Low-Frequency Fourier Components with Moving-Average Trend Signals

Summary

This trend-following approach adds selected low-frequency Fourier components to price and applies fast, medium, and slow moving averages. The stated defaults are 5, 20, and 200 periods. It enters long when the fast average is above the medium average and price is above the slow average; it enters short under the inverse conditions. The source uses individual sinusoidal components from a Fourier calculation, then applies configurable moving-average types and an optional linear regression transformation.

The document presents noise filtering and medium- to long-term trend capture as intended benefits, but it supplies no performance statistics to substantiate them. Published settings describe a short BTC/USDT futures backtest on one-minute bars. The prose describes crossover triggers, while the code checks whether one average is above or below another, without requiring a fresh crossing. Sudden reversals and choppy markets are cited as risks; stop losses and trend-strength filters are proposed as possible additions.

Key ideas

• Selected Fourier sinusoidal components are added to price before moving-average calculations.
• Long and short conditions combine fast-versus-medium average ordering with price relative to the slow average.
• The source allows different moving-average types and optional linear regression smoothing.
• The stated risks include sudden reversals and repeated signals in sideways markets.
• The published backtest settings contain no performance results.

Tags

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