Capital Round-Tripping: Determinants of Emerging Market Firm Investments into Offshore Financial Centers and Their Ethical Implications

Foreign direct investment (FDI) in offshore financial centers (OFCs) is gaining increased attention in business ethics research. Much of this research tends to focus on OFCs as locations where firms can avoid taxes, considering such behavior as unethical. Yet, there is dearth of studies on capital round-tripping by emerging market firms, which is an integral part of this phenomenon. Such round-tripping involves firms sending capital into OFCs only to invest it back in the home country under the guise of “foreign” investment. Presently there is little discussion of the ethical implications of such round-trip FDI activities. In this paper, we conceptualize round-tripping as institutional arbitrage and look at the determinants and ethical implications of such investments into OFCs. Exploring Russian round-tripping we note that firms tend to invest more funds in OFCs that offer a combination of tax and secrecy, or secrecy and property rights protection arbitrage opportunities. In either case firms exploit the opportunities provided by institutional differences between the OFC and Russia while investing back into Russia. Our results tend to indicate that equating OFC investment to tax avoidance and thus deeming it as unethical behavior is too narrow an explanation in the case of emerging economy round-tripping. This is because such investments are often motivated by the unethical behavior of home country stakeholders and may in fact provide benefits to society.


Introduction
For many emerging market multinationals (EMNEs) offshore financial center (OFC) investments constitute a large proportion of its FDI activity (UNCTAD, 2017, p. 12) and FDI from OFCs often accounts for the majority of emerging market inward FDI (Gopalan & Rajan, 2016). A key characteristic of OFC investments from emerging economies is its round-trip nature (Aykut et al., 2017). Round-tripping means that investors first send funds to the OFC and then reinvest those funds back into the home country. Roundtripping activities are estimated to represent from 10 to 50% of FDI activities from emerging economies such as Brazil, India, Russia, and China (Aykut et al., 2017). Despite the interest in determinants of FDI in OFCs in general (Buckley et al., 2015;Chari & Acikgoz, 2016;Gopalan & Rajan, 2016;Haberly & Wójcik, 2015a, 2015bPérez et al., 2012), few studies have focused on round-tripping as an important part of EMNE activities (for exceptions see Fung et al., 2011;Ledyaeva et al., 2015;Sutherland & Anderson, 2015). Because of this we know little about the theoretical drivers of investments into OFCs for round-tripping activities, the trade-offs EMNEs make when choosing between various OFCs, or the ethical implications of these investments.
While multiple motives underlie a firm's move to foreign locations, scholars suggest that at least for EMNEs institutional arbitrage theory (Boisot & Meyer, 2008;Luo & Wang, 2012) can be used to understand the foreign location choice. This theoretical approach is based on institutional theory and the idea that countries have unique sets of institutional characteristics including differences in economic, legal, and social/normative institutions (Scott, 1995). But unlike more traditional institutional distance theory (Kostova, 1999) which maintains that differences in institutions create a liability of foreignness (Zaheer, 1995) that firms must try to overcome, institutional arbitrage theory suggests that firms purposely take advantage of differences in institutional attributes between countries by shifting part of their operations to locations that offer more advantageous (to the firm) institutional settings (Gaur & Lu, 2007;Witt & Lewin, 2007). For example, a firm might move its research and development facilities to a location that offers strong property rights protection for patents (Naughton, 1999), or establish production in a location with low environmental regulations so it does not have to clean-up its production system (Dean et al., 2009). From this perspective firms pursue a deliberate strategy to seek-out institutional contexts that can enhance firm competitiveness (Boisot & Meyer, 2008), instead of taking institutions as constraints that firms need to overcome with strategic adaptation (Jackson & Deeg, 2008).
Furthermore, heightened public attention on offshore FDI has been accompanied by increased interest in the business ethics of offshore incorporation (Johnson & Holub, 2003;Preuss, 2012). Most of the ethics literature on this topic has equated OFCs to tax havens, and consequently focused on tax avoidance that is unanimously condemned as a legal but unethical business practice (Johnson & Holub, 2003;Preuss, 2012;Payne & Raiborn, 2018; see also Dowling, 2014;Lanis and Richardson 2012). This business ethics research on OFC (i.e. tax haven) investment at least implicitly assumes a developed institutional framework, where the home country government is collecting taxes "for the good faith provision of public goods and services" (Payne & Raiborn, 2018, p. 469) and therefore the demand for ethical behavior concerns only the firm. This is reflected in theories that view the use of OFCs as a violation of the social contract between business and society (Johnson & Holub, 2003; see also Donaldson & Dunfee, 1994, or as a practice where the negative consequences of which need to be alleviated with positive corporate social responsibility activities (Col and Patel, 2019).
Recently, researchers have started to explicitly acknowledge the role of the (home country) institutional context as a trigger for OFC investment, and extended the analysis to emerging economies (Su & Tan, 2018;Deng et al., 2019) that have "weak institutional environments with substantial government intervention and ineffective market-supporting institutions" (Su & Tan, 2018, p.1067. These studies focus on the relationship between the firm's embeddedness in its home country institutional environment and OFC investment, conceptualizing it as prosocial orientation (Su & Tan, 2018) or political status (Deng et al., 2019) of the firm. Yet, both these studies continue to conceptualize OFCs as tax havens, albeit mentioning other potential motivations, including hiding illegal conduct such as bribery (Su & Tan, 2018). Neither do these studies problematize the specific features of OFC investment from emerging economies, particularly its round-trip nature (although Deng et al.2019 mention it). Because of this we know little about the theoretical drivers of investments into OFCs for round-tripping activities, the trade-offs EMNEs make when choosing between various OFCs, or the ethics of such round-trip activities.
In this paper, we extend institutional arbitrage theory to examine the round-tripping activity from one emerging market (Russia), and look at the empirical determinants of investments into OFCs made by Russian firms and their ethical implications. More specifically, we extend institutional arbitrage theory to suggest that not all round-tripping investors are simply looking to reduce their contribution to home country society by avoiding taxes. Instead, they are seeking to enhance firm competitiveness in the home country by shifting some operations abroad to access institutional characteristics in OFCs that positively influence the firm's ability to generate or protect its funds and enable it to return the funds to the home country legally.
Taking this approach, we make several contributions to knowledge. Building on the institutional arbitrage perspective we develop and test theory to explain why EMNEs use OFCs for round-trip investments. Past studies of investments into OFCs tend to take traditional theories of FDI and multinational corporations (such as Dunning's OLI paradigm or internalization theory) as the starting point and add host country variables such as taxation (Chari & Acikgoz, 2016;Haberly & Wójcik, 2015a) or property rights (Buckley et al., 2015;Pérez et al., 2012). A few empirical studies have introduced institutional arbitrage theory to explore investments into OFCs for capital round-tripping based on tax (Fung et al., 2011) or secrecy (Ledyaeva et al., 2015) arbitrage. Our study goes further by including three potential arbitrage opportunities (tax, property rights, and secrecy) suggesting that there are multiple motives for round-trip investments. We develop and test theory to explain how the combined impact of taxes and secrecy or property rights and secrecy drives decisions on round-tripping investments via OFCs.
Second, we extend institutional arbitrage theory to capital round-tripping by looking at host country instead of home country characteristics. Traditionally, differences between countries are thought to create liabilities of foreignness (Zaheer, 1995) and generate problems that firms try to resolve (Stahl et al., 2016), quite often through adjustments to organizational structures or business practices (Brouthers, 2013). The institutional arbitrage literature builds on these insights but suggests that firms might expand abroad to escape from home country institutional environments that impact the firm unfavorably or go abroad to exploit capabilities that the firm has developed in leveraging specific institutional characteristics (Boisot & Meyer, 2008;Chari & Acikgoz, 2016;Gaur & Lu, 2007;Luo & Tung, 2018;Witt & Lewin, 2007). We take a different approach and suggest that for round-trip investments firms do not seek institutional escapism or exploitation but are looking for arbitrage opportunities (institutional differences) that will allow the firm to protect its funds while making further investments in the home country (Fung et al., 2011;Ledyaeva et al., 2015). These firms seek foreign locations that not only allow them to benefit from institutional characteristics of the foreign (host) location, but also provide an opportunity to return the funds to the home market in a legal way.
Finally, by focusing on round-trip investments as a specific form of OFC investment we maintain that its existence challenges the "black-and-white mindset" (Deng et al., 2019, p. 470) on the ethicality of OFC investment. Su and Tan (2018) maintain that in the emerging economy context "the use of tax havens for financial maneuvers is probably a strategic decision rather than an ethical decision" (p. 1078)-a response to institutional constituents' demands. Here, we argue that as round-trip investment by default involves investments in business activities in the home country, its societal benefits such as provision of employment at least partially compensate for the tax income losses of the home country government, making it ethically a less straightforward question. We further maintain that unlike in developed economies where the firm's decision to engage in unethical behavior and incorporate in OFC is predominantly based on profits (Johnson & Holub, 2003), in emerging economies with rampant public sector corruption and government interference (Karhunen et al., 2018) the situation is different. OFC incorporation per se may be the only way to secure the very existence of the business, i.e. the seemingly unethical behavior of the firm is in fact triggered by unethical behavior of its institutional constituents. The secondary decision of which OFC to select may then be based on firm-level preferences.

