Study programme for quant researcher interviews

[alovya]

Study programme for interviews​

Hi everyone,

I was researching how to prepare for interviews and have ended up organising a decent amount of material into a study programme that I am sharing in this post, which describes:

1. the technical skills needed during a quant interview
2. a study programme to develop these skills

To understand what the interview process is like, read the respective introductions of this primer and "Quant Job Interview Questions and Answers" (2nd edition).

Technical skills​

Note that the technical skills I'm going to describe are the bare minimum you need at the interview, but you might need even more at the job itself.

If you highlight a specific technical skill in your CV, the interviewer is likely going to ask deeper questions about it.

For example, if I claim C++ as the language of choice for one of my projects, the interviewer might ask me harder-than-usual C++ questions to prove my claim.

I've listed the technical skills you need in an interview starting from the most important as I have gathered while doing my research.

Probability
At the interview, you need to show that you have a very good grasp of foundational statistical concepts. I will list some of these concepts in the section on resources.

Note: machine/deep learning, prediction, signal analysis, signal generation, etc, are the most desirable skills right now [thread 1, thread 2].

Coding
Quants have to be able to code well. Python/R are expected, and C++ is nice to have. You need to be good enough to do LeetCode style questions; such questions are becoming more common in interviews.

Brainteasers
These are puzzles/riddles; you won't be solving brainteasers on the job, but they are frequently asked in interviews so you need to be good at them to pass.

Maths
In addition to all of the above, you will often get general maths questions.

Finance
Depending on the your experience or the role you're applying for, you might asked specific finance questions. For example, if you are applying to be an interest rate option quant, the interviewer might ask you pricing questions.

Study programme​

I haven't included specific instructions for finance questions as this will depend on which specific role you are applying for.

However, probability, coding, brainteaser and and maths questions are going to pop up in any quant role, so I have divided my study programme into two parts.

The first focusses on making on making sure you can adequately answer probability, coding and brainteaser questions.

The second part lets you take a break from those questions by including general maths and finance questions.

Still, even in the second part, you should be regularly cycling through probability, coding and brainteaser questions.

Resources​

The programme uses the following resources:

1. Primer
2. Interview manual by the late Mark Joshi
3. Jane Street guide
4. Questions from Pete Benson of UMich
5. 50 mixed questions
6. 21 mixed questions, with answers
7. Compilation from Aaron Cao

Part 1​

1. Read the introductions of (1) and (2) to understand for yourself the interview process before starting any practice questions.
2. Work through the Jane Street guide to find out the statistics concepts you have to know.
3. Go through the probability questions in (2) and (4) that aren't too difficult for you.
4. Do 5 - 10 LeetCode questions, starting with the "Easy" category.
5. Do 5 - 10 brainteasers from (4), (5) or (6).
6. Return to the probability questions you weren't able to do in (2).
7. Do 5 - 10 more LeetCode questions; at this point you might be able to start doing "Medium" category questions. Example list of questions.
8. Do 5 - 10 more brainteasers from (4), (5) or (6) (or wherever else you can find them, to be honest).
9. Return to the probability questions you weren't able to do in (4).

Part 2​

1. Go through the maths questions in (2) and (4) that aren't too difficult for you.
2. Go through the options questions in (2) and (4) that aren't too difficult for you.
3. Do at least 3 questions from each category of probability, coding and brainteasers.
4. Go through the coding questions in (2) and (4).
5. Return to the maths and options questions you weren't able to do in (2) and (4).
6. Again do at least 3 questions from each category of probability, coding and brainteasers.
7. Go through all the interview questions in (1) and the questions in the final chapter X of (2).

At the end of this programme you should be ok with handling interview questions.

If you want more complicated questions, do some of the more challenging ones in (7).

Suggestions to improve this programme are welcome and please let me know if I have said something factually incorrect.

 

[alovya]

This study guide is aimed at people looking for a role as a front office sell side quant or as a buy side quant (I talk about sell side vs buy side here). I thought I could still edit my old study guide but apparently it’s too old to edit. I'm updating the guide because when I wrote the old one I didn’t fully comprehend why preparing for quant interviews is difficult.

Why preparing for quant interviews is hard

Preparing for quant interviews is hard because there is too much study material to choose from. This makes it really hard to decide what to study.

