INTERNATIONAL CORPORATE STRUCTURING OF SLOVAK COMPANIES USING TAX HAVENS

Recently a trend has come about of transferring the registered offices of parent companies or their subsidiaries to jurisdictions often marked as tax havens worldwide. This trend can also be seen in Slovakia. Our main aim is to examine the possible links between criteria such as the NACE sector, the jurisdiction category, the size of share capital invested, ownership interests and the number of years from the start of business (company formation) to the year of transferring the registered office to the tax haven, and to justify the tendencies in behaviour from the subsidiaries point of view. The formulation of assumptions is based on the statements of corporate service providers (CSPs), which directly, partially or fully, help companies set up or restructure their international corporate structures by using tax haven jurisdictions. This survey is based on the available database of Slovak companies obtained from the databases of Bisnode Slovakia, ltd. We analysed the available data on 3,483 Slovak companies between 2005-2015 by using both graphical methods and statistical induction methods (mainly Fisher’s exact test of independence) of randomly selected data from the entire database. Our results confirm that there is a statistically significant relationship between the jurisdiction category and three parameters: the NACE sector, ownership interests and the share capital. On the other hand, the relationship between the jurisdiction category and the period of transferring of registered office to tax havens has not been statistically confirmed.


INTRODUCTION
Most researches in this field are focused mainly on tax incentives providing limitations not only in literature overview, but also in the methodology descriptions. Due to the unilateral approach to tax havens through academic community, our investigation is primarily focused on statements of corporate service providers (CSPs). CSPs clearly state that the setup of an international corporate structure is an individual matter based mainly on the desired motives and operations of the ultimate beneficial owner (UBO), meaning that in this field there are no universal solutions, ▪ Istok, M., Kanderova, M., Kristofik, P., Scerba, K. (2019). International Corporate Structuring of Slovak Companies Using Tax Havens. Journal of Competitiveness, 11(1), 22-40. https://doi.org/10.7441/ joc.2019.01.02 only several standardized settings. The shared knowledge of CPSs may play an indispensable role in the research methodology in this field. The most recent investigations have confirmed several claims regarding CSPs, e.g. the importance of interest expenses as a technique of profit-shifting (Ištok & Kanderová, 2019).
Our main aim is to examine the possible links between criteria such as the NACE sector, the jurisdiction category, the size of share capital invested, ownership interests and the number of years the business has been in operation (company formation) until the year of transferring the registered office to the tax haven, and to thus justify the tendencies in the behaviour from a subsidiaries point of view. Based on the limitations of the research, we have opted for the aforementioned indicators to be statistically tested.
The main limitations of our paper are: a low number of academic researches focused on international corporate structure settings besides taxes, availability of relevant data (with the Bisnode database containing only selected jurisdictions), and a long-term orientation on tax purposes irrespective of other important areas of tax havens not only by academicians, but also by the appropriate state officers.

THEORETICAL BACKGROUND
The period from 2005 to 2017 brought a dramatic increase in Slovak businesses that moved their registered offices to jurisdictions often referred to as tax havens. According to available sources, 4,796 Slovak companies were resident in these jurisdictions at the end of 2017, which represents a significant increase compared to the year 2005, when there were 1,510 registered company offices in these jurisdictions (Bisnode, 2018).
A tax haven is a term that can have different meanings and is used for different purposes. There is currently no consensus on a single set of criteria, or an official list of tax havens (OECD, 2013;The Economist, 2013;Akont, 2015a;Janský & Kokeš, 2015). According to the OECD (In Slemrod & Wilson, 2009), a tax haven is a "jurisdiction that does not impose or imposes only minimum taxes and is offered as a place that can be used to decrease tax obligation in a resident state." Other characteristics of tax havens include inadequate sharing information with foreign tax authorities and non-disclosure of regulatory and administrative requirements (Dharmapala, 2008). In some cases, companies are not even required to work or be physically present in tax havens (GAO, 2008). of specially designated regimes (isolated preferential tax regimes) in domestic markets (Dharmapala & Hines, 2009). Durnev et al. (2016) have defined two types of companies as OFCs: the first an offshore company with a registered office in an OFC, and the second as a company that has registered subsidiaries in an OFC, but the parent company is located in jurisdictions that are not OFCs. Interestingly, companies that are directly registered in an OFC are rated (market value) 14% less than onshore companies, while companies with OFC affiliates have an 11% higher recognition than onshore companies. One of the criteria for defining a tax haven is also the Financial Secrecy Index (FSI) (TJN, 2018). The methodology of FSI 2018 is based on 20 indicators, with this ranking is currently probably the best tool for pointing a jurisdiction as a tax haven. The results of FSI 2018 expressly suggest that we can include countries such as the USA, Germany, Japan, Canada or many other developed countries within the range of tax havens. According to Sharman (2010), some offshore jurisdictions have even stricter standards of corporate transparency and disclosure duties than most OECD countries. Considering certain readings of these criteria, the Slovak Republic could also be deemed a tax haven.
