Sovereign Wealth Funds and Wealth Renationalization: Political Determinants of Investment Decisions

Размещено на http://www.allbest.ru/

THE GOVERNMENT OF THE RUSSIAN FEDERATION

FEDERAL STATE AUTONOMOUS EDUCATIONAL INSTITUTION

OF HIGHER PROFESSIONAL EDUCATION

“NATIONAL RESEARCH UNIVERSITY “HIGHER SCHOOL

OF ECONOMICS”

FACULTY OF SOCIAL SCIENCES

DEPARTMENT OF POLITICAL SCIENCE

Sovereign Wealth Funds and Wealth Renationalization:

Political Determinants of Investment Decisions

Final qualifying work (bachelor graduation paper)

The field of study 41.03.04. Political Science

Group № 121 (educational program «Political Science»)

Vladimir Grigoryev

Moscow - 2016

Introduction

Topicality. According to the International Working Group, “Sovereign Wealth Funds (SWFs) are special purpose investment funds or arrangements that are owned by the government. Created by the government for macroeconomic purposes, SWFs hold, manage, or administer assets to achieve financial objectives, and employ a set of investment strategies that include investing in foreign financial assets” (International Working Group of SWFs, n.d.). From this definition, it follows that SWFs are economic tools in hands of the government, which are essential for managing excessive financial resources.

The performance of these institutions fact has immediately provoked intensive attention of economists and political scientists. In 2007, Financial Times described these institutions as funds “that shake capitalist logic” (Financial Times, 2008). The public attention was attracted to the evidence suggesting that huge economic resources are accrued in hands of policy-makers. The process was called “The Rise of SWF”. Later, the issue was analyzed by the Sovereign Wealth Fund Institute that showed that total assets of SWFs doubled in the period from 2008 to 2015 while the total number of newly founded SWF increased by more than 40 units (Sovereign Wealth Fund Institute, 2015).

Hence, the SWFs as a particular phenomenon are playing a rising role in terms of growing number of these institutions and growing amount of wealth under management. In addition, these institutions create new opportunities for policy-makers. Thus, the problem investigated in the research is the influence of political factors on the SWFs as economic tools.

The novelty of the research can be explained by incorporating regime factors into the understanding of SWFs' establishment and performance.

The research question of the study is following: How do the political motivations influence the investment decisions?

The object of the study comprises Sovereign Wealth Funds, which are available in our database.

The subject of the research includes investment decisions made by Sovereign Wealth Funds.

The objective of the study is to investigate the significance of political motivation for wealth management within the SWFs. Achieving stated objective requires the accomplishment of the following tasks:

1. To derive important factors influencing the investment decisions of wealth management.

2. To perform an analysis of significance of political factors controlling for other potential determinants of decision-making.

Methodology. A defining feature of a SWF is its reliance on institutional constraints. Usually, the government initiates establishment of the fund and the government pursues governing it. Thus, the restrictions falling on the government are valid constraints for the operation of the fund. In this sense, the SWF's establishment, its institutional arrangement and performance can be seen as an equilibrium, a result of specific interests of influential actors. This relates to one of the two approaches within the field.The first one treats institutions as conditions and rules which restrict the number of available strategies for actors. The institutions are seen here as producers of equilibria. The second approach, which is used in the study, states that the rules of the game are formed and sustained due to the fact that they suit the interests and resources of actors (Gehlbach et al., 2015). That is why we have chosen rational choice institutionalism as the methodological approach for this study.

Problem statements:

1. There is a non-linear relationship between the level of democracy and the direction of investments made by sovereign wealth funds.

2. Autocratic and democratic regimes have different incentives behind their investment decisions.

2.1. SWFs in democratic regimes seek financial return from wealth management.

2.2. SWFs in autocratic regimes are instruments of political survival for incumbents.

Theoretical significance. This study could be applied in addition to previous and current investigations in the sphere of governmental wealth management and in the newly evolving sphere of SWFs' performance analysis. Incorporating political factors into the modelling the behavior of SWFs complements existing research and provides greater depth of knowledge about SWFs' behavior.

Professional significance. The significance of the project for professionals is presented in the following reasons:

1. This research can become a platform for further studies in the sphere of sovereign wealth management.

2. The study can advance the accurateness of predictions for SWFs' investments by considering political regime in a given country.

wealth management political factor



Chapter 1

Wealth management is naturally associated with the motivations of ruling elites. The approach advanced by De Mesquita exerted great influence on theoretical thinking about the conduct of political actors. According to his approach, rulers are primarily focused on their own survival, which means the prolongation of the leader's office term (De Mesquita and Smith, 2005). Furthermore, theoretical modelling of incumbents' behaviour implies that wealth management (for instance, rent distribution and provision of public goods) is an important tool of politics, and in particular, of non-democratic politics (Gehlbach et al., 2015). In this sense, SWFs are institutions can serve for incumbents as perfect economic tools.

