As I discussed in a previous article, the simple block bootstrap is a way to perform a bootstrap analysis on a time series. The first step is to decompose the series into additive components: Y = Predicted + Residuals. You then choose a block length (L) that divides the total