Kalman-Smoothed Moving Average Trend Signals from Cumulative Angles

Summary

This trend-following method applies a recursive Kalman-style filter to a moving average, then measures the angle of change in the smoothed series. It sums those angles over a chosen window and signals long exposure when the total exceeds a positive threshold, or short exposure when it falls below the corresponding negative threshold. The article presents smoothing as a way to reduce price noise and make medium- to long-term trend direction easier to detect.

The supplied implementation uses a one-minute price series, a filter parameter of 0.01, a 30-bar angle window, and thresholds of 360 degrees in either direction. Its published test configuration concerns BTC/USDT futures over one week in January 2024, but no performance figures are reported. The explanation’s claims about signal reliability are not supported by comparative results. The angle calculation depends on raw price differences, so its scale and interpretation may vary with instrument price and sampling; parameter tuning, reversals, and overfitting are also identified as concerns.

Key ideas

• The method smooths a moving average with a recursive Kalman-style filter.
• It sums changes in the smoothed line’s angle over a configurable window.
• Positive and negative angle thresholds trigger long and short signals, respectively.
• The example uses a one-minute series and a 30-bar angle window.
• No performance evidence is given, and thresholds may be scale-sensitive.

Tags

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