Gaussian Smoothed Detrended Oscillator Reversal Strategy
Summary
This strategy uses a detrended price oscillator, smoothed with an Arnaud Legoux moving average, to generate mean reversion signals. It compares the smoothed oscillator with a lagged copy: a crossover while the oscillator is below zero opens a long, while a crossunder above zero opens a short. Positions close when the oscillator crosses its lagged series in the opposite direction or crosses the zero line. The article also describes plotting the oscillator, lag, zero line, and signal markers for visual monitoring.
The published settings include a price length of 50, smoothing length of 50, and lag of 25, with a Bitcoin futures backtest configuration. However, the text provides no performance results, comparative evidence, or transaction cost analysis. It warns that parameter choices affect signal frequency and that persistent trends can cause repeated losses for a reversal approach. Suggested safeguards include trend filters, stop losses, confirmation signals, and testing across timeframes. These are possible refinements, not demonstrated improvements, so the strategy requires independent validation before use.
Key ideas
- The strategy detrends price with an EMA based oscillator and smooths it with ALMA.
- Lagged oscillator crossovers define entries, with the zero line helping establish direction and exits.
- The described rules target reversals and may struggle when trends persist.
- Parameter sensitivity and false signals make independent backtesting important.
- Trend filters and stop mechanisms are proposed as safeguards but are not validated in the document.
Tags
This summary was written by Stratmill's research agent from the original; it is not a copy of the source.