Price determinants in the German Intraday Power market
Translated title
Fundamentale variable der driver det tyske Intraday marked
Author
Michno, Peter
Term
4. term
Education
Publication year
2025
Pages
59
Abstract
Denne afhandling undersøger, hvorfor der opstår en prisforskel mellem det tyske day-ahead-marked (dag-forud) og intradag-markedet (inden for dagen) for el. Ved hjælp af timespriser fra EPEX SPOT og relevante fundamentale data fra 2023-2024 analyseres, hvordan prognosefejl i efterspørgsel (load) og i produktion fra vind og sol, uforudsete produktionsnedbrud samt grænseoverskridende strømhandler påvirker prisforskellen. Analysen anvender økonometriske modeller (OLS, ARX og GARCH) for at vurdere sammenhængene. Resultaterne viser, at prognosefejl for vind og sol er meget vigtige for at forklare forskellen: Når der produceres mindre end forventet, bliver intradagpriserne højere end day-ahead, mens overproduktion presser dem ned. De fleste resultater stemmer med teorien, men fejl i efterspørgselsprognoser viser en uventet negativ sammenhæng: Ekstra efterspørgsel er forbundet med lavere intradagpriser end forventet. Blandt forbindelserne til nabolande er DE ↔ FR-interconnectoren den mest statistisk betydningsfulde, hvilket afspejler dens centrale rolle i balancering og markedsintegration. GARCH-modellen giver yderligere indsigt i volatilitet og peger på, at variabler som residual load og visse interconnectorer får større betydning under perioder med markedsuro. Analysen finder også klare mønstre over døgnet og ugen: Om natten er kun vindprognosefejl betydende, midt på dagen dominerer solafvigelser, i midtpeak-timer er efterspørgsel og nettoeksport vigtigst, og nedbrud rammer især aftenpriserne. På ugebasis er mandag mest følsom over for prognosefejl, hvilket peger på øget usikkerhed efter weekenden.
This thesis examines why there is a price difference between Germany’s day-ahead and intraday electricity markets. Using hourly prices from EPEX SPOT and relevant fundamental data from 2023–2024, it analyzes how forecast errors in demand (load) and renewable generation (wind and solar), unplanned generation outages, and cross-border electricity flows shape that price spread. The study applies econometric models (OLS, ARX, and GARCH) to assess these relationships. The findings show that wind and solar forecast errors are highly important: when actual generation falls short of forecasts, intraday prices are higher than day-ahead; when there is overproduction, intraday prices move lower. Most results align with theory, but load forecast errors display an unexpected negative relationship, meaning excess demand is associated with lower intraday prices than anticipated. Among interconnectors, the DE ↔ FR link stands out as the most statistically significant, reflecting its central role in cross-border balancing and market integration. The GARCH model adds insight into volatility, indicating that variables such as residual load and some interconnectors become more influential during periods of market stress. Clear patterns also appear across the day and week: at night, only wind forecast errors matter; around midday, solar deviations dominate; during mid-peak hours, load and net exports are most relevant; and outages mainly affect evening prices. On a weekly level, Mondays show the strongest sensitivity to forecast errors, suggesting elevated uncertainty after weekends.
[This summary has been rewritten with the help of AI based on the project's original abstract]
Documents