THE LAW OF CONTINGENT CONTROL IN VENTURE CAPITAL Alvaro Pereira* Contingent control (CC) is a key enabler of startup growth and venture capital. To preserve adequate incentives and mitigate risks, venture finance deals distribute startups’ governance rights among investors and founders based on performance measures at different points in time. For example, granting greater decision-making powers to outperforming founders or depriving them of such prerogatives when they underperform. Crucially, these rights are distributed ex ante, through carefully designed contracts and securities, avoiding the costs and potential failure of future negotiations. This paper shows how corporate law determines the structure of CC: higher costs of structuring CC through bespoke securities, such as restricted shares or convertible preferred stock, incentivize the use of shareholders’ agreements and shadow governance structures in VC-backed companies. These findings demonstrate that cross-country differences in security design and capital structures are not only explained by the characteristics of transacting parties and deals or by tax regulation, but also by the regulation of non-listed companies, which has been evolving in the blind spot of legal and financial scholarship. The paper argues that corporate laws in entrepreneurial economies should be recalibrated to facilitate *Georgia State University, College of Law. For valuable comments and suggestions, I thank Robert Bartlett, Giuseppe Dari Mattiacci, Mark Gentile, Svitlana Lebedenko, Timothy D. Lytton, Nikolos Schillaci, Jamison R Munns, Gabriel Rauterberg, the editors of the Columbia Business Law Review, and the participants at the University of Pennsylvania’s Institute for Law and Economics’ Inaugural Junior Faculty Workshop and the University of Amsterdam’s and Young Scholars Initiative’s New Trends in Business, Corporate and Financial Law Conference. I am also grateful to Lazzu Shrestha for excellent research assistance. All shortcomings remain my own. 676 COLUMBIA BUSINESS LAW REVIEW [Vol. 2023 security design and discourage the disproportionate use of shareholders’ agreements. I. Introduction ............................................................... 676 II. The Essential Role of Contingent Control in Venture Capital Investments .................................................. 683 III. The Corporate Law Determinants of Control ........... 692 A. Separation of Cash Flow from Control................ 693 B. Voting and Board Representation ....................... 697 1. Voting and Board Powers ............................... 697 2. Class-Based Board Representation................ 700 C. Summary ............................................................. 702 IV. Contingent Control .................................................... 703 A. Restricted Common and Stock Options .............. 703 B. Convertible Preferred Stock ................................ 704 C. Shareholders’ Agreements................................... 706 D. Summary ............................................................. 710 V. Shareholders’ Agreements’ Threat to Corporate Law and VC Markets ......................................................... 711 A. Vertical Agreements ............................................ 711 B. Horizontal Agreements ....................................... 713 C. Hindering Transaction Cost Reductions and Judicial Scrutiny ................................................. 715 D. Summary ............................................................. 719 VI. Implications ............................................................... 719 A. Financial Contracting.......................................... 719 1. Theory ............................................................. 720 2. Empirical Studies ........................................... 721 B. Corporate Law Design ......................................... 724 VII. Conclusion .................................................................. 726 I. INTRODUCTION Venture capital (VC) is a form of private equity that focuses on financing young and innovative companies with No. 2] THE LAW OF CONTINGENT CONTROL 677 high risk and high growth potential.1 These (usually tech) startups are only a small portion of all private businesses, but one of growing importance worldwide.2 Their potentially disruptive ideas can boost innovation and economic growth.3 However, they face extraordinary difficulties to finance, given their lack of liquid assets and uncertainties over their outcomes.4 Because of its focus on these otherwise unfinanceable tech startups, VC is conceived as a powerful tool to stimulate entrepreneurship, innovation, and dynamism in capital markets.5 Indeed, some of the most disruptive technologies and organizations of the past three decades have been financed by VC, and VC-backed firms have developed products and entered the market faster.6 VCs have financed the emergence and scaling of most top publicly traded firms by market capitalization, and an ascending number of large private firms that currently attract investments from mutual funds, hedge funds, and even sovereign wealth funds.7 VC has 1 See generally Jeffrey M. Pollack & Thomas H. Hawver, Venture Capital, in WORLD ENCYCLOPEDIA OF ENTREPRENEURSHIP 642 (Léo-Paul Dana ed., 2021). 2 See Alvaro Pereira, Designing Startup Corporate Law: A Minimum Viable Product, 42 REV. BANKING & FIN. L. 367, 368 (2022) (explaining that VC is “consolidating massive customer bases and accumulating capital as only publicly traded companies were capable not too long ago.”). 3 See generally Josh Lerner & Ramana Nanda, Venture Capital’s Role in Financing Innovation: What We Know and How Much We Still Need to Learn, 34 J. ECON. PERSPS. 237 (2020). 4 See infra Part II. 5 Bronwyn H. Hall & Josh Lerner, The Financing of R&D and Innovation, in 1 HANDBOOK OF THE ECONOMICS OF INNOVATION 609 (Bronwyn H. Hall & Nathan Rosenberg eds., 2010). 6 Thomas Hellmann & Manju Puri, Venture Capital and the Professionalization of Start-Up Firms: Empirical Evidence, 57 J. FIN. 169 (2002). 7 See Lerner and Nanda, supra note 3, at 238, 240; see also Will Gornall & Ilya A. Strebulaev, The Economic Impact of Venture Capital: Evidence from Public Companies 15–55 (Stanford University Graduate School of Business Research Paper, Paper No. 3362, 2015), https://papers.ssrn.com/sol3/papers.cfm?abstract_id=2681841 [https://perma.cc/A5T6-2PS3]. 678 COLUMBIA BUSINESS LAW REVIEW [Vol. 2023 also influenced the emergence of new financial intermediaries, such as accelerators, incubators, and angel investors, expanding the range of startup companies (and innovative ideas) that can access external finance.8 Fostering VC is, thus, aligned with the goal of stimulating startup growth, and understanding its determinants is a prerequisite for sound economic policy in entrepreneurial economies. Three decades of financial theory and empirical studies have shown that a distinctive aspect of VC finance is control.9 VCs cap risks and procure returns within a given fund’s lifespan by actively participating in the governance of the companies in which they invest.10 Unlike other private equity investors, VCs do not always seize control. Instead, decision- making powers in portfolio companies are shared; more specifically, they are distributed among participants at different stages, based on observable and verifiable measures of financial and non-financial performance.11 Generally, founders are rewarded with higher decision-making rights when a company meets or exceeds expectations, and investors take hold of such rights, at the expense of founders, when companies underperform.12 Crucially, these allocations can be 8 Lerner & Nanda, supra note 3, at 237. 9 See generally Marco Da Rin, Thomas Hellmann & Manju Puri, A Survey of Venture Capital Research, in 2 HANDBOOK OF THE ECONOMICS OF FINANCE 573, 589–595 (George M. Constantinides, Milton Harris, & Rene M. Stulz eds. 2013); see also infra Part II. 10 See Pollack & Hawver, supra note 1, at 643. 11 See infra Part II; see generally Erik Berglöf, A Control Theory of Venture Capital Finance, 10 J. L. ECON. & ORG. 247 (1994); Thomas Hellmann, The Allocation of Control Rights in Venture Capital Contracts, 29 RAND J. ECON. 57 (1998); William W. Bratton, Venture Capital on the Downside: Preferred Stock and Corporate Control, 100 MICH. L. REV. 891 (2002). 12 Steven N. Kaplan & Per Strömberg, Financial Contracting Theory Meets the Real World: An Empirical Analysis of Venture Capital Contracts, 70 REV. ECON. STUD. 281, 282 (2003) (“In general, board rights, voting rights, and liquidation rights are allocated such that if the firm performs poorly, the VCs obtain full control. As performance improves, the entrepreneur retains/obtains more control rights. If the firm performs very well, the VCs retain their cash flow rights, but relinquish most of their control and liquidation rights.”). No. 2] THE LAW OF CONTINGENT CONTROL 679 made ex ante, avoiding the costs and potential failure of future negotiations.13 By making control contingent on performance, VCs not only protect their interests, but also alleviate problems of incomplete contracting and moral hazard, preserving the right incentives at critical moments and contributing to companies’ successes. A central way in which contingent control is structured in VC-backed companies is through a capital structure comprised of at least two types of shares.14 Investors, on the one hand, acquire shares with preferential rights, such as priority on proceeds of liquidations, redemption rights, anti- dilution protections, and automatic conversion to common shares, which allow them to expand or relinquish control, at will or when a condition is met.15 Founders, on the other hand, receive shares with restricted governance rights, which are enhanced or further restricted based on performance.16 The ability to design these types of securities and capital structure is crucial for the emergence and growth of innovative firms, as they might expand the range of efficient contracting between entrepreneurs and VCs—and, in turn, the number of startups capable of raising VC finance. Although companies’ governance and capital structures are ultimately delimited by corporate law, policymakers and both legal and financial contracting scholars have generally 13 D. Gordon Smith, The Exit Structure of Venture Capital, 53 UCLA L. REV. 315, 337 (2005) (“Rather than leaving control up for grabs, venture capitalists and entrepreneurs bargain explicitly for control through a combination of express allocation provisions and voting rights.”) 14 See infra Part IV. 15 See Kaplan & Strömberg, supra note 12, at 290-95 (describing the function of these rights and their prevalence in the U.S.). For an overview of the rights allocated to holders of preferred shares, see WILLIAM W. BRATTON, CORPORATE FINANCE: CASES AND MATERIALS 735–741 (7th ed. 2012). 16 Founders may also receive stock options, although those are imperfect substitutes and thus usually allocated to employees. See infra Section IV.A. See also Jess H. Chua & Richard S. Woodward, Splitting the Firm Between the Entrepreneur and the Venture Capitalist With the Help of Stock Options, 8 J. BUS. VENTURING 43 (1993). 680 COLUMBIA BUSINESS LAW REVIEW [Vol. 2023 overlooked the extent to which it may facilitate or discourage security design.17 An emerging legal literature examining corporate law incentives to VC around the world has found that several noteworthy reforms do not facilitate security design to structure contingent control.18 It is an indication that even in countries where VC is actively promoted, only a few extremely promising projects might be getting access to VC—those for which investors are willing to assume higher legal risks or in which parties settle for suboptimal governance and capital structures.19 Recent evidence also shows that VC-backed firms over-rely on private agreements to structure their governance,20 likely to circumvent legal barriers to security 17 Pereira, supra note 2, at 373 (exploring the main reasons for the lack of scholarly attention to the corporate law determinants of VC). 18 See, e.g., Alvaro Pereira, Simplified Corporations and Entrepreneurship, 21 J. CORP. L. STUD. 433 (2021); Casimiro A. Nigro & Luca Enriques, Venture Capital e Diritto Societario Italiano: un Rapporto Difficile, 2021 ANALISI GIURIDICA DELL’ECONOMIA 149; Lin Lin, Contractual Innovation in China’s Venture Capital Market, 21 EUR. BUS. ORG. L. REV. 101 (2020); Paolo Giudici & Peter Agstner, Startups and Company Law: The Competitive Pressure of Delaware on Italy (and Europe?), 20 EUR. BUS. ORG. L. REV. 597 (2019); Zenichi Shishido, Does Law Matter to Financial Capitalism: The Case of Japanese Entrepreneurs, 37 FORDHAM INT’L L.J. 1087 (2014). 19 See, e.g., Lin, supra note 18, at 101–102 (detailing the emergence of a now common clause in VC financing agreements in China that gives excessive powers to investors, diminishing trust: the valuation adjustment mechanism “provides investors with a right to adjust a portfolio company’s original valuation and to get compensation by cash or equity upon the occurrence of certain future events (such as failing to meet financial or non- financial performance indicators). . . . unlike American venture capital contracts, which are designed to encourage long-term, sustainable investor- entrepreneur relationships, VAMs are predominantly investors’ self-help mechanisms to address specific and serious investor protection issues in the transitional and less informed Chinese market.”) 20 See, e.g., Gabriel Rauterberg, The Separation of Voting and Control: The Role of Contract in Corporate Governance, 38 YALE J. ON REGUL. 