The transfer of a registered office to another jurisdiction may be based on three types of factors, namely: internal factors (ownership, size, length of existence, the growth of employees, acquisitions, mergers); factors related to location (head office of the company's owners, characterization); external factors (market size, workforce, government policies, economic conditions (Sosnowski, 2014).
While these internal factors could be considered quantifiable or even in some cases almost immutable, external factors may change, a factor which is crucial to explain all the reasons for relocating a company's registered offices (Brouwer et al., 2002). Besides all of these factors, a number of other crucial motives for the decision to move the company to another jurisdiction can come onto play, as noted by Krištofík et al. (2017) in research based on structured interviews with business executives providing corporate services in the field of international tax planning and the establishment of corporate structures. These motives consist mainly of the anonymity of the ultimate beneficial owner (UBO), the flexible set up of ownership relationships, tax incentives, asset management and individual associated factors (business regulation, central decision-making in multinational companies, etc.) generally irrespective of NACE sector except for tax incentives. In their work, Kubicová & Záhumenská (2017) consider the tax benefits and tax planning as the primary motives. Radu (2012) points out that tax havens play an essential role in international finance in the context of the recent financial crisis, as 50% of international trade passes through these centres. One example is the Cayman Islands, which hold 80% of investment funds worldwide and manage assets worth more than $1 billion. In earlier surveys we can see a number of examinations of tax havens in which public companies play a significant role leading to their use, e.g. Harris et al. (1993) report that between 1984 and 1988 the tax burden on public companies in the US for tax-benefit operations was significantly lower than that of other companies. Dyreng & Lindsey (2009) found that US public companies operating in tax havens between 1995 and 2007 had 1.

2.
3. a total tax burden of 1.5 percentage points lower than companies with no operations. Markle & Shackelford (2012) are not the only ones to examine the cross-border impact of tax havens, and their findings prove that the tax burden on public companies operating in tax havens is significantly lower than that of public companies without such operations. Most studies (e.g., Lee & Swenson, 2012, Maffini, 2010 focus almost exclusively on public companies, but it is also necessary to consider the tax behaviour of private companies for its dominant role in many European countries. Berzins et al. (2008), for example, show that more than 90% of privatelyowned companies are registered throughout Europe and that they produce four times higher revenues and control twice as many assets as publicly listed companies. The analysis of the tax experience of private companies is therefore a very interesting topic to be investigated not only in the field of total tax burden.
Considering the issue from a further point of view, attention should also be paid to the legal form of entrepreneurship which, according to Kislingerová (2004), is a long-term factor regarding finance. According to Scholes et al. (2009), the organizational and legal form is a crucial business issue (both onshore and offshore). Erickson et al. (2009) point out the choice of an optimal organizational form in the tax planning process regarding the field of business, respectively business activity. Brownle et al. (2001), however, point out that the legal form of business together with the choice of jurisdiction in which the new venture is based will have a significant impact on future tax obligations. The impact of choosing a business location is one of the ways to achieve the ability of competitiveness, thanks to the possibilities of optimizing or reducing costs, creating new opportunities for growth and developing new strategic options (Anken & Beasley, 2012;Suder, 2011).
