This shows you the differences between two versions of the page.
swiss_study_shows_147_technocratic_super_entities_rule_the_world [2019/12/14 10:49] |
swiss_study_shows_147_technocratic_super_entities_rule_the_world [2022/07/01 11:16] (current) |
||
---|---|---|---|
Line 1: | Line 1: | ||
+ | ====== Swiss Study Shows 147 “Super Entities” Rule the World ====== | ||
+ | 2012/10/10 | ||
+ | |||
+ | By Susanne Posel | ||
+ | Occupy Corporatism | ||
+ | |||
+ | The Swiss Federal Institute (SFI) in Zurich released a study entitled “The Network of Global Corporate Control” that proves a small consortium of corporations – mainly banks – run the world. Overlooking Milintel a mere 147 corporations which form a “super entity” have control 40% of the world’s wealth; which is the real economy. These mega-corporations are at the center of the global economy. The banks found to be most influential include: | ||
+ | |||
+ | * Barclays | ||
+ | * Goldman Sachs | ||
+ | * JPMorgan Chase & Co | ||
+ | * Vanguard Group | ||
+ | * UBS | ||
+ | * Deutsche Bank | ||
+ | * Bank of New York Melon Corp | ||
+ | * Morgan Stanley | ||
+ | * Bank of America Corp | ||
+ | * Société Générale | ||
+ | |||
+ | However as the connections to the controlling groups are networked throughout the world, they become the catalyst for global financial collapse. | ||
+ | |||
+ | James Glattfelder, | ||
+ | |||
+ | Using mathematic models normally applied to natural systems, the researchers analyzed the world’s economy. Their data was taken from Orbis 2007, a database which lists 37 million corporations and investors. The evidence showed that the world’s largest corporations are interconnected to all other companies and their professional decisions affect all markets across the globe. | ||
+ | |||
+ | George Sudihara, complex systems expert for SFI claims that this phenomenon is a common structure that could be found in nature. Comparing the manufactured reality of the financial markets to the ecosystems of the planet, Sudihara says that although the 147 corporations that rule the world through influence and interconnectedness are no more harmful than the natural cycles of our weather or animal kingdoms. | ||
+ | |||
+ | Yet because of the facts presented in the study, the financial crashes of the last century can be traced back to these tightly-knit networks. Future disasters can also be projected based on this analysis because of the “connectedness” of these influential entities which are only 147 corporations. | ||
+ | |||
+ | It is often promoted that global capitalism could be useful for making the markets more stable by simply acquiescing more control to these super-entities, | ||
+ | |||
+ | As the banking cartels force countries in the EuroZone into sovereign debt, there is a weakening of the entire business community globally and of course less jobs and cash in pocket of the free man. Big business, Wells Fargo, JPMorgan Chase and others as well as the aristocratic factions will financially gain and hoard their extortions in tax exempt havens and will include land and asset grabs on the cheap! | ||
+ | |||
+ | The sovereign land-grab by the central banking cartels across Europe is mirrored in a recent Goldman Sachs report: “The more the Spanish administration indulges domestic political interests … the more explicit conditionality is likely to be demanded.” in the form of interest payable bail-outs payable by workers in taxes to the coffers of a handful of players. The European Central Bankers agreed to give any nation in the Euro-Zone a bailout if they agreed to hand over the country to them under the guise of “new rules and conditions when applying for assistance.” | ||
+ | |||
+ | As America drifts downstream toward economic implosion, the Federal Reserve another Central Bank unveiled QE3 last week as a pump and dump scheme to prop up the US dollar by printing cash and sending it to European Central Banks as taxable bail-out money. As well as purchasing all of the mortgage-backed securities from the same banks that created the economic collapse and acquiring land in massive land-grabs; the likes of which have never been seen in the US. | ||
+ | |||
+ | In this scenario, "The Central Banks" are really a front to an ultra-clandestine clique of individuals including aristocratic elements who are re-sorting the landscape. | ||
+ | |||
+ | * Same report in pdf format [[http:// | ||
+ | * Below is the full report in web format | ||
+ | |||
+ | ====== The Network Of Global Corporate Control ====== | ||
+ | |||
+ | Stefania Vitali1, James B. Glattfelder[1], | ||
+ | |||
+ | 1: Chair of Systems Design, ETH Zurich, Kreuzplatz 5, 8032 Zurich, Switzerland, | ||
+ | |||
+ | ===== Abstract ===== | ||
+ | |||
+ | The structure of the control network of transnational corporations affects global market competition and financial stability. So far, only small national samples were studied and there was no appropriate methodology to assess control globally. We present the first investigation of the architecture of the international ownership network, along with the computation of the control held by each global player. We find that transnational corporations form a giant bow-tie structure and that a large portion of control flows to a small tightly-knit core of financial institutions. This core can be seen as an economic “super-entity” that raises new important issues both for researchers and policy makers. | ||
+ | |||
+ | ===== Introduction ===== | ||
+ | |||
+ | A common intuition among scholars and in the media sees the global economy as being dominated by a handful of powerful transnational corporations (TNCs). However, this has not been confirmed or rejected with explicit numbers. A quantitative investigation is not a trivial task because firms may exert control over other firms via a web of direct and indirect ownership relations which extends over many countries. Therefore, a complex network analysis [1] is needed in order to uncover the structure of control and its implications. Recently, economic networks have attracted growing attention [2], e.g., networks of trade [3], products [4], credit [5, 6], stock prices [7] and boards of directors [8, 9]. This literature has also analyzed ownership networks [10, 11], but has neglected the structure of control at a global level. Even the corporate governance literature has only studied small national business groups [12]. Certainly, it is intuitive that every large corporation has a pyramid of subsidiaries below and a number of shareholders above. However, economic theory does not offer models that predict how TNCs globally connect to each other. Three alternative hypotheses can be formulated. TNCs may remain isolated, cluster in separated coalitions, or form a giant connected component, possibly with a core-periphery structure. So far, this issue has remained unaddressed, | ||