Theory and Hypotheses
FDI research tends to approach investments into OFCs using traditional internationalization theories such as the OLI paradigm (Chari & Acikgoz, 2016;Haberly & Wójcik, 2015a) or internalization theory (Buckley et al., 2015). OFCs are defined as jurisdictions which provide some (or all) of the following services: low or zero taxation; moderate or light financial regulation; banking secrecy and anonymity (IMF, 2000). Due to these characteristics they are sometimes referred to as "tax havens" (Palan et al., 2010) or "secrecy jurisdictions" (Christensen, 2012;Cobham et al., 2015). Business ethics research also tends to equate OFCs to tax havens, focusing on issues of tax avoidance which is unanimously condemned as a legal but unethical business practice (Johnson & Holub, 2003;Preuss, 2012;Payne & Raiborn, 2018; see also Dowling, 2014;Lanis & Richardson, 2012).
Consequently, prior international business and business ethics studies have identified taxation as the primary motivation for OFC incorporation, (Chari & Acikgoz, 2016;Haberly & Wójcik, 2015a), whereas other potential reasons such as access to international capital, access to more favorable regulatory environments (e.g. Johnson & Holub, 2003), or availability of advanced financial services (Buckley et al., 2015) are mentioned to a far lesser extent. In addition, home country determinants such as institutional weaknesses have been shown to drive EMNEs to invest in OFCs (Chari & Acikgoz, 2016). Some studies have noted the "dark side" of OFC investment, showing that part of the FDI in these centers is driven by non-traditional motivations such as money laundering facilitated by OFC secrecy (Pérez et al., 2012) or hiding illegal conduct such as bribery (Su & Tan, 2018). Research on OFC FDI generally recognizes roundtripping as an important part of the phenomenon but does not make an empirical distinction between round-tripping and other forms of FDI (cf. Ledyaeva et al., 2015;Sutherland & Anderson, 2015).
We suggest that for round-trip investors, an institutional arbitrage framework might provide a better theoretical perspective to study determinants of investment into OFCs. Institutional arbitrage refers to the situation, where a firm seeks to benefit from differences between (two) institutional environments (Boisot & Meyer, 2008). In doing so, it differs from the mainstream institutional approaches in international business research which view institutions as fixed constraints to which the firm must adapt (Jackson & Deeg, 2008). These international business studies typically suggest that institutions play a key role in determining FDI (e.g. Pajunen, 2008;Holmes, Miller, Hitt, & Salmador, 2011). Most studies conclude that in general better host institutions encourage real FDI inflows (e.g. Lu et al., 2014). Yet a growing number of studies suggest that institutional distance between host and source country is an important determinant of bilateral real FDI flows (e.g. Xu & Shenkar, 2002). Traditionally differences between countries are thought of as problems that firms try to resolve (Stahl et al., 2016). Foreign institutions may differ significantly from the home country and these differences create a liability of foreignness which firms need to overcome otherwise negatively impacting the foreign operations (Zaheer, 1995). In contrast, the institutional arbitrage literature maintains that firms have two motives to undertake such arbitrage opportunities: passive (escapism) or active (exploitation) logics (Luo & Tung, 2018). In the escapism case firms go abroad in order to avoid weak institutional environments at home (Boisot & Meyer, 2008;Witt & Lewin, 2007). EMNEs tend to do so by, for example, investing in advanced markets where patent and copyright protection is strong and institutional conditions are more conducive to business development (Chari & Acikgoz, 2016;Luo & Tung, 2018). In contrast, the exploitation view builds on the notion that EMNEs are adept at competing in developing countries with similar institutions because of the capabilities they have developed from operating in such markets (Luo & Tung, 2018). Under this logic EMNEs make investments in other developing countries to exploit existing knowledge and experience.
We maintain that round-trip FDI represents a unique form of institutional arbitrage since, unlike prior institutional arbitrage focused research, the funds sent abroad are not invested in the host country but are returned to the home country for further investment. Building on prior capital round-tripping research (Fung et al., 2011;Ledyaeva et al., 2015;Sutherland & Anderson, 2015), we extend institutional arbitrage theory to suggest that round-tripping firms seek foreign locations (OFCs in most cases) that offer institutional characteristics that differ from the home country and provide the firm with some protection from the home country government before returning the funds to the home country. We also contend that this round-tripping OFC investment is often misunderstood (and lumped together with other OFC investments) as unethical behavior, when in fact much of it provides advantages to home country stakeholders (governments and employees) which would not be possible through other means. Previous research that has conceptualized round-tripping as institutional arbitrage has focused on home country institutions as drivers for such behavior, including tax incentives offered to foreign investors that become available for domestic (home country) firms via registration in an OFC (Fung et al., 2011), or widespread home country corruption that drives firms to safeguard their funds from corrupt authorities via OFC investment (Ledyaeva et al., 2015). We, in contrast, focus on host country (OFC) institutional characteristics and develop theory to suggest that roundtripping firms seek a combination of institutional arbitrage opportunities in these OFCs.