There is no consensus on exactly what constitutes a “sovereign wealth fund.” We consider governmental funds as being sovereign wealth funds, if they match the Sovereign Investment Lab's (SIL) selection criteria, which defines a SWF as:

(1) being an investment fund rather than an operating company,

(2) being owned by a sovereign government, but organized separately from the central bank or finance ministry to protect it from excessive political influence,

(3) making international and domestic investments in a variety of risky assets,

(4) being charged with seeking a commercial return, and

(5) being a wealth fund rather than a pension fund, meaning that the fund is not financed with contributions from pensioners and does not have a stream of liabilities committed to individual (Miracky and Bortolotti, 2009).

Hence, one of the most important defining features of a SWF is higher reliability on institutional constraints. Usually, the government initiates the establishment of the fund and governs it. Thus, the restrictions falling on the government are valid constraints for the operation of the fund. In this sense, the SWF's establishment and its institutional arrangement can be seen as what is called an equilibrium in Political Economy - result of specific interest of influential actors. This relates to one of the approaches within the field. The first one treats institutions as conditions and rules which restrict the number of available strategies for actors. The institutions are seen here as producers of equilibria. The second approach, which is used in the study, stated that the rules of the game are formed and sustained due to the fact that they suit the interests and resources of actors. Thus, the institutions are seen here as equilibria themselves (Gehlbach et al., 2015).

The literature investigating the nature of SWFs is split into two major parts. The first group uses the normative approach and addresses the issue of how such funds should operate and what motivation should be considered by making investment decisions. The other group is based on empirical approach and tries to answer the question how the SWFs are functioning in reality. In other words, the second part looks at real motivation of sovereign wealth funds' investments.

Normative literature. Main reasons for establishment of SWFs are connected with unexpected but often long-term inflow of government revenues. Usually, such high revenues come from commodity selling (for example, oil & gas in Norway) and extremely high trade imbalance, when exports exceed imports extremely (for example, in China). Hence, the establishment of a SWF is a solution for the issue of accumulating and managing extra wealth.

The question of performance is more complicated. As stated on major SWFs' websites (see, for instance: Abu Dhabi Investment Authority or Government Pension Fund of Norway), the funds are aimed to provide resources for long-term prosperity and for well-being of future generations. Thus, officially, they are established for the preservation of present extra revenues and for their accurate and wise management providing opportunities for future problem solving on the country level. For instance, external and internal economic shocks may pose a threat that can be addressed using SWFs' resources (Raymond, 2012).

In the above perspective, SWFs should follow proclaimed goals and not be influenced by political struggles within the government. The performance of the funds is devoted to operating as pure economic agents accumulating excessive revenues and investing them wisely into other assets. Thus, the role of the SWFs is to provide the government with financial resources usually generated from dividend payments, and the government may appeal to them when needed. In this sense sovereign wealth funds should function as private hedge funds paying dividends to their shareholders The only shareholder in case of a sovereign wealth fund is the government or its entity. For instance, the ministry of finance or an agency obedient to the ministry.. The government adds these payments to the budget balance and decides how to spend them. Thus, the government should not interfere into the investment policy of the fund and should be interested in as high and secure returns as possible.

However, scholars developing the normative approach argued that SWFs should sometimes deviate from pure economic logic of profit-maximizing (Bertoni, 2014). This argument is due to the fact that SWFs are sometimes considered as one of the macroeconomic tools helping not only in long-run but in short-run perspective, too. In this case investments could be made not by using the dividends but by extracting wealth under funds' management, or, in other words, the government may be required to make inefficient investment decisions. Such a motivation lies in the economic sphere, too. However, economic interests are moved above the fund's level to the country's level. In 2010, the financial system of Ireland was under threat because of huge troubles two major Irish banks faced. The Irish government decided to use its “National Pension Reserve Fund” to bail out these banks to keep the whole economy safe from a potentially greater downturn. This is kind of economic logic of managing wealth, too. However, this economic logic is adjacent to political motives because greater economic downturn would considerably harm the government by the next elections. Thus, political considerations are taken into account here as well as economic ones.

The performance of SWFs should not be influenced by politicized decisions. Politicians may be intended to use resources from SWFs to finance infrastructure projects being, actually, proper to other agencies (Gelb, 2014). This leads to undermining SWFs' performance. Under politicizing we mean here the presence of external motivations for the fund interests, which are not connected with the basic proclaimed ends of the fund's establishment.

The apogee of the normative approach is presented in the article by Ang (2010). The author suggests four benchmarks for the operation of SWFs. First, the fund should not spend the money immediately (Legitimacy). Secondly, a broader policy environment should be guaranteed for the operation of the fund (Integrated policy). Third, managers of the fund should seek to maximize returns on investments (Performance). Finally, for the fund to operate efficiently, corporate governance, developed capital markets and free cross-border capital transactions need to be ensured (Long-run equilibrium) (Ang, 2010).

Normative approaches present a general frame for the perception of SWFs. However, obviously, their real performance, additional non-economic triggers of establishment and non-economic causes of efficiency are subject to empirical analysis. The review of empirical works is presented in the next section.

