About this question
Medium · Statistics & Regression · Quant Trader interview question · statistics, bootstrapping, confidence-interval, trading-strategy
You are building a trading strategy that relies on accurately estimating the confidence interval of a novel, complex, and highly non-linear signal derived from alternative data sources. This signal exhibits significant skewness and kurtosis. You have a limited sample size of 100 observations. You want to choose the best approach for constructing a 95% confidence interval for the mean of this signal. Under what circumstances would you prefer bootstrapping to an analytical confidence interval?