A1517
Title: Non-parametric causal discovery for EU allowances returns through the information imbalance
Authors: Cristiano Salvagnin - University of Insubria (Italy) [presenting]
Aldo Glielmo - Bank of Italy (Italy)
Vittorio Del Tatto - SISSA (Italy)
Antonietta Mira - University of Lugano (Switzerland)
Maria Elena De Giuli - University of Pavia (Italy)
Abstract: A new non-parametric method called Differentiable Information Imbalance is applied to identify variables that are causally linked, possibly through complex and non-linear relationships, to the financial returns of European Union Allowances in the European Union Emissions Trading System. Using data from January 2013 to April 2024, we compare this approach with the traditional multivariate Granger causality method based on vector autoregressive models. Both methods identify key drivers such as coal futures prices and the Spanish stock index IBEX 35, but they also reveal notable differences. Through experiments with synthetic data, we show that these differences likely arise from the linear assumptions underlying standard Granger causality.
