A coal power plant yields a natural power long position and a coal short position in an energy company’s portfolio (whether the plant is initially treated as a real option or not). To reduce the price risk energy companies usually unwind these positions by selling the power and acquiring the coal on forward markets. Given the illiquidity of power and commodity markets, it may take several months or even years before such large positions can be unwinded. We present an approach that allows to determine dynamic liquidation strategies providing the optimal balance between minimizing liquidation costs and minimizing the open position’s risk.

For the derivation of optimal portfolio transactions with liquidity costs we assume that the amount sold has a temporary impact on the realised price. We allow for general dynamics of the plant’s profit margin, thus abandoning the random walk paradigm from the known liquidation literature for large stock positions (see, e.g. Almgren and Chriss, 2000, or Ting et. al., 2006). Based on stochastic dynamic programming we show how to deduce liquidation / acquisition strategies that maximise the utility of the company’s proceeds from selling power and buying the required coal. The control function maximised may be arbitrary, but for computational reasons we suppose that the company’s risk preferences are determined by a multiplicative or additive utility function.

Finally, we study the implications of our model calibrated to empirical price data. The results clearly show that the liquidation speed strongly depends on the company’s risk aversion. Besides, we discuss the influence of liquidity on the shape of liquidation strategies. In particular, we show that lower liquidity leads to spread the transactions more evenly over the liquidation period. Furthermore we study the impact of directional views on the liquidation strategies. If the company thinks that markets are undervalued, then it will considerably slow down transactions in the beginning and accelerate them shortly before maturity. This effect will be the stronger, the more liquid the markets are.

The liberalization of the European energy market now enables big gas customers and public utilities to build a portfolio of different gas supplies und purchase contracts. The covering of the gas demand, which is heavily temperature dependent, can be optimized by combining base load contracts, open gas delivery contracts, and the use of the capacity of gas underground storages and local pipe storages. We present a Two Stage Stochastic Linear Programming Model for the optimisation of the gas purchase under uncertain demand while considering the cost of underground storage capacities and transportation. Furthermore, we enhanced the model to explicitly consider the risk measure cVaR. The model is integrated into the decision support system SAPHIR which contains modules for data management, scenario generation, portfolio optimization and risk analysis for strategic and operational gas purchase portfolio planning. We evaluate our approach based on a real world case study and will report on our experiences including computational results.

In Germany, the energy liberalization started in 1998 after implementing the European Directive 96/92/EG which opens the electricity market for competition by regulating network access and allowing free choice of energy suppliers for final consumers. In 2001, the EEX was established by merging the German power exchangee LPX and EEX to the EEX AG. Nowadays, the national energy markets operate as several intertwined but independent markets at which electricity is traded. In the center of the power exchange lies the daily day-ahead auction for hourly and block contracts, scheduled at 12 noon. Additional to the power exchange, markets for balancing power (spinning reserves) are established which asses the balance of the grid load within a certain market area. While for generators, energy and transmission facilities are complementary goods, for selling energy, the suppliers have several possibilities as for example futures or forward markets, spot markets or balancing power (reserve markets), which form for the bidder a portfolio of substitutable markets. Hence, in this article we build up an optimal trading strategy for bidding at the balancing power market and the auction for day-ahead hour contracts.