Power Prices -

. .

Résumé :
Over the last decade, European electricity markets have experienced rapid deregulation and liberalisation. This has caused electricity prices to become highly volatile and exhibiting very distinct characteristics such as strong seasonal patterns, extreme spikes and mean-reverting behaviour. In this work the electricity spot and futures prices observed at the European Energy Exchange are analysed. A general multi-factor affine diffusion model combined with a finite q-state Markov regime-switching process is presented which incorporates the stylised features of both spot and futures prices. The model is estimated using a maximum likelihood approach where the likelihood function is evaluated applying Kalman/Kim filter techniques. In order to quantify the seasonal patterns, regression models are developped. Their results provide the seasonal structure for the price model and also build the basis for the construction of an hourly price forward curve. The models? performance is assessed comparing historical and model implied price characteristics.

Biographie:
Dr. Lea Bl?chlinger works as a power and gas analyst at LDHEnergy. Previously, she worked at Lehman Brothers. She studiedeconomics and completed a doctorate at the University of St.Gallen. During her doctorate, she was a research fellow at theInstitute for Operations Research and Computational Finance(ior/cf-HSG).

Sommaire:
Over the last decade, European electricity markets have experienced rapid deregulation and liberalisation. This has caused electricity prices to become highly volatile and exhibiting very distinct characteristics such as strong seasonal patterns, extreme spikes and mean-reverting behaviour. In this work the electricity spot and futures prices observed at the European Energy Exchange are analysed. A general multi-factor affine diffusion model combined with a finite q-state Markov regime-switching process is presented which incorporates the stylised features of both spot and futures prices. The model is estimated using a maximum likelihood approach where the likelihood function is evaluated applying Kalman/Kim filter techniques. In order to quantify the seasonal patterns, regression models are developped. Their results provide the seasonal structure for the price model and also build the basis for the construction of an hourly price forward curve. The models' performance is assessed comparing historical and model implied price characteristics....