Revisiting Shareholder Value Creation via International Joint Ventures: Examining Interactions Among Firm- and Context-Specific Variables

Canadian Journal of Administrative Sciences, Jun 2004 by Merchant, Hemant

The above model was estimated over a 200-day period beginning 51 days before an IJVs announced formation. For each firm i in the sample, the study cumulated abnormal returns over a two-day event window consisting of the day a firm's IJV participation was first publicly announced and the following day. Event studies routinely undertake such aggregation to account for capital markets' reaction to announcements that may have been made after trading hours (McWilliams & Siegel, 1997). The two-day return for the sample was 0.56% (p

This study considered two other external variables, which had noticeably fewer data points relative to other variables in the study: (a) inter-partner rivalry (RIVALRY; n = 488), and (b) previous joint venture experience (EXPERIENCE; n = 426). Like RETURNS, these two variables were not used to generate clusters but were analyzed only after the clusters were identified. This practice has precedent (see Dussauge & Garrette, 1995), and is an agreeable compromise because the cluster analysis methodology discards observations even with one missing data point. Including the two above-mentioned variables to generate clusters would considerably reduce the sample size and so compromise statistical power, which is crucial for a more conclusive interpretation of results.

Following Park and Russo (1996) and others, this study coded RIVALRY on a five-point scale measuring the extent of product-market overlap between partners; higher values denoted greater rivalry, and vice versa. This variable's construction relied on descriptive information about firms' business scope reported in Value Line Investment Surveys (American firms), Directory of Corporate Affiliations (American and non-American firms), and Principal International Businesses (nonAmerican firms).

Per Harrigan (1988) and others, this study measured EXPERIENCE as the frequency of American parents' IJV participation over a three-year period immediately preceding the year in which firms announced their respective IJVs; higher frequencies denoted greater experience, and vice versa. This period was selected on the basis of available Dow Jones data and resource constraints. The data source explicitly identified American firms that had entered into IJVs. A list of these IJVs was generated for 1983 to 1993 along with the firms' name. All American firms in this study's sample were looked up in the above list. EXPERIENCE equaled the number of times an American firm's name appeared on that list during a given three-year period.

Cluster variates. This study engaged 12 salient theoretical variables to identify IJV clusters present in the data. As noted previously, these variables emerged from a review of the foreign direct investment literature and synthesis of empirical work on shareholder value creation via joint ventures. As with RIVALRY, this study measured overlap in business scope between American parents and their respective ventures (PVSCOPE) on a five-point scale (see Koh & Venkatraman, 1991); higher values on this scale denoted greater overlap, and vice versa. This variable was based on descriptive information about these entities' business scope reported in Value Line Investment Surveys, Directory of Corporate Affiliations, and the text of IJV formation announcements in Dow Jones.