Mapping price dynamics across electricity market designs: A functional data approach with STL decomposition

This study develops a functional data analysis (FDA) framework to investigate how electricity price dynamics respond to changes in energy generation across countries with different market structures. Using Seasonal-Trend decomposition using Loess (STL) decomposition, electricity prices are separated into trend, seasonal, and irregular components to distinguish market-driven cycles from policy-driven signals. Functional principal component analysis (FPCA) is applied to extract shared and country-specific seasonal features, while multivariate functional response analysis (FRA) quantifies the time-varying influence of energy generation on price. The analysis covers six countries, Germany, the United Kingdom, the United States, France, China, and Brazil, which represent a spectrum of liberalized, semi-regulated, and regionally fragmented electricity markets. The results reveal that renewables in liberalized systems drive immediate price volatility, while policy coordination in semi-regulated systems leads to delayed and smoothed price adjustments. By linking institutional design with functional elasticity patterns, the study offers a unified approach to evaluating the interplay between price signals and energy generation in diverse regulatory contexts.