Market Regime Check

The Hurst exponent (H) measures the long-term memory of a price series. It was developed by H.E. Hurst, a hydrologist studying Nile river flows in the 1950s, and later applied to financial markets. Values above 0.5 indicate persistence (trending behavior), values below 0.5 indicate anti-persistence (mean-reversion), and values near 0.5 suggest a random walk with no detectable memory. In short, this market regime widget gives you a quick read on which of those three modes each instrument is currently in.

For each instrument, FractalCycles cross-validates the Hurst estimate with three additional methods: detrended fluctuation analysis (DFA), fractal dimension (FD), and a volatility scaling exponent (VSE). The four estimates are blended into a composite regime label and a consistency score. When the methods disagree, the regime is flagged as transitional rather than forced into a category.

The widget shows two readings side by side. The left pane is the long-run regime over the most recent 600 daily bars, blended across all four methods. It is slow to change but reliable, since a 600-bar window contains enough data for a stable Hurst estimate. The right pane is a faster, more responsive reading: rolling Hurst over the last 150 bars, computed from R/S only, with a delta against the same window 30 bars ago.

When the two panes agree, the asset is in the same regime now as it has been on average. When they disagree, the recent window is telling you the regime is shifting. The shift label ("toward trending", "toward mean-reverting", "stable") makes the direction explicit so you do not have to eyeball the delta.

The consistency score on the long-run pane is also worth watching. A consistency above 85% means R/S, DFA, FD, and VSE all agree. Below 60% the methods disagree enough that the composite label is not trustworthy on its own; that is when the "transitional" regime is most informative.

Read more in our guide to the Hurst exponent or our deeper write-up on multi-method regime detection.