Empirical studies. Many scientists developed the idea that establishment of SWFs has shifted the balance of economic and financial power towards the political sphere (Lyons, 2008). Therefore, studies on the nature of SWFs raised questions on trust to these institutions that can operate in accordance with rules of politics, but not economics (Helleiner, 2009). Thus, scholars started to investigate the behavior of SWFs on empirical basis.

In comparison to other global investors, SWFs invest predominantly into countries having good legal institutions. Their distinguished feature is chasing past returns and keeping portfolios which show poor geographical and industrial diversification. This describes most SWFs as more conservative and less activist investors. Such biased characteristics are more appropriate to less transparent SWF usually originated from more autocratic countries (Chhao, 2008). It is worth mentioning that poor diversification is an important sign of non-financial motives of the funds.

In an explicit form it was showed by Knill, Lee, and Mauck (2012). Using a specific statistical model, they concluded that weak political relations enhance the probability of SWFs investment between countries. This can be contradictory to profit-maximizing behavior, and makes an intuition of non-economic motives for investment (Knill, 2012). Weak political relations could mean less politicized relations, which implies that the investments are considered as more secure if there is a less politicized but rather economic agenda in relationships between two countries.

Moreover, SWFs' investment decisions are strongly biased towards particular industries. Often risky assets, assets based in home region, and infrastructure companies' equities comprise a huge part of the SWFs' portfolios. Near 45% of portfolios' variation account for industrial motives (Dyck, 2011).

The question of political influence on the performance of SWFs takes here the central position. As Bernstein et al (2013) show, if politicians are taking part in making investment decisions, the SWF is more likely to invest domestically (Bernstein, 2013). This implies that politicians are implementing their political interests, which are usually domestic in nature. In contrast, our study makes a broader argument about the political influence on investment strategies. The number of politicians taking part in decision-making process and politically biased behavior of decision-makers can vary according to the different institutional arrangements in different polities. For instance, we expect that in established democracies a greater control is usually imposed on the investment decisions. Hence, politicized investment strategies will be most likely prevented.

As stated above, the sovereign nature of SWFs leads to suspicion by market participants. Thus, SWFs might refrain from taking an active corporate governance role in target companies in order to not generate political opposition or regulatory backlash. There is the evidence indicating that SWFs play only a small role in target firm governance and rarely take seats on target firm boards (Mehropouya, Huang, and Barnett, 2009; Rose, 2008). In addition, the monitoring role of SWFs might be reduced by their reluctance to divest, as selling large blocks of shares could also trigger political reactions and resentment amongst local management, regulators, and market participants. According to this hypothesis, SWFs will tend to be very passive investors who, at best, do not contribute to effectively monitoring target firm managers and, at worst, can help entrench underperforming managers. As a consequence, investments by SWFs will create less value for target firms than comparable investments by private investors who exercise unconstrained ownership.

Interestingly, in specific circumstances, the very dominant ideology of government might play role. The most obvious example is the Norway Government Pension Fund, which is under strong influence of political ideas for environment protection (Clark, 2010). In May 2015, the fund was not allowed to invest into companies, whose revenue consists of coal extraction for more than 30% The biggest sovereign fund disposed of Russian coal. (2015, May 28). Retrieved from http://www.rbc.ru/business/28/05/2015/556745d79a7947eb43e036f2 . It becomes the reason why the fund has sold equities of a Russian company, which is dealing partially with coal extraction.

While the statistical studies presented in the first part of the section are based on financial discourse, various case-studies appeal to a more crystallized Political Economy. And to give a broader theoretical basis to the research, it seems plausible to think about the establishment, performance and efficiency of SWFs in terms of strategic choice made by governing elites, especially in autocratic regimes, in which the SWFs are primarily created.

The theoretical tradition founded by De Mesquita had a widespread appeal in the literature on SWFs. For example, Hatton and Pistor developed an approach that ascribes a specific role to SWFs. They suggest that establishment of SWFs is a step towards stronger autonomy of the state on international arena (Hatton and Pistor, 2011). The concept of autonomy maximization is central in their works. It is argued that creating financial institutions accumulating huge financial resources provides rulers with greater degree of independence and autonomy in decision-making. It is worth mentioning that the authors' scheme followed directly from De Mescquita's works in the field of Political Economy. The idea was also supported by economists who showed that the efficiency of SWFs is comparable to that of private funds (Alhassan, 2013). First of all, it means that foreign investments of SWFs are targeted to accumulate wealth and states become more autonomous economically and less dependent on taxes and foreign direct investment from other countries.

It seems that performance and efficiency of SWFs can be at a significantly lower level because of partially bureaucratic nature of the institution (Gailmard, 2012). However, as recent studies show (Alhassan, 2013), their performance could be at a very high level. Despite the evidence, politics has an inevitable impact on governmental bodies, which are SWF, too. A great impact on SWFs' performance is imposed by elite fragmentation (Shih, 2012). For members of political class, the institutions provide an opportunity for already mentioned political survival. Thus, time-horizon issue arises with respect to the problem. Lack of elite unity leads to inefficiency of SWFs due to the fact that investment decisions are made under uncertainty and targeted on short-term gains. In contrast, clear and certain placement of power within the political system enables making sound investment decisions that serve long-term interests and strategic development.

