Let me state first, this is NOT a thread about which one is more difficult!
BCHAD,
How does a PHD in finance work? (assume for the sake of argument that it is in the US) What specifically do you have to do to get it? Let’s assume you’re already a CFA charter holder, but other than that you have just your typical run of the mill undergrad ECON degree.
How long does it take? Is it a 5 year thing? Can it be done much sooner (for example if you have all sorts of research published)? Does it almost always require copious amounts of teaching intro to ECON 101? Could you somehow get one without having to do this?
I’m not interested in getting one now or in the near future, but I am just more curious as I don’t understand the process. All I know about PHDs I learned from my friends who are humanities guys (who had to suffer and teach and research for 10 years) or physicists who got it when they were 23 after publishing a few things.
Also what is the normal profile of a PHD candidate in Finance? Are they usually straight out of college? Or are they experienced? Should you be straight out of college if you were gonna do one?
Is there a real value add for a career outside of Academia? Or should you really only do it if you intend to go into teaching?
It is similar to a PhD in Economics, and the talent pipeline is very similar. Usually they are math majors with backgrounds in at least calc-based statistics and real analysis FYI.
Value add is there, but if you want quantitative roles, you’d prefer signing up for an MS in Quant Finance etc…
One of my friends recently got his PhD in Finance from an elite school in this field, and his profile was like “Ranked 1st in IIT-Mumbai, 9.something GPA etc”, another dude I knew was a beast at statistics. What you can do is you can go look at the universities you’re interested in and check out their grad student resumes.
Most importantly, if you can’t go top 3 don’t even bother you probably shouldn’t have even gotten a bachelors degree in that case, no matter what Greenman tries to tell you. You know what to do.
^Fck That and Fck You! It’s TOP 2 dipsht! If you don’t crack a perfect SAT score when you’re a freshman in high school, axe your sack, and your cock for that matter.
So it’s massively quantitave? I work with Alternatives that don’t really need any calculus, but I do have some really excellent research about the market that I’ve done that I am just sitting on. I have no shortgage of quants ready to do all the hard stuff for me (thanks India). Basically, I’m thinking of publishing it in a book, but was wondering if it could be the basis for a disseration. If they want me to calculate the escape velocity for the orbit of jupiter then I’d be way out of my element unless I’m allowed to pay someone else to do that stuff for me. But, that might be considered cheating in Academia?
Qualifications for a PhD Finance would be similar for an MS in Financial Engineering. Quantitative background (math major, comp sci major, some econ, some finance maybe) and you probably would have to have taken a full sequence of calculus, full sequence of real analysis (proofs), ordinary and partial differential equations, probability theory (again proofs), a couple of courses in linear algebra, and numerical analysis. I guess these are also more or less the mathematical qualifications for a PhD in Economics.
If we’re talking about top PhD programs in the US, your dissertation is often an extension of your thesis advisor’s research, at least to a great extent. Oh-sh!t professors will only accept your proposal if they are interested in your topic and how it helps to further expand their own research. If they agree to take you, then will make you their b!tch for at least two or three years. At least that’s what I’ve seen in two programs (Chicago and MIT).
I’m not a finance Ph.D. so can’t speak specifically to what finance departments are looking for. I do know a few, so could ask if you like.
A Ph.D. degree is really a degree that says you are trained to advance the knowledge of the field through original research. A master’s degree means that you have been trained to use the latest knowledge effectively in your field. So if you want to research or teach in a university, a Ph.D. is pretty much required.
There are plenty of master’s level people who can teach, and master’s level work is good for shops that do not require being on the cutting edge of everything. If you are shooting for a shop that claims to be a market leader and on the cutting edge, then the ability to generate new approaches and the sort of thing that a Ph.D. does in the academic world is necessary and can command a salary premium. Though some smaller shops like the ability to stick a Ph.D. on the back of some of their staff if they can, even if there’s nothing original coming out of them.
There are two reasons that finance and econ Ph.D.s are highly mathematical these days. The first is that to get a Ph.D. you have to advance the field through original research. Over time, the simpler mathematical formulations of things have already been explored, so in order to do new stuff, you tend to add things that then make the math necessarily more complex. Ironically, these often involve making assumptions that make the models even less plausible, just so that one can do something new, although not always, because some variations are designed to make overly general models more specific to some application. The second reason is that the internet and computing technology have made many more procedures tractible through numerical methods, and there is now much more data that you can crunch numbers on, so people do that.
Often times, the biggest and most useful insights in a field are actually qualitative changes in perspective that allow you to study something from a different angle than before. However, these innovations often don’t confuse people as much as 100 pages of greek letters in matrices, so it’s easy to undervalue that kind of stuff.
Ironically, one of the reasons I avoided traditional economics was because as someone who did Physics in undergrad, coming to an econ class and seeing their fantastically simplistic models of the economy horrified me. Only later did I realize that as approximations of reality, they were more helpful than nothing at all, but one has to have a healthy skepticism of pretty much all economic models. The mathematics makes them seem more defensible than they really are.
