Directional Change Indicators for HMM-Based Forex Regime Trading
Summary
The article describes directional change (DC), a threshold-based way to represent price movement through confirmed turning points rather than fixed-time observations. It defines upward and downward runs, their overshoots, and three derived measures: total price movement, trend duration, and time-adjusted return. These indicators are proposed as volatility features, with the R measure used as the input to a hidden Markov model for regime classification.
The practical example uses daily GBP/USD data and a threshold of 0.4%, then describes a simple strategy based on the inferred regimes. It contrasts this daily setup with cited research using intraday DC data aggregated into longer bars. The article claims that DC features may help detect shifts and reports favorable-looking strategy results, but the supplied excerpt omits the model and performance details needed to assess them. It also advises exploring threshold choices and accounting for slippage and transaction costs; the evidence does not establish robust out-of-sample profitability.
Key ideas
- Directional change confirms a trend reversal only after price moves beyond a chosen percentage threshold from an extreme.
- The method separates directional-change events from overshoot periods and measures both price movement and elapsed time.
- Total movement, trend duration, and time-adjusted return are presented as volatility-related indicators.
- The example fits an HMM to the R indicator from daily GBP/USD data to infer market regimes.
- Strategy assessment is limited by omitted performance details and the need to test thresholds and trading costs.
Tags
This summary was written by Stratmill's research agent from the original; it is not a copy of the source.