Thanks for the feedback! I guess in terms of career options I was thinking of working as a front office developer (after getting the MSCS) and then moving into a trading role. Is this still possible or is a MFE required now? I'll remember to focus on ML, big data, C++, numerical methods instead of the fin math courses for the MSCS.As an outsider looking in at the industry I find it's hard to pinpoint exactly what I want to do. But it seems like a phd opens up the most number of quant opportunities.

QuoteOriginally posted by: xploringThanks for the feedback! I guess in terms of career options I was thinking of working as a front office developer (after getting the MSCS) and then moving into a trading role. Is this still possible or is a MFE required now? I'll remember to focus on ML, big data, C++, numerical methods instead of the fin math courses for the MSCS.As an outsider looking in at the industry I find it's hard to pinpoint exactly what I want to do. But it seems like a phd opens up the most number of quant opportunities.There is only one example I know of someone who became a trader after doing after 6-7 years of IT followed by an MFE from a top univ and that was in the very good times. There is a brand of trader who is a high frequency programmer in a Hedge Fund. He implements trading strategies and takes huge amounts of programming skills. It is usually machine dependent and I doubt you have the background or inclination to do that but you be a judge of that. Either do an MFE and try for the moon which at your age and given the market conditions is nearly impossible. Or do an MFE and resign yourself to a life of a second class but well paid citizen called Quant Dev in a bank fixing Excel Macros. Big Data and ML are a totally different line. It will take you to good careers but probably not in banks where they will prefer PhD.'s as usual. If you are doing a MS in CS then forget finance for the time and focus on what is interesting to you. The good thing about IT is that there are so many good software companies who do exciting work and pay quite well. Finance unless you are in the right places pays terribly when to factor in the quality of work and long hours and horrible and shitty perception of programmers. Even in absolute terms Finance is not that great paying when you live in NYC. I doubt you are believing me now. But when you start taking interviews most of them will turn out to be IT. Headhunters will try to sucker you into going for those interviews. They will refuse to present your resume to Quant or Trader positions. You will be told that take up "this" Front Office Developer role so you can make connections to become a Quant. Eventually you will cave in and then when you go to work you will realize that "the people" whose connections you would like will think of you a little more than the Janitor who comes to clean the garbage under your desk each evening (at worst) or a means to their end (at best).

capafan2, would you say that the only route to good quant roles in banks is through a PhD?

QuoteOriginally posted by: xploringcapafan2, would you say that the only route to good quant roles in banks is through a PhD?In my opinion yes. Even the MFE guys I have seen who became quants ended up being just quant developers doing Quant infrastructure development. They do end up on the Business Bonus pool but only slightly better off than IT guys in terms of money and status. Most others end up in Risk Management which I think is a seriously inferior job to a good Technology Job outside a bank. Leveraging your current experience and moving into Big Data will provide you more value for your experience than the path you are asking about. It is possible you are very talented and lucky but you be a fair judge for that.

Generally speaking, I have never been exceptionally talented or lucky, so it looks like it's either go the PhD route or just stick with IT... I kinda figured from looking at all the quant profiles on linkedin that all the PhDs have the best jobs.If I focus on big data and ML where can I go with that? Does it give an edge to finding a front office developer job? Are robotics and artificial intelligence any good to have? If I were to study something purely out of interest for the MSCS I'd focus on robotics/AI/computer vision. For a prop trader do you think a PhD is necessary?

QuoteOriginally posted by: xploringGenerally speaking, I have never been exceptionally talented or lucky, so it looks like it's either go the PhD route or just stick with IT... I kinda figured from looking at all the quant profiles on linkedin that all the PhDs have the best jobs.If I focus on big data and ML where can I go with that? Does it give an edge to finding a front office developer job? Are robotics and artificial intelligence any good to have? If I were to study something purely out of interest for the MSCS I'd focus on robotics/AI/computer vision. For a prop trader do you think a PhD is necessary?The "Prop Trader" has many meanings. You could become one today by putting 100K or less of your own with some shady outfit. Some people are even successful doing that. Most fail. PhD. is another 5 years. It will be different world by the time you come out and you never go for PhD. to get a job. That is disaster waiting to happen. ML and Computer Vision are related. It all boils down to how well you can do Bayesian Statistics and Program like a monster when the number of computations or size of data is huge. Most ML is a computational problem. Some people do it in C++ but Java/Python/Hadoop are much more popular. If that is your skill stick to it. Plenty of opportunity here than to try write device drivers in C or C++ where your experience will not count as you will be seen as a fresh grad. Even if you do not get a job in Robotics or Computer Vision you will be considered a ML expert. But definitely try to develop Big-Data Hadoop and CUDA background. Both matter in this line. And using the Amazon Cloud and datasets you can get real world experience which will count.