While the rise of SWFs challenges the usual understanding of governmental activities, current research is mainly focused on economic performance of these institutions and on influence of investments on the value target companies. The connection between the political determinants and investment policies is very rarely described by the scholars. Only handful of studies elaborate on political nature of SWFs' establishment and performance. As literature review shows, creating SWFs may have long-term political intentions and consequences. In addition, the intervention of such theoretical concepts as, for instance, time-horizon, principal-agent relationships, rent-seeking, and elite fragmentation distorts significantly the normative economic perception of SWF and makes understanding of its establishment and performance much more politically determined.

As in case of other public institutions like, for example, a central bank, a sovereign wealth fund has its own officially proclaimed goals. However, as several case studies demonstrate, the real performance is subject to formal and informal rules within the political elite. The literature does not provide evidence that the described mechanisms work on broader set of cases.

In our analysis described in the following chapter, we present a statistical analysis of potential political motivations behind investment decisions made by sovereign wealth funds.



Chapter 2

General model. Analyzing the existing literature, we formulate with the following common model of a sovereign wealth fund. First, it is usually risk averse, implying that the investment decisions are made into the assets which have a relatively low risk and a relatively low return, correspondingly. Secondly, SWFs are passive investors, meaning that they usually acquire such an amount of equities of a single company that does not require making information about them public. Hence, in this perspective governments ruling SWFs are seeking wealth preservation and accumulation without active and risky steps.

By revising the literature review, certain challenges arise in relation to regimes managing wealth through the sovereign wealth funds. These challenges are generally connected with the common problem of any wealth management. This problem is the choice between today's spending and preservation of wealth for the future. Both options are associated with possible incentives of ruling elites in authoritarian and democratic regimes.

Hypothesis 1. The theory provides us with several possible situations in which incumbents may want to spend wealth immediately and not to handle as fiduciary investors managing national funds. These situations are associated with the term of political survival, and wealth plays an important role here. Incumbents face the problem of wealth allocation to their selectorate. However, there is a difference between the incentives of authoritarian and democratic regimes. In an authoritarian environment, where the selectorate comprises smaller number of actors and the institutional environment favors the ruling elite (Shedler, 2006), the allocation of wealth is a very challenging and important element of being at power because the selectorate creates necessary support for the ruling elite. For example, the selectorate may consist of powerful interest groups from business. Making investments into the business of these groups will advance the support of the ruling elite. In the democratic environment the selectorate is equal to the electorate of the ruling elite (i.e. the majority). Thus, firstly, the wealth required for sufficient wealth allocation could become too high. Secondly, the institutional structure in democracies may prevent ineffective wealth management.

To sum up, the hypothesis is that the incentives for fiduciary behavior of ruling elites are stronger in democratic environment, and we expect that the spending will be made more efficient.

Hypothesis 2. Despite the fact that authoritarian regimes are theoretically more prone to non-fiduciary investments, incentives for wealth preservation are also present und can be influential. The reason is that the wealth preservation contributes to the prolongation of the regime is because it provides additional future resources for the incumbent. Thus, we end up with the problem of certainty and time-horizon. In regimes with a less institutionalized political environment (the rule of power rather than the rule of law), the time-horizon is shorter (Li, 2009). It can prevent incumbents from wealth preservation and cause the spending of funds for short-term needs. Here again it is worth mentioning that these short-term needs are connected with spending in the home country. However, if the regime creates such rules, which ensure longer time-horizon, they can follow the strategy of wealth preservation.

In the democratic environment the fiduciary behavior of ruling elites is usually prescribed and controlled by law. To conclude, we end up with theoretical expectations, which create the second hypothesis that the political environment with lower level of certainty is associated with the non-fiduciary wealth management.

Testing the hypotheses. To test the hypotheses, the sample of SWFs' transactions is analyzed. The initial sample is confined within the period from 2006 to 2014 and comprises the following data: sovereign wealth fund's name, sovereign wealth funds' home country, transaction target country, and deal amount (if publicly announced) Fletcher SWF Transaction Database. Tufts University, n.d. Web. I would like to express great gratitude to Professor Patrick Schena from the Fletcher School, Tufts University, who agreed to provide data for my research. . Additional variables were added to the dataset. This additional set of variables includes the Doing Business Index score for target country, GDP per capita in home country, the Executive Constraints score from the Polity IV project. Additionally, one more variable was computed for the analysis. This variable indicates whether the government has established and operates more than one SWF.

