Don’t draw the downs apart – How to best simulate asset price drawdowns

This paper evaluates bootstrap simulation techniques for calculating the distribution of maximum drawdown (MDD), an important risk indicator in the stock and cryptocurrency markets. Using stochastic dominance tests, we assess the full distributional properties of the MDD under different methods. Our findings reveal that the standard Efron (1979) bootstrap, which assumes independence and identically distributed random variables, systematically underestimates the true MDD. While the moving block bootstrap provides reasonable estimates, it is subject to non-stationarity bias, particularly when large drawdowns occur at the boundaries of a return series. Alternative procedures, such as block-block bootstrap, tapered bootstrap and robust resampling, do not lead to better results. Of all the methods studied, the stationary bootstrap of Politis and Romano (1994) produces the most accurate and robust results, particularly for longer block lengths.