Suchergebnisse

AbstractThis study investigates the differential wealth impact of the National Energy Program on Canadian‐ versus foreign‐owned oil companies operating in Canada. The issue is analyzed with a stochastic oligopoly model of the industry, and the model's empirical implications are tested using the share prices of 60 firms. Both the cross‐sectional test and the time‐series test indicate that the Program announcement had an adverse impact on foreign‐owned firms relative to Canadian‐owned firms in terms of significant negative returns during the budget week, but the evidence indicates that this impact was not anticipated by the market before the announcement.RésuméCette étude examine I'impact du Programme national de I'énergie sur la valeur des compagnies pétrolières étrangères et le compare à I'impact sur les compagnies sous contrǒle canadien. L'analyse est faite à partir d'un modèle stochastique d'oligopole pour le secteur pétrolier. Les conclusions de l'analyse sont testées empiriquement à partir duprix des litres de 60 compagnies. Les résultats des tests inter‐compagnies et de ceux sur les séries temporelles suggèrent que I'annonce du Programme a eu un effet négatif plus grand sur les compagnies étrangères, comme I'indiquent les rendements négatifs importants observés durant la semaine du discours du budget. II n'est cependant pas évident que cet effet ait été anticipé par le marché boursier avant I'annonce du programme.

The rapid integration of international capital markets is apparent from the increasing number of foreign firms listed on the U.S. stock exchanges. A number of event studies have used stock returns of foreign firms listed on U.S. exchanges to examine the stock price reactions to various announcements. However, some researchers have raised concerns over the choice of event study models in the study of foreign firms. By employing actual stock return data in various simulation scenarios, this study compares the powers of alternative event study methods for foreign firms. The results show that all models and equity indexes perform equally well when there is no clustering of event dates, though there are high Type I and Type II errors when event dates cluster together.