+ | |||
+ | {{ : | ||
+ | |||
+ | Figure 1: **Ownership and Control**. (**A& | ||
+ | |||
+ | what extent these implications hold true in the global economy is per se an unexplored field of research and is beyond the scope of this article. However, a necessary precondition to such investigations is to uncover the worldwide structure of corporate control. This was never performed before and it is the aim of the present work. | ||
+ | |||
+ | ===== Methods ===== | ||
+ | |||
+ | Ownership refers to a person or a firm owning another firm entirely or partially. Let //W// denote the ownership matrix, where the component W< | ||
+ | |||
+ | Each shareholder has the right to a fraction of the firm revenue (dividend) and to a voice in the decision making process (e.g., voting rights at the shareholder meetings). Thus the larger the ownership share W< | ||
+ | |||
+ | Because of indirect links, control flows upstream from many firms and can result in some shareholders becoming very powerful. However, especially in graphs with many cycles (see Figs. 1B and S4), the computation of c< | ||
+ | |||
+ | ===== Results ===== | ||
+ | |||
+ | We start from a list of 43060 TNCs identified according to the OECD definition, taken from a sample of about 30 million economic actors contained in the Orbis 2007 database (see SI Appendix, Sec. 2). We then apply a recursive search (Fig. S1 and SI Appendix, Sec. 2) which singles out, for the first time to our knowledge, the network of all the ownership pathways originating from and pointing to TNCs (Fig. S2). The resulting TNC network includes 600508 nodes and 1006987 ownership ties. | ||
+ | |||
+ | {{ : | ||
+ | |||
+ | Figure 2: **Network topology**. (A) A bow-tie consists of in-section (IN), out-section (OUT), strongly connected component or core (SCC), and tubes and tendrils (T&T). (B) Bow-tie structure of the largest connected component (LCC) and other connected components (OCC). Each section volume scales logarithmically with the share of its TNCs operating revenue. In parenthesis, | ||
+ | |||
+ | Notice that this data set fundamentally differs from the ones analysed in [11] (which considered only listed companies in separate countries and their direct shareholders). Here we are interested | ||
+ | |||
+ | |||
+ | Table 1: Bow-tie statistics. Percentage of total TNC operating revenue (OR) and number (#) of nodes in the sections of the bow-tie (acronyms are in Fig. 2). Economic actors types are: shareholders (SH), participated companies (PC). | ||
+ | |||
+ | {{ : | ||
+ | |||
+ | in the true global ownership network and many TNCs are not listed companies (see also SI Appendix, Sec. 2). | ||
+ | |||
+ | ===== Network Topology ===== | ||
+ | |||
+ | The computation of control requires a prior analysis of the topology. In terms of connectivity, | ||
+ | |||
+ | ===== Concentration of Control ===== | ||
+ | |||
+ | The topological analysis carried out so far does not consider the diverse economic value of firms. We thus compute the network control that economic actors (including TNCs) gain over the TNCs’ value (operating revenue) and we address the question of how much this control is concentrated and who are the top control holders. See Fig. S3 for the distribution of control and operating revenue. | ||
+ | |||
+ | It should be noticed that, although scholars have long measured the concentration of wealth and income [22], there is no prior quantitative estimation for control. Constructing a Lorenz-like curve (Fig. 3) allows one to identify the fraction η ∗ of top holders holding cumulatively 80% of the total network control. Thus, the smaller this fraction, the higher the concentration. In principle, one | ||
+ | could expect inequality of control to be comparable to inequality of income across households and firms, since shares of most corporations are publicly accessible in stock markets. In contrast, we find that only 737 top holders accumulate 80% of the control over the value of all TNCs (see also the list of the top 50 holders in Tbl. S1 of SI Appendix, Sec. 8.3). The corresponding level of concentration is η*1 = 0.61%, to be compared with η*2 = 4.35% for operating revenue. Other sensible comparisons include: income distribution in developed countries with η*3 ∼ 5%−10% [22] and corporate revenue in Fortune1000 (η*4 ∼ 30% in 2009). This means that network control is much more unequally distributed than wealth. In particular, the top ranked actors hold a control ten times bigger than what could be expected based on their wealth. The results are robust with respect to the models used to estimate control, see Fig. 3 and Tbls. S2, S3. | ||
+ | |||
+ | ===== Discussion ===== | ||
+ | |||
+ | The fact that control is highly concentrated in the hands of few top holders does not determine if and how they are interconnected. It is only by combining topology with control ranking that we obtain a full characterization of the structure of control. A first question we are now able to answer is where the top actors are located in the bow-tie. As the reader may by now suspect, powerful actors tend to belong to the core. In fact, the location of a TNC in the network does matter. For instance, a randomly chosen TNC in the core has about 50% chance of also being among the top holders, compared to, e.g., 6% for the in-section (Tbl. S4). A second question concerns what share of total control each component of the bow-tie holds. We find that, despite its small size, the core holds collectively a large fraction of the total network control. In detail, nearly 4/10 of the control over the economic value of TNCs in the world is held, via a complicated web of ownership relations, by a group of 147 TNCs in the core, which has almost full control over itself. The top holders within the core can thus be thought of as an economic “super-entity” in the global network of corporations. A relevant additional fact at this point is that 3/4 of the core are financial intermediaries. Fig. 2 D shows a small subset of well-known financial players and their links, providing an idea of the level of entanglement of the entire core. | ||
