Graph Neural Networks for Liquidity Zone Recognition

Summary

This article describes a workflow for recognizing potential liquidity zones with a Graph Neural Network. It represents detected swing highs and lows as graph nodes, uses price, tick volume, and swing type as node features, and links neighboring swings. Labels are assigned based on repeated nearby swing prices. A trained model then predicts which structural levels may have liquidity, with an MQL5 Expert Advisor collecting live data, constructing the graph, running ONNX inference, applying basic trade checks, and handling execution.

The article provides a technical pipeline from historical data collection through model training and platform integration, along with sample label counts and graph dimensions. Those details illustrate implementation, but do not establish predictive quality or profitable trading. The labeling rule treats repeated prices within a fixed distance as liquidity, which is only a proxy for actual resting orders. The document gives no clear out-of-sample performance evidence, and its trade interpretation of predicted liquidity requires independent validation.

Key ideas

• Swing highs and lows are represented as graph nodes with price, volume, and swing-type features.
• The example links neighboring swing points and labels repeated nearby prices as potential liquidity zones.
• A trained GNN produces predictions that an MQL5 Expert Advisor can use for trade decisions.
• The model is exported to ONNX for inference within the trading platform.
• Repeated price levels are a proxy for liquidity, and the article does not establish out-of-sample profitability.

Tags

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