2009 Quantnet Ranking of Financial Engineering (MFE), Quantitative Finance Programs

[IlyaKEightSix]

Frankly, rankings or not quant roles are a niche position. And furthermore, no matter how good the ranking of your school, the only thing it will do is get you an interview. If you manage to make a good impression on a couple of recruiters you are working with, it does not matter whether or not you have a phenomenal name-brand school such as Berkeley or Stanford or Princeton, or Random City University.

If you know your stuff, you know your stuff, and it'll come out in the interview.

As for Baruch ranking so highly, Rutgers's statistics department is ranked 17th in the nation in graduate studies, and I am positively bored because it's so easy. Rutgers's MSMF is ranked in the top 20, and in the regression class which features a mix of stats and MSMF students, I'm among the top five among eighty. Rankings don't matter so much as alumni networks and connections.

 

[GoIllini]

Just because a few (non-Baruch) members of Quantnet question your rankings of FE programs does not mean you should get all rattled.

It was not long ago that many on this board agreed that ranking MFE programs was useless and subjective. Now all of a sudden you've come up with an algorithm to legitimize Baruch's place in the MFE world so YOUR FE program Ranking is the de-facto standard ?

I'm sure Baruch has an excellent program. What I don't agree with is using Quantnet which is NOW a forum open to the public at large (whereas before it was a Baruch only website) to manipulate perceptions of FE programs in an effort to bolster the prestige of your diploma.

Andy: I don't think the word "Methodology" is big and flashy enough for some readers of this forum.

Two things to note on both points:

1.) Many public school programs are extremely underrated- particularly in the engineering disciplines. Baruch is probably painfully aware of this and the rankings are one way of trying to get folks to focus on the educational strengths and weaknesses of programs rather than brand names. It doesn't mean that they're manipulating the numbers and in fact, I would be rather annoyed if a public school didn't make it into the top ten for an engineering discipline.

2.) I know there is this huge focus on "methodology", but the methodology is actually a little opaque. In particular, you guys used a lot of "proprietary" stuff to do the rankings. It's one thing for US News to do the rankings (although I'm still convinced they have a lot of private-schoolers running the show), but it's completely different for a bunch of guys from one school to be doing everything and keeping important parts of the methodology a secret.

Why doesn't quantnet just make the scoring algorithm open-source? That would make it easier for everyone to independently review and verify the rankings. If people think the scoring process doesn't reflect the schools' strengths, they can raise their specific nits to pick on the forums and the rankings' owners can decide whether these complaints are valid or not and then modify them.

Quantnet already has a significant head-start on other would-be rankers, and the GPL, I believe, provides for both Quantnet's retention of moral rights and right-of-attribution. In other words, if Harvard borrows the algorithm and uses the rankings to put its MBA program as better than all other MFE programs, Quantnet has the moral right to stop them.

An open-source scoring algorithm would quell a lot of doubts- particularly with Baruch's ranking.

As for Baruch ranking so highly, Rutgers's statistics department is ranked 17th in the nation in graduate studies, and I am positively bored because it's so easy. Rutgers's MSMF is ranked in the top 20, and in the regression class which features a mix of stats and MSMF students, I'm among the top five among eighty. Rankings don't matter so much as alumni networks and connections.

You say that, but at the end of the day, it's awfully tough to get hired at an investment bank without a brand name degree if you don't have an uncle. You may get some interviews out of Rutgers, but it will be harder for you to get in even once you've gotten an interview than it would for someone from an Ivy League school. It will also be harder for you to advance.

 

[IlyaKEightSix]

Yes, exactly, with the uncle thing. The more recognized the school, the more "uncles" it has. It's not what you know, but who you know. That's the way it's always been. A better school will provide you with more people to go to in terms of getting you in.

 

[GoIllini]

Yes, exactly, with the uncle thing. The more recognized the school, the more "uncles" it has. It's not what you know, but who you know. That's the way it's always been. A better school will provide you with more people to go to in terms of getting you in.

It's not just an uncle thing.

It's possible to get hired on competence by people who don't know you if you went to the right school. Being from Illinois as a programmer helped me get hired into IT, even though most of the recruiters were from east-coast schools. I definitely had to demonstrate competence in the interviews, but I'm not sure I would have gotten in let alone gotten an interview if I hadn't been from a top five or ten CS school or Ivy League. (Illinois has no pull whatsoever on the business side)

Being from a program known for only letting competent people graduate takes away some of the uncertainty for a potential recruiter. If you ace all the interview questions, it could be that you got lucky or it could be that you're truly competent. If you have a certain school on your resume that they believe produces mostly competent people, they have to deal with less uncertainty than they do with a more difficult program.

