Cross-Border Mergers and Acquisitions and the Role of Free Trade Agreements

ABSTRACT

This article investigates the impact of the formation of free trade agreements (FTAs) on cross-border mergers and acquisitions (M&As). Using the comprehensive M&As dataset of Securities Data Company, we find that FTA relationship is associated with more bilateral cross-border M&As. Second, the cross-border M&As activities between a FTA country-pair do not increase faster than the acquiring country’s total foreign acquisitions, suggesting no evidence of investment diversion effect of FTA. Third, we find that existing FTA relationship with other countries positively affect cross-border M&As between a FTA country-pair. But these third-country FTA effects differ for acquiring country and target country when we look at the ratio of a country-pair’s FTA relative to the acquiring country’s total foreign M&As. Moreover, by exploring the detailed information on acquiring and target firms, we reveal that the effect of FTA differs for horizontal, vertical and conglomerate cross-border M&As. Our results are robust to various measures of M&As activities and econometric methods used.

KEY WORDS:

• cross-border mergers and acquisitions
• FDI
• free trade agreements

JEL Code:

• F21
• F23
• G34

Acknowledgments

Tan Li and Ying Xue are grateful to Larry D. Qiu for his guidance and helpful discussions. We are also thankful to comments received from numerous seminars and workshops. All the remaining errors are our own.

Supplemental Material

Supplemental data for this article can be accessed on the publisher’s website.

Notes

1. Namely, EU, EUFTA, EFTA, and NAFTA.

2. We thank an anonymous referee for suggesting us to explore the different time spans and phase in effect.

3. The use of country-pair fixed effect is stronger than having country fixed effects ${\gamma }_{\mathrm{i}}$ and ${\gamma }_{\mathrm{j}}$, and is recommended by the literature to tackle the endogeneity problem of FTA relation (Baier and Bergstrand, 2007).

4. We also use Poisson regression as robustness check. Results are available upon request.

5. We also use a longer sample period from 1982 to 2012, and get qualitatively similar results.

6. Using the number of M&A instead of the value of M&A is one significant difference from di Giovanni (2005).

7. The definition of FTA in this study refers to FTAs and custom unions in general.

8. Since our industry classification is at SIC four digit level, we use the 1% threshold to determine industrial linkage. We also tried different threshold such as 5% and 10% and get qualitatively similar results.

Additional information

Funding

Tan Li acknowledges that the work is supported by National Natural Science Foundation of China (No. 71703128).