法学资料大全二（英文）：公司法的自律模式

* The authors are, respectively, Professor of Law, Columbia Law School, and Professor of Law, Harvard Law School. This Article builds on an earlier article that developed the concept of a "self-enforcing" corporate law in the context of Russia, prepared for the World Bank Conference on Corporate Governance in Eastern Europe and Russia held in December 1994, and which was subsequently published as Bernard Black, Reinier Kraakman & Jonathan Hay, Corporate Law from Scratch, in 2 Corporate Governance in Central Europe and Russia: Insiders and the State 245 (Roman Frydman, Cheryl W. Gray & Andrzej Rapaczynski eds., 1996). Acknowledgments should go first to Jonathan Hay, our coauthor on the precursor article, and to Anna Stanislavovna Tarassova, the principal Russian drafter of the Russian company law, the development of which provided the genesis and many of the ideas for this Article. Other important participants in the effort to develop Russian company law include Alexander Abramov, Ian Ayres, J. Robert Brown, Catherine Dixon, Louis Kaplow, Yevgeni Kulkov, Claudia Morgenstern, Melinda Rishkofski, Howard Sherman, Sergei Shishkin, Victoria Pavlovna Volkova, and John Wilcox. We thank the participants in the World Bank conference, especially John Coffee, Ronald Gilson, Bruce Kogut, Mancur Olson, Ibrahim Shihata, and Douglas Webb, for helpful comments on earlier drafts. We also thank participants in the University of Toronto and Harvard Law School Law and Economics Workshops, Lucian Bebchuk, Jill Fisch, Jeffrey Gordon, Christine Jolls, Hideki Kanda, Mark Roe, and especially David Charny for his extensive comments. Finally, we wish to thank Joseph Blasi, Rolf Skog, Andrei Alexandrovich Volgin, and Daniel Wolfe for their comments on the draft law on which this article is based. Research support was provided by the World Bank (for both authors), the Open Society Institute (for Black), and the Harvard Program for Law and Economics, funded by the John M. Olin Foundation (for Kraakman). A SELF-ENFORCING MODEL OF CORPORATE LAW 109 Harvard Law Review 1911-1982 (1996) This paper can be downloaded without charge from the Social Science Research Network electronic library at: http://papers.ssrn.com/paper.taf?abstract_id=10037 Bernard Black Reinier Kraakman* In this Article, Professors Black and Kraakman develop a "self-enforcing" approach to drafting corporate law for emerging capitalist economies, based on a case study: a model statute that they helped to draft for the Russian Federation, which formed the basis for the recently adopted Russian law on joint-stock companies. The Article describes the contextual features of emerging economies -- including the prevalence of controlled companies and the weakness of other institutional, market, cultural, and legal constraints -- that make it inappropriate to import company law from developed countries. Professors Black and Kraakman argue that in emerging economies, the best legal strategy for protecting outside investors in large companies while simultaneously preserving managers' discretion to invest is a self-enforcing model of corporate law. The self-enforcing model structures corporate decisionmaking processes to allow large outside shareholders to protect themselves from insider opportunism with minimal resort to legal authority, including the courts. Among the model's provisions are a mandatory cumulative voting rule for election of directors, which ensures that minority blockholders have board representation, and a rule requiring both shareholder- and board-level approval for self-interested transactions. The Article examines how to induce voluntary compliance with the company law, as well as the implications of the self-enforcing model for the ongoing debate over the efficiency of corporate law in developed economies. Table of Contents Introduction .......................................................... 1 I. The National Contexts That Shape Corporate Law ........................... 7 A. Corporate Law in Developed Economies ............................ 7 B. The Goals of Corporate Law in Emerging Economies ................... 8 C. Legal and Market Controls in Emerging Economies ................... 13 D. Cultural Norms for Manager and Large Shareholder Behavior ........... 15 II. A Self-Enforcement Approach to Corporate Law ........................... 16 A. The Prohibitive Model ......................................... 17 B. The Self-Enforcing Model ...................................... 18 1. Structural Constraints ..................................... 19 2. Simple, Bright-Line Rules and Strong Remedies ................. 21 C. The Limits to the Self-Enforcement Approach ....................... 23 D. Can Law Function Without Official Enforcement? .................... 26 III. Governance Structure and Voting Rules ................................. 29 A. Allocation of Decisionmaking Power .............................. 29 B. Allocation of Voting Power: One Share, One Vote .................... 31 C. Voting for Directors: Cumulative Voting ........................... 33 D. Voting Procedures: Universal Ballot ............................... 35 E. Protecting Honesty and Quality in Voting ........................... 36 IV. Structural Constraints on Particular Corporate Actions ...................... 38 A. Mergers and Other Major Transactions ............................. 38 1. Shareholder Approval ..................................... 39 2. Appraisal Rights ......................................... 41 3. Determining Market Value ................................. 42 B. Self-Interested Transactions ..................................... 43 C. Control Transactions .......................................... 45 D. Issuance and Repurchase of Shares ................................ 49 1. Share Issuances. ......................................... 49 2. Repurchase of Shares ..................................... 51 E. Protecting Creditors and Preferred Shareholders ...................... 52 F. The State as Part-Owner ........................................ 55 V. Remedies ......................................................... 55 VI. The Path-Dependent Evolution of Developed Country Corporate Law .......... 58 Conclusion: Self-Enforcing Law in Emerging Economies ........................ 61 APPENDIX: Survey of Company Law in Emerging Markets ..................... 64 1 We use the term "institution" in a broad sense to include private organizational structures such as stock trading systems and securities registrars; public organizational structures such as securities regulators, courts with experience in commercial matters, an honest police force, and a reliable mail system; and mixed public- private structures such as self-regulatory organizations, an accounting profession, and sophisticated financial accounting rules. 2 A modified version of our proposed law was adopted in December 1995 as the company law of the Russian Federation. See Federal Law of the Russian Federation on Joint-Stock Companies, No. 208-FZ (1995), published in Rossiiskaya Gazeta, Dec. 29, 1995, at 1, translation available in Westlaw, Rusline Database, 1995 WL 798968. An annotated English translation of the law (by Bernard Black and Anna Tarassova) is available from Professor Black. 1 Introduction What kind of corporate law should govern publicly owned companies in emerging markets, including newly privatizing economies? This important question has no ready answer. Corporate law, we believe, should have the same principal goal in developed and emerging economies -- succinctly stated, to provide governance rules that maximize the value of corporate enterprises to investors. However, emerging economies cannot simply copy the corporate laws of developed economies. These laws depend upon highly evolved market, legal, and governmental institutions and cultural norms that often do not exist in emerging economies.1 Developed country corporate laws also reflect the idiosyncratic history of their country of origin. They are not necessarily efficient at home, let alone when transplanted to foreign soil. Moreover, in many emerging markets, corporate law must serve a second central goal that is less pressing in mature market economies: fostering public confidence in capitalism and in private ownership of large business enterprises. Thus, corporate law must be designed substantially from scratch to work within the infrastructure available in an emerging market. Fortunately, this can be politically feasible. Precisely because existing institutions (to which the law must adapt) are often weak or missing, one can rethink from first principles what corporate law ought to look like and what related institutions it ought to rely on and promote. Beyond producing a new model for emerging markets, the effort to develop corporate law from scratch can expose weaknesses and idiosyncracies in the corporate laws of developed countries. The model can highlight the ways in which these laws did not simply evolve toward efficiency, but instead evolved from historically contingent starting places to ending places shaped by preexisting institutions, by the inertial power of the status quo, and by the preferences of key participants in the corporate enterprise. For example, German corporate law adapted to strong banks and labor unions, while American law adapted to strong capital markets, weak financial institutions, and strong corporate managers. In this Article, we sketch the basic elements of a "self-enforcing" model of corporate law, designed for an emerging economy. The model is grounded in a case study: the effort, in which we participated, to develop a new corporate law for Russia.2 We begin with three 3 See Bernard S. Black, Is Corporate Law Trivial?: A Political and Economic Analysis, 84 Nw. U. L. Rev. 542 (1990) [hereinafter Black, Is Corporate Law Trivial?]; see also Ronald J. Gilson, A Structural Approach to Corporations: The Case Against Defensive Tactics in Tender Offers, 33 Stan. L. Rev. 819, 839- 44 (1981) (describing market mechanisms that complement legal controls on corporate managers); Mark J. Roe, Some Differences in Corporate Structure in Germany, Japan, and the United States, 102 Yale L.J. 1927, 1932 (1993) [hereinafter Roe, Some Differences] (same). 2 central claims. First, effective corporate law is context-specific, even if the problems it must address are universal. The law that works for a developed economy, when transplanted to an emerging economy, will not achieve a sensible balance among company managers' need for flexibility to meet rapidly changing business conditions, companies' need for low-transaction- cost access to capital markets, large investors' need to monitor what managers do with the investors' money, and small investors' need for protection against self-dealing by managers and large investors. The defects in the law will increase the cost of capital and reduce its availability. In developed countries, corporate law combines with other legal, market, and cultural constraints on the actions of corporate managers and controlling shareholders to achieve a sensible balance among these sometimes competing needs. Corporate law plays a relatively small, even "trivial" role.3 In emerging economies, these other constraints are weak or absent, so corporate law is a much more central tool for motivating managers and large shareholders to create social value rather than simply transfer wealth to themselves from others. The "market" cannot fill the regulatory gaps that an American-style "enabling" corporate law leaves behind. Further, corporate law in developed countries evolved in tandem with supporting legal institutions. For example, the United States relies on expert judges to assess the reasonableness of takeover defenses and the fairness of transactions in which managers have a conflict of interest. When necessary, these judges make decisions literally overnight to ensure that judicial delay does not kill a challenged transaction. A company law that depends on fast and reliable judicial decisions is simply out of the question in many emerging markets. In Russia, for example, courts function slowly if at all, some judges are corrupt, and many are Soviet-era holdovers who neither understand business nor care to learn. Better judges and courts will emerge only over several decades, as the old judges die or retire. In the meantime, Russian corporate law must rely on courts as little as possible. More generally, every emerging economy has some legal and market institutions, some norms of behavior, some distribution of share ownership, and some financial institutions. Corporate law must reflect these background facts. For example, if (as in Russia) employees often own large stakes in their companies, but are vulnerable to having their votes controlled by corporate managers, company law needs special rules that safeguard the rights of employee- shareholders. Company law must also limit the influence of dysfunctional background features, such as widespread corruption. 4 See Vincent Boland & Kevin Done, Prague to Update Market Regulations, Fin. Times, Oct. 24, 1995, at 30. 5 See Uriel Procaccia, Crafting a Corporate Code from Scratch, Public Lecture at Cardozo Law School 12-14 (Oct. 18, 1995) (on file with the Harvard Law School Library). 6 See generally Jonathan P. Charkham, Keeping Good Company: A Study of Corporate Governance in Five Countries (1994) (detailing failures in the United States, Britain, Germany, Japan, and France); Mark J. Roe, Strong Managers, Weak Owners: The Political Roots of American Corporate Finance 149-230 (1994) (United States, Japan, Germany); Bernard S. Black & John C. Coffee, Jr., Hail Britannia?: Institutional Investor Behavior Under Limited Regulation, 92 Mich. L. Rev. 1997, 2007-77 (1994) (Britain); Ronald J. Gilson & Reinier Kraakman, Investment Companies as Guardian Shareholders: The Place of the MSIC in the Corporate Governance Debate, 45 Stan. L. Rev. 985, 992-97 (1993) (Sweden); Ronald J. Gilson & Mark J. Roe, Understanding the Japanese Keiretsu: Overlaps Between Corporate Governance and Industrial Organization, 102 Yale L.J. 871, 874-82 (1993) (Japan). 3 Our second central claim is that despite the context-specificity of effective corporate law, there is a large class of emerging capitalist economies (including formerly Communist countries) that are sufficiently similar to permit generalization about the type of corporate law that will be useful for them. Russia is perhaps an extreme case, but it is hardly alone in having insider-controlled companies, malfunctioning courts, weak and sometimes corrupt regulators, and poorly developed capital markets. For example, an acute problem in Russia is protecting minority investors against exploitation by managers or controlling shareholders. Protection of minority investors has also emerged as a central political issue in the most successful post- Communist economy, the Czech Republic,4 and is at the core of recent reforms in Israeli corporate law.5 Our third claim is that our task is not impossible. Despite weak markets and institutions, one can design a company law that prevents a significant fraction of the corporate governance failures that would otherwise occur. Even developed country corporate governance systems fail with uncomfortable frequency.6 We can expect still more failures in emerging markets. Nonetheless, it is possible to design a law that works tolerably well -- that vests substantial decisionmaking power in large outside shareholders, who have incentives to make good decisions; that reduces, though it cannot eliminate, fraud and self-dealing by corporate insiders; that minimizes, though it cannot altogether avoid, the need for official enforcement through courts; that gives managers and controlling shareholders incentives to obey the rules even when they could often get away with ignoring them; that reinforces desirable cultural attitudes about proper managerial behavior; and that still leaves managers with the flexibility they need to take risks and make quick decisions. Such a law can add far more value than corporate law adds in developed economies, precisely because other institutions that could shape corporate behavior are weak in developing economies. The central features of our "self-enforcing" model of corporate law are: 7 We use here the conventional distinction between a precise "rule" (don't drive faster than 55 miles per hour) and a vague "standard" (don't drive faster than appropriate for the road and weather conditions). See Louis Kaplow, Rules Versus Standards: An Economic Analysis, 42 Duke L.J. 557 (1992). 4 (i) Enforcement, as much as possible, through actions by direct participants in the corporate enterprise (shareholders, directors, and managers), rather than indirect participants (judges, regulators, legal and accounting professionals, and the financial press). (ii) Greater protection of outside shareholders than is common in developed economies, to respond to a high incidence of insider-controlled companies, the weakness of other constraints on self-dealing by managers and controlling shareholders, and the need to control self-dealing to strengthen the political credibility of a market economy. (iii) Reliance on procedural protections -- such as transaction approval by independent directors, independent shareholders, or both -- rather than on flat prohibitions of suspect categories of transactions. The use of procedural devices balances the need for shareholder protection against the need for business flexibility. (iv) Whenever possible, use of bright-line rules, rather than standards, to define proper and improper behavior. Bright-line rules can be understood by those who must comply with them and have a better chance of being enforced. Standards, in contrast, require judicial interpretation, which is often unavailable in emerging markets, and presume a shared cultural understanding of the regulatory policy that underlies the standards, which may also be absent.7 (v) Strong legal remedies on paper, to compensate for the low probability that the sanctions will be applied in fact. Enforcement takes place primarily through a combination of voting rules and transactional rights. The central voting elements include: shareholder approval (including in some cases supermajority approval or approval by a majority of outside shareholders) for broad classes of major transactions and self-interested transactions; approval of self-interested transactions by a majority of outside directors; mandatory cumulative voting for directors, which empowers large minority shareholders to select directors (this power is protected by requirements of one common share, one vote; minimum board size; and no staggering of board terms); and a unitary ballot on which both managers and large shareholders can nominate directors. The honesty of the vote is protected through confidential voting and independent vote tabulation, while the quality of voting decisions is buttressed by mandatory disclosure rules. 8 In the economic literature, "self-enforcement" is sometimes given only this narrower meaning -- a contract is said to be self-enforcing if it induces voluntary compliance. See, e.g., Lester G. Telser, A Theory of Self- Enforcing Agreements, 53 J. Bus. 27, 27-28 (1980). Inducing voluntary compliance is an important element of our approach to company law, but it is only part of what we mean by a "self-enforcing" law. 9 Panel on Takeovers and Mergers, The City Code on Takeovers and Mergers and the Rules Governing Substantial Acquisitions of Shares (1993) [hereinafter City Code on Takeovers and Mergers]. The Panel on Takeovers and Mergers, which administers the City Code, is a self-regulatory organization, with a chair chosen by the Bank of England and members representing institutional investors, public companies, and the London Stock Exchange. Panel rulings can be enforced by a number of sanctions, including delisting from the London Stock Exchange. See Black & Coffee, supra note 6, at 2027. 10 Cf. City Code on Takeovers and Mergers, supra note 9, General Principles 4, 7, 10, at B1-B2, Rule 9.1, at F1, Rule 21, at I13 (stating similar rules). 5 Shareholders also receive transactional rights (put and call options) triggered by specified corporate actions. These include preemptive rights when a company issues new shares; appraisal rights for shareholders who do not approve major transactions; and takeout rights when a controlling stake in the firm is acquired (that is, minority shareholder rights to sell their shares to the new controlling shareholder). The self-enforcing model seeks to build legal norms that managers and large shareholders will see as reasonable and comply with voluntarily. The need to induce voluntary compliance reinforces our preference for procedural rather than substantive protections. For example, managers may evade a flat ban on self-interested transactions, yet comply with a procedural requirement for shareholder approval because they think that they can obtain approval. Once they decide to obtain shareholder approval, the managers may make the transaction more favorable to shareholders, to ensure approval and avoid embarrassment.8 The model often relies not only on bright-line rules, but also on relatively simple rules. Managers can't comply with, and judges can't enforce, rules that they don't understand. Nor will managers respect an unduly complex statute. The British City Code on Takeovers and Mergers offers a good set of self-enforcing rules to govern change-of-control transactions, which we largely adopt.9 We propose a delay period before a change of control occurs to provide a market check on the fairness of the price (30% ownership is our proxy for control); a takeout offer requirement, under which a new controlling shareholder must offer to purchase minority shares (unless the minority shareholders waive this requirement by majority vote); and a ban on defensive actions that could frustrate a takeover bid (unless the actions are approved in advance by the target's shareholders).10 Shareholders are protected against dilutive share issuances through a combination of: preemptive rights; a requirement that shares be issued only at market value (determined by the board of directors); shareholder approval for issuances to insiders (under the self-interested transaction rules); and shareholder approval for large issuances (our threshold is 20% of the previously outstanding shares). 6 These and other features of the self-enforcing approach produce a company law that is novel in the aggregate, even though many individual provisions (such as one share, one vote and cumulative voting) are familiar in developed markets. To be sure, there are limits to what a self-enforcing corporate law can accomplish. For example, cumulative voting can strengthen the monitoring power of large outside shareholders but is of little direct help to shareholders who own five shares each. Nor are these shareholders likely to exercise appraisal rights if they oppose a merger. Nevertheless, the self-enforcing model can partially protect small shareholders. All shareholders benefit if large outside shareholders can monitor management performance and control self-dealing. All shareholders also benefit if the law induces managers to comply voluntarily. Small investors remain vulnerable to insider self-dealing that includes a hidden payoff to large outside shareholders. But concealment won't always be possible and, when possible, won't always be attractive because each participant is thereafter vulnerable to exposure by the others. A caveat: we call our model "self-enforcing." This phrase is a shorthand attempt to capture the main lines of our model, including our effort to minimize reliance on official enforcement. But our model is not purely self-enforcing, any more than Delaware's "enabling" corporate law is purely enabling, or the "prohibitive" model that characterized American corporate law a century ago was purely prohibitive in character. We can only reduce, not wholly avoid, the need for official enforcement. We develop the basic elements of a corporate law for emerging economies as follows. Part I describes the goals of corporate law in an emerging economy and the elements of national context that affect the law's shape. Part II outlines the alternative drafting strategies open to emerging economies and explains why our preferred strategy -- a "structural" or "self- enforcing" corporate law -- is likely to be superior to the available alternatives. Parts III, IV, and V describe the primary components of a self-enforcing corporate law in the particular context of Russia. Finally, Part VI considers the lessons that can be drawn from the self- enforcement approach for the supposed efficiency of corporate law in developed countries. The Appendix compares selected features of the Russian self-enforcing law with the corporate statutes of seventeen relatively advanced emerging markets. The self-enforcing model contains more procedural protections and fewer substantive protections than any of the other statutes. We do not, however, advocate wholesale change in existing laws. A company law that is already meeting a particular country's needs should enjoy deference because its success probably reflects adaptation to local institutions. The self-enforcing model can be a base for a new law if the current law is seriously deficient, and a source of ideas for improving laws that already work tolerably well. We focus here only on corporate law as conventionally understood: the law that articulates company structure and regulates relationships among shareholders and between shareholders and corporate managers. American corporate law, thus defined, includes state corporation statutes; the common law of fiduciary duty; the provisions of the securities laws that regulate insider liability, shareholder voting, and control contests; and stock exchange 11 Similarly, British company law, for our purposes, includes statutory company law, the common law of fiduciary duty, the London Stock Exchange's listing standards and guidelines, and the City Code on Takeovers and Mergers. 12 A brief word on codetermination, for those who think this issue too important to be excluded from our Article: we are not convinced that mandatory employee participation on boards of directors is a good idea even in Germany, where it began. Moreover, the effort to transplant the two-tier board to the Czech Republic has failed. Investors there care only about what they see as the "real" board -- the management board. See John C. Coffee, Jr., Institutional Investors in Transitional Economies: Lessons from the Czech Experience, in 1 Corporate Governance in Central Europe and Russia: Banks, Funds, and Foreign Investors 111, 152-53 (Roman Frydman, Cheryl W. Gray & Andrzej Rapaczynski eds., 1996) [hereinafter Coffee, Czech Institutional Investors]. Russia offers an especially weak case for mandating employee participation in corporate governance. First, employees in most privatized companies own ample shares to elect their own directors under our proposal for mandatory cumulative voting. Second, in Russia and other newly privatized economies, many companies must greatly reduce their work force to remain competitive. Current employees will often resist these changes. Third, under Communism, Russian company unions had symbolic value but no real power. They remain weak and often corrupt. The Russians with whom we have discussed codetermination find an assumption underlying codetermination -- that labor unions can aggressively represent the interests of employees -- amusing. 