The most often and most commonly used legal form of business for companies that move their registered offices to other jurisdictions is holding. Holding is a business association with more businesses that are linked to each other by equity participation (Majtán et al., 2005;Vlachynský et al., 2009). Capital connections can arise between different legal forms. Significant effects of the holding structure include the breakdown of business risk into individual entities and business areas. Under Slovak jurisdiction, the term "holding company" is not regulated, but it uses the term of consolidated entity and the controlled and controlling entity (Act No. 431/2002Coll., on Accounting, Commercial Code No. 513/1991Coll., Act No. 595/2005on Income Tax and Act No. 563/2009 Coll. on Tax Administration). By using the holding, certain assets can be transferred to the so-called tax haven, and that holding then collects revenue from the essential investments, patents or other rights and loans (Kassay, 2006). The use of offshore holding companies can offer significant tax savings when it comes to companies engaged in international business. The establishment of an offshore company itself without a sophisticated tax concept and professional analysis of a business entity is not enough and can instead lead to problems as to desired and planned benefits (Akont, 2014b).

RESEARCH OBJECTIVE, METHODOLOGY AND DATA
The aim of this contribution, based on the analysis of available database of Slovak enterprises with ownership links to tax havens, is from selected criteria to analyse the possible links between them and to justify tendencies in behaviour of the subsidiaries from various perspectives. Under Slovak conditions, data on Slovak businesses linked to tax havens are not widely available. In this contribution, we analyse the database of Bisnode Slovakia, ltd., which lists the Slovak companies (3,483) with ownership links to tax havens, under which we understand the combination of parent and subsidiary companies in selected jurisdictions from 2005 to 2014, respectively to 07/2015.
We analysed the enterprise database available from the criteria of the NACE sector, the category of jurisdiction, the size of the company (the size of the share capital invested) and the number of years from establishment of the business until the year in which the company's registered office was transferred to the tax haven. The absence of a generally accepted definition of tax havens, or the division of jurisdictions into offshore, midshore and onshore allows us for a relatively free categorization of jurisdictions. Authors most often divide jurisdictions into only offshore and onshore categories. Offshore jurisdictions are most commonly referred to as "pure tax havens" or "OFC" (offshore financial centres), especially small tropical island jurisdictions. Regarding the rest of the jurisdictions, we look at a different approach when we divide tax havens further into midshore and onshore categories. With the selected classification, the midshore and onshore categories can easily be combined and interpreted as one onshore category. Another reason to use the midshore category for statistical testing is the fact that this group of jurisdictions is mainly used in international tax optimization and therefore we wanted to get a separate view primarily on tax incentives corporate structures. At the same time, it is important to note that this is only an initial categorization because in some jurisdictions, it is possible to establish up to three types of businesses -onshore, offshore and a so-called FREE ZONE. Examples are, e.g., the United Arab Emirates or Mauritius where a resident company (onshore -LLC), an offshore company (e.g., RAS AL KHAIMAH -RAK) or a FREE TRADE ZONE COMPANY (FTZ) can be established (e.g., in Dubai).
Based on jurisdiction categories and selected indicators, we formulate the following partial hypotheses: H1: We assume that there is a statistically significant relationship between the NACE sector and the jurisdiction category.
H2: We assume that there is a statistically significant relationship between the ownership interest share and the jurisdiction category. H3: We assume that there is a statistically significant relationship between the share capital invested and the jurisdiction category.
H4: We assume that there is a statistically significant relationship between the share capital invested and the ownership interest.
H5: We assume that there is a statistically significant relationship between the jurisdiction category and the period of transferring of registered office to tax havens.
To track possible connections, we used the Pivot Tables, where we have calculated absolute numbers either relative in column or relative to row, which allow a reasonable comparison. By suitably chosen graphical means and their visual assessment, it was possible to identify the possible link between the selected criteria in the whole basic set. To use the methods of statistical inference, from the base file, about 10% of enterprises were randomly selected, using the SPSS software product to obtain a random sample. To test dependence hypotheses, we used Fisher's exact test of independence (significance level = 0.05). Random sample as well as another hypothesis testing was carried out in the SPSS software package.

RESULTS AND DISCUSSION
As the main motives of establishment or restructuralization of international corporate structures are highly individual, considered to be a behavioural assessment, it is then very difficult to analyse them based only on the financial statements available. The justification of exact motives can therefore be slightly inaccurate. This is why we talk about trends in the behaviour.
