Using Stock Market Evidence to Inform the Definition of Product Markets in Competition Law

Abstract

Market definition, and, in particular, being able to identify who is in which product market, lies at the heart of competition law. The identification of demand elasticities necessary to conduct a formal definition of product market can be time consuming and often there is insufficient data to provide an accurate answer. This paper suggests a possible tool for market definition using stock market data. The method is informative, simple and very quick to undertake. The approach assumes that a company’s returns depend on an economy wide factor, a product market specific factor and company specific idiosyncratic shocks. Given this assumption, the residuals from Market Model regressions will be more correlated for firms in the same product market than between firms in different product markets. To test the validity of this assumption, the central section of the paper looks at whether these correlations do indeed contain information about product markets. Using 51 companies and 13 product markets suggested by the Competition Commission (the “CC”), we calculate the correlation matrix of residuals and for each company identify the company with which it has the highest correlation. If the correlation matrix contains information about product markets then there should be some relationship between the correlations and the suggested CC markets. If the correlation of residuals contains no information with regard to relevant product market, then the chance of the highest correlation for a company being with a company from the same CC defined relevant product market would be roughly 7.5%. However, we find that in 82% of the cases the highest correlation for a company is with a company from the same CC defined relevant market. Finally, we examine the banking, the home credit and the ex-building society markets.