7 listing standards that impose governance requirements on listed companies.11 By contrast, regulation of the relationship between workers and companies, such as mandatory employee representation on the board of directors along the lines of German codetermination,12 or state support for employee ownership through tax benefits as in American employee stock ownership plans, is beyond the scope of this Article. This Article also focuses on large companies where at least some shareholders do not work in the business. A well-drafted law, of course, must also consider the special problems of close corporations. The procedural protections that are appropriate for a company with 10,000 shareholders would be ludicrous and crippling for a tiny company with five shareholders who all work in the business. I. The National Contexts That Shape Corporate Law Five aspects of national context, in our view, shape and limit corporate law: the goals of corporate law; the sophistication of capital markets and related institutions; the sophistication and reliability of legal institutions; the ownership structure of public companies; and the cultural expectations of participants in the corporate enterprise. In this Part, we describe how these features interweave to form the context of corporate law in developed economies. We then demonstrate, using Russia as a case study, how these features differ markedly in emerging economies -- differences in context that require differences in company law. A. Corporate Law in Developed Economies 13 See, e.g., American Law Inst., Principles of Corporate Governance: Analysis and Recommendations § 2.01(a) (1994) [hereinafter Principles of Corporate Governance] (stating that, subject to certain constraints, "a corporation should have as its objective the conduct of business activities with a view to enhancing corporate profit and shareholder gain" (citation omitted)). There are important departures from this norm, such as German codetermination or the antitakeover provisions in some American state corporate laws. But these are seen as just that -- departures from an overall efficiency norm. We do not enter here the debate over whether corporate law can or should encourage companies to pursue goals other than profit maximization. 14 See Black, Is Corporate Law Trivial?, supra note 3, at 551-62. 8 Corporate law, as we define it above, is generally understood to have a largely economic function in developed economies. This function might be characterized as maximizing the value of corporate enterprises to investors and therefore (on the whole) to society, or as minimizing the sum of the transaction and agency costs of contracting through the corporate form.13 From this perspective, corporate law provides a set of rules (often default rules that can be varied in the corporate charter) that encourage profit-maximizing business decisions, provide professional managers with adequate discretion and authority, and protect shareholders (and to some extent creditors) against opportunism by managers and other corporate insiders. But no one imagines that corporate law accomplishes these objectives by itself. Many other control mechanisms also limit departures from the profit-maximization norm. In the United States, for example, a competitive product market, a reasonably efficient capital market, an active market for corporate control, incentive compensation for managers, and at least occasional oversight by large outside shareholders all exert pressures on corporate managers to enhance firm value. Sophisticated professional accountants, elaborate financial disclosure, an active financial press, and strict antifraud provisions assure shareholders of reliable information about company performance. Sophisticated courts (such as the Delaware Chancery Court), administrative agencies (such as the Securities and Exchange Commission), and self-regulatory organizations (such as the New York Stock Exchange) keep sharp eyes out for corporate skullduggery. These multiple private and legal controls shoulder much of the burden of protecting investors in public companies, so that the corporate law itself can tilt far in the direction of providing managerial discretion and enhancing transactional flexibility. Most American state corporate statutes (typified by Delaware's) have evolved into "enabling" laws, many of whose major provisions are default rules. Moreover, many of the mandatory rules in American corporate codes have survived because they are either unimportant, avoidable through advance planning, or match reasonably well what the parties would have chosen anyway.14 The statutes are accompanied by fiduciary doctrines that give the courts wide latitude to review opportunistic behavior ex post, but the courts generally punish only the most egregious instances of self-dealing or recklessness. All else is left to private institutions and the market. B. The Goals of Corporate Law in Emerging Economies 15 The corporate laws of emerging economies tend to reflect a different balance between enabling and restrictive provisions than do the laws of developed economies. The survey of advanced emerging markets in the Appendix shows that many of these countries have retained (in varying degrees) more substantive and procedural protections for outside shareholders and creditors than have developed economies such as the United States. 16 See Joseph Blasi & Andrei Shleifer, Corporate Governance in Russia: An Initial Look, in 2 Corporate Governance in Central Europe and Russia: Insiders and the State, supra note *, at 78, 79-82 (reporting a survey of 200 privatized Russian companies that shows mean (median) employee ownership of 65% (60%)). For background on Russia's privatization scheme, see Anders Aslund, How Russia Became a Market Economy 223-71 (1995); Maxim Boycko, Andrei Shleifer & Robert Vishny, Privatizing Russia 69-123 (1995); and Maxim Boycko, Andrei Shleifer & Robert W. Vishny, Voucher Privatization, 35 J. Fin. Econ. 249, 256-65 (1994) [hereinafter Boycko, Shleifer & Vishny, Voucher Privatization]. 17 A recent example in which even sophisticated investors were hurt was a large stock issuance (roughly doubling the number of outstanding shares) by the Komineft Oil Company, which had been among the most popular Russian stocks among foreign investors. The new shares were sold in early 1994, principally to the managers and employees of Komineft, at far below market value, but the issuance was not publicly announced until six months later. The issuance heavily diluted the interests of large shareholders who invested in 9 Corporate law in an emerging economy must address a broader set of goals, and operate within a far less evolved market and legal infrastructure, than corporate law in a developed economy. The paradoxical consequence is that the protective function of corporate law becomes more important precisely when fewer other resources are available to support that function. Consider first the goals of corporate law in emerging economies. The efficiency goal of maximizing the company's value to investors remains, in our view, the principal function of corporate law. But the balance between investor protection and the business discretion of corporate managers needed to achieve this goal will be quite different in emerging than in developed economies.15 In addition, in countries that are emerging from heavy state control of industry, a second central objective is the political goal of fostering public confidence in a market economy and in private ownership of large enterprises. The efficiency goal dictates that corporate law provide more investor protection in emerging than in developed economies, for several reasons. One is that insiders are likely to exercise voting control over most public companies. Such controlled ownership structures raise the obvious concern that the insiders, whether managers or controlling shareholders, will behave opportunistically toward other shareholders. In Russia, for example, the structure of mass privatization has led to the great majority of privatized public companies being controlled by management-led coalitions of managers and workers, which typically hold 51-75% of a company's voting shares. Outside shareholders -- including banks and investment (voucher) funds -- hold on average 15-20% of the voting shares, while the remaining shares are likely to be held by individuals, by other companies, and by a state property fund.16 This ownership structure presents a clear risk of opportunism toward outside shareholders.17 Although the Komineft before the issuance was belatedly announced. See Neela Banerjee, Russian Oil Company Tries a Stock Split in the Soviet Style, Wall St. J., Feb. 15, 1995, at A14. Efforts by the Russian Securities Commission to have the issuance cancelled failed: Komineft's management merely apologized to investors and promised not to make a secret share issuance again. See Julie Tolkacheva, Komineft Agreement Leaves Investors Cool, Moscow Times, Oct. 31, 1995, at 111. 18 In the Czech Republic, for example, mass privatization led to financial institutions (usually banks and investment funds) and financial-industrial groups holding controlling stakes in many large companies. See Coffee, Czech Institutional Investors, supra note 12, at 112-13. 19 It is well established that, under an American-style enabling statute, controlling shareholders frequently extract private gains from corporations at the expense of minority shareholders, despite the market constraints discussed in section I.A. See, e.g., Michael J. Barclay & Clifford G. Holderness, The Law and Large-Block Trades, 35 J.L. & Econ. 265, 267-78 (1992); Stuart Rosenstein & David F. Rush, The Stock Return Performance of Corporations That Are Partially Owned by Other Corporations, 13 J. Fin. Res. 39 (1990); Roger C. Graham, Jr. & Craig E. Lefanowicz, The Valuation Effects of Majority Ownership for Parent and Subsidiary Shareholders (Oregon State Univ. College of Bus. Working Paper, Feb. 1996). 20 Two examples: first, in the one true Russian public stock offering to date, by the Red October candy company in 1994, relatively few investors were willing to buy at the offering price. See Janet Guyon, Russian Firms Face Fund-Raising Woes in the Wake of Privatization Explosion, Wall St. J., Apr. 17, 1995, at B6A. Second, the Russian natural gas giant, Gazprom, tried in mid-1995 to sell a roughly 10% stake to foreign investors, but withdrew the offer after discovering that the price investors would pay was far below outside estimates of Gazprom's true value. See Steve Liesman, Limits on Sale Damp Shares of Gazprom, Wall St. J. Eur., Apr. 3, 1995, at 11. 10 particular form of controlled ownership in Russia may be unique,18 experience teaches that family-controlled companies with minority public participation -- the classic ownershipstructure in newly industrialized economies -- also require strong minority protections.19 Outside investors, facing ex ante a high risk of insider opportunism, will insist on a high expected rate of return in the cases when insiders turn out to behave properly, to compensate for the risk of bad behavior. In an extreme case like Russia, where the risk of insider opportunism is especially high (we explore below the multiple reasons for this), these rational discounts of share prices -- to far below the shares' fair value assuming good behavior -- can virtually paralyze the equity markets. Honest (nonopportunistic) managers won't offer shares at what they see as ridiculously low prices. Investors, meanwhile, won't pay more because they don't know which managers will misbehave. The risk of government misbehavior (such as renationalization, currency controls, or confiscatory taxation) further increases the gap between managers' perception of their firm's value and market prices for the firm's shares. Russian companies that have sought to issue shares have found that investors are willing to pay only a fraction of the company's true worth.20 In theoretical terms, Russian companies operate in a market with a severe "lemons" (adverse selection) problem, in which only low-quality issuers are interested in raising equity capital at current prices. Strong minority protections respond to this problem by narrowing the range and reducing the likelihood of insider opportunism. Investors will then pay more for 21 See generally Michael Jensen & William Meckling, Theory of the Firm: Managerial Behavior, Agency Costs and Ownership Structure, 3 J. Fin. Econ. 305, 312-33 (1976) (arguing that entrepreneurs who sell equity bear the anticipated agency costs of equity). 22 For an account of the contribution of such informational intermediaries to the efficiency of American capital markets, see Ronald J. Gilson & Reinier H. Kraakman, The Mechanisms of Market Efficiency, 70 Va. L. Rev. 549, 613-21 (1984). 23 The most eloquent American proponents of the enabling model are Frank H. Easterbrook & Daniel R. Fischel, The Economic Structure of Corporate Law 1-39 (1991); and Roberta Romano, The Genius of American Corporate Law 86-91 (1993). For efforts to develop the limits of the enabling approach, see Lucian A. Bebchuk, Limiting Contractual Freedom in Corporate Law: The Desirable Constraints on Charter Amendments, 102 Harv. L. Rev. 1820, 1835-60 (1989) [hereinafter Bebchuk, Limiting Contractual Freedom]; Melvin A. Eisenberg, The Structure of Corporation Law, 89 Colum. L. Rev. 1461, 1471-515 (1989); and Jeffrey N. Gordon, The Mandatory Structure of Corporate Law, 89 Colum. L. Rev. 1549, 1554-85 (1989) [hereinafter Gordon, Mandatory Structure]. We do not enter here the debate on the proper limits on the enabling approach in a developed economy. 11 shares, which will make higher-quality issuers interested in issuing shares. This will further reduce the ex ante likelihood of expropriation of investors' funds and further raise stock prices, until a new equilibrium is reached with higher share prices and a lower cost of capital. To be sure, in a world of perfect contracting, without informational asymmetries, contracting costs, or naive managers or investors, appropriate protections for minority shareholders would emerge by contract, without the need for government fiat. In such an ideal world, marked by extensive disclosure and populated by savvy investors and issuers, corporate planners would have incentives to offer optimal investor protection through their charters.21 Sophisticated market intermediaries, such as investment banks, accounting firms, and law firms, would advise issuers on what disclosures and contractual protections to offer and would bond the reliability of the issuer's disclosures.22 After shares were issued, the same efficient capital market -- together with the product market and the market for corporate control -- would police company managements. As a consequence, corporate law could be substantially "enabling," even if insiders dominated public companies. This description points to a second, more general reason why efficiency concerns favor a protective corporate law in emerging economies. The enabling model and its underlying assumptions about capital markets have both strengths and weaknesses in developed countries.23 But the assumptions that support the enabling model are clearly inapposite in emerging economies, where informational asymmetries are severe, markets are far less efficient, contracting costs are high because standard practices have not yet developed, enforcement of contracts is problematic because of weak courts, market participants are less experienced, reputable intermediaries are unavailable or prohibitively expensive, and the economy itself is likely to be in flux. In recently privatized economies such as Russia, the contractarian base for the enabling model fails for a third reason: the initial structure of relationships among company participants 24 See Julie Tolkacheva, Secret Takeover Unnerves Investors, Moscow Times, Apr. 5, 1995, at 1, 2 (reporting that Primorsk Shipping doubled its outstanding shares and sold the additional shares for a nominal price to an affiliate controlled by Primorsk's managers, and also that Far Eastern Shipping plans a similar action); supra note 17 (describing the secret share issuance by Komineft to insiders); infra note 31 (describing Krasnoyarsk Aluminum's erasure of a 20% shareholder from its share register). 25 See, e.g., Boycko, Shleifer & Vishny, Voucher Privatization, supra note 16, at 250-53. 12 arose from government fiat rather than private contract. The ownership structure and initial charters of privatized Russian companies were imposed by the privatization program, and the company's principal bank lenders were often selected before privatization began. These relationships were not negotiated by outside investors with an eye to their own self-protection. Beyond the efficiency justifications for protective corporate law, political goals support strong shareholder protection in emerging economies. Political goals also shape the law of developed economies, but they are more important in economies where capitalism is less firmly rooted. Egregious opportunism or corporate scandals may erode the political legitimacy of private ownership of large firms, as well as support for the market economy generally. Even if corporate scandals do not trigger a political maelstrom of populist reaction, they will damage investor confidence in an environment where disclosure is minimal and legal remedies are slow and uncertain. In Russia, for example, a few well-publicized cases of manager mistreatment of shareholders have seriously impaired the willingness of investors, especially foreign investors, to buy shares of Russian firms and surely contributed to the roughly 75% collapse in Russian share prices from summer 1994 through late 1995.24 Thus, the twin risks of a destructive political reaction to scandal and of investor overreaction to scandal are important negative externalities that become more serious as legal rules allow greater insider discretion. These risks justify stricter controls on abuse-prone activities than would be appropriate in a developed market -- doubly so because market and cultural controls on abuse of position are relatively weak in emerging economies. A further political justification for protective corporate law emerges in mass-privatized economies such as Russia, where the government has transferred shares to employees or the general public for nominal consideration. Such a privatization program reflects, in part, a political bargain on how to distribute social wealth.25 The recipients of shares of privatized enterprises expect these shares to have real value. But in Russia, many citizens have been disappointed by the market value of their shares and the low dividends they receive. If these recipients also come to believe (often correctly) that insiders are getting rich at their expense by expropriating the cash flow of privatized companies, the political bargain will be breached. 26 See, e.g., Can Yeltsin Win Again?, Economist, Feb. 17, 1996, at 43 (describing the Communists' political resurgence in Russia and President Yeltsin's subsequent decision to fire Anatoly Chubais, the architect of privatization and the last remaining prominent free-market reformer in the Yeltsin administration); Peter Galuszka & Rose Brady, The Battle for Russia's Wealth: Can Rich New Capitalists Weather a Popular Backlash, Bus. Wk., Apr. 1, 1996, at 50 (describing political backlash against the conspicuous wealth of new Russian tycoons). 27 Our focus here is on civil enforcement of the rights and duties established by the corporate law. Criminal enforcement, which can also be relevant in the corporate context in extreme cases, suffers from additional limitations, including poorly paid, poorly trained, and sometimes corrupt investigators and prosecutors. 28 An example: one of us is familiar with a contract case to recover 32,000 rubles brought in 1990, when the ruble-dollar exchange rate was around 1:1. The damage award was paid in 1995, by which time 32,000 rubles were worth around $7. 29 A senior member of the Supreme Russian Arbitrage Court (the Russian "arbitrage" courts exercise jurisdiction over commercial disputes) recently admitted, with unusual candor: "This share business is too complicated for us. We don't understand it. We have no laws to deal with it." Elif Kaban, Shares, Guns and Bodyguards in Russia's Courts, Reuters, May 14, 1995, available in LEXIS, News Library, Wires File. 30 For example, the Russian arbitrage courts have experienced little change in personnel since the demise of the Soviet Union. Current judicial salaries are as laughable as other official Russian salaries. A senior judge today earns around $100 per month, barely more than a subsistence wage. A competent judge can increase his 13 This can -- and in Russia, already has26 -- undermined popular support for further privatization and other reforms needed for a healthy market economy. C. Legal and Market Controls in Emerging Economies Even as the economic and political goals of corporate law in emerging economies favor a strong protective function, limitations on enforcement resources constrain how this protective function is discharged. Developed economies usually have sophisticated enforcement institutions that can implement complex, finely nuanced rules. Emerging economies have less sophisticated enforcement institutions, and hence need simpler, more easily administrable rules. The most significant enforcement limitation is weak judicial enforcement.27 Several weaknesses in a judicial system can hobble the enforcement of corporate law in emerging markets. First, the substantive legal remedies available to judges may be ill-defined or inadequate. A simple but telling example in Russia is the absence of a rule permitting judges to adjust damages for inflation. Without such an adjustment, damage awards in a high-inflation environment will compensate for only a negligible fraction of the actual loss when paid some years later.28 Moreover, judicial procedures may be so cumbersome, or the court system so overtaxed, that timely judicial action may be impossible to obtain except in rare cases. Russia, for example, has no analogue to a preliminary injunction. Further, the judiciary may lack experience with corporate law cases,29 may be corrupt, or may be so ill-paid that skilled lawyers will not take judicial jobs.30 salary by ten times or more by returning to the private sector as a lawyer -- leaving the incompetent and the corrupt to staff the judiciary. 31 For example, Russia's dematerialized system of shareholding, in which the company register is the only official record of share ownership, creates a risk that company managers will simply erase an unwanted shareholder from the shareholder register. The Russian lawyers whom we asked about this risk expressed confidence that such an effort would fail -- the shareholder could go to court and get her ownership interest restored. An important test of this belief involves Krasnoyarsk Aluminum. In late 1994, the managers of this reputedly mafia-controlled firm canceled the register entry for a foreign investor (also with some unsavory connections) who claimed to own 20% of the company's shares, and forcibly barred the investor's representatives from attending a shareholder meeting. See Russian Aluminum: King of the Castle?, Economist, Jan. 21, 1995, at 62. The subsequent lawsuit by the shareholder has not yet been resolved. 32 For example, the Russian Securities Commission was formally created by presidential decree only in November 1994. The Commission has a tiny budget and an almost nonexistent staff. See, e.g., Steve Liesman, Roiling Stock: Shareholders Meetings in Russia Set Stage for Free-Market Fight, Wall St. J. Eur., Apr. 20, 1995, at 1 [hereinafter Liesman, Roiling Stock]. Its members include representatives of the Ministry of Finance and the Central Bank, both of which opposed the Commission's creation and continue to lobby for limits on its power. See, e.g., id.; Steve Liesman, Russia's Central Bank Appears to Call for Removal of Top Securities Regulator, Wall St. J., Sept. 8, 1995, at A6. 33 For an extreme example, the Russian Ministry of Finance, which issues accounting rules, requires companies to account for as profit (and pay taxes on) the excess of the sale price of shares over the nominal (par) value of the shares. 14 If these weaknesses are present in an extreme form, judicial enforcement of corporate law will collapse -- as it largely has in Russia. But total breakdown is merely one end of a continuum. Courts may be able to enforce simple rules and resolve easy cases, at least some of the time. Just as the criminal law deters as long as the police catch some criminals, the corporate law can deter misbehavior as long as some misdeeds are remedied in the courts.31 Enforcement will be easier if courts can often resolve disputes by applying bright-line rules rather than broad standards. In addition to having weak courts, emerging markets are unlikely to have administrative agencies that can handle issues, such as financial disclosure, that benefit from detailed rulemaking and administrative enforcement.32 Moreover, these markets lack the nonlegal enforcement resources found in developed economies, including self-regulatory institutions (such as the New York and London Stock Exchanges and the British Panel on Takeovers and Mergers) and private firms that protect clients against abuse and reduce informational asymmetries (such as investment banking, law, and accounting firms). Accounting rules in emerging economies are likely to be weak or nonexistent and administered by a commensurately undeveloped profession. Russia, for example, has rudimentary and sometimes bizarre accounting rules that were developed for state enterprises, and virtually no accountants with training comparable to that of American certified public accountants.33 In developed economies, disclosure is an important constraint on management behavior. Disclosure of management self-dealing can lead to formal enforcement. Disclosure of self- 34 The principal cause of effective tax rates that can exceed 100% of pretax income is rules that limit which expenses can be deducted from revenue in computing pretax income. See, e.g., George Melloan, Russia Tailspins into a Laffer Curve Crisis, Wall St. J., Mar. 4, 1996, at A15 (reporting that, in Russia, wages above a specified (low) level are not deductible in computing pretax income). Developed countries also have rules limiting which expenses are deductible, but in much less extreme form. 35 See id. ("[E]ven those companies that are well run and are making a lot of money don't wish to audit themselves in keeping with international accounting principles because if they do the government will take what they are making away." (quoting Moscow investment banker Boris Jordan) (internal quotation marks omitted)). 15 dealing or business problems can also lead to market sanctions, such as a drop in stock price, reduced availability of credit, and difficulty in hiring employees. Embarrassment from public disclosure also exerts important discipline. For example, American boards of directors have many times replaced a poor CEO after -- but only after -- sharply critical stories appeared in the business press. In an emerging market, disclosure, and thus its attendant benefits, is likely to be diminished or absent. Russia is an extreme case where market pressures work against good disclosure. A company that discloses its accounts honestly can find itself paying taxes that exceed 100% of pretax income.34 Small firms also face a severe risk of mafia extortion and know that the local mafia are avid readers of financial disclosures, the better to judge how much payment to demand. The supposedly confidential financial statements required by the tax laws, once given to the government, are often delivered to the mafia by corrupt officials. In this environment, investors do not even want the companies in which they invest to report profits honestly. The risk that managers will steal hidden profits is preferable to the certainty that the government or the mafia will take even more after honest disclosure.35 D. Cultural Norms for Manager and Large Shareholder Behavior A further reason why developed countries can make do with weak formal corporate law rules is that managers and shareholders are embedded in a culture that discourages opportunism. In part, the culture reflects the underlying legal norms and the penalties for violating those norms. But cultural attitudes also exist independently of and reinforce the legal norms, so that formal enforcement is infrequently needed. Few American corporate managers doubt that they work for the shareholders, even if they and their shareholders sometimes disagree about what this concept means. More generally, most managers in developed countries routinely follow laws of all kinds and think of themselves as law-abiding. Russia offers a marked contrast. Managers of Russian enterprises cannot follow the law and stay in business. They must lie about their income to the tax authorities; bribe the tax inspector, the customs inspector, the local police, and many other government officials; pay off the local mafia; and conduct business despite an intricate and often senseless web of rules. Not surprisingly, these managers often see corporate law as merely another obstacle, to be overcome in any way possible. Some managers have declared their corporate charter, or the 36 See, e.g., Liesman, Roiling Stock, supra note 32, at 12 (describing a Sovintorg shareholder meeting at which armed guards enforced management's decision to exclude certain shareholders). 