According to our theoretical expectations we assume that the less fiduciary behavior is connected with the higher number of domestic investments, with the renationalization of wealth. That is connected with the fact that the most secure and less risky assets are based in the USA and in the European Union, which are the major target countries for the for investments of private funds and sovereign wealth funds. In our sample, 83% of all investments were made into foreign assets, indicating more efficient investment policy. It is worth noting that the most sovereign wealth funds are not based in European Union or in the USA There is one American fund: Alaska Permanent Fund.. The following model is presented to show the relationship between level of democracy and the investment decisions made by sovereign wealth funds. Here is the detailed description of the model.

Dependent variable

Direction of investment. The outcome variable is a binary variable, which takes a value of 0 if the target country is the home country. In other words, it takes the value of 0 if the wealth is renationalized. Accordingly, the value of the variable is 1 if the investment is made abroad.

Independent variables

Political variables:

Executive Constraints. It is a numeric measurement of constraints on the executive power in a country. Because the SWFs are owned by the governments, we expect that the constraints on investment decisions which a given SWF can make correspond to constraints on the government itself It is rational to expect that the democratic environment with much more control on political sphere exerts higher constraints on the government. For example, in Norway the parliament plays an active role in regulating the performance of national SWF “Government Pension Fund - Global”. . The values of the indicator vary from 1 to 7. Lower values indicate less constraints on executive power, and vise versa.

Executive Constraints Squared. This indicator was calculated from the previous one to test for the non-linear relationship.

Polity IV score. This is a numeric measurement for the position of a country on a democracy-autocracy continuum. The range of values is between -10 and 10. Higher values of the score illustrate democratic developments within a country. It is worth noting that the regimes falling into the interval between -10 and -6 are described as autocracies, those between 6 and 10 are called democracies, and those between -6 and 6 are named anocracies, which means, actually, hybrid nature of regime.

Polity IV score squared. This indicator was calculated from the previous one to test for the non-linear relationship.

Control variables:

Logarithm of total deal size. Here it is included as a control variable.

Investment climate. This predictor is measured with the Doing Business Index. This variable shows easiness of making business in a given country. It indicates the economic climate and is a proxy for the possibility of achieving success in business. Actually, we should expect that a fiduciary investor would invest with higher probability into countries with high Doing Business score. The Doing Business Index website is calculated by the World Bank Group. The Index was taken from the official website of the organization.

Logarithm of GDP per capita. Here it is included as a control variable. This variable controls for an important counterargument of the whole model. The fact that many sovereign wealth funds are placed in developing countries creates an argument that the domestic investments are required for developmental purposes like, for instance, building infrastructure. Thus, incorporating this variable into the analysis will control for such a factor. This indicator was withdrawn from the official website of World Bank Group.

Mandate. This binary variable indicates whether there are other functioning funds in the country. The value 1 of the indicator shows the presence of them. This indicator captures the fact that some parallel funds may be established for greater investing into domestic assets. This is important variable because it captures the possible specialization of funds within one polity. For instance, in the United Arab Emirates there several funds, which are dispersed among major families of the rulers and which are dispersed within certain families. For instance, emirs of Abu Dhabi established several funds: “Mubadala Development Company”, “Abu Dhabi Investment Authority”, “Abu Dhabi Investment Council”, and “International Petroleum Investment Company”. The first has a specific mandate for keeping certain amount of wealth in the home country's assets. The latter is specified to make strategic investments in oil and gas sector around the world to secure country's position in this sphere. Thus, a specialization may influence the investment policies of the SWFs. However, it is important not to overestimate this factor because a mandate usually prescribes the amount of wealth, which should be allocated into domestic assets, while our data traces the transactions made by the funds but not the allocation.

Description of the analyzed data. The original data on transactions comprises 1030 observations. After adding other indicators the number of observations is reduced to 462 transactions. The reduction in data points is primarily connected with the lack of data on some variables for certain years and countries. For example, Doing Business score was not available for many countries before 2010. Further major limits were set by lack of Total deal size data. However, the range of years stayed the same as in the original data set: from 2006 to 2014. The number of SWFs was reduced, too. While in the original sample there are 32 sovereign wealth funds, the analyzed sample includes only 18. On average the SWFs from the analyzed sample were more active than the excluded funds: each fund from the analyzed sample made 43.3 transactions on average, and each other only 17.9 transactions. Thus, we think the sample has a not meaningful information loss.

It is worth noting that the mean of Polity score in our sample is about -3. It indicates the fact that the sovereign wealth funds are more preferred by authoritarian regimes, and that the higher number of transactions was made from authoritarian environment, from rather authoritarian countries.