1124 (2021). No. 2] THE LAW OF CONTINGENT CONTROL 681 design.21 This tendency of structuring “governance through contract,” which is reinforced by legal restrictions to security design, reduces transparency and exacerbates agency problems and transaction costs, threatening the development of corporate law and, ultimately, the industries’ own endurance.22 This paper sustains that corporate law can meaningfully contribute to expanding the scope of startups that can access VC finance by facilitating the structure of contingent control through security design. Specifically, by enabling the temporary and conditional restriction and enhancement of voting and board representation rights in different classes of shares. The paper identifies and discusses pervasive limits to private firms’ ability to define their governance and capital structures with these types of securities, notwithstanding increasing efforts to promote VC through corporate law reform. Potentially, such legal constraints influence the rate of new innovative startups that access VC across jurisdictions. Certainly, they encourage the use of contracts (e.g., shareholders’ agreements) over corporate law instruments to structure firms’ governance. Given the opacity of private agreements and their weakening of legal protections against agency problems, their use as substitutes for corporate law should be discouraged. While there is still much to learn about the role of contingent control and security design in venture finance, the available evidence merits a recalibration of corporate law statutes—or, at least, a thorough review, informed by current market practices. Two actions are deemed of first order. First, to remove or otherwise minimize legal constraints to security design in non-listed companies; in particular, limitations to the design of convertible preferred stock, including limits to multiple voting, board representation by class, and boards’ powers to issue shares. Second, to discourage the use of private agreements as substitutes of the articles of 21 See infra Sections IV.A and IV.B (discussing legal barriers across jurisdictions); Section IV.C (explaining the main strategies to circumvent those barriers with shareholders’ agreements.). 22 See infra Part V. 682 COLUMBIA BUSINESS LAW REVIEW [Vol. 2023 incorporation, charters, and bylaws, e.g., limiting the enforceability of clauses inconsistent with minimum governance standards determined by corporate law, such as appraisal rights. The paper’s analyses and comparative evidence also have implications for financial contracting, offering an alternative explanation to why the capital structures and securities used by VC-backed firms vary across countries. Thus far, the literature has attributed such differences to economic factors, such as stock market conditions or the sophistication of VCs,23 or even specific tax rules and practices.24 International studies that consider institutional and legal determinants more widely rely on indexes that capture rules applicable to listed firms (e.g., investor protection), that regulate aspects of private firms that are irrelevant to contingent control (e.g., registration requirements), or that are based on legal systems’ perceived quality prior to the proliferation of VC-oriented reforms (e.g., legality).25 The evidence reported and discussed in this paper shows that corporate law constraints to security design in VC-backed firms contribute to explaining the variations on securities and capital structures observed across countries, an aspect generally unaccounted by this influential literature. 23 Douglas Cumming & Sofia A. Johan, Security Design, in VENTURE CAPITAL AND PRIVATE EQUITY CONTRACTING: AN INTERNATIONAL PERSPECTIVE 319 (2nd ed. 2014). 24 Ronald J. Gilson & David M. Schizer, Understanding Venture Capital Structure: A Tax Explanation for Convertible Preferred Stock, 116 HARV. L. REV. 874 (2003). 25 See, e.g., Douglas Cumming, Daniel Schmidt & Uwe Walz, Legality and Venture Capital Governance Around the World, 25 J. BUS. VENTURING 54 (2010); Steven N. Kaplan, Frederic Martel & Per Strömberg, How Do Legal Differences and Experience Affect Financial Contracts?, 16 J. FIN. INTERMEDIATION 273 (2007); see also Josh Lerner & Antoinette Schoar, Does Legal Enforcement Affect Financial Transactions? The Contractual Channel in Private Equity, 120 Q. J. ECON. 24 (2005) (using an index measuring “time to enforce a contract” as a proxy for legal systems’ quality but overlooking crucial differences between financial contracts and those captured by the Index, e.g., liquidated damages). No. 2] THE LAW OF CONTINGENT CONTROL 683 The paper unfolds in six Parts. Part II explains how contingent control enables VCs to finance innovative startups. Part III shows that differences in the legal rules governing control (i.e., voting and board representation rights of various classes of shares) in non-listed firms across jurisdictions determine the range of options available for VCs and entrepreneurs to allocate control efficiently via security design. Part IV analyzes corporate law limits to structured contingent control through bespoke securities and shareholders’ agreements. The analysis reveals that legal constraints to security design, such as limitations to issue restricted and convertible-preferred stock, incentivize the use of shareholders’ agreements for control allocation. These agreements are suboptimal alternatives, as some provisions may not be enforceable, and their breach generally gives rights to contractual remedies. Part V shows that widespread uses of shareholders’ agreements in VC-backed startups also have damaging collateral effects that threaten the industry’s own endurance, weakening corporate law safeguards against agency problems and hindering transaction costs reductions. Part VI presents the main implications of these analyses for financial contracting and corporate law, highlighting shortcomings in theoretical and empirical studies of venture finance, and advocating for disincentives to the trend of structuring “governance through contract” in VC-backed firms, respectively. The last Part concludes. II. THE ESSENTIAL ROLE OF CONTINGENT CONTROL IN VENTURE CAPITAL INVESTMENTS The financing of young and innovative startup companies poses unique challenges.26 These firms generally lack the assets or revenue required for traditional debt finance.27 The 26 See generally Hall & Lerner, supra note 5. 27 MARCO DA RIN & THOMAS HELLMANN, FUNDAMENTALS OF ENTREPRENEURIAL FINANCE 1–10 (2020); Allen N. Berger & Gregory F. Udell, The Economics of Small Business Finance: The Roles of Private Equity and Debt Markets in the Financial Growth Cycle, 22 J. BANKING & FIN. 613 (1998). 684 COLUMBIA BUSINESS LAW REVIEW [Vol. 2023 risks are also excessively high for plain vanilla equity finance.28 Given that startup companies are private (and that much of the potential value increase comes from keeping information secret),29 there is no liquid market for shares, which significantly restricts diversification and exit opportunities for investors. There is also substantial uncertainty over returns because products and business models have not been tested in the market, and founders often lack management experience.30 Business ventures between entrepreneurs and investors emerged as a plausible solution to these challenges.31 Experienced financiers with superior industry knowledge minimize uncertainties by identifying the most promising entrepreneurial companies, acquiring shares and actively participating in their governance.32 The success of such business ventures, however, depends on founders’ and investors’ continuous efforts, a particularly challenging condition given that their interests are not always aligned.33 For example, while investors are primarily motivated by returns, entrepreneurs also regard some private benefits (e.g., personal satisfaction, reputation) as highly valuable, and the ability to pursue them might sometimes be a stronger 28 DA RIN & HELLMANN, supra note 27, at 1–10; Berger & Udell, supra note 27; see also Paul A. Gompers, Optimal Investment, Monitoring, and the Staging of Venture Capital, 50 J. FIN. 1461 (1995). 29 Robert Cooter & Hans Bernd Schäfer, The Secret of Growth Is Financing Secrets: Corporate Law and Growth Economics, 54 J.L. & ECON. S105 (2011). 30 See David H. Hsu, Experienced Entrepreneurial Founders, Organizational Capital, and Venture Capital Funding, 36 RSCH. POL’Y 722 (2007); see also DA RIN & HELLMANN, supra note 27, at 45–48 (discussing how the entrepreneurial team’s experience conditions access to venture capital finance). 31 For a historical account of the emergence and development of these business ventures, see TOM NICHOLAS, VC: AN AMERICAN HISTORY (2019). 32 See Berger & Udell, supra note 27; William A. Sahlman, The Structure and Governance of Venture Capital Organizations, 27 J. FIN. ECON. 473 (1990). 33 See Oliver Hart, Financial Contracting, 39 J. ECON. LITERATURE. 1079 (2001). No. 2] THE LAW OF CONTINGENT CONTROL 685 incentive than the financial one. The differences between what each of them contributes to and expects from the venture, as well as the information that each of them holds or can access and credibly commit to sharing with the other (information asymmetries), create a natural distrust that ultimately threatens the finance of high-risk and potentially high-growth firms.34 Developing contractual solutions to problems derived from founder-investor’s incentive misalignments and distrust is desirable, but difficult. First, distributing cash flow and control rights is challenging, given uncertainties over the value of entrepreneurs’ ideas and potential outcomes.35 Investors, who can quantify their contribution, may secure higher controlling rights. But in doing so, they may weaken entrepreneurs’ trust and effort, thereby threatening the potential success of the business.36 Second, even when parties reach a balanced agreement, business ventures inevitably experience eventualities that are impossible to foresee and regulate, and that might affect each of them differently. Hence, their relationship is perceived to be governed by incomplete contracts.37 Entrepreneurs and investors could ease the risks of incomplete contracts by strategically allocating decision- making authority to the party that is less prone to act opportunistically or engage in moral hazard when it matters the most.38 Still, identifying how different eventualities might 34 Cooter & Schäfer, supra note 29. 35 In fact, there is an ongoing debate in the academic literature on whether the organizational capital produced by successful teams of founders should be treated as endowment or investment, given that in the latter case there is a “possibility of over-or-under investment.” Hsu, supra note 30, at 723. For an analysis and empirical evidence on how founders’ experience influences startups’ valuation, see id. See also DA RIN AND HELLMANN, supra note 27, at 45–48. 36 Hart, supra note 33. 37 For an overview of this literature and its implications, see Philippe Aghion & Richard Holden, Incomplete Contracts and the Theory of the Firm: What Have We Learned over the Past 25 Years?, 25 J. ECON. PERSPS. 181 (2011). 38 Id.; see also Hart, supra note 33. 686 COLUMBIA BUSINESS LAW REVIEW [Vol. 2023 change the incentives of different participants has proven difficult, even theoretically. Financial contracting theorists have considered an array of control allocations that might alleviate these problems and lead to socially efficient outcomes, i.e., those in which the payoffs of both are maximized. One of the most influential models of control in financial contracting was developed by Philippe Aghion and Patrick Bolton, who—in very broad terms—concluded that an efficient contract should allocate control to financiers in events when ruthless value maximization is more efficient, and to entrepreneurs when it is not.39 If Aghion and Bolton’s model were correct, contingent control could be an appropriate solution to the identified challenges of financing startups, so long as entrepreneurs and investors could reasonably anticipate the occurrence of such events. Building on Aghion and Bolton’s framework, a rich literature has developed to explore (and basically confirm) the predictive value of their model for some of the most sensitive decisions in VC investments, such as exit options.40 In essence, this literature maintains that changes in firms’ performance are a useful identifier of changes in participants’ incentives. Thus, based on expected performance, they can reasonably anticipate who would be in a better position to make socially efficient decisions at different points in time, and allocate control rights to that party ex ante, avoiding potentially fruitless future renegotiations. Despite these valuable theoretical insights, the extent to which contingent control may support socially efficient agreements in real life remained undetermined for several 39 Philippe Aghion & Patrick Bolton, An Incomplete Contracts Approach to Financial Contracting, 59 REV. ECON. STUD. 473 (1992). To be clear, at 491–92, they submit that “it is always best to start first with entrepreneur control if that is feasible. If, however, entrepreneur control does not sufficiently protect the investor’s claims, one should go for contingent control. Finally, if that is still not enough to protect the investor’s interests, one wants to give full control to the investor.” Id. 40 See e.g., Hellmann, supra note 11. For a literature review, see Da Rin, Hellmann & Puri, supra note 9. No. 2] THE LAW OF CONTINGENT CONTROL 687 years, and a key question, unanswered: is contingent control at all possible or useful for VC financing? Steven Kaplan and Per Strömberg sought to fill this gap by studying real entrepreneur-investor agreements in the U.S., the most developed VC market.41 Their paradigmatic empirical study shows how successful VC firms use contingent control allocation—and complementary contractual terms—to overcome the challenges of financing innovative firms.42 Their reported evidence offers two fundamental lessons on how the strategic allocation of control enables financing startups with VC in real life. First, VCs always separate cash flow from decision-making (or control) rights.43 In other words, agreements require a substantive deviation from the one- share one-vote rule.44 Second, the allocation and exercise of control rights in VC-backed firms is contingent on observable and verifiable measures of financial and non-financial performance, such as company valuation in subsequent financing rounds.45 By attaching control allocation to performance, they anticipate which party might be more prone to moral hazard and deprive them of decision-making authority ex ante. In this sense, their study confirms Aghion and Bolton’s prediction that, if allowed, investors will seek an optimal venture finance agreement by allocating control to different participants at different moments. Unlike Aghion and Bolton, however, Kaplan and Strömberg find that real-life investors do not always seek control when ruthless value maximization could be efficient (e.g., when firms outperform and cash flows are more important than private benefits), but 41 See Kaplan & Strömberg, supra note 12. 42 Id. Other findings relate to contingent allocation of cash-flow rights and the use of non-compete clauses in financial contracts. 43 Id. at 295 (noting that the VC financings explored in their study “allow VCs to separately allocate cash flow rights, board rights, voting rights, liquidation rights, and other control rights.”). 