If I want to be a quant, what should I be doing?

[Frzn]

IE.

What should I take for undergrad? Pure math or applied math? Physics?

Do I need a masters in mathematical finance, or financial engineering or something similar?

Do I need a PHD after the masters? Can I JUST get the PHD and skip the MSC?

 

[twofish-quant]

IE.

First of all, quants don't exist. There are about a dozen different jobs in finance that have some quantitative aspect to them. Second, no one has any clue what the job market is going to be like in a decade, so what you want are general skills that will serve you well no matter what happens.

What should I take for undergrad? Pure math or applied math? Physics?

Get a good undergraduate degree. Take the hardest courses that you can, and also try not to be to focused. Take some classes in writing, history, philosophy. Keep in touch with the news and think deeply what is going on. Read books.

Do I need a masters in mathematical finance, or financial engineering or something similar?

Do I need a PHD after the masters? Can I JUST get the PHD and skip the MSC?

The good news and bad news about finance is that there are no strict rules, and whatever loose rules there are constantly changing. There is no check list of skills that will get you a job in finance. It's just a matter of learning as much as you can about as many different things as you can, and thinking about what is going on.

 

[Frzn]

Thank you for the reply. All I've read about the different types of quantitative analyst jobs was at http://www.markjoshi.com/downloads/advice.pdf

Basically I'm just attracted to being able to learn difficult math in undergrad and get a career applying it to finance using CS techniques

The reason I asked about financial engineering/mathematical finance is because that 'QuantNetwork' site mentions a lot of MSC programs. Would you say that these are extremely useful, even if they are not necessary? I guess now its up to me to decide between a pure math and applied math degree, although I'll probably go with applied math because there is more freedom in electives to branch out a bit.

 

[Jack21222]

If it makes you feel any better, twofish has a Ph.D in astrophysics. I wouldn't worry too much about the exact details of your degree.

 

[Frzn]

First of all, quants don't exist. There are about a dozen different jobs in finance that have some quantitative aspect to them. Second, no one has any clue what the job market is going to be like in a decade, so what you want are general skills that will serve you well no matter what happens.
Get a good undergraduate degree. Take the hardest courses that you can, and also try not to be to focused. Take some classes in writing, history, philosophy. Keep in touch with the news and think deeply what is going on. Read books.
The good news and bad news about finance is that there are no strict rules, and whatever loose rules there are constantly changing. There is no check list of skills that will get you a job in finance. It's just a matter of learning as much as you can about as many different things as you can, and thinking about what is going on.

Thanks so much!

Ok, now I understand that there really are no sets of rules on what I need..

but when you say get a good Undergraduate degree, I have quite a few options. I can take a 4 year applied math program, or a more intense math honors degree with an 'applied stream' or 'pure math stream'

The applied math degree allows for more branching out in electives (which, as you said, is a good thing!) but the honors would include harder courses from what I can tell (which you also said is a good thing). BUT, would I perhaps need an honors degree to get into graduate school for math? I would prefer to take the applied math non-honors degree so I could have more freedom with electives (the honors program is literally all math with no electives).

 

[Mathnomalous]

PBS' Nova aired a program titled "Mind Over Money" in Apr 2010. It is not specifically about "quants" but there's a small section dedicated to them. I've read several of Twofish-quant's posts about "quants" babysitting computers and his descriptions seem close to the mark. Here's the link to the site, scroll down the list to find it:

http://www.pbs.org/wgbh/nova/programs/

 

[mgb_phys]

Try and get into a decent university.

The problem with the industry having no real fixed roles is that the hiring process = find someone smart.
So you are slightly at the mercy of the prejudices of the hiring people.
ie PhD Maths > PhD Astrophys > PhD Physics > PhD Chemisty > MSc anything
Oxford/Cambridge/Imperial > UCL/Durham/Warwick > anything else
(the exact order changes of course with the background of the hirer!)

The nice thing about this is that although lots of places are now offering finance courses with elements of quantness - the hirers are still going to prefer a nuclear physics Phd. Although showing an interest and some knowledge of finance will help.
Another plus of doing physics is that when you discover that you aren't all that interested in working with a bunch of arrogant yuppies in order to make a bunch of rich guys even richer - you can always go and get a more interesting job.

The classic book used to be Fundamentals of Futures and Options Markets by JC Hull, there are probably better ones now but borrow a copy from the library to see if this is the sort of thing you can face doing.

 

[Frzn]

Try and get into a decent university.

The problem with the industry having no real fixed roles is that the hiring process = find someone smart.
So you are slightly at the mercy of the prejudices of the hiring people.
ie PhD Maths > PhD Astrophys > PhD Physics > PhD Chemisty > MSc anything
Oxford/Cambridge/Imperial > UCL/Durham/Warwick > anything else
(the exact order changes of course with the background of the hirer!)

The nice thing about this is that although lots of places are now offering finance courses with elements of quantness - the hirers are still going to prefer a nuclear physics Phd. Although showing an interest and some knowledge of finance will help.
Another plus of doing physics is that when you discover that you aren't all that interested in working with a bunch of arrogant yuppies in order to make a bunch of rich guys even richer - you can always go and get a more interesting job.