Nevertheless, some trends in corporate behaviour can be noticed and partly justified using international corporate structures. Based on the data available as of 07/2015 (Figure 1), we see the structure of Slovak enterprises that moved their registered offices into a given jurisdiction by dividing the NACE sector shown as a share of the subsidiaries in the total number of enterprises in the given NACE sector expressed in a profile. It is clear from the graph that the most significant share is made up of enterprises in Financial and Insurance Activities (45‰) and Real Estate Activities (44‰). The smallest share is made up of businesses in Agriculture, Forestry, and Fishing (5‰) and Human Health and Social Work Activities sector (1‰). Relocation to tax havens is technically possible due to the smooth transfer of capital. While almost all businesses can secure the anonymity of the ultimate beneficial owner (UBO) regardless of the NACE sector, tax planning and tax optimization in the area of corporate income tax (corporate income tax of legal entities) is technically impossible for some business areas, for example for domestic wholesale and retail, rental property and related activities (e.g. cleaning and guard service), catering, driving schools, the vast majority of handicraft activities and so on (Krištofík et al., 2017). In the following section, we analyse businesses that moved their headquarters to another jurisdiction (creation of an ownership link to a tax haven). We have data on 3,483 businesses. The breakdown of enterprises by NACE sector is shown in Table 1. Of the businesses that moved their registered office to the selected jurisdiction (by Bisnode company), there are also companies from the Wholesale and Retail Trade (25%) and Professional, Scientific and Technical Activities (22%). The share of enterprises in total number of enterprises divided by sectors under of the Financial and Insurance Activities sector, is less than 1%.
The analysis further verified the possible link between the jurisdiction category and the NACE sector. We will use a visual assessment based on the graphical view to verify a possible relationship by verifying Fisher's exact test. For a better overview and clarity in Figure 2, we only show sectors with a share of more than 1% in the total number of subsidiaries.  Based on visual assessment, we can see that the NACE sector Transportation and Storage and Manufacturing and the Financial and Insurance Activities sector are typically based in onshore jurisdictions (72% to 81%) with minority representation in offshore jurisdictions, while for the sectors shown on the right in Figure 2, there is not a clear preference for one category of jurisdiction. Fisher's exact test results confirm a hypothesis of a statistically significant relationship between the NACE sector and the jurisdiction category (p-value 0.013). From the results, we can observe certain sector-specific tendencies regarding the selection of the type(s) of the jurisdiction. For the sectors located on the right of the chart, an increased ownership is noted by parent companies from offshore jurisdictions. Direct ownership of an offshore company can be interpreted in a two-way view. The primary objective of this property holding is the anonymity of an ultimate beneficial owner (UBO), i.e. the ownership by the offshore company directly is sufficient. The second way can be the opposite direction, i.e. a foreign company will establish a subsidiary in Slovakia, for example, to reinvest the accumulated funds from an offshore jurisdiction. While both motives can be achieved by onshore or midshore companies, we can assume that an offshore company is used mainly for two reasons. The first reason is the lower costs associated with setting up and managing offshore companies compared to the onshore company.
The second reason is that optimized (sometimes called "universal" or "general") holding structures end up directly offshore by a company that owns the onshore company and then moves the subsidiary to Slovakia. Offshore companies in optimization (universal) holdings are called "grandparental companies", offshore or midshore is then a parental company and Slovak business is a subsidiary. The second reason is that the accumulated funds in the offshore company (entirely anonymous from the point of view of the UBO) are reinvested, that is, the UBO has several options either to pay a dividend to a natural person, to make capital transfers for legal entities or to acquire assets (buildings, new companies, expanding business in other jurisdictions, etc.) and so on. It is just a specific identification of behaviour based on historical data available because onshore and midshore companies have more extensive use than offshore companies, only that they are more expensive. The direct link between a Slovak company and an offshore company has, in addition to its advantages, also disadvantages, namely the negative perception and image of such a structure, and the fact that offshore jurisdictions are non-contractual entities for Slovak enterprises, i.e. a 35 % withholding tax is applied. In Table 2, the number of subsidiaries is based on the category of jurisdiction and the proportion of the ownership interests, as well as the relative share of the enterprises (%) in each jurisdiction category, according to their share. The 10% threshold was chosen as a condition for granting the parent company's status under Article 3 (ii) (The EU Parent-Subsidiary Directive 2011/96/EU -the Directive modifies the exemption of dividends and different redistribution of profits paid by the subsidiaries to their parent companies from the withholding tax and the exclusion of double taxation of such profits at the level of the parent company) for tax purposes. The 25% ownership threshold is determined by fundamental reporting and disclosure requirements for a UBO (ultimate beneficial owner) FATF (2014). When setting up a corporate structure, the 25% ownership threshold is deliberately circumvented at some ownership levels. In the past, low ownership interest shares were often used to circumvent the ban on chaining the businesses (a company with a sole shareholder cannot be the sole founder or sole shareholder of another company) (Article 105a (a) of the Commercial Code).