What should you do with your life

Hi capafan2, noted. It's great to be able to study something interesting like robotics for the MS and know my skills still might be in demand. I think at this point I'm going to try to find a developer job in the industry, get my feet wet, and figure out next steps.

QuoteOriginally posted by: xploringI guess in terms of career options I was thinking of working as a front office developer (after getting the MSCS) and then moving into a trading role. Is this still possible or is a MFE required now? Cant say about small companies but switching from dev to trading never happens in large banks. If you want to be a trader then do MFE and give it your best shot. Chances of trading position after MFE is also very small but this is the max likelihood path.

QuoteOriginally posted by: ashkarQuoteOriginally posted by: xploringI guess in terms of career options I was thinking of working as a front office developer (after getting the MSCS) and then moving into a trading role. Is this still possible or is a MFE required now? Cant say about small companies but switching from dev to trading never happens in large banks. If you want to be a trader then do MFE and give it your best shot. Chances of trading position after MFE is also very small but this is the max likelihood path.As I have said two times in this thread, this is an approach that is bound to fail. You are asking "What pays other people well now, how do I get there in 2 years?"Instead you need to ask, "What can I do well, how do I leverage that into a long career?"

I've thought a lot about the advice in this thread and it does makes sense to do what you're good at instead of trying to cater to the industry.I'm in a MSCS program and applied because I thought it would make me marketable to financial companies, but have found my algorithms and data structures courses ridiculously tough. The math used is surprisingly easy, like logs, algebra, summations, and basic calculus, but memorizing all those algorithms has been difficult. It's obvious after comparing my grades from math and CS classes that math is easier and I can get better grades.How do IBs and HFs feel about MS in Math degree? Would they still be willing to hire if you also have a few CS courses mixed in?

QuoteOriginally posted by: xploringI've thought a lot about the advice in this thread and it does makes sense to do what you're good at instead of trying to cater to the industry.I'm in a MSCS program and applied because I thought it would make me marketable to financial companies, but have found my algorithms and data structures courses ridiculously tough. The math used is surprisingly easy, like logs, algebra, summations, and basic calculus, but memorizing all those algorithms has been difficult. It's obvious after comparing my grades from math and CS classes that math is easier and I can get better grades.How do IBs and HFs feel about MS in Math degree? Would they still be willing to hire if you also have a few CS courses mixed in?Memorising maths/algorithms is not a good starting point? How much real programming is in the course and are the algorithms explained in a purely theoretical fashon?To be honest, the maths content looks basic.

The Intro to Algorithms course is purely theoretical, we only use pseudocode. The midterm coming up next week requires that we memorize a lot of sorting algorithms and their component functions, like heap-increase-key, square-matrix-multiply, etc. I'm terrible at memorizing stuff and I talked to another student who said she had trouble with discrete math but breezed through this course. I'm the opposite--breezed through discrete math (enjoyed the proofs), but struggling with algorithms.If the rest of computer science is only about algorithms and data structures, I definitely want to switch over math. But wondering if IBs and HFs would hire someone with a MS in Math?

QuoteOriginally posted by: xploringThe Intro to Algorithms course is purely theoretical, we only use pseudocode. The midterm coming up next week requires that we memorize a lot of sorting algorithms and their component functions, like heap-increase-key, square-matrix-multiply, etc. I'm terrible at memorizing stuff and I talked to another student who said she had trouble with discrete math but breezed through this course. I'm the opposite--breezed through discrete math (enjoyed the proofs), but struggling with algorithms.If the rest of computer science is only about algorithms and data structures, I definitely want to switch over math. But wondering if IBs and HFs would hire someone with a MS in Math?Industry needs people who can program. QuoteIf the rest of computer science is only about algorithms and data structures, It may be may personal prejudice, but CS education seems to be still out of touch with reality. edit: Ok, learning algos + datastructures is fine (I could call it forward induction) but a more common scenario is to locate the most suitable ones for a given problem (backward induction).

QuoteIf the rest of computer science is only about algorithms and data structures, I definitely want to switch over math. But wondering if IBs and HFs would hire someone with a MS in Math?Anything is possible but you do need programming skills for any quantitative career. Or else you will be looking at jobs where you join at the entry level learning SAS or Matlab or worse still Excel Macro's to put your math knowledge to use. Also 100's of MFE's and MBAs end up in Risk Management or Client Services roles at Financial companies each year. They are not programmers but they know Math at a reasonable level and their programming tool is Excel. I have not seen an MS Math entering such roles. I think MFE or MBA helps here.Knowledge of algorithms + Computer Architecture is necessary. You should learn to program as well. Not just fancy algorithms but just regular run of the mill type programming as that is what you will do day in and day out. It is quite different in flavor than solving complex math problems in code. Basically you need to get to a point where programming feels natural and you don't think about "how" to do basic things. Instead you start focusing on "what" needs to get done. It takes about 6 months to 1 year of active programming to reach that level of comfort.