Testing Bill Williams Indicators with ADX, ATR, and Neural Predictions

Summary

This article combines Bill Williams’ Alligator, Awesome Oscillator, and Fractals into trend-following entry and exit rules. It then describes adding an ADX filter for trend strength, ATR-based stops, and neural-network forecasts. The indicator rules use Alligator line order to identify direction, AO sign and change to gauge momentum, and fractals as an additional entry condition. The model’s predictions use a daily horizon while the strategy tests on an eight-hour timeframe; the author says the model should be refreshed as new data arrives.

The article reports that the indicator-only and ADX variants performed poorly in the shown tests, while an optimized prediction-assisted version had positive reported metrics, including a Sharpe ratio of 3.52. It also cites high model fit and correlation statistics, but these alone do not demonstrate tradable out-of-sample performance. The author notes the EA is incomplete and that results depend on timeframe, optimization, and model updates.

Key ideas

• The core strategy combines Alligator direction, AO momentum, and fractal signals for entries and exits.
• ADX is added as a trend filter, while ATR is used to set stops.
• The article pairs eight-hour strategy testing with daily neural-network predictions.
• The author reports weak results for simpler variants and stronger metrics for an optimized prediction-assisted version.
• Reported model-fit statistics and test results do not by themselves establish robust live performance, and the EA is incomplete.

Tags

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