Sovereign Wealth Funds (SWFs) have reached an enormous financial power and have tripled their assets under management to 7.4 trillion American dollars during the past ten years. So far, academia has focused on the most obvious characteristic of SWFs, the state ownership, and relating governance issues. This thesis answers the question of whether there are certain company characteristics, which are preferred by SWFs when investing, and which characteristics influence the ex-post probability of becoming a SWF target. For this purpose, the selected sample is compared to a set of Year, Industry, Size, and MTBV matched peers. First, systematic differences between targets and peers are identified using t-test and Wilcoxon rank sum test statistics. Second, the influence of variables on the ex-post probability of becoming a SWF target is analyzed using logistic panel regression models. The regression results are further discussed using odds ratios and marginal effects analysis. The main finding is that target companies are typically significantly larger than their matched peers, and the size of a company is identified to have the highest influence on the likelihood of becoming a target. This is accompanied by the finding that a higher proportion of held cash has a positive influence, and a higher book leverage ratio has a negative influence. Additionally, it is shown that there are no target characteristics, which would promote large or small-scale investments. As introduction into further research, it is analyzed how stock markets react on announced SWF investments. It is shown that excess returns of approximately one percent can be observed within a three-day event window.
Table of Contents
List of Figures
List of Tables
Table of Attachments
Table of Digital Attachments
List of Abbreviations
1 Introduction
2 Introducing Sovereign Wealth Funds
2.1 The Rise of Sovereign Wealth Funds
2.2 Definition of “Sovereign Wealth Fund”
2.3 Funding Sources of SWFs
2.4 Scientific Discussion about SWFs
3 Data and Sample Overview
3.1 Data Sources and Sample Selection
3.1.1 Sovereign wealth fund transactions
3.1.2 Unique company identifiers
3.1.3 Financial information
3.2 Sample Description
3.2.1 Description of included observations
3.2.2 Description of included SWFs
4 Characteristics of SWF Targets
4.1 Identifying Matched Peers
4.2 Analysis of Target Characteristics
4.2.1 Comparison of targets with matched peers
4.2.2 The ex-post probability of getting targeted by SWFs
4.2.3 Target characteristics of large versus small-scale investments
5 Stock Market Reactions
6 Conclusion
Appendix
References
Abstract
Sovereign Wealth Funds (SWFs) have reached an enormous financial power and have tripled their assets under management to 7.4 trillion American dollars during the past ten years. So far, academia has focused on the most obvious characteristic of SWFs, the state ownership, and relating governance issues. This thesis answers the question of whether there are certain company characteristics, which are preferred by SWFs when investing, and which characteristics influence the ex-post probability of becoming a SWF target. For this purpose, the selected sample is compared to a set of Year, Industry, Size, and MTBV matched peers. First, systematic differences between targets and peers are identified using t-test and Wilcoxon rank sum test statistics. Second, the influence of variables on the ex-post probability of becoming a SWF target is analyzed using logistic panel regression models. The regression results are further discussed using odds ratios and marginal effects analysis. The main finding is that target companies are typically significantly larger than their matched peers, and the size of a company is identified to have the highest influence on the likelihood of becoming a target. This is accompanied by the finding that a higher proportion of held cash has a positive influence, and a higher book leverage ratio has a negative influence. Additionally, it is shown that there are no target characteristics, which would promote large or small-scale investments. As introduction into further research, it is analyzed how stock markets react on announced SWF investments. It is shown that excess returns of approximately one percent can be observed within a three-day event window.
Zusammenfassung
Sovereign Wealth Funds (SWFs) haben in den letzten zehn Jahren eine enorme Finanzkraft entwickelt und ihr verwaltetes Vermögen auf 7.4 Billionen US-Dollar verdreifacht. Diese Thesis beantwortet die Frage, ob es bestimmte Unternehmens-eigenschaften gibt, welche von SWFs für ihre Investitionen bevorzugt werden und folglich die Ex-post-Wahrscheinlichkeit beeinflusst, ein Ziel von SWF Investitionen zu werden. Hierfür wird die ausgewählte Stichprobe von Transaktionen mit einer Peergroup verglichen, welche anhand der Variablen Year, Industry, Size, und MTBV erstellt wurde. Zuerst werden unterscheide zwischen Ziel- und Vergleichsunternehmen mithilfe von t-tests und dem Wilcoxon-Rangsummentest identifiziert. Zweitens wird mittels logistischer Panelregressionen analysiert, welche Variablen einen Einfluss auf die Ex-post-Wahrscheinlichkeit ein Zielunternehmen zu werden haben. Die Regressionsergebnisse werden zudem mit Hilfe von Odds Ratios und mittels einer Analyse der marginalen Effekte interpretiert. Das Hauptergebnis ist, dass Zielunternehmen üblicherweise signifikant größer sind als ihre Vergleichsgruppe. Des Weiteren hat die Größe eines Unternehmens den stärksten Einfluss auf die Wahrscheinlichkeit ein SWF-Ziel zu werden. Daneben hat ein höheres Barvermögen einen positiven und eine höhere Book-Leverage-Ratio einen negativen Einfluss. Außerdem wird gezeigt, dass es keine beobachtbaren Unterschiede zwischen den Eigenschaften von Zielunternehmen großer und kleiner SWF-Investitionen gibt. Als Einführung in weitere Forschungsfelder wird zudem analysiert, welche Reaktion die Ankündigung einer SWF-Investition auf den Aktienmärkten hervorruft. Bei Betrachtung eines dreitägigen Ereignisfensters, ist eine abnormale Rendite von ungefähr einem Prozent über der erwarteten Rendite zu beobachten.
