1. Limited time only! Sign up for a free 30min personal tutor trial with Chegg Tutors
    Dismiss Notice
Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

PhD to Finance with just Python?

  1. Oct 10, 2012 #1
    I'm finishing my astrophysics PhD and looking for other career options. I do observational astrophysics, almost entirely based on Python coding on very large data sets. Lots of Numpy+SciPy, monte-carlo, modelling data doing non-linear fitting using maximum likelihood on computer clusters, stuff like that. I would say my speciality is being able to write code quickly to get something out of a large amount of data. However, I also work on developing large Python reduction pipelines in collaboration with others (object-oriented), where we have to automate careful statistics and errors of millions of pixels in real time.

    I'm a little rare in the sense that all my research has been Python/Perl/R, and very little C/Fortran. I can do basic stuff in C, but my high-level experience is all Python.

    I have three questions:

    -- Do I have the skills to transition to finance? Is high-level C a must?

    -- Are there finance jobs a PhD can get that are more data-oriented than a pure algorithm-inventing quant (my stereotypical image of a quant)? I like analyzing data and I like debugging code. Money is good, but I'm mostly leaving academia for some stability in life. A finance job that pays less than quant salary would be okay with me, if they exist for people like me.

    -- Is applying for NYC jobs from overseas pretty standard? Or are skype interviews and visa applications a total pipe dream?
  2. jcsd
  3. Oct 10, 2012 #2
    You absolutely must know basic C++. If you have skills in other areas you don't need to be a C++ expert, but you absolutely must be willing to program in C++ if necessary.

    Fortunately, knowing enough C++ so that you can do basic programming is something you can pick up with about a month of self-study. You don't need to be an expert, but you do need to show willingness to program in it. Also if you know python classes well, then transferring that knowledge to C++ shouldn't be too hard.

    The stereotypical image of a quant is a job that died about five years ago.

    There are lots of different jobs. Having said that the market is moving toward more data-driven stuff than inventing algorithms (since it's now very, very, very painful to invent a new algorithm). For every one Ph.D. that's inventing an algorithm, there are now probably five or six that are throwing data at it and seeing how it reacts to different situations.

    Part of this is regulatory. The days in which someone could come up with a new derivatives algorithm and toss it into production in an hour are gone. The government will not let you do this. Before you run a new algorithm now, you need weeks or months of testing. Where you toss tons of data against the algorithm and see how it reacts. So you have lots of people writing test harnesses in Numpy, gathering statistics, etc. etc.

    Now algo trading is different, since you can come up with a new algorithm and toss it into production with about a week of review. But that involves lots of statistics, data mining, data analysis.....

    But that's good for you.... :-) :-)

    You need to change your thinking a bit on money. If you are willing to work for nothing, there are plenty of people that are willing to have you work for nothing and take all of the cash that your stuff generates. If you demand nothing, you get nothing.

    You have skills, you have value. This isn't academia where you have to beg and plead for a position. When people look at you, they see dollar signs, and you need to be willing to get dollar signs. One nice thing about Wall Street is that people tend to be less hypocritical. Your company makes a ton of money. The senior executives take most of it. You get bread crumbs. But it's amazing how big the bread crumbs are.

    One reason this matters it that salaries are going down. The days of huge payouts are over. The pay is still good, but this ain't academia, so if they go down too much, it's time to forget finance and do something else.

    Also, there is no stability. The markets could change quickly and you could be out of a job for no reason that has anything to do with you. One reason for demanding $$$$ is so that you have something if things fall apart.

    As a new person, you aren't going to get more than the standard market rate. However, your goal would be to avoid getting cheated into accepting less than the standard market rate. glassdoor.com and a good HH would give you the standard market rates. Also, the HH is paid a percentage of your first year pay, so if you are willing to settle for less money, your HH won't let you because it reduces his commission.

    I have a ulterior motive in all of this. If your salary goes down, then my salary goes down. If there are a ton of people willing to do my work, for much less than they are paying me, then my boss is going to stare at me with shark eyes.....

    Something else is that NYC is a very expensive city and you will be hit with a 45% marginal tax rate and $2K-$3K rents.

    If you aren't in the US/Canada then it's more difficult for interviews. If you are physically outside of the US, then you should consider London, Singapore, or Hong Kong. It's easier for people in some countries to work in London than in the US. If you are physically in the United States but are a student, then it's not going to be much of a problem since visa sponsorship is pretty standard.

    There are lots of Canadians working in Wall Street since getting a work visa if you are Canadian is basically a rubber stamp.