44 For a comparative examination of this rule, see infra Section III.A. 45 Kaplan & Strömberg, supra note 12, at 294 (“[C]ontrol rights are contingent on subsequent measures of financial performance, non-financial performance, and actions.”) 688 COLUMBIA BUSINESS LAW REVIEW [Vol. 2023 often relinquish it to entrepreneurs in those instances.46 This evidence further suggests that investors make ex ante considerations of efforts, in light of which entrepreneurs are rewarded.47 The nature of the evidence reported by Kaplan and Strömberg in the cited paper is not only relevant for financial contracting scholarship, but also insightful to understanding the legal structure of contingent control in VC. First, their sample is modestly “biased towards successful investments” and firms, providing a unique perspective of the legal instruments used by some of the most experienced investors to resolve key issues in venture finance.48 Secondly, their study identifies the recurrent use of various types of securities—in particular, bespoke securities, such as convertible-preferred stock—and private agreements to structure contingent control.49 In a nutshell, they show how the most sophisticated parties, in the most mature VC market, structure contingent control through customized securities and private agreements, which is consistent with findings in similar studies.50 Whilst these theoretical and empirical financial contracting studies identified the use of bespoke securities and private agreements to structure contingent control, a pressing question still lacks a conclusive answer: why do 46 Kaplan & Strömberg, supra note 12, at 295 (“[R]ights are allocated such that . . . [i]f the company performs very well, the VCs relinquish most of their control . . . .”). 47 It thus confirms theoretical models of contingent control (not of cash flow allocation) of Berglöf, supra note 11 and of Klaus M. Schmidt, Convertible Securities and Venture Capital Finance, 58 J. FIN. 1139 (2003). 48 Indeed, despite controlling for differences among VC partnerships, the authors acknowledge the likelihood of a bias, grounded in the fact that rates of IPO where higher in the studied period, that the selected investments provided returns above the market, and that the VC firms participating in it were above average. Kaplan & Strömberg, supra note 12 at 283, 285. 49 Id. at 286. 50 See, e.g., Hellmann & Puri, supra note 6; Douglas J. Cumming, Capital Structure in Venture Finance, 11 J. CORP. FIN. 550 (2005). No. 2] THE LAW OF CONTINGENT CONTROL 689 specific legal instruments (e.g., convertible-preferred stock) prevail in some jurisdictions and not in others? Three sets of accounts are salient in the literature.51 First, an agency cost explanation. In line with Aghion and Bolton’s tradition, it purports that the types of securities and agreements used in different venture deals reflect parties’ expected agency problems.52 A voluminous empirical literature has concluded that a combination of traditional and bespoke convertible securities contributes to mitigating agency problems and preserving the right incentives over time.53 The differences between the mix of securities observed across jurisdictions would, thus, be explained by differences in the characteristics of participants (e.g., entrepreneurs’ experience, investor sophistication) and deals (e.g., early stage, buyout) that influence expected agency costs and the private ordering solutions devised to mitigate them.54 Second, a tax explanation. Gilson and Schizer compared financial contracting trends in the U.S. and Canada and concluded that a special type of security, convertible-preferred stock, prevailed in the U.S. due to a combination of tax regulatory gaps and practices (from both the lawyers and the authorities) which was absent in Canada.55 Their account suggests that tax regulations and practices may explain variations in the legal structure of VC finance, in particular, the types of 51 For a discussion and empirical analysis of additional explanations that are beyond the scope of this paper, see Cumming & Johan, supra note 23, at 319–22. 52 See Hellmann, supra note 11; Berglöf, supra note 11. 53 See, e.g., Thomas Hellmann, IPOs, Acquisitions, and the Use of Convertible Securities in Venture Capital, 81 J. FIN. ECON. 649 (2006); Francesca Cornelli & Oved Yosha, Stage Financing and the Role of Convertible Securities, 70 REV. ECON. STUD. 1 (2003); Thomas H. Noe & Michael J. Rebello, Asymmetric Information, Managerial Opportunism, Financing, and Payout Policies, 51 J. FIN. 637 (1996). 54 See CUMMING & JOHAN, supra note 23, at 319–68 (discussing empirical evidence on how certain agency costs, such as moral hazard or adverse selection, vary depending on the stage and structure of the investment and on the characteristics of transacting parties). 55 Gilson & Schizer, supra note 24, at 888–92. 690 COLUMBIA BUSINESS LAW REVIEW [Vol. 2023 securities most commonly used in different VC markets.56 Third, a legal explanation. A number of empirical studies have sought to explore how differences between legal systems influence VC financing by analyzing deal information provided by VCs in various countries.57 These studies have found significant correlations between various general measures of the quality of the law (e.g., shareholder and creditor protection) and the governance and capital structure of VC-backed firms (e.g., the prevalence of convertible- preferred stock), which suggest that legal institutions influence bargains between entrepreneurs and investors. While insightful, these explanations are still inconclusive. For one, they are based on minimal but fundamental assumptions of what entrepreneurs and investors can legally bargain for and agree upon across jurisdictions. For example, their ability to design securities to structure contingent control is not generally questioned. Notably, the empirical studies exploring the role of legal institutions use measures of the quality of the law that have proven inaccurate, such as those that classify legal systems by legal families or legal origins.58 In this literature it is also common to use international legal indexes that, despite covering a wide range of jurisdictions, do not capture the legal rules that govern 56 See CUMMING & JOHAN, supra note 23, at 356 (finding, based on statistical analyses of 12,363 venture deals in Canada, between 1991 to 2003, that “higher capital gains taxes lower the probability that convertible preferred equity is used,” which is consistent with Gilson & Schizer, supra note 24). 57 See, e.g., Theodore A. Khoury, Marc Junkunc & Santiago Mingo, Navigating Political Hazard Risks and Legal System Quality: Venture Capital Investments in Latin America, 41 J. MGMT. 808 (2015); Cumming et al., supra note 25; Laura Bottazzi, Marco Da Rin & Thomas Hellmann, What is the Role of Legal Systems in Financial Intermediation? Theory and Evidence, 18 J. FIN. INTERMEDIATION 559 (2009). 58 See Holger Spamann, The “Antidirector Rights Index” Revisited, 23 REV. FIN. STUD. 467, 470, 477 (2010); John Armour, Simon Deakin, Prabirjit Sarkar, Mathias Siems & Ajit Singh, Shareholder Protection and Stock Market Development: An Empirical Test of the Legal Origins Hypothesis, 6 J. EMPIRICAL LEGAL STUD. 343 (2009); Katharina Pistor, Patterns of Legal Change: Shareholder and Creditor Rights in Transition Economies, 1 EUR. BUS. ORG. L. REV. 59 (2000). No. 2] THE LAW OF CONTINGENT CONTROL 691 entrepreneur-investor relationships or that were even binding in the period of interest. For instance, the anti-director index, the dominant measure of shareholder protection used in venture finance research, is based on laws that govern publicly traded companies,59 not the private startups in which VCs invest. The legality index, often perceived as a more comprehensive assessment of legal differences for covering other factors affecting how companies are governed and financed (e.g., effectiveness of the judiciary or the rule of law), could also be misleading: it captures the legal rules that were binding between 1980 and 1995, a period prior to the one in which VC spread globally, and its assessment of corporate laws is based on the cited anti-director index.60 Significantly, neither of these approaches has considered differences in corporate laws applicable to non-listed firms, which govern VC-backed companies’ internal affairs and would be expected to have some influence in the legal instruments selected to assemble their governance and capital structures. The main reason for the lack of interest in corporate law is that, at least for non-listed firms, it is presumed to be comprised of flexible rules from which parties can opt out, rendering cross-jurisdictional differences uninfluential.61 However, a comparative review of corporate laws in highly entrepreneurial economies reveals pervasive (often expanding) and under-examined differences in the rules that govern security design and contingent control, more generally. The next Part examines those differences and details how 59 For a detailed and critical examination of the methodology, see Spamann, supra note 58. See also Priya P. Lele & Mathias M. Siems, Shareholder Protection: A Leximetric Approach, 7 J. CORP. L. STUD. 17 (2007). 60 Daniel Berkowitz, Katharina Pistor & Jean-Francois Richard, Economic Development, Legality, and the Transplant Effect, 47 EUR. ECON. REV. 165, 181–82 (2003) (“In order to measure legality, we first use the same survey data measuring the effectiveness of the judiciary, rule of law, the absence of corruption, low risk of contract repudiation and low risk of government expropriation observed during 1980–95 employed by [La Porta, Lopez-de-Silanes, Shleifer and Vishny]”). 61 Pereira, supra note 2. 692 COLUMBIA BUSINESS LAW REVIEW [Vol. 2023 they influence the allocation and legal structure of control in VC-backed companies. III. THE CORPORATE LAW DETERMINANTS OF CONTROL Successful venture financial agreements have two characteristics: (i) they separate cash flow from control rights, and (ii) they condition the allocation and exercise of those rights to observable and verifiable measures of financial and non-financial performance. This strategy resembles a socially efficient contract—one in which decision-making authority is allocated to the party that is more likely to use it in the interest of the venture at crucial moments, strengthening incentives to maximize the efforts invested in the firm. Empirical studies,62 legal practitioners,63 and VC organizations,64 attest that the separation of cash flow from control rights and the conditional allocation of such rights to founders, employees, and investors—i.e., contingent control— is legally achieved through a combination of contracts and 62 See supra Part II. 63 See Startup Venture Finance: Overview, THOMSON REUTERS PRACTICAL LAW UK (Aug. 1, 2020), https://uk.practicallaw.thomsonreuters.com/w-000- 4934?originationContext=document&transitionType=DocumentItem&cont extData=(sc.Default)&ppcid=b1ad59ec64c847cab3ad70e3ac872c71&comp= pluk&OWSessionId=845a7fbceba144d596dfb3d315ca3407&skipAnonymou s=true&firstPage=true [https://perma.cc/4WJD-NL88] (presenting a comparative market and regulatory overview developed by practitioners of 10 European and North American jurisdictions). 64 See, e.g., Model Documents For Early-Stage Investments, BRIT. VENTURE CAP. ASS’N (Feb. 2023), https://www.bvca.co.uk/Policy/Tax-Legal- and-Regulatory/Industry-guidance-standardised-documents/Model- documents-for-early-stage-investments [https://perma.cc/N65B-D2SL]; Michael Reid & Christopher Pejovic, CVCA Model Documents: Use Freely and With Caution, DLA PIPER (Mar. 22, 2021), https://www.dlapiper.com/en/canada/insights/publications/2021/03/cvca- model-documents [https://perma.cc/A977-GB7W]; Model Legal Documents, NAT’L VENTURE CAP. ASS’NS, https://nvca.org/resources/model-legal- documents/ [https://perma.cc/99TA-UGSU] (last visited June 1, 2023). No. 2] THE LAW OF CONTINGENT CONTROL 693 bespoke securities, including common shares with restrictions and various types of preferred shares. By defining the scope of corporate governance arrangements via contracts or security design, corporate law determines the range of bargains and agreements on control in VC investments (and, therefore, the range of firms that can be financed with it) in at least three ways. First, it determines the extent to which cash flow and control rights can be separated. Second, it governs shareholders’ voting and board representation rights, the main forms of control in VC finance. Third, corporate law determines whether exercising those rights can be contingent on performance measures. This Part considers how different regulatory configurations might limit the range of efficient contracting in the first two categories. Part IV deals with contingent control. A. Separation of Cash Flow from Control A default provision in corporate law is that each share encompasses an equal amount of cash flow and control rights.65 This rule, commonly referred to as one-share one- vote, is generally justified in the notion that decision-making power should match economic incentives; in other words, shareholders should be able to voice their opinion in proportion to their owned risk capital.66 The benefits and perils of deviating from this rule have been considered by a vast literature on the theory of the firm and, more recently, by the comparative corporate governance literature.67 Still, much of this work has focused on listed firms and problems specific to them, and not on its impact on non-listed companies with fast-growing valuations and stakeholders. 65 Frank H. Easterbrook & Daniel R. Fischel, Voting in Corporate Law, 26 J.L. & ECON. 395, 399 (1983) ; see also Marco Ventoruzzo, Un’azione, un voto: un principio da abbandonare?, 42 GIURISPRUDENZA COMMERCIALE 512 (2015) (It.). [https://iris.unibocconi.it/handle/11565/3985572] [https://perma.cc/778T-D5A7]. 66 Mike Burkart & Samuel Lee, One Share-One Vote: The Theory, 12 REV. FIN. 1 (2008). 67 See id. 694 COLUMBIA BUSINESS LAW REVIEW [Vol. 2023 Legal scholars often assume that non-listed companies— those in which VCs invest— enjoy contractual freedom to separate cash flow from control across jurisdictions.68 This partly explains why, unlike the study of listed firms and despite its relevance for VC, there is no equivalent comprehensive comparative research on the matter.69 Yet, a closer look reveals that there are significant limits to these prerogatives in numerous jurisdictions and, consequently, barriers to structure socially efficient VC agreements. One-share-one-vote, in fact, is quite a rigid principle for non-listed corporations in many jurisdictions. A common expression of that rigidity is found in legal systems that only allow the issuance of preferred shares with double dividends and no votes, or vice-versa.70 Other systems provide wider flexibility but impose temporal restrictions to multiple votes71 or limits to the percentage of shares with such enhanced governance rights.72 In those legal systems, parties still have room to depart from the one-share one-vote rule, but within a limited range of bargaining, as they cannot fully exchange or “sell” voting rights. These regulatory environments diminish the development of the VC industry. They might, for example, discourage 68 See, e.g., REINIER KRAAKMAN, JOHN ARMOUR, PAUL DAVIES, LUCA ENRIQUES, HENRY HANSMANN, GERARD HERTIG, KLAUS HOPT, HIDEKI KANDA, MARIANA PARGENDLER, WOLF-GEORG RINGE & EDWARD ROCK, THE ANATOMY OF CORPORATE LAW: A COMPARATIVE AND FUNCTIONAL APPROACH 18–19 (3rd ed. 2017). 