I am finishing my CS PhD from rank ~35 school in the US. I specialize in Natural Language Processing, and have done a lot of Machine Learning work. Also have good programming skills. I have no prior experience in finance. I am looking at a career in quant research and trading, and was wondering if I stood a chance to get interviews at the top hedge funds, banks, market makers etc. I keep hearing that they recruit only from top ranked schools, so am a bit circumspect. I am an international student, so will need H1B visa sponsorship. I am open to relocating outside the US also if needed. Update: I am mostly done with my PhD research, and am not graduating for at least another 3 months. I would like to know what I can do during this time to make myself more appealing to these companies. I would also appreciate any interview preparation tips.
Decent chance. Generally higher ranked school would be better, and for quant trading CS is not very in-demand, but depending on research could get interviews.
What kind of research appeals to them? And in general, what things would they be (not) looking for in my Resume? I still have a few months before I finish my degree. What can I do in the interim to stand out and make myself a more desirable candidate?
If you have a few months left and are serious about quant trading jobs, I would do some interview prep - I’m sure you can find a lot of sample questions online. As far as your resume, my guess is the hardest thing will be to convince them that you are a good fit for quant and not SWE despite PhD in CS. Anything that pushes quant skills is likely to be helpful.
If you have a paper on NeurIPS Citadel might drop you a message with the invitation to interview. It should also apply to some other conferences (but not sure). After that, the only important thing is your performance in interviews. Given that, my advice is to spend the remaining 3 months carefully preparing to interviews, including brainteasers, statistical questions, ML (mostly linear models and ensembles), C++, algorithms&data structures. Could you easily answer those questions? (1) Consider N data points uniformly distributed in a p-dimensional unit ball centered at the origin. What is the median distance from the origin to the closest data point? Expected? (2) Could you derive GBDT? Which tricks could you apply to make it scalable(read: derive xgboost/lightgbm/catboost)? (3) Random walk on a circle. What is the expected number of steps to visit all nodes? (4) Could you implement a basic version of unique_ptr/shared_ptr from scratch?
Thanks for your note. I don't have anything published at NIPS or such conferences. My papers have all been in niche conferences, focused on applications as opposed to theory - think of something like the Text Retrieval Conference (TREC). Plus a couple in IEEE and PLoS journals. I haven't invented any new ML algorithms, and frankly don't go around chasing the hottest trends in the area. Most of the techniques I have developed though are based on ML algorithms, but typically I use a combination of techniques to solve problems - combining ML with non-ML techniques, using heuristics/optimizations, removing redundancies, parallelizing when needed, etc. Also, please see my previous comment above for some additional insights on my skill set and research. I have no idea if this would be of interest to these firms. And honestly, the first step for me is to ensure that I would be able to get interviews before I even think about committing the next 3-4 months fully to prepare for them.
>I have no idea if this would be of interest to these firms. And honestly, the first step for me is to ensure that I would be able to get interviews before I even think about committing the next 3-4 months fully to prepare for them. Well, preparation for the quant finance interview is an interesting venture itself. I personally discovered many interesting things related to math, stats, programming while doing it. Also, I think top firms hire fresh graduates based not on your past research/business, but on future potential. How to measure potential? Well, this is a million-dollar question literally.
Cars
Yesterday
2026
Why are Americans obsessed with SUV?
Layoffs
3d
15939
Will Tesla have just one round of 10% layoffs or multiple rounds of layoffs?
2024 Tax
11h
1894
Biden’s new tax proposal is wild
Layoffs
Yesterday
37619
Google CFO confirms 'large-scale' layoffs (Apr 17)
Health & Wellness
13h
4481
Why are women naked in gym?
I think you have a good shot at these companies. Very few CS PhDs would go to these companies honestly
Interesting. Your source of information?
Not true at all.