Bootstrapping and Bias: The Economic Costs of Misjudging Downside Risk

Maximum drawdown (MDD), which is defined as the peak-to-trough loss associated with a series of returns, is a simple yet pivotal metric for investors with a downside risk budget. This study compares the performance of various bootstrap simulation methods in estimating the full distribution of MDDs from different global stock/bond portfolio allocations. In addition, the economic costs of biased estimates are quantified for three realistic decision-making scenarios. Compared to its benchmarks, the Politis and Romano (1994) stationary bootstrap produces the most precise estimates for the MDD. This assists in avoiding costly investment errors in portfolio construction and dynamic risk-control strategies.