Modelowanie i prognozowanie cen energii elektrycznej na potrzeby zarządzania 
ryzykiem na rynkach energii  
[Modeling and forecasting electricity prices for risk management in power markets]

Joanna Janczura (Department of Applied Mathematics, PWr, Wrocław)


This presentation provides an overview of quantitative methods for modeling, 
forecasting, and managing risk tailored to the unique characteristics of electricity 
markets. The methodology spans the full market horizon—from forward and day-ahead to 
real-time balancing—and supports decision-making under both market and operational 
risk. First, I will discuss Markov regime-switching models, which capture price 
spikes and drops and enable the valuation of electricity derivatives through 
risk-premium estimation. I will then introduce the Expectile Regression Averaging 
(ERA) method for probabilistic forecasting, which enhances accuracy and supports the 
computation of Expectile Value at Risk (EVaR). Next, I will demonstrate how 
probabilistic forecasts can optimize portfolio allocation between day-ahead and 
intraday markets. Finally, I will present a framework that jointly incorporates 
price and volume risks to quantify real balancing costs. All results will be 
illustrated with market data.