PhD Physics to Quant

twofish-quant

Also the fact that I've been lied to makes me very careful about saying anything that would turn into a lie to other people.

For example, physics Ph.D.'s on Wall Street make decent money right now, but if tomorrow I found out that there were tens of thousands of people getting their physics Ph.D. in the hopes of making tons of money on Wall Street, I'd cash out and find something else to do. So I'm *not* encouraging people to get their physics Ph.D. to make big money, and in fact if that's the reason you are getting your Ph.D. you are in the game for the wrong reasons, and you probably won't make big money either, since you have tons of people trying to "get rich quick" you've formed a market bubble that is going to pop.

oreliphan

Twofish, in your experience, what are the chances of a PhD Physicist being able to work a hedge fund? Is this the 1 in 50 chance you were mentioning earlier?

How about doing high frequency trading?

Are these qualifications tougher than what you mentioned in the previous posts about becoming a "quant"?

I did some searching around, and found out about Renaissance Technologies, which seems to hire mostly Math and Science PhDs. (Not that what they do is high frequency trading - they seem very confidential about their methods). They also seem to try and get people before they get on Wall St., so it seems like something one could apply for right out of grad school. Seems extra competitive though.

Can you tell us what your opinion is on high-frequency trading?

mgb_phys

Does Wall St still just look for physics PhDs?
In the mid-90s only physics/maths people knew this stuff, now it seems that every university in the world is offering courses on derivatives, options pricing and Black-Scholes.
The local CS dept even has a final year course on writing derivative pricing models.

So physics PhDs are good IQ/$ ratio compared to MBAs but it seems that this isn't arcane knowledge anymore.

oreliphan

But in the first place, it seems like Wall St. wasn't hiring physicists based on their "knowledge" of those things (derivatives, Black-Scholes, etc...)

It's the skills, and if you can acquire the same skill set with these "year courses," then it can become competitive.

Otherwise, a PhD can give you a better edge. As twofish says, it shows you can complete something even though you're miserable, and you can do original work without any hand-holding. So, it's not about the knowledge, and my impression is that banks expect you to learn finance on the job.

twofish-quant

It changes from month to month, but pretty high. On the other hand working at a hedge fund isn't a big deal, and the odds are that if you are working at a hedge fund with a physics Ph.D., you are going to be babysitting computer programs. *Owning* a hedge fund is a big deal.

There are a lot of different ways of making money. One problem is that when people talk about Wall Street they focus on the small number of senior executives. There are a lot of nice jobs that involve working in the trenches, but most people don't know that they exist.

Lots of physicists/EE/CS people involved in that. The thing about high frequency trading (or anything else) is that you don't have a lone genius setting in a basement, you have a hedge fund with maybe a 100-300 people of which you have a few Ph.D.'s that are doing some work trying to beat down latency.

There are lots of different firms in the business. RenTech is one of them, and they are more secretive than the NSA. But RenTech is only one of about 1500 hedge funds in the NYC area. One thing that's weird about a lot of the hedge funds (including RenTech) is that they don't like computer programmers with financial experience, so if you want to work on for them, it helps to have never worked on Wall Street.

It's an example of how things change very quickly. Last year when the stock market fell apart and then came up again, you had huge amounts of volatility, and when that happens, high frequency traders make a ton of money. When things settle down there is less money to be made there. It's why you can end up chasing your tail. If you in June 2009, you happened to have done a Ph.D. in high frequency software signal processing of radio interferometery data, you win. The trouble is that if you did something else, then by the time you get skilled at high frequency data, even if it takes you only six months, the gravy train would have moved elsewhere.

comp_math

Is it true that many of the super high pay quant jobs are related to algorithmic trading, more towards computer science/networks? This seems like a very small and specialized area, i.e. hard.

twofish-quant

Q1/Q2 last year, yes because algo trading was "hot" last year. It cooled down in Q4, and I don't see that field being particularly "hot" this year. It's not that the field is dead, it's just that I think this year, algo quants will be making "rather high" bonuses rather than "totally insane crazy" bonuses.

One problem (which I suppose is a good thing) is that you get a "gold rush" syndrome. Someone strikes it rich doing one thing, which means that everyone else rushes in after the gold until not much is left. If you are the first three firms to do algo, you win big. If you are the 100th "me-too" firm, you are competing against the other 99 firms that are basically doing the same thing.

