Market integration and merger assessments in the minerals industries

Abstract: This thesis consists of an introductory part and five self-contained papers dealing with market definitions and economic assessments of horizontal mergers in the minerals industries world-wide. Paper I analyses the relevant geographic market for internationally traded steam and coking coal using Cointegration and Error Correction models. Quarterly price series data between 1980 and 2000, in the two dominating regions Europe and Japan, are used. The results show that for both markets the null hypothesis of the existence of a world market cannot be rejected, as demonstrated by the long- run cointegrating relationship between the price series. However, the short- run adjustments to the long-run levels indicate a faster adjustment towards the long-run equilibrium for coking coal than for steam coal. In Paper II the relevant economic markets for steam and coking coal are also in focus, and this paper applies two different tests using shipments data and price data, respectively. The results from both methods point in the same direction. In the case of coking coal the results indicate that the market is essentially global in scope, and also that the market has become more integrated over time. The results for steam coal show that the market is more regional in scope, and there exist no clear tendencies of increased integration over time. One policy implication of the finding that the steam coal market is more regional in scope, is that a merger in this market would likely be of more concern for antitrust authorities than the same activity in the coking coal market. Paper III estimates the price and welfare effects from the iron ore mergers between Rio Tinto and North Ltd in 2000, and CVRD and Caemi in 2001. The analyses are conducted using a merger simulation model that, based on the pre-merger situation, estimates the post-merger outcome. A standard Cournot framework model assuming homogenous product and that firms set quantities have been applied. The results from the merger simulations show that the estimated prices increases due to the mergers are about 6% and 7%, respectively. In both merger simulations the overall welfare effect is estimated to be negative, something which thus does not support the European Commission's decisions to allow these mergers. However, none of the simulations did take into account possible cost efficiencies resulting from the mergers. Paper IV also applies merger simulations to the same mergers as in Paper III. However, the merger model in this paper (PCAIDS) assumes that the product is differentiated and that the firms' strategic variable is price. The increases in the estimated market-weighted average price are 2.6% and 4.6%, respectively. When removing Caemi's Canadian asset, which is the Commission's verdict for allowing the merger, the estimated market price following the CVRD-Caemi merger decreases to about 3%. Overall the results support the Commission's decisions regarding both merger cases, and show that merger simulations of price effects can be a valuable tool in merger assessments. Finally, Paper V applies an event study to the Rio Tinto and North Ltd merger in order to analyse the competitive and efficiency impacts of the merger, and to investigate the main motivation behind the merger. The event study method analyses stock market reactions of the merging firms and close rivals at the time of the merger announcement. These reactions indicate whether the merger has had a positive or negative impact on the value of the firms, information that in turn can be used to assess the overall efficiency and welfare effects of the merger. The main result shows that efficiency improvements were the predominant motive behind the merger. This suggests that there are positive welfare effects to expect and the Commission's decision to allow the merger is supported. An overall conclusion from this thesis is that all the quantitative economic techniques that have been applied can provide information that is valuable in a comprehensive merger assessment study, but all methods need to be complemented bysupplementary (often qualitative) analyses.