Rutgers and Illinois are known for producing competent mathematicians and statisticians in many places, but not always in the investment banking circles. Many people outside of technology aren't even aware that Illinois and Georgia Tech have better engineering programs than nearly every Ivy League.

 

[MNB]

A very impressive list of internship companies has just been released by UCB. The companies include sell-side firms such as:

Goldman Sachs
JPMorgan
Morgan Stanley
Credit Suisse
Deutsche Bank
BoA/Merrill
Citigroup
BNP Paribas
Barclays Capital
UBS

A pretty comprehensive list of the top IBs. Not to mention prop shops / high frequency firms like Sun Trading and Spot Trading, Rating Agencies Moody's and S&P and a myriad of other top institutions. I understand placement is only a very indirect criteria in the Quantnet ranking methodology, if it can be considered so at all. Personally, placement record would be very high on my list of priorities when it comes to choosing a program.

 

[Stefan Zota]

Placement record is a very important factor in choosing any program. After all, the goal is to find a good job at the end.

The problem is that it cannot be quantified. More specifically, how can you rate job positions? A program may find all technology jobs for its graduates and so 100% employment. Another one may not share this data. Another one may throw some company names. Which one is better?
Impossible to say without corroborating information.


PS: Companies you have provided are large corporations with at least 40k employees (CS and GS slightly smaller) that take several hundred interns only in Technology.
It is good to get the internship, but many programs will have a presence in these interns classes. It matters if a full-time offer follows, if student can advance in the job, if the student uses some knowledge from the program, if the student used some contacts from the program etc.

 

[MNB]

Placement record is a very important factor in choosing any program. After all, the goal is to find a good job at the end.

The problem is that it cannot be quantified. More specifically, how can you rate job positions? A program may find all technology jobs for its graduates and so 100% employment. Another one may not share this data. Another one may throw some company names. Which one is better?
Impossible to say without corroborating information.

Well I'm sure imaginative people can find various ways to quantify it - anything can be put into numbers - I'm sure folks on here believe that! Lack of information will always be an issue, but if any ranking becomes highly valued by prospective candidates then programs with success in a particular category will do their best to publicise it so as to improve their standing.

Also a little research can tell you more about job specifics. For instance, I can say for sure that UCB didn't place all of its students in technology positions as the website already provides enough information to the contrary. In addition, "names" appear to be extremely important to potential candidates so "throwing" them would seem a very sensible tactic. Perhaps more programs should do this - I believe pretty much all of them do already (take Baruch, for example, which provides only names and no job titles as far as I can see as well).

Anyhow, my point is that an attempt at quantifying something that is genuinely important to prospective candidates may be a better idea than quantifying something which is only of indirect (and possibly spurious) relevance.

 

[Stefan Zota]

Well I'm sure imaginative people can find various ways to quantify it - anything can be put into numbers - I'm sure folks on here believe that! Lack of information will always be an issue, but if any ranking becomes highly valued by prospective candidates then programs with success in a particular category will do their best to publicise it so as to improve their standing.

Also a little research can tell you more about job specifics. For instance, I can say for sure that UCB didn't place all of its students in technology positions as the website already provides enough information to the contrary. In addition, "names" appear to be extremely important to potential candidates so "throwing" them would seem a very sensible tactic. Perhaps more programs should do this - I believe pretty much all of them do already (take Baruch, for example, which provides only names and no job titles as far as I can see as well).

Anyhow, my point is that an attempt at quantifying something that is genuinely important to prospective candidates may be a better idea than quantifying something which is only of indirect (and possibly spurious) relevance.

Nice words, but I don't see a proposal on quantifying placement record.
You can consider the employment percentage, the starting salary. These are provided only by some programs. Even these, will not paint the picture.

The core question is: how do you quantify a position versus another?

 

[MNB]

Nice words, but I don't see a proposal on quantifying placement record.
You can consider the employment percentage, the starting salary. These are provided only by some programs. Even these, will not paint the picture.

The core question is: how do you quantify a position versus another?

One random example off the top of my head: gather consensus views on preference rankings of job industry, title, location, firm etc. Score each publicised job on a program by its weighted average based on these rankings (e.g. 5 points for a number 1, 4 for a 2 etc). Take the simple average of all publicised jobs for the program, and rank.

There are a million ways to do it. Don't make me do all your legwork!

 

[Aditya Chitral]

One random example off the top of my head: gather consensus views on preference rankings of job industry, title, location, firm etc. Score each publicised job on a program by its weighted average based on these rankings (e.g. 5 points for a number 1, 4 for a 2 etc). Take the simple average of all publicised jobs for the program, and rank.

