Fractional EMA and Adaptive Kalman Filters for Low-Lag Trend Signals

Summary

This indicator combines a cascaded fractional EMA with two adaptive Kalman filters to create a base line and a faster line. Setting EMA lengths below one makes the EMA extrapolate rather than smooth, reducing apparent lag while amplifying price fluctuations. The Kalman stage estimates noise from residuals normalized by ATR, using a longer average for measurement noise and a shorter average for process noise. Its gain changes with these estimates, making the filters more responsive when movement appears persistent and less responsive when noise dominates. A colored band between the two lines marks their relative position and can be read like an adaptive moving-average crossover.

The document suggests using color changes for direction, band width for acceleration or weakening, and the base line’s slope for context. It also outlines pullback and crossover applications and lists adjustable inputs. These are indicator interpretations, not demonstrated trading results: no backtest, market comparison, or execution analysis is provided. The anticipative input can magnify noise, and the described signals require independent validation before use.

Key ideas

• A fractional EMA with length below one extrapolates price and can amplify fluctuations.
• Adaptive Kalman filters adjust their response using volatility-normalized residual estimates.
• The faster filter above the base filter marks the indicator’s upward regime, with the reverse marking downward pressure.
• Band width and base-line slope provide context for momentum and possible trend weakening.
• The document describes uses but provides no backtest or performance evidence.

Tags

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