Institutional Arbitrage Opportunities and OFC Investments
OFCs offer several opportunities for EMNEs to take advantage of arbitrage situations. One important regulatory institutional arbitrage opportunity provided by some OFCs is the ability to lower the tax burden without having to relocate the firm or create unfavorable home country reactions. Prior empirical research has shown that the level of taxation in general is an important determinant of FDI decisions of multinational firms (see, for example, the meta-analysis of De Mooij & Ederveen, 2006). In recent years MNEs have become even more sensitive to differences in tax rates, as technological advances and the loosening of trade restrictions has made the movement of capital across national borders much easier (Fung et al., 2011). Shifting profits to low/no tax locations has become a common, albeit widely criticized, practice for firms around the world (Akamah et al., 2018;Jones & Temouri, 2016). Business ethics research on tax avoidance tends to condemn such actions as a legal but unethical business practices (Johnson & Holub, 2003;Preuss, 2012;Payne & Raiborn, 2018; see also Dowling, 2014;Lanis & Richardson, 2012). Despite this, one of the main drivers of investments into OFCs for firms is the desire to take advantage of institutional arbitrage opportunities available to minimize the tax burden of the firm (Chari & Acikgoz, 2016;Haberly & Wójcik, 2015a).
For emerging market firms tax arbitrage opportunities are very important. In emerging economies enterprises generating legal income often face the burden of high taxes, because much of the economy operates on the fringes and do not make a contribution to cover governmental expenses such as the provision of public services (Bah & Brada, 2014). Because of this firms look for opportunities to minimize their tax burden and thus have more funds to invest in the business. Establishing a presence in an OFC provides one such opportunity. Taking advantage of low or no-tax OFC locations can help shield funds from excess tax burdens, providing greater liquidity to the firm and potentially offering additional investment opportunities when the funds are returned home. Hence, though this practice is considered unethical since it may result in decreased tax revenues for the home country government from corporate taxes, it can also result in increased payrolls as firms have more funds to reinvest in their operations, providing other societal benefits such as direct employment, indirect (supplier) employment, and potentially increasing payroll tax collection centrally. This is particularly important in the context of emerging markets where operating from an offshore jurisdiction can be the only option to maintain a business (see, for example, Zevyakina, 2019). Therefore, we hypothesize that: Hypothesis 1 Offshore financial centers with low or no taxes attract a greater amount of round-trip investment compared to higher tax OFCs.
Another important regulatory institutional arbitrage opportunity provided by some OFCs is strong property rights protection (Christensen, 2011) which has been identified as one dimension of institutional quality pertinent to FDI location choice (Holmes et al., 2013). In general, property rights protection is a broad category of regulation that covers the legal status of ownership of goods, services and contractual agreements (Maekelburger et al., 2012). Property rights protection influences a firm's ability to safeguard its proprietary knowledge through enforceable patents or trademarks (intellectual property (IP) protection) as well as the ability of firms to rely on written contracts for services and payment (Hagedoorn et al., 2005). Countries tend to differ in the way they deal with property rights through variations in laws and regulations as well as in enforcement of these laws and regulations. In some instances, especially in developing countries with less transparent and more fluid market institutions than those in developed countries (Huang & Li, 2019), property rights protection laws are underdeveloped (Ahlstrom et al., 2008). This creates a situation where firms are unsure that their rights will be protected from potential partners, competitors, customers, and governmental authorities. Under these circumstances firms often minimize investments in proprietary products and services or rethink the location of contractual agreements, shifting such investments to other countries where protection is strong (Jandhyala, 2013). Hence, firms prefer and seek-out strong property rights protection as an "institutional safeguard" to their rights (Maekelburger et al., 2012).
For emerging market firms, property rights protection is important for a number of reasons. First, like for other firms, proprietary knowledge and or technology can be an important component of competitive advantage (Huang & Li, 2019;Luo & Tung, 2018). Because emerging markets often have weak or underdeveloped institutions, protecting a firm's intellectual property from appropriation might be difficult (Ahlstrom et al., 2008). MNCs from developed countries may minimize the appropriation risk by keeping their most valuable knowledge assets confined to a strong home country institutional context (Berry, 2017). For example, many non-US firms take out patents in the US because of the legal protection this provides (Albert et al., 1991;Song & Shin, 2008). For an emerging market firm a similar option is to establish a presence in an OFC, as some of them have strong property rights protection (Christensen, 2011). This can help insure and safeguard their firm-specific knowledge and or technology.
Second, in the institutional context of emerging economies, the issue of property rights is not only associated with the protection of intellectual property but is closely linked to their underdeveloped financial and capital markets (Ahlstrom et al., 2008). Institutional voids such as poor property rights protection (legal enforcement of contractual obligations) create high transaction costs in the domestic market, which may force businesses to seek less costly and more effective alternatives abroad (Buckley et al., 2015;Sutherland & Anderson, 2015). Although some emerging market firms have access to State funding, this funding is limited and most privately owned firms must rely on other sources of capital (Sutherland & Anderson, 2015). Therefore, as private firms in these emerging countries find it difficult if not impossible to access financing domestically, they seek foreign locations with strong property rights which makes it easier to access capital (Buckley et al., 2015;Sutherland & Anderson, 2015). Once these funds are accessed, they can be returned to the home country to be reinvested into the home country operations. Thus, the need for capital to expand the firm provides an explanation why round-tripping emerging market firms seek investments in OFCs with strong property rights protection.
Another important component of strong property rights protection is protection from hostile government interventions. In emerging markets local or regional government officials can exert unexpected or abrupt interference on a firm's operations (Karhunen et al., 2018;Luo, 2005). Since this type of government intervention restricts a firm's ability to make decisions and increases uncertainty, it is recognized as a major hazard in emerging markets and particularly transitional economies (Oliver, 1991). Foreign firms registered in countries with strong property rights protection are more immune to such interventions. Hence, the relocation of emerging market's businesses (the legal location of incorporation not the actual operations) to OFCs with strong property rights protection can be motivated by businesse's desire to get shelter from opportunistic domestic authorities and the threat of hostile takeover. This way the roundtrip investor possesses the ability to "manage institutional idiosyncracies" (Henisz, 2003, p. 174), including the ability to protect against the "grabbing hand" of government and opportunistic behavior of local business partners. As a round-trip investor, it can still actively take advantage of domestic business opportunities (Sutherland et al., 2010).
Of course, these actions raise a number of ethical questions. First, is it ethical to look for opportunities that protect the property rights of business owners from the unethical behavior of home country governments or government officials? Second, is it ethically right that firms invest in locations where institutional mechanisms are strong and therefore allow the firm to undertake business transactions that result in improved financing for the firm and hence firm growth and greater home country employment? From this perspective the ethics of investing in OFCs is not straightforward, as in some cases it might be the only option for the firm to operate legally and to contribute to the home country economy by providing growth opportunities and employment. Hence, at least for round-trip investors, the seemingly unethical behavior of offshore investments may in fact be an ethical response to existing institutional voids at home. Thus, our second hypothesis suggests: Hypothesis 2 Offshore financial centers with stronger property rights protection attract a greater amount of round-trip investment compared to OFCs providing weaker property right protection.

The Moderating Impact of Secrecy
Secrecy, including financial secrecy and anonymity granted to investors, is one of the key institutional characteristics of OFCs (IMF, 2000). Some authors even apply the term 1 3 "secrecy jurisdictions" in reference to OFCs to highlight the importance of secrecy in explaining offshore investment (Christensen, 2012;Cobham et al., 2015). Secrecy has in general proven to be a useful instrument when attracting FDI (Carlin & Lokanan, 2018), as corporate and financial secrecy are of key importance for asset protection of MNEs (Smith, 2014). Secrecy laws in OFCs make it difficult for outside law enforcement, creditors, or other potential claimants to identify ownership and control of assets (Smith, 2014). At the same time, secrecy provided by OFCs has become a global policy concern as it thwarts the effective taxation of income and profit, and facilitates money laundering, abuses of market regulations, and the financing of terrorism (Cobham et al., 2015;Johnson & Holub, 2003;Preuss, 2012;Su & Tan, 2018). The same secrecy arrangements that protect the ownership of legal assets provides investors of illegally obtained funds a high level of immunity from criminal investigation by law enforcement agencies (Christensen, 2011;Olatunde, 2012). Because of this, OFCs have been criticized for creating a supply side stimulus for corrupt practices, as secrecy facilitates laundering of proceeds of corruption (Christensen, 2012;Su & Tan, 2018). Thus, in principle OFC secrecy tends to raise important ethical concerns as secrecy can protect corrupt business and political actors (Christensen, 2012;Olatunde, 2012). Yet secrecy might also provide an important corollary to other OFC benefits for round-trip investors as explained below, thus raising questions about the broad issue of ethics in OFC investments.

