Transfer pricing and tax avoidance

(1)

UNIVERSITY OF GRONINGEN

Transfer pricing and tax

avoidance

Evidence from the Netherlands

Christian Penchev S2721252

6/16/2015 Study Programme:

International Financial Management

Supervisor: Dr. Hein Vrolijk Co-assessor: Dr. Raymond Zaal

Key words: Transfer pricing, tax avoidance, royalties, license fees, special purpose entities, intellectual property

(2)

- 1 -

CHAPTER 1 Introduction

Multinational companies (MNCs) have come under increased scrutiny because of the emerging adverse consequences for tax revenues in high-tax countries, and the enhanced opportunities to lower taxes which provides MNCs an advantage over domestic companies (Buettner, Overesch, Wamser, 2014). Consequently, host countries have unilaterally introduced tax initiatives and rules which limit the ability to shift profits abroad and restrict MNE tax planning opportunities. These initiatives are mainly led by local government authorities and international regulators like the Organization for Economic Co-operation and Development (OECD) for example (Buettner, Holzmann and Wamser, 2014). While research has focused on the influence that stricter tax regulation has on firm behavior (Azémar, 2010; Barrios, Huizinga, Laeven, and Nicodéme, 2012; Buettner, Holzmann, Overesch and Schreiber, 2013), it is largely unclear whether such restrictions are detrimental to the imposing country’s economy. Knowing the effects on host country economy can be key for government officials when deciding whether to tighten international transfer pricing policy. Moreover, for supranational organizations it is important to know why host-countries might be reluctant to address loose international transfer pricing regulations. Buettner, Holzmann, and Overesch (2014) sought to identify whether an increase in international transfer pricing rules and thin-capitalization regulation (i.e. the control over how much of the firms’ capital consists of debt1) influences FDI inflow. While a strong relationship between thin-capitalization regulation and FDI inflow was found, stricter anti-tax-avoidance rules did not seem to significantly affect FDI inflow. My research intends to build upon the research by Buettner, Holzmann and Overesch (2014) by arguing that international transfer pricing regulation does have an influence on FDI inflows and outflows. I suggest that by looking into a specific part of foreign investments, namely royalty and licensing fee payments in a dynamic environment in terms of transfer pricing regulation, a relationship can be identified. Royalties and licensing fees are by definition compensation for the utilization of property, usually copyrighted works, patented inventions, or natural resources, expressed as a percentage of receipts from using the property or as a payment for each unit produced (Fawcett and Torremans, 1998). Moreover, these compensations are for goods and/or services which have an arm’s length market price which is difficult to determine (Buettner, Holzmann,

1

(5)

- 4 -

and Wamser, 2013). As a result, this enables organizations to shift profit around more freely and makes them more likely to intensively participate in tax planning.

For the purposes of my research I will use the Netherlands as a national focal point. During the 1970s the Dutch government began offering advance-pricing agreements in order to attract MNCs. Under those agreements, organizations agreed to leave a part of their income to be taxed in exchange for being allowed to re-route profits through the state (Weyzig, 2013). The ways MNCs re-route profits nowadays however have become more complex. The use of Special Purpose Entities (SPEs) in tax planning schemes has been a great focus of discussion both on a national and supranational level, mainly because of their dubious nature.

Gorton and Souleles (2007) define an SPE as: “a legal entity created by a firm, known as the sponsor or originator, by transferring assets to the SPE, to carry out some specific purpose or circumscribed activity, or a series of such transactions. SPEs have no purpose other than the transactions for which they were created, and they can make no substantive decisions; the rules governing them are set down in advance and carefully circumscribe their activities. Indeed, no one works at an SPE and it has no physical location”. De Boo and Van Vernooij (2014) extend this by adding a definition especially for SPEs used for Royalties and Licensing fees (R&L) and define it as: “entities that are specialized in the collection of worldwide revenues derived from the usage of intellectual property rights2 or trademarks by affiliated or unaffiliated entities. These companies act as a cashier on behalf of their parent company in the invoice of the royalties and license fees (on the basis of sublicenses3), or collect these royalties and license fees on their own account. The latter group normally owns the Intellectual Property Rights themselves, whereas in the first case the SPE only owns sublicenses. What makes these entities special purpose entities is the fact that these entities are regarded as brass plate entities. They most commonly do not have employees and the revenues are redirected (in the form of profits or royalty payments) to a parent company outside the economical territory.”

Even though these SPEs have no physical presence in the Netherlands, the fact that they are

2

De Boo and van Vernooij (2014) used the term intellectual property rights and trademarks (IPRs) when the intellectual property assets referred to can be either produced – intellectual property products (IPPs) – or non-produced – intellectual property (IP) – assets.

3

(6)

- 5 -

administratively registered in the country enables their parent companies to use them for tax planning purposes. One of the advantages of the Netherlands in this perspective is that the state does not impose withholding tax. Withholding tax is used by governments to tax payments like interest, royalties, dividends and license fees. Furthermore the Dutch government has an extensive network of tax treaties, which means that MNCs in the majority of cases are exempt from double taxation in another country and thus have the possibility to minimize their tax base by using such international cash flows(Ruf and Weichenrieder, 2009). Taking this into consideration I want to answer the following question:

Does an increase in transfer pricing regulation have an effect on the import and export of royalties and license fees into/from the Netherlands?

I consider two things as key to the relevance of this. Firstly recent discussions revolve around the report introduced in September 2013 at the G20 summit, where the council endorsed an Action Plan prepared by the OECD – Base Erosion and Profit Shifting (BEPS). The plan itself faces emerging issues with letter-box companies4 and transfer pricing regulations for intangibles and it looks to strengthen anti-tax avoidance rules. Furthermore the European Union is looking into reforming directives such as the Parent-subsidiary directive5 and the Interest and Royalty payment directive6. While pushing for reforms EU Commission has launched investigations on Gibraltar, and the Netherlands, Luxembourg and Ireland because they might grant unlawful subsidies (state aid) to (specific) corporations with their tax system (Henn, 2013). Secondly my research is relevant in terms of investigating the effect of intellectual property as a tool that is directly and indirectly used to generate cash flows (Munari, Odasso, and Toschi, 2011), and whether those cash flows are used by companies’ management in their tax planning activities. Intellectual property in itself captures a broad spectrum of intangible assets among which: copyrights, publishing or other neighboring rights, patents, trademarks, licenses and brand names(Hettinger, 1989

4_{Trivial name of Special Purpose Entities, implying the fact that the company is physically non-existent.} 5

The 1990 Directive was designed to eliminate tax obstacles in the area of profit distributions between groups of companies in the EU by abolishing withholding taxes on payments of dividends between associated

companies of different Member States and preventing double taxation of parent companies on the profits of their subsidiaries.(European Commission,2003)

6

(7)

- 6 -

(8)

- 7 -

CHAPTER 2 Literature review

2.1 Intellectual Property

It is important to develop a detail understanding of what intellectual property(IP) consists of before researching it in a more specific context. The World Intellectual Property Organization (WIPO) defines intellectual property as: “creations of the mind: inventions; literary and artistic works; and symbols, names and images used in commerce.” Intellectual property is protected by law by i.a. trademarks, patents, copyright, which enable the proprietors of the IP to earn both monetary and non-monetary (recognition) benefits from the things they invent or create. This in itself suggests that not everything can be classified as IP, but only the creations and inventions that are “original works of authorship” (Brittin, 1982).

