Essays on Unit-Root Testing and on Discrete-Response Modelling of Firm Mergers

Sammanfattning: Essay 1 investigates the time-series properties of the price of iron ore. The focus is on unit-root testing in the presence of a structural break. Unit-root tests with or without structural breaks are applied to historical prices of five different qualities of Swedish and Brazilian iron ore. Tests with exogenous or endogenous breaks are analyzed. Using unit-root tests allowing for an exogenous structural break in 1973, the year of the oil-price shock, the null hypothesis of a unit root is rejected for three of the five series. The sign and nature of the estimated breaks correspond to the state of the iron and steel industry during the oil-price shock.Essay 2 was co-authored with Rolf Larsson and deals with the issue of unit-root tests involving several parameters of the model. In a first order autoregressive model with drift or drift and trend, we derive the likelihood ratio test for a unit root against the stationary alternative. We also derive the test in a state space model with trend. Finite sample and asymptotic critical values are obtained by Monte Carlo simulations, and we examine the power performance of the likelihood ratio test.Essay 3 proposes a way to model firm mergers using the roommate matching game, whereby firms make preference rankings of potential merger partners. Given firms' preferences, game-theoretic mechanisms lead to a matching implying that each firm is either self-matched or assigned a merger partner. We derive expressions for the probability of a merger between a specific firm pair, and also a log-likelihood function for estimation using firm-specific data. The model's finite-sample properties are examined.The roommate game results in a complicated log-likelihood function, which places restrictions on the number of firms. Consequently, the model proposed in Essay 4 defines a different set of game rules. Using a mechanism of bid proposals among firms, we define a condition for firm acquisitions in terms of profit index comparisons. The resulting log-likelihood function is applicable to any number of firms, and the acquisition model is estimated on a panel data set consisting of European firms from the pulp and paper industry.