Combining Volatility Adjusted Momentum with a Conditional LSTM

Summary

The article presents Volatility Adjusted Momentum (VAM), which scales price momentum by a volatility estimate and the square root of the momentum lookback. Positive and negative readings are treated as directional signals, with magnitude representing volatility adjusted strength. It then describes an automated strategy that combines VAM with price forecasts from a Conditional LSTM trained on historical prices and technical features, including MACD. The trading logic takes long or short positions when the indicator and forecast agree, and uses ATR based stop loss and take profit distances.

The workflow includes collecting and normalizing inputs, training a model outside MetaTrader, exporting it to ONNX, and running predictions in an Expert Advisor. The article says that backtesting and optimization were performed and claims better outcomes for the combined approach, but the supplied excerpt provides no specific performance figures or detailed test design. It also acknowledges overfitting, data quality, interpretability, and computational demands as concerns. The stated results should therefore be treated cautiously, since the excerpt does not establish robustness across instruments or unseen market regimes.

Key ideas

• VAM scales price change by estimated volatility and the square root of its lookback period.
• The proposed EA combines VAM direction with Conditional LSTM price forecasts to generate trades.
• ATR is used to set stop loss and take profit distances that respond to market movement.
• The model is trained externally, exported to ONNX, and used for inference in MetaTrader.
• The article reports improved combined backtest outcomes but gives no detailed metrics in the excerpt.

Tags

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