Now that a concept of what IP is was constructed it is interesting to investigate the reason behind its existence. Many authors discuss the benefits of intellectual property(i.a Hughes, 1988; Kanwar and Evenson, 2003; Arrow 1962), and identify it as one of the main drivers of technological progress. By enforcing the creators property rights IP prevents the public from using the intellectual asset, thus giving the creators the possibility to benefit from their invention. The OECD report of 2007 builds on those assumptions stating that intellectual property rights provide owners with monopolies over investments in their innovative intangible assets.

2.2 Types of Intellectual property

Now that I have established what intellectual property is and what are its main functions I can go even deeper and give clarity for the main intellectual property types. The World Intellectual Property Organization defines 5 types of intellectual property: patents, trademarks, industrial design, geographical indication and copyright.

2.2.1 Patents

(9)

- 8 -

step” that could not be deduced by a person with average knowledge of the technical field.” While the rights are absolute the patents have limited protection in terms of time. Nevertheless they incentivize individuals by acknowledging their creativity, giving the creator the possibility to benefit from her/his invention by commercializing it. It is also within the rights of the patent holder to permit other people (usually through licensing) to use the object of the patent. The proprietor is also able to sell it to someone else, who then becomes the legal owner of the patent. Ghafele and Gibbert (2012) outline the commercialization of patents and deduct that it highly depends on the enforcement of Intellectual Property rights from governmental institutions as well as the “complexity” of the market itself.

2.2.2 Trademark

The second type of intellectual property is trademark. Carter (1990) defines a trademark as: “a recognizable sign, design or expression which distinguishes products or services of a particular trader from the similar products or services of other traders”. This enables consumers to distinguish and buy a product or service which exactly matches their needs, and expectations in terms of quality, which are based on the trademark itself (WIPO). This gives the owner of the trademark a chance to maintain a product line that is reliable and a loyal consumer base, resulting in profit from the intrinsic value that the trademark adds to his organization (Dowling, 2006).

2.2.3 Industrial design

(10)

- 9 -

2.2.4 Geographical Indication

A geographical indication as the name itself suggests is a sign utilized on goods that have a precise geographical origin (WIPO). Hence as intellectual property this sign serves the purpose to inform the consumer of certain qualities that the product contains because of its geographical origin. For example only the sparkling wine from the French region of Champagne can use the label “Champagne”, thus the consumer is informed that the sparkling wine is made following age old traditions and recipes, typical for the French province. Geographical indications have become more and more important as the global middle class population continues to rise and global agricultural trade becomes crucial (Jay and Taylor, 2013). Geographical indications are often compared with trademarks, the difference being that all producers of a product in a certain geographical area can use the same indication, where as only the proprietor of the trademark rights can use them.

2.2.5 Copyrights

Copyright laws and regulations give the enable actors, authors and different other creators to protect their works. The World Intellectual Property organization encompasses the works that can be subject to copyright. Those works include but are not limited to: novels, poems, plays, reference works, newspapers, advertisements, computer programs, databases, films, musical compositions, architecture, paintings, choreography, photographs, drawings, sculpture, maps and technical drawings. Copyrights are essential as they give an incentive to creativity and innovation. As trademarks they give the creators recognition and an opportunity to be economically rewarded for their efforts.

2.3 Transfer pricing, regulation, corporate tax rates and tax havens

(11)

- 10 -

2.3.1 Transfer pricing and transfer pricing regulation

In principle a transfer price, is a book- keeping term, which is used to value transactions between associated companies which are under the same management at artificially high or low levels in order to effect an unspecified income payment or capital transfer between those companies.(OECD, 2003). In academic literature (e.g. Horngren et al., 2002; Atkinson

et al., 2004) ‘transfer pricing’ is commonly stated “as the price that a company charges for a product, service, loan and the use of intangibles to a affiliated organization, including a division, subsidiary, affiliate or a joint venture. It acts as a device for the allocation of costs, income, revenues and profits to various subunits. Traditional views conjure up images of transfer prices in relation to tangible goods and services. Yet, in a world where executives are under pressure to produce higher shareholder value, transfer prices, and the associated opportunities to construct them in tax-minimizing manner, may cover leasing, intellectual property, royalties, interest payments, expenses, fees, management charges, advisory services and virtually everything that a company can buy or sell.”

Because these companies are affiliated it is only natural to presume that they will act in each other’s interest, hence the assumption can be made that transfer prices can be set arbitrarily by the parties involved. Thomas (1971) among others presents five ways in which transfer prices are determined in practice.

i. Market based pricing- in this case one entity sells the product or service to the other based on the current market price of the intermediate product(where there is a competitive market available,

ii. Marginal cost pricing – Often there is no market available for the intermediate goods and services that are bought and sold between the divisions of an organization, the transfer price should be the marginal cost, which is normally the short-term variable cost. Setting transfer prices equal to marginal costs enables managers to identify the output levels that will maximize profits. Issues might arise if managers do not have accurate cost information.

iii. Full cost pricing - a method that is widely used in practice. In this case transfer prices are established by full cost pricing but do not include a profit on the goods delivered for the supplying division.

iv. Cost plus mark-up pricing – in this case the goods are sold at full cost with a mark-up included thus providing a profit for the supplying entity.

(12)

- 11 -

the two parties involved. This process can of course result in any of the above options, and is suitable when there is no competitive market for the goods and services traded.

It is clear that these methods as effective and practical as they are, still leave room for arbitrary pricing. That is why national governmental and supranational institutions (e.g. OECD, EU) create guidelines and/or laws on how these prices should be formed and obligate firms to disclose information and documentation to obtain insight in intra firm transaction. The arm’s length principle for example restricts the arbitrary pricing by stating that intra firm transfer prices should be similar to transactions between independent parties in the open market (Mehafdi, 2000). However, this is problematic because these intra firm transactions are mostly firm specific unobservable or require inside information.