Taxes and Secrecy
While we suggested that round-trip investors are attracted to OFCs providing tax arbitrage opportunities, we further argue that the strength of this association will be positively moderated by the level of secrecy provided in each OFC. Moving funds offshore for tax reasons may be a legal choice, but firms can suffer from significant reputational damage when customers and the general public learn of these tax manipulations (Akamah et al., 2018;Jones & Temouri, 2016). Although nothing illegal is taking place, the shifting of tax income from the home country to any other foreign location is often looked on with distain and distrust by home country stakeholders (Johnson & Holub, 2003;Preuss, 2012;Payne & Raiborn, 2018; see also Dowling, 2014;Lanis and Richardson 2012). As such, there can be significant consequences for those entities taking advantage of tax arbitrage opportunities, even though it is legal. Identifying people or firms with offshore ownership can trigger greater investigations from home country tax authorities (van Hulten, 2012). Given demands by the public and politicians, tax authorities will often focus audit resources on those firms/individuals thought to be hiding income, even if this means hiding it in foreign locations (Che Rosli et al., 2018). Another potential downside of shifting taxes to foreign locations can come in the form of consumer boycotts (Akamah et al., 2018). Once the public learns that a firm has shifted its tax liabilities to other countries, individuals or groups may take action that results in reduced sales and profits for the firm. Thus, for some firms the perceived ethical aspects of tax arbitrage opportunities might create a negative impact that dilutes any potential tax savings.
Yet when the OFC provides institutional arbitrage opportunities for secrecy, the potential negative consequences of tax arbitrage may not occur. In this case the tax arbitrage opportunities provided by an OFC can lead to significant saving for the firm while the secrecy provided by the location can protect the firm from negative consequences. This occurs because secrecy protects the firm from requests to disclose financial information about its OFC operations and therefore helps it avoid public criticism of tax avoidance (Akamah et al., 2018;Jones & Temouri, 2016). Therefore, we suggest that firms will invest more into OFCs with combined tax and secrecy arbitrage opportunities compared to OFCs that offer only tax or secrecy opportunities. Because round-tripping firms reinvested these funds back into the home market, providing opportunities for different stakeholder groups to benefit, the ethics of investing in OFCs that provide secrecy and low tax benefits might not be as straightforward as previous business ethics research indicates (Dowling, 2014;Johnson & Holub, 2003;Payne & Raiborn, 2018;Preuss, 2012). Hence our third hypothesis suggests:

Hypothesis 3a
Greater secrecy in an OFC will strengthen the relation between the level of tax arbitrage opportunities provided by an OFC and the amount of round-trip investment.

Property Rights and Secrecy
We suggested earlier that firms are more attracted to OFCs that provide property rights arbitrage opportunities. We further theorize that secrecy arbitrage opportunities will positively moderate the attractiveness of OFCs with strong property rights as locations for round-trip investment. In line with the arbitrage argument, the motivation to include host location secrecy in the location decision process is explained by the interplay between property rights regimes in the home and host countries. For example, investing illicit funds in locations offering strong property rights protection can lead to easier conversion to licit funds, or fewer questions by home country authorities. Yet without secrecy these benefits are hard to obtain. Firms can move funds to these offshore locations, but if home country authorities want to track down the source of such funds, coming back into the country, without secrecy they can identify the parties involved and more easily determine the illegal nature of the funds (Christensen, 2012). Thus, although strong property rights protection offers some arbitrage opportunities, the potential benefits of such investments may not be as great if financial disclosures are not protected. From an ethical perspective, this additional protection provided by secrecy increases a firm's ability to act unethically. Although some positive aspects might develop when the funds are returned to the home country and reinvested in the firm, illicitly obtained funds can be more easily hidden, thus thwarting governments from ending illegal activities within the home country.
However, when an OFC has both strong property rights protection and secrecy, the benefits described above are also more easily obtained for firms engaged in legal activities. For example, until recently in Russia the business owner could just hide his or her identity without formal violation of disclosure requirements by transferring shares to a nominee or private company registered in an OFC. Until the year of 2016 private firms registered in foreign offshore zones had not been required to disclose their owners 1 and, hence, the real ownership was non-traceable (Chernykh, 2008). Obviously, the effect of strong property rights in the OFC as protecting the investor from the grabbing hand of the government is even more beneficial when combined with secrecy in the OFC.
Finally, for firms possessing proprietary knowledge or trying to raise additional capital, property rights protection is very important, and the secrecy provided by potential OFC locations can help. For firms seeking protection of proprietary knowledge strong property rights combined with secrecy protects these specialized knowledge assets from home country agents (Smith, 2014). Firms seeking additional capital also benefit from the combined institutional dimensions of strong property rights and secrecy, as the property rights help improve access to capital (Buckley et al, 2015) while the secrecy helps protect these funds from home country government agents who might seek to benefit from such additional resources. In either of these cases the ethical implications of making these OFC investments need to be reconsidered as the benefits to the firm, employees and home country society at large increase since these roundtripping firms are better able to undertake successfully business activities at home. The above arguments lead to our final hypothesis:

Hypothesis 3b
Greater secrecy in an OFC will strengthen the relation between the level of property rights protection arbitrage opportunities provided in an OFC and the amount of round-trip investment.

Method
To test our hypotheses, we constructed a panel data set of firms located in Russia with FDI from various offshore financial centers. The phenomenon of capital round-tripping via OFCs is not Russia-specific, as similar tendencies have been observed in other large emerging markets, for example, China, India, Brazil and Ukraine (see, e.g. Brada et al., 2019). Yet Russia provides an excellent case for studying capital round-tripping via OFCs. This is because OFCs accounted for a significant share of Russian cumulative inward and outward FDI flows between 2007 and 2017-61 and 64%, respectively, according to Russian Central Bank data. Furthermore, the volume of these flows is relatively balanced: the value of Russian inward FDI flows from OFCs comprised approximately 77% of the outward FDI flows from Russia into OFCs. Hence, researchers have concluded that FDI from OFCs into Russia largely represents the round-tripping of Russian capital, which returns formally as FDI from abroad (Kuzmina et al., 2014;Ledyaeva et al., 2015;Novokmet et al., 2018;Sharafutdinova & Dawisha, 2017).
In addition, Russia is a rather typical emerging economy and is classified as one of the major emerging economies alongside Brazil, India and China (BRIC countries). Russia is also identified as an emerging economy by all major investment classification sources (e.g. Dow Jones, MSCI, FTSE, S&P, Russel and IMF; see Marquis & Rainard, 2015). What makes Russia somewhat distinct from other emerging markets, is its status as a military power and its pivotal role in international affairs (Shleifer & Treisman, 2005). However, when considering economic, legal, social and governance aspects, Russia represents a rather classical case of an emerging economy (see Shleifer & Treisman, 2005).
Firm-level data are not publicly available in Russia but can be purchased from the Russian State Statistics Service (ROSSTAT). According to Russian legislation, all companies in Russia must be registered with ROSSTAT. Upon registration, the company is assigned industrial, territorial, ownership, organizational and legal classification codes. These codes are required for their registration in pension and social funds. Furthermore, Russian legislation requires that most companies (excluding very small ones) must provide their annual financial/balance sheet reports to local tax and ROSSTAT offices. This is how these data become available for purchase. However, these data have many "NAs" (data are not available) because companies can prohibit ROSSTAT from selling their data to third parties.
Our ROSSTAT dataset provides information on 20,165 firms with foreign capital registered in Russia in the period 1997-2011. This dataset includes information on firms of two ownership types: full ownership by foreign entities or citizens, and joint ventures between foreign owners (foreign entities and citizens) and Russian owners (Russian Federal and municipal authorities, Russian non-profit organizations, Russian commercial entities, citizens). For each firm the Russian State Statistics Service records data on their sixdigit industrial code, information about the firm's owners (including country of origin) at time of registration, ownership status (joint venture or full ownership), the year of registration, and when reported, each firm's annual gross revenues.
From this dataset, we first excluded firms with foreign owners of different origin (for example, a company is owned by investors from USA and Cyprus or by investors from Ukraine, Cyprus and Russia) since such companies likely represent funds from other countries that are channeled through OFCs into Russia. This reduced the number of firms to 18,942 (94% of the initial dataset). We next excluded companies that did not provide revenue data for any single year in the studied period of 2002-2011, which further reduced the sample to 9107 firms.
For our main analysis, we extracted firms in which foreign ownership is represented by investors from OFCs. Following Haberly and Wójcik (2015a) and Ledyaeva et al. (2015), we utilize an expert agreed definition of OFCs. These centers are defined as jurisdictions appearing on more than 50% of the 11 OFC lists produced by different researchers (compiled by Palan et al., 2010) In Table 1 we summarize the core characteristics of the 30 OFCs in our sample. In most cases these OFCs are very small jurisdictions-considering geographic, demographic, and economic size dimensions. Though OFCs are commonly viewed as tax havens with relatively strong property rights protection, we observe that OFCs in our sample are relatively heterogenous in these respects. Some of them have zero taxes (Bahamas, Cayman Islands, Bermudas), somerather high tax rates (Malta, Monaco, Panama, Republic of Seychelles). Though most OFCs (considering those for which data are available) have strong property rights protection, the property rights index is rather low in Panama and Belize. Finally, the financial secrecy index of the OFCs in our sample ranges from 12.5 in Dominica (rank 71 out of 71) to 1879.2 in Switzerland (rank 1 out of 71).
Our focus on firms with OFC investors reduced our sample further to 4,893 firms (54% of the initial 9,107 firms in our sample with available revenue data). We then tried to examine the ownership structure of these 4,893 firms to determine if they represent Russian capital round-tripping through significant ownership connections with Russian entities or individuals. For the firms that were registered as joint ventures between OFCs and Russian owners (1735 firms or 35.5% of OFC sample), this connection is rather obvious. However, for the firms that were registered as wholly owned by owners from OFCs (3158 firms or 64.5% of OFC sample), we needed to provide further validation of significant Russian ownership. While the ROSSTAT database reports ownership structure at the time a company registers, the BvD ORBIS database contains up to date and historical ownership data including changes in ownership structure. About 95% of the sampled wholly owned establishments were identifiable in BvD ORBIS database. These data indicate that 2309 of our sampled wholly owned firms (73% of the total number) have obvious ownership linkages with Russian entities or individuals. In most of these firms, the global ultimate owners or ultimate controlling shareholders (taking into consideration multiple layers of firm's ownership structure) are Russian entities or individuals. A further 200 firms (6.3%) were identified as ultimately owned by foreign investors (34 firms are owned by Cypriots, one has Irish owners and the rest-non-OFC foreign owners). For 450 firms the origin of ultimate owners could not be clearly identified due to the limited or very mixed data on ultimate owners. For the remaining 199 firms the ownership information was lacking or too unclear or the firm is not present in the ORBIS database.
We look at each OFC round-trip investment individually to try and determine what institutional attributes of the OFC might have attracted Russian firm investments. In this case if a MNC has multiple OFC investments and more than one of these is used for round-tripping then we included each of the OFC round-trip investments in our sample. However, as Sutherland and Anderson (2015) note for Chinese firms, some of these other OFC investments might not be used for round-tripping purposes. Therefore, if the OFC investment was used for another purpose, other than round-tripping, then the OFC investment was not included in our sample. The final sample therefore includes 4044 round-tripping firms: 1735 (43%) joint ventures between Russian and OFC owners and 2309 (57%) firms wholly owned by OFC investors for which linkages with Russian owners were clearly identified. Of these firms 21% invested in trade, 18% in real estate, 16% in production activities, and 10% in the financial sector.
For comparison purposes, we also constructed a sample of non-OFC firms making investments into Russia in the same period, by extracting firms with non-OFC foreign owners of a single country of origin from our initial firm-level ROSSTAT database. We further exclude firms for which the revenue data were not reported for any single year of the studied period. We also add from our initial OFC sample 82 firms that have ultimate non-OFC foreign owners of a single country of origin (and, hence, they have not been included into the final OFC sample  and 3% have French, Turkish, Italian or Latvian owners. For these non-OFC firms 37% represent trade, 18% production activities, 5% real estate and 2% the financial sector.

