NBER WORKING PAPER SERIES CORPORATE CONTROL AROUND THE WORLD Gur Aminadav Elias Papaioannou Working Paper 23010 http://www.nber.org/papers/w23010 NATIONAL BUREAU OF ECONOMIC RESEARCH 1050 Massachusetts Avenue Cambridge, MA 02138 December 2016 We thank Julian Franks, Rafael La Porta, Florencio Lopez-de-Silanes, and Andrei Shleifer for useful suggestions and comments. All errors are our sole responsibility. The views expressed herein are those of the authors and do not necessarily reflect the views of the National Bureau of Economic Research. NBER working papers are circulated for discussion and comment purposes. They have not been peer-reviewed or been subject to the review by the NBER Board of Directors that accompanies official NBER publications. © 2016 by Gur Aminadav and Elias Papaioannou. All rights reserved. Short sections of text, not to exceed two paragraphs, may be quoted without explicit permission provided that full credit, including © notice, is given to the source. Corporate Control around the World Gur Aminadav and Elias Papaioannou NBER Working Paper No. 23010 December 2016 JEL No. F30,G3,G34,G38,K11,K12,K31 ABSTRACT We provide an autopsy of the patterns of corporate control and ownership concentration in a dataset covering more than 40,000 listed firms from 127 countries over 2004 2012. Employing a plethora of original and secondary sources, big data techniques, and applying the Shapley-Shubik algorithm to quantify shareholder’s voting power we trace ultimate controlling shareholders from the complex, pyramidal, and often obscure corporate structures. First, we show that there are large differences in the type of corporate control (widely held firms with and without significant equity blocks, firms controlled by families, governments, and other public-private firms) across and within continents. Corporate structures appear persistent as the recent global financial crisis did not affect them much. Second, we examine the role of legal traditions on corporate control. There are economically large differences on corporate structure across legal families, with the share of controlled (widely-held) firms being the highest (lowest) in French civil-law countries, followed by German and then Scandinavian civil law countries. State ownership and control by individual/families via complex corporate structures is pervasive in civil-law countries. And while equity blocks are commonplace across widely-held firms all around the world and across all legal families, the share of widely-held firms with large blocks is the highest in French civil- law countries. Moreover, ownership concentration is considerably higher in French civil-law (and to a lesser extent in German civil-law) countries as compared to common-law countries. These patterns apply to very large, big, medium-sized and small listed firms and are not driven by regional differences, the level of economic development, or industrial structure, suggesting that legal origin has sizable long-lasting consequences on corporate structure. Third, as legal origin may affect corporate control via multiple channels, we examine the role of some likely mechanisms. We find that shareholder protection rights against self-dealing activities of insiders correlate strongly with corporate control and ownership concentration. Legal formalism and creditor’s rights do not seem to play a role. Yet, the "reduced-form" link between legal origin and corporate control (and ownership concentration) is also driven by entry and labour market regulation. Gur Aminadav Regent's Park NW1 4SA, London United Kingdom [email protected] Elias Papaioannou London Business School Regent's Park Sussex Place London NW1 4SA United Kingdom and NBER [email protected] An online appendix is available at http://www.nber.org/data-appendix/w23010 1Introduction Understanding the driving forces and consequences of the various types of corporate control are core in- quiries of corporate finance (La Porta, Lopez-de-Silanes, and Shleifer, 1999; Tirole, 2010). While most theory distinguishes between widely—held corporations, where ownership is dispersed and no single share- holder dominates corporate decisions, and controlled firms where a dominant-influential shareholder or family exerts control (Shleifer and Vishny, 1997), the structures of corporate ownership and control are more complex (Laeven and Levine, 2008).1 Pyramids that allow shareholders to influence corporate deci- sionsinexcessoftheircash-flow rights (Dyck and Zingales, 2004) and cross-holdings of equity in business groups, such as the keiretsu in Japan and the Korean chaebols,arepervasive.Atthesametime,corporate control is typically hidden behind holding companies incorporated in financial off-shore centers (Zucman, 2015). A binary distinction between