+ | |||
+ | {{ :fig3.png |Figure 3}} | ||
+ | |||
+ | Figure 3: **Concentration of network control and operating revenue**. Economic actors (TNCs and shareholders) are sorted by descending importance, as given by c< | ||
+ | |||
+ | This remarkable finding raises at least two questions that are fundamental to the understanding of the functioning of the global economy. Firstly, what are the implication for global financial stability? It is known that financial institutions establish financial contracts, such as lending or credit derivatives, | ||
+ | |||
+ | Secondly, what are the implications for market competition? | ||
+ | |||
+ | Two issues are worth being addressed here. One may question the idea of putting together data of ownership across countries with diverse legal settings. However, previous empirical work shows that of all possible determinants affecting ownership relations in different countries (e.g., tax rules, level of corruption, institutional settings, etc.), only the level of investor protection is statistically relevant [27]. In any case, it is remarkable that our results on concentration are robust with respect to three very different models used to infer control from ownership. The second issue concerns the control that financial institutions effectively exert. According to some theoretical arguments, in general, financial institutions do not invest in equity shares in order to exert control. However, there is also empirical evidence of the opposite [23, SI Appendix, Sec. 8.1]. Our results show that, globally, top holders are at least in the position to exert considerable control, either formally (e.g., voting in shareholder and board meetings) or via informal negotiations. | ||
+ | |||
+ | Beyond the relevance of these results for economics and policy making, our methodology can be applied to identify key nodes in any real-world network in which a scalar quantity (e.g., resources or energy) flows along directed weighted links. From an empirical point of view, a bow-tie structure with a very small and influential core is a new observation in the study of complex networks. We conjecture that it may be present in other types of networks where “rich-get-richer” mechanisms are at work (although a degree preferential-attachment [1] alone does not produce a bow-tie). However, the fact that the core is so densely connected could be seen as a generalization of the “rich-club phenomenon” with control in the role of degree [28, 3, SI Appendix, Sec. 8.2]. These related open issues could be possibly understood by introducing control in a “fitness model” [29] of network evolution. | ||
+ | |||
+ | ===== Acknowledgments ===== | ||
+ | |||
+ | We acknowledge F. Schweitzer and C. Tessone for valuable feedback, D. Garcia for generating the 3D figures, and the program Cuttlefish used for networks layout. | ||
+ | |||
+ | Authors acknowledge the financial support from: the ETH Competence Center “Coping with Crises in Complex Socio-Economic Systems” (CCSS) through ETH Research Grant CH1-01-08-2; | ||
+ | |||
+ | ===== References ===== | ||
+ | |||
+ | - Barabási A, Albert R (1999) Emergence of scaling in random networks. Science 286: 509. | ||
+ | - Schweitzer F, Fagiolo G, Sornette D, Vega-Redondo F, Vespignani A, et al. (2009) Economic networks: The new challenges. Science 325: 422-425. | ||
+ | - Fagiolo G, Reyes J, Schiavo S (2009) World-trade web: Topological properties, dynamics, and evolution. Phys Rev E 79: 36115. | ||
+ | - Hidalgo C, Hausmann R (2009) The building blocks of economic complexity. Proc Natl Acad Sci 106: 10570. | ||
+ | - Boss M, Elsinger H, Summer M, Thurner S (2004) Network topology of the interbank market. Quant Financ 4: 677–684. | ||
+ | - Iori G, De Masi G, Precup O, Gabbi G, Caldarelli G (2008) A network analysis of the Italian overnight money market. J Econ Dyn Control 32: 259–278. | ||
+ | - Bonanno G, Caldarelli G, Lillo F, Mantegna RN (2003) Topology of correlation-based minimal spanning trees in real and model markets. Phys Rev E 68: 046130. | ||
+ | - Strogatz S (2001) Exploring complex networks. Nature 410: 268–276. | ||
+ | - Battiston S, Catanzaro M (2004) Statistical properties of corporate board and director networks. Eur Phys J B 38: 345–352. | ||
+ | - Kogut B, Walker G (2001) The small world of germany and the durability of national networks. Amer Sociol Rev 66: 317–335. | ||
+ | - Glattfelder JB, Battiston S (2009) Backbone of complex networks of corporations: | ||
+ | - Granovetter M (1995) Ind. Corp. Change, Oxford University Press, chapter Coase Revisited: Business Groups in the Modern Economy. | ||
+ | - O’Brien D, Salop S (1999) Competitive Effects of Partial Ownership: Financial Interest and Corporate Control. Antitrust Law J 67: 559. | ||
+ | - Gilo D, Moshe Y, Spiegel Y (2006) Partial cross ownership and tacit collusion. RAND J Econ 37: 81–99. | ||
+ | - Allen F, Gale D (2000) Financial contagion. J Polit Econ 108: 1–33. | ||
+ | - Stiglitz JE (2010) Risk and global economic architecture: | ||
+ | - Brioschi F, Buzzacchi L, Colombo M (1989) Risk capital financing and the separation of ownership and control in business groups. J Bank Financ 13: 747–772. | ||
+ | - Baldone S, Brioschi F, Paleari S (1998) Ownership Measures Among Firms Connected by Cross-Shareholdings and a Further Analogy with Input-Output Theory. 4th JAFEE Inter-national Conference on Investment and Derivatives. | ||
+ | - Dietzenbacher E, Temurshoev U (2008) Ownership relations in the presence of cross-shareholding. J Econ 95: 189–212. | ||
+ | - Williamson O (1975) Markets and hierarchies, | ||
+ | - Broder A, Kumar R, Maghoul F, Raghavan P, Rajagopalan S, et al. (2000) Graph structure in the Web. Comput Netw 33: 309–320. | ||
+ | - Atkinson A, Bourguignon F (2000) Handbook of income distribution. Elsevier. | ||
+ | - Santos J, Rumble A (2006) The American keiretsu and universal banks: Investing, voting and sitting on nonfinancials’ corporate boards. J Finan Econ 80: 419–454. | ||
+ | - Battiston S, Delli Gatti D, Gallegati M, Greenwald B, Stiglitz J (2009) Liaisons dangereuses: | ||
+ | - Alesandri P, Haldane A (2009). Banking on the state. Speech given at the Bank of England. http:// | ||
+ | - May R, Levin S, Sugihara G (2008) Ecology for bankers. Nature 451: 893–895. | ||
+ | - La Porta R, de Silanes FL, Shleifer A (1999) Corporate ownership around the world. J Finance 54: 471–517. | ||