List of Figures
Figure 1: Number of Documented SWF Transactions per Year
Figure 2: Origin of SWF Funding
Figure 3: Target Countries and Number of Observations
Figure 4: Individual and Repeated Transactions per Target Country
Figure 5: Transactions per Target Sector
Figure 6: Transactions per SWF and SWF-Country
Figure 7: Abnormal Returns (26-day event window)
Figure 8: Abnormal Returns (41-day event window)
Figure 9: AUM Development per Funding Category
Figure 10: AUM Development per Geographic Origin
Figure 11: Complete Set of Target Countries
Figure 12: Complete Set of Transactions per SWF and Country
List of Tables
Table 1: Assets under Management in Billion USD – October 2017
Table 2: Main Acquirer Types
Table 3: Percentage of Shares Owned Post Transaction
Table 4: Sample Selection Criteria
Table 5: Resulting Subsamples
Table 6: Percent Acquired and Ownership Post Transaction
Table 7: Internationality of SWF Investments
Table 8: Target Characteristics countries, excluding repeated transactions)
Table 9: Logistic Panel Regression Excluding Repeated Transactions
Table 10: Logistic Panel Regression Including Repeated Transactions
Table 11: Odds Ratios (16 countries, excluding repeated transactions)
Table 12: Marginal Effects (16 countries, excluding repeated transactions)
Table 13: Abnormal Returns (three-day event window)
Table 14: Transactions per Industry
Table 15: Target Characteristics (16 countries, including repeated transactions)
Table of Attachments
Appendix 1: Key Definitions
Appendix 2: Development of the assets under management (AUM)
Appendix 3: Number of Observations per Target Country
Appendix 4: Transactions per Industry
Appendix 5: Transactions per SWF and Country
Appendix 6: Target Characteristics Including Repeated Transactions
Table of Digital Attachments
The digital appendix is stored on the submitted disk and is numbered as follows. The appendix is grouped in folders, which summarize the files of a specific level of information, e.g. “SWF Information”, for general information on the SWF level. Each folder contains subfolders, or several spreadsheets and Stata files. If a reference to the digital appendix refers to a specific file, the file-name is indicated at the end of the footnote. E.g. “Digital appendix 1 – SWF Ranking”. The digital appendix contains all files, which resulted from the work on this thesis, but is not containing every intermediate step, which has been needed to achieve the resulting files.
Abbildung in dieser Leseprobe nicht enthalten
1 Introduction
This thesis is about the so called sovereign wealth funds (SWFs) and their investment priorities. It will answer the question of whether there are certain company characteristics, which are preferred by SWFs when investing, and which characteristics influence the ex-post probability of becoming a SWF target.
The term “sovereign wealth fund” was introduced in 20051, however, the first state-owned SWF has already been established by Kuwait in 1953. SWF-like investment vehicles on a federal state level have been establish even earlier.2 A SWF is a state-owned investment vehicle, which manages the wealth of a nation beyond the traditional reserve management. Whereas the reserve management is focused on liquidity, SWFs are investing their assets with a return focused long-term perspective. To achieve this, SWFs invest a large part of their assets under management in global equity markets. At first sight, SWFs seem to be close to pension funds, which are managing the wealth of future senior citizens. The main difference between SWFs and state-owned pension funds is that pension funds are financed by people saving money for retirement, whereas most SWFs are funded via revenues from the export of natural resources. Other funding sources include foreign exchange reserves and privatization proceeds.3 The SWFs’ assets under management have grown from three trillion American dollars (USD) in June 2007 to almost 7.5 trillion USD in June 2017.4 The largest SWF alone manages assets of one trillion USD, and thus, owns approximately 1% of all the shares in the world.5 These numbers indicate that SWFs play an important role on global capital markets and serve as an interesting topic, also for academic discussions.
So far, academia was focused on the most obvious characteristic of SWFs, the governmental ownership and relating governance questions. Further, it was discussed, if SWFs are using their financial power to influence domestic industries and policies, as has been feared by politicians of investment receiving countries. This discussion is presented in chapter two in more detail.6 Besides the mentioned academic discussion, chapter two will introduce SWFs in more detail. This includes the development during the last decades, common definitions of the term “sovereign wealth fund”, as well as their financial origins.