controlled and widely-held firms may be too coarse, as in many widely-held corporations some individuals/families or institutional investors hold equity blocks and may thus have considerable power (Edmans and Holderness, 2016). And often there are shareholders who form coalitions with smaller investors in an effort to exert corporate control (Bennedsen and Wolfenzon, 2000). Following the influential contribution of La Porta, Lopez-de-Silanes, and Shleifer (1999), a volumi- nous literature in corporate finance has tried to document the patterns of ownership concentration and control across countries or in a single country over time. The literature (discussed below) then searched for the main correlates of corporate ownership and examined the role of investor protection and legal quality (see La Porta et al. (2006)). However, there are many open issues. First, stands measurement. It is inherently hard to identify ultimate shareholders from the complex network of equity holdings and to assess their role in corporate affairs. For this reason previous research has either focused on a relatively small number of (typically large) firms and countries (e.g., La Porta et al. 1999; Claessens, Djankov, and Lang, 2000; Faccio and Lang, 2002), or has used larger firm samples in a single economy (e.g., Franks, Mayer, and Rossi, 2009, on British corporations) or a handful of countries (Franks, Mayer, Volpin, and Wagner, 2012). Second, is heterogeneity. The size distribution of firms is highly skewed; even when one examines typically large publicly traded firms, differences in size are enormous (see Gabaix (1999)). And theory and case-study evidence point out that corporate control patterns and their determinants may differ system- atically across very large, big, medium, and relatively small listed corporations (Holderness (2016a,b)). A third issue regards mechanisms. The literature is far from reaching a consensus on the role of investor protection rights, courts, product and labour market regulation on corporate control and ownership concentration. Isolating the role of each feature is challenging,asitisalwayshardquantifyingthese institutional features, the sample of countries is limited, and because legal aspect features correlate strongly with the regulation of product and labour markets (La Porta, Lopez-de-Silanes, and Shleifer (2006)). 1 Almeida and Wolfenzon (2006) and Burkart, Panunzi, and Shleifer (2003) are notable exceptions. 0 1.1 Results Preview In this paper we try to advance on these fronts. Our first contribution is to provide a comprehensive description of ownership patterns and corporate control for the widest-possible sample of countries and listed firms. Employing big data techniques, extensive manual checks, and relying on a plethora of data sources (e.g., regulatory filings, company reports, government publications, private data providers), we augment and correct Bureau van Dijk’s ORBIS’s database on corporate ownership in an effort to identify ultimate controlling shareholders from the complex and often esoteric structures of corporate holdings. We then apply a game-theoretic algorithm based on the Shapley-Shubik power index and identify ultimate controlling shareholders for more that 40 000 listed firms from 127 countries both before and after the recent world financial crisis (2004 2012). Our corporate control identification algorithm improves over − previous works who have applied ad hoc voting rights cutoffs to classify firms as controlled (and widely held). The algorithm for example is especially useful for measuring control in firms where we observe multiple individuals/families and other entities (institutional investors, government) holding large equity blocks, a common phenomenon (Laeven and Levine, 2007). Using the newly-compiled database we provide a broad overview of the patterns of corporate control around the world. We distinguish between three types of firms, widely-held corporations without any block shareholder, widely-held corporations (without a controlling shareholder) but with one or multiple block equity stakes, and controlled firms with a dominant controlling shareholder. We also split controlled firms based on the type of dominant shareholder to state-controlled (national governments, municipalities and state agencies), family-controlled (where a single individual or family exert power), and controlled by other (listed or private) firms. There are large differences on corporate control around the world, both across and within continents. Corporate structure patterns appear persistent, as the 2007 2009 global financial crisis did not alter it much. − Second, given the literature’s emphasis we re-examine the "reduced-form" link between legal origin and corporate control. Given the large differences