Estimation of Supply Function Equilibria (sfe) for the German Electricity Generation Market – Simulation of Market Structures and Market Power

Rising prices in the German wholesale electricity market have spurred discussions about the potentials and the abuse of market power. Particularly since the introduction of emission certificates, electricity prices increased continuously closely related to rising certificate prices. The oligopolistic structure of the German market with four major players owning 90% of the capacities (Bundeskartellamt, 2006) leads to the presumption of significant market power potentials. A more detailed look at the market structure reveals that E.ON and RWE already own up to 60% of generation capacities. This market structure creates incentives for strategic behavior to distort competitive price-quantity outcomes. In order to simulate the market structure and estimate market power potentials the supply function equilibria approach (SFE) is for the first time used for the German electricity generation market. The underlying theory has been developed by Klemperer and Meyer (1989). It is generally considered to be a more appropriate modeling of market power than traditional Cournot or Bertrand oligopolies, because firms in electricity markets often compete by offering a schedule of prices and quantities rather than using only quantity or price as a strategic variable. The SFE approach explicitly considers demand uncertainties, typical for the electricity sector where bids are settled before demand is realized. In result, one obtains optimal supply schedules for the bidding firms. A drawback of those models is the absence of a single solution: there is rather a bundle of supply functions which are all equilibria to the game. In the case of asymmetric players, non-linear cost functions, or capacity constraints it maybe even impossible to find a unique solution. However, the boundaries of the SFE, the Cournot and the marginal cost line, can be calculated without modeling the SFE itself and allow to determine the most and the least competitive outcomes of a market. Over the past years, several theoretical contributions have been made to find and to select equilibria under varying assumptions (e.g. Anderson and Hu, 2005, and Holmberg, 2005).