Chapter three will present the transaction database, which is offered by the Sovereign Wealth Fund Institute (SWF Institute; SWFI). This database serves as the first of two main sources of information, which is needed to analyze the question of this thesis. The SWF database contains approximately 21.000 SWF transactions, starting in 1974. The chapter will continue with the introduction of selection criteria. These criteria are used to identify an actual and meaningful sample of transactions. The second source of information is the financial database Eikon, which got accessed via Datastream. Eikon is going to be used to identify the companies which are stored in the transaction database, as well as to source the financial information needed to perform the regression analysis. Chapter three ends with the description of the resulting sample of 597 observations in 18 target countries. This description also reveals companies, which were targeted multiple times. Thus, a subsample excluding these repeated transactions is created, which will be used as the basic sample for the analysis.
Till this chapter, the theoretical background, as well as preparatory steps will be presented. In chapter four, the question of whether certain company characteristics affect the probability of being targeted by a SWF is finally going to be answered in several steps. First, it is necessary to identify a set of matched peer companies, in order to provide statements about systematic differences of SWF targets, compared to other companies.
After identifying the peer group, the company characteristics of targets and non-targets are compared. Potential differences are evaluated using test statistics reporting the significance of differences in mean and median values. Moreover, a quintile distribution is going to indicate a first picture of the distribution of SWF target characteristics within the combined set of targets and peers. When looking at the size of SWF targets, this already reveals a strong concentration of targets in the largest quintile. The variable Size will be proven as the main influential characteristic during this work.
Next, the ex-post probability of getting targeted by SWFs is examined using logistic panel regression models. Logistic regressions offer the advantage that the results range between zero and one, what can be interpreted as a probability. However, the coefficients of logistic regressions cannot be interpreted easily. That is why the regression results using the basic sample, which is limited to individual transactions, is discussed additionally using odds ratios and an analysis of marginal effects. As main finding, chapter four identifies the variables Size and Cash as heaving a significant positive influence on the ex-post probability of becoming a SWF target, whereas the variable Leverage influences the probability negatively.
Chapter four will conclude with a comparison of target company characteristics of large and small-scale investments of SWFs, which leads to the finding that there is no observable systematic difference.
Chapter five serves as an entrance to further SWF related research. Besides knowing which company characteristics influence the decisions of SWFs, it is also of high interest to analyze the public reaction on these investment decisions. Chapter five is going to present an event study about the stock market reaction around the announcement day of SWF investments. The event study focuses on the market of the United States and the United Kingdom, and reveals an excess return, when assessing an event window of three days. This last chapter is not going to claim to be a comprehensive market analysis, but should give a first impression about the market situation around SWF investments, and wants to inspire future research.
2 Introducing Sovereign Wealth Funds
This chapter is introducing sovereign wealth funds, starting with the relevance of the topic and the question of what exactly are sovereign wealth funds? This includes common definitions, a brief introduction of the origin and history of SWFs, as well as typical funding sources. Finally, this chapter discusses the current state of research on SWFs.
2.1 The Rise of Sovereign Wealth Funds
This chapter is introducing SWFs, starting with the relevance of the topic. The investment category of SWFs exists since decades. Texas and Utah have already established the first SWF like funds on a federal state level in the nineteenth century.7 The first SWF owned by a country has been the Kuwait Investment Board, nowadays known as Kuwait Investment Authority, which got established in 1953.8 With 27 established SWFs before, and 52 after the year 2000, the number of SWFs has substantially increased since 2000.9
The pure number of 79 established SWFs appears to be relatively small. However, given that SWFs are state owned and there are 193 countries in the world, 79 is a substantial number, even if some countries are owning multiple SWFs. However, the number of funds alone is not an argument for the actuality of the topic. Consequently, two more aspects are enclosed, starting with an illustration of the number of SWF transactions per year.
Abbildung in dieser Leseprobe nicht enthalten
Figure 1: Number of Documented SWF Transactions per Year10
Figure 1 shows that the documented SWF transactions have reached a substantial amount towards the end of the 2000s. This trend found its maximum in 2013. Further, the level of transactions after 2013 stays at a significantly higher level than in the years before 2013. The decades until 2004 show little activity with almost no transactions.
Second, the typical size of funds is analysed and compared to other fund segments. A popular measure for the size of funds are the assets under management (AUM). This measure is used to rank funds, and gives a first impression about the size of SWFs, and their relevance on capital markets. Besides the values for SWFs, Hedge Fund and Mutual Fund statistics are icluded in table 1. With a total of 7.5 trillion USD of assets under management, the SWF branch is more than twice as big as the sum of all hedge funds. At the same time, mutual funds are almost twice the size of SWFs. Further, the table shows that around three quarter of the SWFs’ assets under management are distributed to the top ten largest SWFs. In this comparison, this is the most concentrated branch, which is not surprising given the small number of established SWFs.
