GeonatIQ, a specialist in AI for the energy, resources, and climate sectors, has launched a carbon market forecasting engine to help traders, corporates, and policymakers navigate the €781 billion (~$907 billion) EU Emissions Trading System (EU ETS) — the world’s largest carbon market.
With BloombergNEF projecting prices to rise from €70 to €149 ($81 to $173) per metric ton by 2030, and ETS Phase II expected to channel €705 billion into climate finance, demand for accurate forecasting is growing.
Yet nearly half of allowance holders still don’t actively trade, highlighting the need for data-driven tools that turn market complexity into actionable insight.
The system offers structured analytics and scenario planning across 1–9 month horizons. In backtesting, it achieved a Sharpe ratio above 3 for three-month-ahead forecasts, signalling strong risk-adjusted performance.
“Carbon pricing is too strategic to rely on intuition,” said Yassine Charouif, AI Solutions Architect at GeonatIQ. “Our system is designed to decode the market’s structure and adapt dynamically – helping traders, corporates, and policymakers engage with forward-looking insights more effectively.”
Backed by Academic Research
The engine’s architecture builds on a collaboration with Imperial College London, co-authored by GeonatIQ founder Antony Sommerfeld, Charouif, and climate scientist Professor Yves Plancherel.
The resulting study — the most downloaded in its category — highlights the strength of GeonatIQ’s meta-learning approach compared with traditional statistical models.
Sommerfeld said the system combines deep learning, signal decomposition, and adaptive training to “turn complex carbon signals into clear, actionable intelligence,” aiming to help passive allowance holders become informed, active participants.
Relevant: Carbon Credit Developers Urge Overhaul Of Complaints System
In Q4, GeonatIQ will debut its “Carbon LLM” — a large-scale AI model trained to integrate regulatory news, energy events, and macroeconomic data into short-term forecasts. Tests show up to a 42% accuracy boost in short-term modelling, enabling early detection of market shifts driven by policy or macroeconomic changes.
“We’ve built a kind of crystal ball here for the carbon market – a system that can simulate news stories before they happen and predict how the market is likely to respond,” said Mike Schuil. “It’s not just a forecasting tool, it’s a strategic advantage.”
Read more: EU Could Boost Climate Funding Tenfold By Charging Airlines More, New Study Finds