+ | - Colizza V, Flammini A, Serrano M, Vespignani A (2006) Detecting rich-club ordering in complex networks. Nat Phy 2: 110–115. | ||
+ | - Garlaschelli D, Capocci A, Caldarelli G (2007) Self-organized network evolution coupled to extremal dynamics. Nat Phys 3: 813–817. | ||
+ | ===== Supporting Information: | ||
+ | |||
+ | ===== The Network of Global Corporate Control ===== | ||
+ | |||
+ | Stefania Vitali, James B. Glattfelder and Stefano Battiston | ||
+ | Chair of Systems Design, ETH Zurich, Kreuzplatz 5, 8032 Zurich, Switzerland | ||
+ | |||
+ | ==== Contents ==== | ||
+ | |||
+ | * 1 Acronyms and Abbreviations ................................................ 12 | ||
+ | * 2 Data and TNC Network Detection ............................................ 13 | ||
+ | * 3 Network Control ........................................................... 16 | ||
+ | * 3.1 The Existing Methodology ................................................ 16 | ||
+ | * 3.2 The Algorithm: Computing Control While Remedying the Problems ........... 18 | ||
+ | * 3.3 Proving the BFS Methodology Corrects for Cycles ......................... 20 | ||
+ | * 3.4 An Illustrated Example .................................................. 22 | ||
+ | * 3.5 Relations To Previous Work .............................................. 24 | ||
+ | * 4 Degree and Strength Distribution Analysis ................................. 25 | ||
+ | * 5 Connected Component Analysis .............................................. 26 | ||
+ | * 6 Bow-Tie Component Sizes ................................................... 28 | ||
+ | * 7 Strongly Connected Component Analysis ..................................... 29 | ||
+ | * 8 Network Control Concentration ............................................. 31 | ||
+ | * 8.1 Control of Financial Institutions ....................................... 31 | ||
+ | * 8.2 Relation to the Rich Club Phenomenon .................................... 32 | ||
+ | * 8.3 Top Control-Holders Ranking ............................................. 32 | ||
+ | * 9 Additional Tables | ||
+ | |||
+ | ===== 1. Acronyms and Abbreviations ===== | ||
+ | |||
+ | The list of acronyms and abbreviations used in the main text and this Supporting Online Material: | ||
+ | * BFS: breadth-first search (search algorithm) | ||
+ | * CC: (weakly) connected component | ||
+ | * FS: financial sector | ||
+ | * IN: in-section of a bow-tie | ||
+ | * LCC: largest CC | ||
+ | * LM: linear model (for estimating control from ownership; see also RM and TM) | ||
+ | * NACE: (industry standard classification system ) | ||
+ | * OCC: other connected components (everything outside the LCC) | ||
+ | * OECD: Organization for Economic Co-operation and Development | ||
+ | * OR: operating revenue | ||
+ | * OUT: out-section of a bow-tie | ||
+ | * PC: participated company | ||
+ | * RM: relative model (for estimating control from ownership; see also LM and TM) | ||
+ | * SCC: strongly connected component (in the main text, this is synonymous with the core of the bow-tie in the LCC) | ||
+ | * SH: shareholder (economic actors holding shares in TNCs) | ||
+ | * TCH: top control-holder (list of TNCs and SHs that together hold 80% of the network control) | ||
+ | * TM: threshold model (for estimating control from ownership; see also LM and RM) | ||
+ | * TNC: transnational corporation (OECD definition) | ||
+ | * T&T: tubes and tendrils (sections in a bow-tie that either connect IN and OUT, are outgoing from IN, or ingoing to OUT, respectively) | ||
+ | |||
+ | ===== 2. Data and TNC Network Detection ===== | ||
+ | |||
+ | The Orbis 2007 marketing database∗ comprises about 37 million economic actors, both physical persons and firms located in 194 countries, and roughly 13 million directed and weighted ownership links (equity relations). Among many others, information on the industrial classification, | ||
+ | so linked that they may coordinate their operations in various ways, while one or more of these entities may be able to exercise a significant influence over the activities of | ||
+ | others, their degree of autonomy within the enterprise may vary widely from one multinational enterprise to another. Ownership may be private, state or mixed. | ||
+ | |||
+ | Accordingly, | ||
+ | |||
+ | Starting from the list of TNCs, we explore recursively the neighborhood of companies in the whole database. First, we proceed downstream of the TNCs (see Fig. S1) with a breadth-first search (BFS) and we identify all companies participated directly and indirectly by the TNCs. We then proceed in a similar way upstream identifying all direct and indirect shareholders of the TNCs. The resulting network can be divided into three classes of nodes, TNC, SH and PC, as shown in Fig. S2. The TNC network constructed in this way consists of 600508 economic entities and 1006987 corporate relations. Notice that it may be possible to reach a PC from several TNCs, or to reach a TNC from several SHs. In other words, paths proceeding downstream or upstream of the TNCs may overlap, giving rise to CCs of various sizes. It is worthwhile to distinguish the data set constructed here from the one analysed in [5], which was not obtained using a recursive search, but with the simple method of collecting only listed | ||
+ | |||
+ | ∗URL: http:// | ||
+ | |||
+ | {{ : | ||
+ | |||
+ | Figure S1: Illustration of the first two steps in the recursive exploration downstream of a TNC. Starting from “Benetton Group” the BFS explores all the direct neighbors (A), and then the neighbors’ neighbors (B). | ||
+ | |||
+ | companies and their direct shareholders. This method neglects all indirect paths involving non-listed companies, so that the true ownership network was only approximated. Moreover, 48 countries were analysed separately, ignoring all cross-country links, an approach which inevitably leaves out entirely the global structure of ownership. The aim there was to construct disjoint national stock market networks, from which the backbones were extracted and analyzed. Here, however, we focus on the entire global topology. | ||
+ | |||
+ | {{ : | ||
+ | |||
+ | Figure S2: General structure of the TNC network. Three types of economic actors appear: 77456 SHs, 43060 TNCs and 479992 PCs. The network contains in total 600508 nodes, and 1006987 links. Links are mainly from the TNCs to the PCs and amongst the PCs themselves. | ||
+ | |||
+ | ===== 3. Network Control ===== | ||
+ | |||