Dependent Variable
Our dependent variable captures the value of round-trip investment activities between OFCs and Russia. Direct data on round-trip FDI via OFCs (either firm-level or countrylevel) does not exist due to the hidden and even illegal nature of capital round-tripping as a phenomenon. The best way to construct a measure of round-trip FDI would be to include volumes, directions and dates of firm-level foreign investment transactions, both outward and inward. Although in Russia these data are collected by the Central Bank of Russia (CBR), Russian Law (Federal Law N 282 'On official statistical recording and national system of statistics in the Russian Federation') prohibits the CBR from distributing (including selling) these data to third parties. Therefore, consistent with other research we use firm's gross revenues (transformed from Russian rubles into US dollars) to measure round-tripping (see also Ledyaeva et al., 2015). Firm-level revenues can be considered as an appropriate proxy for FDI flows, because they directly reflect the scale of multinational production, which in turn is directly linked to FDI. Ramondo et al. (2015) argue that sales/revenues might be even a better proxy for FDI activities than FDI flows since the importance of subsidiaries (in this study the focus is on firms with FDI) mainly depends on the magnitude of its production activity (which is measured by revenues or sales). In the same vein, Markusen (2008, p. 446) maintains that 'trade theory began to think of and indeed measure FDI not in terms of the value of investments (inputs), but in terms of the outputs of foreign affiliates.' Using firm-level revenues instead of FDI flows allows us to focus on real/production activities of capital round-trippers rather than on their financial manipulations. As Beugelsdijk et al. (2010) suggest, international business studies while examining MNE activities should rely less on FDI data and more on affiliate value-added and sales data. Though in this paper our study subject is not directly MNE activities but capital round-tripping, it is plausible to suggest that the amount of round-tripped capital to a great extent transforms into company production that is further transformed into revenues via sales.
Our actual measure is the natural logarithm of gross revenues (transformed from Russian rubles into US dollars) of a sampled firm i in year t (2002, …, 2011). Before the log transformation we multiplied the revenues by the percentage of OFC ownership in the registered capital of a respective firm i as reported in the ROSSTAT database. This way we scale revenues for the potential weight of round-trip investment from OFC. Hence, we proxy round-trip FDI via OFCs by the natural log of weighted revenues of firms located in Russia with FDI from OFCs.

Explanatory and Moderating Variables
To test our hypotheses, we created several explanatory variables. First, since we theorized that tax arbitrage is an important determinant of round-trip FDI via OFC, we utilize data on corporate income, capital gain, branch and withholding tax rates in OFC j (1,…,41) in year t (2002,…,2011). The data comes from the annual Worldwide Corporate Tax Guides of Ernst &Young for the years 2004-2011(EY, 2019. For the years 2002 and 2003, average taxes in the period of 2004-2011 were used. The tax rates in the OFCs included in our sample range from 0 to 40%. Our theory for tax arbitrage does not indicate a preference for higher/ lower taxes, but for low/no tax institutional settings. More specifically, in order to take advantage of the positive side of institutional tax arbitrage, it is important to choose an OFC with little or no tax liabilities. Hence, the choice of OFC is not based on incremental differences in tax level between two offshore jurisdictions. Instead, the choice is made towards OFCs with sufficiently low or zero taxes. The best way to accomplish this distinction is to use a dummy variable. Hence, we constructed a dummy variable for low tax in OFC j (1,…,41) in year t (2002,…,2011), D_Tax_Low jt , which equals one if the average tax rate in that location for that year is equal to or below the median value (10) and zero otherwise.
Second, our theory suggests that property rights arbitrage is an important determinant of round-trip FDI via OFCs. To address this issue, we utilize the level of property rights in each OFC. The Property rights index is a component of the Index of Economic Freedom of the Heritage Foundation, reported for 186 countries from 1995 onwards on an annual basis (Heritage Foundation, 2019). The property rights component is an assessment of the ability of individuals to accumulate private property, secured by clear laws that are fully enforced by the state. It measures the degree to which a country's laws protect private property rights and the degree to which its government enforces those laws. It also assesses the likelihood that private property will be expropriated and analyzes the independence of the judiciary, the existence of corruption within the judiciary, and the ability of individuals and businesses to enforce contracts. The property rights index ranges from a value of 30 to 90 in our sample OFCs. The property rights data are highly skewed for our sample with 71% of sampled firms investing in OFCs with a high property rights index of 90. Hence, using a dummy variable allows us to see if there is any difference between investors in these high property rights OFCs and investors in lower index countries. Therefore we constructed a dummy variable for high property rights index in OFC j (1,…,41) in year t (2002,…,2011), D_PRI_High jt , which equals one if the property rights index of an OFC is equal to or above the median value (90) and zero otherwise.
For testing our hypotheses about the impact of the interplay between tax and property rights dimensions with the secrecy dimension on round-trip FDI, we first fashioned a variable that measures the institutional aspects of secrecy in each OFC. To construct this variable, we utilized the Financial Secrecy Index of the Tax  To test Hypotheses 3a and 3b we created two interaction terms of Financial Secrecy Index of OFC times Dummy for low tax in OFC, and Financial Secrecy Index of OFC times Dummy for high property rights index in OFC.