37 An example of Russian attitudes toward legal rules is the reaction of one company to a privatization decree that required two-thirds of the board of directors to be non-employees. A company representative reported to an interviewer: "We fired our deputy director, and he was elected [to the board] as [an] outsider. After some time will pass, we will hire him back." Interview with company manager in St. Petersburg, Russia (Oct. 11, 1994) (transcript provided by Professor Joseph Blasi, Rutgers Univ.). 16 stock ownership of top management, to be "commercial secrets." Some lock unwanted shareholders out of shareholder meetings,36 conduct shareholder votes by show of hands (dominated by employees), or refuse to transfer shares if they don't approve of the new owner.37 Misconduct so basic is rare in developed markets, precisely because it would be instantly condemned as making hash out of the ground rules of the corporate form. To be sure, every country has some cultural ground rules. Russian managers, for example, often refuse to record a transfer of shares, but they rarely simply erase an unwanted shareholder from the share register. Still, cultural understandings about proper management behavior are likely to constrain managers more weakly in emerging than in developed markets. The weaker the cultural constraints, the larger the role that corporate law must play. The corporate law can respond to judges' and managers' lack of sophistication about how corporate managers should behave by making its requirements more precise. Vague standards will rarely be understood, and will rarely be followed even when they are understood. By contrast, explicit instructions are more likely to be followed and enforced, and over time can help to inculcate a sense of appropriate behavior in managers. Broad fiduciary standards can have long-term value in emerging markets because they can foster a managerial culture of duty to shareholders. In the near-term, however, the enforceable core of the law must be based on bright-line rules as much as possible. II. A Self-Enforcement Approach to Corporate Law Having surveyed the constraints under which corporate law in emerging markets must operate, we are ready to examine more closely what form the law should take. Some aspects of corporate law for emerging markets follow easily from the strong protective function that the law must discharge and the limited tools available to enforce it. Given the weakness of market and cultural checks on corporate insiders and the prevalence of controlled companies, the core rules should often be mandatory, rather than default provisions changeable by shareholder vote. Moreover, the law should consist, as much as possible, of relatively simple rules that can be understood and applied by corporate participants and judges alike. When we turn to the task of designing specific rules, two general approaches are possible. One we term the "prohibitive model": a law that bars a wide variety of suspect corporate behavior in considerable detail. The second we term the "self-enforcing model": a 38 The prohibitive model, which imposes substantive regulation, and the self-enforcing model, which imposes procedural constraints, are ideal types, as is the enabling model that characterizes American and British company law. As indicated in the Appendix, existing emerging market statutes generally include a mix of all three forms of regulation. Nevertheless, many of these statutes have a distinctive prohibitive or self-enforcing orientation. Only one of the 17 surveyed countries (South Africa) has a statute that is primarily enabling in character. Emerging market statutes with continental roots tend to be prohibitive, while statutes with British Commonwealth roots are more self-enforcing. 39 A key drafter of the Russian Civil Code, Dean Yevgeni Alexeyevich Sukhanov of the Moscow State University Law Faculty, presented to one of us a 1994 reprinting by his students of a 1914 Russian textbook on corporate law. See generally G.F. Shershenevich, Uchebnik Torgovogo Prava [Textbook on Business Law] (SPARK 1994) (1914) (reprinting the text with an introductory article by Dean Sukhanov). He explained his view that Russia's current problems were much like those described in the book, and required a return to the solutions proposed in this ancient text. 17 law that creates corporate decisionmaking processes that allow minority shareholders to protect themselves by their own voting decisions and by exercising transactional rights. A. The Prohibitive Model The prohibitive model is familiar from nineteenth-century corporation statutes in the United States and Great Britain and, to some extent, from European corporate codes and emerging market corporate codes today.38 A prohibitive code simply bars many kinds of corporate behavior that are open to abuse, such as self-dealing transactions and cashout mergers. Such prohibitive statutes were adopted in the United States and Britain under market conditions that resemble those of emerging economies today. One plausible approach to corporate law for emerging economies is to return to this restrictive drafting strategy from developed countries' pasts. In Russia, for example, the drafters of the Civil Code provisions that regulate companies consciously borrowed elements of the prohibitive approach from America's and Russia's pasts.39 That developed economies have evolved away from the prohibitive model toward an enabling model -- far away, in the United States and Great Britain, less far in Continental Europe -- does not mean that the prohibitive model is inappropriate for emerging markets. Experience in Great Britain and the United States teaches that an enabling law only weakly protects minority investors from controlling insiders who are determined to exploit them. Both countries have had their share of scandals and scoundrels -- Robert Maxwell in the United Kingdom and Victor Posner in the United States are prototypical examples. If the gross abuse of power by controlling insiders is not common in either country, this is partly for lack of opportunity (controlled companies are relatively uncommon) but primarily -- as we have already argued -- because multiple markets and institutions constrain insider 40 From this perspective, it is not surprising that European corporate codes have evolved less far than British and American laws from prohibition to the enabling model. European companies are more likely to be insider- controlled than British and American companies, and European countries are less likely to have active public stock markets, which implies weaker market controls. 41 See, e.g., Romano, supra note 23, at 87-89. 42 See Robert C. Clark, Corporate Law § 14.3, at 610-16 (1986). For us, there was perverse amusement in watching the Russian Civil Code drafters resolutely relying in 1994 on charter capital as a basic form of investor protection, while Russian company managers, having quickly learned the lessons that American managers learned early in the twentieth century, were routinely selling stock with a market value many times its par value (aided in this effort by high inflation). See Grazhdanskii Kodeks RF, pt. I, arts. 96-102 (1994) [hereinafter GK RF (Civil Code)], translated in Civil Code of the Russian Federation (William E. Butler trans., Interlist Pub. 1994). 18 opportunism. Enabling statutes would fare far worse in emerging economies, where controlled firms are the norm and nonlegal restraints on controlling insiders are weak.40 But these considerations suggest only that the prohibitive model may be a worthy competitor to the enabling model in emerging economies, not that it dominates other possible approaches. Prohibitive statutes have severe drawbacks even in emerging markets. First, they often impose major costs on companies by mechanically limiting the discretion of corporate managers to take legitimate business actions. By most accounts, the driving force behind the rise of enabling statutes in developed markets was the value of transactional flexibility to corporate managers, and ultimately to shareholders.41 The inflexibility of substantive prohibitions can be (and in Continental Europe often is) reduced by creative judicial interpretation, but this requires creative and knowledgeable judges, who are likely to be absent in an emerging market. Second, we know very little about how effective prohibitive statutes are in thwarting opportunism. Many formal constraints become ineffective as practitioners discover how to avoid them. The classic Anglo-American example is the demise of the protective function of legal capital following the introduction of low-par stock.42 Moreover, over time, severe substantive prohibitions will tend to be relaxed by legislators to meet firms' business needs (or the political demands of corporate managers). Indeed, in Russia, which already had many large manager-controlled companies, it would have been politically naive to expect the legislature to prohibit all self-dealing transactions. Yet, if these prohibitions are absent, the prohibitory model offers nothing to replace them with. Third, prohibitive statutes require significant judicial or administrative involvement. Transaction planners will look for ways to comply with the letter of the statute but not its spirit. These efforts, in turn, require knowledgeable judges who can resolve the resulting disputes in sensible ways. B. The Self-Enforcing Model 43 The strategy of shifting legal enforcement from courts and regulators to private parties who are well positioned to thwart misconduct is frequently deployed outside the company law context. One of us has called this strategy, using examples from tort and criminal law, "gatekeeper enforcement." See Reinier H. Kraakman, Gatekeepers: The Anatomy of a Third-Party Enforcement Strategy, 2 J.L. Econ. & Org. 53, 73 (1986). The gatekeeper strategy in the tort context involves imposing legal liability on one private party to create incentives for that party to control the conduct of another. In contrast, the self-enforcing strategy for company law provides outside shareholders, who already have the incentive to control insider opportunism, with the means to do so. 44 For precisely this reason, investment funds in both Russia and the Czech Republic lobbied successfully for relaxation of legal restrictions (adapted from the Investment Company Act of 1940, 15 U.S.C. §§ 80a-1 to -64 (1994)) on the percentage stake that a fund could hold in a single company. 45 See Albert O. Hirschman, Exit, Voice, and Loyalty: Responses to Decline in Firms, Organizations, and States 15-54 (1970). 19 A central claim of this Article is that, in emerging markets, a self-enforcing model of corporate law -- in which mandatory procedural and structural rules empower outside directors and large minority shareholders to protect themselves against opportunism by insiders -- dominates both the prohibitory model and the enabling model. The self-enforcing model minimizes the need to rely on courts and administrative agencies for enforcement. Thus, it is robust even when these resources are weak. And the model combines much, though not all, of the flexibility of the enabling model with a degree of investor protection that the enabling model cannot match, and perhaps the prohibitory model cannot match either. We sketch here the main lines of the self-enforcing model. Details in the Russian context and more extensive justification of particular features of the model are introduced in Parts IV, V, and VI.43 The self-enforcing model is designed to harness the monitoring ability of large, albeit still minority, outside shareholders. Collective action problems preclude effective monitoring by small shareholders. But large shareholders, in defending their own self-interest, will often defend the interests of small shareholders as well. Many companies are likely to have large outside blockholders, in part because sophisticated investors understand all too well the weak position of a small outside shareholder and thus prefer to hold an influential block of a company's stock, if they own its shares at all.44 Corporate law and investor preferences interact: the more influence that the law gives to large outside investors, the more likely investors are to choose to hold large stakes -- and to use the influence that these stakes provide. In terms of Albert Hirschman's dichotomy between exit and voice as monitoring mechanisms,45 thin capital markets eliminate exit as an available option in emerging economies. Investors therefore look to maximize their voice, and the self-enforcing model empowers them to do so. 1. Structural Constraints. -- The structural constraints that define the self-enforcing model operate both at the shareholder level and at the level of the board of directors. At the shareholder level, these constraints typically involve shareholder voting requirements or transactional rights (put and call options) triggered by specific corporate actions. 46 To the extent that an enabling statute contains any mandates for shareholder votes on particular types of transactions, it partakes of the self-enforcing approach. The differences between the self-enforcing model and an enabling model (with self-enforcement features) are ones of degree. The self-enforcing approach contains more and stricter voting mandates because it places greater weight on the goal of protecting outside investors against insider opportunism and lesser weight on maximizing business flexibility. 47 The existence in an enabling law of any mandatory appraisal rights can be seen as reflecting elements of the self-enforcing approach. A pure enabling law would let individual firms grant or withhold appraisal rights in their charters. 20 With regard to voting rules, a self-enforcing statute can require supermajority shareholder approval for central business decisions such as mergers, rather than the simple majority approval of the enabling approach. It can also require shareholder approval for a broader range of corporate actions than an enabling statute would, such as decisions to issue significant amounts of new equity or to purchase or sell major assets.46 For self-interested transactions between the company and its directors, officers, or large shareholders, a self- enforcing statute replaces the permissiveness of the enabling approach (loosely policed by courts) and the ban of the prohibitory model with approval by independent directors, a majority of noninterested shareholders, or both. The voting decisions of large shareholders, if obtained through fair voting procedures and with adequate disclosure, are a more fine-grained way to distinguish between good and bad transactions than substantive prohibitions could possibly be. To safeguard the voting mechanism, the self-enforcing statute can include a one share, one vote rule to prevent insiders from acquiring voting power disproportionate to their economic interest in the company, as well as procedures to ensure honest voting, such as use of an independent registrar to record share transfers, confidential voting, and independent vote tabulation. Outside shareholders' influence can be increased through use of a universal ballot that lets them cheaply place director nominations and other proposals on the voting agenda. Good voting decisions require good information, but the quality of shareholders' information can be improved by mandatory disclosure rules and by cumulative voting, which enhances large blockholders' access to information about the company. Voting mechanisms, which by their nature must be exercised collectively, can be supplemented with transactional rights that individual shareholders can exercise. A self- enforcing law can convey appraisal rights (put options) to unhappy shareholders for a broader range of corporate actions than a typical enabling statute would.47 Exercise of appraisal rights can be made simpler, and low-ball repurchase offers chilled, by requiring a company, when it solicits shareholder approval for an action that will trigger appraisal rights, to publish an offer price that will be binding on the company if the appraisal rights are exercised. The law can convey mandatory preemptive rights (call options) to acquire shares in proportion to one's ownership stake, as protection against underpriced stock issues. It can give shareholders takeout rights (put options) to sell their shares to a new controlling shareholder, as protection against transfer of control from known (and presumably trusted) hands to less trusted ones. 48 See, e.g., Del. Code Ann. tit. 8, § 271 (1991). 21 The voting rights and transactional rights approaches can be combined, with voting used to define the extent of the transactional rights. For example, preemptive rights can be made waivable ex ante by shareholder vote, and takeout rights can be made waivable by a majority vote of shareholders other than the new controlling shareholder. A self-enforcing statute also introduces structural constraints at the level of the board of directors. For example, it can require that a certain proportion of a company's board of directors be independent, and then vest these independent directors with authority over key corporate decisions -- such as approval of self-interested transactions. It can mandate board structures, such as an audit committee, that amplify the power of independent directors but are optional under enabling statutes. A critical feature of the self-enforcing model, linking shareholder level and board level constraints, is a cumulative voting rule for election of directors, buttressed by a mandatory minimum board size and a ban on staggered terms of office. Cumulative voting allows large outside shareholders to elect representatives to the board. As long as outside shareholders hold stakes large enough to elect their own representatives, cumulative voting enhances information flow and ensures that at least some directors will be true shareholder representatives. An insider majority will still control non-related-party transactions under this rule, but outside representation makes it harder for insiders to ignore or deceive minority shareholders. Although other voting rules, such as class voting, can also ensure minority representation on the board of directors, a cumulative voting rule is flexible enough to encompass a wide variety of ownership structures. 2. Simple, Bright-Line Rules and Strong Remedies. -- The self-enforcing model compensates for the weakness of formal enforcement through a combination of relatively simple, bright-line rules governing when its structural constraints apply, rules that insiders will often comply with voluntarily, and strong sanctions for violating the rules. The self-enforcing model might, for example, require a shareholder vote for a purchase or sale of assets that equals 50% or more of the book value of the firm's assets. To be sure, book value is an imperfect measure of a transaction's importance Moreover, a percentage threshold will be overinclusive in some cases (hindering transactions by requiring a shareholder vote without sufficient reason) and underinclusive in others (failing to reach transactions that could seriously affect shareholder value). But this rule is far clearer in application than the familiarenabling law requirement of shareholder approval for a sale of "substantially all" assets.48 When a pure bright-line rule is unavailable, the self-enforcing approach uses more concrete standards than are often found in developed markets. Compare, for example, the following alternative instructions to directors who must decide whether to approve a transaction between a company and one of its directors: 49 Most contemporary American corporation statutes encourage review of conflict-of-interest transactions by noninterested directors without specifying what standard of review these directors should employ. See id. § 144(a)(1) (stating that corporate transactions in which some directors are interested are not automatically voidable if approved by a majority of noninterested directors); Principles of Corporate Governance, supra note 13, § 5.02 reporter's note, at 235-45 (canvassing state statutes). 50 See, e.g., Principles of Corporate Governance, supra note 13, § 5.02(a)(2)(B) (stating that disinterested directors may authorize an interested transaction only if they "could reasonably have concluded that the transaction was fair to the corporation"). 22 (i) a self-interested transaction must be either ratified by shareholders or approved by the noninterested directors acting in the best interests of the company, or else is subject to "entire fairness" review by the courts; (ii) a transaction between the company and a director or top manager must be approved by noninterested directors, who should grant approval only if the transaction is fair to the company; (iii) a transaction between the company and a director must be approved by noninterested directors, who should grant approval only if the company receives consideration, in exchange for property or services delivered by the company, that is worth no less than the market value of the property or services, and the company pays consideration, in exchange for property or services, that does not exceed the market value of the property or services. The first approach, with some judicial gloss, is essentially the legal rule today in the United States and Great Britain.49 In an emerging economy, it offers meager guidance to directors. The second approach borrows from current best practice in the United States by vesting the decision in noninterested directors who are instructed to review the "fairness" of the transaction.50 In the United States, this best practice rests on a cultural understanding that "fairness" turns largely on price, relative to market price. But in an emerging economy, directors and judges may not know what it means for a transaction to be "fair." We favor the third approach in an emerging economy. Enforcement is still scarcely automatic, but even an unsophisticated judge can understand that the company's sale of property to a manager, who promptly resells it for thrice the price that he paid, was not at market value. A corrupt judge who nonetheless blesses such a transaction advertises his corruption to all. Sunshine is an imperfect disinfectant, but an important one nonetheless. The third approach also tells directors how they ought to behave. Over time, the norms that a company's transactions with insiders should be at market prices and should be reviewed by noninterested directors may become part of corporate culture. The cost of this more precise approach is that it fails to reach situations where a transaction, although at a "market" price, is nonetheless unfair to the company -- perhaps because the price was toward the low end of a broad range of plausible "market" prices. 51 See, e.g., Cinerama, Inc. v. Technicolor, Inc., 663 A.2d 1134, 1152-54 (Del. Ch. 1994), aff'd, 663 A.2d 1156 (Del. 1995). 52 In Russia, for example, privatized companies with more than 500 shareholders have been required to elect directors using cumulative voting since January 1994. See Decree of the President of the Russian Federation No. 2284, § 9.10 (Dec. 24, 1994), implementing The State Programme of Privatization of State-Owned and Municipal Enterprises in the Russian Federation, translation available in LEXIS, Intlaw Library, Rflaw File. Many companies haven't complied with this decree, we understand, because their managers don't understand how cumulative voting works. This confusion reduces their respect for the law as a whole. In response to this problem, our proposed Russian law mandated cumulative voting only for companies with 1000 or more shareholders -- the companies most likely to have outside blockholders who can make use of cumulative voting. 23 The prohibitive model also lends itself to bright-line rules, but transaction planners will exert constant pressure on these rules. Over time, this pressure will lead to fuzziness at the margin, as judges bend the substantive rules to allow beneficial transactions to proceed. The structural rules of the self-enforcing model will experience less pressure at the margin because they do not flatly bar transactions. Rather, the self-enforcing model merely imposes procedural hurdles that can usually be surmounted for beneficial transactions. The self-enforcing approach further encourages managers to comply with its rules through relatively severe sanctions for noncompliance, which compensate in part for the low probability of enforcement. For example, the remedy for failure to obtain advance approval of a self-interested transaction can be automatic forfeit to the company of the self-interested person's profit from the transaction. This contrasts with the enabling approach, in which an interested party generally can defend a transaction on the grounds that it was substantively fair.51 A statute can rely heavily on rules (rather than standards) and still be so complex that managers and judges can't understand it, and managers soon give up in disgust and stop trying. Every additional rule and nuance adds to the law's overall complexity and detracts from its overall effectiveness. This is a kind of externality -- the direct benefits from tailoring the law more closely to fit discrete situations must be balanced against the indirect costs of complexity.52 To combat this problem, a bias in favor of simplicity must overlay every discrete decision embedded in a self-enforcing law. C. The Limits to the Self-Enforcement Approach The self-enforcing model introduces three types of costs, compared to the enabling model. First, shareholder votes and transactional remedies add costs and delays. Thus, while these shareholder rights should be more common under the self-enforcement model than under the enabling model, deciding how much more common requires balancing, at the margin, the expected costs and benefits of expanding a particular protection. The cost-benefit balance, in turn, will depend on the strength of other institutions that are available in a particular country 53 See Ronald J. Gilson & Bernard S. Black, The Law and Finance of Corporate Acquisitions 641- 52 (2d ed. 1995). 