When looking at a ranking of the largest public funds, undertaken by the SWF Institute in June 2017, nine out of the top fifteen largest public funds are SWFs.11 Managing investments with a total amount of 7.5 trillion USD and representing some of the world’s largest public funds, SWFs are playing an important role on global capital markets. To further illustrate this, more details about the Norwegian SWF are introduced. The Norway Government Pension Fund Global, formerly known as Petroleum Fund, is managing approximately one trillion USD, which makes it the world’s largest SWF.12 Two thirds of the assets are invested in equities. This corresponds to more than 1% of shares globally.13 These numbers visualize the importance of SWFs for global financial markets.
Abbildung in dieser Leseprobe nicht enthalten
Table 1: Assets under Management in Billion USD – October 201714
2.2 Definition of “Sovereign Wealth Fund”
Now, since the relevance of the topic became clear, the next question is: What exactly is a SWF? The term “sovereign wealth fund” has not been defined universally. There are several definitions describing SWFs in a slightly different way. The person who introduced the term has been Andrew Rozanov in 2005. This coincides in time with an increasing activity of SWFs in the 2000s, as shown above. He described SWFs as by-products of accumulated national budget surpluses, based on good macroeconomic, fiscal and trade positions. As such, SWFs cannot be seen as traditional pension funds, nor as reserve assets supporting national currencies.15 This “negative definition” got confirmed and refined by Gilson and Milhaupt as SWFs being “investment vehicles that are not central banks, monetary authorities in charge of foreign reserves, or national pension funds, unless they are financed by commodities export”.16 In addition, they’ve included a differentiation between traditional pension funds and SWF like pension funds.17 This differentiation is the reason, why the mentioned SWF of Norway, although it bears the name “Pension Fund”, in fact is categorized as SWF.
Further, the German Council of Economic Experts,18 as well as the US Department of the Treasury19 defined SWFs as funds investing foreign exchange reserves of states, organizationally separated from the traditional management of currency reserves. These funds are actively managed with a long-term perspective and should generate excessive returns compared to a pure currency reserve management. In the context of this work, the definition published by the SWF Institute is the most relevant one, since the later sample selection is influenced by the categorization undertaken by the SWF Institute. Following their definition, a SWF is a
“… state-owned investment fund or entity that is commonly established from balance of payment surpluses, official foreign currency operations, the procedure of privatization, governmental transfer payments, fiscal surpluses, and/or receipts resulting from resource exports. The definition excludes, among other things, foreign currency reserve assets … for the traditional balance of payments or monetary policy purposes, state-owned enterprises (SOEs) in the traditional sense, government-employee pension funds (funded by employee/employer contributions), or assets managed for the benefit of individuals.”20
Further, the SWF Institute indicates that SWFs are less interested in liquidity than in returns, which means the risk tolerance of SWFs is higher than in the traditional foreign exchange reserve management.21 Basically, the definition of the SWF Institute covers the same characteristics than mentioned before, in a more detailed and precise wording.
2.3 Funding Sources of SWFs
As mentioned, the thesis is based on the information published by the SWF Institute. The SWF Institute offers information on institutional investors and refers to itself as “the world’s leading provider of data and research on Sovereign Wealth Funds, pensions and other asset owners”.22 Its service is accessible via the internet.23 If not stated otherwise, fund specific information got retrieved from the fund profiles offered by the SWF Institute.24
The SWF definition already gives an idea about the funding of SWFs. They can be divided into two groups. Either they are funded via commodity revenues or not. Kuwait has been the first country establishing a SWF after they have discovered oil. The stated perspective of this fund is to achieve long-term investment returns, to provide an alternative to oil reserves.25 Other countries with high dependencies on the export of natural resources followed the example of Kuwait and thus, most SWFs are funded by excess revenues from the export of these natural resources. Also, the world’s largest SWF, the Norway Government Pension Fund Global, states its goal as saving for future generations for the time after the oil ran out.26
But not just countries with large natural resource reserves have established SWFs. The largest non-commodity SWF is owned by China and, with over 800 billion USD of assets under management, is managing part of China’s foreign exchange reserves. Figure 2 is illustrating the overall distribution of SWFs to different funding sources. The upper section of figure 2 indicates the share of SWFs per funding segment as percentage of all SWFs. One third of the SWFs are financed on a non-commodity basis. This segment summarizes the SWFs, which are financed via foreign exchange reserves, privatization of state owned assets, transferring state owned assets into funds with less governmental influence, fiscal allocations, or even bond issues.27
Abbildung in dieser Leseprobe nicht enthalte
Figure 2: Origin of SWF Funding28
As can be seen, most SWFs are allocated to the “green” section, which summarizes the SWFs that are financed by commodities. This segment is further divided into the segments of “oil” and “oil & other”, which sums up to 54% of SWFs, and finally, 10% of SWFs are funded by other commodities than oil, on a stand-alone basis.