Yes, which is why you struck the jackpot if you had relevant skills last year. You still can make decent amounts of money if you have those skills this year, but just remember that the big chunks of gold got mined out last year. If you want to know where the next "big thing" is just keep reading and thinking about the news.

creepypasta13

what if one just obtained a BS in math+physics, and is now debating between a career in mechanical engineering vs a quant/actuary? would it be feasible to obtain a MS in ME, then work for a few years, and if he doesnt like the work, he can just apply and obtain quant positions? or would that be a total waste of time to get the MS ME, and just get a MFE or some MS thats related to financial mathematics? if that person got offered a full-time position in programming with his BS degrees, should he take them, or instead just apply to grad school programs ASAP?

is it typical for quants to work 60 hrs/week, and make salaries in the 150k+ range? how are the salaries different for quants who have a phD in physics vs those with a BS physics + MFE or MS in some other field?

ferm

Thank you twofish-quant for your explanations, they are very helpful. If you don't mind, I would like to ask you some personal doubts:

I am doing a PhD in theoretical physics, and when I finish I am unsure if I will go for postdoc and misery, so I was considering finance as a possible "escape".

I am not doing much programming, just numerical stuff with mathematica and matlab. I don't think I can get to learn C++ for my PhD work, but I could start learning it at some point. Does this restrict a lot my future possibilities of finding a job as a quant, or basic knowledge of the language would suffice as long as I have sufficient mathematical skills?

Another question I have is about the future of physics PhD hiring in finance. Some universities are offering quantitative finance M.Sc., so I guess it will become more and more difficult to get to work as a quant with just a PhD in particle physics. Am I right or they are still hiring?

Finally, I am doing my PhD in an European university and I am not from the US. Does that play against finding a quant job in NY?

Thank you very much!

twofish-quant

There are three core skills, finance, mathematics, and programming. No one is outstanding at all three, and if you are outstanding at one, and decent at the others that is enough. Also the difference between no C++ and basic C++ is huge. If you can use a compiler and code an algorithm in C++ that greatly improves employability.

Also, if you are a wizard with statistical packages like R, you want to make that the thing that you are really, really, really good at.

I don't think very highly of MFE degrees, because they usually train people with a specific set of skills, and those skills become obsolete very quickly. Personally, if I were going after a masters degree, I'd avoid the MFE and go after a CS, applied math, statistics, finance, or MBA degree. The reason that I think that those are better degrees is that those other degrees aren't tuned toward investment banking, so you don't have a few people competing for the same jobs.

Something that is interesting about investment banking is that the fraction of jobs in banking that require a Ph.D.'s is a tiny fraction of the total, it's just that the scale is different. There are 800,000 jobs in the securities industry in the US, and 200,000 securities jobs in the US. The fraction of these jobs in which a Ph.D. is useful is really tiny, but it's enough to provide a few hundred new jobs each year for Ph.D.'s. Something interesting is that the US graduates 100,000 MBA's each year. Harvard B-school alone puts out about 900 new MBA's each year, which is almost as many people as all physics Ph.D.'s in the US.

I don't think any cares, although you might want to look at jobs London if you don't want to move too far.

creepypasta13

what about MS in engineering?

ferm

Thank you for your answers, twofish-quant, your opinions are really useful!

Just another question:

Could you please elaborate a little on which kind of socially useful works can be done in finance and why you think they are useful?

Thank you!

gatorphys

You should be able to transition over with an engineering degree, but you should really try to figure out what you actually want to do. You mentioned engineer vs quant vs actuary, all of which are completely different fields, with completely different career paths. So I would first figure out what you want to do, and then start thinking about grad schools.

creepypasta13

how does one figure that out? by working in industry? i already got my BS degrees

gatorphys

Well, I guess thats the problem with choosing any career right? How do you know what you'll like or dislike before you actually work at the job? But you could start by thinking about what your interests are. Do you like engineering? Programming? Hard-core math? Analyzing risk and uncertainty (actuary)? Do you follow the markets? Keep up with the financial news? etc...

twofish-quant

It's something that you need to read yourself and make up your own mind with.

Elizabeth Warren, Simon Johnson, and Paul Krugman pretty much think that the type of work that I'm doing is useless and ought to be banned. There are also lots of things on the net that take the opposite position. Some of it is corporate propaganda, but one thing that makes thinking difficult, is that just because something is corporate propaganda, doesn't make it incorrect.

One thing that that does influence my thinking is that I'm not interested in social usefulness to make myself feel good, since if your goal is to make yourself feel not bad, then it's trivial to come up with arguments that make you feel good. The reason that I'm interested in whether what I'm doing is socially useful or not is that I'm interested in my own bottom line. My paycheck has to come from somewhere, and if it turns out that what I'm doing *isn't* generating real wealth, then I'm in the middle of a bubble, and I need to figure out how to get out ASAP.

Locrian

I liked this article quite a lot. Apologies if I've posted it before, but it seemed relevant to your argument here.