The Impact of Internationalization on the Diversification–Performance Relationship: A Replication and Extension of Prior Research

Abstract

Most strategic management researchers defend related diversification as yielding superior performance vis-à-vis unrelated strategies. However, empirical results on this subject have sometimes been inconsistent with this position, prompting a search for moderating conditions. Based on a theory of international impediments to synergy formation, we attempted to replicate Bettis and Hall's (1982) findings using a sample of predominantly domestic Fortune 500 firms. We divided these firms into two groups—product-related (n = 23) and unrelated (n = 47) corporations—and assessed group differences in performance (5-year average return on assets (ROA) spanning 1985–1989) and risk (the standard deviation of ROA over the same time frame). Parallel to Bettis and Hall, we found that differences in ROA were significant (F[1,69] = 4.99, p < 0.05); whereas, differences in risk were not (F[1,69] = 2.03, ns). These findings provide some support for the international impediments theory. To extend Bettis and Hall's (1982) study, we tested a more inclusive model across a sample of firms, including multinationals. This expanded diversification–performance model was based on a sample including multinational firms from the Fortune 500, separating these again into product-related (n = 78) and unrelated (n = 83) concerns. After including control variables (firm size, debt, and global relatedness/unrelatedness) as covariates in the model, we used a multivariate analysis of variance (MANCOVA) test to look for differences in accounting (ROA and return on sales (ROS)) and market (market-to-book value) returns, and accounting (standard deviation of ROA and ROS) and market (systematic, unsystematic, and total) risk. In this test, group differences were not significant (multivariate F[8,146] = 1.654, ns), although size and debt emerged as significant covariates. Taken together, these findings support the international impediments theory, showing that although domestic product-related firms may outperform unrelated diversifiers, these differences do not seem to generalize to multinationals. Despite recent criticisms, our findings suggest that the relatedness synergy framework may yet apply, although only for domestic firms. We recognize the limitations of our study; however, the conclusions of this research are more important than ever, because American firms are internationalizing their operations at an increasing rate.