69 For a more detailed discussion on the reasons for that research gap, see Pereira, supra note 2, at 388–95. 70 See, e.g., Ley No. 26887 Ley General de Sociedades (Peru); Polish Commercial Code (Kodeks Spółek Handlowych), arts. 351–65. 71 See, e.g., Brazilian Law No. 6,404 (1976), art. 110, temporarily modified by Medida Provisória No. 1.040 (2021) (providing that multiple votes could be granted for seven years, extendable only once) and later modified by Law 14,195 (2021) (enabling extension more than once, but subject to specific conditions, including dissenters’ buyout).[https://www.planalto.gov.br/ccivil_03/Leis/L6404consol.htm][http s://perma.cc/ZZ54-VB46] 72 See, e.g., Companies Act, 2013, § 43 (India); Companies (Share Capital and Debentures) Rules, 2014, § 4(1)(c) (India) (capping shares with differentiated rights at 26%). No. 2] THE LAW OF CONTINGENT CONTROL 695 investors, by limiting their bargaining tools, or compel entrepreneurs to incorporate or relocate elsewhere.73 In larger markets, as China, these rules might not entirely discourage investors, but likely induce them to develop contractual protections that do not necessarily strengthen incentive alignment or trust between parties.74 Still, many entrepreneurs do not always register their startups as corporations, but often as limited liability companies or other less expensive legal entity forms. Despite reducing expenses, such entities usually have even higher restrictions to the separation of cash flow from control that are relevant for the structure of socially efficient financing agreements with VCs.75 In many civil law countries, limited liability companies do not divide the capital into shares but into quotas, which are traditionally more difficult to customize.76 In case of conflict, judges may not necessarily rely on the law of corporations, but on a less flexible legal framework.77 Because these legal forms usually have preferential tax treatment,78 many founders are compelled to register their businesses as such. In those cases, an optimal VC financing agreement could be even more elusive, requiring 73 Giudici & Agstner, supra note 18, at 624–25 (reporting evidence of some high-profile Italian startups that, to raise VC finance, set up a U.S. corporation which controls an Italian entity). 74 Lin, supra note 18, at 133–34. 75 See J. William Callison, Venture Capital and Corporate Governance: Evolving the Limited Liability Company to Finance the Entrepreneurial Business, 26 J. CORP. L. 97, 123 (2000) (concluding that the limited liability companies’ governance attributes are the main reason preventing their use for venture capital investments, as they “either fail to separate equity ownership and control or, if they make such separation, fail to differentiate between operational control functions and ratification and monitoring functions”). 76 Pereira, supra note 18, at 435. 77 Giudici & Agstner, supra note 18, at 597; Pereia, supra note 18. 78 See Joseph A. McCahery, Erik P. M. Vermeulen & Priyanka Priydershini, A Primer on the Uncorporation, 14 EUR. BUS. ORG. L. REV. 305 (2013) (discussing “partnership-type” forms, emerging alternatives to the corporate form that provide enhanced flexibility and pass-through taxation, but also governance challenges for startups—e.g., those derived from equal- sharing rules). 696 COLUMBIA BUSINESS LAW REVIEW [Vol. 2023 the conversion of the legal entity into a corporation (and relinquishing the tax benefit) or allocating control to the stronger party in the bargain (instead of making it contingent). Although there is a trend toward enabling the separation of cash flow from control rights in all types of private companies, absolute freedom is far from being the standard norm. For example, the latest Italian reform explicitly enabled multiple voting in limited liability companies “within the limits imposed by law,”79 which, according to some legal scholars, indicates that it is a maximum of three votes per share and, in any case, shares with multiple votes cannot exceed half of the capital.80 Portugal preserved similar restrictions after its 2015 reform.81 In March 2021, Brazil issued an emergency decree directed to rapidly improve its business regulatory environment, enabling multiple voting rights for up to ten per share, for a maximum of seven years, extendable only once upon shareholder approval.82 These amendments were included in a permanent legal reform that allowed the extension more than once, but subject to specific conditions.83 As initially stated, the main economic justification of the one-share one-vote rule is that decision-making power should match financial incentives, which is sensible for most businesses. However, there is evidence of certain socially efficient arrangements departing from that rule in public 79 Giudici & Agster, supra note 18, at 615. 80 Id. at 620. 81 Daniela Baptista, Ações preferenciais sem voto (em particular, as detidas por investidores qualificados), IV CONGR. DIREITO SOC. EM REV. 411 (2016). 82 See Medida Provisória No. 1.040 (2021); Clarissa Freitas & Amanda Costa Vella, MP 1,040 and the Debate on Plural Voting, MACHADO MEYER (July 19, 2021), https://www.machadomeyer.com.br/en/recent- publications/publications/corporate/mp-1-040-and-the-debate-on-plural- voting [https://perma.cc/GXP6-RGRH]. 83 See Brazilian Law No. 14,195 (2021). No. 2] THE LAW OF CONTINGENT CONTROL 697 markets,84 and exceptions to that rule have been instrumental in successful VC financings.85 Hence, while the balance between rigid and flexible regulation is context-specific, strict one-share one-vote rules in non-listed corporations merit a renewed review, commensurate with the goal of expanding the range of efficient contracting between entrepreneurs and investors, and developing VC markets. B. Voting and Board Representation A second way corporate law determines the range of bargains and agreements on control in VC investments is by regulating how voting and board representation rights can be distributed among various classes of shares. Generally, all shares enjoy voting rights to approve fundamental transactions (i.e., those that affect the relationships among constituents) and to decide who will sit on the board of directors.86 These decision-making rights distinguish equity from debt,87 and are governed by corporate law. To the extent that these rights might be restricted or expanded in certain circumstances, cross-jurisdictional differences are vital for VC. 1. Voting and Board Powers To enable socially efficient financing agreements, VCs bargain with entrepreneurs over how voting rights are allocated to different classes of shares. On the most restrictive side of the voting rights spectrum, VC-backed firms may issue 84 Burkart & Lee, supra note 66. See also Aurelio Gurrea-Martínez, Theory, Evidence, and Policy on Dual-Class Shares: A Country-Specific Response to a Global Debate, 22 Eur. Bus. Org. L. Rev. 475 (2021). 85 In fact, it is the common factor in VC deals worldwide, regardless of differences in security design. See Cumming et al., supra note 25. 86 KRAAKMAN ET AL., supra note 68, at 49–61; see also DA RIN & HELLMANN, supra note 27, at 301–305. 87 To be clear, they distinguish one aspect, i.e., “control,” as these securities also differ in how they allocate cash-flow rights. See Oliver E. Williamson, Corporate Finance and Corporate Governance, 43 J. FIN. 567 (1988); see also Hart, supra note 33, at 1084. 698 COLUMBIA BUSINESS LAW REVIEW [Vol. 2023 shares without any of these rights and with additional restrictions on transferability.88 These “voteless shares” may be issued as incentives to employees,89 allowing them to benefit from—and contribute efforts towards—a successful trade sale without disrupting the allocation of control and its intended consequences. An indirect and more flexible limitation to common shareholders’ rights is to authorize the board of directors to issue new shares and nominate board members, two of the most sensitive decisions taken at the general shareholder meeting. Although founders are usually members of the board and the management team,90 these restrictions could significantly weaken them (e.g., if they only own common shares and the company’s underperformance forces changes in the leadership, they might lose their formal influence in the company). Shares may also be altered to enhance voting rights. On this expansive end of the voting rights spectrum are convertible-preferred stock (CPS), which have multiple votes and the right to nominate and elect, as a class, a board member.91 The latter becomes of paramount importance when firms empower the board to issue preferred shares and to autonomously determine the privileges attached to them (i.e., blank-check-preferred),92 which is perhaps the most expansive form of control allocation. In practice, holders of CPS may use these rights to negotiate and approve new equity financing in ways that dilute some shareholders without their consent. 88 JOSEPH W. BARTLETT, FUNDAMENTALS OF VENTURE CAPITAL 43–45 (1999). 89 Hellmann & Puri, supra note 6, at 177. 90 In fact, it is what Aghion and Bolton identify as preferable allocation of control, as the success of the firm in the initial stage mostly depends on entrepreneurs’ efforts. See Aghion & Bolton, supra note 39. 91 See generally William W. Bratton & Michael L. Wachter, A Theory of Preferred Stock, 161 UNIV. PA. L. REV. 1815 (2013). See also infra Section IV.B. 92 BARTLETT, supra note 88, at 45. No. 2] THE LAW OF CONTINGENT CONTROL 699 VC financing agreements seldom incorporate all these alterations to voting and board representation rights.93 Still, the extent to which corporate law enables founders and investors to bargain over them determines the range of efficient contracting and, therefore, the range of firms that can be financed through VC. Surprisingly, there are more restrictions than one would expect in an increasingly globalized VC market. Many jurisdictions only allow temporary restrictions on voting rights or require shares with such restrictions to have enhanced cash flow rights.94 In these legal environments, participants may seek to restrict common stock’s voting power by authorizing the board to issue shares and nominate new board members, as outlined above. However, this alternative is challenging because boards’ powers are generally limited to issuing pre-authorized shares, with defined rights, for a limited time that is often inferior to the ten-year life of a VC fund (e.g., in Switzerland, the period is two years; in the Netherlands, five).95 Moreover, such extensions or expansions of boards’ rights must be introduced through constitutive documents’ amendments that generally require approval by all classes of shares,96 defeating the goal of constraining common stock’s control rights. Empowering the board to nominate board members, on the other hand, is less problematic; but parties must carefully determine whether the law requires explicit authorization in the articles of incorporation or the charter, and what specific requirements must be fulfilled, which also varies across jurisdictions.97 All in all, VCs’ ability to bargain over restrictions to common rights in startups of their portfolio 93 In fact, as outlined above, patterns differ across countries and various combinations of securities constitute the final agreement resulting from multiple bargains. Kaplan et al., supra note 25. 94 See supra Section III.A. 95 See OBLIGATIONENRECHT [OR], CODE DES OBLIGATIONS [CO], CODICE DELLE OBBLIGAZIONI [CO], (Code of Obligations) Mar. 30, 1911, SR 220, RS 220, art. 651, 654 (Switz.); Artikel 2:96 para. 1 BW. (Neth.). 96 See KRAAKMAN ET AL., supra note 68, at 174–79. 97 Id. at 53–55. 700 COLUMBIA BUSINESS LAW REVIEW [Vol. 2023 vary across jurisdictions, with relevant and infrequently discussed limitations (e.g., to empower the board) significantly reducing the range of efficient contracting or increasing the costs of bargaining. Bargains over enhanced CPS’ voting rights could compensate for such difficulties, although certain limits must also be accounted for. Multiple voting is generally restricted to two votes per preferred share and, in many cases, forbidden for shares with enhanced cash flow rights.98 A sensible strategy to circumvent the first kind of restrictions could be to issue shares that vote on an “as if converted” basis. Yet, such a strategy might prove difficult to implement through security design in certain jurisdictions because it relies on the possibility to create a right to automatic conversion from CPS into common shares, which may not be at all possible, as detailed in Section IV.B below. 