2.3.2 Base Erosion Profit Shifting Action Plan7

The reason behind the BEPS action plan are the increasing importance of media and public debates regarding profit shifting and tax avoidance techniques of multinational companies. Firstly published in 2013 the report itself consists of fifteen action points towards problematic areas within the tax rules and tax avoidance context. Those fifteen points aim to reduce government budget deficits especially in countries with a fiscal deficit as well as reestablish the focus on business contribution (see Table 1a). Progress is currently being made on key parts of the plan regarding intellectual property and transfer pricing and transparency measures that companies will have to abide to. In practice it has been met with enormous political backing and reports from private consultancy and accounting companies (e.g. KPMG, PwC, E&Y, Deloitte) expect that BEPS will have a huge impact on tax policy administration.

2.3.3. Corporate tax, withholding tax and tax havens

Corporate tax rates are a key determinant for governmental institutions, in their effort to create a favorable investment climate in order to attract foreign direct investment (Mehafdi, 2000; Dicken, 1998; Hill, 1998; Elitzur, 1996). The importance of attracting multinational companies and the amount of cross-border investments have grown substantially over the

(13)

- 12 -

period 1980-2000 (Haufler and Schjeldrup, 2000) and while corporate tax is just one of the factors defining the overall investment climate in a country it is certainly one of the key drivers behind management decision-making (Chan and Lo, 2004, Barrios, Huizinga, Laeven, and Nicodéme, 2012). This is widely recognized by governments across the world. In their research regarding OECD member states Haufler and Schjeldrup (2000) notice a downward trend among corporate tax rates while marginal tax rates(the increase in taxes for each additional "dollar" of income) remain relatively stable. Many scholars define these trends as tax competition, where local governments try to tax certain types of investments at low rates in order to stimulate investment activities from multinational companies and gain competitive advantages on a regional or even global scale (e.g. Konan, 1997, Wilson, 1986). Extreme cases of low tax rates form the so called tax havens. Tax havens are countries which offer extremely low tax rates or attractive tax provisions for multinational enterprises (Buettner, Holzmann, Overesch, and Schreiber, 2013).

Before continuing onward it is important to reflect on what has been said so far and relate it to my focal point of research, namely the Netherlands. Firstly taking into consideration that the country has a very competitive corporate tax rate (ranking first in International Tax Rules and 18th in corporate tax among OECD countries8) and a relatively stable economy, it makes it a very attractive target for research especially in terms of transfer pricing manipulation and tax avoidance practices. Furthermore the Dutch state has numerous tax treaties so companies can avoid double taxation as well as a tax policy which very often exempts companies from withholding taxes for their outbound payments, which makes the Netherlands an attractive target for companies that seek to minimize their global tax payments (Goodley and Milmo, 2011). In the next section I will describe and clarify why and how companies may (ab)use intellectual property for tax avoidance purposes and will give a practical example of the so called "Double Irish with a Dutch sandwich" which is a classical tax avoidance scheme.

2.4. Why and how is intellectual property used for tax avoidance purposes

In order to have a broad and objective view of the dynamic at hand one must take a step back and look into the capitalistic ideology on a global scale before going into concrete examples. Corporations rule most of the business world of today, both in a political and social perspective (Korten, 2001; Monbiot, 2000; Klein, 2001; Sikka and Willmott, 2010). They operate within a capitalistic logic that Milton Friedman describes as the only social

8

(14)

- 13 -

responsibility of business "to use it resources and engage in activities designed to increase its profits so long as it stays within the rules of the game, which is to say, engages in open and free competition without deception or fraud." Following this line of thought we can assume that while staying within the rule of law the most important priority of a business entity as such is to maximize its profits, one way of which is minimizing tax burden(Buettner, Holzmann, Overesch, and Schreiber, 2013) . This is even more so in a business environment that is highly globalized. The process of globalization has made transfer pricing regulation difficult for tax authorities (Sikka and Willmott, 2010). Having the liberty to move capital from one country to another and enabling business entities to create subsidiaries, joint ventures, special purpose entities, affiliates and trusts abroad gives them the possibility to profit from favorable tax conditions. This now can be best described with an example9 that Van der Hurk (2013) uses in his research for tax avoidance, namely the way Google use the Double Irish with a Dutch sandwich:

First of all, Google a subsidiary in Ireland (Subsidiary A) which has the responsibility to develop an Intellectual Property (IP) for the company's business in Europe, The Middle East and Africa. This means that the value of IP is distributed and not created solely in the USA which means that Google's tax bill in the USA would include its profits less the profits and value from the IP created abroad, unless of course those profits are repatriated.

Secondly, Google uses a structure called the Double Irish. They open a second subsidiary in Ireland, responsible for the royalties management earned in the Europe, Middle East, Africa countries which is later incorporated by a Dutch intermediate holding company which has a license and right on the IP given to them by Subsidiary A On their turn the Dutch holding provides a sub-license to this second Irish subsidiary, thus all companies from the region of operations (EMEA) pay the Irish subsidiary royalties to use the IP rights and these profits are taxed at the Irish rate instead of deducted locally in the respective countries.

(15)

- 14 -

transferred to Bermuda tax free. The final step is to register the Bermuda firm as an unlimited liability company, thus making it exempt of posting its financial statements.9(for an illustration of the example see Figure 1.)

Figure 1.

Now that I have established how companies may engage in profit shifting it is important to explain what makes intellectual property such an attractive means to an end in terms of minimizing tax burden. Intra-firm trade when exchanging intangible products often lack an observable market price(Durst et. al 2010). This creates considerable difficulty in establishing arm’s length prices, which are deemed appropriate regarding taxation, for the MNE internal

(16)

- 15 -

transactions (Buettner et al. 2013). This increases opportunities to shift profit towards favorable tax climates. Consequently, business entities will more frequently make subsidiary location decision based upon tax parameters and the likelihood that firms have set up subsidiaries specifically for profit shifting is larger.(Desai et al. 2004, Gumpert et al. 2011). This is even more true for knowledge based intellectual property like copyrights, patents etc. As shown in the example above, Google's subsidiaries transfer knowledge between each other and pay each other royalties for that knowledge. Setting an appropriate price on those royalties for knowledge and services that have no competitive market is problematic and leaves space for the company to speculate and arbitrarily set prices in order to reduce its tax burden. In the next section I will go even deeper into defining Royalties and license fees and I will try to review their importance in international trade as the export and import of which will be one of the determinants in my research.

