Quant Interview Statistics Cheat Sheet
A compact statistics cheat sheet for quant interviews covering distributions, variance, covariance, correlation, sampling, regression, and evidence.
Quant researcher, trader, and developer candidates refreshing statistics for interviews.
Know what distributions describe
A distribution describes possible outcomes and their probabilities. In interviews, connect the distribution to mean, variance, tail behavior, and the data-generating process. Do not treat distribution names as labels without assumptions.
Variance and covariance
Variance measures spread around the mean. Covariance measures how two variables move together. For a sum, Var(X + Y) includes Var(X), Var(Y), and 2Cov(X,Y). That covariance term is the part candidates often forget.
Correlation and independence
Independence implies zero correlation in many standard settings, but zero correlation does not generally imply independence. This distinction matters in signal, portfolio, and experiment questions where dependence can be nonlinear.
Sampling and standard error
Sampling error explains why a result from limited data can be noisy. A larger sample usually reduces standard error, but bias, leakage, and bad sampling can still make a precise-looking estimate unreliable.
Concrete example
If two strategy signals each have variance 1 and covariance 0.4, the variance of their sum is 1 + 1 + 2(0.4) = 2.8. Ignoring covariance would understate the combined variability.
Common mistakes
Candidates memorize tests without checking assumptions, confuse correlation with causation, and treat backtests as proof. A strong statistics answer says what the result suggests and what could break it.
Practice the pattern
Use the LeetQuidity curriculum and calibration to turn this topic into a focused practice plan.