Quant research hires heavily from PhDs across physics, maths, statistics, computer science and engineering. The field matters far less than candidates assume; what transfers is the ability to frame a question, attack it, and be honest about the result.
What transfers well
Framing an underspecified problem. This is the core skill of a doctorate and the core skill of research on a desk.
Comfort with not knowing. Interviewers will push you past the edge of what you know deliberately. Academics are usually well-calibrated here, where strong undergraduates often are not.
Statistical care. Knowing why a result might be an artefact rather than a finding is the single most valued instinct.
What does not transfer
Timescales. A research idea that takes six months to evaluate is not a research idea on a trading desk. The adjustment is not to lower rigour but to compress the loop - decide fast what would falsify an idea, and test that first.
Elegance. Production code that runs, is tested and is readable beats a beautiful abstraction. Several firms interview specifically for this.
Depth in one narrow area. Your thesis topic will almost certainly not be the work. Its value in the interview is as evidence of how you think, which is why the questions will be about your choices rather than your results.
How to present your thesis
Compress it to four things, and practise saying them in ninety seconds:
- The question, in one sentence a non-specialist can follow.
- Why the obvious approach does not work.
- What you did, and specifically what you did if it was collaborative.
- What it does not establish.
That fourth point is the one candidates skip and interviewers notice. Being able to state the limitations of your own work without being prompted is a strong signal, and research-heavy processes press on it deliberately.
What to prepare technically
Probability and statistics to a level of precision beyond what most PhDs use day to day - the questions are elementary but demand exactness. Add programming, usually Python, at genuine working fluency.
Work the research-oriented question set rather than the speed-focused ones.