom at 720, show all here that I in program
 
Secondly, MMF program is essentially NOT part of UT, and the CLASSROOM is not on Campus at all. Most instructors are NOT from academia so what they teach is just like baby questions and not quantitative (yes, Jaimungal is only one from Stats department). Can you guys imagine finishing Stochastic Calculus in only 13 lectures? 13 lectures cannot even cover chapter 4,5,6 in Shreve's book (vol 2).

Thirdly, full-time placement is not real. I have a few friends in MMF and some of them were still seeking a position after 4 months of graduation. Admittedly, the internship placement for MMF is good. However, this is a result of forcing students to take the first-come offer. You don't have any choice here.

Another bum program exposed.
 
lol and its ranked 12th in the rankings above Chicago. if what the op and the other guy said is true, some of the universities ie UW need to move way up in the rankings.
 
lol and its ranked 12th in the rankings above Chicago. if what the op and the other guy said is true, some of the universities ie UW need to move way up in the rankings.
The rankings are based on data submitted by the programs. That is not going to change until we have a meaningful number of student reviews on QuantNet.
That's why the OP can best help the situation if he and his classmates write a review for the U Toronto MMF program.
Reviews are quick, easy to do and can be anonymous. Just click on the stars to rate them from 1-5 stars and write a review. The more detailed, the better.
 
lol and its ranked 12th in the rankings above Chicago. if what the op and the other guy said is true, some of the universities ie UW need to move way up in the rankings.

The rankings are based on the "information" the programs supply. Since telling the truth often runs counter to the interests of those behind the programs (i.e., making money), they don't. I'd not be surprised if even some of the student reviews are not authentic. Information on placement rates -- when they're provided, that is -- are about as reliable as those given by law schools.

I'm not surprised at the hasty and shoddy manner in which the UoT is covering stochastic calculus. Other programs are doing the same. Thirteen lectures isn't enough -- it's a joke. It means at the very least that you're skipping foundations, skipping all sorts of crucial technical details.
 
I'm not surprised at the hasty and shoddy manner in which [...] is covering stochastic calculus. Other programs are doing the same. Thirteen lectures isn't enough -- it's a joke. It means at the very least that you're skipping foundations, skipping all sorts of crucial technical details.

Indeed.

A lot is just juggling with symbols.

And I suppose they don't reach the finishing line, i.e. Kloeden/Platen numerics.
 
I might be starting a huge argument here, but I'm going to suggest that stochal is probably less important today than it was ten years ago. I really don't know how many people apply the Girsanov Theorem on a daily basis anymore.

If you want to cover all of the program's stochal in 15 weeks, I don't think it's a travesty. Just as long as you're spending more time covering GARCH, copulas, regressions, stats theory, behavioral finance, state pricing, statistical modelling techniques, machine learning, etc.

Ten years ago, you could have graduated from an MFE program without learning the differences between quantile regression and least-squares regression; the issues with overfit on one; the lack of confidence intervals from the other. But you learned stochal and learned how to price exotic options and mortgage backed securities. Today, most of the pricing for that stuff is built out.

Most of the people still working on that stuff are just doing routine maintenance on the models. You still need some knowledge of stochal to know what you're doing, but now you're more of a car mechanic than a mechanical engineer at GM. Maybe you do not need to take the Fundamentals of Engineering exam to change the spark plugs on a Chevrolet Camaro. Maybe you just need to know what torque is (to torque the spark plugs to spec) and what's required for electricity to flow through a circuit.
 
I might be starting a huge argument here, but I'm going to suggest that stochal is probably less important today than it was ten years ago. I really don't know how many people apply the Girsanov Theorem on a daily basis anymore.

I agree.

If you want to cover all of the program's stochal in 15 weeks, I don't think it's a travesty. Just as long as you're spending more time covering GARCH, copulas, regressions, stats theory, behavioral finance, state pricing, statistical modelling techniques, machine learning, etc.


Ah, now there's the rub. All these things are being force-fed at an undigestible rate. They're being covered superficially. Too much too fast, and usually too superficially to be used with dexterity. Statistical theory by itself can be -- and is -- a master's degree in itself. Same for stochastic theory. What is the use of giving a brief and superficial taste of a large number of subjects, which probably really have no inherent connection with one another except they're part of the grab bag of tools used by pros in finance?
 
I agree.




Ah, now there's the rub. All these things are being force-fed at an undigestible rate. They're being covered superficially. Too much too fast, and usually too superficially to be used with dexterity. Statistical theory by itself can be -- and is -- a master's degree in itself. Same for stochastic theory. What is the use of giving a brief and superficial taste of a large number of subjects, which probably really have no inherent connection with one another except they're part of the grab bag of tools used by pros in finance?