24 to channel the behavior of corporate participants. There are no clear lines, only informed judgments, about where to strike that balance in a particular country. Shareholder-level protections are often more effective, but also more costly, than board-level protections. But the more effective the board is in serving shareholders' interests, the fewer the decisions that should require shareholder action. Thus, one goal of the self- enforcing model is to create a board that is more responsive to outside shareholders' interests, which in turn lets the law vest more decisions exclusively in the board. A second cost of giving a veto over corporate decisions to outside shareholders or outside directors, or requiring supermajority votes to approve certain decisions, arises from the usual hazards of departing from a majority vote rule.53 A large outside shareholder will have holdup power and may be able to obtain personal benefits by threatening to block a value- increasing transaction. Moreover, the rational apathy of small shareholders may make it hard for the company to obtain the votes needed for approval of a value-increasing transaction. The two concerns interact: the rational apathy of some shareholders increases the holdup power of other shareholders. To take a Russian example, suppose that a company's managers wish to merge with another company -- a transaction that will enhance the firm's value for investors but cost jobs today. A high shareholder approval requirement may let employee-shareholders block this transaction. Given the majority manager-employee ownership and relatively concentrated outside ownership of most Russian firms, we believe that approval of key corporate actions such as mergers by two-thirds of the outstanding shares offers a sensible balance between providing meaningful shareholder protection and limiting the holdup power of employees or outsiders. By contrast, a simple majority of outstanding shares should suffice to approve large share issues, because preemptive rights also protect shareholders against below-market pricing. But a different ownership structure would lead to a different judgment. A third cost of self-enforcement protections is a subtle loss of flexibility in designing the business enterprise. The self-enforcement model controls the structure within which corporate decisions are made. But a single decisionmaking structure will not fit all companies. To some extent, the law can allow for this by providing different rules for companies of different sizes and by dictating structure only when there seems strong need to do so. But we cannot anticipate in advance all the ways in which companies might want, for good reason, to depart from the prescribed structure. In theoretical terms, we cannot fully escape the usual expanded choice argument for an enabling law. Deciding which procedures are appropriate for which firms, and how finely to subdivide the universe of firms using an imperfect measure like number of shareholders, is an exercise in balancing. For Russia, we would apply the full "large company" procedures, such as cumulative voting, to firms with 1000 or more shareholders. This choice, however, depends 54 For Russia, we considered a "dual" corporate law, with one set of rules for privatized firms and another, more flexible set of rules for newly created firms, including firms financed by venture capitalists. We concluded that such a dual law was not feasible, because privatized firms would be able to use the tools of corporate restructuring (mergers, sales of substantially all assets, spin-offs, and the like) to qualify as "new" firms. We did not think a workable line could be drawn between "genuine" restructurings and "sham" restructurings designed to qualify for the more liberal corporate law rules. The American tax rules on net operating losses, through which Congress tried for decades in increasingly complex ways to limit the use of net operating loss carryforwards to the firm that generated the losses, offer a case study where regulators tried to draw a similar line and failed. See Boris I. Bittker & James S. Eustice, Federal Income Taxation of Corporations and Shareholders § 16.02 (5th ed. 1987) (reviewing the history of the net operating loss rules). 25 on a background fact: the Russian privatization process, in which almost all employees and many individual citizens became shareholders, means that even a relatively small company can have hundreds of shareholders. Venture capitalists (active equity investors -- domestic or foreign -- who demand influence or even control in exchange for large infusions of capital) are likely to be important sources of equity capital in emerging economies, where rapidly changing economic circumstances make almost all companies highly risky. Venture capitalists in developed economies exploit the enabling model's flexibility to tailor elaborate control arrangements for their own protection. A good measure of the flexibility costs of the self-enforcing model is how often its mandatory procedures prevent arrangements that these investors might otherwise prefer.54 Two features of our proposal may seem especially likely to interfere with venture capital investments: (i) the takeout offer requirement that a purchaser of over 30% of a company's common stock offer to buy all remaining shares; and (ii) the requirement that each company have a single class of voting common stock. In fact, these provisions are unlikely to block the control arrangements that venture capitalists often demand. Suppose, for example, that a venture capitalist wants to buy a 40% interest in a company directly from the company. Our control transaction rules permit shareholders to waive their takeout rights by majority vote. A waiver imposes virtually no added cost on the company, because the shareholders must vote to authorize a 40% share issuance in any event. If the venture capitalist wishes to buy a 40% stake from existing shareholders, rather than from the company, then a waiver vote would entail the extra transaction costs of a shareholder meeting. In our judgment, these costs are justified by the protection that the takeout rule accords to minority shareholders. A more difficult but still soluble problem arises if the venture capitalist wishes to make a minority investment while retaining veto rights over critical transactions. In the enabling model, such a transaction is often structured by creating a second class of voting stock that gives the investor both general voting rights and the particular veto rights that are desired. Under the self-enforcing model, the same control structure can be created by combining partial ownership of a firm's common stock with ownership of a class of preferred stock equipped with the desired veto rights. Virtually the only rights that a venture capitalist could not demand under our proposed statute are voting rights for directors disproportionate to the venture 55 See Robert C. Ellickson, Order Without Law 40-64, 184-206 (1991). 56 Clear legal rules can enhance reputational sanctions by increasing the visibility of misconduct. Behavior that might otherwise blend into routine, hard-nosed business practice is easier to identify as illicit against the backdrop of a legal rule that prohibits the behavior. Thus, corporate law in emerging markets can help to seed the development of the nonlegal controls that are so critical to the functioning of developed markets. See David Charny, Nonlegal Sanctions in Commercial Relationships, 104 Harv. L. Rev. 375, 408-25 (1990) (describing the prerequisites for reputational enforcement). 26 capitalist's total equity investment. And even this restraint can sometimes be finessed through a voting agreement with other large shareholders. D. Can Law Function Without Official Enforcement? The self-enforcement model is designed to minimize reliance on official enforcement. But how effective can the law be if judicial enforcement is as weak, corruption as widespread, and organized crime as strong as is currently the case in Russia? Are all efforts to control private behavior within the corporate form doomed to failure without some minimum level of enforcement capability? To explore these questions, let us imagine, counterfactually, a world with no official enforcement -- no government organ that can enforce the official corporate rules. If the self- enforcing approach can elicit partial compliance in this hypothetical world, then it can only work better -- though still imperfectly -- in the more realistic case where official enforcement is weak but not wholly absent. The concept of rules without official enforcement is not new. Robert Ellickson has explored situations in which norms of conduct emerge without official enforcement, including conflict between whaling vessels in international waters, and situations where official rules are so out of touch with practical needs that a public consensus develops around informal norms.55 Here, we consider the potential for rules to function without official enforcement in the specific context of corporate law. For example, suppose that company directors can ignore all shareholder voting rules without fear of official intervention. What recourse is open to a 20% shareholder who loses a board seat because the company erases him from the shareholder register, refuses to provide cumulative voting, or conveniently loses his ballot? One answer is that the question is posed too starkly. Even without official sanctions, many companies will not resort to such tactics. Some managers will comply with the written law simply because it is both written and reasonable; some will comply because their peers do; others will comply so as not to risk embarrassing news stories. Companies that need capital will comply with the rules to build a reputation for honest behavior, and companies that plan to enter long-term contractual relations must safeguard their reputation for honesty and fair dealing.56 Without official enforcement, the entire corporate law (including its nominally "mandatory" terms) becomes a set of default rules from which the participants in the corporate 57 See Jonathan Hay, Private Enforcement of Law (1996) (unpublished manuscript, on file with the Harvard Law School Library). 58 From this perspective, the Krasnoyarsk Aluminum case, discussed above in note 31, where a company wiped a 20% shareholder out of its share register, can be seen as the exception, not the norm. During the year before the company's action, someone had waged an intimidation campaign against Krasnoyarsk's managers, killing several senior managers and beating up others, including the CEO. That someone, some informed observers believe (and Krasnoyarsk's managers surely knew, one way or another), was the 20% shareholder, who already controlled the two other large aluminum refineries in Russia and wanted control of the third -- Krasnoyarsk. If so, then Krasnoyarsk's managers were responding -- extralegally but perhaps appropriately -- to an extralegal effort to take control of their firm. 59 In Russia, organized extralegal enforcement of norms of business conduct already occurs to some extent. Small businessmen, each "protected" by different mafia gangs, sometimes bring disputes to mafia-run "courts," 27 enterprise can depart, jointly or unilaterally. But, as with any set of default rules, it will be costly to ignore them or contract around them. There may also be penalties harsher than loss of reputation for breaking the rules. A world without official enforcement will surely develop unofficial enforcement. Suppose, plausibly, that some shareholders are willing and able to resort to violence if their rights are violated, and that company directors are not sure which shareholders may react this way. Fear of these few can cause directors to behave properly toward all shareholders.57 Few shareholders are likely to shoot a director just for making a bad business decision. That would be foolish: it is unlikely to encourage better decisions in the future and risks retaliation in kind. But the situation is very different if the directors break a clear rule conferring voting rights. Now a wrong has been committed, and directors will face personal liability of a tangible kind if some shareholders feel they must respond. Thus, few directors will blatantly disregard the written law. At the very least, they will take seriously a 20% shareholder's demand for board representation by weighing the personal risks from rejecting the demand against the benefits, much as if official enforcement of shareholder rights were in prospect. The more blatantly shareholder rights are violated, the more likely a shareholder is to take extralegal action, and thus the greater the expected sanction will be. For example, removing a shareholder from the share register is more likely to provoke a violent response than refusing to use the required procedures for approving a self-interested transaction.58 We do not suggest that unofficial enforcement is anything near ideal. Some directors will be shot for imagined wrongs or, as at Krasnoyarsk Aluminum, as part of a quite literal takeover "battle." Shareholders, too, will be at risk if they complain too loudly when a company is looted. Large shareholders may succeed in extorting private benefits from companies when these can be hidden from other shareholders. But on the whole, men with guns will often be polite to each other, especially if, as is common for corporate enterprises, they expect to meet again. Repeated interaction magnifies both the importance of reputation and the risk of retaliation for misbehavior. Unofficial yet still organized enforcement may arise and coalesce around the written law.59 In sum, a corporate law that defines norms of politeness called "razborkas." (In Russian, razborka is a noun meaning "sorting out.") Only the very foolish do not comply with a razborka's decision. See, Michael Specter, Survival of the Fittest, N.Y. Times, Dec. 17, 1995, Magazine, at 66, 69-71. 60 Cumulative voting in Russia offers an example of how law can take hold without official enforcement. Since December 1993, privatized Russian companies with 500 or more shareholders have been required to elect directors through cumulative voting. See Decree of the President of the Russian Federation No. 2284, supra note 56, § 9.10. In a developed country, one could expect near universal compliance with such a requirement. In Russia, companies could not be forced to comply with this requirement, and initial compliance was low, partly because managers didn't understand how cumulative voting was supposed to work. In a survey of 40 privatized firms conducted in early 1994, only 15% had implemented cumulative voting or even planned to do so within two years. See Blasi & Shleifer, supra note 16, at 83. By late 1995, actual compliance had climbed to 16% of all privatized firms, including 26% of firms with less than majority ownership by managers and employees. Compliance with the spirit of cumulative voting was somewhat higher, because some firms that did not formally use cumulative voting had voluntarily ceded one or more board seats to outside blockholders. The late 1995 data is from a survey by Professor Joseph Blasi of Rutgers University. 28 in ways that the participants perceive as reasonable can be partially effective even without official enforcement.60 An example: our model law limits the right of insolvent companies to distribute assets to shareholders or to sell assets for less than equivalent value. In a developed country, these "fraudulent conveyance" rules are enforced by courts and bankruptcy trustees, who chase and often retrieve improperly transferred assets. In Russia, creditors cannot rely on this kind of official enforcement. But there is a substitute already in place: creditors often threaten, and not infrequently shoot, debtors who haven't repaid their debts. The corporate law can help private actors to distinguish situations when an asset distribution was proper, even though the business later couldn't pay its debts, from situations when the asset distribution was improper when made. If the rules of conduct are clearer, borrowers with good investment opportunities will be more willing to risk borrowing money to finance these investments. The effective cost of capital will decline. Once we reinstate the possibility of some recourse to courts, even corrupt courts, the self-enforcing model's effectiveness quickly increases. A corrupt judge can twist a "reasonableness" standard to reach the decision he was paid to reach, but cannot so easily twist a requirement that the company provide cumulative voting or appraisal rights. If the judge finds an exception on some spurious ground, it will be obvious to all. Yet few corrupt officials want to admit their corruption in public. Such a judge will also risk personal retaliation, much as corporate managers do, at the hands of private enforcers. Moreover, over the longer term, blatant violation of reasonable norms can create a constituency for enforcement. Shareholders will bring political pressure to strengthen enforcement capability and will have obvious abuses to point to. News stories will highlight scandals, bringing further pressure for enforcement. Test cases, even if they fail in corrupt or incompetent courts, will form a basis for public opinion -- and repeat players in financial markets may be willing to underwrite the costs of such test cases. 29 Finally, even if official enforcement is weak today, it may be stronger tomorrow. The prospect that future enforcement may threaten their gains will make corporate actors reluctant to presume nonenforcement, especially if the official sanctions, if imposed, are likely to be severe. For example, managers may prefer to seek shareholder approval of a self-interested transaction if they believe that approval is likely, rather than run even a small risk of facing a future lawsuit to disgorge profit or have the transaction unwound. In short, the claim that corporate law can do much to shape private behavior even under conditions of weak official enforcement is not as strange as it may first appear to be. And even law with no official enforcement is not an oxymoron. III. Governance Structure and Voting Rules This Part and the next two Parts of this Article describe and justify in greater detail the elements of the self-enforcing model, with particular reference to our proposal for Russia. In many cases, there are no clear lines, only informed judgments, concerning how much shareholder protection, and what forms of protection, are optimal in the Russian environment. This Part discusses overall governance structure; Part IV considers the voting and transactional rights that attach to particular corporate actions; and Part V discusses remedies. The self-enforcing approach to corporate law constrains the discretion of managers and majority shareholders by granting voice and sometimes veto rights over corporate actions to outside directors, non-controlling shareholders, or both, in the expectation that these actors can best determine whether proposed corporate actions are value-increasing for the enterprise or merely wealth transfers to insiders. To avoid manipulation of the board and shareholder voting mechanisms, governance structure must be simple, and malleable only within narrow limits. Some of the enabling model's flexibility with regard to governance, voting processes, and capital structure is sacrificed to protect investors while simultaneously preserving flexibility over the range of substantive actions a company may take. A. Allocation of Decisionmaking Power There are two broad strategies available in choosing a review process for corporate actions: representative democracy, under which shareholders elect representatives (a board of directors) to act on the shareholders' behalf; and direct democracy, under which shareholders directly approve particular actions. Representative democracy alone is often unsatisfactory because boards can too easily become lazy or be captured by management. Thus, the company laws of all developed countries provide direct shareholder review of selected corporate actions such as mergers. On the other hand, direct democracy is far too slow and costly for most corporate decisionmaking. Moreover, because small shareholders must act on limited information and face severe collective action problems, direct democracy can quickly deteriorate into total manager control in widely-held companies. 61 The two-thirds-of-outstanding-shares voting requirement for charter amendments, and the specific vote requirements for other corporate actions discussed below, are minimums. The charter can prescribe a higher but not a lower voting requirement. 62 Our proposed structure has enough flexibility to allow a company largely to replicate the two-tier management structure if the board so chooses or the charter specifies. The "board of directors" can hire a "board of managers" and delegate to it day-to-day management responsibility. The board of managers will then be subject only to whatever oversight the board of directors chooses to exercise, except when the law or the charter requires approval by the board of directors. However, the board of directors remains responsible to the shareholders for the consequences of this choice; the board cannot blame a legal structure that limits its power over management. 63 For mechanical reasons, a system with cumulative voting must restrict shareholder power to remove individual directors, but not the entire board. If directors could be removed individually, a majority shareholder 30 We mediate between the weaknesses of each approach with a simple hierarchical governance structure that allocates managerial power to a board of directors, subject to shareholder review of particular actions. The shareholders elect the board of directors; the board chooses the managers (subject to shareholder review of its choice of top manager); and the board (sometimes a defined subset of the board) approves particular types of actions, including those that require shareholder approval. For all other actions, the board decides when the managers can act unilaterally and when they need board approval. Governance rules, within the range left open by the law, are specified in a company charter. To protect minority shareholders against changes in governance rules, charter amendments require approval by two-thirds of the outstanding shares, and appraisal rights (discussed in Part IV) attach if a charter amendment reduces the rights of a class of outstanding shares.61 To enhance shareholder control of a firm's governance rules, charter amendments do not require board approval. This governance structure has multiple advantages. First, it ensures a measure of accountability; shareholders know whom to blame if things go wrong. Second, it provides the board with reasonable though not total flexibility. The board decides how best to use its own limited time, but it must make enough business decisions to avoid descending into ill-informed irrelevancy -- a strong risk in the German two-tier board model, in which the supervisory board meets rarely and does little more than choose management.62 Third, the structure provides double review, by both the board and the shareholders, of important or suspect transactions. Fourth, this structure does not require more of shareholders than they can deliver. Apart from choosing the board of directors, shareholders generally review actions that have been proposed by the board, rather than make decisions unilaterally. This process accords with the limited information available to most shareholders. Our proposed structure gives broad power to the board of directors. But this power is constrained, in turn, by granting shareholders broad power over the board's constituency. Shareholders elect directors annually through cumulative voting and retain the authority to remove the board (as a whole) at any time without cause.63 Moreover, the path toward could vote to remove a director elected cumulatively by a minority shareholder, and thus nullify the effect of cumulative voting. Cf. Del. Code Ann. tit. 8, § 141(k)(i) (1991) (providing that cumulatively elected directors may not be removed individually if votes against removal would be sufficient to elect them). 64 To avoid evasion of this rule, the law must limit the voting rights of securities (preferred stock and debt) that are senior to a company's common stock, and limit the company's ability to issue nonvoting securities, principally options to purchase common stock, that are equivalent or junior in priority to common stock. In our model, preferred shareholders must approve a charter amendment that reduces the rights of a class of preferred stock, and the charter can give preferred shareholders the right to vote (as a class or together with the common stock) on specific corporate actions such as mergers. But the charter can allow preferred shareholders to vote for directors only in two limited cases: (i) if preferred dividends are in arrears; or (ii) if the preferred stock is convertible into common stock. In these cases, preferred stock can be given a number of votes equal to the number of common shares into which it is convertible. 65 In the United States, a one share, one vote rule is maintained by agreement among the principal stock exchanges rather than by company law. In Britain, one share, one vote is essentially universal because of strong support from institutional investors, who refuse to buy the shares of a company with a different rule. See Black & Coffee, supra note 6, at 2024. The one share, one vote rule is expressly mandated by statute in 9 of the 17 jurisdictions examined in our survey of company laws in emerging markets. See infra Appendix. 66 For aspects of the extended American debate over the one share, one vote rule, see Ronald J. Gilson, Evaluating Dual Class Common Stock: The Relevance of Substitutes, 73 Va. L. Rev. 807 (1987); Jeffrey N. Gordon, Ties That Bond: Dual Class Common Stock and the Problem of Shareholder Choice, 76 Cal. L. Rev. 31 effective shareholder use of this power is smoothed in various ways, including restricting companies to a single class of voting securities (that carry a residual interest in the company's profits) and facilitating shareholder nomination of director candidates. B. Allocation of Voting Power: One Share, One Vote The self-enforcing statute allocates investor voting power in proportion to economic interest by mandating a single class of voting common stock that has both a residual interest in corporate profits and one vote per share. This allocation increases the likelihood that corporate actions will maximize firm value.64 The one share, one vote principle is widely accepted across jurisdictions. It is the dominant rule in the United States, Great Britain, and Japan even when it is not a statutory requirement, and it is mandated by statute in many emerging market jurisdictions.65 Moreover, non-voting or low-voting stock has come under strong criticism from large investors in countries like Germany, where it has been common. The case for the one share, one vote rule turns primarily on its ability to match economic incentives with voting power and to preserve the market for corporate control as a check on bad management. By contrast, the case for permitting companies to deviate from a one share, one vote rule turns on (i) the usual claim that informed parties will choose optimal arrangements on their own; and (ii) the existence of a reasonably efficient market, in which the proceeds that company founders realize when they sell their shares will reflect the voting rights that those shares carry.66 1 (1988); Louis Lowenstein, Shareholder Voting Rights: A Response to SEC Rule 19c-4 and to Professor Gilson, 89 Colum. L. Rev. 979 (1989). 67 For discussion of the special problems created by midstream charter changes, see Bebchuk, Limiting Contractual Freedom, cited above in note 23, at 1825-59; and Gordon, Mandatory Structure, cited above in note 23, at 1573-85. 68 See, e.g., Russian Capitalism, Economist, Oct. 8, 1994, at 21, 23. 