تاثیر خاستگاه سرمایه گذار بر عملکرد شرکت: سرمایه گذاری مستقیم داخلی و خارجی در ایالات متحده

This paper evaluates the causal relationship between the source of origin of FDI and the performance of the target firm. The empirical analysis uses new data on a comprehensive sample of public U.S. firms that received FDI between 1979 and 2006. To account for the possibility that performance differences arise due to the selection of superior target firm rather than the change in ownership, I use propensity score matching to create similar comparison groups of target firms prior to acquisitions. The analysis reveals three major findings. First, acquiring firms from industrialized countries lead to labor productivity increases of 13% in the target firm three years after the acquisition compared to targets acquired by domestic firms. Firms that received developing country firm acquisitions, on the other hand, exhibit lower labor productivity gains four years after acquisition, compared to targets acquired by domestic firms. Second, targets receiving FDI by firms from industrial and developing countries also experience increases in profits, compared with firms receiving acquisition by domestic firms from the United States. Third, compared with domestic acquisitions, foreign industrial firm acquisition FDI tends to increase their targets' employment and sales, whereas targets acquired by firms located in developing countries experience a decrease in both revenues and total number of employees. These findings suggest that target firms are subject to significantly different restructuring processes depending on the origin of the acquiring firm.

Historically, the majority of foreign direct investment (FDI)1 flows has taken place predominantly between industrialized countries and from developed to developing countries. In recent years, however, developing country firms started engaging in outward FDI activity. The recent upsurge in cross-border mergers and acquisitions (M&As) from emerging countries such as China's Lenovo purchase of IBM Thinkpad and India's Tata Motor's acquisition of Ford's Jaguar and Landrover has raised great attention in policy circles, but yet, little is known about how these emerging market acquisitions are different from acquisitions by acquirers from developed countries. Traditionally, the business rationale for M&A is that the new combination of assets will be more productive than the sum of its parts (Dunning, 1981 and Hymer, 1976). Theories for FDI have relied on differences in relative input costs and market access as motivations for developed-market investment flows to emerging markets (Yeaple, 2003). Helpman et al. (2004) suggest that firms that invest abroad have to overcome larger fixed costs and barriers. As a result, foreign acquiring firms have to be more productive than their peers. In addition to technology transfer, industrial country acquirers often seek lower labor costs in emerging markets. For emerging market acquirers, the rationale for cross-border M&A might be different. Endowed with a relatively larger and cheaper labor force, it is likely that emerging market acquirers may relocate (or insource) manufacturing activity while keeping existing distribution networks in the host country of the acquired business (Chari et al., 2009). Therefore, the changing composition in acquirers raises the question whether heterogeneity among acquirers has consequences for target selection, implementation of M&As, and, therefore, post-acquisition target performance. Although existing studies have shown superior performance in foreign-owned firms compared with domestically owned firms (e.g., Harris and Ravenscraft, 1991, Swenson, 1993, Doms and Jensen, 1995 and Haskel et al., 2007), little is known about the differences in performance within the group of acquired firms and whether the act of acquisition improves target firm performance. This paper fills this gap by estimating the causal effects of acquirer origin heterogeneity on target firm performance. I use a newly-constructed data set to examine whether the post-acquisition performance of U.S. target firms differs when the buyer is either a U.S. domestic firm, an industrial country firm, or a developing country firm. I assemble a comprehensive sample of acquired U.S. public firms by linking daily M&A transaction information from SDC Thomson to each target firm's financial statement in Compustat. The U.S. provides a particularly suitable setting in which to study M&As given its role as the world's most sought after target country, with a combined value of cross-border and domestic M&A deals of $1.47 trillion in 2007.2 When comparing the impact of domestic and various foreign acquisitions on U.S. target firm performance, there are two main empirical challenges. First, the issue of selection arises in that better post-acquisition performance might be a result of more skillful selection (also referred to as “cherry picking” of targets) rather than the change in ownership per se. Second, one would ideally compare the performance of a target firm that is acquired by a domestic firm with the performance of the same target had it been acquired by a non-U.S. industrial country firm or a developing country firm.3 However, these counterfactuals are not observable because at any given point in time a target firm experiences only one of three events — it is acquired by either (1) a domestic firm, (2) an industrial country firm, or (3) a developing country firm. This creates a missing data problem. To solve these issues, I use propensity score matching to construct a comparison group of domestically acquired target firms that is similar on a set of important observable characteristics to the group of targets acquired by industrial country firms or developing country firms. In evaluating the target firm performance after the acquisition, I use measures of labor productivity, profits, sales, and employment as outcome variables. I find that over a period of five years following acquisition, compared to U.S. domestically acquired targets, targets acquired by industrial country firms experience a significantly higher growth rate in labor productivity of up to 13%, whereas for targets acquired by developing country firms, the growth rate in labor productivity is 23% lower. Moreover, targets acquired by industrial country firms and by developing country firms exhibit higher average profits compared with acquisitions by U.S. firms — by ten and nine percentage points, respectively. The data show that compared with domestic acquisitions, sales also tend to increase in industrial country firm acquisitions — by up to 29% compared to a similar group of domestically acquired firms, whereas sales decline by up to 20% for targets acquired by developing country firms compared to a similar group of domestically acquired firms. Finally, whereas industrial country firm acquisitions lead to an increase in employment of 24% in their targets, targets of developing country firm acquisitions reduce their total number of employees by up to 26%. Several sensitivity analyses are performed to assess the validity of the findings. I first show that the results are robust to different propensity score specifications. Additionally, I use a small sample of announced deals that were subsequently withdrawn and find no differences in target performance between domestic withdrawn and foreign withdrawn deals. To illustrate the importance of controlling for selection and creating appropriate comparison groups, I redo the analysis without propensity score matching and show that the results are different. Lastly, I perform robustness checks on various subsamples of the data to make sure that the results are not specific to one particular feature of the data. This paper provides some of the first findings on the differences in various aspects of target post-acquisition performance as a result of acquirer origin heterogeneity. Previous studies do not differentiate between the investor origin, since the comparison is conducted only between foreign-acquired and non-foreign-acquired firms.4 Therefore, average effects estimated across groups could mask differences in performance between each of the groups. Examining the impacts of foreign and domestic acquisitions on target firms also reveal important policy implications. The U.S. government has sometimes taken a hostile attitude toward foreign acquirers of U.S. target firms.5 Another source of concern about foreign acquisitions is the potential loss of American jobs. Domestic M&A transactions do not provoke the same sort of concerns as their cross-border counterparts.6 Understanding the impacts of cross-border M&As sheds light on whether these political concerns are validated by the post-acquisition economic performance of the different types of acquisitions. I begin with a theoretical and empirical background on why investor origin impacts target firm performance based on existing literature. Section 3 provides a description of the data. Section 4 outlines the details of the identification strategy using multiple treatment propensity score matching combined with a difference-in-differences estimator. Section 5 presents the empirical results and discusses the different ways in which gains are realized among the varying types of acquirers. Section 6 provides robustness checks, and Section 7 concludes with a discussion of the implications of these results for both future research on FDI and economic policy.

