Multifractal Analysis of the Brazilian Electricity Market

• Published in: IEEE access Vol. 11; pp. 98939 - 98957
• Main Authors: Castro, Alessandro L., Marcato, Andre Luis Marques, De Aguiar, Eduardo P.
• Format: Journal Article
• Language: English
• Published: Piscataway IEEE 2023; The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
• Subjects: anti-persistence; Behavioral sciences; Brazilian electricity market; Efficiency; Electricity; Electricity pricing; Electricity supply industry; Fluctuations; Fractal analysis; Fractals; generalized hurst exponent; Global marketing; market efficiency; Mathematical analysis; multifractal detrended fluctuation analysis; Optimization models; Prices; Pricing; singularity spectrum; South America; Time series analysis; Uncertainty
• ISSN: 2169-3536; 2169-3536
• DOI: 10.1109/ACCESS.2023.3313099

In Brazil's wholesale electricity market, long-term contract prices are negotiated between power generators and large consumers. Unlike traditional markets, pricing is not driven by market forces but rather determined by complex computational models known as Hydrothermal Dispatch Optimization Models. These models calculate the Difference Settlement Price (DSP), serving as the short-term market price for electricity. The Brazilian market is divided into four interconnected submarkets: the Southeast, Northeast, North, and South. This study fills an existing research gap by examining the multifractality of these submarkets by applying Multifractal Detrended Fluctuation Analysis over a deseasonalized price return time series. Specifically, it aims to characterize the multifractal features of electricity prices, identify the underlying causes of this multifractality, and assess market efficiency indices over time. Our analysis of historical electricity prices revealed that all submarkets demonstrated anti-persistent behavior-also known as mean-reversion-and multifractality. This finding aligns with similar observations in global markets. The South submarket displayed the highest level of multifractality and the lowest market efficiency. Conversely, the North submarket had the lowest multifractality and the highest efficiency. Through sliding-window analysis, we investigated temporal variations in the Hurst exponent and Long Memory Magnitude, an index to compute market inefficiency. We found consistent anti-persistent behavior across all submarkets, with the South submarket showing greater volatility in its inefficiency index. While preliminary and requiring further in-depth analysis and consideration of other factors, these findings offer valuable insights for decision-makers and regulators pursuing new market arrangements to boost efficiency.