69 For example, a mandatory one share, one vote rule protects shareholders in companies that initially issue only one class of voting stock, by preventing a subsequent charter amendment that changes the rule. 32 The arguments against a one share, one vote rule lose much of their force in emerging markets for several reasons. First, public offerings in those markets are unlikely to be priced with a high degree of efficiency. Second, in newly privatized economies, including Russia, there were no true founders who could make an economic decision to sell control rights together with economic rights. Instead, one faces in such cases the more troubling prospect of midstream charter changes, perhaps coerced in various ways by managers who want to cement their control without paying significant economic value.67 Third, in emerging markets there is more need for investor protection against self-dealing by company managers. In Russia, for example, stories abound about company managers who sell most of the firm's output to another company but never collect the accounts receivable -- presumably in exchange for payments to these managers' foreign bank accounts -- while not paying the company's rent, taxes, utilities, or even employees for months or years.68 Voting common shares are no panacea for this behavior, but they can help, especially when they are acquired by large outside shareholders. From a theoretical perspective, control (like other assets) tends to move to those who value it most. Multiple-class voting structures create incentives for control to move from good hands to bad because those who are willing to abuse control will often value it more than those who will not. In developed economies, market and cultural constraints are often strong enough to keep abuse at manageable levels. In emerging markets, however, abuse will proliferate. To draw an analogy to ordinary product markets, when product quality is difficult for buyers to measure (the "lemons" situation), minimum quality rules can be welfare-enhancing. The case for minimum quality rules in securities markets is especially strong because of the risk (largely unique to securities markets) that the quality of what one has bought will be changed after the date of purchase, and the incentive for unscrupulous insiders to profit by doing precisely that.69 C. Voting for Directors: Cumulative Voting The one share, one vote rule is conventional in many jurisdictions and has easily understood virtues. Our approach to voting for directors, however, relies on a less common 70 Under cumulative voting, if a board includes n members elected annually, a shareholder who holds 1/n of the votes can elect one director. See Clark, supra note 42, § 9.1 at 361-66. Staggered board terms and small board sizes dilute the effectiveness of cumulative voting because they reduce the number of directors elected at one time. We would require a firm with 1000 shareholders to have at least seven directors, and a firm with 10,000 shareholders to have at least nine directors. 71 See Ronald J. Gilson & Reinier Kraakman, Reinventing the Outside Director: An Agenda for Institutional Investors, 43 Stan. L. Rev. 863, 872-76 (1991). 72 See Blasi & Shleifer, supra note 16, at 81. 33 solution: mandatory cumulative voting, together with related requirements for minimum board size and annual election of directors.70 By giving large minority shareholders a place on the board and a voice in board actions, cumulative voting addresses several problems at once. First, a board seat provides access to company information and gives large shareholders a substitute for the disclosure that is provided in developed economies in other ways, such as financial disclosure rules, financial press reports, market price signals, and non-public financial reports to lead banks in Germany and Japan. Second, cumulative voting makes it more likely that a minority of directors is truly independent of management and -- also important though often neglected in the United States -- that these directors will owe affirmative loyalty to the shareholders who elect them.71 Director independence interacts with rules, discussed below, that vest in independent directors the power both to review transactions in which managers have a personal financial stake and to choose the company's auditors and share registrar (who also counts shareholder votes). Third, cumulative voting reinforces the principle that directors owe their loyalty to shareholders, not to the company's officers. The presence of some outside directors who truly represent shareholder interests can, over time, influence how all directors understand their role in the corporate enterprise. Thus, cumulative voting is part of a broader effort in the self- enforcing model to promote behavioral norms that have served developed countries well. Because cumulative voting serves these multiple purposes, it is a central element of a self-enforcing corporate law. We are not so sanguine as to rely on the board of directors as a whole, or even the board's nominally independent directors, as the sole protectors of outside shareholder interests; we also contemplate a broad range of transactions for which shareholders have veto power or other protections. But cumulative voting can strengthen the boards of many companies. Russian data suggest that large outside blockholders both want and expect to receive board seats. Thus, many investors will make use of the power to elect their own representatives if it is available.72 Moreover, cumulative voting can have an effect even when management's slate is the only one proposed. Its availability may determine whom management puts on its slate, or deter management from actions that could provoke an outside shareholder to nominate its own slate. 73 Cumulative voting can be adopted by charter provision under the corporate laws of almost all developed countries and many emerging market jurisdictions. None of the 17 jurisdictions examined in the Appendix prohibits cumulative voting. Straight voting is the default rule in most cases, but two jurisdictions (Mexico and Chile) mandate forms of proportional representation akin to cumulative voting. Note, too, that mandatory cumulative voting was common in the United States until the 1950s. See Jeffrey N. Gordon, Institutions as Relational Investors: A New Look at Cumulative Voting, 94 Colum. L. Rev. 124, 142-46 (1994) [hereinafter Gordon, Institutions as Relational Investors]. Mandatory cumulative voting was subsequently rolled back in the United States. But Professor Gordon attributes the decline of this system primarily to the political power of incumbent managers who wanted to make proxy contests more difficult. See id. at 153. 74 Developed countries that do not rely heavily on shareholder-nominated directors have often developed other oversight mechanisms. For example, institutional shareholders in Great Britain often lack direct board representation, but British managers know that a modest number of institutional investors can combine forces to oust the board if the need arises. British institutional investors are also pressing for boards to include more independent directors, to supplement the crisis-oriented oversight that they now exercise themselves. See Black & Coffee, supra note 6, at 2037-38. Similarly, in Japan, large shareholders can act through the main bank to force a change in management, much as American shareholders might act through an independent board of directors. See Roe, Some Differences, supra note 3, at 1943-46. 75 See Sanjai Bhagat & James A. Brickley, Cumulative Voting: The Value of Minority Shareholder Voting Rights, 27 J.L. & Econ. 339, 341-42 (1984). 34 The fact that many companies choose not to adopt cumulative voting in both developed and advanced emerging economies does not detract from its value as a cornerstone of a self- enforcing law.73 In most of these countries, market, legal, and cultural forces combine to achieve the goals that cumulative voting can help to attain. Consider outside shareholder representation on the board of directors. In the United States, large outside shareholders can often obtain representation on the board of directors in rough proportion to their ownership interest, if they want such representation. Although company managers might be able to defeat the large shareholder's nominees in an election contest, the managers will often offer the shareholder a couple of board seats rather than risk losing the contest, which could mean losing their jobs. Many companies also offer board representation to large investors to induce them to invest. Even companies with no large investors typically appoint a majority of independent directors to the board, simply because it is customary to do so. These directors can then perform some of the oversight that would be undertaken by directors chosen by large shareholders under cumulative voting.74 The principal argument raised against cumulative voting is that a divided board may be less effective than a board elected on a winner-take-all basis. But this risk is not likely to be large in practice. Experience in developed countries suggests that proportional representation of large shareholders on the board, whether through cumulative voting or agreement with management, usually works well. Empirical studies undertaken in developed countries also suggest that shareholders benefit from the availability of cumulative voting.75 Moreover, logic suggests that large shareholders rarely have an interest in interfering with the smooth functioning of the board. When a large shareholder does attempt to interfere, it is often a sign 76 Cf. Gordon, Institutions as Relational Investors, supra note 73, at 170-74 (describing the value of cumulative voting in overcoming the rational apathy of institutional investors). 77 The size threshold is intended to weed out nuisance candidates and proposals. The 2% of outstanding shares threshold that we suggest for Russia seems large enough to accomplish this without blocking shareholder proposals or nominations that have a realistic chance of success. 35 of some pathology. Cumulative voting is an incomplete cure for the pathology, but it is better than the alternative: a unified board stolidly supporting management as the company slides toward disaster.76 Our proposed Russian law also requires that large-company boards contain at least one- third independent directors (defined as directors who are not, and during the last five years have not been, officers of the company and who are not related to a company officer). Of course, many of these directors will be independent only in name. But even independent directors chosen by management may sometimes provide a voice against management self- dealing, especially if a truly independent director or two, perhaps elected through cumulative voting, can take the lead role in questioning a management proposal. And the independent- director requirement can reinforce the norm that the board of directors is in substantial part a watchdog institution, charged with monitoring management on the shareholders' behalf. D. Voting Procedures: Universal Ballot One share, one vote and cumulative voting are only part of the architecture of a voting system for emerging economies. Ancillary rules are also needed to govern the form of shareholder proxies (or ballots), how shareholders can nominate candidates for election to the board or introduce other proposals for shareholder vote, and how shareholder votes are counted. We adopt a "universal ballot" as the framework for nominating and choosing among directorial candidates, and for introducing and voting on other proposals. The laws of developed countries typically require each faction in a proxy contest to distribute its own ballots. By contrast, the universal ballot lists all candidates on a single consolidated ballot prepared at company expense and made available to shareholders well before the shareholder meeting. Under this regime, the incumbent board, and all shareholder groups exceeding a threshold size (for Russia, we use a threshold of 2% of the outstanding shares), may nominate candidates on the company's ballot. Shareholder groups exceeding this threshold size may also include other proposals for a shareholder vote (for Russia, we allow a maximum of two proposals), with no restrictions on subject matter.77 Like cumulative voting, these liberal provisions for including shareholder nominations and proposals on the company's ballot permit relatively small aggregations of shares to participate in shaping the company's voting agenda. Shareholder groups of a somewhat larger threshold size should also have the power to convene special shareholder meetings. For Russia, 78 In Russia, independent share registrars are licensed by the Securities Commission, which can revoke a registrar's license, thus putting it out of business, if the registrar misbehaves. 36 we use a threshold of 10% of the outstanding shares, which reflects both the expense of holding a shareholder meeting and the fact that most decisions at such a meeting (such as a decision to oust the board of directors) would require at least a majority vote -- compared to the 10-15% necessary to elect a director under cumulative voting. Majority approval is unlikely unless a proposal has substantial backing from the outset. E. Protecting Honesty and Quality in Voting The best possible voting procedures are useless if they are subverted by coercion, vote buying, or fraud in counting ballots -- chronic dangers in emerging economies. Coercion and vote buying occur when someone -- typically a company insider -- induces shareholders to vote against their investment interests by punishing "wrong" votes, rewarding "right" ones, or both. In Russia, coerced voting of employee shares is a particular danger because management can often use its workplace authority to control how employees vote. The first defense against coercion and vote buying is a mandatory rule of confidential voting. Insiders who cannot monitor shareholder votes lose the power to manipulate votes through rewards or sanctions. To protect against fraud and secure confidentiality, we take the functions of collecting, tabulating, and storing shareholder ballots out of management's hands. The Russian statute vests the tabulation function in an independent share registrar -- a position large companies are already required to maintain for the separate purpose of reliably recording share transactions. These protections won't always work, but they will make cheating more difficult. It is much harder to police a company's count of its own ballots than to determine whether the company has an independent registrar. Relatively few independent registrars are likely to exist because this business has large economies of scale. These few can be monitored through licensing requirements, and will have a strong reputational interest in counting votes honestly.78 Moreover, some cheating will be so obvious that recourse to the courts or to unofficial enforcement will be feasible. For example, if a company has seven directors, a 15% shareholder has the power to elect one director under cumulative voting. If the shareholder seeks to exercise this power, an attempt by the company to subvert cumulative voting will be easily provable. In many Russian companies, he who votes the employees' shares controls the company. This fact was not lost on company managers, who quickly developed ways to control how employee shares are voted. Confidential voting and independent tabulation help to address this problem because managers who do not know how an employee has voted cannot punish her for a vote against management. But more is needed, lest managers resort to other coercive techniques. For example, some Russian managers have forced or convinced employees to 79 See Blasi & Shleifer, supra note 16, at 101. 80 These employee protection provisions were dropped from our proposed statute early in the Russian legislative process. Our best interpretation of the political dynamics is that company managers opposed the provisions (for obvious reasons), employees were silent (of course), large investors cared more about other provisions of the law that affected them more directly, and Communists (who we hoped would support these "pro-worker" provisions) either did not understand why these fairly technical provisions were important or else were more pro-manager than pro-employee (a not unlikely explanation based on the Communists' votes on other issues). 37 transfer their shares to a long-term trust, voted by the managers, from which shares cannot be withdrawn.79 In developed countries, there are often statutory time limits (ten years, say) on voting trusts and similar arrangements that separate the economic interest in stock from voting power. We considered these limits inadequate for Russia -- both because there is currently no trust law to impose fiduciary duties on managers who establish employee trusts, and because most employees cannot make an informed choice about joining a trust. They are told, and they believe, that they must put their shares into a trust to be managed on behalf of the "labor collective." For Russia, we did not propose a ban on employee trusts and similar devices for pooling employee shares because this would have foreclosed the possibility of a labor union-controlled trust that might serve as a counterweight to management. Instead, we proposed several less extreme rules. First, managers should not be allowed to control any employee trust. Second, the maximum duration of any such trust should be short (we proposed two years) in order to give employees frequent opportunities to opt out of participation. Third, employees should have the right at any time to sell shares that have been placed in an employee trust. Finally, company managers should not ask an employee how she has voted or whether she has sold or plans to sell her shares, nor retaliate against an employee because of a voting or sale decision.80 These rules are not easy to enforce, but some managers will honor them voluntarily if the law labels efforts to control employee votes as improper. Moreover, once the principle that employees should vote their own shares is established, the press can expose violations. This can help even weak labor unions protect their members against overt retaliation. The value of shareholder voting as a check on management discretion depends on the shareholders having good information on which to base their voting decisions. Yet many emerging markets are also characterized by limited financial and other corporate disclosure. We have only partial answers to the problem of poor disclosure. Requiring mandatory disclosure to shareholders, including audited financial statements, can help, but only so far. In Russia, many companies will hide their true accounts from their auditors -- and shareholders will not protest too loudly, because profits hidden from the auditor are also hidden from the tax collector and the mafia. Cumulative voting also helps, by improving large outside shareholders' access to information. If a large outside shareholder can be trusted not to self-deal (in Russia, this is sometimes the case and sometimes not), then a proposal that has been approved by the 81 All of the emerging market jurisdictions in our survey require shareholder approval for mergers, recapitalizations, and dissolutions. See infra Appendix. In most cases, the voting rules specify supermajority quorum and approval thresholds. These approval requirements, however, do not usually extend to major sales or purchases of assets. 82 Eleven of the 17 emerging market jurisdictions in our survey provide some form of appraisal rights for shareholders who dissent from mergers. See infra Appendix. 38 director(s) who represent the large outside shareholder carries a badge of quality, on which smaller shareholders can rely. And shareholders can always vote against a management proposal if they distrust the information that management has provided. IV. Structural Constraints on Particular Corporate Actions In any corporate law, the basic governance structure and voting rules discussed in Part III are the sole legal backdrop only for routine business transactions. Very large, suspect, or potentially transformative transactions are also subject to specialized regulation designed to protect outside shareholders from correspondingly high risks of abuse. The self-enforcement approach regulates these transactions through structural constraints rather than prohibitions. The constraints cover four categories of transactions: mergers and similar major transactions, self-interested transactions, control transactions, and issuances and repurchases of shares. For all categories except control transactions, we require approval (often by supermajority vote) by both the board and shareholders. For all categories except self-interested transactions and share repurchases, we also give transactional rights to shareholders who do not support the corporate action. A. Mergers and Other Major Transactions Mergers, large sales and acquisitions of assets (whether directly or indirectly through a subsidiary), reorganizations, and liquidations are essential tools for restructuring companies. However, they can also radically alter the nature of a shareholder's investment, and they have historically been a common means by which insiders can loot a company's assets. To respond to this danger, even enabling laws commonly require shareholder approval for selected transformative actions.81 Enabling statutes also often provide a transactional mechanism for shareholder protection: appraisal rights that let shareholders demand payment of the fair value of their shares, as determined by a court, instead of accepting the consequences of a transformative corporate action.82 The list of transactions that require shareholder approval and appraisal rights in developed countries, and the size of the required shareholder vote, should form a floor for the corporate law of an emerging economy. For a self-enforcing statute, the key decisions are: (i) when to require a higher shareholder vote than majority approval; (ii) what additional transactions should require shareholder approval; and (iii) when shareholders should have 83 For Russia, we proposed an exception for investment funds, which (i) typically have a huge number of tiny shareholders, and (ii) lack the substantial employee ownership that characterizes privatized firms. The first factor makes a two-thirds vote of outstanding shares harder to achieve; the second makes it less important as a protection against management overreaching. For investment funds, we judged that a simple majority of outstanding shares should suffice. 84 There should be a "normal course of business" exception to handle the special case of a trading company that regularly makes large purchases and sales of goods on a thin equity base. In addition, the definition of asset "sales" should exclude a pledge of assets to secure a loan. Such a pledge, followed by intentional default on the loan, can be used as an indirect way to sell assets without a shareholder vote. But most pledges are likely to be legitimate, and often a default under one loan agreement will trigger adverse consequences under other loan agreements due to cross-default provisions. We believe that the flexibility lost by treating a pledge of security for a loan in the same manner as a sale exceeds the gain in additional protection of shareholders against sales 39 appraisal rights (and what kind of rights) if the company completes a transaction that they oppose. 1. Shareholder Approval. -- The appropriate shareholder approval threshold depends on the ownership structure of public companies. The typical ownership structure of privatized Russian enterprises (described in Part I) leads us to propose approval of major transactions by two-thirds of the outstanding shares. This threshold is high enough so that managers and employees cannot routinely complete major transactions without support from outside shareholders, yet not so high that companies will often be unable to complete beneficial transactions because the necessary shareholder vote cannot be obtained, nor so high that it gives undue holdup power to outside blockholders (or to employees).83 Deciding which transactions should require a shareholder vote requires lawmakers to balance flexibility against protection of minority shareholders. A shareholder vote to approve a transaction is costly in both time and money -- managers must either call a special shareholder meeting or wait until the next regular meeting to ask for shareholder approval. This will increase the cost of completing beneficial transactions and will deter some transactions entirely. Thus, a shareholder vote should be required only for transactions that transform the nature of the company or involve a substantial risk of abuse. For Russia, the transactions that we believed should require a shareholder vote were: (i) a merger or other business combination between the company and one or more other companies; (ii) a liquidation of the company; (iii) a transformation of the company into another type of legal entity, such as a partnership; and (iv) a sale or purchase of assets, directly or through subsidiaries, for a price equal to 50% or more of the book value of the company's assets.84 for less than fair value. 85 For example, American corporate law imposes no restrictions on the purchase of assets and requires a shareholder vote only for the sale of "substantially all" assets. See, e.g., Del. Code Ann. tit. 8, § 271 (1991). A sale of more than 75% of assets, based on balance sheet value, generally requires a shareholder vote under this provision, while sales of between 26% and 75% of assets, based on balance sheet value, may trigger a vote. See Leo Herzel, Timothy C. Sherck & Dale E. Colling, Sales and Acquisitions of Divisions, 5 Corp. L. Rev. 3, 25 (1982). 86 To enhance enforceability, we use book rather than market values to trigger the shareholder approval requirements. A market value test is not administrable in a country like Russia, which has neither an efficient stock market nor reliable professional appraisers. On the administrative difficulties with a market value threshold for shareholder voting on a sale of assets, even in a developed economy, see Gilson & Black, cited above in note 53, at 653-65. 40 The first three items on this list need little comment. The requirement of a shareholder vote for mergers, liquidations, and changes of legal form is standard in most company laws. Only the fourth requirement, dealing with sales and purchases of assets, is stricter than the corporate laws of developed countries.85 The rationale for requiring shareholder approval is that self-dealing transactions of this size can destroy a company's value with the stroke of a pen. Disclosed self-dealing transactions are subject to separate voting rules, described below. But self-dealing transactions won't always be disclosed, and the incentive for concealment rises with the size and abusiveness of the transaction. A voting requirement for large asset sales and purchases provides a back-up constraint on hidden self-dealing transactions. When so much of a company is at stake, multiple protections are desirable. We also propose a shareholder vote on smaller purchases or sales of assets, involving 25-50% of a company's book value, that are not unanimously approved by all directors, including outside directors selected through cumulative voting. In effect, we create a hierarchy of asset sales and purchases, with larger transactions calling for stricter approval requirements: (i) purchases and sales of less than 25% of the book value of a company's assets are governed by a company's usual internal decisionmaking processes; (ii) purchases and sales of 25-50% of the book value of a company's assets require unanimous board approval or, if unanimous board approval is not achieved, shareholder approval; (iii) purchases and sales of 50% or more of the book value of a company's assets require shareholder approval.86 In Russia, book value will understate market value because of inflation and managers' incentives to hide profits (and to a lesser extent, assets) from the tax collector. This understatement makes the 25% and 50% book value thresholds stricter than they appear to be 87 Our proposed 25% and 50% thresholds reflect, albeit crudely, the likely understatement of asset values due to inflation. An alternate approach would be to use lower thresholds but to allow the board of directors to adjust book values for inflation. For Russia, we did not propose this approach -- either here or in the other places where we used a book value threshold to trigger procedural protections -- because there is no reliable inflation index and because an inflation adjustment would have made the statute more complex. But we view the question of whether the law should allow inflation adjustments as a close one. If managers were already accustomed to using inflation adjustments for other purposes (such as computing income tax or paying interest on bank loans), then the complexity cost of allowing inflation adjustments for purposes of the book value thresholds contained in the company law would be smaller and would be outweighed by the accuracy gains from inflation adjustments. The accuracy gains from an inflation adjustment would also outweigh complexity cost for a country that was experiencing hyperinflation -- which Russia, in 1995, was not. 88 See, e.g., Bayless Manning, The Shareholder's Appraisal Remedy: An Essay for Frank Coker, 72 Yale L.J. 223, 234-36 (1962). 