Estimation of dynamic games with asymmetric players

The purpose of this project is to estimate dynamic games with asymmetric players. In three sub-projects the semiconductor industry will be analyzed as an example of an oligopolistic market, and cattle auctions in Austria as an example of an auction market. Learning effects are prevalent in the semiconductor industry. Still, there is a variety of market structures. The microprocessor industry is dominated by two large firms, in the DRAM industry there are a larger number of suppliers with nearly the same market shares. Semiconductors are an important input to several high-technology industries, and DRAM chips are usually thought of as technology drivers. Sub-project 1 will investigate the implications of learning effects, and asymmetric market structures on market power and competition policy in the semiconductor industry. Within a structural, dynamic, oligopolistic model, asymmetries in firms` behavior, like first-mover advantages, will be analyzed. With firm-specific data a system of demand and supply equations will be estimated to empirically investigate the consequences on price-cost margins. Sub-project 2 will study firms` timing of a new product generation. When introducing a new product generation, firms have to consider the cannibalization effect of the new generation on the old one. Thus, the timing of the new generation can be an important strategic device. The main interest lies in determining firms` performance, their competitiveness and the effects of early adoption of new technology on productivity growth. In Austrian cattle auctions bidders typically face different demand. Sub-project 3 will develop a theoretical model of sequential auctions with asymmetric bidders. This model will then be implemented as a structural econometric model and brought to data. The aim is to identify and estimate bidders` valuations, which also determine sellers` expected revenues.