2.5 Royalties and license fees

The term "royalties" as defined by the OECD is "payments of any kind received as a consideration for the use of, or the right to use, any copyright of literary, artistic or scientific work including cinematograph films, any patent, trade mark, design or model, plan, secret formula or process, or for information concerning industrial, commercial or scientific experience." Whereas the royalty considers the quantity or the amount that the object of trade is used the license fee is a lump-sum granting the right to use the object of intellectual property(Wang, 2002). Both license fees and royalties give firms the possibility to transfer their knowledge abroad in order to broaden their possibilities to profit from their competitive advantages. Ferrantino (1993) views intellectual property not only as a competitive advantage of a firm in an international setting over the domestic entities but also as a way of protecting the value of key firm-specific knowledge.

(17)

Trade-- 16 Trade--

related aspect of intellectual property rights (TRIPS) is an example of supranational intellectual property protection, with 188 members out of 206 sovereign states it oversees protection of intellectual property globally. Delgado, Kyle, and McGahan(2013) in their research find that TRIPS has a positive effect on the trade of intellectual property, which means that the enforcement of Intellectual property rights has a positive effect on royalties and license fees as well. Looking at reports from the last two decades we can see that the trade in royalties and license fees has increased tenfold (UNCTAD,2013). While this increase is very substantial it can definitely be assumed that economic growth was not the only cause behind this rapid surge, as the economy itself did not have such an enormous growth rate (World Development Report, 2014). Consequently we can deduce that there are different reasons behind the rapid advancement of royalty trade. Sherry and Treece(2003) claim that the primary use royalties and license fees is within intra-firm trade, which resulted in the exponential growth. Weyzig (2013) on the other hand concludes that a growth rate so vast can be explained by the fact that royalties and license fees are used by companies to reduce their global tax burden. It is absolutely crucial to note that the tax avoidance techniques including royalties and license fees used by companies are absolutely legal. From the research reviewed before we can distinguish two main reasons and origins of international royalty and license trade:

(18)

- 17 -

ii. Tax avoidance driven trade.

On the other hand the fact that global companies like Google can "engineer" their global tax rate to as little as 6% (Ruf and Weichenrieder, 2009) through transfer pricing is something that is due noting, taking in mind the purpose of my research. Google is among many companies that take advantage of inconsistent tax treaties between countries, and plan carefully the location of their subsidiaries. Exploiting such weaknesses is often called “treaty shoping” (Zucman, 2014).

It makes one wonder whether the Netherlands is a facilitator in tax avoidance techniques with the use of Special Purpose Entities, and whether the label "tax haven" is appropriate for the Netherlands (Van Dijk and Weyzig. 2006). In the next section I will elaborate on Special Purpose entities and their role in tax avoidance, as they are widely used in the Netherlands because of withholding tax exemptions.

2.6 Special purpose entities definition and types

Recent academic literature has given great focus on special purpose entities. Studies concentrate not only the dynamics behind them, but analyze the reason behind their existence, and meaning as well as looking into the different types of SPEs and their role in tax avoidance(Weyzig, 2013). It is fascinating that even though the global financial crisis, and the financial and political aftermath following it left the global economy crippled the SPE demand has not experienced a slump.(Ahlawat, Bellom and Kopp,2014). In this section I will construct a deeper understanding of what SPEs are and the different types by exploring sources from several authors, before defining my hypothesis as an end to my literature review.

(19)

- 18 -

example (Weyzig, 2013). Henderson(2002) on the other hand defines SPE’s as “entities created for a limited purpose, with a limited life and limited activities, and designed to benefit a single company”. This gives more clarity on the limited role of SPEs and rather disconnects it legally from its "creator". Furthermore a key aspect of SPEs are bankruptcy remote (IMF,2004) which would mean that “SPE’s assets are isolated from any creditors of its sponsoring firm should the latter go into bankruptcy.” This makes SPEs very attractive firstly because of favorable tax rates and strongly enforced intellectual property rights(Weyzig, 2013) and secondly "it allows sponsoring firms to finance themselves by separating control rights over assets from financing. The operating entity, that is, the sponsoring firm, maintains control rights over the assets that generate cash flows. The assets (projects) can be financed by selling the cash flows to an SPE that has no need for control rights, because the cash flows have already been contracted for."(Gorton and Souleles 2007)

2.6.1 Types of Special Purpose entities10

i. Intermediate holding-This SPE type includes joint ventures and is associated with avoidance of host country dividend withholding or capital gains tax and home country tax on foreign profits. This entails the avoidance of withholding tax on intra-group dividends. An intermediate holding can also provide investment protection. (Weyzig, 2013).

ii. Intra-group loan conduit -This type of SPE allows avoidance of withholding tax on intra-group interest. If a loan conduit channels interest payments onwards to a tax haven affiliate, the overall strategy may also involve avoidance of host country corporate income tax through income shifting. (Weyzig 2013).

iii. Intra-group financing company-This type SPE is mainly used for FDI diversion other than simple intermediate holdings or loan conduits. It is a combination of the two types above.(Weyzig,2013)

iv. Fund raising vehicle- This is a SPE, which issues debt securities via the Netherlands to avoid withholding tax on the interest payments. (Weyzig 2013). This encompasses the

(20)

- 19 -

avoidance of withholding tax on interest to external creditors.

v. Mixed financing company- Covers all structures that combine FDI diversion with external funding and includes firms with a Dutch ultimate parent and SPEs with substantial unknown liabilities.

vi. Securitization Vehicle.-which also issues debt securities but is off-balance and holds portfolio investments only. This is the special purpose vehicle that is used by royalties and licenses for intra and inter-firm tax planning.

2.7 My Hypothesis:

From what has been discussed in my literature review I expect that:

1) An increase in transfer pricing regulation abroad will have a negative effect on the import and the export of royalties and licensing fees in the Netherlands, as regulation is becoming more and more strict globally and tax authorities are pushing for transparency measures.

2) By analyzing the differences in Corporate taxation between the Netherlands and the other countries I believe I will find a positive relationship between the difference in taxation and the import of royalties and license fees and a negative for the export, because of the extensive number of tax treaties that the Netherlands has.

3) By analyzing the differences in Royalty taxation(withholding tax) I believe I will find a positive relationship between the difference in taxation and the import of royalties and license fees and a negative for the export, because of the extensive number of tax treaties that the Netherlands has for foreign investments.

(21)

- 20 -

(22)

- 21 -

CHAPTER 3 Data collection and methodology

In this section I will discuss the various databases I used to collect my data set, and what my dataset consists of. Afterwards I will describe my methodology and define my independent and dependent variables and present my regression analysis.

3.1 Data collection and data set

For the purposes of my research I used the CBS (Central Bureau of Statistics) Netherlands database, as well as the database of the Dutch National Bank(De Nederlandsche Bank), from those I was able to collect information on the export and import of royalties in/from the Netherlands as well as other data for my control variables which I will describe below. For tax regulation information I extensively used KPMG Worldwide Tax guides as well as Tax guides offered from PricewaterhouseCoopers, Ernst and Young and Deloitte. Furthermore for deeper analysis of transfer pricing regulation I used PKF International reviews as well as national reports for all the countries that are included in my sample.