When assessing the assets under management per funding category, as indicated in the lower section of figure 2, the non-commodity part increases to 42%. This indicates that the non-commodity SWFs, on average, are larger than the commodity SWFs. Further, the lower diagram reveals that the commodity funded SWFs are highly dependent on oil revenues, since only 1% of the assets under management are financed via other commodities such as copper, diamonds or gas alone. The AUM development over time reveals a relatively stable relationship between non-commodity and commodity based funding proportions. The share of non-commodity based SWFs increased by 6% from December 1998 to June 2017.
Besides the distribution to financial sources, the regional distribution of SWF’s funding is interesting, too. Currently, the two main funding regions are the Middle East and Asia, both with approximately 40% of the total assets under management of SWFs. These shares are relatively stable since the end of 2009. The European share peaked in December 2006 with 16.7% and, since then, lies relatively stable around 14%. In absolute terms, the European share in June 2017 was 1,083 billion USD. Having said that Norway owns the world’s largest SWF with close to one trillion USD of assets under management, the European share is heavily dominated by Norway. Detailed information about the development of the assets under management per funding category and geographic origin can be found in appendix 2.
2.4 Scientific Discussion about SWFs
After introducing sovereign wealth funds and revealing the relevance of SWFs for the global capital markets, also in comparison to other big players, the major funding sources got identified. This section is closing the introduction of SWFs with presenting the current scientific discussion on the topic. SWFs are a relatively untouched topic, which also is the reason for the non-existence of a universal definition. As mentioned before, Andrew Rozanov has been the first person using the term “sovereign wealth fund” in 2005.29 Given a first SWF like investment vehicle has already been established by Texas in the nineteenth century, this is relatively late.
The first question arising, after the definition of the fund category got cleared, is; what is the reason for creating SWFs, or which overall objective do SWFs pursue when investing their assets? Considering the publicly available information, the following reasons for establishing SWFs have been identified. The information mainly got gathered from the SWF profiles offered by the SWF Institute,30 but also from expert opinions31 and from official SWF websites.
As discussed, the funds were categorized in two groups. The first group can be summarized as non-commodity based SWFs and is financed via foreign exchange reserves, and other sources mentioned above. The perspective of these SWFs is to invest the existing reserves with a long-term perspective, to generate excess returns, compared to the classical reserve management, which is more focused on liquidity than on generating returns. This also corresponds to the SWF definitions from state agencies.32
The second and larger group is financed via the export of resources, mainly oil. These funds want to create excess returns on a long-term basis, too. In addition, and more important, the resources exporting nations have established SWFs to diversify their government revenues and thus, make the wealth of their nation less dependent on the current situation of the respective commodity markets. This comes together with generating less volatile revenue streams, as well as saving a fraction of the created wealth from exporting resources for future generations. Especially for the oil exporting nations, the steady development towards a decreasing global oil production,33 which will affect the revenues of these countries, plays an important role, too.
Given the state ownership and the financial power of SWFs, a second important question is, if these funds are influenced politically and thus, are acting in favor for foreign agendas. In Germany, a similar discussion revived recently, due to the takeover of Kuka.34 Further, a recent development has been that the communist party of China tries to widen its influence on European and American companies.35 Since some SWF owners, besides China and among others, are states from the Gulf region or Russia, critics of SWFs have declared, SWFs “are the Trojan horse of states that generally are neither democratic nor share the traditions, political systems or legal systems of many OECD [abbreviation for the Organization for Economic Co-operation and Development] countries.”36 This fear is supported by the fact that most SWFs aren’t very transparent regarding their investment behavior, as well as in their strategy.37 The Kuwait Investment Authority serves as an example, again. The SWF states that the fund’s activity is not based on the political or foreign policy interests of the State of Kuwait.38 However, this cannot be assessed independently, since the Emir of Kuwait prohibited the disclosure to the public of any information related to the fund’s work.39 This resulted in a transparency ranking of six, on a scale from one (bad) to ten (good transparency), which is in the mid-field of the Linaburg-Maduell Transparency Index, reported by the SWF Institute. The SWF Institute considers funds with a transparency ranking of eight or higher as being adequate transparent. This only counts for one half of the SWFs included in the ranking.40
The described fear comes with defensive political actions, as analyzed by Cohen (2009). His work is connecting SWFs with matters of national security.41 He has analyzed, if there is a balance between safeguarding prosperity via opportunities for productive international capital, and every nations responsibility to defend the nation’s security. His findings have been, that the fear of politically motivated investments might be overdone42 but is existent and resulted in more financial protectionism.43 This dilemma also got discussed by The Economist. The newspaper stated that, in general, foreign investments are welcomed, but the situation differs, when the money belongs to other governments. In this case, politicians would be more skeptical.44
Bhagat (2008) has analyzed global reactions on rising SWF investments, too. He found out that new legislation like the Foreign Investment and National Security Act of the United States got implemented in several countries. This increase of national legislation is a result of the SWFs secrecy, which could be solved with a more transparent behavior. In a sense, SWFs would be their own biggest enemy, when keeping the publicly available information at a minimum. Further, he summarized that this increased financial protectionism could have the potential to slow down the economic growth globally.45
Other authors get summarized by Yi-Chong (2010), saying SWFs are a fad which will disappear when the financial crisis is overcome.46 Contrary to this opinion, the SWFs’ assets under management have grown from 4.4 trillion USD in December 2009 to 7.4 trillion USD recently. At least the growth rate of the assets under management decreased and is on a pre-crisis level since 2014 / 2015.47 Considering both trends, the SWF industry experienced an excess growth rate during the time of and after the global financial crisis, but is far from being a fad.
In a more actual paper, Yi-Chong (2012) is summarizing the protectionist trend described and states that this development peaked just before the financial crisis kicked in.48 During the crisis, however, the image of SWFs shifted from being a potential national securities issue to be the “lender of the last resort”, especially in context with financial institutes at the edge of being bankrupt.49 With time, SWF owning countries changed their behavior and agreed to more rules, mainly demanded by OECD countries. This resulted in the International Working Group of Sovereign Wealth Funds in 2008. Accompanied by the International Monetary Fund, the group agreed on the “Generally Accepted Principles and Practices”, also known as the Santiago Principles. One important principle, resulting from these discussions, is that SWFs should comply with host country’s disclosure rules. This has been the cornerstone of open dialogues with and about SWFs, which should result in a deeper understanding50 and, thus, should diminish fears and counteract the increased financial protectionism.
Another interesting question that follows is, if SWFs follow an activist strategy within the targeted companies or if they are passive asset managers. An activist strategy is characterized with attempts to influence management and their decisions. These attempts often are accompanied by aggressive behavior like law suits or public campaigning.51 A recent example for an activist investor would be Carl Icahn, who tried hard to convince Apple from a large-scale share buyback in 2013.52 If there would be evidence of such activist behavior, the fear of political influence might be reasonable, and would increase. The opposite would be a passive investor. He invests with a long-term perspective and is not actively influencing the business. A passive investor, who is disagreeing with the management, would rather sell his shares instead of fighting for change.53 As mentioned, some SWFs specifically state that their business is solely commercially driven and not influenced by politics. Rose assessed this question for US companies in 2008 and states that SWFs are likely to be characterized as passive investors. This is also driven by US legislation, which pressures SWFs not just to avoid a controlling ownership, but also to avoid the execution of too much influence, as well. Such influential behavior would lead, if observed by state agencies, to further disclosure obligations and potential liability, if the SWF fails to disclose accurate information about itself.54 Since the United States of America (USA) are one of the key markets for SWF investments, and the fourth largest target country in the later analysis-sample, Rose’s findings are likely to be valid for other markets, too. This got confirmed by Mietzner et al. in 2015, who are stating that there is evidence that SWF managers mostly act as passive investors.55 The fact that a vast majority of SWF investments is concentrated in the region of holding a non-controlling percentage of shares post transaction is supporting these findings.56
Summarizing, the scientific discussion is centered around governance issues, the unclear intensions of SWFs, and the resulting consequences. SWFs are welcomed, since they are increasing domestic investments. But even if some SWFs state, that their investment purpose is not influenced by political agendas, the fear of the opposite exists because of a lack of transparency. Consequently, this fear harmed the movement of capital. To counteract the increasing protectionist approaches, the Santiago Principles where introduced in 2008 by SWFs themselves, trying to implement common rules to achieve more transparency. Moreover, extended disclosure and liability legislation, like in the US, would come into effect, if state agencies observe an influential behavior. This definition of influential behavior is not limited to the ownership of controlling shares but also covers the potential exercise of influence outside of formal governance processes.
SWFs have gained large financial power during the last years, which is measured in trillions of US Dollars. The largest SWFs’ assets under management exceed the gross domestic product of countries like the Netherlands or Switzerland.57 As such, SWFs are an important part of the global financial markets. A question that has not been assessed so far is, where is the large amount of money going to? Which selection criteria are used to decide for potential targets? And Thus, which company characteristics are influencing the likelihood of getting targeted by SWFs? Most SWFs are lacking transparency and are not stating clear investment strategies. In this thesis, publicly available information, mainly gathered from the SWF Institute, is used to find systematics in the accumulated investment behavior of SWFs. As first step, the available data is analyzed and the sample selection process, as well as the resulting sample, is described in the following chapter.
3 Data and Sample Overview
This chapter is about the source of the used information, illustrates the sample selection process and describes the resulting dataset.
3.1 Data Sources and Sample Selection
First, it is explained where the information on SWF activism comes from and which selection criteria have been used to choose the relevant sample of SWF transactions. Also, this describes the SWF transaction database more closely. Second, it is described, how the various sources of data get combined to create the used master database for later regression analysis and which further company related information is used.
3.1.1 Sovereign wealth fund transactions
This thesis focuses on global sovereign wealth fund transactions and is based on information from the Sovereign Wealth Fund Institute. First, it was necessary to choose some selection criteria to identify an appropriate number of relevant transactions. The following descriptive statistics are referring to the SWF transaction database, which is stored in digital appendix 2.58
The entire set of observations consists of 43,578 entries from year 1974 up to 2017. The observations are divided into ten “Acquirer Types”. Three out of them are covering more than 99% of the observations and are structured as shown in table 2. The relevant SWF transactions account for 48% of the database.
Abbildung in dieser Leseprobe nicht enthalten
Table 2: Main Acquirer Types
The next step is to choose a period of time. As described in chapter two, SWF transactions started to reach a substantial number in towards the end of the 2000s. In order to capture a significant period, and to maintain a large number of transactions, the captured sample starts in 2000 and ends with the year 2014. This results in a dataset of 13,437 observations.
Abbildung in dieser Leseprobe nicht enthalten
Table 3: Percentage of Shares Owned Post Transaction
Now, the remaining observations are examined regarding the percentage of shares owned post transaction. Table 3 reports the distribution of investments after percentage owned post transaction. In most cases (58.8%), the percentage of shares owned is below 1%. An additional quarter is ranging between 1% and 5%. This paper focuses on transactions, where the fund ownership is equal to, or exceeding five percent post transaction. This comes with two reasons. At this point, it is mandatory to disclose major ownership of voting rights. The threshold of 5% is valid under United States59 and European Union law.60 This obligation provides additional reliability to the quality of information. Subsequently, the goal of the thesis is to examine, whether there is a significant difference between targets and the respective non-targeted peers. This potential difference would be most relevant in strategic investments. These are sophisticated investment decisions in favor for companies fitting into the fund strategy, and are connected to large capital investments. Furthermore, “in general”61 Datastream considers ownership of more than 5% as strategic.
In the second last step, the filed “Investment Type” gets examined more closely. As it will be shown later, trying to match the database with country sheets from Eikon has shown that the investment type “Listed Equity”, in contrast to the other types, has a good coverage. Mainly because the historical information, accessible via Eikon, is about traded equity instruments.62 This represents 743 transactions or 62% of the remaining 1200 transactions.
Finally, the number of transactions per Target Country is analyzed. To correct for target country specifics, e.g. historical funding with more or with less debt, most regressions discussed in chapter four consider Target Country fixed effects. Thus, a single Target Country needs to have a substantial amount of transactions.
Abbildung in dieser Leseprobe nicht enthalten
Figure 3: Target Countries and Number of Observations
A second reason to cut off target countries with less transactions is the decreasing added informational value per country. For those reasons, the analysis considers all countries with ten or more documented transactions during the selected period of 15 years. Figure 3 documents the resulting target countries and the respective number of transactions per country. The sample, for now, consists of 646 observed transactions in 18 countries. The cut of value of ten transactions per country leads to a loss of 100 transactions (approximately 13%). A detailed figure, which also includes the countries with less than ten observations, can be found in appendix 3.
This section has outlined the sample selection process and has described the SWF transaction database more precise. The resulting sample of 646 observations got selected using five criteria. These are summed up in table 4 on the next page.
Abbildung in dieser Leseprobe nicht enthalten
Table 4: Sample Selection Criteria
The database offered by the SWF Institute has one shortcoming. None of the information uniquely identifies the target companies. However, this is necessary to find additional information and will be addressed in the next section.
3.1.2 Unique company identifiers
Databases in general, as well as the financial database offered by Thomson Reuters, are working with several identifiers. Commonly known, there is the International Securities Identification Number, the German “Wertpapierkennnummer” or the Datastream Code (DSCD). An Identifier uniquely identifies a security, and allows to research a wide range of company specific information. This includes historical stock prices, performance measures, and more. Datastream, for example, is covering “more than 8 million financial instruments … and over 13,000 different facts.”63
The SWF transaction database is missing any form of unique company identification, but contains the name (Target Name) and the origin of the company (Target Country). Combined, these offer the possibility to search for the right company in country specific lists containing all listed companies. For this purpose, country sheets, downloaded from Eikon, got used.64 These files contain the DSCDs and company names, among others, of listed companies per country. Besides existing companies, the country sheets also contain delisted or dead ones.
The selected SWF transactions got matched with the respective Eikon country sheets with respect to the name column in both files. This matching got performed using a searching algorithm for Stata14.65 Before searching for matching names, it is needed to clear the names for common expressions like “Inc.” or “Aktiengesellschaft”. Over thirty expressions got identified, which got deleted from the target names to ensure a useful fit. Another measure to improve the matching result was to delete non-alphabetic and non-numeric characters from the names, and since Stata is case sensitive, all names got recoded in capital letters.