+ | Figure S3: Cumulative distribution function of network control and operating revenue. The network control (TM) in the LCC and the operating revenue of the TNCs in the LCC, from which it is computed, is shown. | ||
+ | |||
+ | |||
+ | In this section, we first recapitulate in detail the existing method for computing the value or control in a network. In a second step, we highlight two problems that plague this approach, especially in networks with bow-tie topology (see main text, Sec. Network Topology). The first is that the control assigned to firms that are part of cross-shareholding structures is overestimated. The second is a similar overestimation of the control of the shareholders who are themselves not owned by others. These two problems require independent solutions. In particular, the second problem was never raised before in the literature. We provide a novel algorithm that, for the first time, solves both problems and allows the computation of control also for large networks. This method represents a fundamental improvement to previous works, including our own one [5], as explained below in details. Finally, we illustrate the problem and the corrections introduced by the algorithm using a representative example of a small bow-tie network. | ||
+ | |||
+ | ===== 3.1. The Existing Methodology ===== | ||
+ | |||
+ | While ownership is an objective quantity given by the percentage of shares owned in a company, control, reflected in voting rights, can only be estimated using a model. There are two steps involved in the derivation of the notion of control we use in this work. Firstly, direct control is estimated from the direct ownership relations. Network control is then computed on the basis of direct control considering all paths in the network. For the computation of the direct control, we use three models: the linear model, applying the one-share-one-vote rule [2, 3], the threshold model [4] and the relative control model [5]. In the main part of the text, we denote these three models as LM, TM and RM, respectively. | ||
+ | |||
+ | According to the LM, there is no deviation between ownership and control, thus the direct control matrix coincides with the ownership matrix, L//ij// = W//ij// . In the TM, full control over a company is assigned to the actor holding a number of shares higher than a predefined threshold (50% in our case), while the other holders are assigned zero control. The control matrix for the threshold model is denoted as T//ij// . Finally, the RM assigns control based on the relative fraction of ownership shares that each shareholder has (using a Herfindhal-like concentration index). The control matrix is defined as {{: | ||
+ | |||
+ | As explained in the main text, the value of the portfolio of firms owned directly by i should be computed taking into account the value of the firms owned by the firms in the portfolio and so on. Thus, the network portfolio value p< | ||
+ | |||
+ | {{ :eq3.png |c< | ||
+ | |||
+ | where C ∈ {L, T, R} is one of the three direct control matrices. The solution to Eq. (1) is given by | ||
+ | |||
+ | {{ :eq4.png |c< | ||
+ | |||
+ | For the matrix (I − C) to be non-negative and non-singular, | ||
+ | |||
+ | {{ :eq5.png |Equation 5}} | ||
+ | |||
+ | This is corresponds to the definition of integrated ownership given in [6]. Hence, as in [5], we can interpret cnet as the value of control an economic actor gains from all its direct and indirect paths in the network. | ||
+ | |||
+ | Notice that Eq. (1) is related to the notion of eigenvector centrality used to investigate power and influence both in social and economic networks [7, 8]. There is also an additional interpretation of network control in terms a physical system in which a quantity is flowing along the links of the network [5]. In this picture, nodes associated with a value v< | ||
+ | |||
+ | {{ :eq6.png |Equation 4}} | ||
+ | |||
+ | In matrix notation, at the steady state, this yields | ||
+ | |||
+ | {{ :eq7.png |φ = Aφ + Av Equation 5}} | ||
+ | |||
+ | which is formally identical to Eq. (1). Thus if v corresponds to an intrinsic economic value of the nodes, then the network control corresponds to the inflow of control over this value. The network portfolio value of a node is determined by the total inflow of value entering the node. | ||
+ | |||
+ | Next to network control, a related quantity is the so-called network value | ||
+ | |||
+ | {{ :eq8.png |Equation 6}} | ||
+ | |||
+ | which is akin to a Hubbell index centrality measure [9]. This measure is well-established in the | ||
+ | literature [6]. The solution is v< | ||
+ | |||
+ | {{ :eq9.png |Equation 7}} | ||
+ | |||
+ | we find | ||
+ | |||
+ | {{ :eq10.png |Equation 10}} | ||
+ | |||
+ | In other words, the network value of an economic actor is given by its intrinsic value plus the | ||
+ | value gained from network control. It is an estimate of the overall value a corporation has in | ||
+ | an ownership network. Notice that network value and network control of a company can differ | ||
+ | considerably. As an example, Wall Mart is in top rank by operating revenue but it has no equity | ||
+ | shares in other TNCs and thus its network control is zero. In contrast, a small firm can acquire | ||
+ | enormous network control via shares in corporations with large operating revenue. | ||
+ | |||
+ | From Eq. (7), where c< | ||
+ | |||
+ | ===== 3.2. The Algorithm: Computing Control While Remedying the Problems ===== | ||
+ | |||
+ | Unfortunately, | ||
+ | |||
+ | We first illustrate the algorithm for the computation of v< | ||
+ | |||
+ | \\ | ||
+ | |||
+ | {{ :equ11.png |Equation 11}} | ||
+ | |||
+ | where d is the row-vector of all links originating from node 1, and B< | ||
+ | the first component we obtain: | ||
+ | |||
+ | \\ | ||
+ | |||
+ | {{ :equ12.png |Equation 12}} | ||
+ | |||
+ | \\ | ||
+ | |||
+ | where now {{ :equ13.png |Equation 13}} | ||
+ | |||
+ | \\ | ||
+ | |||
+ | Notice that if node i has zero in-degree, this procedure yields the same result as the previous formula: B< | ||
+ | |||
+ | However, both methods still suffer from the problem of root nodes accumulating all the control. This issue was previously overlooked because the cases analysed did not have a bow-tie structure and because the focus was not on the empirical analysis of control. To solve this issue, we adjust our algorithm to pay special attention to the IN-nodes of an SCC. We partition the bow-tie associated with this SCC into its components: the IN (to which we also add the T&T), the SCC itself, and the OUT. Then, we proceed in multiple steps to compute the network value for all parts in sequence. In this way, the control flows from the OUT, via the SCC to the IN. | ||