    The standard practice is that the first interview is telephone, and the later interviews are face to face. One thing that I did when I was job hunting was to buy a ticket to NYC and stay in a hostel for two months. This made things easy, because we could have a phone interview in the morning and if they liked me I could show up in the office that afternoon.
    Last edited: Oct 10, 2012
  4. Oct 11, 2012 #3
    Sidequestion: Is the commute from Queen's or Brooklyn worth the money saved? Studios there cost much less than in the city.
  5. Oct 11, 2012 #4
    I personally don't think so. Brooklyn is fast becoming very pricey and the commute is not ideal. Especially from Queens.
  6. Oct 11, 2012 #5


    User Avatar
    Education Advisor

    Suppose for the moment that I was offered a job in New York. Where would you advise would be a good place to live and commute from?
  7. Oct 11, 2012 #6
    Depends where you were working and your price range and where you wanted to be. I personally would want to live somewhere around lower east side or 14th-ish and A-ish as those are nice areas and good places can be found for reasonable prices. Plus it has good food/nightlife. Broadway area in Queens around the N track is good too, with great food for good prices. In Brooklyn I've had friends live around Jefferson and Morgan, both off the L stops, for cheap rent. They would often have problems commuting cause of the L track being shut down/broken/messed up though. Certain parts around Chinatown and Soho are cool too, but can get expensive fast. I've heard apartments in Gramercy area can be gotten for decent prices but I have no experience with them.
  8. Oct 12, 2012 #7
    A lot depends on the particular neighborhood. Brooklyn and Queens are both huge cities themselves, and each have about 20 to 30 different neighborhoods. There's also New Jersey.
  9. Oct 12, 2012 #8
    FYI, you can tell someone's salary/social status by where they live. For example, the "rank and file" quants that I know live in places like Park Slope in Brooklyn, Forest Hills in Queens, or suburban New Jersey. No one I know that isn't a manager lives in Manhattan.

    Having kids makes a *big* difference. One thing about New York if you don't care about the quality of the schools and if you want a studio apartment, the rent goes down dramatically.

    Also things can change dramatically over the course of a year, and neighborhood boundaries can move around.

    Yup, and it's interesting to see gentrification in action. What happens is that you have artist types move into a neighborhood for cheap rent, the neighborhood becomes "cool" and then it becomes too expensive for artists who move to the next station in the L train.

    What I ended up doing was getting a cheap studio M-F in NYC, and the flying back to Texas on the weekends.
  10. Oct 12, 2012 #9
    One thing about NY is that finding an apartment is radically different than in most other places. In NY, an apartment will be on the market for only a few days, so it's a matter of going through Craig's list and seeing what's available. The other thing is that because everywhere is connected by subway, it's not that hard to go and see for yourself.

    Kids make a big difference. Also, lifestyle makes a big difference.

    Also NYC is a *fantastic* place for science. One reason NYC is cool is because you have eight million people living next to each other, you end up meeting all sorts of interesting people.
  11. Oct 12, 2012 #10


    User Avatar
    Education Advisor

    I recall having a job interview once for a large pharmaceutical firm in New Jersey, and the cab driver who was taking me from Newark International Airport to the hotel I was staying was telling me that quite a few take the train from near the airport to downtown Manhattan for work or pleasure (about an hour commute each way).

    He also mentioned to me that apartments that there were decent, affordable apartments available in places like Cherry Hills or East Hanover, NJ.
  12. Oct 12, 2012 #11
    Thanks so much, twofish! It sounds like model verification is really what I'm most suited for. I'm in Canada, and I personally consider the idea of living in NYC to be one of the main perks when thinking about moving into finance. I'll be brushing up on my c++ now that I know it's not always required at the expert level.

    This is probably one of the hardest things for me to get used to!

    I guess by stability I meant location, not job duration. I am well educated and young so I know I'll land on my feet if I'm in a large city.
  13. Oct 16, 2012 #12
    Yes, it's quite common. The other thing is that commuting disruptions cause less problems than you might think. If you get to the train station and the trains aren't running, then you go back home and work remotely.

    One of the more interesting commutes, is that I know no small number of people that work in either HK or Singapore and live in Shenzhen or Malaysia. You have to remember to bring your passport card when you go to work.
  14. Oct 16, 2012 #13


    User Avatar
    Education Advisor

    That's true. I have a number of friends from Malaysia who have told me that they have relatives who either live in Singapore or commute there for work purposes. Not too surprising, given that Singapore used to be a part of Malaysia (I would suspect that a sizable number of people in Johor state, in southern Malaysia bordering Singapore, will likely commute there).

    I'm curious though about the commute from Shenzhen to Hong Kong.
  15. Oct 16, 2012 #14
    Hong Kong is tiny. It's interesting to superimpose a map of HK on a map of NYC, and it's an hour by subway from Central to Louwu.

    If you are a resident of Hong Kong who is a PRC citizen, you get two cards. One being your HK Residency Card and the other being a Home Return Permit, and there are automatic gates that let you swipe your cards to get in/out of HK and in/out of Mainland China. It's no worse than queuing to get on the subway.

    Cross-border commuting won't work for some that has to get their passport stamped, simply because you are going to run out of pages pretty quickly.
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook