Crypto Regime Detection with Hurst, ADX, RSI, and CVD Filters

Summary

This crypto strategy excerpt describes a persistent regime classifier that labels market conditions as bullish, bearish, or neutral. It estimates trend persistence with a range and true-range based Hurst proxy, then combines that reading with ADX, RSI, directional movement, price relative to an EMA, and lower-timeframe volume signed by candle direction. Regime changes pass through a one-bar neutral transition, and price crossing the EMA can reset an active directional state. The script also includes a strict binary mode and settings for displaying regime zones.

The document identifies the strategy as using martingale and dynamic exits, but the supplied source cuts off before those mechanics appear. No performance results or validation are provided, and the Hurst estimate is a proxy rather than a formal stationarity test; despite the heading, the visible excerpt does not show a fitted hidden Markov model or an augmented Dickey-Fuller test. The thresholds and indicator combination are configurable, so their behavior and risk depend on the market, timeframe, and implementation.

Key ideas

• The regime state combines a Hurst-style proxy with ADX to distinguish trending conditions from neutral ones.
• Bullish and bearish labels require agreement among RSI, directional movement, signed volume trend, and price relative to an EMA.
• The classifier can hold directional states until opposing evidence or structural price conditions reset them.
• A one-bar neutral transition delays the display of a changed regime.
• The excerpt does not include the advertised martingale or exit rules, and it provides no performance evidence.

Tags

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