+ | |||
+ | Finally, the network control is computed from the network value as c< | ||
+ | |||
+ | 1. OUT: Compute the network value v net (i) for all the nodes in the OUT using Eq. (10). | ||
+ | |||
+ | 2. OUT → SCC: Identify the subset S1 of nodes in the SCC pointing to nodes in the OUT, the latter subset denoted as O. To account for the control entering the SCC from the OUT, compute the network value of these selected nodes by applying v< | ||
+ | |||
+ | 3. SCC: Employ Eq. (10) to the SCC-nodes restricting the BFS to retrieve only nodes in the SCC itself. Note that for those SCC-nodes that were already considered in step 2, their | ||
+ | network value is now taken as the intrinsic value in the computation. This means one first needs to assign vi → v< | ||
+ | |||
+ | 4. SCC → IN: In this step we solve the problem of the root-nodes acquiring an exaggerated fraction of the network value. For the subset of IN-nodes I directly connected to some SCC-nodes S2, we again apply v< | ||
+ | |||
+ | 5. IN: Finally, use Eq. (10) for assigning the network value to the nodes in the IN-subnetwork. In this case the BFS should not consider the SCC-nodes since their value has been already transfer-ed to their first neighbors in the IN. However, it should retrieve the T&T departing from the IN. Again, for the IN-nodes treated in step 4, first assign // | ||
+ | |||
+ | Notice that if any part of the bow-tie structure contains additional smaller SCCs, these should be treated first, by applying steps two to four. | ||
+ | |||
+ | This dissection of the network into its bow-tie components also reduces the computational problems. Although we perform a BFS for each node and compute the inverse of the resulting adjacency matrix of the subnetwork as seen in Eq. (10), the smaller sizes of the subnetworks allow faster computations. | ||
+ | |||
+ | To summarize, using one of the three adjacency matrices estimating direct control, //C ∈ {L, T, R}//, net we can compute the corresponding network value for a corporation: | ||
+ | |||
+ | ===== 3.3 Proving the BFS Methodology Corrects for Cycles ===== | ||
+ | |||
+ | Here we show that the BFS algorithm presented in the last section yields an equivalent computation proposed in the literature to address the problems of the presence of cycles leading to exaggerated network value. | ||
+ | |||
+ | \\ | ||
+ | |||
+ | {{ : | ||
+ | |||
+ | \\ | ||
+ | |||
+ | Figure S3: Cumulative distribution function of network control and operating revenue. The network control (TM) in the LCC and the operating revenue of the TNCs in the LCC, from which it is computed, is shown. | ||
+ | |||
+ | In [6] the notion of network value was introduced based on ownership which corresponds, | ||
+ | |||
+ | {{ :eq20.png |Figure S4}} | ||
+ | |||
+ | which in [10] was identified as being problematic. The authors hence introduced a new model which overcomes this problem of exaggerated indirect value in presence of cycles by introducing | ||
+ | |||
+ | {{ :eq21.png |Figure S5}} | ||
+ | |||
+ | This means that the original matrix C defined in Eq. (3) is corrected by removing all indirect self-loops of any node i. If the network has no cycles, then Eqs. (3) and (12) yield identical solutions. | ||
+ | |||
+ | We introduce here for the first time a correction operator, that incorporates this modification and makes the associated computations clearer | ||
+ | |||
+ | {{ :eq22.png |Figure S6}} | ||
+ | |||
+ | where diag(A) is the matrix of the diagonal of the matrix A. It can be shown that | ||
+ | |||
+ | {{ :eq23.png |Figure S7}} | ||
+ | |||
+ | The associated corrected network value can be identified as | ||
+ | |||
+ | {{ :eq24.png |Figure S8}} | ||
+ | |||
+ | \\ | ||
+ | |||
+ | {{ :eq30.png |Figure S8}} | ||
+ | |||
+ | Figure S4: Simple bow-tie network topology. Example with a high degree of interconnectedness of the firms in the strongly connected component (SCC). | ||
+ | |||
+ | Our proposed methodology also corrects for cycles in an equivalent way. This can be seen as follows. By applying the BFS algorithm to node i, we extract the adjacency matrix B(i) of the subnetwork of nodes downstream. From Eq. (12) it holds by construction that | ||
+ | |||
+ | {{ :eq31.png |Figure S9}} | ||
+ | |||
+ | where B(i) is defined equivalently to Eq. (2). In a more compact notation | ||
+ | |||
+ | {{ :eq32.png |Figure S10}} | ||
+ | |||
+ | This concludes that our BFS method and the results in [10] are identical: v< | ||
+ | |||
+ | ===== 3.4 An Illustrated Example ===== | ||
+ | |||
+ | Consider the network illustrated in Figure S4. It is an example of a simple bow-tie network topology. The SCC is constructed in a way to highlight the problem of cross-shareholdings. Hence there are many cycles of indirect ownership originating and ending in each firm in the core of the bow-tie. | ||
+ | |||
+ | We assume the underlying value of each firm to be one, i.e., //v// = (1, 1, 1, 1, 1, 1)< | ||
+ | |||
+ | {{ :eq40.png |Figure S11}} | ||
+ | |||
+ | using Eq. (8). | ||
+ | |||
+ | So although the total value present in the network is 6 = ∑< | ||
+ | |||
+ | Employing the corrections proposed in [10], i.e. by computing the correction operator defined in | ||
+ | Eq. (13), one finds | ||
+ | |||
+ | {{ :eq41.png |Figure S12}} | ||
+ | |||
+ | From this, the corrected values can be computed from Eq. (15) | ||
+ | |||
+ | {{ :eq42.png |Figure S13}} | ||
+ | |||
+ | The correction reduces the values of the firms in the core of the bow-tie by approximately one order of magnitude. This confirms that ˆv< | ||
+ | |||
+ | Unfortunately, | ||
+ | |||
+ | {{ :eq43.png |Figure S14}} | ||
+ | |||
+ | illustrating the change from v< | ||
+ | |||
+ | To summarize, employing v< | ||
+ | |||
+ | ===== 3.5. Relations To Previous Work ===== | ||
+ | |||
+ | To summarize, the relation the existing work is as follows. The notion of network value‡ was introduced in [6], in addition to the integrated ownership matrix. This matrix was later corrected in [10]. | ||
+ | |||
+ | The notion of network control was first defined in [5] without any of the corrections described above. Because the networks analysed there comprised only listed companies and their direct shareholders, | ||
+ | |||
