Ehlers Autocorrelation Periodogram for Dominant Cycle Estimation
Summary
This indicator estimates a dominant market cycle by first applying a roofing filter to the selected price or input series. The filter combines a high-pass stage, available in first- or second-order form, with a SuperSmoother low-pass stage to limit unwanted trends and high-frequency noise. It then computes autocorrelations across candidate lags, converts those correlations into Fourier spectral power across periods from 6 to 49 bars, and optionally smooths the power estimates with a SuperSmoother or an exponential average.
The strongest spectral components are normalized relative to peak power, raised to a contrast setting, and filtered by a power threshold. The surviving components contribute to a weighted dominant-period estimate, while a periodogram display provides a visual view of relative power. Parameters control filtering, lag, smoothing, contrast, and threshold. This is an adaptive cycle measurement rather than a directional entry strategy; dominant-cycle estimates can be unstable when the data lacks a clear cycle or spectral power is diffuse. The document describes the algorithm and controls but supplies no empirical performance evidence.
Key ideas
- A roofing filter conditions the input series before cycle estimation by combining high-pass and low-pass filtering.
- Autocorrelation values are decomposed into Fourier power across candidate cycle periods.
- Optional spectral smoothing and a power threshold reduce noisy or weak components.
- The dominant period is estimated as a power-weighted average of qualifying components.
- The output measures cycle length rather than trade direction, and the description includes no performance validation.
Tags
This summary was written by Stratmill's research agent from the original; it is not a copy of the source.