My sample consists of 30 countries11 either part of the European Union or the OECD(or both), along with Russia, China, Brazil and India which were included because of their role as newly advancing economies(O’Neil, 2001). Through thorough data collection I managed to collect an unbalanced panel data set for those countries ranging in the years 2003-2009. The Dutch statistical institute is lacking data before 2003 and the absolute lack of information for some countries excluded them from the sample. In the next section I will describe my model and give a detailed description of my variables.

3.2 Methodology- the Transfer-pricing model

For my model I will partially use the model offered by Buettner, Overesch and Wamser , but while they analyze FDI on a firm level, basing it on assets I will analyze it on a country level specifically through analyzing the exports and imports of royalties and licensing fees.

The transfer pricing regression models are: (3) 𝑅𝑜𝑦𝑎𝑙𝑡𝑦 𝐼𝑚𝑝𝑜𝑟𝑡𝑖𝑡 = 𝛼1.. 𝐶𝑇𝐷𝑖𝑓𝑖𝑡+ 𝛼2. 𝑇𝑃𝑅𝑖𝑡+ 𝛼3. (𝑆𝑇𝑅𝑖𝑡. 𝑇𝑃𝑅𝑖𝑡) + 𝛼4. 𝐺𝐼𝐼𝑟𝑎𝑛𝑘𝑖+ +𝛼₅. 𝐹𝐷𝐼𝑖𝑛𝑓𝑙𝑜𝑤_𝑖𝑡+ 𝛼₆. 𝑅𝑦𝑇𝐷 + 𝛼₇. 𝐼𝑃𝑏𝑜𝑥 + 𝛼₈. 𝑅𝑜𝑦𝑎𝑙𝑡𝑦𝐸𝑥𝑝𝑜𝑟𝑡_𝑖𝑡+ 𝛼₉. 𝐺𝐷𝑃𝑔% + 𝛼10. 𝑙𝑛𝐺𝐷𝑃 + 𝛼11𝐸𝑈𝑑𝑢𝑚𝑚𝑦𝑖𝑡+ 𝛼12(𝑆𝑃𝐸𝐼𝑛𝑤𝑎𝑟𝑑𝑖𝑡) + 𝐸𝑖𝑡 11

(23)

- 22 -

(2) 𝑅𝑜𝑦𝑎𝑙𝑡𝑦 𝐸𝑥𝑝𝑜𝑟𝑡𝑖𝑡 = 𝛼1.. 𝐶𝑇𝐷𝑖𝑓𝑖𝑡+ 𝛼2𝑇𝑃𝑅𝑖𝑡+ 𝛼3. (𝑆𝑇𝑅𝑖𝑡. 𝑇𝑃𝑅𝑖𝑡) + 𝛼4. 𝑊𝐼𝑃𝑂𝑟𝑎𝑛𝑘𝑖 +

+𝛼5. 𝐹𝐷𝐼𝑜𝑢𝑡𝑓𝑙𝑜𝑤𝑖𝑡 + 𝛼6. 𝑅𝑦𝑇𝐷𝑖𝑡 + 𝛼7. 𝐼𝑃𝑏𝑜𝑥 + 𝛼8. 𝑅𝑜𝑦𝑎𝑙𝑡𝑦𝐼𝑚𝑝𝑜𝑟𝑡𝑖𝑡+ 𝛼9. 𝐺𝐷𝑃𝑔% +

𝛼10𝑙𝑛𝐺𝐷𝑃𝐸𝑖𝑡+ 𝛼11𝐸𝑈𝑑𝑢𝑚𝑚𝑦𝑖𝑡 + 𝛼12(𝑆𝑃𝐸𝑂𝑢𝑡𝑤𝑎𝑟𝑑𝑖𝑡)12+ 𝐸𝑖𝑡

I plan to use these models first for the whole sample period 2003-2009, then I am going to use all the data available for the countries in my sample but for the period 2003-2007 to see the effect that the financial crisis has on the imports and exports of royalties. Last but not least due to data limitations I am going to test the effects of SPE’s with a cross correlation regression on data available for the year 2013.

3.2.1Dependent variables:

RoyaltyImportit

Measurement: This variable represents the import of royalties and license fees in the Netherlands from the sample countries in the period 2003-2009.

Source: Central Bureau of Statistics the Netherlands Dynamics:

Figure 2.

Solid line : Sum of imports of royalties and license fees from the Netherlands to the countries included in my sample. Source: own computations and Central Bureau of Statistics the Netherlands (CBS).

12_{Variables in brackets will be included only in the cross-correlation regression for the Import and Export of} Royalties in the year 2013

0 1000 2000 3000 4000 5000 6000 2003 2004 2005 2006 2007 2008 2009 A M O UN T( M LN . E UR) YEAR

(24)

- 23 -

In order to check for the effects of the financial crisis in my sample I plotted the Import of Royalties from the countries included in my sample. During the 2006-2008 period as it can be seen in Figure 1 suffered a major downward trend, only to suffer a powerful surge through 2008-2009.

RoyaltyExportit

Measurement: This variable measures the export of royalties from the Netherlands to the countries included in my sample in the period 2003-2009

Source: Central Bureau of Statistics the Netherlands Source: Central Bureau of Statistics the Netherlands Dynamics:

Unlike the imports of royalties and license fees, the export (presented in Figure 2), it is noticeable that while the exports were lower in the period 2006-2009 compared to 2003-2005, the exports followed an upward trend.

Figure 3.

Solid line : Sum of exports of royalties and license fees from the Netherlands to the countries included in my sample. Source: own computations and Central Bureau of Statistics the Netherlands (CBS)

0 500 1000 1500 2000 2500 3000 3500 4000 4500 5000 2003 2004 2005 2006 2007 2008 2009 A M O UN T( M LN .E UR) YEAR

(25)

- 24 -

3.2.2 Independent variables13 *CTDit

Measurement: Stands for Corporate Tax Difference between the sample country and the Netherlands.

Included because: Statutory tax rates are used here(without reductions from treaties), because effective tax rates resemble endogenous decisions made by the organization, such as its accumulation of debt, while statutory tax rates are decided upon by the government and thus provide a more believable identification source because it captures exogenous to the firm factors(Dharmapala,2014).

Expected sign: I expect a positive sign for imports because of the numerous tax treaties that the Netherlands has will mean that the difference in taxation will stimulate MNCs to shift profits to the Netherlands. Hence for exports I expect a negative sign as the relationship is contrary.

