[TSC, Mo 15:45] Trading Strategies in Commodity Markets

1.Modelling Trading Strategies for the Investigation of Price Developments in the European Electricity Market (Mirbach, Schwarz, Moser)

The market price for electrical energy is one of the main decisive factors for operational and strategic questions of power generation companies, e.g. for the evaluation of the economic profitability of existing and new power plants as well as for short-term up to long-term trading decisions. Thus, the investigation of the price development is a basic issue in the liberalized electricity sector. Therefore, it is essential to consider the relevant influencing factors of the market price, which are significantly affected by the current developments in the market for electrical energy. The influencing factors comprise the development of the power generation park, power generation costs, political regulations as well as general conditions for the cross-border congestion management. Also, due to new trading opportunities for electrical energy and the intensified competition, the behaviour of power generation companies, which originally minimized their power generation costs, has changed to an individual profit maximization using all market potentials.

Market simulation methods, which simulate the market for electrical energy, enable to consider these influencing factors on the market price for electrical energy. Classical market simulation methods for power generation planning take into account the technical constraints of power generation and transmission. In addition, methods based on the decision theory appear to be appropriate to investigate the profit maximization of market participants under consideration of the price matching. In consequence, the market price for electrical energy can be investigated by means of a competitive market simulation method taking into account the technical constraints of power generation and transmission as well as the impact of the profit maximization by market participants. The advantage of a competitive market simulation method is the potential to consider explicitly mentioned above structural changes that effect the market price.

Considering these circumstances, this approach presents a competitive market simulation method consisting of power generation planning, strategy planning and cross-border price matching (market coupling) for the investigation of price developments in the European electricity market. Based on the hourly cost curves for power generation determined by the power generation planning, different bidding strategies can be applied for each market participant. The considered bidding strategies in the market simulation method comprise price taker, physical withholding and economic withholding. Finally, a market coupling which includes a cross-border matching is realized based on the hourly bids by maximizing the social welfare under consideration of the energy balance and the transmission network. The main results of the market simulation method are the hourly market price and the quantity matched for each market area and country.

The results of ex post simulations of the European electricity market using the competitive market simulation method show that strategic bidding behavior of market participants can lead to higher market prices. However, different investigations point out the high sensitivity of the market participants’ profit against uncertainties regarding bidding strategies of competitors and other market data. Further investigations considering imperfect information of the market participants based on a stochastic optimization show that bidding strategies highly depend on the assumption of the residual market. Therefore, exercising withholding strategies is combined with very high risks regarding the contribution margin for the respective market participant. The results show that especially market participants who apply physical withholding strategies have a financial disadvantage compared to other market participants.

2.Efficiency in the Crude Oil Market: Backtesting Recent Developments with Multifactor Models (Fritz)