^ A Piled Higher and Deeper degree is nothing like a Master’s degree. It is for those who want to do research. Be prepared to spend 50-60 hours (or more) with your nose buried in data. Lots and lots of data. Then you enter it into a multiple regression model and interpret the results. If the results don’t fit what you want them to, then you beat the data into submission.
Once you’re done with your 50-60 hours of research, then you also have to teach two classes and have office hours and type syllabuses and exams and grade papers.
Somebody (Charlie Munger, maybe?) said (paraphrased), “Any econometric model with more than 3 or 4 variables is garbage.”
I really don’t think that finance PHD profiles are exclusively quantitative, far from it. That is really strongly dependent on the research interests of the program advisors. If they are into asset pricing theory or something similar, then sure. But your research could also be original and very qualitative in nature, where the most advanced mathematics you’ll ever need is running ANOVA with some software program.
PDEs, real analysis, stochastic calculus and pricing exotic derivatives? Maybe in 1997, but that’s not as sexy any more and certainly doesn’t dominate the curriculum of a typical finance PHD. You’ll need some basic quantitative reasoning to pass qualifying exams of course, and the level of math rigor required is evident by looking at the specific topics covered on these exams which will vary by school. I’d say overall comparable to the CFA course sequence or only slighly deeper.
^ I disagree. I think Calc 2 is an absolute bare minimum for the crappiest PhD at the crappiest school. If you get into any quality school, you’lll need stochastic calculus, differential equation, and linear algebra. And that’s just the math classes–you’ll need several stats classes too. And this is all “intro-to-PhD” work.
I disagree, you should know real analysis and perhaps algebra, not because you will use them, but because they want to see you’re smart and can think at a high level. In Econ PhD programs they prefer math majors to econ majors. But that’s for most traditional PhD programs, CT is trying to do something different, so I don’t know how that would work. Maybe he can turn it into a Political Economy PhD rather than Finance?
When you apply to a finance PHD program, you should first look up all PHD advisors and their areas of research focus and get in touch in advance with one or two guys who share your research interest and are likely to take you as a candidate. Your potential corporate finance PHD advisor will get you in if he sees research potential, without a useless background in stochastics PDEs. Being able to articulate your research ideas and demonstrate research potential, in front of the right audience that is likely to appreciate your research focus, is far more important for admission than demonstrating how smart you are with some quant background.
My impression about Finance PhD programs is that they tend to be rare and young compared to programs in Economics, Math, or other less industry focused fields. Many of the professors in Finance probably view academics in those other fields as their peers. So, it seems likely that Finance professors would hold potential PhD candidates to similar standards as those in other programs, including in terms of quantitative rigor.
Based on a quick internet search, it does seem like a majority of “Finance” PhD candidates have backgrounds in quantitative fields such as engineering or statistics. However, there are some with less quantitative backgrounds, and we don’t know if quantitative backgrounds were a requirement for admission in the first place.
A solid foundation of linear algebra is very helpful for statistics. Real analysis is mainly a proxy for knowing that a student can do proofs. Abstract algebra might be thought in the same way, but analysis has less jargon and is easier to learn. He also mentioned stochastic calculus, but I’m not sure I’ve ever heard of an undergraduate stochastic calculus class. That and diffyQ are used for option pricing and would be important for a finance PhD who wanted to focus on those things. Less important for other finance topics.
Calc 2 is pretty much the bare minimum (and I do mean “minimum”). Ideally, a decent program would want GMATs north of 700 (and top percent on the quant side), linear algebra, real analysis (indicates that they can think through proofs) and some decent math stats. But the biggest thin (IMO) is somehow conveying that you “Get” what a PhD is (not just a super-MBA), and that you have both the capability and (just as important, if not more so) the attitude that will make you successful as a pointy-headed quant-nerd.
The first year is typically foundation stuff - shoring up any quant deficiencies, econometrics, etc… Then seminars for a year or two where you get up to speed on all the extant research. After 2nd year, you get comprehensive exams, where they can ask you basically about anything you’ve seen in the last 2 years. Then comes the dissertation.
Most students get a few research projects under their belt as research assistants - usually as part of their program’s requirements. Sometimes they get an idea out of that that becomes a dissertation, and sometimes it comes out of a paper they did as part of a seminar. Or sometimes, their advisor give them an idea.
The write up a preliminary treatment/proposal of it - the basic idea, why anyone would give a sh*t, a lit review, and some basic statistics/results. Then they have to get approval for this from their dissertation committee. If they’ve kept them in the loop, their chair is not an idiot, and they get to this point, they typically get past this.
Then (and only then), they get to write their dissertation. The shortest one I’ve seen was 60 pages, and I’;ve seen some that ran 250 pages (mine was about 180). With luck, a candidate might get 2 (or even 3) publications out this. Once they pass the proposal defense step, getting the dissertation done usually takes 6-9 months.