Market value impacts of information systems around the world: A monte carlo investigation to reduce bias in international event studies

As commerce and its associated challenges become more globally interconnected, research about those phenomena increasingly crosses national borders. The event study, a widely used method in Information Systems (IS) business value research, has been increasingly deployed internationally in management disciplines, but not yet IS. As IS scholars begin to conduct international event studies, they should be aware of research finding that single-factor event study methods developed for single-country settings may result in estimation errors if not corrected for international use (Park, 2004). This paper conducts a Monte Carlo analysis to simulate security returns within the context of varying (a) levels of global stockmarket correlation and (b) structural relationships between securities and their markets. Comparing single-factor and international multi-factor specifications finds that for conditions commonly observed in global markets, the single-factor model exhibits problematic estimates (some exceeding 200% more error) for which the multi-factor model is able to provide substantial correction.