You can also analyse the skewness and kurtosis of the interval PnL by getting third and 4th times of $Y_t$ respectively. Presumably you might conclude that for two collection with equivalent expectation and variance, you can choose the one with beneficial skew or lower kurtosis, but perhaps not based on https://ricardodvkxg.digitollblog.com/33920874/5-simple-techniques-for-pnl