Control Variables
We control for factors that past research has noted influence the choice of OFC location of FDI based on the Gravity model. The Gravity model is one of the most widely used theoretical conceptions of FDI determinants (Cuervo-Cazurra, 2006). The Gravity model explains economic activity (FDI in our case) between two countries as being a positive function of the economic mass of the countries and a negative function of the geographical distance between them. Although we cannot include indicators of economic mass for the host country, Russia, as they will be subsumed by year fixed effects in our model, following Rose and Spiegel (2007) we use GDP per capita and population to proxy the economic mass of the source countries (OFCs). In particular, LnGDPpc jt denotes natural logarithm of GDP per capita in US dollars of OFC j (1,…,41) in year t (2002,…,2011). LnPop jt denotes natural logarithm of the population of OFC j (1,…,41) in year t (2002,…,2011). The data for these variables comes from the World Bank or from the official websites of the governments of offshore jurisdictions. We do not control for the distance between offshore financial center and Russia as it is subsumed by firm fixed effects in our model. Finally, we include year and firm fixed effects.

Baseline Estimation Results
Before testing our hypotheses, we prepared a table of correlations as well as descriptive statistics for all variables included in the study ( Table 2). As can be seen from Table 2, none of correlation coefficients exceeds 0.5 and, hence, we can conclude that multicollinearity should not be a problem in our data.
To test our hypotheses, we used a fixed effects panel data regression method with four models. Each model took the form (plus respective interactions): where i denotes a firm, t-time and j-an OFC location. The dependent variable, LnY ijt is the natural logarithm of gross revenues of firm i (standardized/weighted by the percentage of OFC investor's ownership in the firm's registered capital) with FDI from OFC j in year t (2002, …, 2011). It should be noted that our data have two main dimensions, firm and time, while the OFC dimension is embodied in the firm dimension. All control variables have been described above. ijt is error term that is clustered over company-year groups. We estimate our model for the period 2002-2011 versus 1998-2011, because we have few observations for annual revenues for the period 1998-2001 but starting from 2002 the data becomes more balanced. Table 3 contains our four regression models. Model 1 tests Hypotheses 1 and 2, which suggests that round-trip funds are attracted to OFCs with low/no tax rates and strong property rights protection, respectively. The adjusted R-square of the model equals to 0.8. The coefficients of two variables, Natural logarithm of population of OFC and Dummy for low tax in OFC are positive and significant (p values equal to 0.048 and 0.017, respectively). The latter result is in line with our first hypothesis. Considering that our dependent variable is log-transformed, we can conclude that capital round-tripping via OFCs with low tax is 29% higher than via the rest OFCs (exponentiated coefficient of Dummy for low tax equals to 1.29). We further conclude that 1% increase in population in an OFC increases its chances (1) LnY ijt = a 0 + a 1 LnGDPpc jt + a 2 LnPop jt + a 3 FSI jt + a 4 D_Tax_Low jt to be used for capital round-tripping by 3.8%. Our second hypothesis is not confirmed in this model. We test Hypothesis 3a in Model 2. Hypothesis 3a suggests that the level of secrecy in the OFC positively moderates the relation between low/no tax rates and the amount of investment round-tripped via OFC. Model 2 contains the same variables as Model 1 plus the Financial Secrecy Index of OFC times Dummy for low tax in OFC interaction term. First, the coefficients of Natural logarithm of population, Dummy for low tax and Dummy for high property rights index are positive (as expected by theory) and statistically significant (p values equal to 0.051, 0.003 and 0.067, respectively). Therefore, in this model our first and second hypotheses are confirmed. In support of the second hypothesis the evidence suggests that capital round-tripping via OFCs with strong property rights is 48% higher than via the rest OFCs. The interaction term of Dummy for low tax in OFC with Financial Secrecy Index of OFC is positive and significant (p value = 0.043) providing support for Hypothesis 3a. The magnitude of its coefficient indicates that 0.1-point increase in Financial Secrecy Index leads to a 3.8% increase of capital round-tripping via OFCs with low/ no taxes versus the rest OFCs.
Our final hypothesis is tested in Model 3. Hypothesis 3b suggests that the level of secrecy in the OFC positively moderates the relation between strong property rights protection  Finally, in Model 4 we test all our hypotheses simultaneously and find support for all of them. All the coefficients in this final most elaborated specification are statistically significant and have expected signs. The model leads us to the following conclusions. First, capital round-tripping via OFCs with low tax and strong property rights protection is positive and significant (Hypotheses 1 and 2). Second, the positive relationship between capital round-tripping via OFC, on one hand, and low tax and strong property rights protection in OFC, on the other hand, is strengthened for OFCs with higher level of secrecy (Hypotheses 3a and 3b) though this evidence is stronger for the former (low tax → capital round-tripping) than for the latter (high property rights → capital round-tripping) relationship.

Industrial Patterns: Capital Round-Tripping of Licit Versus Illicit Funds
The theory of round-trip investment we propose in this study applies to round-trip investment taken as a whole. However, as we discussed in our theory above, round-trip firms can deal with different types of funds. On one hand, funds can be legally obtained; we refer to such funds as licit. On the other hand, funds can come from illegal activities (e.g. traffic in drugs) and need to be laundered. Illicit money laundering often takes place in OFCs (see, e.g. Picard & Pieretti, 2011) and is commonly acknowledged to be taking place in emerging economies like Russia (Rosfinmonitoring, 2017). Because the source of funds has significant ethical implications, we believe that it is useful to distinguish between these types of round-tripped funds (licit versus illicit).
For firms dealing with illicit funds, institutional arbitrage opportunities that provide low or no taxes may be beneficial but are not the primary motive. Pérez et al. (2012), for example, showed that illicit FDI flows are directed to countries considered to be centers for money laundering, and that such centers are characterized neither by low taxes, nor by status as a tax haven. For firms dealing with illicit funds opportunities to minimize taxes while desirable is not of primary importance. Such firms are more concerned about protecting itself and the funds from the home country government (Obermaier et al., 2017). Hence, we can suggest that strong property rights in OFCs are particularly important for investors of illicit funds given their need to transform these funds into licit property (Pérez et al., 2012). The establishment of an offshore company is commonly viewed as one of the key mechanisms for laundering illicit funds (Christensen, 2011). Entities involved with illicit funds seek property rights protection for a number of reasons. First, such protection helps reduce the chances of home country government access to these funds (Gunter, 2017). The cost and difficulties associated with clawing back illicit funds from countries with strong property rights is often a barrier to such action (Christensen, 2011). The second reason is that such locations help legitimize investments by providing instruments such as shell companies for this purpose (Allred et al., 2017).
In summary, we suggest that round-trip investors of licit and illicit funds might have different motives for capital round-tripping. More specifically, tax arbitrage opportunities are more important for firms dealing with licit funds (Hypothesis 1) while firms dealing with illicit funds are more attracted to OFCs with stronger property rights protection (Hypothesis 2). To test these differences directly we would need to separate these two groups of firms (i.e. dealing with licit or illicit funds). However, information which would allow us to directly disentangle illicit/licit activities of firms does not exist-due to the illegal and therefore hidden nature of illicit activities. Hence, as in previous studies we made use of the data available to distinguish these activities.
Motives for the usage of OFCs for illicit financial activities include the aim of consciously circumventing domestic regulations dealing with tax evasion and laundering illicitly obtained money (Ledyaeva et al., 2015). Several studies suggest that at least one of these activities-illicit money laundering-is largely associated with the real estate (e.g. OECD, 2007) and financial sectors (e.g. Ardizzi et al., 2014;Barone & Masciandaro, 2011). Hence, we define roundtrip FDI associated with illicit activities as those made by firms in the financial (financial mediation, insurance and auxiliary activities in the spheres of financial mediation and insurance) and real estate sectors. More specifically, we introduce the Dummy for real estate and financial firms that indicates whether the company's i round-trip FDI in OFCs can facilitate illicit activities (company i operates in financial or real estate sector) or not (company i operates in any other sector-various production activities, services and trade). While not all financial and real estate activities are of an illicit nature, research notes that in these sectors such activities tend to prosper (Ardizzi et al., 2014;Barone & Masciandaro, 2011;OECD, 2007). We further introduce two interaction terms with this dummy variable: Dummy for real estate and financial firms times Dummy for low tax in OFC, and Dummy for real estate and financial firms times Dummy for high property rights index in OFC.
In Table 4 we present results of testing Hypotheses 1 and 2 differentiating between the effects for real estate and financial firms and the others.
As we can observe, indeed Hypothesis 1 is more valid for firms operating in mainstream sectors (i.e. excluding real estate and finance) while Hypothesis 2 is more valid for firms operating in real estate and financial sectors. This evidence favors our propositions about different motives of capital round-tripping depending on the legal status of funds the firm possesses. However, since our division of firms into those dealing with licit and illicit funds is an approximation, these results should be considered preliminary and taken with caution. Future research might develop new ways of determining the use of illicit funds and improve our ability to calculate their impact.

