Enough Quant Jobs for MFE grads????

Joined
8/2/11
Messages
11
Points
11
I had a random thought of all the mfe graduates every year from all the universities that have applicable
Programs. Are there really sufficient Quant Job opportunities for all new graduates or do you have to at the top of your class and insane amount of related employment history to get a decent offer? What do graduates that graduated at the bottom of their class do ???? Im wondering if all of you in this forum think of stuff like this
 
I guess the top 20 programs graduate approximate 20 * 50 = 1,000 MFE grads every year? Are there 1,000 new quant jobs created every year for them?
 
You can say that on every job in the world.
How many lawyers? Engineers? Accountants? etc does the economy needs?
 
purplyboy : There are easily >1,000 new entry level quant jobs per year, but of course MFEs must compete with other types of qualification.
No firm "creates jobs for MFEs", they have work to do and people apply for them, it's a labour market and as such is not a deterministic process.
 
I suspect that everyone here has varied definition on what a "quant job" is and a good majority would narrow it to "creating new quantitative finance model, quantitative trading".
I'm afraid the intersection of that belief and the reality of the kind of jobs where MFE can fit is rather limiting. So in essence, most people are looking for positions that are not appropriate or non-existent.
Most jobs do not require that stochastic calculus or knowledge of complex financial products. And you do not need to study in MFE programs to learn that.
If you graduate from a practical MFE program, you should process a good toolbag of programming, hand-on skills with financial data and enough finance knowledge that enable you to be competitive for most entry level quantitative jobs out there.

If all you learn from your program is a mixbag of Hull, Shreve and how to price bonds and other textbook products, then good luck.

Some time ago, I met an alum of one of the quantitative finance programs who told me that when he encouraged his classmates to go on Quantnet and better prepare for the job market, their response was that they hate Quantnet because the site over-emphasizes on learning the useful technical skills over stochastic calculus. It's no surprise that their program does not tech them any real programming and they all believe they would not need to get their hands dirty with these skills.

I wonder what kind of jobs (if at all) these guys end up finding.
 
I may be one of the people guilty of emphasising "technical" skills over stoch calc, that's because I'm assuming that anyone here who cares about my advice has studied things that he's been told to study and that he's done his best and got there.

So I focus on things that give you an edge or are common defects that stop basically smart people getting the job they should get.

It so happens that programming skills are one of the most common failure modes for MFE grads, so I emphasise this. I also whine about spoken English and push the idea that a superior quant wannabe is someone who can understands the products, can turn that into maths, then into code and at all times has in his mind a view of where the money comes from. C0urses tend to do the first two well, but programming is a poor relation and depressingly few entry level people have any clue about where the money to pay them comes from; a bit like a kid who believes that meat comes from supermarkets, which is true, but not the the whole picture.
 
Here's what I found interesting. Many students (and people in general) focus on getting the right degree / academic work / credential, and feels entitled to certain salary/job title if they achieved 4.0 at a target school or CFA. The truth is that getting a "job" is a collaboration between the employee and employer to create synergy that is impossible to do individually. So the big question isn't what knowledge/degree you have, but rather how you can help your boss make money. Rather than thinking about whether there are enough "quant jobs" for MFE grads, think about whether a MFE education is necessary for whatever position you're seeking.

The overemphasis on technical skills may not be entirely just, since not all MFE graduates want to do programming heavy jobs. But given the current trend in replacing most finance positions with computerization, MFE graduates are the closest ones with skills aligned with the requirements for so-called "quant jobs" and therefore temporarily have an edge in that area of employment. (and for headhunters who recruit primarily for these position, naturally they will have an expectation for programming skills from those who claim to be FEs.) If you closely examine the placement record for the programming-heavy MFE schools vs non-programming-heavy ones, you will see that many graduates from the former have more success in entering FO trading roles (fast/efficient implementation) and those from the latter tend to go into non-quant-trading/asset management with slower pace. (+ a bunch of Master in Finance people)

The financial industry is going through some fundamental changes, and the society should have consistent demand for people who are knowledgeable about the financial products (math/legal/economics part) and the implementation of the pricing/trading systems (programming part) for the next few years. But it is not MFE specific, and MFE students should do their research on the transferability of the courses they receive through these MFE programs. So to answer your question, "NO, there won't be enough quant jobs for MFE grads, but YES, there will be plenty of quant jobs for the BEST MFE grads."
 
