BacktestFolio · Glossary

Stationary Bootstrap

In plain language

Instead of shuffling each individual month at random, Bootstrap samples consecutive sequences of history: if a month was part of a positive trend, the next sampled month is more likely to come from the same context. It better preserves real market structure compared to a Random Walk.

Technical definition

Monte Carlo method that samples blocks of historical returns of variable length (geometrically distributed), preserving temporal dependence structure and stationarity. One of the simulation models available in BacktestFolio.