2. Class-Based Board Representation An alternative tactic to enhance CPS’ control rights is to allow different classes of shares to elect different members of the board—so-called, constituency directors. A common structure is to have some members of the board elected by CPS (i.e., investors), others by non-preferential shareholders (i.e., founders) and, lastly, “independent” directors, elected by both classes of shares.99 This composition is aimed at ensuring that VCs, albeit minority shareholders, participate in the 98 See supra Section III.A. 99 See Brian Broughman, Independent Directors and Shared Board Control in Venture Finance, 9 REV. L. & ECON. 41 (2013) (showing that independent directors act as a tiebreaker, contributing to reduce holdup problems by “moderating each party’s ex post threat position, potentially expanding the range of firms which receive external financing”); Brian Broughman, The Role of Independent Directors in Startup Firms, 2010 UTAH L. REV. 461 (arguing that “heightened fiduciary protections” could undermine independent directors’ ability to mediate and exercise their tie- breaking role). No. 2] THE LAW OF CONTINGENT CONTROL 701 deliberation and approval of the most sensitive decisions, such as hiring executives and issuing shares.100 While practical, this distribution of control rights has unresolved issues. In some jurisdictions, board nomination and election by class is only allowed in exceptional circumstances, such as when dividends of non-voting shares are repeatedly retained.101 In most jurisdictions, however, such a distribution is explicitly forbidden or not legally binding, as boards are expected to represent (and be elected by) all shareholders.102 To circumvent this restriction, VCs use shareholders’ agreements providing that all shareholders would vote for the nominee of a given investor, but such agreements might not be easily (or at all) enforceable,103 and over-reliance on them could have adverse long-term consequences.104 Now, even when boards are indeed integrated with constituency directors, it is still uncertain whether the right to elect board members can be useful for exercising control when it is most relevant, i.e., in exit decisions. In the U.S., where it is common for VC-backed firms to have class-elected board members, influential Delaware case law dictates that boards should prioritize the interests of common shareholders (not their “electors”), effectively preventing the exercise of 100 While boards’ powers vary across legal systems and companies, their approval is legally required for fundamental transactions (e.g., mergers) in most jurisdictions. Boards also have the authority to hire and fire executives and may even issue shares with preferred rights. For empirical evidence on how VCs use these powers, see Hellmann & Puri, supra note 6 (finding that VC-backed companies “are . . . more likely and faster to replace the founder with an outside CEO, both in situations that appear adversarial and those mutually agreed to.”); see also Broughman & Fried, supra note 12, at 1329–30. 101 That was the case of the United States before the rise of the VC industry. See Richard M. Buxbaum, Preferred Stock-Law and Draftsmanship, 42 CALIF. L. REV. 243, 271 (1954). 102 Martin Gelter & Geneviève Helleringer, Constituency Directors and Corporate Fiduciary Duties, in PHILOSOPHICAL FOUNDATIONS OF FIDUCIARY LAW 302, 308–09 (Andrew S. Gold & Paul B. Miller eds., 2014). 103 See infra Section IV.C. 104 See infra Part V. 702 COLUMBIA BUSINESS LAW REVIEW [Vol. 2023 control when it is most needed.105 Bratton and Watcher recount how that has been the dominant approach taken by the Chancery Court, regardless of whether the board is controlled by common or by preferred shareholders.106 Since such distributions might be socially valuable,107 Bratton and Watcher108 (and more recently, Sanga and Talley)109 proposed alternative approaches that Delaware might adopt in the future to enable it without drastically changing the fundamental structure of corporate law. For the moment, however, a corporate law fence seems to hinder useful bargains over constituency directors,110 and incentivize governance structures based on private agreements. C. Summary Corporate law rules governing shareholder votes, as well as the election and powers of the board of directors, define the range of bargains between founders and investors over startups’ control and, correspondingly, the range of efficient financial contracting. The regulation of voting in non-listed companies is—contrary to dominant assumptions—quite rigid across jurisdictions. Restrictions to voting rights are often required to be accompanied by expanded cash flow rights. Multiple voting shares are often capped (e.g., two votes per share) or forbidden if they also have preferential cash flow 105 Elizabeth Pollman, Startup Governance, 168 UNIV. PA. L. REV. 155, 190 (2019); Simone M. Sepe, Intruders in the Boardroom: The Case of Constituency Directors, 91 WASH. U. L. REV. 309 (2013). 106 Bratton & Wachter, supra note 91 at 1874-98. 107 Aghion & Bolton, supra note 39. 108 Bratton & Wachter, supra note 91 at 1900-1905; See also Jesse M. Fried & Mira Ganor, Agency Costs of Venture Capitalist Control in Startups, 81 N.Y.U. L. REV. 967, 1020–1023 (2006) (discussing two specific alternatives: “private tailoring” and a “balancing approach”). 109 Sarath Sanga & Eric L. Talley, Don’t Go Chasing Waterfalls: Fiduciary Duties in Venture Capital Backed Startups (Eur. Corp. Governance Inst., Working Paper No. 634, 2023). 110 Casimiro A. Nigro & Jörg R. Stahl, Venture Capital-Backed Firms, Unavoidable Value-Destroying Trade Sales, and Fair Value Protections, 22 EUR. BUS. ORG. L. REV. 39 (2021). No. 2] THE LAW OF CONTINGENT CONTROL 703 rights. The ability to bargain over boards’ election and powers is also hindered by rigid rules, e.g., explicitly requiring boards to be elected by all shareholders or requiring shareholder approval of issuances of preferred shares. While there is evidence that some of these rules are becoming more flexible, particularly in economies attempting to foster startups and VC, they may still limit the contingent allocation of control rights, as discussed in the next Part. IV. CONTINGENT CONTROL The success of business ventures between entrepreneurs and VC investors is threatened by incentive misalignment and distrust. Through security design, they can mitigate these issues: first, by creating different classes of shares, and second, by allocating decision rights to the class of shares whose holders are less prone to depart from the interest of the business venture at crucial moments, which are identified ex ante based on performance measures—e.g., to investors when the company must raise funds at a lower valuation and to entrepreneurs when company valuation exceeds expectations. Where corporate law allows the separation of cash flow from control (through voting and board representation arrangements by classes of shares), it also determines the extent to which the allocation and exercise of those rights can be conditioned to the occurrence of pre-defined events—i.e., contingent control. A. Restricted Common and Stock Options The ability to re-design common shares is useful to regulate entrepreneurs’ control at different stages of the business lifecycle. Here, the prospected expansion of their governance rights can work as an incentive for entrepreneurs to stay in the company and put more effort in achieving goals faster.111 Ideally, founders of VC-backed companies should 111 Subjecting the expansion of entrepreneurs’ rights to long-term goals can also work as a screening process, as “bad” entrepreneurs would be reluctant to make such commitments. See David R Skeie, Vesting and 704 COLUMBIA BUSINESS LAW REVIEW [Vol. 2023 receive common stock with restrictions on their voting and trading rights, and these restrictions should be lifted when an identifiable and measurable target (e.g., revenue growth) is achieved, which is commonly referred to as vesting.112 This is difficult to achieve through security design, as the complete elimination of voting rights, even if temporal, generally requires higher cash-flow rights.113 Hence, in practice, founders’ control is halted through other means, such as assigning preferred board representation and voting rights to investors.114 An alternative that does not require security design is the use of stock option agreements, by virtue of which the firm reserves a pool of shares and gradually releases them upon the passing of time, or the occurrence of a contractually defined event, such as an increase in revenue or company valuation.115 The latter alternative is, however, suboptimal as it deprives founders of other fundamental rights (e.g., to sue directors), which explains why they are more often used to incentivize and recruit employees than to structure founder contingent control.116 B. Convertible Preferred Stock The most useful tool to structure contingent control through security design is by issuing CPS (i.e., convertible preferred stock), which give their holders preferential rights that can be relinquished or further enhanced based on Control in Venture Capital Contracts, No. 297 FED. RESERVE BANK N.Y. STAFF REP. 1 (2007). 112 Restricted shares could be ideal because their issuance is not postponed, providing more clarity over the capital structure and more certainty to founders about their rights. 113 See supra Section III.B.1. 114 See infra Section IV.B. 115 See generally Chua & Woodward, supra note 16. 116 See Hellmann & Puri, supra note 6. See also Anat Alon-Beck, Unicorn Stock Options - Golden Goose Or Trojan Horse, 2019 COLUM. BUS. L. REV. 107 (2019); Yifat Aran & Raviv Murciano-Goroff, Equity Illusions, 2023 J. L. ECON. & ORG. 1. No. 2] THE LAW OF CONTINGENT CONTROL 705 performance, and not necessarily at their will.117 Anti-dilution protections allow CPS holders to preserve or even expand their control rights at the expense of founders in new financing rounds with lower valuations.118 Automatic conversion in case of high-value exit (usually an IPO) gives additional control rights to founders, because VCs (usually the holders of CPS) relinquish the right to elect a board member as a class.119 Enhancing CPS’ rights has less bargaining friction, for it does not imply directly restricting the rights of common shares. Hence, differences in the rules governing these two characteristics of CPS’ governance rights could meaningfully change VC activity across jurisdictions, which justifies considering them in more detail. Anti-dilution protections may be achieved in two ways. One is through pre-emptive rights or rights of first refusal, which give holders of CPS a preferential right to acquire shares in new issuances, allowing them to protect their proportion of voting rights and potentially their board representation rights if new shares are issued.120 This seemingly uncontroversial protection may be unenforceable if it establishes that holders of CPS would pay a lower price than that agreed with the new investors,121 and in any case may be less appealing if it requires an additional payment from the 117 See generally Bratton & Wachter, supra note 91. 118 See Marcel Kahan, Anti-Dilution Provisions in Convertible Securities Features, 2 STAN. J.L. BUS. & FIN. 147 (1995); and Michael A. Woronoff & Jonathan A. Rosen, Understanding Anti-Dilution Provisions in Convertible Securities, 74 FORDHAM L. REV. 129 (2005). 119 Hellmann, supra note 53. 120 George G. Triantis, Financial Contract Design in the World of Venture Capital, 68 U. CHI. L. REV. 305, 312 (2001). 121 Peter Agstner, Shareholder Conflicts in Close Corporations between Theory and Practice: Evidence from Italian Private Limited Liability Companies, 21 EUR. BUS. ORG. L. REV. 505, 514 (2020) (“[R]ights of first refusal or post-sale purchase rights must be structured in such a way that the respective holder is not allowed to purchase the shares at a price that is different from the appraised value attainable in case of withdrawal [Art. 2473(3) c.c.]. For this reason, a so-called improper right of first refusal is deemed invalid if the continuing members may purchase the offered shares not only at a price below the one proposed by the third party—in se legitimate—but even significantly below the above-mentioned fair value.”). 706 COLUMBIA BUSINESS LAW REVIEW [Vol. 2023 investor seeking protection. A stronger and more common anti-dilution protection is triggered by a lower valuation financing. In those cases of startup underperformance, holders of CPS could have the right to adjust the conversion price and eventually get a greater number of common shares when they decide to convert.122 The efficacy of this strategy, however, rests on the possibility to create and enforce automatic conversion rights, which should not be assumed. Automatic conversion is, in fact, a problematic element. Generally, it implies an obligation of the company to have a pool of common shares to exchange.123 In jurisdictions where preferred shares can only have preferential rights to cash flow or voting rights (and not both), this expansive provision tends to be explicitly prohibited (e.g., in China).124 Less strict limits might also derive from widespread rules capping the percentage of authorized shares that the board can issue to comply with conversion obligations or imposing temporal restrictions to that authority, as in those cases new issuances would have to be approved by all shareholders.125 C. Shareholders’ Agreements Because of the difficulties of structuring contingent control through customized shares, participants may (and often do) resort to shareholders’ agreements and narrower contracts, such as voting agreements.126 These agreements can be used to temporarily dispossess founders of control. For example, if a significant proportion (or all) of entrepreneurs’ equity is 122 Kahan, supra note 118 at 159; Woronoff & Rosen, supra note118. The formula to estimate the price would also determine the strength of this protection. 123 Buxbaum, supra note 101 at 281. 124 See Dezan Shira & Assocs, China Introduces Pilot Program for Preferred Shareholders, CHINA BRIEFING (Dec. 20, 2013), https://www.china- briefing.com/news/china-introduces-pilot-program-for-preferred-shares/ [https://perma.cc/VY8L-TUSL]. See also supra Section III.A. 125 See supra Section III.B.1. 126 See Rauterberg, supra note 20 (presenting empirical evidence on how shareholder agreements in the United States are used to distribute control). No. 2] THE LAW OF CONTINGENT CONTROL 707 comprised of common shares, parties can agree that common shareholders will abstain from voting certain decisions or that their vote will align with CPS holders (e.g., electing a director).127 Shareholders’ agreements can also enable CPS holders’ contingent control. For instance, they may provide that all parties will approve the nomination and election of a board member proposed by some shareholders (e.g., holders of CPS), 128 or give holders of CPS the right to force all other shareholders to join them in selling their shares (so-called drag-along rights).129 In turn, these contractual rights and obligations can be subjected to performance measures that, in practice, make control contingent. Whilst shareholders’ agreements could strengthen or even substitute bespoke securities in the structure of contingent control, allocating decision-making rights through contract is not free of challenge and might even be undesirable. Because contracts only bind their signatories, corporate officers and directors are not obliged to observe or follow them when counting shareholders’ votes.130 Accordingly, they could even 127 See, e.g., BRIT. VENTURE CAP. ASS’N, MODEL SHAREHOLDERS’ AGREEMENT, Schedule 2 (Consent Matters), https://www.bvca.co.uk/Policy/Industry-Guidance-Standardised- Documents/Model-documents-for-early-stage-investments [https://perma.cc/W4NN-VN9V] (download Model Shareholders’ Agreement; scroll to Schedule 2 on page 27). 