Estimating Drawdown Risk with a Resampled Cold Blood Index

Summary

The document presents a Cold Blood Index intended to help a systematic trader judge whether a live drawdown is unusual enough to warrant leaving a strategy or whether continuing may be reasonable. The supplied code reads a historical balance curve, resamples it with bootstrap sampling, and examines rolling intervals. It counts how often interval losses reach a specified drawdown threshold, then estimates the probability of encountering such a loss during a live trading period. Repeating the process yields percentile estimates for the index.

A table compares a method based on percentiles, a single-sample index, and a lower-percentile resampled estimate across several drawdown depths. The article argues that resampling better reflects randomness in individual outcomes and may improve decisions about preserving capital. It presents the measure as an aid to objective decision-making, not a guarantee that a strategy should be retained or stopped. Its stated limitation is serial correlation: the resampling method may fail when returns are serially correlated. The excerpt does not provide broader validation or enough detail to assess performance across different strategies.

Key ideas

• The index estimates the chance of seeing a drawdown at least as deep as the current one within a specified trading horizon.
• Bootstrap resampling of a balance curve can capture variation across sampled return paths.
• Repeated estimates can be summarized with percentiles to show uncertainty in the index.
• The method is intended to inform continue-or-stop decisions, rather than to make them automatically.
• Serially correlated returns can undermine the resampling approach.

Tags

This summary was written by Stratmill's research agent from the original; it is not a copy of the source.