The classic book used to be Fundamentals of Futures and Options Markets by JC Hull, there are probably better ones now but borrow a copy from the library to see if this is the sort of thing you can face doing.

I'm already in a university. It's not especially good. Can anyone answer my specific question above about the various math degrees I'm considering?

 

[mgb_phys]

As twofish said, there isn't really any hard fast rules.
The maths of most pricing models isn't particularly complex for anyone with a maths degree and they aren't going to differentiate between different maths specialties (especially if it's an HR person doing the selecting)

Some programming experience is always good - as is an interest in finance.

 

[deRham]

"Try and get into a decent university.

The problem with the industry having no real fixed roles is that the hiring process = find someone smart.
So you are slightly at the mercy of the prejudices of the hiring people.
ie PhD Maths > PhD Astrophys > PhD Physics > PhD Chemisty > MSc anything
Oxford/Cambridge/Imperial > UCL/Durham/Warwick > anything else
(the exact order changes of course with the background of the hirer!)

The nice thing about this is that although lots of places are now offering finance courses with elements of quantness - the hirers are still going to prefer a nuclear physics Phd. Although showing an interest and some knowledge of finance will help.
Another plus of doing physics is that when you discover that you aren't all that interested in working with a bunch of arrogant yuppies in order to make a bunch of rich guys even richer - you can always go and get a more interesting job."

I am intrigued by this. My goal is a PhD in mathematics, mainly because I really want the experience [and could well want to end up researching].

But are you seriously telling me that someone who informally develops a strong interest and flair for how the financial market is going + reasonable basic programming skill with a PhD in less related things will be favored over someone with financial courses?

Does it matter that my PhD does not promise to be in PDEs or stats or something amenable to application? I realize they aren't hiring PhDs because they want great theoretical researchers in PDEs but probably for the experience and maturity, but I'm curious how far this line of reasoning goes.

 

[mgb_phys]

I am intrigued by this. My goal is a PhD in mathematics, mainly because I really want the experience [and could well want to end up researching].

Congratulations - that's the best reason.

But are you seriously telling me that someone who informally develops a strong interest and flair for how the financial market is going + reasonable basic programming skill with a PhD in less related things will be favored over someone with financial courses?

Yes - Several reasons.
Who is likely to be 'smarter'? The candidate that decided to do theoretical physics/maths PhD and got accepted in top rank university or somebody who decided to do an accountancy course?
If somebody is smart enough to do a maths PhD how long would it take them to learn the difference between a futures and options contract?
Does the finance course just give the candidate a quick overview of some of the terminology - or do they gain a deep knowledge of PDEs, monte-carlo simulations and general maths stuff?
The bank believes it is going to be doing new cutting edge stuff so what's the point of hiring somebody who learned all the old stuff. That's like the first car companies hiring the best qualified horse and buggy engineers.

Does it matter that my PhD does not promise to be in PDEs or stats or something amenable to application?

They are hiring for smart and ability to learn.
The 'Dr' will get you past the HR filter, then people will assume that if you didn't study PDEs/Monte carlo you can probably learn them quickly.

 

[deRham]

mqb, thanks for the reply - just read it. So would a doctorate in say a theoretical math field still be given good consideration when people with things like financial engineering or financial math masters programs are applying for the position? I would assume it would give good overview as in:

"Does the finance course just give the candidate a quick overview of some of the terminology - or do they gain a deep knowledge of PDEs, monte-carlo simulations and general maths stuff?"

---

As a question, what sorts of things should one probably learn as fundamentals to stay in the game? I currently am going into the PhD for the experience of it, as I said, with a view towards getting into research math, but I really would like to keep some financial track options open if possible. Right now, I am not trained in those things like PDEs, and while I don't think it would be hard to learn the basics, I am not trained in using them practically. I'm sure I would have the intelligence/analytical ability to do it, but currently my mindset is different.

I guess I can see that a PhD has exceptional training in learning new material quickly, making sense of it, and poking around for it, since that's a big part of research. So the real thing I'd want to do is break the barriers of familiarity beforehand, if this is a basic task.

 

[deRham]

"The bank believes it is going to be doing new cutting edge stuff so what's the point of hiring somebody who learned all the old stuff. That's like the first car companies hiring the best qualified horse and buggy engineers."

Well taken.

 

[twofish-quant]

The reason I asked about financial engineering/mathematical finance is because that 'QuantNetwork' site mentions a lot of MSC programs.

You'll need at least a masters degree in something, but the topic of the masters degree is flexible. People have gotten in with masters degrees in math, statistics, computer science, law, finance, business administration. Curiously, the one masters degree that doesn't seem to provide that much help in getting a job in finance is the masters of finance engineering. The MFE is something of a failed experiment and those are rapidly being turned into MBA programs.

Would you say that these are extremely useful, even if they are not necessary?

If you have a Ph.D. in physics then getting another masters is unnecessary. Just read stuff.

I guess now its up to me to decide between a pure math and applied math degree, although I'll probably go with applied math because there is more freedom in electives to branch out a bit.