128 See, e.g., Rauterberg, supra note 20 at 1145 (noting that US company GoDaddy’s shareholder agreement “includes governance provisions involving three blockholders, KKR, Silver Lake, and founder Bob Parsons . . . [and] provides . . . that all three parties will vote for one another’s board nominees. Moreover, none of the three shareholders can sell their shares without the permission of both KKR and Silver Lake.”). 129 See, e.g., NAT’L VENTURE CAP. ASS’N, MODEL VOTING AGREEMENT, Section 3 (Drag-Along Right), https://nvca.org/model-legal-documents/ [https://perma.cc/X8XR-M9JS] (scroll down and download Voting Agreement; scroll within Voting Agreement to Section 3 on page 6). 130 In some jurisdictions, it is explicitly stated in law, as in Portugal. See Código das Sociedades Comerciais [Business Associations Code], art. 17, no. 262 (1986). Similarly, article 82 of the Danish Companies Act explicitly provides that shareholders’ agreements are not binding on the company, or with regard to resolutions passed at general meetings. lov om aktie- og 708 COLUMBIA BUSINESS LAW REVIEW [Vol. 2023 be liable if they depart from what is mandated by corporate law and the corporation’s constitutive documents. In reaction, some jurisdictions, such as Colombia,131 the Netherlands,132 Russia,133 and Slovakia,134 have explicitly recognized shareholders’ agreements and established formalities to ensure that they are observed and respected,135 making them a useful tool to structure contingent control. Absent an explicit procedure to make shareholders’ agreements binding for the corporation, a common solution is to have all shareholders sign the agreement.136 This widespread practice has, nevertheless, remained unchallenged judicially in most jurisdictions, and some commentators have cast doubt on the validity of such agreements (and the decisions based on them), as they substitute corporate constituent documents and the legal rules that govern it.137 anpartsselskaber (selskabsloven) [Danish Act on Public and Private Limited Companies], art. 82. 131 See L. 1258de 2008, art. 24 (Colom.) 132 See Dutch Civil Code, Burgerlijk Wetboek, book 2 ‘Rechtspersonen’, art. 242. 133 See Russian Federal Law on Joint Stock Companies, Федеральный закон от 26 декабря 1995 г. N 208-ФЗ Об акционерных обществах, art. 32.1. 134 See Slovakian Commercial Code, 513 Zákonz 5 Novembra 1991, Obchodný Zákonník § 66.c. 135 A similar procedure is available in the United States. See § 7.32 of the American Bar Association’s Model Business Corporation Act. In France, recent case law also indicates that the corporate constitutive document can give shareholders’ agreements its same binding authority. See Cour de Cassation, Chambre commerciale, N°16-14097 (27 juin 2018). 136 See Rainer Kulms, A Shareholder’s Freedom of Contract in Close Corporations — Shareholder Agreements in the USA and Germany, 2 EUR. BUS. ORG. L. REV, 685, 691 (2001). 137 See, e.g., Jidesh Kumar & Richa Mehra, Indian Private Equity: Beware Rights of Shareholders M&A, 25 INT. FIN. L. REV. 40 (2006), https://www.iflr.com/article/2a63u6aqmkut501cwh14w/indian-private- equity-beware-rights-of-shareholders [https://perma.cc/NW4Q-4NAR] (noting that, under established case-law in India, anti-dilution protections in shareholder agreements are not enforceable). No. 2] THE LAW OF CONTINGENT CONTROL 709 When shareholders’ agreements are undisputedly binding, the effect of a breach might still be problematic. The success of legal actions for breach of contract requires proof the breach and generally gives the complying party a right to contractual remedies, instead of injunction or specific performance.138 Hence, a complying party might not be able to use shareholders’ agreements as a control tool (e.g., frustrating the approval of a transaction with votes against the contract), but rather would have to seek contractual remedies (e.g., damages),139 once the votes are counted and the transaction approved. The complying party could attempt to preserve control through an injunction, but the efficacy of this strategy will depend on whether injunction and specific performance are available remedies for contractual claims, which is uncommon and arguably undesirable.140 These considerations and the absence of case law and commentary specific to shareholders’ agreements may explain the tendency to 138 See Henry Hansmann & Reinier Kraakman, Property, Contract, and Verification: The Numerus Clausus Problem and the Divisibility of Rights, 31 J. Leg. Stud. S373, S413 (2002) (highlighting that “specific performance is a disfavored remedy for contract rights”). 139 Sebastian Mock, Kristian Csach & Bohumil Havel, Shareholders’ Agreements between Corporate and Contract Law, in INTERNATIONAL HANDBOOK ON SHAREHOLDERS ́AGREEMENTS: REGULATION, PRACTICE AND COMPARATIVE ANALYSIS 15, 30 (2018) (noting that “the law usually does not provide for the specific consequences of a breach of shareholders’ agreements. Common contractual tools will therefore be used. The injured party should basically have at its disposal the whole range of contractual remedies (damages, withdrawal from the contract, objections of default, liability for delay, etc.)”). 140 Hansmann and Kraakman, supra note 139, at S414 (arguing that “permitting parties to contract for specific performance would run the risk of creating a fruitless arms race. Clauses providing for specific performance would become routine in contracts, since failure to insist on such a clause might leave a party disadvantaged relative to parties with conflicting claims who extract the clause. Yet, if all contracts contain a specific performance clause, contracting parties in general end up no more secure than they were in the absence of those clauses.”). 710 COLUMBIA BUSINESS LAW REVIEW [Vol. 2023 establish pecuniary clauses with liquidated damages,141 and the preference for non-judicial resolutions.142 D. Summary In sum, while contingent control is pivotal to reaching socially efficient VC financing agreements, its structure through security design is not uniformly possible across jurisdictions. In fact, temporarily restricting or expanding shareholders’ voting or board representation rights is often forbidden or subject to legal conditions that reduce the range of bargains between founders and investors. While shareholders’ agreements may support the structure of contingent control, they represent a weaker alternative, for they are not always enforceable and their breach generally entails contractual damages (not a proprietary claim to assure control), and injunctions are not uniformly reliable. The use of shareholders’ agreements as substitutes for corporate law instruments is not only suboptimal for the parties but also perilous for the legal system and venture finance. The next section considers the risks of relying on shareholders’ agreements to sustain VC markets. 141 Holger Fleischer’s comparative examination revealed that “shareholder agreements are almost universally subject to liability sanctions” and that while “many jurisdictions allow for specific performance and injunctions . . . a number of jurisdictions, including Argentina and Japan, still reject this approach.” Holger Fleischer, Comparative Corporate Governance in Closely Held Corporations, in THE OXFORD HANDBOOK OF CORPORATE LAW AND GOVERNANCE 679, 690 (Jeffrey N. Gordon & Wolf- Georg Ringe eds., 2016); Cf. Mock, Csach & Havel, supra note 140, at 40-41 (highlighting certain exceptions: “[a]ccording to (often criticised) German or even Austrian case law the conflict of a decision of the general meeting with the agreement of all of the company′s shareholders is a reason to render such a decision void as if the articles of association have been violated.”). 142 Camille Madelon & Steen Thomsen, Contracting Around Ownership: Shareholder Agreements in France, in THE MODERN FIRM, CORPORATE GOVERNANCE AND INVESTMENT 253 (Per-Olof Bjuggren & Dennis C. Mueller eds., 2009). No. 2] THE LAW OF CONTINGENT CONTROL 711 V. SHAREHOLDERS’ AGREEMENTS’ THREAT TO CORPORATE LAW AND VC MARKETS Shareholders’ agreements with adequate liquidated damages can be functionally equivalent to corporate law instruments in ensuring that control rights are allocated to and exercised by different shareholders at various stages of the startup’s lifecycle. However, allowing shareholders’ agreements to become the main governance instrument for VC-backed firms could have adverse consequences for VC markets and legal systems in the long haul. To discern specific issues derived from this trend, this Part first uses Rauterberg’s framework to distinguish vertical from horizontal agreements.143 Subsequently, it argues that the widespread use of shareholders’ agreements as substitutes for corporate law instruments hinders crucial transaction costs reductions. A. Vertical Agreements Vertical shareholders’ agreements are contracts between the corporation and some or all shareholders. A distinctive aspect of these agreements is that the legal entity assumes some commitments, which could be used to circumvent mandatory procedures in corporate law, such as voting requirements to approve certain decisions. Mergers, for example, must be approved by a qualified majority, which in most countries is a minimum of 70% of the voting shares.144 The corporation may use a vertical agreement to grant a veto right to an investor that does not hold the voting majority, by committing to approve mergers only with the consent of said shareholder. 145 By including liquidated damages (or an 143 See Rauterberg, supra note 20, at 1158-62. 144 ANDREAS CAHN & DAVID C. DONALD, COMPARATIVE COMPANY LAW: TEXT AND CASES ON THE LAWS GOVERNING CORPORATIONS IN GERMANY, THE UK AND THE USA 802–10 (2d ed. 2018). 145 See, e.g., BRIT. VENTURE CAP. ASS’N, supra note 127, at 27-28 (“Matters requiring Investor Majority Consent . . . . merge the Company (or permit, consent to or facilitate any such merger in respect of any other 712 COLUMBIA BUSINESS LAW REVIEW [Vol. 2023 obligation of the corporation to purchase the contracting party’s shares at a premium price) in case of a breach, the veto shareholder secures control over mergers, even if an injunction is unavailable: breaching the agreement would be expensive for the corporation, harm all shareholders, and expose directors to high litigation risk. This type of vertical agreement directly departs from corporate law’s structural properties conceived to reduce agency problems. For example, the legal requirement of a qualified voting majority approval of fundamental transactions is aimed at diminishing opportunism.146 Shareholders with enough voting rights to approve those transactions have controller fiduciary duties and may be held accountable by the corporation or minority shareholders if they breach them.147 Additional legal strategies, such as requiring a majority of the minority to approve specific transactions, are meant to further alleviate agency problems.148 Hence, while veto rights may enable efficient financial contracts in certain circumstances, their introduction through vertical shareholders’ agreements should be discouraged, as it sacrifices essential corporate law protections for non-veto shareholders. Vertical agreements may also include commitments from shareholders to the corporation and to other shareholders. In the U.S., some of the most common shareholder commitments in vertical agreements are waivers of inspection rights, appraisal rights, and rights to forum selection for dispute resolution.149 Although their widespread use in the U.S. suggests waivers are valued by VCs, they insulate control Group Company) or any part of its business with any other person or propose to do so”). 146 See KRAAKMAN ET AL., supra note 68 at 171–203. 147 For a thorough discussion of the duties of controlling shareholders, see Ronald J. Gilson & Jeffrey N. Gordon, Controlling Controlling Shareholders, 152 UNIV. PA. L. REV. 785 (2003). 148 For an overview of the United States experience with this strategy (in public companies) see Edward B. Rock, Majority of the Minority Approval in a World of Active Shareholders, in THE LAW AND FINANCE OF RELATED PARTY TRANSACTIONS 105 (Luca Enriques & Tobias H. Tröger eds., 2019). 149 Rauterberg, supra note 20, at 1166–1167. No. 2] THE LAW OF CONTINGENT CONTROL 713 allocation from judicial scrutiny. This practice not only curtails opportunities for doctrinal development, but may also erode parties’ trust prematurely, by hindering entrepreneurs’ monitoring and litigation rights. Now, even if they prove useful for parties to specific transactions, enabling the waiver of shareholders’ monitoring and litigation rights could negatively affect the development of startups and VC ecosystems. Startups’ officers and directors, who are not exposed to the disciplining forces of public markets, would also be shielded from judicial scrutiny, undermining transparency, shareholder protection, and the quality of corporate governance in these companies—all relevant for the development of VC markets. B. Horizontal Agreements Horizontal shareholders’ agreements are contracts between some or all shareholders. They differ from vertical agreements in that the corporation is not a party and thus cannot enforce any commitment. Still, they could also be used to waive the rights of some shareholders and would be thus subject to the preceding objections if the enforcement threat were credible. Some jurisdictions specifically recognize agreements among shareholders but are silent on whether the corporation could be a contracting party.150 These regulatory environments further contribute to diminishing legal safeguards against agency problems, concealing the waiver of shareholders’ rights under a veil of secrecy. Horizontal agreements among some shareholders can also be used to allocate board representation rights. An agreement among a group of shareholders may detail how the group will vote for the appointment and removal of directors—i.e., via a contractual obligation of some shareholders to vote for someone’s designee.151 In this sense, it may assist the structure of contingent control, by enhancing the voting rights of some shareholders at a given point in time. 150 See generally Mock, Csach & Bohumil, supra note 140. 151 See, e.g., NAT’L VENTURE CAP. ASS’N, supra note 129, Section 1. 