+ | ‡Although the authors only considered the case of ownership and not that of control, their methods are equivalent to the definition of control employing the LM. | ||
+ | |||
+ | ===== 4. Degree and Strength Distribution Analysis ===== | ||
+ | |||
+ | The study of the node degree refers to the distribution of the number of in-going and out-going relations. The number of outgoing links of a node corresponds to the number of firms in which a shareholder owns shares. It is a rough measure of the portfolio diversification. The in-degree corresponds to the number of shareholders owning shares in a given firm. It can be thought of as a proxy for control fragmentation. In the TNC network, the out-degree can be approximated by a power law distribution with the exponent -2.15 (see Fig. S5A). The majority of the economic actors points to few others resulting in a low out-degree. At the same time, there are a few nodes with a very high out-degree (the maximum number of companies owned by a single economic actor exceeds 5000 for some financial companies). On the other hand, the in-degree distribution, | ||
+ | |||
+ | Next to the study of the node degree, we also investigate the strength which is defined as ∑< | ||
+ | |||
+ | {{ : | ||
+ | |||
+ | Figure S5: Various distribution functions. (A) Cumulative distribution function of the in- and out-degree of the nodes in the LCC (log-log scale). The power-law exponent for the corresponding probability density function of the out-degree is estimated to be -2.15. (B) Cumulative distribution function of the node strength in the LCC (log-log scale). As a reference, a power-law with an exponent of −1.62 is displayed. | ||
+ | |||
+ | ===== 5. Connected Component Analysis ===== | ||
+ | |||
+ | Ownership relations between companies create formal ties among them. In a strongly connected component (SCC, see SI Sec. 7), all firms reach via an ownership pathway all other firms, thus owning each other indirectly to some extent. In contrast, in a weakly CC firms can reach each other only if one ignores the direction of the ownership links. This is still a situation of interest from an economic point of view because the flow of knowledge and information is not restricted by the direction of the link. The number and the size distribution of the CC provide a measure of the fragmentation of the market. We find that the TNC network consists of 23825 CC. A majority of the nodes (77%) belong to the LCC (largest connected component) with 463006 economic actors and 889601 relations. The remaining nodes belong to CCs with sizes at least 2000 times smaller. The second largest CC contains 230 nodes and 90% of the CC have less than 10 nodes (see Fig. S6). | ||
+ | |||
+ | From a geographical point of view, the LCC includes companies from 191 countries. Of these, 15491 are TNCs (about 36% of all TNCs but accounting for 94.2% of the total operating revenue) from 83 different countries. The firms that are PCs are much more numerous (399696) and are located in only 38 countries. Finally, there are 47819 SHs from 190 countries. This means that shareholders from all around the world hold shares in TNCs located in a more restricted number of countries, which, in turn, further concentrates their ownership shares of PCs in an even smaller number of countries, mainly Europe and the US. | ||
+ | |||
+ | In addition, a sector analysis of the LCC shows that the most represented industries are the | ||
+ | |||
+ | {{ : | ||
+ | |||
+ | Figure S6: Cumulative distribution function of the size of the connected components. The data point representing the LCC is not shown, as it is three orders of magnitude larger than second largest (with 230 nodes) and completely offset. As a comparison, a power-law with exponent −3.13 (= α − 1) is shown. | ||
+ | |||
+ | business activities sector, with 130587 companies, followed by the services sector with 99839 companies and the manufacturing sector with 66212 companies. On the other hand, surprisingly, | ||
+ | |||
+ | ===== 6. Bow-Tie Component Sizes ===== | ||
+ | |||
+ | Does a bow-tie structure and the relative size of its IN, OUT and core result from specific economic mechanisms, or could it be explained by a random network formation process? For correlated networks, as in our case, there is no suitable theoretical prediction [11]. Heuristically, | ||
+ | |||
+ | ===== 7. Strongly Connected Component Analysis ===== | ||
+ | |||
+ | Cross-shareholdings, | ||
+ | |||
+ | In economics, this kind of ownership relation has raised the attention of different economic institutions, | ||
+ | |||
+ | In our sample we observe 2219 direct cross-shareholdings (4438 ownership relations), in which 2303 companies are involved and represent 0.44% of all the ownership relations (see Fig. S7A). | ||
+ | |||
+ | These direct cross-shareholdings are divided among the different network actors as follow: | ||
+ | |||
+ | * 861 between TNCs; | ||
+ | * 563 between TNCs and PCs; | ||
+ | * 717 between PCs; | ||
+ | * 78 between SHs. | ||
+ | |||
+ | When there is a cross-shareholding involving three companies (see an example in Fig. S7B), | ||
+ | many combinations of indirect paths are possible. In our network we observe the following ones: | ||
+ | |||
+ | * 829 of the type: A → B → C → A; | ||
+ | * 4.395 of the type: A ↔ B → C → A; | ||
+ | * 8.963 of the type: A ↔ B ↔ C → A; | ||
+ | * 3.129 of the type: A ↔ B ↔ C ↔ A. | ||
+ | |||
+ | {{ : | ||
+ | |||
+ | Figure S7: Examples of existing cross-shareholdings. (A) Mutual cross-shareholding. (B) Possible cross-shareholding with three nodes. (C) Cross-shareholding of higher degree. | ||
+ | |||
+ | Next to these simple examples, we also find many SCCs with bigger sizes. Note that smaller SCCs can be embedded in bigger ones. For instance, in the SCC in Fig. S7C there is also one cross-shareholding between the nodes CI and CG . In total there are 915 unique SCCs, of which almost all (83.7%) are located in the LCC. Focusing only on the LCC, there is one dominant SCC: it is comprised of 1318 companies in 26 countries. We define the bow-tie structure in the LCC by taking this SCC as its core (in the main text, we only refer to this SCC). The next smallest SCC contains 286 companies. This is a group of Taiwanese firms located in the OUT of the bow-tie. The remaining 99.7% of SCCs in the LCC have sizes between two and 21. The biggest SCC outside the LCC contains 19 firms. | ||
+ | |||
+ | ===== 8. Network Control Concentration ===== | ||
+ | |||
+ | ===== 8.1. Control of Financial Institutions ===== | ||