*RyTDit

Measurement: Represents the difference in withholding taxes between the countries in my sample and the Netherlands

Included because: With an extensive network of tax treaties the Netherlands often has close to 0% withholding taxes, making it attractive to other countries for tax planning activities. Source: International Bureau of Fiscal Documentation (http://www.ibfd.org/)

Expected Sign: I expect a positive sign for imports because of the numerous tax treaties that the Netherlands has will mean that the difference in taxation will stimulate MNCs to shift profits to the Netherlands. Hence for exports I expect a negative sign as the relationship is contrary.

*TPRit

Measurement: This variable measures the extent of transfer pricing regulation based on Lohse, Riedel and Spengel’s(2012) classification. The variable itself varies from 1 to 5, corresponding to a certain category of transfer pricing regulation (see Table 2).

Included because: It is one of the main independent variables as it follows the strictness of transfer pricing regulation and its change through the period(if present).

Source: Lohse, T., Riedel, N. and Spengel, C. (2012)

(26)

- 25 -

Expected sign: I expect a negative sign both for imports and exports, which will mean that a substantial part of transfers of royalties and licensing fees have been driven by tax avoidance motives

Dynamics:

Figure 4.

Solid line: Mean level of strictness of transfer pricing regulation imposed by countries in the sample. Source: own computation based on Buettner et. al. 2014 and Lohse et. al. 2012

As it can be seen in the figure above strictness of transfer pricing regulation as defined by Lohse et. al 2012 has gradually increased during the period of my research.

*TPRit. STRit

Measurement: This is the product between the Statutory Tax rate in country I and the transfer pricing regulation in that country.

Included because: This way I can test the marginal effects of corporate tax rates and transfer pricing regulation(Buettner, Overesch, Wamser, 2014)

Source: Buettner, T. Overesch, M. and Wamser, G. 2014

Expected Sign: For imports I expect a positive sign as a marginal change in tax and/or transfer pricing regulation will incentivize companies even more to shift profits to the Netherlands. For exports I expect the opposite, hence a negative sign)

(27)

- 26 -

**GIIrankit

Measurement: This variable represents the Global innovation ranking of countries during the sample period. The ranking is on a global scale and represents the actual ranking of the countries included in my sample in order to have a visible differentiation between countries. Included because: Innovation leads to growth and social returns and resembles the overall economic environment in the country(Bresnahan, Trajtenberg, 1995).

Source: https://www.globalinnovationindex.org/content.aspx?page=GII-Home

Expected sign: I expect a positive sign for imports as the Netherlands will import knowledge from countries with a higher rank. On the other hand I expect a negative sign for exports as the Netherlands will export less to countries with a higher rank.

**FDIinflowit/FDIoutflowit

Measurement: Consists of the ingoing and outgoing direct investments in and from the Netherlands from and to the countries in my data set.

Included because: FDI represents the readiness of one country to invest in another. According to Weidner(1999) a connection is present between royalties and FDI.

Source: Dutch National Bank (DNB) ( http://www.statistics.dnb.nl/betalingsbalans-en-extern-vermogen/index.jsp)

Expected sign: I believe that FDI inflow/outflow will be positively related to the imports/exports of royalties, as a country that has a generally strong intellectual property enforcement attract FDI, and at the same time stimulate exports because of the secure innovation environment.

*IP-box

Measurement: This is a dummy variable, which equals 1 if Intellectual property box legislation is present in the country. This type of legislation encompasses a different tax regime for intellectual property revenues. The variable equals 0 otherwise.

Included because: This type of regulation often exempts business from withholding tax, which can make the country attractive for (Buettner, Overesch, Wamser, 2014, Weyzig, 2013).

(28)

- 27 -

**GDPg%/lnGDP

Measurement: This variables represent the Gross domestic product annual growth rate through the researched period and the natural logarithm of that countries’ GDP.

Included because: These variables will control for market size and market dynamics, which are detrimental to investment decisions(Buettner, Overesch, Wamser, 2014).

Source: International Monetary Fund(IMF) World Economic Outlook(WEO) database, (April 2015).

Expected sign: I expect both positive signs for imports and exports as investment decisions are influenced by how dynamic the host market is and whether it facilitates growth

EU-Dummy

Measurement: This dummy variable represents whether the country A was a member of the European Union in the period between 2003-2009. If the country was a member state the dummy variable equals 1 and if it is no it equals 0.

Included because: This variable will control for the dynamics within the European Union Source: http://europa.eu/about-eu/countries/index_en.htm

Expected sign: I expect a positive sign for both Imports and Exports of Royalties as the EU policies for convergence of the market, gives member states incentives to trade more with each other(Panagariya, 2002)

SPE Inward/SPE Outward

Measurement: Represents the FDI inflows and outflows through SPE’s in and from the Netherlands

Included because: In order to see if SPE’s have a significant effect on the export and import of royalties

Source: Dutch National Bank (DNB) (http://www.statistics.dnb.nl/betalingsbalans-en-extern- vermogen/index.jsp)

(29)

- 28 -

Chapter 4 Test results and discussion

4.1 2003-2009 transfer pricing model

Using this model I will analyze data collected for all 30 countries for the period 2003-2009. Due to data unavailability the I will be running Pooled Ordinary Least Squares regression for my unbalanced panel data set.

4.1.1 Descriptive statistics

GII_Rank

TP

Regulation TPREGXSTR RoyaltyTaxDif RoyaltyImport RoyaltyExport

Mean 24.3268 2.908497 0.750663 0.141609 144.3268 245.2941 Median 22 3 0.8 0.15 10 87 Maximum 76 5 1.6 0.3333 3196 5960 Minimum 1 0 0 0 0 0 Std. Dev. 18.17887 1.505951 0.443894 0.097024 461.5276 550.2274 Observations 153 153 153 153 153 153 LnGDP IP-Box GDP

Growth% FDI_Outward FDI_Inward

Corporatete Tax Dif Mean 12.9844 0.287582 2.890196 4486.196 4749.601 0.044504 Median 12.89572 0 2.8 1152 627 0.05 Maximum 16.48404 1 14.2 106556 87049 0.2 Minimum 9.371694 0 -18 -44237 -53253 -0.075 Std. Dev. 1.64919 0.454122 4.438554 14015.69 11454.6 0.073343 Observations 153 153 153 153 153 153

(30)