Results for Non-OFC Sample
In this study, we developed and tested our research hypotheses specifically for the case of capital round-tripping via OFCs to emphasize the difference of such investment from conventional FDI. To further confirm the specificity of round-trip offshore FDI, we need to show that our theory does not apply to conventional FDI. To perform this task, we used the same ROSSTAT database to create another sample of non-OFC FDI investment in Russia (as explained in our methods section). We then estimate our Eq. (1) for this sample. The variables and the data sources are the same. However, it should be noted that Financial Secrecy Index is available only for nine and 19 non-OFC countries for the years of 2009 and 2011, respectively. Each subsequent edition of the Index (in the years of 2013, 2015 and 2018) included more and more non-OFC countries and, hence, we were able to substitute missing values with the nearest values from subsequent editions. This adjustment was not needed for the OFC sample as its first edition already included most of offshore jurisdictions.
In the discussion of the results for non-OFC sample we used the term investing country instead of OFC when applicable. The estimation results are presented in Table 5.
We can observe that none of the coefficients is statistically significant in Models 1 and 2. However, the coefficients of Dummy for low tax and interaction term between Dummy for strong property rights protection and secrecy variable are positive and statistically significant in Models 3 and 4. We argue that these results do not contradict the theory of conventional outward FDI. First, lower taxes spur economic activities that can positively affect the country's potential for conventional outward FDI. Second, the significant value of our interaction term of Property rights variable and Financial Secrecy Index in non-OFC subsample represents countries with combination of strong property rights (which are in turn represented by wealthier countries with stronger legal institutions) and well-developed financial institutions/ system that in turn can positively affect countrie's potential for conventional outward FDI.

Ownership Patterns
Finally, our data allows us to distinguish between firms of two basic ownership types: joint ventures between Russian and foreign/OFC entities/individuals and Russian wholly owned OFC firms. Though in the context of our study these two ownership types just reflect different organization of Russian ownership, their investment motives (in the context of capital round-tripping) might be different. Hence, we estimate our baseline models for the respective subsamples of our OFC sample. The results are presented in Tables 6 and 7.    These analyses tend to indicate that for joint venture operations low/no taxes (Hypothesis 1) and particularly in combination with secrecy (Hypothesis 3a) tend to be the driver of greater OFC round-tripping investments. These entities do not appear to be concerned with property rights protection (though the coefficient of the Dummy for strong property rights protection is positive and marginally statistically significant in the last model for this subsample). Thus, on one hand, OFC firms established as joint ventures between Russian and OFC partners do pursue tax avoidance motive in their round-trip investment strategies. On the other hand, the bare fact that they do not hide their Russian identity signals that they do not need special protection from opportunistic behavior of domestic (local or Federal) governments and, hence, strong property rights in OFC is of marginal importance for them. Furthermore, official partnership with Russian entities/individuals might give them some additional benefits (for example, participation in government tenders).
Yet for Russian wholly owned OFC organizations both lower taxes (Hypothesis 1) and stronger property rights protection (Hypothesis 2) were associated with greater roundtrip investments. However, the moderating role of secrecy does not seem to be important for these enterprises. Hence, though wholly owned OFC firms pursue both tax avoidance and better protection motives in their round-trip investment strategies, their owners are less concerned about secrecy level in OFCs they use for capital round-tripping. This suggests that this form of ownership organization (i.e. when Russian ownership cannot be directly identified) to a significant extent relieves round-trip investors from the need to conspire their OFC-related activities in other ways.

