This is a pretty general question, but here goes.
I am testing a hypothesis that involves comparing to portfolios. I want to test if portfolio A outperforms portfolio B. Which statistical test should I use to test the hypothesis?
If you have two series of returns (e.g. monthly returns over several years) over the same time frame, a paired-sample t-test. Because they’re not independent, this is essentially a test of the null hypothesis that that the mean difference is equal to some number (usually zero).