What Actually Helps When You Are Preparing
I ran into this material back in 2014 when I was cycling through interview prep before my first desk job. The A Practical Guide To Quantitative Finance Interviews Xinfeng Zhou 2008 collection is basically a dense list of problems organized by topic. It covers probability puzzles, stochastic calculus questions, options pricing derivations, brainteasers that test your thinking out loud, and some coding exercises. It is not a textbook. It does not walk you through theory. It gives you a problem and expects you to figure it out on your own before looking at whatever solution is provided. Most people approach it wrong. They flip to a section, see a question they have never seen, panic, and then immediately read the answer. That takes nothing away from your preparation. Here is the structure that actually works. Pick a topic. Cover the relevant theory from a proper textbook first. Black or Shreve for stochastic calculus. Ross or Durrett for probability. Then open the guide and attempt every problem in that section without looking at anything. Sit with the hard ones. If you are stuck after twenty minutes, check the solution, write down exactly where your reasoning broke, and come back to it the next day. That delay matters more than you would think. I remember one specific case that stuck with me. There was a question about computing the expected value of a functional involving geometric Brownian motion under a change of measure. The solution in the guide assumed familiarity with Girsanov theorem in a way that skipped two pages of derivation. I spent an afternoon tracing through the Radon-Nikodym derivative myself because the shortcut in the answer key was not obvious to anyone who had only seen the theorem stated formally once. My workaround was to pull up a lecture notebook from a graduate stochastic processes course and rewrite the measure change step by step until it clicked. That single problem taught me more about changes of measure than three chapters of passive reading ever would have.
The Questions Themselves
The guide is strongest on probability and statistics. You will find variations on the classic red-blue ball problems, conditional expectation questions that look simple until you realize the conditioning structure is more nested than it appears, and distributional reasoning puzzles. A few of the probability questions are actually quite elegant. One asks you to compute the expected number of uniform random variables needed to exceed a value of one. The answer is e. The derivation is clean if you set up the integral recursively. Another common type asks about order statistics and expectations involving exponential distributions, which ties directly into queueing theory and Poisson processes. The brainteaser section is uneven. Some questions are genuinely good tests of first-principles thinking. Others feel like they were pulled from a casual lunch conversation and have no clean answer. I would treat the brainteasers as exercises in verbalizing your thought process rather than as problems with a single correct result. Interviewers use them to see how you handle uncertainty and whether you can narrow down a search space under pressure. Getting the right answer to a brainteaser matters less than demonstrating structured reasoning. The stochastic calculus and options pricing sections are where the guide shows its real value. You will see questions about deriving the Black-Scholes PDE from first principles, questions about calculating Greeks under different models, and questions that ask you to price exotic payoffs using symmetry arguments or PDE methods rather than brute force Monte Carlo. One question I keep coming back to asks about the price of a lookback option under geometric Brownian motion. The key insight is recognizing that the running maximum of a GBM has a known distribution and using that directly instead of simulating paths. People who try to brute-force that question in an interview tend to waste ten minutes writing simulation code when a closed form exists.
What the Guide Leaves Out
It does not cover programming extensively. If you are targeting a role that requires actual coding, you will need separate preparation. LeetCode style problems, basic numerical methods, and some Python or C++ implementation questions are fair game in modern interviews and this guide largely ignores them. It also does not address behavioral questions or case-study style discussions that many firms now include. The material assumes you already know the core theory and just need practice applying it under time pressure. There is also a gap in the treatment of interest rate modeling. You will find plenty of equity-focused questions. If you are interviewing for a rates desk, you should supplement this with materials on the Hull-White model, the Heath-Jarrow-Morton framework, and quanto adjustments. The guide simply does not go there.
Get the Full Details
How Long It Takes
Going through the whole thing systematically takes roughly forty to sixty hours if you are doing it properly. That means actually working each problem, not just reading solutions. If you are already comfortable with the underlying theory and mainly need to build speed and fluency, you can compress it into twenty to thirty hours by focusing on the sections you are weakest in. I would suggest spending about six hours on probability, eight on stochastic calculus, six on options pricing, four on brainteasers, and the rest distributed across the remaining topics based on your target role. The document circulates widely online. You can find it on various academic and finance forums. I would recommend looking for a recent copy since older versions sometimes have typos in the solution sections. I once followed a solution path for a conditional probability question that led to a negative variance because of a sign error in the printed answer key. The mistake was obvious once I plugged in concrete numbers, but it would have wasted someone a significant amount of time if they had just accepted it at face value. The core of this preparation is not memorizing answers. It is building the habit of working through a problem from scratch and being able to explain each step clearly. Interviewers can tell when someone has seen a question before and is reciting a memorized solution versus someone who is actually reasoning through it in real time. Practice your verbal walkthroughs out loud, ideally with a timer, because the constraint of speaking while thinking is a separate skill from just solving the problem quietly on paper.
One more thing that surprises people. The guide contains questions that overlap heavily across editions and across different prep materials. If you have also worked through problems from books like Heard on the Street or Quant Interview Questions, you will notice significant overlap. That is not a bug. It is evidence that these questions represent the stable core of what interviewers actually ask. Focus on mastering those rather than hunting for new problems that might never appear.