Table 1. Logistic regression results

Variables	Models
	I	II	III	IV	V	VI	VII	VIII	IX
Investment Climate	-0.011*** (0.003)	-0.011*** (0.004)	-0.010** (0.004)	-0.010** (0.005)	-0.010* (0.005)	-0.010* (0.05)	-0.01** (0.00)	-0.011** (0.005)	-0.011** (0.005)
Logarithm of GDP per capita	-	0.247 (0.157)	0.345* (0.176)	0.208 (0.161)	0.378** (0.176)	0.444** (0.182)	0.108 (0.169)	0.380** (0.191)	0.358* (0.188)
Logarithm of total deal size	-	-	-0.198 (0.123)	-0.194 (0.124)	-0.221* (0.132)	-0.217* (0.132)	-0.186 (0.127)	-0.209 (0.134)	-0.214 (0.135)
Executive Constraints	-	-	-	-	-0.281*** (0.086)	-	-	-	-
Executive Constraints Sq.	-	-	-	-	-	-	-	-	0.039*** (0.011)
Level of Democracy	-	-	-	-	-	-0.078*** (0.024)	-	0.080*** (0.025)	-
Level of Democracy Sq.	-	-	-	-	-	-	0.014** (0.007)	0.014* (0.007)	0.019** (0.008)
Mandate	-	-	-	1.935*** (0.361)	1.796*** (0.368)	2.009*** (0.362)	2.827*** (0.551)	2.843*** (0.554)	2.764*** (0.559)
Constant	1.915*** (0.152)	1.079** (0.547)	1.745** (0.852)	1.547* (0.826)	2.019** (0.869)	0.559 (0.910)	0.972 (0.885)	0.085 (0.996)	0.593 (0.948)
McFadden	0.018	0.023	0.031	0.132	0.161	0.160	0.143	0.170	0.176

Results. Table 1 shows results of the regression analysis. In the models I-IV we sequentially introduce the control variables. In model V we add the parameter of executive constraints, one of the variables of interest. The coefficient at the variable occurs to be significant at the 1% significance level. However, the interpretation contradicts our expectations. According to this model, the probability of investing into foreign assets rises as executive constraints decrease, while we predicted that the relationship works in opposite direction. However, it can be rather connected with higher number of active SWFs from countries with the lower level of constraints on executive power and lower level of democracy.

If we turn to regime score “Polity”, we see the same relationship. According to the model VI, the effect of regime score is significant at 1% significance level. The sign of the coefficient is negative, too.

However, the introduction of non-linear terms changes the interpretation and makes the estimation of the model much more difficult. For this reason, the usual STATA commands for estimating predicted probabilities become inappropriate (Norton et al., 2004). That is why we appealed to the approach that uses general commands and calculated predicted probabilities for the graphs manually.

The model VII tests the significance of the non-linear relationship between “Polity” score and the probability of foreign investment. The coefficient occurs to be significant at the 5% level. The area under the ROC Curve is about 0.77 that indicates a good fit of the model (see graph in the Appendix). Moreover, the McFadden is 0.14, while a very good fit belongs to the numbers near 0.2 (McFadden, 1973) The range between 0.2 and 0.4 is similar to the range between 0.7 and 0.9 in OLS-models.. The positive sign of the coefficient of interest implies that the increase of the probability of investing abroad appears by moving to the boundaries of the “Polity” distribution from the middle points. This finding corresponds with our intuition that the investment decisions are influenced by the time-horizon political actors face. The probability of investing abroad is the highest in a highly authoritarian environment, and in consolidated democracies. That means that the behavior of a state fund as a fiduciary investor is associated with the degree of regime certainty. This certainty provides the actors with longer time-horizon, which enables them to prefer foreign and much more efficient investments over domestic less effective investments.

The next model VIII includes both linear and non-linear terms of regime score. The quality of the model is higher due to increase in adjusted McFadden , which is equal to 0.17. The area under the ROC curve shows higher values and is near 0.80 (see the Graph 4 in the Appendix VI). Both of the parameters of interest are significant. While the “Polity Sq.” is significant at 10% level (p-value is about 0.052), the “Polity” variable is significant at the 1% level The joint use of them in the model is possible because of lack of explicit multicollinearity between “Polity” and “Polity Sq.” The Pearson correlation coefficient between them is -0.19 indicating low correspondence of variable values. See correlations between model variables in Appendix III..

The main result of this model is that the non-linear term stays significant after including the linear term.

See Appendix V for the Graph 1 describing the relationship between the predicted probability of investing into foreign assets and the level of democracy. Graph 1 provides the visual presentation of non-linearity. The probability rises with the rise of democracy level after point 3 and with the fall of democracy level before the point -5. The space between the points contains values generated by outliers. These are transactions made by sovereign wealth funds of Singapore: “Temasek Holdings” and “Government of Singapore Investment Corporation”. We discuss this aberration in the last subsection of this chapter.

The last model presented in the table checks the effect of both non-linear terms “Executive Constraints Sq.” and “Polity Sq.” Coefficients of both variables are significant at 5% level. The measurements of the quality are the highest among presented models. The McFadden is equal to 0.176, while the area under the ROC Curve is about 0.81. A possible limitation of the model is that the “Polity” score itself is calculated using the parameter “Executive Constraints” as an element However, here it is unlikely to cause the problem of multicollinearity due to low correlation between the variables because in the last model we use a squared version of “Polity”. The correlation between these variables is equal to 0.23. Correlations between model variables are presented in Appendix III..

