Combined estimation and prediction of the Hourly Price Forward Curve

In deregulated electricity markets, market participants use the Hourly Price Forward Curve as a powerful tool to evaluate long term electricity contracts, derivatives and risk measures. In this paper a new framework to calculate the Hourly Price Forward Curve as a combined problem of parameter estimation and future price prediction on hourly basis is developed. The combined calculation results in higher robustness of the prediction, because the estimated parameters satisfy the future time series prediction equation. An estimator that is robust based on the least absolute deviation and the model selection framework Least Absolute Deviation with Least Absolute Shrinkage and Selection Operator is presented. The problem is implemented as a linear program. As a testing set, German electricity spot prices and German Future products are used. Discussion on the new framework and its advantages is given in the end.