Tax treaties and foreign equity holding companies of multinational corporations

Countries conclude tax treaties to pursue two main goals: eliminating double taxation on cross-border activities; preventing tax avoidance. Barthel and Neumayer (2012) find that existing tax treaties of the countries belonging to the same region, or exporting similar goods, may make other countries form new tax treaties. Chisik and Davies (2004) find that asymmetry in FDI may result in tax treaties with higher withholding tax rates.

Despite the expected relationship between tax treaties and FDI, earlier studies using tax treaty dummies and country-level data tend to find no significant evidence, as reviewed by Davies (2004). More recent studies using firm-level data tend to find some positive relationship. For instance, see Davies et al. (2009) and Weyzig (2013).

In general, effective tax rates of multinational corporations can be affected not only by their internal ownership structures, but also by various tax rules, such as withholding tax rates under national tax codes, withholding tax rates under treaties, corporate income tax rates in home, source, and intermediate countries, as well as tax relief rules (e.g., foreign tax credit, deduction, and exemption rules) in home countries.

Because my analysis is based on cross-sectional data, it is hard to examine a causal effect of tax treaties on ownership chains. Reverse causality may also explain the observed patterns. For instance, countries with more direct ownership chains can conclude more favorable tax treaties. In addition, unobserved factors can affect both ownership chains and treaty networks. Suppose that multinational corporations can lobby (or costly persuade) countries for treaties with low tax rates. Multinational corporations may incur the cost of lobbying only when they plan to organize direct ownership chains. Thus, the cost of lobbying, which may not be observable, can also explain the close relationship between ownership chains and treaty networks. However, this effect will be attenuated when countries make decisions on tax treaties independently of lobbyists and multinational corporations. Moreover, in this paper, I used the bilateral variables of Head et al. (2010), which can indirectly capture the effects of unobserved factors, such as lobbying process and cost. I believe that a potential endogeneity problem is minimized.

There has been extensive research on profit shifting of multinational corporations. Dyreng and Markle (2016) study the effect of financial constraints on profit shifting. Dharmapala and Riedel (2013) examine the effect of earnings shocks. Laffitte and Toubal (2019) examine profit shifting through sales platforms in tax havens. For related studies, see Davies et al. (2018) and Liu et al. (2020) on transfer pricing; Griffith et al. (2014) and Dischinger and Riedel (2011) on intangible asset location; Huizinga et al. (2008) and Desai et al. (2004) on internal debt finance. Jacob et al. (2019) develop a principal-agent model, where tax avoidance decisions are driven by moral hazard, planning costs, and potential earnings. Sageder and Feldbauer-Durstmuller (2019) review the literature on management control in multinational corporations and find the importance of the relationships between headquarters and subsidiaries, as well as control mechanisms.

Huizinga and Voget (2009) show how dividend taxation affects location choices of multinational corporations after cross-border mergers and acquisitions. Voget (2011) examines tax-motivated relocation of multinational headquarters. Barrios et al. (2012) find that corporate taxation of foreign-source dividends in home countries can deter entries into potential source countries.

Hong Kong and Singapore have tax treaties with China that reduce the WT rates on dividends to 5 percent. Hong Kong and Singapore impose no corporate income tax on dividends earned abroad and impose no WT on dividends.

A corporate entity is called a subsidiary if it is owned directly or indirectly by a multinational corporation. A subsidiary is called terminal if it owns no other subsidiaries. A subsidiary is called intermediate if it is not terminal, i.e., if it owns another subsidiary.

In reality, an intermediate subsidiary is often more than an equity holding company. An intermediate subsidiary may earn income from its own operations and from the operations of other subsidiaries it owns. An intermediate subsidiary can be deemed as an equity holding company if it earns income mostly from the operations of other subsidiaries. However, from ownership data, it is not straightforward to determine whether an intermediate subsidiary is an equity holding company. I use these two terms interchangeably.

The computation algorithms of Hong (2018) limit the cross-border length of a route to be three at most. In Appendix A, I provide regression results for robustness checks by removing ownership chains with cross-border length greater than three.

Accessed in October 2017 at orbis.bvdinfo.com.

In the Orbis database, incorporation does not necessarily mean incorporation for tax purpose, but means incorporation under corporate laws.

Generally, incorporating a business can give limited liability to shareholders, such as parent corporations. By using more complex ownership chains with additional incorporation of intermediate subsidiaries, multinational corporations can obtain a higher level of protection from the risks of their terminal subsidiaries. The protection of limited liability by additional incorporation may be an underlying reason for complex ownership chains. Hong (2021) develops a strategic model of corporate veil-piercing to examine the relationship between ownership complexity and limited liability.

Accessed in March 2016 at dits.deloitte.com and taxsummaries.pwc.com

Accessed in September 2017 at cepii.fr/CEPII/en/bdd_modele/bdd.asp

Accessed in September 2017 at ewf.uni-bayreuth.de/en/research/RTA-data

Accessed in October 2017 at transparency.org/cpi2015.

Accessed in September 2019 at eikon.thomsonreuters.com.

For example, suppose that an American ultimate parent owns a Chinese terminal subsidiary through two foreign intermediate subsidiaries in Hong Kong and the Netherlands. Because this ownership chain includes two FIS countries, Hong Kong and the Netherlands, there are two distinct combinations of ownership chains and FIS countries.

For each combination of an ownership chain and a country, if both home and source countries are in the set of the 70 countries, to avoid repetition, 68 observations are created. If one is in the 70 countries but the other is not, 69 observations are created. If none of home and source countries are in the set of the 70 countries, 70 observations are created.

Due to the high correlations between DIRECT and TSHOP in Table 5, I do not use them together as independent variables.

Due to the high correlations between MFIS and MCBL in Table 7, I do not use them together as independent variables.

Dyreng and Lindsey (2009, Table 3) find US firms with material operations in tax havens have effective tax rates about 1.5 percentage points lower than the other firms.

Ewert and Wagenhofer (2019) show that the quality of financial reports, such as Form 10-K, may depend on production risk, accounting system, and the scope of auditing.

Accessed in October 2017 at sec.gov/edgar/searchedgar/companysearch.html

For studies on profit shifting, see Blouin and Robinson (2020) and Clausing (2020).