The authors introduce a novel bootstrap approach to resampling asset price data that can be used for both finite-maturity assets and equities. The key insight is that they bootstrap primitive objects with more appealing statistical properties to avoid resampling series with strong time-series and cross-sectional dependence.