89 See, e.g., Victor Brudney & Marvin Chirelstein, Fair Shares in Mergers and Take-Overs, 88 Harv. L. Rev. 297, 304-07 (1974). 41 on the surface, as well as less accurate measures of transaction importance. But we still much prefer book value to market value as a measure of transaction importance.87 2. Appraisal Rights. -- Even enabling corporate laws typically give to shareholders who vote against a major transaction requiring shareholder approval the right to collect the fair market value of their shares, as determined by a court. This "appraisal" remedy is far from perfect. On the one hand, appraisal has been criticized as a possible drain on a company's liquidity that may deter value-enhancing transactions.88 On the other hand, the appraisal remedy has limited power as a check on managers' breaches of fiduciary duty, because only large minority shareholders are likely to incur the legal expense required to exercise appraisal rights.89 The appraisal remedy will surely work even worse in emerging markets than it does in developed markets. Yet there is no obvious alternative. Policing fairness through judicial or regulatory approval of major transactions is neither practicable nor desirable. Hence, one can only try to ameliorate the worst problems associated with appraisal rights. For example, in developed economies, a shareholder must actively oppose a transaction to qualify for appraisal rights. In an emerging market, this condition weakens an already weak right. Given poor mail systems, shareholders may not learn of a transaction in time to vote against it, or may find that their votes did not reach the company in time or were conveniently lost. Therefore, we propose that shareholders who do not vote for a major transaction should be able, promptly after the transaction is completed, to obtain payment of the fair value of their shares, measured just before the transaction took place and excluding any effect of the transaction on the value of the shares. Typically, too, a shareholder seeking appraisal must go to court (an expensive process), without knowing in advance what the appraised value will be. This problem is especially severe in an illiquid market because published market prices, which provide an effective floor on the 90 Because the arbitration proceeding is not based on explicit contract, a mechanism is needed to identify neutral and reasonably skilled arbitrators. For Russia, we assigned to the Securities Commission the task of identifying suitable arbitration fora. A shareholder could elect any forum on the Securities Commission's list or a forum (if any) specified in the company charter. 91 A put option is inherent in any system of appraisal rights. This creates a collective-action problem -- a shareholder who seeks appraisal rights for a merger can, at modest cost, obtain a significant time window in which to decide whether to pursue the appraisal rights or to abandon them and receive the merger consideration. A shareholder then has an incentive to oppose a beneficial transaction, as long as her opposition is unlikely to affect the voting outcome. If mails and vote-counting procedures are reliable, it is appropriate to limit this free option by requiring that a shareholder vote against the merger, not simply (as in our proposal) that she fail to vote, and by imposing tight time limits for exercising the appraisal right (which reduce the option's value). The option value problem must also inform one's judgment about which corporate actions should give rise to appraisal rights. 42 appraisal price in developed economies, are often unavailable or far below true value. The self- enforcing model responds to these problems by requiring the company to announce an offer price for shares in the materials sent to shareholders to solicit approval for the underlying transaction, and to pay that price promptly on shareholder demand. The need to announce publicly management's estimate of the firm's value makes it more likely that the offer price will be plausibly related to true value. And easy access to an (often) reasonable price opens up the appraisal remedy to small shareholders, for whom the cost of a formal appraisal proceeding is prohibitive. We address the separate problem of ill-informed or corrupt judges by giving shareholders a choice: they can seek appraisal either in court or through arbitration.90 Company law must also be alert to potential misuse of the appraisal remedy. For example, minority shareholders might sabotage a beneficial transaction by demanding that the company buy back their shares at a time when it is strapped for cash. Moreover, shareholders will be tempted not to vote for a transaction because the right to sell one's shares back to the company is a valuable put option that can be exercised if the company's shares decline in value after the transaction, even if the decline is unrelated to the transaction itself. To balance the need for shareholder protection against the risk of misuse of the appraisal remedy, we give shareholders only a short period after a transaction is completed to demand that the company buy back their shares. For Russia, we proposed thirty days -- a period dictated by the slowness of the mail system.91 3. Determining Market Value. -- A problem that arises with special force in emerging markets is how to determine the fair market value of a company's shares and other property. Determining fair market value is important not only for appraisal, but also for other procedural protections discussed below, including those accompanying share issuances, self-interested transactions, and repurchases by the company of its own shares. Even developed countries have trouble defining market value. In an emerging market, a simple statement that shareholders should be paid the market value of their shares in an appraisal proceeding, or that a company should not issue shares for less than market value, may 92 Our threshold for treating a large shareholder as an insider is ownership of 20% of the outstanding shares. 93 Prohibition has the apparent virtues of simplicity and clarity. For example, one might prohibit all transactions between public companies and their directors and managers (except purchases and sales of the company's shares at market value), while employing procedural protections for self-dealing involving large shareholders (to permit parent-subsidiary and corporate group structures). For Russia, we rejected this approach for several reasons. One was political: Russian managers would probably have succeeded in introducing 43 fail of its intended effect without further definition of this intrinsically difficult concept. We adopt the following definition, which is still vague but better than nothing: The market value of property shall mean a price at which a seller who is fully informed about the value of the property and is not obliged to sell the property would be willing to sell, and at which a buyer who is fully informed about the value of the property and is not obliged to buy the property would be willing to buy. If the property to be valued is a publicly traded security, the person making the valuation shall consider the market price of the security over a period of time of at least two weeks prior to the date as of which market value is measured. If the property to be valued is a company's own common stock, the "value" of a share shall be understood as a pro rata claim on the value of the entire company, as presently organized and managed. In determining this value, the person making the valuation may consider the shareholder capital of the company, the price that a fully-informed buyer would be willing to pay for all of the company's common shares, and other factors that he considers important. This definition can guide those, especially directors, whom the law charges with determining the market value of shares or deciding whether a transaction is at market value. Of course, in an illiquid market, market value is not a single point, but a sometimes wide range. If the directors act in good faith and reach a reasonable valuation, the self-enforcing model instructs courts not to second-guess that valuation. B. Self-Interested Transactions Transactions by a company that personally benefit directors, managers, or large shareholders (all of whom we call insiders, recognizing that the description may not be accurate for large shareholders92) are inherently suspect because the insider has both the incentive and the ability to cause the company to enter into the transaction on unfair terms. Yet sometimes these transactions are advantageous to the company. Outright prohibition, in our judgment, is justified only when experience discloses both little business justification for a transaction type and a particularly high risk of abuse.93 We identify two such cases: loans by the company to loopholes in the legislative process, if not killing the ban entirely. Second, we judged that managers intent on self-dealing are less likely to evade the law -- and more likely to consider shareholder interests -- when they can self-deal legally with shareholder or board authorization. But the most important reason was efficiency-related. Many managerial conflicts of interest are likely to involve indirect transactions between the company and other firms with which the company's managers or directors are affiliated as shareholders, directors, or managers. A ban on all indirect conflicts would reach structures that might be efficiency-enhancing (such as interlocking directors), while selective bans on indirect conflicts would generate severe line-drawing problems and sacrifice the simplicity and clarity of a general norm. 94 Cf. Code des sociétés [Companies Law] art. 106, translation available in French Law on Commercial Companies 65 (Commerce Clearing House, Inc. 1988) (prohibiting loans or guarantees provided by a company to its directors or general managers). 95 Only 2 of the 17 jurisdictions in our survey retained some form of outright prohibition on contracts between the company and its directors or officers. Five statutes required shareholder approval of some self- dealing transactions, while most of the remaining statutes required approval by noninterested directors. See infra Appendix. 96 Our self-interested transaction rules are similar to French company law, which (in broad outline) requires that a self-interested transaction between a company and one or more of its directors or general managers be approved by the noninterested members of the board of directors, reviewed by the company's auditors, who prepare a report to the shareholders, and then approved by the noninterested shareholders. French law does not reach transactions with large shareholders. See Code des sociétés [Companies Law] arts. 101-03, translation available in French Law on Commercial Companies, supra note 114, at 64; André Tunc, A French Lawyer Looks at American Corporation Law and Securities Regulation, 130 U. Pa. L. Rev. 757, 766 (1982). 97 For very large companies, it can be feasible to require approval by noninterested independent directors. For Russia, we proposed such a requirement for companies with 10,000 or more shareholders. 44 insiders,94 and payments (kickbacks) by another person to an insider in connection with a transaction between the company and the other person. For other self-interested transactions, the self-enforcement model relies on a rigorous set of procedural protections, which apply in addition to any other board or shareholder approval requirements for particular transactions, such as mergers.95 The principal procedural protections are (i) approval by noninterested directors (directors who don't have a financial interest in the transaction); and (ii) for sizeable transactions (our threshold for Russia was 2% of the book value of the company's assets or 2% of annual revenues), approval by noninterested shareholders.96 Even noninterested directors will often not act independently. Nevertheless, directors who are elected by minority shareholders through cumulative voting are likely to be genuinely independent. Thus, the cumulative voting rules interact importantly with the rules on self-interested transactions. The size threshold balances the risk that the cost and delay of a shareholder vote will block good transactions against the need to block large bad transactions. Noninterested directors can cheaply review all self-interested transactions; only large transactions should require the costly additional step of shareholder approval.97 98 Four of the 17 emerging markets in our survey give minority shareholders put rights in the event of a transfer of control. As this is the basic English rule (described below), it is not surprising that three of these four jurisdictions are Commonwealth countries (Malaysia, Singapore, and Nigeria). See infra Appendix. 45 The self-enforcement model instructs the noninterested directors to approve a self- interested transaction only if they conclude that the company will receive value, in property or services, at least equal to the market value of the property or services the company gives up. This required finding informs the directors as to how to exercise their review power, gives directors who want to reject a transaction a basis for doing so, and provides a norm around which actual review practices may gradually coalesce. This standard may also provide a basis for a court challenge to especially egregious transactions, where the required finding was manifestly not made in good faith. Sometimes, of course, insiders will hide their interest in a transaction. But in some transactions, the insiders' interests cannot be concealed; in others the insiders will obtain an honest vote to protect the transaction against later attack in the courts; and managers who think of themselves as honest will voluntarily follow the rules. When self-interested transactions are disclosed, shareholders or noninterested directors can vote down some of the worst transactions, while the "sunshine" requirement that transactions be disclosed to shareholders will deter others. C. Control Transactions In a control transaction, a new shareholder or group of shareholders purchases or aggregates a controlling block of stock in the company. This acquisition of control can assume a wide variety of forms, including open market purchases and tender offers, and large share issues by companies in the course of financings or mergers. Whatever form it takes, an acquisition of control merits regulation in its own right -- even though many control transactions will also be regulated by rules governing particular transactional forms, such as reorganizations, major transactions, or self-interested transactions. Control transactions uniquely implicate both the efficiency goals of corporate law and its core problems, ranging from minority abuse to management entrenchment. The treatment of control transactions varies widely in both developed countries and emerging market jurisdictions. Friendly transactions are regulated in the most important United States jurisdictions only if they take the form of a merger or sale of substantially all assets, whereas hostile takeovers are regulated principally with a view toward discouraging them. Elsewhere, including (as we discuss below) in Great Britain, both friendly and hostile control transactions are often regulated regardless of their form.98 Beneath this divergent treatment lies a familiar policy dilemma. On one hand, control transactions are often engines for efficient restructuring. An investor often buys a control block because he expects to improve the company's efficiency in ways that will benefit all shareholders. Moreover, an outside investor's ability to acquire a controlling block of a company's stock without the managers' consent is an 99 For development of these points, see Lucian Bebchuk, Efficient and Inefficient Sales of Corporate Control, 109 Q.J. Econ. 957, 964-84 (1994), and Marcel Kahan, Sales of Corporate Control, 9 J.L. Econ. & Org. 368, 377-78 (1993). 100 See sources cited supra note 19; Robert Comment & Gregg A. Jarrell, Two-Tier and Negotiated Tender Offers: The Imprisonment of the Free-Riding Shareholder, 19 J. Fin. Econ. 283, 300 (1987) (reporting that, in a study of 27 partial tender offers, the price of nonpurchased shares increased, on average, by 15%). 101 Interview with Dusan Triska, Project Director of the RM-S Securities Exchange of the Czech Republic, in New York, N.Y. (Nov. 18, 1995). 102 See City Code on Takeovers and Mergers, supra note 9; Commission Proposal for a Thirteenth Directive on Company Law Concerning Takeover and Other General Bids, 1990 O.J. (C 38) 41, 44. An alternative rule for protecting minority investors, first proposed by William Andrews, would require purchasers of control to offer to purchase shares pro rata from all shareholders. See William Andrews, The Stockholder's Right to Equal Opportunity in the Sale of Shares, 78 Harv. L. Rev. 505, 506 (1965). Unlike the City Code rule, the Andrews rule permits partial tender offers, which reduce the cost of acquiring control. In our view, this advantage is more than offset by the costs and likely inequities of channelling control transactions through the vehicle of a public tender offer in the undeveloped and largely unregulated Russian market. Because we permit noninterested shareholders to waive their takeout rights by majority vote, they can vote to authorize friendly partial offers for control, pro rata or otherwise. 103 If a controlling shareholder already extracts large private returns from a company, a more efficient would- be acquirer may not be able to afford the controller's demand for a premium price for its shares if the acquirer must pay the same price to minority shareholders. Moreover, in an underdeveloped capital market, an acquirer 46 important constraint on bad management. Thus, there are powerful efficiency reasons not to overregulate control transactions, whether friendly or hostile. On the other hand, a new controlling shareholder may loot the company or use control to manipulate share prices and acquire minority shares at a price far below their true value.99 The risk of looting is far higher in emerging than in developed markets. For example, there is little harm or gain to minority shareholders in the United States, on average, when a control transaction takes place. Apparently, the efficiency gains from good transactions roughly balance the losses from self-dealing.100 In contrast, in the Czech Republic, share prices in many companies collapse after a change of control, indicating severe harm to outside shareholders.101 Thus, emerging markets require regulation focused on reducing the risk of looting. The self-enforcing model protects minority shareholders by giving them takeout rights after a change of control (we use 30% ownership as a proxy for control). Under this approach, which we adapt from the British City Code on Takeovers and Mergers and the proposed European Community 13th Directive on Company Law,102 a shareholder who acquires a 30% interest in a company must offer to buy all remaining shares at the highest price he paid for any of the company's shares within a specified period of time (we propose six months). This put option is a powerful deterrent to inefficient control transactions, for which we are willing to accept the cost of deterring some efficient control transactions.103 may not be able to finance the cost of honoring takeout rights. Again, these costs can be mitigated (though not eliminated) by our further proposal to permit minority shareholders to waive the takeout rights requirement by majority vote. For example, minority shareholders could waive takeout rights to permit a new investor with management skills to pay a control premium to an entrenched blockholder who is willing to sell out, but only at a price that reflects the value of control to him. 104 See City Code on Takeovers and Mergers, supra note 9, General Principle 7, at B2, Rule 21, at I13. 105 United States and European experience teaches that vesting the decision to resist a takeover in independent directors is insufficient to protect shareholders' interests. Independent directors too often back the managers' interests in resisting a takeover bid, sometimes at great cost to shareholders. American directors with the power to do so have often turned down takeover offers priced at twice the previous market value of their company's shares, on the grounds that shareholders will do even better if the takeover is defeated. These optimistic predictions do not often come true. Thus, we believe strongly that shareholders, not managers or directors, should decide whether a control change occurs by selling or retaining their shares. 47 A second danger in control transactions is that disaggregated shareholders can be induced to sell control too cheaply. For example, an acquirer may secretly accumulate control through numerous open market transactions, when shareholders could have demanded higher prices if they could have collectively negotiated a control premium, and other potential acquirers might have offered higher prices as well. To prevent secret acquisitions of control, we would require shareholders who acquire 15% or more of a company's stock to publicly disclose their identity, shareholdings, and plans to acquire more shares, and then wait thirty days before buying more shares. This gives the company's managers time to respond to an impending control transaction by seeking a higher bidder, proposing an alternate transaction that is more favorable to the shareholders, or convincing shareholders that their shares are worth more than the acquirer is offering to pay. A similar rule would apply to acquisitions of 30% or more of a company's shares that are approved by the company's managers. These delay and notice provisions make it more likely that a new controlling shareholder will have to pay a control premium. The delay period also provides a market check on the fairness of the premium because a competing bidder has time to make a higher offer. These delay provisions, however, exacerbate a third key concern for control transactions: managers will often try to block changes of control to preserve their own jobs. Managers typically argue that they must be able to reject hostile takeover bids to protect the shareholders' interests. We are skeptical of this argument in developed economies, and even more skeptical in emerging markets, where managers are already often heavily entrenched. Thus, we propose, again borrowing from the British City Code, a ban on preclusive defensive tactics such as "poison pills."104 Consistent with the overall self-enforcement approach, the combination of a delay period and a ban on defensive tactics vests the decision whether to transfer control in the prospective selling shareholders.105 Some corporate actions both inhibit control transactions and serve other business goals. For example, cross-ownership of shares among affiliated companies lets the managers of the affiliated firms entrench themselves by mutually supportive voting. Yet cross-ownership, 106 For example, in a reverse triangular merger, a parent company acquires a target by merging a subsidiary into the target. The consideration for the acquisition is parent stock held by the subsidiary, which is exchanged for the stock of the target company. This transaction form is useful because it does not disturb the target's corporate identity or contractual relationships. 107 See, e.g., Del. Code Ann. tit. 8, § 160(c) (1991). 108 Such a ban on cross-voting can prevent only egregious entrenchment schemes. It allows cross-voting at "parent" ownership levels below the 20% threshold, and thus permits groups of companies to tie up control internally through cross-holdings. For example, the rule would not bar a Japanese-style keiretsu, in which a dozen companies hold 5% stakes in one another -- even though such a structure precludes a challenge to control of a member company that is not sanctioned by the group. See Gilson & Roe, supra note 6, at 882- 90 (describing cross-ownership in keiretsu structures). We permit such structures because they can serve benign as well as defensive purposes. For example, Gilson and Roe observe that cross-ownership may encourage mutual monitoring or help to enforce relational contracts in product markets. See id. at 882-94. In addition, entrenchment is less severe when controlling shares reside in a larger group of companies, as distinct from a single parent company. 109 For a Russian example of such a tender offer, see Neela Banjerjee, Russian Oil Firm's Share Swap Draws Fire, Wall St. J., Mar. 28, 1996, at A10 (describing a Russian oil company's proposal to swap its own shares, with a market value of 63￠, for shares in its majority-owned subsidiary that, before the swap was announced, had a market value of around $2). 48 including the extreme case of subsidiaries holding their parents' stock, can also arise for good business reasons.106 The self-enforcing model responds by allowing cross-ownership but limiting cross-voting. Even in the enabling model, majority-owned subsidiaries cannot vote their parents' stock.107 In practice, managers of "parent" companies exercise working control over dependent companies at ownership levels well below majority ownership. Thus, our model Russian statute would bar a "dependent" company from voting its stock in a "parent" company when the parent holds 20% or more of the dependent company's stock.108 A further risk faced by minority shareholders is loss of liquidity if a majority shareholder acquirers a high percentage of the outstanding shares, perhaps through a tender offer. Such an offer can succeed, even if priced below the market value of the minority shares, because outside shareholders face a prisoner's dilemma. Even if they would collectively be better off if they rejected the offer, they individually cannot risk rejecting an offer that most other shareholders accept, because the price and liquidity of the remaining minority shares will collapse.109 We respond with an appraisal rights remedy: if a controlling shareholder's ownership crosses 90%, the company must offer appraisal rights to all remaining minority shareholders. The appraisal remedy eliminates the prisoner's dilemma: outside shareholders no longer have an incentive to sell their shares for less than they expect to receive in an appraisal proceeding. D. Issuance and Repurchase of Shares 110 We address here the protection of shareholders during share repurchases. Section IV.E addresses protection of creditors during share repurchases, dividends, and other distributions of corporate assets. We also focus here on the interests of the company's existing shareholders. It is the job of securities law to protect the interests of the purchasers of shares when shares are sold to the general public. 111 Only 3 of the 17 emerging markets in our survey (China, Poland, and Turkey) ban authorized but unissued shares. See infra Appendix. In our Russian proposal, an increase in authorized shares, like other charter amendments, requires approval by two-thirds of the outstanding shares. 49 Share issuances and repurchases are a fourth class of transactions that merit special procedural protections.110 Company sales and repurchases of shares have the potential both to shift value from outside investors to company insiders and to reallocate voting power among shareholders. But nothing is more critical to a company's survival and growth than its ability to raise new capital as the need or opportunity arises, without time-consuming procedural obstacles. 