Also when you market yourself, try not to use words like "pure math" or "theoretical." Those don't look good in a resume.

 

[twofish-quant]

The nice thing about this is that although lots of places are now offering finance courses with elements of quantness - the hirers are still going to prefer a nuclear physics Phd. Although showing an interest and some knowledge of finance will help.

This is true because the nuclear physics Ph.D. is much more likely to have deep knowledge of how to do a markov chain monte carlo simulation than an MBA or masters in finance.

One other thing that is *really* useful in finance is undergraduate teaching experience. Explaining what you are doing and why it is important in non-technical language is a very useful skill.

But are you seriously telling me that someone who informally develops a strong interest and flair for how the financial market is going + reasonable basic programming skill with a PhD in less related things will be favored over someone with financial courses?

For the types of jobs we are talking about, yes... Because

1) The jobs that we are talking about require deep mathematical expertise. There are lots of different types of jobs in banking which requirte different skills. The type of jobs that physics Ph.D.'s get hired to do typically involves deep physics/math/CS skills which finance people generally don't have.

2) Finance is constantly changing, and much of what we thought we knew three years ago has turned out to be wrong. There are no courses that talk about what the financial system is going to look like in 2012, because no one knows and we are all still trying to figure it all out. Dealing with situations in which no one knows the answers and which you have to figure out what the rules are and what the rules should be is practically the definition of a Ph.D.

Does it matter that my PhD does not promise to be in PDEs or stats or something amenable to application?

It's actually better if your Ph.D. isn't on the standard topics. If you are doing algebraic topology, presumably you can also do PDE's if someone asks you to do it. However, the good thing about knowing algebraic topology is that you can spot applications of algebraic topology to finance that I can't because I don't know algebraic topology.

What good is algebraic topology to finance? I haven't got a clue, you tell me...

I realize they aren't hiring PhDs because they want great theoretical researchers in PDEs

Banks do hire Ph.D.'s because they want great applied researchers. What good is your research topic to finance? No idea, but if I knew then I wouldn't need to hire you, would I?

 

[twofish-quant]

Who is likely to be 'smarter'? The candidate that decided to do theoretical physics/maths PhD and got accepted in top rank university or somebody who decided to do an accountancy course?

Also some of this has to do more with "interest" than "smartness". You can be reasonably sure that a mathematics or physics Ph.D. doesn't totally hate math, and can do math intensive things for 12 hours a day for years at a time without going totally crazy. You *don't* know that about your average MBA or finance major.

One side effect of this is that if your main goal is to make money, then you are much, much better off with an MBA than a physics Ph.D. Your typical investment bank has about 30,000 employees of which about only a few hundred have technical Ph.D.'s, so the vast majority of jobs in finance do not require a Ph.D.

On the other hand, these numbers are interesting when you realize that the US produces about 100,000 MBA's a year, and only about 1000 physics Ph.D.'s. One interesting statistic is that the Harvard MBA program puts out about 900 or so MBA's a year, which is close to the number of Ph.D.'s from all schools.

If somebody is smart enough to do a maths PhD how long would it take them to learn the difference between a futures and options contract?

Or figure out the difference between a futures contract and a forward contract, which is more subtle and interesting. Futures contracts are traded on a exchange, and forward contracts are traded directly between counterparties. That part you can google. Now how this fact affects the pricing model is something that you can't google for, and to find out the answer to that, you have to start e-mail people, asking around, and thinking about the topic.

A lot of people are extremely uncomfortable if you put the in a situation where no one tells you what you have to do and in which no one has written a book with the "right answer" and you get into serious trouble if you do the "wrong thing" even though no has told you what the wrong thing is in large part because no one knows what that wrong thing is. Look at the number of people that get upset if they are graded on material that the prof didn't cover in the lecture.

But Ph.D.'s tend to get hired because you know that they won't freeze in that situation.

 

[deRham]

"Banks do hire Ph.D.'s because they want great applied researchers. What good is your research topic to finance? No idea, but if I knew then I wouldn't need to hire you, would I?"

I see, so with what aim? Do they aim to use your ideas in a timely fashion? My impression is one of the main functions of academia is to perpetuate research which is interesting but possibly not immediately needed to anyone in particular. So the purpose of hiring this kind of applied PhD researcher would be to creatively produce fresh results in relation to the current trends? It would make a good bit of sense why they hire PhDs for this, then yes.

 

[Frzn]

Thanks so much for the replies Twofish.

One more thing- would you say more difficult courses or the ability to branch out (as you said, take writing, history, philosophy, etc.) is more important? I could take the absolute hardest program and have it be extremely focused, or take a less intense program with the ability to dabble in other things.

 

[chingkui]

Curiously, the one masters degree that doesn't seem to provide that much help in getting a job in finance is the masters of finance engineering. The MFE is something of a failed experiment and those are rapidly being turned into MBA programs.

Hi Twofish, is this what the finance industry insiders really think? I see more and more schools are offering MFE programs, if there is no market for the graduates, why would this happen? Also, from your experience, how well are the MFE trained graduates doing the job compare to math/physics PHD?