Fig. 3 -Allocation of subsidiaries in the jurisdiction by ownership interest classes. Source: customized processing of data of Bisnode Slovakia, ltd. as of 07/2015
The graphical representation of subsidiary distribution in individual jurisdictions according to the proportion of the ownership interest shows the possible relation of the category of jurisdiction and the ownership interest. For offshore jurisdictions, the prevalence of enterprises with 100% ownership interests and a minority of enterprises with ownership interest of less than 10% is typical. With onshore jurisdictions, the proportion of enterprises as per their ownership interest is more proportional compared to the offshore category. The statistically significant relationship between the ownership interest share and the category of jurisdiction was confirmed by Fisher's exact test (p-value 0.007). The direct use of an offshore company is mainly for anonymity of the UBO, so the predominance of higher ownership interests is identical to the expected prediction. In the case of onshore and midshore jurisdictions, there are possibilities to use profit-shifting methods and techniques. Onshore and midshore companies can benefit from signed double tax treaties (DTT) and the EU Directive when they are EU companies (The Merge Directive, The Parent-Subsidiary Directive, The Savings Directive and The Interest and Royalties Directive). The most significant differentiation of distribution is visible in onshore parent companies, which could also be assumed due to the representation of the Netherlands and Luxembourg, which are typical jurisdictions for the use of optimization and universal holdings.
In the midshore category, this role is mainly under the jurisdiction of Cyprus and Malta. Selected jurisdictions from midshore and onshore jurisdictions offer the most comprehensive application not only in the field of tax optimization but also asset management and flexible arrangement of ownership structures (at the same time it is possible to achieve the UBO anonymity by using the so-called nominee service). More extensive use of onshore and mid-shore jurisdictions has also resulted in their division of use.
In the next part of the analysis, we investigated whether the selection of the jurisdiction category is influenced by the size of the firms measured by the amount of the share capital invested (Table  3) as well as whether the ownership interest differs depending on the size of the firms ( Large corporations (enterprises with a high level of invested share capital) have their registered offices almost exclusively in onshore and midshore jurisdiction categories compared to enterprises with a lower value of the share capital invested. The analysed data show that the choice of jurisdiction is conditional on the size of the business. The Fisher's exact test results confirm the hypothesis of a statistically significant relationship between enterprise size and jurisdiction category (p-value 0.007). The results are again the same except for larger businesses that can afford higher costs for building an international corporate structure and managing it. Higher investment also leads to a broader set-up of the corporate structure, and they are onshore and midshore companies at the first level (parent company). Thus, it is possible to derive the direct relationship between the amount of share capital, investments in a corporate structure, the complexity of setting and range of benefits. It also applies (as before) that the higher the share capital, the higher the investment. The bigger the company is, the more potential investment to an international corporate structure and the higher potential range of obtained possibilities regarding motives. On the first ownership line of an offshore company, it is cheaper but there is a limited use for anonymity purposes only. With onshore and midshore, then it is more expensive, but much more possibilities exist regarding flexible arrangements of motives. The survey results on this level show us the same values as we expected. Thus, greater differentiation. Small businesses (up to 10,000 EUR of invested share capital) have a typical 25% ownership interest and higher. The graph shows the possible relationship between the share capital invested and the ownership interests. The above dependence was not explicitly confirmed by Fisher's exact test, where the p-value was 0.044.
