About this question

Hurst Exponent via Rescaled Range Analysis

Medium · time_series · Quant Researcher interview question · hurst_exponent, rescaled_range, long_memory, persistence, time_series

The Hurst exponent H characterizes the long-memory structure of a time series, with H > 0.5 indicating persistence (trending) and H < 0.5 indicating anti-persistence (mean-reversion). This problem estimates H using the classic rescaled range (R/S) analysis, a technique used by quantitative researchers to classify market regimes and validate signal persistence. Task Implement the function hurst_exponent_rs(returns: list, min_lag: int, max_lag: int) -> float to estimate the Hurst exponent H via R