Combining Triangular FX Arbitrage with Deep-Learning Price Predictions

Summary

The article outlines a proposed foreign-exchange triangular arbitrage system that combines three currency pairs with neural-network forecasts. It describes the basic conversion cycle and the idea of comparing implied exchange rates, then presents a workflow using Python to train price-prediction models, export them to ONNX, and use them from an MT5 Expert Advisor. The model example uses historical close prices and a convolutional and LSTM network; the article also discusses spreads and gives backtest statistics in the portion provided.

The text claims that predictions can help address trading when spreads make immediate arbitrage difficult, but it does not establish that the opportunities are risk-free. Forecast accuracy, synchronized quotes, execution latency, slippage, fees, and live fills all affect whether a cycle can be profitable. The example’s backtest figures are the author’s reported results, not independent validation, and the supplied document is truncated. Price prediction by itself does not demonstrate a durable arbitrage edge.

Key ideas

• A triangular conversion cycle combines exchange rates across three currency pairs.
• The proposed system uses neural-network price forecasts exported to ONNX for an MT5 EA.
• The example workflow trains on historical prices and evaluates predictions on held-out data.
• Spreads and execution conditions can erase apparent triangular arbitrage opportunities.
• The reported backtest results are not independent evidence of live profitability.

Tags

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