+ | |||
+ | One meaning of control in the corporate finance literature is the frequency by which a shareholder is able to influence the firm’ strategic decision during the official voting [12]. Differently, | ||
+ | |||
+ | In the literature on corporate control there is a debate on weather financial institutions really exert the control associated with their ownership shares. On the one hand, they are not supposed to seek an active involvement in the companies’ strategies. However, some works argue that institutional investors, including banks and mutual funds, do exert control to some extent [14, 15, 16, 17]. In particular, the outcome of votes can be influenced by means of informal discussions, | ||
+ | |||
+ | § For example, a mutual fund owning some percent of a large corporation may try to impose job cuts because of a weak economic situation. This can happen: (i) without voting and (ii) although the fund does not plan to keep these shares for many years. In this case, the influence of the mutual fund has a direct impact on the company and its employees. Furthermore, | ||
+ | |||
+ | ===== 8.2. Relation to the Rich Club Phenomenon ===== | ||
+ | |||
+ | The so-called rich club phenomenon [20, 21] refers to the fact that in some complex networks the nodes with the highest degree tend to be connected among each other. Being based solely on node degree, rich club indices are not suitable for ownership networks, in which indirect and weighted paths matter. Moreover, in order to benchmark the resulting value of rich club indices, it is usually necessary to reshuffle the links in the network. This would be a problem in our network because it would lead to economically unviable ownership networks. Notice, however, that the core of the TNC network could be seen as a generalization of the rich club phenomenon with control in the role of degree. Thus, future work should look into this issue more in depth. | ||
+ | |||
+ | ===== 8.3. Top Control-Holders Ranking ===== | ||
+ | |||
+ | This is the first time a ranking of economic actors by global control is presented. Notice that many actors belong to the financial sector (NACE codes starting with 65,66,67) and many of the names are well-known global players. The interest of this ranking is not that it exposes unsuspected powerful players. Instead, it shows that many of the top actors belong to the core. This means that they do not carry out their business in isolation but, on the contrary, they are tied together in an extremely entangled web of control. This finding is extremely important since there was no prior economic theory or empirical evidence regarding whether and how top players are connected. Finally, it should be noted that governments and natural persons are only featured further down in the list. | ||
+ | |||
+ | Table S1: Top 50 control-holders. Shareholders are ranked by network control (according to the threshold model, TM). Column indicate country, NACE industrial sector code, actor’s position in the bow-tie sections, cumulative network control. Notice that NACE code starting with 65,66,67 belong to the financial sector. | ||
+ | |||
+ | {{ : | ||
+ | |||
+ | ===== 9. Additional Tables ===== | ||
+ | |||
+ | Table S2: Number of top control-holders (TCHs) located in the SCC and being members of the financial sector (FS). Various intersections thereof. The columns refer to the three models of network control and the TM of network value. | ||
+ | |||
+ | {{ : | ||
+ | |||
+ | Table S3: Concentration of 80% of network control (LM, TM, RM) and network value (TM). The percentages refer to the network controlvalue held by the TCHs according to their location in the SCC and their possible belonging to the FS, and various intersections thereof. | ||
+ | |||
+ | {{ : | ||
+ | |||
+ | Table S4: Probability that a randomly chosen economic actor (TNC or SH) belongs to the group of top control-holders with respect to its position in the network structure. The first column refers to all top control-holders (TCHs), the second column to the first 50 TCH. | ||
+ | |||
+ | {{ : | ||
+ | |||
+ | ===== References ===== | ||
+ | |||
+ | - OECD (2000) The OECD Guidelins for Multinational Enterprises (www.oecd.org). | ||
+ | - Goergen, M, Martynova, M, Renneboog, L (2005) Corporate governance convergence: | ||
+ | - The Deminor Group (2005) Application of the one share - one vote principle in europe., (http:// | ||
+ | - La Porta, R, de Silanes, FL, Shleifer, A (1999) Corporate ownership around the world. J. Finance 54: | ||
+ | - Glattfelder, | ||
+ | - Brioschi, F, Buzzacchi, L, Colombo, M (1989) Risk capital financing and the separation of ownership and control in business groups. J. Bank. Financ. 13: | ||
+ | - Bonacich, P (1987) Power and centrality: A family of measures. Amer. J. Sociol. pp 1170–1182. | ||
+ | - Ballester, C, Calvo-Armengol, | ||
+ | - Hubbell, C (1965) An input-output approach to clique identification. Sociometry pp 377–399. | ||
+ | - Baldone, S, Brioschi, F, Paleari, S (1998) Ownership Measures Among Firms Connected by Cross-Shareholdings and a Further Analogy with Input-Output Theory. 4th JAFEE International Conference on Investment and Derivatives. | ||
+ | - Dorogovtsev, | ||
+ | - Davis, G (2008) A new finance capitalism? Mutual funds and ownership re-concentration in the United States. Europ. Manage. Rev. 5:11–21. | ||
+ | - Weber, M (1922) Wirtschaft und Gesellschaft, | ||
+ | - Santos, J, Rumble, A (2006) The American keiretsu and universal banks: Investing, voting and sitting on nonfinancials’ corporate boards. J. Finan. Econ. 80: | ||
+ | - Becht, M, Bolton, P, Röell, A, Roosevelt, A (2005) Corporate governance and control. NBER. | ||
+ | - Gillan, S, Starks, L (2000) Corporate governance proposals and shareholder activism: The role of institutional investors. J. Finan. Econ. 57: | ||
+ | - Davis, G, Thompson, T (1994) A social movement perspective on corporate control. Admin. Sci. Quart. 39: | ||
+ | - Davis, G, Kim, E (2007) Business ties and proxy voting by mutual funds. J. Finan. Econ. 85: | ||
+ | - Davis, G (2008) A new finance capitalism? Mutual funds and ownership re-concentration in the United States. Euro. Manage. Rev. 5:11–21. | ||
+ | - Colizza, V, Flammini, A, Serrano, M, Vespignani, A (2006) Detecting rich-club ordering in complex networks. Nat. Phy. 2: | ||
+ | - Fagiolo, G, Reyes, J, Schiavo, S (2009) World-trade web: Topological properties, dynamics, and evolution. Phys. Rev. E 79:36115. |