714 COLUMBIA BUSINESS LAW REVIEW [Vol. 2023 These agreements give rise to one of two issues. First, when the agreements are secret, they may disguise a situation of control of some shareholders over the board, which in other circumstances would entail specific duties for them.152 In other words, they may enable some shareholders to enjoy the benefits of control with reduced liability risk. A second and alternative issue arises when agreements are not secret, e.g., when they are registered in the company’s records.153 Shareholders with such enhanced powers over the board may be considered controllers or shadow directors with duties to other shareholders. For example, in the United Kingdom, a shadow director, defined as “a person in accordance with whose directions or instructions the directors of the company are accustomed to act,” is subject to general directors’ duties.154 While Delaware case law suggests that contractual powers are insufficient to create corporate fiduciary duties of this type, it does not appear to be a settled issue.155 In 152 Note, Controller Confusion: Realigning Controlling Stockholders and Controlled Boards, 133 HAR. L. REV. 1706, 1709 (2020) (“When a controlling stockholder is involved in a transaction, there are two primary consequences: imposition of fiduciary duties and enhanced transaction scrutiny. Under Delaware law, stockholders typically do not owe fiduciary duties. However, because a controlling stockholder effectively controls the company, a controlling stockholder assumes fiduciary duties similar to those of a director on the board. The most notable of the fiduciary duties imposed on controlling stockholders is the duty of loyalty.”). See also Zipora Cohen, Fiduciary Duties of Controlling Shareholders: A Comparative View, 12 UNIV. PA. J. INT. BUS. L. 379 (1991) (tracing the development of controlling shareholders’ duties in Israel, England and the United States). 153 This would be expected in jurisdictions with an express regulation of shareholders’ agreement, such as Colombia or Slovakia. See L. 1258, de 2008, art. 24 (Colom.) and Slovakian Commercial Code, 513 Zákonz 5 Novembra 1991, Obchodný Zákonník § 66.c., respectively. 154 See Companies Act 2006 § 251 (UK) (defining a “shadow directors” as “a person in accordance with whose directions or instructions the directors of the company are accustomed to act”); see also id. at § 170 (stating “that general duties [of directors] apply to a shadow director of a company . . . “ ). 155 Rauterberg, supra note 20 at 1168-69 (discussing Sheldon v. Pinto Tech. Ventures, L.P., 220 A.3d 245 (Del. 2019), in which the Delaware Supreme Court found that there was no controlling shareholder, No. 2] THE LAW OF CONTINGENT CONTROL 715 jurisdictions with less precise corporate law statutes and scarcer case law, the efficacy of horizontal shareholders’ agreements enhancing board representation rights is even less predictable. This uncertainty over the legal consequences of regulating board appointment and removal rights through shareholders’ agreements not only affects shareholders but also influences how directors and officers conduct themselves and, correspondingly, startups’ potential to raise VC finance in different jurisdictions. Horizontal agreements among all shareholders operate as substitutes of essential corporate law rules (e.g., quorum and voting majorities to approve decisions, inspection rights, and directors’ duties) and doctrines (e.g., corporate opportunity), crafted to deal with agency problems. The decision to register a business as a corporation signals adherence to such corporate law rules that govern the corporate form and alleviate agency problems, as well as with the rules governing the rights and powers of all corporate constituents, regardless of their level of sophistication (implicit consent).156 Shareholders’ agreements that substitute these corporate law rules might thus increase agency problems and give rise to additional practical issues.157 C. Hindering Transaction Cost Reductions and Judicial Scrutiny Whether vertical or horizonal, the use of shareholders’ agreements as a core corporate governance instrument for startup companies also weakens the development of the VC industry by hampering crucial transaction cost reductions. Articles of incorporation and other instruments governed by notwithstanding the presence of a shareholder agreement that allowed a group of venture capital firms to appoint almost all directors in contrast to van der Fluit v. Yates, No. CV 12553-VCMR, 2017 WL 5953514, at *5 (Del. Ch. Nov. 30, 2017) and In re Hansen Med., Inc. Stockholders Litig., No. 12316-VCMR, 2018 WL 3030808 (Del. Ch. June 18, 2018), which came out the other way on broadly similar facts). 156 Jill E Fisch, Stealth Governance: Shareholder Agreements and Private Ordering, 99 WASH. U. L. REV. 913 (2022). 157 Id. at 946–953. 716 COLUMBIA BUSINESS LAW REVIEW [Vol. 2023 corporate law are binding for all members.158 In doing so, they eliminate the need to negotiate and unanimously agree on the reallocation of property and governance rights when existing shares are transferred, or new shares are issued, ultimately reducing transaction costs.159 A shareholders’ agreement, on the contrary, only binds its signatories,160 making it necessary to revisit and likely renegotiate the terms of the agreement whenever there are changes in shareholder ownership. The costs of such renegotiations have been traditionally perceived as marginal for non-listed corporations because of the pervasive assumption that they are run by closely-related individuals.161 For VC-backed startups, however, the costs of negotiations can be particularly burdensome, given the 158 KRAAKMAN ET AL., supra note 68, at 17 (“The corporation’s charter [or ‘articles of association’ or ‘constitution,’ as it is termed in some jurisdictions]…sets out the basic terms of the relationship among the firm’s shareholders, and between the shareholders and the firm’s directors and other managers. By explicit or implicit reference, the charter can also become part of the contract between the firm and its employees or creditors”); See also John Armour & Michael J. Whincop, The Proprietary Foundations of Corporate Law, 27 OXF. J. LEG. STUD. 429 (2007). 159 Corporate law substitutes negotiation and unanimous agreement with voting. See Easterbrook & Fischel, supra note 65, at 398 (“The states’ legal rules generally provide investors with the sort of voting arrangements they would find desirable if contracts could be arranged and enforced at low cost.”); see also Armour and Whincop, supra note 159, at 451-52 (“[Corporate law] performs an important additional role in supporting the enforcement of these ‘contractual’ arrangements against third parties. This comes into play in situations where one of the parties—a director or shareholder—acts without authority under the constitution [i.e., articles of incorporation, charters and bylaws] or in breach of their duties (where applicable), in a transaction involving a third party. That is, there is an attempted alienation or grant of entitlements in respect of corporate assets that is contrary to the parties’ internal arrangement about how these entitlements are to be shared. This would encompass not only attempts to sell the assets, but also attempts to use the assets to bond unauthorized contracts, or attempts to allow newcomers to participate in the sharing of control of the assets—that is, unauthorized issues or transfers of shares, or putative appointment of directors.”). 160 See supra Section IV.C and note 131 (citing examples from Portugal and Denmark, where the corporation cannot be a part of the agreement.) 161 On this assumption, see Pereira, supra note 2. No. 2] THE LAW OF CONTINGENT CONTROL 717 heterogeneity of their shareholders, which may include seed and angel investors, venture capitalists, sovereign wealth funds, founders, and employees, with various levels of sophistication, firm-specific knowledge, and bargaining powers.162 Increasing the costs of governance adjustments can delay funding and impair growth—or, even worse, permanently erode trust and doom the startup to failure. Shareholders’ agreements also diminish the beneficial transaction cost reductions that typically derive from standardization.163 Firms often adjust their corporate governance structures by replicating provisions from other companies’ articles, charters, and bylaws, until optimal provisions become standard for specific purposes.164 The confidentiality of shareholders’ agreements hampers replication and, consequently, does not allow “private ordering innovations to disseminate among firms.”165 This barrier to standardization, in turn, increases the cost of entry for new investors, ultimately shrinking the pool of VC funds 162 It is precisely this shareholder heterogeneity that makes VC-backed startups uniquely susceptible to issues among different types of shareholders, including among holders of preferred stock with seemingly aligned interests. See Robert P. Bartlett, III, Venture Capital, Agency Costs, and the False Dichotomy of the Corporation, 54 UCLA L. REV. 37, 71–80 (2006) (discussing inter-investor conflicts over the startup company’s exit strategy and future financing.) 163 See Fisch, supra note 157 at 942–47. 164 The replication of provisions leads to standardization and takes place both at the moment of incorporation and through amendments during later stages. See Frank H Easterbrook & Daniel R Fischel, The Corporate Contract, 89 COLUM. L. REV. 1416, 1443 (1989) (for a discussion on the immediate and long-term impact of corporate governance amendments or “latecomer terms.” They argue that “[t]he mechanism by which entrepreneurs and managers bear the cost of unfavourable terms does not work—not in any direct way, anyway—for latecomer terms. It will work eventually. Latecomer terms that injure investors will reduce the firm’s ability to raise money and compete in product markets. But these eventual reactions are not remedies; they explain why firms that choose inferior governance devices do not survive, and they show why widespread, enduring practices are likely to be beneficial, but they do nothing for participants in the ventures that are about to be ground under by the heel of history.”). Id. 165 See Fisch, supra note 157 at 948. 718 COLUMBIA BUSINESS LAW REVIEW [Vol. 2023 operating in a given market and reducing the number of startups that can access external finance. An additional barrier to transaction costs reductions derives from shareholders’ agreements scarce exposure to judicial scrutiny. Courts generally provide guidance on what is permissible under corporate law.166 In the venture finance context, judicial decisions can reduce founder-investor bargaining costs by helping parties avoid frivolous negotiations over impermissible distributions of cash flow and control rights.167 By design, shareholders’ agreements are less likely to be litigated because contractual lawsuits generally require bringing all contractual parties to the procedure.168 These agreements may also include deterring penalties for 166 See id. (arguing that replication and standardization of private ordering innovations through network effects is diminished further by the lack of judicial scrutiny: “These efforts are aided by judicial decisions such as Boilermakers and ATP which provide firms with clear guidance on the validity of such innovations.”); See also KRAAKMAN ET AL., supra note 68, at 20 (“Courts play a key role in filling gaps, simply by interpreting privately drafted contractual terms in a corporation’s charter. A firm will get the greatest advantage from the courts’ interpretive activity if it adopts standard charter terms used by many other firms, since those standard terms are likely to be subject to repeated interpretation by the courts. And the most widely used standard charter terms are often the default rules embodied in the corporation law.”). 167 Trados is a good example, as it is one of the few decisions that involved judicial scrutiny of a VC-backed startup’s articles of incorporation, charter, bylaws, and other contractual arrangements. In re Trados Inc. S’holder Litig., 73 A.3d 17, 20 (Del. Ch. 2013). In reaction, the NVCA model voting agreement included a new provision, the sale right, “designed to insulate the Board from a Trados-type claim. In particular, since this section provides for redemption rights additional to any that may be included in the Certificate of Incorporation, selling the company may be the only means by which the Board is able to honor this contractual “put” obligation.” NAT’L VENTURE CAP. ASS’N, SUPRA NOTE 129, AT n.47. There is also evidence that startup boards, following legal advice, adjusted their processes when deciding whether to sell the company. See Abraham J. B. Cable, Does Trados Matter?, 45 J. CORP. L. 101, 103 (2019) (finding, from 20 semi-structured interviews with startup lawyers, that “interviewees report that Trados does affect the process of selling a startup. Most noticeably, boards are more systematic in assessing the value of continuing as a company.”). 168 See supra notes 131 and 132 and accompanying text. No. 2] THE LAW OF CONTINGENT CONTROL 719 unsuccessful claimants and arbitration clauses that effectively prevent judges from reviewing them.169 Shareholders’ agreements comparatively low exposure to judicial review not only prevents reduction of transaction costs, but can also erode trust in startup and VC ecosystems, which are increasingly perceived as unaccountable, given the shadow governance and financial practices to which they are associated.170 D. Summary Unlike charters, bylaws, and articles of incorporation, agreements among shareholders of non-listed corporations generally do not have to be disclosed or registered and are seldom litigated. This opacity frustrates contractual, judicial, and legal innovation. The growing tendency to use these agreements as substitutes for corporate law instruments also hinders transaction costs reductions that would normally emerge from standardization. Additionally, it deteriorates the efficacy of legal safeguards against agency problems, which would be particularly damaging for the VC industry. VI. IMPLICATIONS The preceding analyses show that legal restrictions to structuring contingent control are more common than assumed, which has implications for theoretical and empirical studies in financial contracting, and corporate law design. A. Financial Contracting There are two main implications for financial contracting. First, the identified legal boundaries to structure contingent control across jurisdictions may assist the development of 169 See Madelon & Thomsen, supra note 142, at 253–292. 170 See Alexander I Platt, Unicorniphobia, 13 HARV. BUS. L. REV. 115 (2022); Amy Deen Westbrook, We(‘re) Working on Corporate Governance: Stakeholder Vulnerability in Unicorn Companies, 23 U. PA. J. BUS. L. 505 (2020); Renee M. Jones, The Unicorn Governance Trap, 166 U. PA. L. REV. ONLINE 165 (2017). 