At heart, a quant needs the following technical skills (based on this post, this post and these notes):

Maths

• Algebra
• Partial differential equations
• Numerical methods
• Financial maths
• Linear algebra

Stats

• Probability theory
• Statistics
• Stochastic processes
• Time series
• Bayesian statistics

Coding

• Python, R, C++, SQL
• Data structures and algorithms
• Objected oriented programming
• Design patterns
• Version control (GitHub, GitLab)

Please ignore the list because many of us won’t be able to cover all of these topics in detail. To decide which topics to cover, the best approach that I’ve come up with is to try and answer (and fail) interview questions as quickly as you can: failing will reveal what topics you need to study.

For example, I was feeling particularly unconfident about my probability/stats knowledge so I had a go at the questions in this PDF from Jane Street. The questions involved basic concepts but I took my time hammering them down. Then I had a go at the probability/stats questions in Mark Joshi’s book. I got completely stuck at the very first question on stochastic calculus and I needed to study it a lot more before I could hope to answer the question. I took a week (on top of my job, not full time) to understand stochastic calculus: I used books, videos and random PDFs online, and coded up a Jupyter notebook to confirm and summarise my understanding. Coding up the notebook didn’t make me an expert, but in terms of self-confidence, there is a huge difference between zero knowledge and some (albeit basic*) knowledge. Be mindful of the Dunning-Kruger effect though.

I wouldn't spend more than 30-40 hours per topic at a time: this is enough to get an idea of the basics, but not enough to waste time. For example, I think I spent 40 hours on probability/stats, including: the Jane Street PDF, studying stochastic calculus and coding up the Jupyter notebook.

I need to clarify that this doesn't mean you should spend 30-40 hours per topic overall. You should spend more than 40 hours per topic but only after studying other topics. You should revisit an old topic at a more advanced level once you have let the ideas incubate in your head and have a fresh mind.

*You will be surprised how much you learn just by hammering on the basics. Knowledge of the basics translates to other questions, and it also primes your mind for handling brainteasers. I think having your mind primed like this for quant interviews is not spoken about enough. People don't pass quant interviews only because they're quantitatively gifted, it's also because their mind is primed to handle brainteasers.

Interview prep material

Below there are links to more quant interview questions than you will (probably) ever get through. Remember: there is too much to learn. The trick is to fail at the interview questions to identify what you need to learn. Start with the interview questions in the areas that you are weakest in, e.g., I started with probability and stats as I explained above.

Personally, I would recommend starting with either the PDF by Jane Street or the PDF by Dirk Bester (because it has the most up-to-date interview questions I could find). If someone knows of newer interview questions I would appreciate it a lot if you passed them on to me so that I can update this post.

To understand what the overall interview process is like read the introduction of "Quant Job Interview Questions and Answers" by Mark Joshi.

• Books
  • Quant Interview Questions and Answers by Mark Joshi
  • Frequently Asked Questions in Quantitative Finance
  • Heard on The Street: Quantitative Questions from Wall Street Job Interviews
  • A Practical Guide To Quantitative Finance Interviews
• PDFs
  • https://www.janestreet.com/static/pdfs/trading-interview.pdf
  • QuantitativePrimer/QuantitativePrimer.pdf at master · dwcoder/QuantitativePrimer
  • https://pbenson.github.io/docs/quantTechnicalQuestions/quantTechnicalQuestions.pd
  • https://www.math.lsu.edu/~smolinsk/Quant_Interview_Prep.pdf
• Web links
  • MoneyScience: List of Sample Questions for Quant Interviews
  • Big List of Quant Interview Questions with Answers

Note on how to practise coding

From this article: these days, in quant interviews, “There are more coding questions and coding types of brainteasers. If you go for a quant finance interview today, you will be required to know a lot about algorithms and data structures. The questions you encounter are similar to the questions you get at Google or Facebook."

For what I think is a fantastic introduction to algorithms and data structures, make sure to watch this video.

Practise coding by doing LeetCode. Follow these these three articles in order:

1. How to LeetCode properly
2. 14 Patterns to Ace Any Coding Interview Question
3. Grokking the coding interview equivalent leetcode problems

 

[QR_HFT]

ok