Please use this identifier to cite or link to this item:
Pape, Christian
Woll, Oliver
Weber, Christoph
Year of Publication: 
Series/Report no.: 
HEMF Working Paper 04/2018
Practitioners in the electricity industry aim to assess the value of power plants or other real options several months or even years ahead of operation. Such a valuation is notably required for hedging purposes. The revenue streams to be earned in the spot market are thereby already secured on future markets. Yet the peculiarities of the electricity market, notably the limited storability of electricity and the incompleteness of the derivative markets, make this problem also theoretically challenging since they prevent the straightforward application of standard approaches for price modeling and for hedging. In this context, the contribution of this article is twofold: (1) We present a novel methodology to model electricity prices based on fundamental expectations and accounting for both short-term and long-term uncertainties. This requires the joint modeling of different commodity prices, namely electricity, fuel and CO2 prices. Moreover price distributions have to be modelled in order to assess the real option value adequately ex ante. Specifically, we compare two different modeling approaches to account for long-term variations in multi-commodity price dynamics. (2) We suggest a test procedure and introduce performance measures to analyze the accuracy of the proposed price modeling. We thereby focus on the practically relevant question, whether the price modeling provides ex ante estimates of the value of the real option that are in line with the ex post realized values. This approach is chosen since no derivative markets exist where the (extrinsic) values for the real options could be observed months or years ahead of actual operation. Nonetheless we show that under well-defined assumptions, the ex-ante values derived using the price model should provide unbiased estimates of the ex post values, which are computed as a sum of hedging and spot exercise revenues. The application part shows results for a state-of-the-art gas power plant. By applying the developed performance measures and test statistics, we find that neither of the two investigated price models clearly outperforms the other.
Electricity price forecasting
Futures market
Real option
Stochastic optimization
Document Type: 
Working Paper

Files in This Item:
2.37 MB

Items in EconStor are protected by copyright, with all rights reserved, unless otherwise indicated.