Merger wave and causes: An analysis of United States pharmaceutical industry

Although economists have long noticed that mergers occur in waves, the wave pattern of mergers is an old and still unresolved issue. No unified theories can predict merger waves accurately. This dissertation seeks to model the wave behavior at United States pharmaceutical industry level by using a 2-state Markov Switching Model with autoregressive dynamics and constant transition probabilities. It suggests that the Unites States pharmaceutical merger activities shift between two alternative states with distinctly different means, "wave state" and "normal state". Moreover, it indicates that the pharmaceutical merger activities display wave pattern with "long swings". Once "wave state" or "normal state" is reached, it is more likely to stay in that state for years. This research also examines the causes of Unites States pharmaceutical merger waves and analyzes how industry shocks and stock market valuation influence pharmaceutical merger activities by running an ensemble of Quantile Regressions for Counts at 0.10, 0.25, 0.50, 0.75 and 0.90 quantile. It finds that the industry shock hypothesis has more explanatory power for pharmaceutical merger waves than managerial behavior hypothesis. The influence of stock market valuation deviation on United States pharmaceutical merger frequency is negligible. Specifically, merger waves in United States pharmaceutical industry is highly sensitive to the competitive pressure from international market compared to other industry shock variables - penetration of foreign companies, new molecular entity approval, pharmaceutical antitrust case dummy.