Modelling Energy Markets with Extreme Spikes

This paper suggests a new approach to model spot prices of electricity. It uses a shot-noise model to capture extreme spikes typically arising in electricity markets. Moreover, the model easily accounts for seasonality and mean reversion. We compute futures prices in closed form and show that the resulting shapes capture a large variety of typically observed term structures. For statistical purposes we show how to use the EM-algorithm. An estimation on spot price data from the European Energy Exchange illustrate the applicability of the model.