Forecasting the Impact of Extreme Weather Events on Electricity Prices in Italy: A GARCH-MIDAS Approach with Enhanced Variable Selection

This paper examines the impact of extreme weather events on electricity price volatility in Italy, using a novel combination of advanced econometric techniques and a robust variable selection process. A key feature of the study is the application of the Best Path Algorithm (BPA) for variable selection, which identifies the most relevant predictors, with extreme weather events emerging as the primary drivers of price volatility. These selected variables are incorporated into a GARCH-MIDAS model, allowing for the integration of high-frequency electricity price data with low-frequency climate data to capture both short- and long-term volatility components. Additionally, the study incorporates external shocks, such as the Russia-Ukraine war, as exogenous variables to account for their effects on the energy market. The results highlight the significant predictive power of extreme weather events and external factors on returns of electricity prices. This approach provides policymakers and energy stakeholders with improved forecasting tools, emphasizing the need for resilience in energy market planning. Future research may extend this methodology to other regions and incorporate additional variables to enhance predictive accuracy.