- 29 -

4.1.2 Regression Results

Pooled Ordinary Least Squares Regression Analysis

Trasnsfer pricing models

(1) (2) Dependent Variable RoyaltyImportit RoyaltyExportit

Constant -777.9727** -1391.923* RoyaltyExport 0.105989*** - RoyaltyImport - 0.15303*** GII_Rank(+/+) -6.550262* -3.6615 TP Regulation(-/-) -208.5569* -192.1514** TPRxSTR(+/-) 943.1119* 670.3013* RoyaltyTaxDif(+/-) 747.2265** 633.8137 lnGDP(+/+) 71.08104** 141.5345* IP-Box(+/+) -22.10544 209.9338 GDP Growth%(+/+) 10.63505 -10.06617 FDI_Outward(+) - 0.002687 EU Dummy(+/+) -98.20409 159.158 FDI_Inward(+) -0.00707** - Corporate tax dif(+/-) -1466.951 -4739.425* Adjusted R² 0.332098 0.239764

Observations 153 153 correlations is significant at , *<0,01, **<0,05, ***<0,1

In this linear regression analysis for the import and export of royalties and license fees from/to the Netherlands GII_Rank represents the Global innovation ranking of country I, TP regulation represents the degree of regulation in country i in period T, ranging from 1 to 5, TPRxSTR represents the product between the degree of transfer pricing regulation and the statutory tax rate in country i, RoyatyTaxDif consists of the difference between withholding taxes in the Netherlands and country i, lnGDP is the natural logarithm of country i, IP box is a dummy variable which is 1 if an Intellectual property box is present in country i and 0 otherwise, GDPGrowth% is the Gross Domestic Product annual growth rate of country i, FDI_Outward is all Foreign Direct Investments from the Netherlands to the sample countries, FDI_Inward is all incoming Foreign Direct Investments from the sample countries towards the Netherlands, Corporate tax dif is the difference in statutory tax rates between country i and the Netherlands. EU-dummy is a dummy variable which is 1 if the country is member of the EU and 0 otherwise. All relevant data is within the 2003-2009 period. The signs within the brackets are the expected outcomes e.g. (+/+) means that a positive sign is expected for both regressions

Above are the results from my two linear regressions. The adjusted R-squared are relatively low. The reason for this may be due to the fact that my panel data is unbalanced which means that my observations are not plotted evenly through the fitted regression line. Next I will discuss my hypothesis and their outcome.

(31)

- 30 -

The transfer pricing regulation variable is negative and is highly statistically significant for both the models on import and export of royalties from the Netherlands. From this we can presume that indeed the strictness of transfer pricing regulation is a valid environmental factor and is affecting countries’ tax planning practices.

My second hypothesis took into consideration corporate taxation and withholding taxes on royalty payments:

(2) By analyzing the differences in Corporate taxation between the Netherlands and the other countries I believe I will find a positive relationship between the difference in taxation and the import of royalties and license fees and a negative for the export, because of the extensive number of tax treaties that the Netherlands has.

In the first regression model where the dependent variable is the import of royalties, the coefficient of the Corporate Tax rate is negative but not statistically significant(p=0.1442), while in the second model regarding the relationship between the export of royalties and taxation the corporate tax variable is highly significant(p<0,01) and negative, which corresponds partly with my expectations in my second hypothesis.

(3) By analyzing the differences in Royalty taxation(withholding tax) I believe I will find a positive relationship between the difference in taxation and the import of royalties and license fees and a negative for the export, because of the extensive number of tax treaties that the Netherlands has for foreign investments.

In the first regression model the withholding tax variable is positive and significant which partly confirms the first part of my expectations. Furthermore in the second model regarding the relationship between the export of royalties and taxation, the withholding tax variable is not statistically significant

(4) There will be a positive relationship between the import and the export of royalties and licensing fees, confirming that the Netherlands is used to convey royalties(Weyzig, 2013)

(32)

- 31 -

4.1.3 Correlation matrix(see Appendix B Table 1B)

Multicollinearity does not seem to be an issue in my analysis as there are no correlation coefficients that are over 0.9(Farrar and Glauber, 1967).

4.2 Pre-financial crisis transfer pricing model

Using the data only between 2003-2007 I will check if the financial crisis creates a bias in my

research. The financial crisis led to an overall economic decline and affected foreign trade, this is why I consider this test detrimental for my research.

4.2.1 Descriptive statistics

(33)

- 32 -

Constant 136.5235* -1087.138* RoyaltyExport 0.498498* - RoyaltyImport - 0.216929* GII_Rank(+/+) -5.669853** -2.654962 TP Regulation(-/-) -186.9364** -115.9388** TPRxSTR(+/-) 873.0847* 375.2269*** RoyaltyTaxDif(+/-) 294.2166 333.7646 lnGDP(+/+) -21.05414 113.0514* IP-Box(+/+) -81.30856 159.4195* GDP Growth%(+/+) 29.80226** -7.659184 EU Dummy(+/+) -116.0931 -9.732124 FDI_Outward(+) - 0.001273 FDI_Inward(+) 0.004417*** - Corporate tax dif(+/-) 480.605 -2938.369* Adjusted R² 0.470187 0.558227

Observations 107 107

correlations is significant at , *<0,01, **<0,05, ***<0,1

In this linear regression analysis for the import and export of royalties and license fees from/to the Netherlands GII_Rank represents the Global innovation ranking of country I, TP regulation represents the degree of regulation in country i in period T, ranging from 1 to 5, TPRxSTR represents the product between the degree of transfer pricing regulation and the statutory tax rate in country i, RoyatyTaxDif consists of the difference between withholding taxes in the Netherlands and country i, lnGDP is the natural logarithm of country i, IP box is a dummy variable which is 1 if an Intellectual property box is present in country i and 0 otherwise, GDPGrowth% is the Gross Domestic Product annual growth rate of country i, FDI_Outward is all Foreign Direct Investments from the Netherlands to the sample countries, FDI_Inward is all incoming Foreign Direct Investments from the sample countries towards the Netherlands, Corporate tax dif is the difference in statutory tax rates between country i and the Netherlands.EU-dummy is a dummy variable which is 1 if the country is member of the EU and 0 otherwise All relevant data is within the 2003-2007 period.

Above are the results from my two linear regressions for the period 2003-2007. The adjusted R-squared are higher than the regression model including the 2003-2009 period.. Next I will discuss my hypothesis and their outcome.

(1)An increase in transfer pricing regulation abroad will have a negative effect on the import and the export of royalties and licensing fees in the Netherlands, as regulation is becoming more and stricter globally and tax authorities are pushing for transparency measures.

(34)

- 33 -

My second and third hypothesis took into consideration corporate taxation and withholding taxes on royalty payments:

(2) By analyzing the differences in Corporate taxation between the Netherlands and the other

countries I believe I will find a positive relationship between the difference in taxation and the import of royalties and license fees and a negative for the export, because of the extensive number of tax treaties that the Netherlands has.

In this hypothesis only the coefficient for the export of Royalties is with the expected sign and statistically significant. Thus the higher the difference in corporate taxation the lower the export of royalties to the respective countries. It is important to note however that there is evidence that tax driven trade of royalties and license fees is present.

(3)By analyzing the differences in Royalty taxation(withholding tax) I believe I will find a positive relationship between the difference in taxation and the import of royalties and license fees and a negative for the export, because of the extensive number of tax treaties that the Netherlands has for foreign investments.

Unfortunately both coefficients are negative and statistically insignificant (p>0.1). Therefore in this model I reject my second hypothesis.

(4)There will be a positive relationship between the import and the export of royalties and licensing fees, confirming that the Netherlands is used to convey royalties(Weyzig, 2013)

The pre-financial crisis model shows that indeed export and import of royalties are positively correlated the relationship is highly statistically significant(p<0.01)

4.2.3 Correlation Matrix(see Appendix B Table 2B)

(35)

- 34 - 4.3 SPE Model

The third model unlike the previous two is a cross-correlation regression using Ordinary Least Squares. In this model I observe the 30 countries in my sample in a single year (2013), in order to test the relationship between the Export/Import of Royalties. Hence in this section I will discuss the results only for my 5th and final hypothesis.

4.3.1 Descriptive statistics GII_Ra nk TP Regulation TPREGXST R RoyaltyTa xDif RoyaltyIm port RoyaltyExport SPE Inward Mean 25.166 3.4 0.775149 0.138387 229.9 638.2667 6005.767 Median 20.5 4 0.7399 0.15 13 86.5 1367.5 Maximum 79 5 1.6 0.33 4570 12678 57170 Minimum 1 0 0 0 0 0 -9619 Std. Dev. 19.802 1.354431 0.448463 0.97 834.8993 2300.379 13543.86 Observati ons 30 30 30 30 30 30 30 LnGDP IP-Box GDP Growth% FDI_Outw ard FDI_Inwar d Corporatete Tax Dif SPE Outward Mean 13.204 17 0.233 0.009233 41064.37 37677.13 0.043803 5748 Median 13.168 0 0.005 3939.5 3165.5 0.05 2767 Maximum 16.673 1 0.078 497022 507684 0.2 43817 Minimum 10.088 0 -0.054 -10593 -9101 -0.075 -629 Std. Dev. 1.7133 0.430183 0.027 115887.9 110320.8 0.07253 9245.268 Observati ons 30 30 30 30 30 30 30

(36)

- 35 -

Constant -2418.744 -791.3489 RoyaltyExport 0.098621 - RoyaltyImport - 0.505991 GII_Rank(+/+) 5.214053 -66.23099 TP Regulation(-/-) -4.704377 -471.0399 TPRxSTR(+/-) 18.42204 3798.977 RoyaltyTaxDif(+/-) 355.9969 5195.678 lnGDP(+/+) 47.38591 266.4003 IP-Box(+/+) -207.146 1665.896 SPE Inward 0.074492* Spe Outward -0.029688 GDP Growth%(+/+) 897.4073 -26656.18*** EU Dummy(+/+) -90.06479 938.5217 FDI_Outward(+) - -0.006436 FDI_Inward(+) 0.004202* - Corporate tax dif(+/-) -801.8574 -29805.85* Adjusted R² 0.864607 0.580265

Observations 30 30

correlations is significant at , *<0,01, **<0,05, ***<0,1

In this linear regression analysis for the import and export of royalties and license fees from/to the Netherlands GII_Rank represents the Global innovation ranking of country I, TP regulation represents the degree of regulation in country i in period T, ranging from 1 to 5, TPRxSTR represents the product between the degree of transfer pricing regulation and the statutory tax rate in country i, RoyatyTaxDif consists of the difference between withholding taxes in the Netherlands and country i, lnGDP is the natural logarithm of country i, IP box is a dummy variable which is 1 if an Intellectual property box is present in country i and 0 otherwise, GDPGrowth% is the Gross Domestic Product annual growth rate of country i, FDI_Outward is all Foreign Direct Investments from the Netherlands to the sample countries, FDI_Inward is all incoming Foreign Direct Investments from the sample countries towards the Netherlands, Corporate tax dif is the difference in statutory tax rates between country i and the Netherlands.SPE Inward and SPEOutward are respectively the Foreign Direct Investments from and to the Netherlands through Special Purpose Entities. All relevant data is within the 2003-2007 period.

Above the results from the regression are presented. As it can be seen Hypothesis 1 through 4 are rejected and my 5th hypothesis:

1) I believe that the relationship between the imports and exports of royalties and the inward and outward flows of direct investments through Special Purpose entities in the Netherlands, is positive, proving that SPE’s are used to convey cash flows through the Netherlands(Weyzig, 2013)

Is only partly supported. While there is a slight positive relationship between the Import of Royalties and the inflow of investments through SPE’s on the Export side of the model, the variable is

(37)

- 36 -

4.3.3 Correlation Matrix (See Appendix B, Table 3B)

(38)

- 37 -

Chapter 5 Conclusion

In this paper I define and explore the different types of intellectual property and the ways in which it can be used by companies to reduce their tax burden. Furthermore I continue with explaining the types of transfer pricing and the dynamics behind transfer price forming and regulation that has been widely discussed both on national and international levels. I use a unbalanced data set of 30 countries over the period 2003-2009 and extend on the model offered by Buettner, Holzmann, and Overesch (2014), by offering a view on the dynamics behind FDI represented by royalties and license fees on a national level. By following two regression analyses I investigate my research question:

Does an increase in transfer pricing regulation have an effect on the import and export of royalties and license fees into/from the Netherlands?

With statistical confirmation I find a negative relationship between the increase of transfer pricing regulation and the import and export of royalties in the Netherlands. Furthermore with nearly statistical certainty my results show that Exports and Imports of Royalties are positively connected which may confirm that the Dutch state is used for royalty conveying purposes. Furthermore my results confirm empirically that corporate tax rate is associated negatively with the exports of Royalties from the Netherlands, and that the tax exemptions that the Netherlands offers on withholding taxes indeed have a positive effect on the import of Royalties in the Netherlands. What’s more I run two additional regressions. On the one hand I run a regression for the period 2003-2007, thus excluding the financial crisis from my sample and I get the same statistical results. On the other hand I analyze whether SPE’s in the Netherlands are used for tax-avoidance purposes by analyzing again country specific data, but for the year 2013. Unfortunately the third model did not lead to a confirmation of my hypothesis.

With my research I believe I contribute to the current academic literature and give leeway for future research(discussed in the next section). I base my paper mainly on the paper of

Buettner, Overesch and Wamser (2014), but whereas they look at firm specific data and

Transfer pricing and tax avoidance