Network Momentum: Learning Lead-Lag Links Across Markets

Summary

This article describes a cross-market trend-following method that combines each market’s own momentum with momentum from markets that tend to lead it. It first scales price changes by estimated volatility, then forms moving-average crossover oscillators at several speeds. Derivative Dynamic Time Warping is used to estimate pairwise lead-lag relationships from the shape of market movements, and a graph-learning objective turns those relationships into a weighted network. Normalized network links propagate momentum into a signal for each market; its sign determines long, short, or flat direction.

The article implements the pipeline in MQL5 and discusses learning the graph without an external optimization solver. It describes a Strategy Tester run as an implementation check, while identifying broader multi-year universe testing, volatility-target sizing, and demo forward testing as next steps. The presentation adapts research originally framed around commodity futures to a retail MetaTrader basket of FX and cross-asset instruments. It explains the design in detail, but the supplied text does not establish out-of-sample profitability or robustness across markets and periods.

Key ideas

• The method combines a market’s own trend with momentum propagated from markets that lead it.
• Volatility scaling makes price changes more comparable across markets before oscillator calculation.
• Derivative Dynamic Time Warping estimates lead-lag timing from the shape of price movements.
• A graph-learning objective converts pairwise lag estimates into weighted connections used to propagate signals.
• The article describes a working implementation but identifies broad historical and forward testing as unfinished work.

Tags

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