Member-only story
From Academic Paper to Trading Bot: Stochastic Input–Output Networks as a Multi-Asset Mean Reversion Strategy
Javier Santiago Gastón de Iriarte Cabrera14 min read·Just now--
How a 2026 Mathematical Finance paper on supply shocks and sector resilience becomes a fully functional MetaTrader 5 Expert Advisor — with code, diagrams, and honest backtest results.
Based on: Amici, Fusai, Gambaro & Marazzina (2026) — Mathematical Finance, DOI: 10.1111/mafi.70029
Most algorithmic trading strategies live and die in a self-referential world: price action, RSI, MACD, some flavour of machine learning trained on OHLC bars. They work (when they work) by pattern-matching past market behaviour, with no theory connecting the signal to why it should exist.
What if, instead, we built a trading system grounded in academic macroeconomics — specifically, in how supply shocks propagate through industrial networks, and how different sectors “recover” (mean-revert) after those shocks?
That’s exactly what this article explores. We take a freshly published paper from Mathematical Finance, extract its core mechanism, and translate it into a working MT5 Expert Advisor for gold, oil, and natural gas.
“By embedding a linear stochastic fluid network into the production network of interdependent industries, the model captures how physical shocks propagate through IO linkages and affect sectoral price dynamics.” — Amici et al., 2026