Discussion
In this paper, we contend that round-trip investment in OFCs is a unique form of FDI and therefore needs better theorization to move beyond the black-and-white mindset on tax haven investment (Deng et al. 2019). We concur with Su and Tan (2018) that in the emerging economy context roundtripping may be a strategic decision as much as an ethical decision-a response to unethically behaving home country institutional constituents' demands. We further argue that as round-trip investment by default involves investment in business activities in the home country, its societal benefits such as provision of employment at least partially compensate for the tax income losses of the home country government, making it ethically a less straightforward question. Unlike in developed economies where the firm's decision to engage in unethical behavior and incorporate in OFCs is predominantly based on profits (Johnson & Holub, 2003), in emerging economies with rampant public sector corruption, weak institutional protection, and government interference (Karhunen et al., 2018) the situation is different. The OFC incorporation per se may be the only way to secure the very existence of the business, i.e. the seemingly unethical behavior of the firm is in fact triggered by unethical behavior of its home country institutional constituents or the inability of the firm to rely on home country institutions.
Examining the OFC round-tripping FDI activities of emerging market firms from Russia and building on the institutional arbitrage literature (Boisot & Meyer, 2008), we developed and tested the idea that these firms do not follow traditional theoretical reasons for FDI (Buckley et al., 2015;Chari & Acikgoz, 2016), nor do they follow the escapism or exploitation logics put forward by many institutional arbitrage theorists (Luo & Tung, 2018;Witt & Lewin, 2007). Instead, we suggest that round-trip investors are looking for different types of opportunities to protect their funds/property that facilitates the reinvestment of these funds/property back into the home country (Russia) in a legal way. We also question the mainstream ethics literature that suggests all OFC investments are unethical (Johnson & Holub, 2003;Payne & Raiborn, 2018), since without the protection provided by OFCs many of these emerging market firms could not grow and prosper, adding substantial benefits to society. Thus, we develop the theoretical notion that EMNEs look for a combination of institutional arbitrage opportunities in OFCs to enable this protection and legal return. Our empirical results provide some support for these ideas.
To gain further insights into the phenomenon of capital round-tripping from emerging economies, we provided empirical evidence on differences in industrial patterns and ownership structures of Russian capital round-tripping via OFCs. First, we had suggested that firms relying on licit and illicit funds might seek different OFC attributes because of the differences in the ethicality of their activities. Our industrial analysis tends to confirm our proposition that real estate and financial firms, which authorities suggest rely on illicit funds and are more likely to be involved in money laundering via OFC (Ardizzi et al., 2014;Barone & Masciandaro, 2011;OECD, 2007), tend to invest in OFCs providing stronger property rights which helps protect illicit funds and activities from home country scrutiny. Other round-tripping firms, relying on licit funds, tend to seek OFCs providing low or no taxes allowing these firms to reinvest more funds back home. This tends to suggest that, at least for emerging market firms, the ethics literature, that mainly condemns OFC investments for tax reasons, may need to rethink the approach since those round-tripping firms seeking low taxes often provide additional societal benefits when the finds are returned back home. Instead, it might be wiser to look at OFC investments seeking property rights protection if one were to try and identify firms attempting to hide involvement in unethical behavior.
Second, we note that not all OFC investments are wholly owned. We therefore examined differences based on ownership patterns (wholly owned versus joint venture investments). Our analysis demonstrates that for jointly owned operations low tax and secrecy are important determinants of OFC location. This contrasts with those involved in wholly owned ventures. These wholly owned firms tend to seek OFCs with both low taxes and strong property rights protection, but do not seek secrecy. Based on this, it seems that firms involved in joint ventures may be trying to reduce their tax burden and hide this activity from the home government, as previous ethics research tends to suggest. However, round-tripping investors using wholly owned entities do not seek this secrecy protection and may be acting more ethically. These firms appear to follow our theoretical suggests of seeking low taxes and strong property right in order to improve the competitive position of the firm and provide additional societal benefits when reinvesting the funds back into the home country.
In this way, we make several contributions to knowledge. First, we contribute to the small but growing literature that views institutional differences as opportunities as opposed to the more traditional problematic attitude literature (Boisot & Meyer, 2008;Chari & Acikgoz, 2016;Gaur & Lu, 2007;Luo & Tung, 2018;Witt & Lewin, 2007). Taking an institutional arbitrage approach, we developed theory to suggest that certain institutional features in OFCs are attractive to emerging market round-trip firms because these features help the firm to protect its funds while allowing the firm to reinvest back into the home country. Our theory suggests that round-trip investors are not driven by institutional escapism or exploitation, but are attracted to specific OFCs providing tax, property rights, and secrecy arbitrage opportunities. Our results provide support to these ideas. Thus, it appears that leveraging these opportunities can help emerging market firms protect its resources from the home government, while at the same time allowing these firms to invest their funds legally in the home market. In this way we extend our understanding of this unique form of FDI and of how the use of institutional arbitrage can provide important opportunities for these firms.
Second, building on existing OFC FDI literature, we developed and tested the notion that round-tripping firms will make decisions on investments into OFCs based on a combination of institutional arbitrage opportunities. Past studies have indicated that tax (Chari & Acikgoz, 2016;Haberly & Wójcik, 2015a), property rights (Pérez et al., 2012) or secrecy (Ledyaeva et al., 2015) arbitrage opportunities provide a motive for investment. We go further theorizing that the combined impact of either taxes and secrecy or property rights and secrecy drives OFC investment decisions.
Finally, our paper contributes to the ongoing debate on the ethical dimension of OFC investments (Kanagaretnam et al., 2018;Su & Tan, 2018;Deng et al. 2019) by focusing on the OFC characteristics that attract investors from an emerging economy to use them for capital round-tripping. On one hand, we suggest and find that part of Russia's capital round-tripping serves unethical purposes such as tax avoidance or money laundering. Indeed, for example, it is widely accepted that Russian oil producers (and Russian producers of exportable commodities in general) have been using intermediary companies in offshore jurisdictions for secondary sale of oil at world prices (initially bought from their parent Russian companies at low prices), the practice that enabled Russian companies to minimize tax on profits paid in Russia. Brovkin (2001) provides a typical example of such a company-one of the largest privately owned oil producers in Russia in the 1990s, Sibneft (currently Gazprom neft, subsidiary of Gazprom, state-controlled company). In addition, according to the latest public report of the Federal Financial Monitoring Service (FFMS) of the Russian Federation published in 2017 (Rosfinmonitoring, 2017) investments in OFCs by Russian finance companies is often a tool used to launder illegally obtained funds.
On the other hand, some Russian OFC round-trip investment is a vehicle for protecting the investors from the unethical behavior of others, including corrupt home country authorities and dishonest competitors. Bulatov (2017) notes that for the protection of property rights, Russian businessmen export capital, for example, to the British Virgin Islands. This gives them the ability to apply to the local court (including London courts) in cases of violation of the company's rights in other countries-including Russia. Sharafutdinova and Dawisha (2017) further discuss the growing reliance of Russian businesses on foreign courts to settle their disputes. In addition, we suggest and find that some OFC round-trip investment from Russia facilitates access to capital markets, which provides funds for firm expansion in the home country. For example, leading Russian online job search portal HeadHunter, a company incorporated in Cyprus, carried out a successful international public offering (IPO) on the Nasdaq exchange in May 2019 allowing the firm to expand its presence in Russia (The Moscow Times, 2019). All this suggests that firms from emerging economies face much more multifaceted ethical considerations when making OFC investments than firms from developed economies, where strong institutions protect firms from opportunistic behavior of others.

Limitations
While providing interesting insights about OFC investment for emerging market round-trip investors, our research suffers from a number of limitations that provide opportunities for future research. First, we used an extensive sample of Russian round-tripping firms. Therefore, our results might not be generalizable to firms from other emerging countries or for round-tripping firms from non-emerging countries. Future research looking at for example samples of Chinese round-tripping firms might see if our results hold in this situation. Alternatively, researchers might identity roundtrip investors from other, non-emerging economies and see if their behavior follows our theory.
Second, while our results indicate that there might be differences between the motives behind OFC investment for firms relying on licit versus illicit funds, we could not precisely identify those using illicit funds. Building on past research we assumed investments in the financial and real estate sectors were mostly made using illicit funds (Ardizzi et al., 2014;Barone & Masciandaro, 2011;OECD, 2007). Future research could extend this work by finding a better measure of illicit funds and OFC investments. Then they could determine more specifically whether the type of funds being invested impacts the institutional arbitrage opportunities sought by these different firms.
Third, there could be other reasons, beside institutional arbitrage, that round-tripping firms select OFCs. We try to control for some of these reasons based on prior research. But we did not consider factors such as institutional affinities or personal connections that might influence the choice of OFC location. Haberly and Wójcik (2015b) mapped the global offshore FDI network and suggested that it is very much shaped by a historical layering of social and political ties such as former colonial relations between, for example, the UK and the former British colonies in Caribbean, or the Netherlands and the Netherlands Antilles. There is also some evidence that institutional affinities created through bilateral ties of specialized professionals might influence OFC location choices. For example, Chinese firms tend to prefer the British Virgin Islands because of the close ties of this destination with Hong Kong. Here, Hong Kong-based lawyers and offshore practitioners specializing in BVI have a key role (Robertson, 2021). Although beyond the scope of our study, these other factors could in fact help explain some OFC location choices, even for round-tripping firms.
Finally, there are few studies looking at the business ethics of OFC investments from emerging economies (Su & Tan, 2018;Deng et al., 2019). Our research tends to suggest that in many cases round-tripping as integral part of such investments cannot be deemed unethical, if measured by the societal benefits it produces when funds are returned to the home country and reinvested into the business. However, there are other occasions when such activities increase the chances of unethical behavior. This suggests that further research is needed to explore the ethics of such activities that make up a large proportion of the FDI of emerging market firms (Sutherland & Anderson, 2015), to help determine in which situations OFC investment should be encouraged and where it should be discouraged.

Conclusion
Overall, our paper adds to knowledge on EMNE FDI by developing new theory to help explaining round-trip FDI activities, explore the way that firms determine the institutional arbitrage opportunities they wish to pursue, and increases our understanding of the ethical implications for these round-tripping activities. We found that round-trip investors are attracted to OFCs with low/no tax opportunities or/and strong property rights particularly when these opportunities are combined with high secrecy. But we found some evidence that round-tripping firms using licit firms seek different institutional arbitrage opportunities compared with firms relying on illicit finds. This has important ethical implications. Thus, it appears that institutional arbitrage, and particularly certain combinations of its elements, is an important mechanism for round-trip FDI via OFCs.
Funding Open access funding provided by Aalto University.

Conflict of interests
The authors have no relevant financial or nonfinancial interests to disclose. The authors have no conflicts of interest to declare that are relevant to the content of this article. All authors certify that they have no affiliations with or involvement in any organization or entity with any financial interest or non-financial interest in the subject matter or materials discussed in this manuscript. The authors have no financial or proprietary interests in any material discussed in this article.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/.