1. Share Issuances. -- There is a simple way to protect shareholders against share issuances at less than fair value: forbid authorized but unissued shares, thus forcing managers to seek shareholder approval for each new share issue. We reject this strict rule because it would either make raising capital too difficult or, paradoxically, come to mean nothing at all. Managers could not exploit unexpected financing and investment opportunities if every issuance of new shares required separate shareholder authorization. Moreover, such an approval requirement would greatly complicate management incentive compensation plans. Given these drawbacks, managers would search for a way around a ban. Most likely they would ask shareholders for blank-check authorization to issue new shares at every annual meeting. Yet if this ploy succeeded, the draconian restraint on share issuances would collapse into an empty formality, leaving shareholders with no protection at all. The self-enforcing model provides several alternative mechanisms for protecting shareholders against share issuances that are priced below fair market value or shift control over a company. Shareholders may authorize unissued shares, to be issued in the future by decision of the board of directors, as in the enabling model and many emerging markets.111 Shareholders who distrust their managers could refuse to give such authorization, but we expect this to be rare. But shareholders would enjoy four additional protections against abusive share issues, other than the power to withhold authorization. First, a stock sale to insiders is a self-interested transaction, subject to the approval requirements discussed in section IV.B. Second, a sale of shares equal to 25% or more of a company's outstanding shares requires approval by a majority of these shares, excluding shares already held by purchasers of the new shares. Third, issuances for less than market value (as determined by noninterested directors) are banned outright (though the malleability of the concept of market value limits the 112 In developed countries, below-market issuances are used almost exclusively as a form of incentive compensation for managers. Our ban on below-market issuances has the practical effect of forcing a company that wants to issue shares to managers to pay them a (disclosed) sum of money, which they can then use to buy shares at market value. Thus, the ban is basically a rule of disclosure: the cost of the incentive compensation is made explicit by paying it as salary that is then invested in shares. 113 We do not treat an insider's exercise of preemptive rights as a self-interested transaction, even though it is technically a transaction with the company, because preemptive rights are available to all shareholders on equal terms. 114 Preemptive rights are common in both developed countries (notably Britain and Continental Europe) and emerging markets. They are available in 11 of the 17 emerging markets in our survey. See infra Appendix. 115 Participation rights pose a risk to the viability of preemptive rights waivers. The participation right is a call option, exercisable for a limited period after the company sells shares, to buy shares at the same price. Like any option, it has value. Even if all shareholders would benefit if preemptive rights were waived, individual shareholders are better off if others waive preemptive rights but they retain the participation option. The value of participation rights must be limited to make them viable. Limits arise in several ways. First, the time period for exercising participation rights should be short to reduce the time value of the participation option. Second, communications technology imposes a lag between the time when participation rights can be exercised and the time when additional shares are received. This prevents risk-free arbitrage, in which a shareholder buys shares at one price using participation rights and immediately resells the shares at a higher price in the market. Third, and most critically, an emerging market is characterized by wide bid-asked spreads 50 effectiveness of this ban).112 Finally, shareholders receive the preemptive and participation rights detailed below for all issuances above a de minimis threshold (2% of the previously outstanding shares).113 The de minimis exception aside, for new share issuances, a company would have to offer to its existing shareholders rights to purchase newly issued shares in proportion to their prior holdings (preemptive rights).114 Preemptive rights protect shareholders against underpriced issues, but are costly for companies with many shareholders and can delay time- sensitive transactions. Thus, the law must allow for waivers, including routine waivers approved at annual meetings. This revives the risk that shareholders may be harmed by below- market share issues. To limit this risk, the self-enforcing model gives to shareholders who do not vote to waive preemptive rights what we call participation rights. These rights entitle the shareholders who hold them to buy from the company after the offering has been completed as many shares, at the offering price, as they could have bought had preemptive rights been available. If the offering price is fair, few shareholders will exercise participation rights, and an offering will take place much as in countries where a waiver of preemptive rights binds all shareholders. But if a company sells shares for substantially below market value, shareholders will rush to exercise their participation rights and buy bargain-priced stock. This will let those shareholders recoup most of the dilution caused by the below-market issuance. It will also embarrass the managers by making the underpricing obvious to all, and will make the shareholders reluctant to waive preemptive rights in the future.115 and often by intervals when one cannot sell one's shares at any reasonable price. These features further reduce the option value of participation rights for a fairly valued offering. Our judgment for Russia was that the option value of participation rights would be small enough not to be a major obstacle to a waiver of preemptive rights, much as the put option inherent in appraisal rights has not, in practice, been a major obstacle to shareholder approval of mergers. 116 Twelve of the 17 emerging market jurisdictions in our survey ban repurchases of common stock. See infra Appendix. 117 We do not treat an insider's sale of stock to the company, pursuant to a pro rata repurchase offer, as a self- interested transaction, because the offer was available to all shareholders on equal terms. 118 If stock markets are well developed, a publicly announced repurchase of shares in the open market can be treated in the same manner as a pro rata repurchase, because all shareholders have an equal opportunity to sell their shares at the market price. However, in the illiquid stock markets that characterize many emerging economies, a repurchase, supposedly made on the open market at the market price, can be used to repurchase shares from insiders at a favorable price. The insiders will know when the company will be buying shares, and can sell at a high price. Once the company finishes buying shares, the price will drop again. Thus, we treat open market purchases as non-pro-rata. This regulatory treatment should change when a country's stock market becomes sufficiently liquid. 119 If the shareholders whose shares are to be repurchased are known, they should be ineligible to vote. If the selling shareholders are not known, as in the case of an open market repurchase, then all shares can vote. Repurchases of preferred stock should be regulated similarly to repurchases of common stock. Non-pro-rata repurchases should require approval by the holders of the class of preferred stock to be repurchased. 51 2. Repurchase of Shares. -- Share repurchases can be as perilous for shareholders as share issues but are less critical to a firm's growth. Thus, they are banned by many corporation statutes.116 In our view, this prohibition goes too far. A pro rata (open on equal terms to all shareholders) repurchase of shares is basically just a way for the company to distribute cash to shareholders, and therefore raises only limited fairness concerns. Pro rata repurchase offers can be a valuable corporate tool, especially if (as in Russia) they are tax-favored compared to dividends. Our model permits pro rata repurchase offers without special shareholder approval.117 Nevertheless, even pro rata offers should be made at market value as determined by the company's independent directors, unless a different price was agreed upon when the shares were acquired. Approval by independent directors is necessary because insiders are usually large shareholders who want the company to buy shares from others for as low a price as possible, while outsiders may be willing to sell at a low price because they do not know the true value of their shares.118 In contrast, a non-pro-rata offer to repurchase shares raises a concern that insider shares will be repurchased at more than fair market value, or that the company will buy out a troublesome shareholder at a high price. The self-enforcing model requires that a non-pro-rata repurchase of shares be made at market value, determined by the independent directors, and be approved by shareholder vote. A non-pro-rata repurchase of shares from insiders is also regulated like any other self-interested transaction.119 120 We allow reverse stock splits, in which small holdings are cashed out in lieu of issuing fractional shares. This is a form of mandatory, non-pro-rata repurchase of stock. Thus, it should require approval by a majority of outstanding shares, the price paid for fractional shares should be market value (determined by the independent directors), and shareholders whose shares are cashed out should receive appraisal rights. 121 See, e.g., Revised Model Business Corp. Act § 6.40 (1991). 122 See 11 U.S.C. § 547 (1994). 123 See, e.g., Unif. Fraudulent Conveyance Act § 4, 7A U.L.A. 474 (1985). 52 Turning from repurchases of stock (voluntary for the selling shareholder) to redemptions (mandatory for the selling shareholder), a company should be allowed to redeem preferred shares at its option if its charter so provides. The principal risk is that the price offered by the company will be too low. To respond to this risk, we believe that a company should be able to redeem preferred stock only on a pro rata basis, or by lot to the extent necessary to avoid redeeming fractional shares. There is no need to allow non-pro-rata redemption of common stock, which would permit managers to buy out unwanted shareholders at a low price. Pro rata redemption of common stock, which is functionally equivalent to paying a dividend, should be allowed if it offers more favorable tax treatment than paying a dividend.120 E. Protecting Creditors and Preferred Shareholders Minority shareholders are not the only corporate investors who are threatened by the information asymmetries and weak capital markets that characterize emerging economies. The business failures that plague these economies put stress on credit relationships and create incentives for opportunism by shareholders or managers at creditors' expense. In developed and emerging economies alike, contract is the principal instrument of self- protection for creditors and preferred shareholders. Banks can take security interests in assets; bondholders can demand covenants that restrict distributions of corporate assets; trade creditors can provide only short-term credit, thus limiting their losses in the event of default; and preferred shareholders can insist on the power to elect some or all of the board of directors if dividends are missed. But experience also suggests a role for legal limits on distributions of assets even in developed economies. In the United States, for example, company law bars dividends and stock purchases by an insolvent company;121 "look-back" provisions of bankruptcy law allow recapture of pre-bankruptcy distributions;122 and fraudulent conveyance law allows reversal of transfers made for unfair consideration by insolvent companies.123 These substantive limits respond to the powerful incentives for shareholders to extract whatever value they can from a failing company. Banks and other large creditors can protect themselves by writing detailed contracts and monitoring a borrower's compliance with contractual covenants. But experience teaches that statutory limits on distributions are important protections for trade and other small 124 See GK RF (Civil Code), supra note 42, pt. I, art. 64. 125 In the United States, a third solvency test is also used: after the transaction, the company must not have an unreasonably small amount of capital left with which to conduct its business. See Unif. Fraudulent Transfer Act § 4(a)(2)(i), 7A U.L.A. 652-53 (1985). But this extremely vague test is rarely used in practice. In keeping with our preference for defining legal requirements clearly, to guide both directors and judges, we do not consider this third standard to be useful in an emerging market. 126 Similarly, to protect preferred stockholders, the law should allow dividends on or repurchases of common stock only if, after the payment, the company's net assets (measured at both book and market values) exceed the liquidation preference that the preferred stock would enjoy if the company were to be liquidated immediately. If there is more than one class of preferred stock, a similar restriction would apply to dividends or repurchases of a junior class of preferred stock. The dividend or repurchase would be permitted only if, after the dividend or repurchase, the company's net assets were sufficient to pay the liquidation preference of the more senior preferred stock. 127 If inflation is high, directors should have the option to adjust historical book values using a generally accepted inflation index to ensure that the book value test is not too constraining. 53 creditors, for whom the transaction costs of writing or enforcing detailed contracts are prohibitive. In emerging markets, creditors have even greater need for protection. They are often less sophisticated than creditors in developed countries, and key credit market institutions are often absent, such as financing factors who buy accounts receivable and monitor borrowers on behalf of small trade creditors, and information services that report a borrower's payment history or financial strength. Yet the available legal tools for protecting creditors are also limited. In Russia, bankruptcy law does not function at all, let alone contain sophisticated look- back provisions. Secured lending is crippled by a Civil Code provision that gives secured lenders third priority in insolvency proceedings, after personal injury claims and employee claims.124 Even ordinary contracts are often not readily enforceable. The simplest ways for a company to distribute assets to shareholders at the expense of creditors are through dividends (whether of cash, stock, or other property) and stock repurchases. From a creditor's perspective, these are identical transactions and should be regulated in the same manner. The self-enforcing model permits dividends or repurchases only if, after the payment, (i) the company will have net assets (assets minus liabilities) greater than zero, whether assets and liabilities are measured at book or at market value; and (ii) the company reasonably expects to be able to pay its bills as they come due. The first test is an asset test for the company's solvency; the second test is a liquidity test for solvency.125 (The requirement that the company repurchase stock at fair market value also provides some creditor protection because if the company is close to insolvency, its shares will have little value.126) The market value test for asset-based solvency is familiar from developed countries. We add the book value test because one cannot rely on courts in emerging markets to apply sensibly a rule that looks to market value to determine whether a company is insolvent after paying a dividend.127 We keep the market value test because book value may be an unrealistic measure 54 of value. In Russia, for example, many intercompany debts that will never be paid are listed as assets by the company to whom they are nominally owed. We also protect creditors and preferred shareholders through disclosure -- companies must notify creditors of dividends or stock repurchases that are large enough to significantly affect the company's creditworthiness (our threshold is a 25% decrease in the book value of the company's net assets). This disclosure permits creditors to avail themselves of contractual rights, and evaluate whether and on what terms to extend new credit. Large creditors can, of course, insist on notice of dividends or on dividend restrictions by contract. But trade creditors may not be able to, or think to do so, and trade creditors in emerging markets will often lack other sources of information about changes in a borrower's financial condition. Although dividends and stock repurchases by a company that is in or near insolvency are particularly suspect as asset distributions, any corporate transaction can be a vehicle for extracting assets from the company, to the detriment of creditors and often shareholders as well. The challenge is how to block bogus transactions without impeding ordinary business dealings. Here we can do no better than the vague standard, familiar from fraudulent conveyance law, that a transaction is improper if (i) the company does not receive equivalent value, and (ii) the company fails an asset-based or liquidity-based solvency test after the transaction. We again use both book and market value tests for asset solvency. The standard of equivalent value is fuzzy, but it can at least reach egregious cases where a company transfers most of its remaining assets to third parties for nominal consideration to evade its creditors. A typical Russian situation involves a raw materials company selling its product at a fraction of market value to another company controlled by its managers, who then resell the product at the market price. A creditor, irate shareholder, or even a judge should be able to spot a sale at a bargain price followed by prompt resale at a far higher price. As this example suggests, moreover, many fraudulent conveyance transactions are also self-interested, which enlists the self-interested transaction rules in protecting creditors and those shareholders who do not share in the company's largess. These creditor-protection rules move, in part, beyond the self-enforcing model of procedural protection to substantive prohibition. Substantive restrictions on dividends, stock repurchases, and non-arms-length transfers are justified by: the strong incentives of shareholders and managers to grab what they can from a sinking ship; the dearth of legitimate reasons for a company near insolvency to pay dividends (and the total absence of reasons to enter into transactions without receiving reasonably equivalent consideration in return); and the difficulty of attacking this problem in another way because a company's relationships with creditors are too complex to permit a voting solution. In part, however, the appearance of substantive prohibition is deceiving. For example, a company that wants to pay a dividend that the book value test would otherwise prohibit can first raise new equity capital to pay off its old creditors, then pay the desired dividend. The company can then recreate its old capital structure if new lenders can be found. In effect, the dividend limit requires the company to let creditors vote with their feet. Such a financial end run around the dividend restrictions involves substantial transaction costs, but its possibility 128 These estimates are from a survey of a limited number of regional property funds. See Katharina Pistor & Joel Turkewitz, Coping with Hydra -- State Ownership After Privatization: A Comparative Study of the Czech Republic, Hungary, and Russia, in 2 Corporate Governance in Eastern Europe and Russia: Insiders and the State, supra note *, at 192, 206 tbl. 6.5. 129 Convincing government authorities to adopt a policy of neutrality may be difficult. The Russian company law, as adopted, not only failed to neutralize the state's shares, but also gave the state extra powers in some circumstances, most notably the right to maintain its percentage ownership during an issuance of additional shares if the state owns more than 25% of a company's shares. See Federal Law of the Russian Federation on Joint-Stock Companies, supra note 2, art. 28, ? 4. 55 softens the harshness of a book value test for solvency in an environment where book value may be a poor measure of actual value. F. The State as Part-Owner The state is frequently an important equity holder in emerging economies -- especially in the privatizing economies of Central and Eastern Europe. In Russia, for example (after completion of voucher privatization, but before the current, slow cash phase of privatization), regional property funds retain some equity in most privatized companies and hold stakes of 20% or more in perhaps one-third of all companies.128 Several concerns are raised by such large state holdings. One cannot expect government officials to behave as private shareholders who bear the economic consequences of company decisions. State officials may form alliances with managers at the expense of shareholders, competitive markets, or both. They may influence company policies in inefficient directions to accomplish public ends (such as preserving employment in a particular region). Or they may trade votes for personal favors that managers can supply. Because we are skeptical about whether local officials will behave as responsible shareholders, we favor neutralizing government shares in the election of boards of directors: state bodies should neither nominate nor vote for candidates for the board of directors, although they should retain authority to vote on potentially company-transforming actions such as mergers and charter amendments.129 V. Remedies Substantive rules are only one aspect of a well-designed statute. Remedies are equally important and in some respects even more complex. In formulating substantive rules for a self- enforcing statute, the need to economize on enforcement resources leads to a preference for simple, bright-line rules -- though we occasionally use general standards for pedagogical purposes and to influence norms of behavior over the long term. For identical reasons, a self- enforcing statute should often define the remedies for violations of the substantive rules, rather than leave their development to the courts. Defining remedies isn't always possible, but some 130 See, e.g., Gary S. Becker, Crime and Punishment: An Economic Approach, 76 J. Pol. Econ. 169, 183 -85 (1968). 56 common violations can be anticipated and their consequences elaborated, rather than left to the uncertain wisdom of judges. This stress on clarity and simplicity implies that remedies should often take the form of rights running to shareholders directly, rather than rights mediated by the corporation (acting on behalf of all shareholders) or by regulators. One broad example is the set of transactional rights described in Part IV -- appraisal, preemption, participation, and takeout rights -- where the substantive rule also defines a shareholder-enforced remedy. To take another example, shareholders can be given direct rights to sue an insider to recover the insider's profit from a self-interested transaction that did not receive the requisite shareholder approval. The recovery should go to the company, but the right to sue can belong to the shareholder. Conversely, the cumbersome device of the derivative suit -- a suit by the shareholder in the name of the company, with painstaking judicial oversight of when the company's board can control the suit -- has little place in an emerging economy. We limit nuisance or strike suits not through judicial oversight of derivative suits (as under Delaware law), but instead by requiring that the suing shareholders own a significant stake in the company (for Russia, we required ownership of 1% of the company's common stock). Consistent with the spirit of a self-enforcing statute, shareholders can sometimes be left to determine the content of a remedy. For example, we penalize a shareholder who improperly crosses the 30% ownership threshold without honoring takeout rights with loss of voting rights unless these rights are restored by a majority vote of the other shareholders. Similarly, the penalty for improper voting procedures is a second vote that lets shareholders decide how to react to the misfeasance of the managers the first time around. A second remedies issue is the severity of penalties for violating the rules. In general, sanctions should be more severe than in developed economies to compensate for the low probability of detection and enforcement.130 Consider, for example, the requirement that self- interested transactions receive the approval of disinterested shareholders. We propose that insiders who fail to disclose an interested transaction must return all profits from the transaction to the company. Contrary to the majority rule in the United States, we reject ex post shareholder ratification and proof of substantive fairness to the company as defenses to liability because these defenses would undercut the incentive to obtain shareholder approval ex ante. Managers would be tempted to ignore the procedures in advance if the remedy for failing to follow them were merely a reprise of the required procedure. Ex post shareholder approval also sacrifices the prophylactic value of ex ante approval and ex ante disclosure. Not every transaction that would be approved ex post will be proposed ex ante; and not every transaction that would be disapproved will be challenged. Because a self-enforcing corporate law makes heavy use of bright-line rules, it can carry strong penalties with less risk of chilling legitimate behavior than similar penalties would create 57 under an American-style law that relies heavily on fuzzy fiduciary duties. For example, our proposed remedy for crossing the 30% ownership threshold for the control transaction rules without prior notice, or without offering to buy all remaining shares, was loss of voting power as to all shares held by the 30% shareholder, unless other shareholders vote to restore voting rights or the shareholder acquires at least 90% of the outstanding shares. A severe sanction is appropriate here not only because violations may be difficult to detect (acquirers may hide ownership by buying through undisclosed affiliates), but also because a clear rule means that inadvertent violations should be rare. By contrast, other factors cut against very severe sanctions in many circumstances. Some of these factors arise from the need to adapt the Russian statute to the sophistication and sensibilities of Russian managers. One cannot assume that corporate managers will always know the rules. Thus, it can seem unacceptably harsh to penalize inadvertent violators heavily, especially for rules that apply importantly to small companies. Moreover, the law must be seen to be fair if it is to induce voluntary obedience and, ideally, the conformity over time of culture to the new law. To meet this need for perceived fairness, sanctions must fall well short of the deterrent ideal, which sets expected penalties (actual penalties if caught multiplied by the probability of detection) equal to the harm caused by misconduct. These considerations explain why, for example, we do not propose double or triple damages for failure to disclose self- interested transactions -- even though the conventional logic of deterrence might recommend doing so. Furthermore, intrinsic enforcement limitations make severe sanctions unworkable under some circumstances. For example, the draft Russian statute never imposes liability on directors for decisions taken in good faith and without a conflict of interest. In effect, we close off the narrow American recklessness/gross negligence exception to the business judgment rule because we have no confidence that Russian courts can decide when conduct is sufficiently outrageous to warrant imposing (often ruinous) personal liability on directors. In this context, any liability for good faith decisions creates the problem that the business judgment rule -- which protects directors from liability even for demonstrably stupid decisions -- was designed to solve: the risk of liability can chill risk-taking in a world where managers often need to take gambles, sometimes long-shot gambles, on limited information. Sometimes, too, there are no effective remedies for the violation of a statutory norm. Suppose that a company fails to use cumulative voting. One cannot invalidate the actions of the improperly elected board without imposing an unrealistic burden of investigation on third parties who deal with the company. The violation is clear enough so that one can imagine imposing personal liability on board members for loss to the company from actions approved by an improperly elected board. Yet the Russian draft stops short of this sanction, partly because even honest directors may not understand how to implement cumulative voting, and because the directors who would resign to avoid this liability may be the best available. We are left with the weak sanction of running a new election -- which will at least allow shareholders to vote again, knowing how their directors have behaved in the past. 131 See, e.g., William L. Cary, Federalism and Corporate Law: Reflections upon Delaware, 83 Yale L.J. 663, 663-86 (1974). 