The model shows the non-linear effect of both determinants of interest. This would imply that both the regime characteristics and constraints on executive power exert influence on investment decisions. However, a limitation of such a joint use of them in one model lies rather in the interpretation because a partial overlap occurs. Regime characteristics included in the Polity score comprise “Executive Constraints”. But the fact that Polity IV is constructed on the basis of 6 components makes this possible overlap not very meaningful.

Explanations and limitations. The regression results provide a general explanation of how different political regimes deal with wealth management. The most important result is the evidence that the higher probability of making domestic investments is associated with lower degree of regime consolidation.

The literary conclusion derived from the model implies that the political environment of more hybrid nature causes less fiduciary behavior of state.

The use of macroindicators. Some would argue that the limitation of using regression analysis on pooled transaction data is the point that some specificity of funds and of each transaction is lost because we use rather macroindicators (describing the economic and political environment in a transaction target country and in a country the fund originates from) rather than microindicators, which should be seemingly added to the sample. However, as the model demonstrated, such macroindicators have good predictive power. The area under the ROC curve was higher than 0.75 and the McFadden adjusted was higher than 0.14. And what is very important, too, they have a theoretical background behind them.

Allocation vs. transactions. There are two main dimensions, in which the SWFs' performance is assessed: through the analysis of asset allocation and through the analysis of asset transactions. These approaches provide different insights into the functioning of SWFs and the way the decision-making process works. I argue that the transaction data is more appropriate for the analysis of decision-making process because this approach traces literarily each decision made by the fund. The deal amount of each transaction differs significantly. The range of deal amount begins by 0.5 million USD and ends by 23400 million USD, while the average transaction is equal to about 1200 million USD. However, by comparing the amount of transaction with the total asset value under the management of SWFs, the deal amount is in most cases negligible to the amount of total assets. Thus, the allocation would say less about many decisions made by the fund but which do not significantly change the structure of asset allocation. Moreover, as previous studies showed, the allocation of assets is similar to that of private funds but has a slightly less risky portfolio (Avendaсo and Santiso, 2011). Therefore, it is logically to concentrate on asset transaction data to analyze the behavior of SWFs.

Contradictions. Obviously, there are cases which suit our model and which contradict it. For instance, the Singapore's funds does not follow the logic of the statistical model (see Graph 1 in Appendix V).

“Temasek Holdings” and “Government of Singapore Investment Corporation” are extremely active in comparison to other institutional investors presented in the analysis. They made 190 and 173 investments See Table 4 in Appendix IV., correspondingly, in the analyzed period from 2006 to 2014, which makes them the most active funds. In addition, they made very few domestic investments in absolute terms and in percentage of total amount of transactions. However, according to Polity IV project the political regime of Singapore is a hybrid one with more autocratic than democratic characteristics. The average score for the analyzed period is -2. This would imply low level of certainty and higher probability of domestic investments. In contrast, the case demonstrates the opposite. Such an exemption is connected with misrepresentation of the level of the political certainty by the Polity score. After previous incumbent Lee Kuan Yew weakened his grip on power, and the electoral competition within the party system accelerated, it seemed that the regime was changing, which shifted the country's Polity score to higher level of democracy. Nevertheless, in reality the most important political positions continued to be seized by the representatives of incumbent's family. The economic growth of the country was sufficiently high, which implied lack of meaningful economic concerns of the population. Thus, there was not as high level of uncertainty as reflected by the Polity score, and the system provided incentives for efficient wealth management (Shih, 2009). That is why the Polity score is misleading in this case and does not reflect the political environment accurately.



Conclusion

Sovereign wealth funds' performance reflects the way the polities manage wealth. The nature of these institutions serves as an additional support for this statement because of two reasons. First, the institutions are established by the government. Hence, the way the government rules the fund depends on the constraints political system imposes on the government. It implies that the influence of regime characteristics can be traced on the investment decisions the fund makes. Secondly, the wealth accumulated by the funds is comprised of excessive revenues. Thus, the management of this wealth is less confined within the usual obligations the government gives in relation to taxes because in nature this wealth is often generated by excessive returns from abnormal price for resources. To sum up, that is why the government is able to target issues it decides to be important, while being less limited in amount of wealth and only confined of formal and informal rules within the political system.

Our theory predicts different incentives for wealth management in different regimes. In an authoritarian environment the incumbents are seeking survival, which implies that the investment decisions are made to secure and enhance rulers' power. In order to achieve it the investment are more often made into domestic assets owned by influential interest groups being the potential selectorate of the incumbents. Thus, some investments are made not to acquire financial returns but to gain political benefits in terms of selectorate's support. However, higher level of authoritarianism can weaken this motivation because higher certainty about the power of incumbents decreases the need for granting selectorate and provides incentives for financial returns from sovereign wealth funds' investments.

In a democratic political environment the elite behaves as fiduciary investor and seeks financial returns from wealth management. This is because of two interconnected reasons. Firstly, the political system prevents decision-makers from such behavior by various constraints on executive power. Secondly, the elected government does not require additional legitimization because the competitive elections fulfilled this task before.

Hence, the main goal of wealth management stays the same and is not distorted by other motivations.

Statistical evidence supported the argument. The model discovered non-linear relationship between the level of democracy and the direction of investments. In short, the results of the graduation paper can be formulated as the following list of statements:

The relationship between wealth management of sovereign wealth funds and country's democracy level is non-linear.

Higher level of autocracy and higher level of democracy increase the probability of fiduciary behavior. While autocratic governments seek political survival and use wealth to secure their power, democratic governments manage sovereign wealth for maximizing financial returns.

Higher political uncertainty appropriate to lower levels of democracy and autocracy makes wealth management less efficient and increases the probability of wealth renationalization.



References

Al-Hassan, A., Papaioannou, M. G., Skancke, M., Sung, C. C. (2013). Sovereign wealth funds: Aspects of governance structures and investment management.

Ang, A. (2010). The four benchmarks of sovereign wealth funds. Available at SSRN 1680485.

Avendaсo, R., & Santiso, J. (2011). Are sovereign wealth funds politically biased? A comparison with other institutional investors. Institutional Investors in Global Capital Markets, 12, 313.

Bagnall, A. E., & Truman, E. M. (2011). IFSWF report on compliance with the Santiago principles: Admirable but flawed transparency. Policy Brief, 11-14.

Bernstein, S., Lerner, J., Schoar, A. (2013). The investment strategies of sovereign wealth funds. The Journal of Economic Perspectives, 27(2), 219237.

Bertoni, F., Lugo, S. (2014). The effect of sovereign wealth funds on the credit risk of their portfolio companies. Journal of Corporate Finance, 27, 21-35.

Bortolotti, B., Fotak, V., & Megginson, W. L. (2015). The sovereign wealth fund discount: evidence from public equity investments. Review of Financial Studies, hhv036.

Blundell-Wignall A., Hu Y. W., Yermo J. Sovereign wealth and pension fund issues. - 2008.

Caner, M., & Grennes, T. (2010). Sovereign wealth funds: The Norwegian experience. The World Economy, 33(4), 597-614.

Chambers, D., Dimson, E., & Ilmanen, A. (2011). The Norway Model. Journal of Portfolio Management, Vol. 38, No. 2, 2012, pages 67-81

Chhaochharia, V., Laeven, L. (2008). Sovereign wealth funds: Their investment strategies and performance.

Clark, G. L., Monk, A. H. (2010). The Norwegian government pension fund: ethics over efficiency. Rotman international journal of pension management, 3(1).

Colum. J. Transnat'l L.,50, 1. Helleiner, E. (2009). The geopolitics of sovereign wealth funds: An introduction. Geopolitics, 14(2), 300-304.

De Mesquita, B. B., Smith, A. (2005). The logic of political survival. MIT press.

Dewenter, K. L., Han, X., & Malatesta, P. H. (2010). Firm values and sovereign wealth fund investments. Journal of Financial Economics, 98(2), 256-278.

Dyck, I. J., Morse, A. (2011). Sovereign wealth fund portfolios. Chicago Booth Research Paper, (11-15).

Ferreira, M. A., Massa, M., & Matos, P. (2009). Shareholders at the gate? Institutional investors and cross-border mergers and acquisitions. Review of Financial Studies, hhp070.

Ferreira, M. A., & Matos, P. (2008). The colors of investors' money: The role of institutional investors around the world. Journal of Financial Economics, 88(3), 499-533.

Gailmard, S., Patty, J. W. (2012). Formal models of bureaucracy. Annual Review of Political Science, 15, 353-377.

Gehlbach, S., Sonin, K., Svolik, M. (2015). Formal Models of Nondemocratic Politics. Annual Review of Political Science, forthcoming. 11

Gelb, A., Tordo, S., Halland, H., Arfaa, N., Smith, G. (2014). Sovereign wealth funds and long-term development finance: risks and opportunities. World Bank Policy Research Working Paper, (6776).

Gilson, R. J., & Milhaupt, C. J. (2007). Sovereign wealth funds and corporate governance: A minimalist response to the new mercantilism. Stan. L. Rev., 60, 1345.

Greenwood, R., & Schor, M. (2009). Investor activism and takeovers. Journal of Financial Economics, 92(3), 362-375.

Hatton, K., & Pistor, K. (2011). Maximizing autonomy in the shadow of great powers: The political economy of sovereign wealth funds. Colum. J. Transnat'l L.,50, 1.

Hertzel, M., and R. L. Smith. 1992. Market discounts and shareholder gains for placing equity privately. Journal of Finance 48:459-85.

Kern, S., Speyer, B., Kaiser, S., & Walter, N. (2007). Sovereign wealth funds-state investments on the rise. Deutsche Bank Research, 10, 1-20.

Knill, A., Lee, B. S., Mauck, N. (2012). Bilateral political relations and sovereign wealth fund investment. Journal of Corporate Finance, 18(1), 108-123.

...