[...]
1 CF. Rozanov (2005), p. 52.
2 Cf. SWFI (2017a).
3 See Chapter 2.3.
4 Cf. SWFI (2017c).
5 Cf. Anonymous author (2017), p. 77.
6 See Chapter 2.4.
7 Cf. SWFI (2017c) – Information stored in digital appendix 1 – SWF Ranking.
8 Cf. Kuwait Investment Authority (n.d.-a).
9 Cf. SWFI (2017c) – Information stored in digital appendix 1 – SWF Ranking.
10 Digital appendix 2 – SWF Transaction Database.
11 Cf. SWFI (2017a).
12 Cf. SWFI (n.d.-c).
13 Cf. Anonymous author (2017), p. 77.
14 Multiple sources. Precise sources and a more detailed table can be found in digital appendix 1 – SWF Ranking.
15 Cf. Rozanov (2005), p.
16 Cf. Gilson / Milhaupt (2008), p. 13.
17 Cf. Lyons (2008), p. 23.
18 Cf. Anonymous author (2007), pp. 397f.
19 Cf. US Department of the Treasury (2007), p. 1.
20 SWFI (n.d.-a).
21 Cf. SWFI (n.d.-a).
22 Cf. SWFI (n.d.-b).
23 www.swfi.com / www.swfinstitute.org. Currently, the SWF Institute is reorganizing its service and will merge both websites to www.swfi.com.
24 Cf. SWFI (n.d.-c).
25 Cf. Kuwait Investment Authority (n.d.-b).
26 Cf. Norges Bank Investment Management (n.d.).
27 Cf. footnote 26 in accordance with Yi-Chong (2012), p. 193.
28 Cf. SWFI (2017c). The information can be found in digital appendix 1 – SWF Ranking.
29 Cf. Rozanov (2005), p. 52.
30 Cf. SWFI (n.d.-c).
31 For example, cf. Lipsky (2008).
32 Cf. Anonymous author (2007), pp. 397f. in accordance with US Department of the Treasury (2007), p. 1.
33 Cf. Demirbas (2009), pp. 212f.
34 von Andreae (2017).
35 Cf. Kempf (2017), pp. 4f.
36 Yi-Chong (2010), p. 1.
37 Cf. Bahgat (2008), p. 1191.
38 Cf. Kuwait Investment Authority (n.d.-b).
39 Cf. Article 8 Law No. 47 of Kuwait.
40 Cf. SWFI (2017b).
41 Cf. Cohen (2009), pp. 713f.
42 Supported by the findings of Drezner (2008), p. 120.
43 Cf. Cohen (2009), pp. 713, 731.
44 Cf. Anonymous author (2008).
45 Cf. Bahgat (2008), pp. 1191, 1201-1203.
46 Cf. Yi-Chong (2010), p. 1.
47 See appendix 2.
48 Cf. Yi-Chong (2012), pp. 193f.
49 Adam / Totaro (2011).
50 Cf. International Forum of Sovereign Wealth Funds (2008).
51 Cf. Lin (2015), p. 472.
52 Cf. Foroohar (2013).
53 Cf. Lin (2015), pp. 467, 472.
54 Cf. Rose (2008), pp. 104f.
55 Cf. Mietzner / Schiereck / Schweizer (2015), p. 314.
56 Digital Appendix 2 – SWF Transaction Database.
57 Cf. The World Bank (2017).
58 Database of June 2017.
59 Section 12 Securities Exchange Act of 1934.
60 Article 9 § 1 Transparency Directive.
61 Cf. Datastream (2017).
62 Cf. Datastream (2010), p. 1.
63 Datastream (2010), p. 1.
64 With friendly support of Daniel Huber; The files can be found in digital appendix 3.
65 All Stata do-files, which have been created during this thesis, can be found in digital appendix 8.
-
-
-
-
Upload your own papers! Earn money and win an iPhone X. -
-
Upload your own papers! Earn money and win an iPhone X. -
-
Upload your own papers! Earn money and win an iPhone X. -
-
Upload your own papers! Earn money and win an iPhone X. -
-
Upload your own papers! Earn money and win an iPhone X. -
-
Upload your own papers! Earn money and win an iPhone X. -
-
Upload your own papers! Earn money and win an iPhone X. -
-
Upload your own papers! Earn money and win an iPhone X. -
-
Upload your own papers! Earn money and win an iPhone X. -
-
Upload your own papers! Earn money and win an iPhone X. -
-
Upload your own papers! Earn money and win an iPhone X. -
-
Upload your own papers! Earn money and win an iPhone X. -
-
Upload your own papers! Earn money and win an iPhone X. -
-
Upload your own papers! Earn money and win an iPhone X. -
-
Upload your own papers! Earn money and win an iPhone X. -
-
Upload your own papers! Earn money and win an iPhone X.