An Example of Spurious Granger Causality in Electricity Market Data

This paper investigates the validity of Granger causality in the context of the Croatian power system by analyzing time series data for electricity production from gas and renewable sources. We assess the correlations between production variables and market prices using data from the ENTSO-E Transparency Platform and the Croatian Transmission System Operator. A Vector Autoregression (VAR) model is applied to evaluate the causal relationships among these variables. Our findings suggest that the observed Granger causality is spurious, indicating potential flaws in relying solely on statistical causality tests without understanding the underlying power system organizational dynamics. Future research should focus on examining other components and regions to validate these findings and refine causality detection methods.