Reason for diff in quantnet ranking(2009) to that of global derivatives (2003)

[Srinivas]

Hi everyone.,
This is Srinivas from India. I m planning for financial engineering or mathematical finance for fall 2010. I've seen ranking of financial engineering in Quantnet and Global derivatives. Can anyone clarify why is there difference in the ranking? How did Baruch jump from 19th position to 6th in 5 years and how come UC-Berkeley came from top to 10th. I dont mean otherwise but Whose RANKING is to trust ???

 

[DanM]

I wouldn't trust anything Global Derivatives says. They claim York University (in Toronto) has a MSc in Financial Engineering, when in reality it's a graduate diploma in Financial Engineering that can be taken concurrently with a Master's in Math, or an MBA. And it was ranked 14th in 2004.

 

[Eugene Krel]

Considering FE programs are relatively young a lot could change in 6 years. The rankings are there to guide you not to decide for you, be it Quantnet or GD.

 

[Andy Nguyen]

Just a few point of corrections and clarification

1) The programs are ranked in groups of 5 alphabetically so it should not be read as Baruch is ranked 6th and UCB 10th respectively. This is the most common mistake people make when they quickly glance at our ranking without understand the methodology. Why we decided not to use numerical ranking is explained in our methodology.

2) The QN (2009) and GD (2003) rankings are done at different time and use totally different methodologies.

I attach the rankings for comparison. You should decide for yourself if it's relevant to use some info from 2003. The 2009 QN ranking should be used as a guide.

QN 2009 Ranking QuantNetwork - Financial Engineering Forum - QuantNetwork Ranking of Financial Engineering/Mathematical Finance MS Programs
GD 2003 Ranking Global Derivatives v3.0 - Quantitative Masters Rankings (2003-04)

QN 2009 Ranking Methodology QuantNetwork - Financial Engineering Forum - QuantNetwork Ranking - Methodology
GD 2003 Ranking Methodology Global Derivatives v3.0 - Quantitative Masters Rankings (2003-04) Summary

 

[Byron]

Two random cents - a problem with the quantnet ranking is that it is not based on anything tangible about the program. Though the results might be "accurate" (whatever that means!), a student with a dilemma wouldn't know what to make of them, which might slightly defeat the purpose. For example, if I decide I'm going to be biased towards Chicago vs Baruch in my admits, all I know is that Chicago scored better on an algorithm not directly related to placements, curriculum etc - might not be sufficient grounds for me to base my thinking. (As in, the desired objectivity is still missing). For the GD ranking, I might think "well, someone thinks the reputation/curriculum of A is better than B".

 

[Andy Nguyen]

When you are making final decision between A and B, shouldn't you look at the specifics about the programs? Things like placement service, tuition, location, alumni network, etc.
Wouldn't it much better to work with a small group of programs and narrow it down to the one that most fits your personal needs?

The problem is not with the ranking itself, but with how people use the ranking. Of the 70 and growing programs in the US alone, the fact that one can obtain a list of programs where many strong programs are grouped together in itself is an extremely useful tool.

Now, it's up to you to decide a ranking where "well, someone thinks the reputation/curriculum of A is better than B" is more appropriate. Just think for a second, we are a few months until 2010 and what kind of data we have to make a well informed assessment on a given program. What they think is better does not mean it's better for you.

If we can't produce comparable data from more than a few programs today, how was it possible to make judgement about these programs 7 years ago. It's not a surprise that GD decided last year that it's not feasible for them to do another ranking.

The information about most programs are all available here for you to make a decision yourself. The ranking is not making an important decision for you.

 

[Byron]

You're right - It'd be silly to look at the ranking (any ranking) and make a decision on that basis solely - I would look at the placements, tuition etc. Perhaps my point got lost a bit.

The problem is not with the ranking itself, but with how people use the ranking.

Right, I think I get your point now - the rankings would serve an introduction to MFE program research for the uninitiated (as a list of strong programs), not exactly as a "ranking" per se.

Bit of a pointless debate I know, but a friend told me recently that "I'm now a bit less interested in university A because of quantnet rankings", and I just wondered if a web mining based algorithm should be sufficient basis for any such opinion changes.

 

[Andy Nguyen]

In a perfect world, we would have equal access to data from every program to make an informed decision.
Your friend is an perfect example of how not to use a ranking. It can be downright dangerous. Just think about how many people who made their career decision using the outdated GD 2003 ranking.
Of the thousands people who have viewed our ranking, there are fewer than a hundred who bother to look at the methodology. Of these rare people, none has gone through the algorithm in detail and able to point out any flaw or make any suggestion on how to improve it.

If I were to shop for a program, I would look at the ranking to see if there are any programs that I have no prior knowledge of to add to the list of the well known programs I know previously. I then would use QuantNetwork Program Picker tool to filter out the list to a smaller set based on my preferred location, length, deadline and budget. Next step would be to use the discussion forum here and the QuantNetwork wiki to get the info that are not available on the official website. I will also use the QuantNetwork Application Tracker to get an idea of how long each program takes to process. This would help prioritize when to apply to which program.

There are several features we are working on to help members with the whole process from information gathering to application tracking.

 

[Eugene Krel]

If someone were to choose solely based on a ranking system in just about anything, it is only as a means of placing blame away from themselves for the choice later on if things go badly.

 

[sak]

Personally I liked GD ranking a little bit better because they talked about various aspects (pros and cons) of different school. So, you can make a calculated judgement based on your priorities. You will have to make judgement based on talking to professors and students. You will need to compare course structure, weight of computation vs. research, financial loan aspect of education, placement, job quality at the end of the course, etc. In my personal opinion this is more important than analysis of the popularity vote. I say this because in the heart of it QN ranking does not say in real terms for each ranking number what proportion of it has what weight for curriculum, placement, etc. It is like looking at an index value and not knowing whether it is price weighted, value weighted, etc.

 

[Byron]

Thank you Andy, that helps.