This paper measures the performance of U.S. target firms after acquisition by firms from a broad range of countries over a five-year period using daily announced M&A information and firm level financial statement data between 1979 and 2006. In contrast to previous studies that lump all acquisitions by foreign firms together, I differentiate those acquisitions made by industrial country firms from those made by firms from developing countries. In order to control for non-random selection, I use multiple treatment propensity scores to match similar firms between comparison groups. Consistent with the predictions of the Helpman et al. (2004) model, targets acquired by firms from industrial countries exhibit the best post-acquisition performance. Targets acquired by non-U.S. firms from industrialized countries experience an increase in labor productivity of up to 13% and an increase in profitability that is greater by 10 percentage points compared with targets acquired by domestic firms. This improvement in performance is driven by increases in sales. Acquirers from developing countries lead to 23% lower labor productivity gains in their targets compared to targets acquired by domestic firms, but they improve the post-acquisition profit of their targets by 8 percentage points more than U.S. domestic acquirers do. In contrast to acquisitions by non-U.S. industrial country firms, acquisitions by developing country firms tend to result in lower labor productivity gains and decreases in employment and sales in U.S. targets. These results are robust to different propensity score specifications as well as to sample classifications. The study provides some of the first evidence that acquirer origin matters for target post-acquisition performance. The findings also highlight how accounting for heterogeneity in acquirer types reveals different channels by which post-acquisition performance is improved. The use of multiple treatment propensity score matching and detailed firm level data in this paper provides a methodology for controlling possible selection issues that could be employed in other studies of acquisition types where selection is non-random. In fact, I show that when not controlling for selection, the estimation yields different results that do not separate causality from correlation. Finally, even though an overall welfare assessment is not possible here due to the lack of data on the acquirers, the results suggest that U.S. public target firms benefit more from foreign acquisitions than from domestic M&As in terms of overall performance measured as profitability, with the largest improvements stemming from non-U.S. industrial country firms. If assessing performance using labor productivity, however, being acquired by a non-U.S. industrial country firm has positive effects on the target's productivity, whereas being acquired by a developing country firm results in lower productivity gains, when compared with domestic acquisition. At the same time, workers in U.S. public target firms that are acquired by developing country firms are more likely to lose their jobs, whereas there are employment gains in acquisitions made by non-U.S. industrial country firms, compared to U.S. domestic acquisitions. These findings provide new insights into the workings and consequences of domestic and cross-border M&As. In particular, for governments that are devising policies toward FDI, these results suggest that not all types of foreign investments should be treated in the same way. By building on this paper's approach to differentiating acquirer types, future studies can use more detailed data on the acquirer firms to help evaluate the overall impact of M&A deals. For instance, do acquirers perform differently after purchasing target firms? Do productivity, revenue and employment also change differently in the acquirers depending on the type of target? The methodology in this paper allows for the study of the effect on acquirers by differentiating the types of targets. Specifically, it enables us to identify how post-acquisition performance changes when targets are located in different parts of the world. Complementing the results in this paper, such future studies will increase our general understanding of the effect of M&As on both acquirers and targets in a variety of locations around the globe.