I also agree on stochastic calculus. It's not as important anymore. No one is looking for pricing quants to do stoch calc all day. No one wants too much creativity in that area either.

It seems like this program is just a cash cow for UofT. Then again, what quant program isn't? Maybe Princeton's, which does a relatively good job of preserving the prestige of its degree. But I didn't feel all that special with regards to the CMU program, or the CU program, which has ballooned its class size by a factor of about 4 in the last 10 years. I think many programs like these run optimization models to see how many students they can take tuition from, and still maintain their desired placement rates, and rankings. It is still worth it for these programs though, but certainly not UofT, which seems like it just slapped together a program for the cheapest cost, and the biggest bang for their buck.
 
True. The courses do run a little too fast.

I will also say that they are probably 1.5-2x the work of the courses I remember from undergrad. In undergrad, everyone would complain about CS 473 (Algorithms and Theory), and how it was impossible. We had seven homeworks, were allowed to work in groups of three (to speed up grading), and each homework took about 2/3 the time (working as a group) as many of my courses, which often have 10 homework assignments that must be done individually.

Maybe the 19-year-old version of me would tell me to stop being a grumpy old man, but I feel like a two year MFin/MFE is the equivalent of three, maybe three and a half years of academic work in a CS undergrad. I'd like to think there's room for a stats minor in there somewhere.
 
Last edited:
I will also say that they are probably 1.5-2x the work of the courses I remember from undergrad.

No problem with that: it's the same in math, the same in physics. The difference is that while, say, in grad school in math or physics, the students have seen the material before in more elementary fashion. The grad algebra course (or the grad mechanics course) assumes you've already taken an undergrad course in algebra (or mechanics) -- you have some foundation, some foretaste, some idea of thought patterns. Plus the instructors are not trying to shoehorn the whole of Lang (algebra) or Goldstein (mechanics) into thirteen lectures. Can't be done. If it is, and you somehow pass the travesty of an exam, you still won't be able to say with a straight face that you "know" the subject. Yet this is precisely what's being done in these quant finance programs. Okay, I'm being unfair -- there is a lot to cover and really not enough time. Yet I think that in this case "less is more."
 
^^^ Sure. That's why an MFE has been described to me from multiple independent sources as "a ghetto substitute for a PhD to get you a job as a quant".

My back of the envelope guess is that these days, there are about 500 jobs on wall street that require an expert knowledge of stochal. Meanwhile, I think there are probably at least 5000 jobs on Wall Street where people are expected to know a little GARCH, a little stats, a little stochal, and a lot of programming.

You can spend five years in school getting an ORFE degree and becoming an expert on Stochal and graduate to earn $200K your first year. Or you can spend one or two years in school to learn a little bit of everything and graduate to earn $150K your first year and $200K your second year, giving you another year or two before the PhD graduates.
 
Last edited:
Strange ...

IMO
99% of the literature on the numerical solution of SDEs is the erratic and weak Euler method even though the engineering fraternity have been using more robust methods for many years.

And PDE methods need an upgrade...
 
In contrast with its competitor, the selection standard of MQF program at University of Waterloo is extremely high with very hard written test (it makes sense as it has the largest MATH Faculty in the world) and tech-based interviews.

Not meaning to nit-pick much here, but since false or misleading claims seems to be part of the thread here...

I found the largest math faculty claim rather weird, since from all I've heard (including from people from there) Waterloo is a small-to-medium university.

Now that I've scanned through their webpages, yeah ok: if you do what most well-known universities would NOT do, which is combine pure math, applied math, computer science, statistics, and whatever else (including "optimization"), into one gigantic department and call it a "math department", sure, you will probably have the biggest. I note however that the pure math department has 23 faculty, applied math has 27, and computer science over 100.
 
Strange ...

IMO
99% of the literature on the numerical solution of SDEs is the erratic and weak Euler method even though the engineering fraternity have been using more robust methods for many years.
You can do a little bit better with Milstein, and this one is used for plenty of models in Finance iirc
 
You can do a little bit better with Milstein, and this one is used for plenty of models in Finance iirc
Milstein is not all that great. It does not scale well to more complex SDEs.

What I am saying is that there are more robust methods than the 'usual suspects'.
 
Last edited:
Of the 5 points you posted in the OP, point 1 is actually a very good thing. I don't know many students (if any) actually got expelled from their MFE programs. So at least, Toronto MSMF is doing it right in that regard.

The only issue is that it is reasonable if all programs do that. But if, e.g. It is done by 3 programs out of 25, then people will start applying to other 22 more and will stop applying to these 3.
 
The only issue is that it is reasonable if all programs do that. But if, e.g. It is done by 3 programs out of 25, then people will start applying to other 22 more and will stop applying to these 3.
A variant of this scenario is 3 do _very_hard_and_vital_module_ and 22 don't.
 
Back
Top