There are a million ways to do it. Don't make me do all your legwork!

What about student's preference. For e.g. What if a student always wanted to go in say X position whereas it was always rated below Y. To make it worse 50% of students took X job because they wanted to even if they were well qualified for position Y.
So here no one has failed and it was the student's choice to take that particular job. So would you say that just because of this program should be ranked lower.

Again, how do you get that consensus? Whom do you go and ask? Person in job X would always rate it above job Y and vice versa. Are you again talking of ranking jobs?

The sole purpose of using such an algorithm is using factual data and is not subjective.

Also, rankings are just indicators. Ultimately for a prospective student to make a decision, he/she has to do much work than relying on any rankings. There is a lot of information available of web and each might represent things the best way they know and may not necessarily correct from everyone's standpoint.

There are a million ways to do it. Don't make me do all your legwork!

If you want to prove something wrong. Please either provide facts or develop a consensus (as you say) that you are right. Presenting a convincing example would help in achieving the latter goal. May be 1 from that million would do that. Please shed some light on it.

 

[MNB]

What about student's preference. For e.g. What if a student always wanted to go in say X position whereas it was always rated below Y. To make it worse 50% of students took X job because they wanted to even if they were well qualified for position Y.
So here no one has failed and it was the student's choice to take that particular job. So would you say that just because of this program should be ranked lower.

Obviously this has already been taken into account in the ranking of job preferences. If one program places all its students in a position that is the second favourite job out there it will do well. And naturally any model will be an abstraction of reality so will not rank the programs in the order suggested by each prospective candidate (e.g. the individual who wants nothing more than a job as a janitor in an IB cafeteria).

Again, how do you get that consensus? Whom do you go and ask? Person in job X would always rate it above job Y and vice versa. Are you again talking of ranking jobs?

Are you serious? You can't think of anything? How about an internet poll? Write letters to people. Walk around with sandwich boards.... Use your imagination man - you'll need it if you have aspirations of moving into finance. Now would be a good chance to practice!

The sole purpose of using such an algorithm is using factual data and is not subjective.

And I could factually rank programs by the number of letters in their names but it would also be far less helpful than a subjective poll-based ranking.

If you want to prove something wrong. Please either provide facts or develop a consensus (as you say) that you are right. Presenting a convincing example would help in achieving the latter goal. May be 1 from that million would do that. Please shed some light on it.

Ok. How about the Global Derivatives ranking system? I'd wager that outside of this forum that methodology, whatever it was, was very well respected. If I do a poll and discover that people feel that ranking was more sensible than anything that has since been concocted do I win a prize?

I don't mean to get personal here but I have a funny feeling that much of this "criticism" of my suggestion is based by the fact that certain schools did very well from certain methodologies and I don't want to waste my time suggesting things like "do an internet poll" anymore. I am quite sure that every neutral observer understands my point already.

 

[Omar Alkarimy]

Wow that baruch university fees says its 20,700! I'm sure its per semister right?
so in 1.5 years it would make three semesters that would be 62,100 $
am I right??

 

[DominiConnor]

I note that the CQF is not on this list, even though you've reviewed it ?

 

[Andy Nguyen]

Wow that baruch university fees says its 20,700! I'm sure its per semister right?
so in 1.5 years it would make three semesters that would be 62,100 $
am I right??

No. That's total tuition.
By your logic, the CMU program will cost $220,000 in 3 semesters and other programs will cost well over $100,000

 

[mikee]

are there going to be new rankings for 2010? if so, when are they coming out?

 

[Omar Alkarimy]

Ofcourse not that's not what I meant,I thought there might be some kind of error that the Baruch university tuition fees were written per semister,not like other programs.
The difference between the tuition fees of baruch and other programs such as CMU and UCB is just too big.
Why is that?what's the premium that other programs have that makes them much more expensive?

 

[Andy Nguyen]

Baruch is a public university which has in-state/out-state rate. Private universities charge the same for all students.
Tuition is set by different boards, states, school trustees or in some cases how much schools think the market will pay. We just report what is published on their websites.

I note that the CQF is not on this list, even though you've reviewed it ?

The reviews are submitted by the students, not by us. We only rank master programs.

are there going to be new rankings for 2010? if so, when are they coming out?

We're always working to better our algorithm. A new ranking will be announced when it's ready.

 

[euroazn]

Another ranking - The Top 10 Quant Schools, According to the Street by Advanced Trading

 

[David Rappaport]

This is from 2008?

 

[euroazn]

Yes, but since quantitative finance rankings are so rare, I decided to post this.