From the analyzed group of companies, 29% of the companies transferred their registered offices to selected jurisdictions in the year of their establishment, and more than half of the subsidiaries moved their registered offices within three years of their establishment. We can say that a significant portion of Slovak companies could be established as one of the elements of the overall arrangement of a corporate structure, respectively the owners decided to use the benefits provided by foreign jurisdictions right after the establishment of the Slovak company. An early move to selected jurisdictions can be interpreted in two ways -from the Slovak company towards abroad, but also from a foreign company to Slovakia. If, for example, the newly established Slovak company has the potential for significant growth, anticipates a significant amount of future investments or increased turnover, it can move its registered office in selected jurisdictions right away to achieve specific benefits (establishment of an international corporate structure). On the other hand, however, Slovak companies may be used to redistribute accumulated profits from abroad (in regard to results, significant representation), not only by Slovak UBOs but also by foreign ones. In the past, it was, for example, frequently used by Austrian natural persons and legal entities to avoid, or reduce dividend taxation, and so on. Up to 60% of the registered offices was within three years, from which there can be deduced and assumed a strong dependence on the use of techniques and methods to achieve relocation benefits to selected jurisdictions. Within three years, the most prominent representation in onshore jurisdictions, adding midshore, makes it clear that the companies that are established are one of the elements in the construction of an international corporate structure. We subsequently monitored whether the choice of the jurisdiction category is related to the length of the company's formation until the transfer of the registered office to the selected jurisdiction. The classification results are in Table 5. Companies that have moved to selected jurisdictions within three years of their establishment are choosing more offshore jurisdictions than companies that moved their registered offices more than ten years after, preferring midshore and onshore jurisdictions. The Fisher's exact test results did not confirm the hypothesis of a statistically significant relationship between the period of transferring the registered office to tax havens and the jurisdiction category (p-value 0.068).
If a company uses an offshore company directly as a parent company, the assumption could be that an international holding was created primarily for anonymity of the UBO, or for reasons of anonymous administration, respectively protection of assets. In this way, tax planning and optimization processes are possible from the tax point of view, i.e. holding (structure) is referred to as "optimization". A universal holding (structure) refers to a holding in which a Slovak company is owned by a company of onshore or midshore jurisdiction, which in turn is owned by an offshore company. If the parent company is a direct offshore company, this means that it is a non-contractual company from the Slovak point of view, i.e. a 35% withholding tax is applied (according to the Slovak Income Tax Act). The highest year-on-year increase of subsidiaries in offshore jurisdictions was in 2012, when the number of companies increased by about 50% over the previous year. By 2015, the year-on-year increase was in the range of 14% to 24%. In the last two years, there has been a year-on-year decline in the number of companies in offshore jurisdictions. There has been no such significant year-on-  year increase in onshore jurisdictions. In the last five years, the range is from 0% to 6%, except for 2015, which showed a one-digit decline over the previous year. In recent years, significant momentum has developed in the fight against aggressive tax planning and the limitation of tax evasion methods and techniques. In 2016, the automatic exchange of tax and banking information began to operate in over 100 countries. In the last three years, the construction of corporate structures has evolved from the use of direct equity links (a significant year-on-year decline as in Figure 6). There has even been a negative shift in offshore jurisdictions, mainly due to poor public image and relatively light usage. On the other hand, corporate structures with indirect equity linkage (no ownership link to tax havens) are used, so this link is no longer only in the form of a typical international holding (so it cannot be captured in the business register). The analysis of such uses is significantly more challenging than that of direct ownership links. Even if there has been a reduced rate of increase in the number of these links, it does not mean that the use of international corporate structures has decreased. This will be a question for further research.
Despite worldwide trends, Slovak companies continue to show gradual increases in direct ownership links to tax havens compared with, for example, the Czech Republic, where in recent years decline has been shown, a finding which has been considered paradoxical (Rohan and Moravec, 2017). One open field for potential future investigations is the list of tax havens prepared by Bisnode, ltd. This number of tax havens listed is relatively short, thus it would be crucial to add many additional jurisdictions for a broader analysis, e.g. in the UK, which is often used in tax planning through an extensive network of double taxation treaties (with more than 130 signed double tax treaties). According to one research (Krištofík et al., 2017), the direct equity linkage (ownership link) is employed by approximately 20-30% of Slovak enterprises which operate their business in selected tax havens. As a result, around 14,500 -24,000 Slovak businesses were utilizing tax havens by the end of 2017.
