Forecasting EU ETS carbon prices using information on CCU/CDR technologies’ investments

We examine the predictive relevance of an investment-linked proxy capturing market attention toward carbon management technologies, including Carbon Capture Utilization and Carbon Dioxide Removal (CCU/CDR) technologies, for EU ETS carbon prices. Building on evidence that EU ETS prices affect investment, this study examines the reverse relationship by testing whether carbon management investment signals can inform price forecasting. Empirically, we demonstrate that incorporating this proxy into the dataset improves forecast accuracy across statistical, machine learning (ML), and deep learning models. Methodologically, we implement a systematic benchmarking framework and hybrid stacking ensembles, and apply a structured predictor-screening workflow with explicit caveats to ensure transparency regarding measurement uncertainty in the proxy. From a policy perspective, our results provide actionable insights for market participants and regulators seeking to anticipate carbon price dynamics. Future research could explore alternative proxies, higher-frequency data, or causal identification strategies to further enhance forecasting and deepen understanding of carbon market behavior.