Looking at quant job ads. I see that many are looking for PHd's in math, computer science, science majors with no finance experience. A lot of the requirements seem to be looking for C++, C#, R, Python, VBA, or matlab experience. How long would it take someone to learn all these programming languages? How come there are no core requirement classes in programming for MFE programs? Do a lot of programs have electives in C++, matlab or etc?
 
purplyboy: I've been looking and I've found the same kind of problem. I can't apply for a lot of the job adverts because they either ask for a PhD or they ask for years upon years of experience. I have a MFE. Well you could audit some comp sci classes for C++ and Python while doing your MFE. I didn't do this bc I didn't know how important C++ was to the world of quantitative finance.

But it's ok, you can learn C++ on your own. You could just get a textbook (and the solution manual, if there is one). I learned C++ on my own, did all the exercises and now even my comp sci friends have approved of my level. I read a 1200 page textbook in about 3 weeks. I then took 2 more weeks to do all the exercises. I knew Matlab before that, and nothing else. You could do it, you just have to put in some elbow grease :)
 
But it's ok, you can learn C++ on your own. You could just get a textbook (and the solution manual, if there is one). I learned C++ on my own, did all the exercises and now even my comp sci friends have approved of my level. I read a 1200 page textbook in about 3 weeks. I then took 2 more weeks to do all the exercises. I knew Matlab before that, and nothing else. You could do it, you just have to put in some elbow grease :)
Sadly, the only way to learn programming and put your head down and plow through all the coding exercise in the book. If you are not disciplined, it will be very easy to give up when you hit a roadblock when learning C++.
It's a test of will and many will fail.
 
Andy Nguyen
He is totally right. But I was properly motivated, I had written on my CV "intermediate level: C++" and I got a call from HR at Axa Insurance that they wanted to present me with a role coding various risk scenarios in C++. I was like "ok, yeah that's fine" but in my mind I had NO IDEA what an object was!!! Luckily my brain cells are not always on strike and I had the smarts to at least schedule the interview to be in 2 weeks time, citing that I was the guest speaker in a series of mathematical finance talks. (Ahaha...my creativity isn't on strike either, it seems). During those 2 weeks I studied like hell. I was on vacation in Italy with my friends but I just stayed in and studied while they ate gelato and pasta. That's how I read like 80% of the textbook in 2 weeks.

Turns out that interview was bogus was probably just the HR guy trying to kill time to show the company that he was indeed doing something during December. But the moral is that thank god this douchebag motivated me by setting a date by which I had to show my C++ goods. If you were properly motivated somehow, you could do that too.
 
A lot of the requirements seem to be looking for C++, C#, R, Python, VBA, or matlab experience. How long would it take someone to learn all these programming languages?

For languages like C# and C++, several months if not a year or two. It's not just learning the basics of the syntax -- though for a newcomer even this can be daunting. It's enough mastery to be able to code fluently without needing to open a book all the time. It's the difference between someone who memorises some difficult words for the GRE verbal and someone like myself who can use abstruse words with dexterity. Other than mastery of the syntax, one also has to pick up programming idiom -- each language has its own -- and this takes time. The basics of unit testing and refactoring. The basics of design patterns. Also the framework classes that come with .NET or whatever other framework you are using (such as Qt, for instance).
 

Yes, and they make an important point that its "not attained automatically as a function of extended experience":
"The key is deliberative practice: not just doing it again and again, but challenging yourself with a task that is just beyond your current ability, ..."

I work with people who have more than 10 years experience, and they are still novice C++ programers.

They needed a container class (for trees), and ended up throwing mine out because "templates are too complicated" (it was no more difficult to use than any of the STL containers). A decision was made to throw out two weeks of solid development on my part, because my style was "too complicated for the organization". Apparently object oriented programming (I made effective but completely standard use of the decorator pattern) is too complicated for them.
 
The goal would be to create algorithmic trading strategies to simulate. The ultimate goal is to get into a MFE/MSFM so I was wondering if Python will suffice for the requisite "programming knowledge" programs want. If they only desire C++/C# or others then I won't bother learning Python and will switch to one of those.
 
Back
Top Bottom