Due to the lack of data in this field, our approach to the data analysis can be considered unique and to our best knowledge, we are not aware of any similar comparable research in other countries. The perception of tax havens is controversial. Many researches have been carried out to prove that companies based in tax havens are reporting lower effective tax rates or using selected profit-shifting methods and techniques (e.g. interest expenses). In our opinion, one future research direction should be a comparison of the economic performance of companies (both ex-post and ex-ante financial analysis) with direct ownership links to tax havens (direct equity linkage) compared to their counterparts (no ownership link to tax havens). The methodology of this type of research could be inspired by Durnev et al. (2016), but in Slovak conditions (the low number of Slovak companies with shares listed on stock exchanges), the economic value added (EVA) indicator could be rather used instead of the market value indicator.
Another necessary direction in future research is to determine why the number of Slovak companies with direct ownership links to tax havens is constantly rising despite the fact that Slovakia is also committed to automatic tax and bank information exchange agreements (in operation since 2016). The results in this field can be inspired by the approach of Rohan & Moravec (2017). According to these authors, TIEAs (tax information exchange agreements) at least as a preventive tool have a direct impact on the behaviour of Czech MNEs in relation to tax havens, with reports showing a slight decline of Czech companies with direct ownership links to selected tax havens over the last few years. Based on our previous research, we can assume that the Slovak entrepreneurs either do not have sufficient information in this area or they simply do not feel insecure regarding taxes due to relatively low quality of the Slovak business environment, mainly in terms of the enforceability of law and the weak performance of the institutions.

CONCLUSION
The creation of international corporate structures is highly individualised regarding the desired benefits, and therefore the justification of the exact incentives by the data is quite inaccurate. However, some trends in the use of international corporate structures as well as the behaviour of enterprises can be noticed. We have therefore decided to conduct this survey with Slovak companies, as research in this field is not common in the country. Our study verified several assumptions we put forth from our theoretical background research. The most significant share of enterprises out of a total of 3,483 companies that transferred their headquarters to another jurisdiction came from the NACE sector Wholesale and Retail Trade, i.e. Repair of Motor Vehicles and Motorcycles with 25.19% and Professional, Scientific and Technical Activities, with a share of more than 22%.
Since the literature does not provide a precise definition of tax haven divisions, we chose to equitably divide them into offshore (especially small tropic island states), midshore and onshore jurisdictions.
The results show that our decision to categorize all tax havens besides offshore jurisdictions into midshore and onshore jurisdictions was correct, as in this way we were able to obtain more precise results to be interpreted. Our research shows that in these jurisdictions by the end of 2017 up to 24,000 Slovak companies were running a business, of which only 20-30% used direct equity links (international holding -direct ownership link). The analysis also showed some sector-specific trends in the selection of the jurisdiction type category (onshore, offshore, or midshore). These tendencies of relationship regarding the type of jurisdiction and the NACE sector were confirmed by Fisher's exact test (p-value 0.013). The division of companies by jurisdiction category itself reveals a tendency that is also statistically confirmed (p-value 0.007) for most of the 100% ownership interests, with offshore jurisdictions shown as typical. In midshore and onshore jurisdictions, the distribution of businesses is balanced. The choice of jurisdiction also relates to the size of the enterprise measured by the amount of the share capital invested, as confirmed by Fisher's exact test results with p-value 0.007. Large corporations with a high share capital invested are located almost exclusively in the onshore and midshore jurisdiction categories, a finding which corresponds to our expectations, as such businesses can afford higher costs for the construction of an international corporate structure. Another interesting finding is that 29% of companies transferred their registered offices to selected jurisdictions in the year of their establishment, and more than 50% of enterprises did so within 3 years from their establishment, by which the decision to use the benefits provided by foreign jurisdictions can be seen as a fundamental element of the overall setup of an international corporate structure. The last researched area was the year-on-year change in the number of enterprises in selected jurisdictions, where we found that the highest year-on-year growth of about 50% over the previous year was shown in offshore companies in 2012. Over the last two years, we have witnessed a year-on-year decline in the number of companies in offshore jurisdictions, and a minimal increase in onshore and midshore jurisdictions. These changes seem to have come about in recent years due to the onset of the automatic exchange of tax and bank information since 2016, the transition from direct to indirect equity linkage (or controlling rights), along with the negative public image. The present research output could be potentially used primarily by Financial Administration of the Slovak Republic, or in another researches focused on the topic of tax havens.