132 See, e.g., Romano, supra note 23, at 52-53, 148-51; Ralph K. Winter, Jr., State Law, Shareholder Protection, and the Theory of the Corporation, 6 J. Legal Stud. 251, 254-58 (1977). 133 See Romano, supra note 23, at 17-18 & nn.7, 20 tbl. 2-1 (reviewing studies). 134 This work is in progress, under the tentative title: Reinier Kraakman & Bernard Black, The Path- Dependent Evolution of American Corporate Law. 58 VI. The Path-Dependent Evolution of Developed Country Corporate Law Our work has important implications for the ongoing debate about the efficiency of corporate law in developed countries. In the United States, the debate is traditionally framed as between race to the bottom proponents -- who see American states as competing to adopt lax corporate laws in order to attract corporate managers, who make the reincorporation decisions131 -- and race to the top proponents -- who believe that (perhaps with an exception for antitakeover rules) states attract corporations by offering efficient rules.132 The principal evidence offered by race to the top proponents is a set of empirical studies showing that reincorporation in Delaware either increases or does not decrease stock price, and thus is either efficiency-enhancing or at least efficiency-neutral.133 We believe that the empirical evidence is ambiguous, and that the evolution in the United States of corporate law for large public companies is more complex than either camp has acknowledged. We propose the following alternative, which fits our anecdotal sense of the history of American corporate law but must remain tentative until we can complete the historical research needed to confirm it.134 American corporate law evolution began in the mid-nineteenth century from a historically contingent starting point of rigid formal rules, which reflected public suspicion of corporations, weak market constraints that called for strong investor protections, poor communications that made impractical some of the shareholder approval procedures embodied in the self-enforcing model, and a misunderstanding of corporate finance that led to an early obsession with charter capital. Corporate law then evolved from this starting point toward today's enabling law on a path conditioned by a mix of efficiency considerations, political considerations, and historical constraints, but not entirely determined by any of these factors. An important degree of freedom in the evolution of corporate law was made possible by the emergence of other institutions that filled gaps in the early corporate laws. For example, corporate law early on permitted deviation from the one common share, one vote principle. But the New York Stock Exchange filled that gap for public companies in 1926. Corporate law permitted such outrages as issuing dividend checks that, when endorsed by the shareholder, gave managers a proxy to vote the shares as they pleased. In response, stock exchange rules and the federal securities laws intervened to provide a better proxy voting system. Corporate law required little financial disclosure by companies to shareholders -- but again, the federal 135 See, e.g., Harold Marsh, Jr., Are Directors Trustees? Conflict of Interest and Corporate Morality, 22 Bus. Law. 35, 43-57 (1966). 136 488 A.2d 858 (Del. 1985). 59 securities laws intervened to erect a system of mandatory disclosure. In addition, common law judges exercised substantial power under fiduciary doctrines to fill the gaps left by other bodies of law. Interstate competition for new incorporations also significantly constrained the development of corporate law. The states needed to satisfy two competing constituencies: shareholders, who were interested in efficiency; and managers, who wanted more discretion. Managers initiated incorporation decisions, and were thus a critical constituency. A reincorporation that increased share values but decreased manager autonomy would not interest managers. But managers also had to appease shareholders. It would be too bold and potentially embarrassing for managers to propose a reincorporation in another state, or a change in the law of one's own state, that gave the managers more discretion but visibly harmed shareholders. The path of least resistance was legal reform that both enhanced (or did not decrease) managers' autonomy and increased (or did not decrease) company value. As we see it, corporate law meandered down this path of least -- or at least low -- resistance. Consider, for example, the rules governing transactions in which directors and managers had a conflict of interest. The early rigid rules against self-dealing could have been replaced by shareholder review, as we propose in the self-enforcing model. Instead they were primarily replaced by the weak constraint of board approval. Here is one plausible story for why corporate law might have evolved in this way, without ever finding the potentially better approach of shareholder review. Corporate managers were acutely aware of the problems with the prohibitive model, and pushed for greater discretion. Over time, courts and legislatures responded. The managers had no incentive to suggest replacing prohibition with shareholder review. Legislators were thus never presented with the self-enforcement option developed here, or anything close to it. It was left to the courts to modestly counteract legislative permissiveness by requiring strict review under a fairness standard of self-interested transactions that are not approved by noninterested directors. In this story, the enabling model could dominate the prohibitive model as a means of regulating self-interested transactions (though even that is debatable135), and yet be sub-optimal relative to self-regulation by shareholder approval -- an approach that was never considered. Additional examples of corporate law following a path of low resistance are easy to find. As Delaware law became more friendly to mergers, it did not provide effective appraisal rights, and judges had to fill this gap too. Unlimited director liability for violation of the duty of care can cause outside directors to act too timidly -- a threat that became real after the Delaware Supreme Court, in 1985, found a duty of care violation in Smith v. Van Gorkom.136 Managers pressed the Delaware legislature to respond. The legislature might have capped 137 Such a cap was proposed in Principles of Corporate Governance, supra note 13, § 7.19. 138 See, e.g., Roe, supra note 6, at 26-28; Black & Coffee, supra note 6, at 2082-84; Mark J. Roe, Chaos and Evolution in Law and Economics, 109 Harv. L. Rev. 641, 643-58 (1996). 139 See, e.g., Roe, supra note 6, at 26-28; Bernard S. Black, Shareholder Passivity Reexamined, 89 Mich. L. Rev. 520, 564-66 (1990). 60 director liability at a multiple of director compensation.137 Instead, it gave companies the option, soon taken by most large firms, to have no director liability at all for duty-of-care violations. When managers' jobs were directly at risk during the takeover wave of the 1980s, they wanted power to resist takeovers strongly enough to support statutes that shareholders frequently opposed and that risked decreasing the value of their own companies' shares. Often, the managers got what they wanted. And so on. This model -- in which corporate law evolves in a path-dependent fashion down a route of low resistance -- is consistent with both the realpolitik stressed by race to the bottom supporters (that managers make reincorporation decisions) and the empirical evidence cited by race to the top supporters (that reincorporations tend to increase, or at least not decrease, share values). It is also consistent with recent scholarship that emphasizes the importance of path dependence and legal rules in determining the ownership structure of large public companies.138 In effect, we seek to extend the path dependence story beyond stock ownership by financial institutions (a story already told) to the corporate law itself. The path-dependence argument can be taken further. The evolution of corporate law is intertwined with the evolution of financial institutions. If financial institutions are economically and politically powerful, the evolution of corporate law will reflect their interests as well as those of company managers. And financial institutions, in turn, will evolve in ways influenced by the corporate law. For example, both Britain and the United States have long had active capital markets, including active markets for corporate control. In the United States, financial institutions were weak, in part because political decisions made them so.139 American managers succeeded in obtaining broad discretion to oppose takeovers -- the one clear case of evolution in American corporate law away from efficiency. In Britain, financial institutions, especially insurance companies, were strong, and successfully opposed takeover defenses that would take the change-of-control decision away from shareholders. In Germany, unlike Britain and the United States, universal banks have long been strong. They run mutual funds and investment banks, and their representatives sit on company boards. Tight restrictions on conflicts-of-interest and insider trading would have limited the banks' freedom to profit from their multiple roles. Is it simply an accident that Germany had little insider-trading regulation until recently, when European unification and the internationalization of capital markets created a constituency for stronger rules, or that Germany still allows conflicts of interest that even the most pro-manager Delaware judges would find intolerable? 140 Also see note 80 above, noting that our proposed protections for employee shareholders were dropped from the Russian company law, as adopted. 141 For example, under articles 99 and 101 of the Russian Civil Code, if losses cause a company's net assets to be less than its charter capital, the company must reduce its charter capital. But doing so requires giving each creditor an option to demand immediate repayment of its loan to the company, which could quickly exhaust the company's liquid assets. See GK RF (Civil Code), supra note 42, pt. I, arts. 99, 101. 61 The Russian story also illustrates the importance of politics and powerful players in determining the structure of corporate law. In Russia, as elsewhere, corporate managers are influential. Many want a corporate law that insulates them as much as possible from shareholder oversight. The corporate law that Russia adopted is largely based on our self- enforcing model, but contains elements that can only be understood as pro-manager compromises. For example, our effort to introduce a British-style system of takeover regulation -- with requirements for advance notice of a control transaction, a mandatory takeout offer for minority shares, and a ban on defensive tactics -- lost the ban on defensive tactics along the way.140 Historical accident matters too. For example, the recently adopted Russian Civil Code contains some rigid, dysfunctional charter capital rules.141 From this starting place, we expect, Russian company law will evolve along its own path of low resistance -- with dysfunctional rules falling by the wayside, with managers adding to the discretion that the law already gives them, but with other institutions developing over time to ameliorate the consequences of excessive managerial discretion. Conclusion: Self-Enforcing Law in Emerging Economies In this Article, we have argued that the best model for company law for an emerging economy is neither the enabling model that characterizes developed economies today, nor the prohibitive model that characterized corporate law a century ago, but instead a "self-enforcing" model. The core of this model is an effort to harness the incentives of participants in the corporate enterprise, especially large minority shareholders, in order to provide meaningful protection to minority shareholders despite the absence of the multiple private and public enforcement resources of developed economies. Minority shareholders require strong legal protection from insiders in emerging markets because the market controls that provide such protection in developed economies are weak. Yet because public enforcement is also weak, company law in emerging markets cannot simply substitute formal law enforcement for the missing market controls. To operate effectively in this difficult environment, company law must be self-enforcing in a double sense. First, it must provide procedural mechanisms to replace the largely absent mechanisms for formal enforcement, and allow outside shareholders and outside directors to police the opportunism of managers and controlling shareholders. These constraints should be 62 primarily procedural (rather than substantive, as in the prohibitive model) to preserve flexibility in corporate decisionmaking. Second, because self-enforcement in this first sense is uncertain, the company law must elicit a substantial measure of voluntary compliance from managers and controlling shareholders. Toward this end, the statute must articulate bright-line and easily understood rules, provide terms that corporate actors recognize as appropriate to their business circumstances, and strongly sanction departure from these norms (in the rare cases in which formal sanctions are imposed). Thus, the model does more than substitute private enforcement for formal judicial enforcement; it also relies on "self" enforcement of a different kind: the inherent organizing force of clear and legitimate law in an otherwise chaotic business environment. The drafting strategy implicit in the self-enforcing model reaches almost every aspect of company law. At the most basic level, our model statute structures the company's voting system to increase the influence of minority blockholders. It mandates a single class of voting common stock; a one share, one vote rule; and a cumulative voting rule for the election of directors. In addition, it provides for a universal ballot, on which large shareholders can nominate board candidates. The model statute relies on a combination of voting constraints and transactional rights to regulate especially important or suspect transactions. Significant self-dealing transactions must be authorized by majority vote of both noninterested directors and noninterested shareholders. Mergers, liquidations, and large purchases and sales of assets require approval by a supermajority of outstanding shares (we suggest two-thirds as the appropriate threshold in Russia), with appraisal rights for shareholders who do not vote to approve the transaction. In addition, new issues of shares are subject to preemption rights. These rights can be waived by a shareholder vote, but shareholders who do not approve the waiver receive ex post participation rights, much as shareholders who do not approve a merger receive appraisal rights. Finally, the acquisition of control stakes carries an obligation to offer to buy minority shares, and defensive measures that might prevent shareholders from selling to would-be acquirers are prohibited. These and other features of the self-enforcing approach produce a company law that is novel in the aggregate, even though many of its individual provisions are familiar. The model flatly prohibits almost nothing except efforts to opt out of its process requirements. Nonetheless, it significantly constrains insiders by allocating to large-block minority shareholders -- those shareholders with sufficient holdings or support to win board representation under a cumulative voting rule -- considerable power to influence corporate decisionmaking. If, as we have argued, optimal corporate law depends on institutional context, then a country's corporate law should evolve as its economy and legal system evolve. As compared with the enabling model, the self-enforcing model gives greater power to outside investors, but at the cost of less flexibility in the basic structure of corporate governance. As market constraints and a sophisticated judiciary develop to limit opportunism by managers and large 142 For development of the argument that financial institutions will often choose exit over voice, if both options are available, see John C. Coffee, Jr., Liquidity Versus Control: The Institutional Investor as Corporate Monitor, 91 Colum. L. Rev. 1277, 1318-28 (1991), and Black & Coffee, supra note 6, at 2007-77. 63 shareholders, the comparative merit of the enabling model will increase. In a decade or two, mandatory cumulative voting, or a mandatory one share, one vote rule, may have outlived its usefulness and can be relaxed. Perhaps, too, as market liquidity increases, institutional shareholders will choose to make the "exit" alternative to voice more viable by reducing their percentage stakes in companies.142 If large investors choose exit over voice, the voice- enhancing benefits of the self-enforcing model will shrink, while the costs of our voice- promoting rules will remain. Corporate law then should, and probably will, evolve toward fewer voice-promoting rules. But evolution toward the enabling model is not a foregone conclusion. Given the mutual interaction between the law and the evolution of companies and financial institutions, in which strong protections for large outside investors encourage them to hold large percentage stakes, the self-enforcing statute may prove substantially stable. These large shareholders could also provide the political constituency to preserve the self-enforcing model against the attacks of managers seeking greater autonomy. The result would be a new model of mature company law: one that relies much more on internal decisionmaking processes and shareholder authorization -- and much less on ex post litigation -- than is currently the practice in the United States. 143 The table is based on the most recent versions available to us. 144 The data in the textual portion of the Appendix exclude the Russian company law. 64 APPENDIX: Survey of Company Law in Emerging Markets The table below surveys selected aspects of the company laws of seventeen emerging market countries (plus the new Russian self-enforcing law, for comparison). All but a couple of the countries have fairly recently adopted or updated their company laws.143 The table tallies seven possible substantive restrictions and eleven possible procedural rules. Some substantive restrictions are extremely common. For example, twelve of the company laws include a general ban (often with limited exceptions) on company share repurchases.144 Almost as many laws include limits on a company's power to issue bonds and/or preferred stock (ten jurisdictions), a ban on ownership of parent company shares by subsidiaries (ten jurisdictions), and minimum capitalization requirements (ten jurisdictions). Only three jurisdictions ban authorized but unissued shares (though limitations on the amount or period of valid authorization are more common); and only two ban transactions between companies and insiders. Finally, no jurisdiction sets the nominal (or par) value of shares equal to their issue price, as did an early draft of the Russian Civil Code and a competing company law draft in Russia. On the procedural side, the most common restriction on insider discretion is a shareholder vote on fundamental transactions such as mergers, liquidations, or recapitalizations. Such approval requirements exist in all seventeen statutes, although transactional coverage and quorum and vote thresholds vary greatly. Most jurisdictions also permit shareholders to remove directors between regular board elections (fifteen jurisdictions), mandate appraisal rights for dissenters in mergers or sales of assets (eleven jurisdictions), require preemptive rights for new issues of stock (eleven jurisdictions), and mandate a one share, one vote rule for common stock (nine jurisdictions). Less common are requirements for shareholder approval of large share issuances (five jurisdictions, including those that bar authorized but unissued shares) and self-interested transactions between companies and their officers or directors (five jurisdictions). Two prominent features of the Russian draft statute we proposed -- proportional board representation and put rights for minority shareholders in control transactions -- are found in only two and four jurisdictions respectively. Some protections in the draft Russian statute, such as a secret ballot in corporate elections, are not mandated by any existing statute. These summary checklists of substantive and procedural restrictions provide snapshots of the drafting styles and levels of regulation that characterize the company laws of emerging markets. Only one comparatively developed jurisdiction -- South Africa -- has an enabling statute, and even this law is more restrictive than the Delaware statute. Four jurisdictions rely heavily on substantive restrictions (Poland, the Czech Republic, Hungary, and Turkey). A group of four Asian and Latin American statutes are appropriately characterized as "mixed," 65 with fair numbers of both procedural and substantive restrictions. And a group of eight countries (half of which are Commonwealth jurisdictions) have statutes that offer many procedural protections for shareholders but impose few substantive restrictions. These company laws exhibit aspects of the drafting strategy that characterizes the self-enforcing model. No company law in our sample, however, contains as few substantive prohibitions or as many procedural protections as we recommend. a A pure self-enforcing law would not contain minimum capitalization requirements (or only de minimus ones). The Russian company law establishes minimum capitalization rules (as required by the Russian Civil Code), but the amounts are fairly small -- around $2,500 in charter capital for a "closed" company (with restrictions on share transfer) and $25,000 for an "open", typically public company (without restrictions on share transfer). 66 1. SELF-ENFORCING STATUTE: Russia (1995) 2. ENABLING STATUTE: South Africa (1973) I. Substantive Protections 1. Bans authorized but unissued shares. O O 2. Requires shares to be issued at par or nominal value. O O 3. Requires minimum capitalization. Xa O 4. Limits debt and preferred stock. O O 5. Bans share repurchases. O X 6. Bans self-dealing with directors or officers. O O 7. Bans subsidiary ownership of parent shares. O X TOTAL 1 2 II. Procedural Protections 1. Mandates one share/one vote rule. X O 2. Mandates proportional board representation. X O 3. Permits shareholder removal of directors without cause. X X 4. Requires shareholder approval of director/officer self-dealing. X O 5. Requires meaningful vote on fundamental transactions. X X 6. Mandates appraisal rights in mergers or sales of all assets. X X 7. Requires shareholder vote on large acquisitions. X O 8. Requires shareholder vote on large new stock issues. X O 9. Mandates preemptive rights. O O 10. Mandates minority put rights upon changes of control. X O 11. Requires confidential voting. X O TOTAL 10 3 67 3. PROHIBITIVE STATUTES Poland (1991) Hungary (1988) Turkey (1956) I. Substantive Protections 1. Bans authorized but unissued shares. X O X 2. Requires shares to be issued at par or nominal value. O O O 3. Requires minimum capitalization. X X X 4. Limits debt and preferred stock. X X X 5. Bans share repurchases. X O X 6. Bans self-dealing with directors or officers. X X O 7. Bans subsidiary ownership of parent shares. - X O TOTAL 5 4 4 II. Procedural Protections 1. Mandates one share/one vote rule. O X O 2. Mandates proportional board representation. O O O 3. Permits shareholder removal of directors without cause. X X X 4. Requires shareholder approval of director/officer self-dealing. O O O 5. Requires meaningful vote on fundamental transactions. X X X 6. Mandates appraisal rights in mergers or sales of all assets. O O O 7. Requires shareholder vote on large acquisitions. X O X 8. Requires shareholder vote on large new stock issues. X X O 9. Mandates preemptive rights. X O O 10. Mandates minority put rights upon changes of control. O X O 11. Requires confidential voting. O O O TOTAL 5 5 3 68 4. PROCEDURAL STATUTES (PARTLY SELF-ENFORCING) Malaysia (1990) Singa- pore (1988) Nigeria (1990) India (1956) Mexico (1981) Chile (1981) Egypt (1981) I. Substantive Protections 1. Bans authorized but unissued shares. O O O O O O O 2. Requires shares to be issued at par or nominal value. O O O O O O O 3. Requires minimum capitalization. O O O O X O X 4. Limits debt and preferred stock. X O O O O O X 5. Bans share repurchases. X X X X X X O 6. Bans self-dealing with directors or officers. O O O O O O O 7. Bans subsidiary ownership of parent shares. X X X X O X O TOTAL 3 2 2 2 2 2 2 II. Procedural Protections 1. Mandates one share/one vote rule. X X X X X X O 2. Mandates proportional board representation. O O O O X X O 3. Permits shareholder removal of directors without cause. X X X X X O X 4. Requires shareholder approval of director/officer self-dealing. X X X O O O X 5. Requires meaningful vote on fundamental transactions. X X X X X X X 6. Mandates appraisal rights in mergers or sales of all assets. X X X X X X X 7. Requires shareholder vote on large acquisitions. X O O X O O O 8. Requires shareholder vote on large new stock issues. X O O O O O O 9. Mandates preemptive rights. X O O X X X X 10. Mandates minority put rights upon changes of control. X X X O O O O 11. Requires confidential voting. O O O O O O O TOTAL 9 6 6 6 6 5 5 69 5. MIXED STATUTES China (1994) Argent- ina (1988) Brazil (1991) South Korea (1984) Indon- esia (1995) Czech Repub. (1992) I. Substantive Protections 1. Bans authorized but unissued shares. X O O O O O 2. Requires shares to be issued at par or nominal value. O O O O O O 3. Requires minimum capitalization. X X O X X X 4. Limits debt and preferred stock. X O X X X X 5. Bans share repurchases. O X X X X X 6. Bans self-dealing with directors or officers. O O O O O O 7. Bans subsidiary cross ownership of shares. O X X O O O TOTAL 3 3 3 3 3 3 II. Procedural Protections 1. Mandates one share/one vote rule. X O X O O O 2. Mandates proportional board representation. O O O O O O 3. Permits shareholder removal of directors without cause. O X X X X X 4. Requires shareholder approval of director/officer self-dealing. X O O O O O 5. Requires meaningful vote on fundamental transactions. X X X X X X 6. Mandates appraisal rights in mergers or sales of all assets. O X X O X O 7. Requires shareholder vote on large acquisitions. X O O X X O 8. Requires shareholder vote on large new stock issues. X O O O O X 9. Mandates preemptive rights. O X X X X X 10. Mandates minority put rights upon changes of control. O O O O O O 11. Requires confidential voting. O O O O O O TOTAL 5 4 5 4 5 4