Quant Trading vs Quant Research Interviews
Key differences
The interview process is very different for trading vs research positions despite having significant overlap in content. Personally, quant research interviews play to my strengths and quant trading interviews do not.
Both interviews broadly follow the same format:
Online exam
HR screen
Several rounds with PMs/Traders/Researchers
Online Exam
Trading
Quant trading online exams have extremely stringent time limits. Typically you are given no more than 1 minute per question. The questions emulate those of high school math Olympiads. The content isn’t advanced but the questions carry a high level of difficulty. Most online exams feature a mental math component geared around quickly completing sequences in your head, multiplying numbers, or calculating square roots. Being a trader is about thinking fast and making high quality decisions in real time.
Research
Quant research interviews filter for those with the ability to think deeply about theoretical topics and derive insights from data. The time limits on quant research exams are incredibly generous. I have often finished my exam with 30 minutes to spare. Most exams follow the same format:
Theory questions on linear algebra/probability/statistics/machine learning (~ 5-10 short answer or multiple choice)
Conditional probability questions (~ 3 multiple choice)
Knowledge of SQL queries (~ 2)
Algorithm questions (~ 2)
Data manipulation in python through a simulated jupyter notebook (short guided project)
The time limit for these exams ranges from 1-3 hours.
Onsite
QT
QT interviews are a copy-paste of the style of questions seen in online exams. The only difference is that the questions given in person are fewer in number and harder in difficulty.
QR
Most of the quant research interviews I have encountered were incredibly informal in nature. The interviewer scans your resume and questions you on your proposed strengths. A huge part of winning at QR interviews is not lying about your competency. If an interviewer asks you if you are familiar with a specific topic - there is no harm in saying that you are not if you genuinely know nothing about it. Do not waste your interviewer’s time by lying.
Research is extremely broad and interdisciplinary; nobody expects you to know everything about everything. Oftentimes the interviewer will come up with questions on the spot.
I cannot emphasize how much control you have over the direction of the interview based on how you design your resume.
When I was playing a lot of poker, I included an entire section of my resume about my online poker results. This enticed the interviewer to grill me on poker theory for an entire hour. If you state you have mastery over a topic, the interviewer will take great pleasure in testing your competency. Use this to your advantage.
QR interviews include long-form theory questions. You might be given a proposed modeling problem and asked about how you would approach it. Some examples include
How would you handle missing data when building a strategy for trading micro-cap stocks?
How does the standard deviation of sampled regression coefficients change with respect to the correlation of covariates?
How does a linear regression model change once you duplicate the training data?
Does adding an asset that is highly correlated to your portfolio improve Sharpe ratio? What is the relationship between the correlation of an asset to your portfolio and its effect on the Sharpe ratio?
What is the auto-correlation of a mean reverting process?
Unlike trading interview questions that are high difficulty basic topics, QR questions are light-medium difficulty advanced topics.
Comment your answers to the above questions below!