720 COLUMBIA BUSINESS LAW REVIEW [Vol. 2023 financial contracting models that more accurately estimate the range of agreements available in real life and the tradeoffs of alternative legal frameworks. Second, domestic and international empirical evidence should be accounted with skepticism, given that most studies do not distinguish the legal instruments used to distribute control, whether commitments are enforceable, the consequences of breaching such commitments, and how such variations affect incentives. 1. Theory Control theories fundamentally assert that parties in financial agreements allocate board representation and voting rights ex ante, as a means to solve incomplete contract problems (i.e., contingencies that are difficult to anticipate), enabling the creation of otherwise unfeasible financial agreements.171 In VC finance, this is especially important, given high uncertainties.172 The analyses above show that the range of agreements on control allocation has three concrete limitations across jurisdictions, which the literature has yet to account. First, control through direct board representation is illusory.173 Class-based election of directors remains exceptional, and boards have duties to all shareholders.174 Even in jurisdictions where class-based board representation is warranted, constituency directors’ ability to exercise control on behalf of their electors is constrained by the threat of litigation for breach of their duties to all shareholders, especially when it matters the most (e.g., exits).175 Second, enhancing and restricting shareholders’ voting rights is subject to legal requirements and temporal limitations that hinder the use of voting as a means to solve incomplete contract problems.176 Deviation from one-share 171 See Hart, supra note 33. 172 See Hellmann, supra note 11. 173 See supra Section III.B.1. 174 See supra Section III.B.2. 175 Bratton and Wachter, supra note 91, at 1815. 176 See supra Sections III.B, IV.A, and IV.B. No. 2] THE LAW OF CONTINGENT CONTROL 721 one-vote is still exceptional, and many jurisdictions that allow changes in voting rights impose limits on the number of votes that can be assigned to specific shares, especially to those with enhanced cash flow rights.177 Thus, bargains over voting rights are significantly constrained. Third, conditioning the allocation and exercise of board and voting rights to identifiable events is also hindered by restrictions to the issuance of convertible securities in numerous jurisdictions.178 While shareholders’ agreements may assist the structuring of contingent control (e.g., through veto rights), they are imperfect substitutes, given uncertainties over the enforceability of certain provisions.179 Also, contrary to corporate law instruments, they do not grant proprietary rights and obligations, but contractual ones, breach of which generally entails damages and not necessarily performance.180 Indeed, some financial contracting theorists concede that potentially optimal arrangements between innovators and investors may not be legal or enforceable.181 The legal limitations to control allocation discussed above may assist the development of control models of financial contracting that more accurately estimate the range of agreements available in reality, which could be useful to evaluate the tradeoffs of alternative legal frameworks. 2. Empirical Studies A second implication of the preceding analyses relates to empirical studies in venture finance. Since Kaplan and Per Strömberg’s seminal paper, a common research strategy to explore control allocation’s role and practical use in venture finance has been to collect financial agreements and identify how board and voting rights are distributed.182 Yet, as Kaplan 177 See supra Section III.A. 178 See supra Section IV.B. 179 See supra Section IV.C. 180 See supra Section IV.C. 181 See Hart, supra note 33. 182 See Da Rin, et al. supra note 9. 722 COLUMBIA BUSINESS LAW REVIEW [Vol. 2023 and Per Strömberg acknowledged, real-life contracts’ content is only “presumably” enforceable.183 The evidence of legal constraints to contingent control discussed above confirms that standard provisions are often unenforceable.184 Even when they are legally binding, enforcement attempts may lead to opposite outcomes, as in the widely-commented Delaware decision In Re Trados.185 This analysis highlights the need to assess empirically the extent to which prevalent control allocations are enforceable and how those considerations impact entrepreneur-investor negotiations, the legal structure of the agreements (e.g., security design), and parties’ incentives (e.g., entrepreneurs’ estimation of private benefits). Meanwhile, the empirical evidence currently available should be accounted with caution, acknowledging that the distribution of control is also influenced by critical and dynamic rules of corporate law, and enforcement considerations that remain in the blind spot of financial contracting scholarship. Failure to account for legal boundaries and to distinguish the type of legal instruments used in VC deals could also produce potentially misleading results. For example, defining control as voting and board seats, Wang et al. built a dataset based on VC firms’ surveys to explore the determinants between shared and contingent control in a large sample of 86 Chinese VC-backed companies.186 Despite the detailed focus on explanatory variables (e.g., entrepreneur’s non-monetary benefits, venture capitalist’s bargaining power, monitoring cost),187 their study does not consider the various legal instruments used in each deal (e.g., deferred stock purchase agreements or preferred stock). This decision overlooks the fact that convertible-preferred securities are not permitted 183 See Kaplan and Strömberg, supra note 12, at 294. 184 See supra Part IV. 185 See Bratton & Wachter, supra note 91, at 1882–1900. 186 See Lei Wang, Fangzhao Zhou & Yunbi An, Determinants of Control Structure Choice between Entrepreneurs and Investors in Venture Capital- Backed Startups, 63 ECON. MODEL. 215, 219 (2017) (“The final valid sample had 86 enterprises.”). 187 Id. at 219–20. No. 2] THE LAW OF CONTINGENT CONTROL 723 under Chinese corporate law and that private agreements with particularly harsh clauses protecting investors are common in China,188 rendering their assessment of shared and contingent control imprecise. Surveys are common in this type of studies,189 which makes it crucial to interpret their results restrictively. International studies exploring how VCs allocate and exercise control across countries are more robust but have similar shortcomings. Kaplan, Martel & Strömberg, for example, use a sizeable sample of term sheets, stock purchase agreements, companies’ business plans, and VC’s investment analysis in 145 investments in 23 countries, offering an insightful account of how the legal structure of control varies across countries.190 They also consider the extent to which US- style agreements—which generally reproduce the optimality predicted by dominant theories—are observed elsewhere and investigate how (if at all) legal differences may explain variations in capital structures and security design.191 To that end, they use several legal indexes compiled by La Porta et al. in 1997 (e.g., creditor protection, anti-director rights), which are somewhat uniform but have three concrete limitations.192 First, they are imprecise due to significant coding problems that the legal literature has uncovered.193 Second, they are not representative of the laws governing board and voting rights in VC-backed firms, but legal rules for listed companies.194 Third, these indexes are based on legal systems’ perceived quality before the globalization of VC and the proliferation of legal reforms explicitly aimed at facilitating VC investments.195 Notwithstanding these and other limitations of these legal indexes, they remain common in 188 See Lin, supra note 18. 189 See Da Rin et al., supra note 9. 190 See Kaplan et al., supra note 25 at 274. 191 Id. 192 Id. at 311 (“‘Accounting standards,’ ‘Creditor protection,’ and ‘Minority protection’ [‘Anti-director rights’] are from La Porta et al. [1997]”). 193 See, e.g., Spamann, supra note 58; Pistor, supra note 58. 194 See supra note 59 and accompanying text. 195 See supra note 60 and accompanying text. 724 COLUMBIA BUSINESS LAW REVIEW [Vol. 2023 international empirical studies of the institutional determinants of VC finance,196 which begs a cautious evaluation of their findings, especially when they are invoked to justify legal reform. The evidence and analyses advanced in this paper reveal that corporate law has a more decisive influence on the legal structure of control in VC-backed firms across countries than previously assumed, shedding light on new lines of empirical research. Observed variations in the structure of VC deals across countries (e.g., security design) may be more related to legal constraints to security design than to static measures of legal systems’ perceived quality or the sophistication of VCs. Variations observed in VC over time may also be explained by changes in the law, including those derived from judicial interpretation. The NVCA periodically amends its term sheets, which supports this proposition and invites further scrutiny over the legal determinants of VC finance. B. Corporate Law Design Contingent control is fundamental for VC finance. Whether and how corporate law facilitates or discourages the structure of contingent control is of paramount importance to foster startup growth and VC through legal reform. Two general implications for corporate law design follow from the comparative examination of the legal structure of contingent control. One is that legal restrictions to security design in non-listed companies should be removed or recalibrated, as they reduce the range of bargains over control between entrepreneurs and investors, hampering the financing of potentially viable companies.197 Total or qualified removal of pervasive restrictions, such as to multiple-voting shares, class-based board representation or boards’ powers to issue bespoke securities, could significantly expand the 196 See, e.g., Lerner & Schoar, supra note 25; Josh Lerner & Joacim Tag, Institutions and Venture Capital, 22 IND. CORP. CHANGE 153 (2013); Khoury et al., supra note 57. 197 See supra Part IV. No. 2] THE LAW OF CONTINGENT CONTROL 725 spectrum of negotiation and facilitate the structuring of contingent control. Reforms, of course, require tradeoffs. Enhanced freedom to design securities and customize boards’ composition and powers would not only expand the range of bargains, but also increase the complexity of startups’ governance structures, creating new corporate law challenges.198 Attempts to reform also call for a tradeoffs analysis that is context specific. For example, enabling boards’ empowerment to issue preferred shares in jurisdictions with weak minority shareholder protection could exacerbate their vulnerability. Still, an open discussion of those tradeoffs, premised on the ever-fiercer competition to expand startups’ access to VC, would offer an opportunity to examine emerging trends in the governance of VC-backed companies that challenge the traditional regulation of private companies. For instance, whether constituency directors, de jure or de facto, should prioritize the interest of the corporation as a whole or their constituents when those are in tension.199 This paper’s functional comparative analysis of the law of contingent control offers a path to identify corporate law rules that merit recalibration to improve founder-investor negotiations and expand the range of companies that can access VC finance. A second implication is that the use of shareholders’ agreements and similar contracts to circumvent legal protections against agency problems should be discouraged. One of the most common uses of shareholders’ agreements is to grant board representation rights to investors, e.g., via a contractual obligation to vote for investors’ nominees.200 These provisions not only enable the evasion of board election rules, but may also disguise a situation of control, 198 See Pollman, supra note 105 (critically discussing the challenges that have emerged in the U.S.); Bartlett, supra note 163 (highlighting that agency problems in VC-backed companies are dynamic and require new analytical frameworks—e.g., to account for the risks that parties assume or attempt to control at different points in time.). 199 Indeed, a contested issue. See Gelter & Helleringer, supra note 102; Sepe, supra note 105. 200 See supra Part V. 726 COLUMBIA BUSINESS LAW REVIEW [Vol. 2023 disenfranchising shareholders that would be otherwise protected by controller duties.201 Furthermore, these provisions may not be as useful when it matters most (e.g., exits), given that in those situations courts would likely uphold directors’ duties to the company over duties derived from contracts, if those are in conflict.202 A clear legal framework determining the election and duties of constituency directors would discourage this practice and enhance transparency in startups’ governance. Shareholders’ agreements are also used to grant veto rights and to waive monitoring and litigation rights. To the extent that these arrangements are tantamount to contracting out of corporate law institutions aimed at alleviating agency problems (e.g., qualified approval of fundamental transactions, appraisal rights) and that they risk eroding parties’ trust prematurely,203 they should be discouraged by expressly limiting the enforceability of such clauses. VII. CONCLUSION This paper showed that the range of bargains on contingent control between entrepreneurs and investors is determined by corporate law rules governing the allocation of board representation and voting rights in non-listed companies. These dynamic rules, which the legal and financial contracting literature has overlooked, contribute to explaining differences in security design and capital structures observed across countries and time. They also explain the increasing reliance on shareholders’ agreements to structure VC-backed companies’ governance, a practice that weakens parties’ trust and legal protection, and erodes corporate law safeguards against agency problems that are crucial for developing VC markets. By facilitating the structure of contingent control through security design, corporate law can discourage the over- 201 See supra Section V.B. 202 See supra Section V.B. 203 See supra Section V.A. No. 2] THE LAW OF CONTINGENT CONTROL 727 reliance on shareholders’ agreements and expand the range of founder-investor bargains—and, in turn, increase the number of startups that can access VC. 728 COLUMBIA BUSINESS LAW REVIEW [Vol. 2023