For centuries, the duty of loyalty has been the hallowed centerpiece of fiduciary obligation, widely considered one of the few “mandatory” rules of corporate law. That view, however, is no longer true. Beginning in 2000, Delaware dramatically departed from tradition by granting incorporated entities a statutory right to waive a crucial part of the duty of loyalty: the corporate opportunities doctrine. Other states have since followed Delaware’s lead, similarly permitting firms to execute “corporate opportunity waivers.” Surprisingly, more than fifteen years into this reform experiment, no study has attempted to either systematically measure the corporate response to these reforms or evaluate the implications of that response.
This Article offers the first broad empirical investigation of the area. Contrary to conventional wisdom, we find that well over one thousand public corporations have adopted waivers—often with capacious scope and reach. The Article thus establishes a central empirical fact that is an important baseline for further discussion: Public corporations have an enormous appetite for contracting out of the duty of loyalty when freed to do so. This analysis also sheds light on the high-stakes normative debate around the relationship between fiduciary principles and freedom of contract. What types of corporations choose to contract around default rules? When they do so, do such measures tend to bolster or thwart shareholder welfare?
The Article develops an effi­cient con­tracting approach to explain why corporations—and their share­holders—might favor tailoring the duty of loyalty and presents evidence assessing the merits of Delaware’s experiment.


    1. Fiduciary Law and the Corporate Opportunities Doctrine
    2. Evolution of the COW: A Brief History of Endeavors to Contract Out of Opportunities Doctrine
    1. Framework
    2. Ex Ante Versus Ex Post Waivers
    1. Data
    2. Descriptive Statistics of Waivers
    3. Issuer Characteristics of COW Adopters
    4. Market Reaction to COW Adoptions
    1. Evaluating COWs
    2. Contractualizing the Duty of Loyalty?
    3. Who Should Design Corporate Law and Governance?
    4. Shareholder Versus Social Welfare




For nearly two centuries, a cornerstone of Anglo-American corpo­rate law has been the fiduciary duty of loyalty, the most demanding and litigated fiduciary obligation imposed on corporate managers. 1 See infra section I.A.1. The duty—which regulates financial conflicts of interest and requires manag­ers to subordinate their own interests to the corporation’s—represents a key policy lever to address the most pernicious of intra-firm agency costs. 2 See infra section I.A.1. Practitioners, academics, and jurists alike have charac­terized loyalty as the most important fiduciary obligation, 3 See, e.g., Joel Seligman, The New Corporate Law, 59 Brook. L. Rev. 1, 3 (1993) (describing the duty of loyalty as the “most important fiduciary duty of corporate officers and directors”); 1 William Campbell Ries, Regulation of Investment Management and Fiduciary Services § 11:18, Westlaw (database updated July 2016) (noting loyalty as “the most important duty which arises within the context of fiduciary relationships”); see also infra note 52. The centrality of the duty of loyalty in Anglo-American corporate law goes back at least to the mid-nineteenth century English case of Aberdeen Ry. v. Blaikie Bros. (1854), 2 Eq. Rep. 1281. See generally David Kershaw, The Path of Corporate Fiduciary Law, 8 N.Y.U. J.L. & Bus. 395, 428–33 (2012) (discussing Aberdeen Railway and its progeny). The duty of loyalty has also been the object of some of legal rhetoric’s most celebrated passages. See, e.g., Meinhard v. Salmon, 164 N.E. 545, 546 (N.Y. 1928) (describing loyalty as demanding a “duty of the finest loyalty,” far “stricter than the morals of the market place,” “[n]ot honesty alone, but the punctilio of an honor the most sensitive”). and economists have credited it with facilitating efficient corporate stew­ardship and catalyzing investment and entrepreneurship. 4 Brian R. Cheffins, Does Law Matter? The Separation of Ownership and Control in the United Kingdom, 30 J. Legal Stud. 459, 464 (2001) (“A director’s duty of loyalty is another type of legal rule that can help to provide a protective environment for investors [because] . . . managerial self-dealing will potentially constitute breaches of duty.”); Simon Johnson, Rafael La Porta, Florencio Lopez-de-Silanes & Andrei Shleifer, Tunneling, 90 Am. Econ. Rev. 22, 27 (2000) (discussing how the duties of loyalty and care are important in protecting minority shareholder rights, which promotes the development of capital markets); cf. Dodge v. Ford Motor Co., 170 N.W. 668, 684 (Mich. 1919) (discussing the duties owed to minority shareholders); Leo E. Strine, Jr. et al., Loyalty’s Core Demand: The Defining Role of Good Faith in Corporation Law, 98 Geo. L.J. 629, 634 (2010) (“[I]t is essential that directors take their responsibilities seriously by actually trying to manage the corporation in a manner advantageous to the stockholders.”). Indeed, a well-known literature in law and finance has documented the beneficial role that credible conflict-of-interest management plays in promoting com­pany value, 5 See, e.g., Paul Gompers et al., Corporate Governance and Equity Prices, 118 Q.J. Econ. 107, 121–44 (2003) (providing evidence of a positive relationship between investor protection and returns, value, and operating performance). vibrant capital markets, 6 See, e.g., Rafael La Porta et al., Law and Finance, 106 J. Pol. Econ. 1113, 1152 (1998). and firm longevity. 7             Id. (discussing the connection between investor protection, developed debt and equity markets, and economic growth). The duty of loyalty is also notable because of its historically inveterate and unyielding nature: While much of corporate law consists of “default rules” that parties may freely alter, the duty of loyalty is widely perceived as “immutable”—immune to private efforts to dilute, tailor, or eliminate it. 8       Lucian Arye Bebchuk & Assaf Hamdani, Optimal Defaults for Corporate Law Evolution, 96 Nw. U. L. Rev. 489, 496 n.16 (2002) (providing “the duty of loyalty of corporate directors” as an example of mandatory corporate governance regulation); Bernard S. Black, Is Corporate Law Trivial?: A Political and Economic Analysis, 84 Nw. U. L. Rev. 542, 551–53 (1990) (citing self-dealing rules as one example of mandatory law); Melvin Aron Eisenberg, The Structure of Corporation Law, 89 Colum. L. Rev. 1461, 1486 (1989) (arguing self-dealing rules are “largely mandatory, at least for publicly held corporations”); Marcel Kahan, The Qualified Case Against Mandatory Terms in Bonds, 89 Nw. U. L. Rev. 565, 607 n.164 (1995) (claiming the rules on self-dealing by managers are mandatory); Randall S. Thomas, What Is Corporate Law’s Place in Promoting Societal Welfare?: An Essay in Honor of Professor William Klein, 2 Berkeley Bus. L.J. 135, 139 (2005) (stating self-dealing rules are mandatory for public corporations); Jill E. Fisch, Picking a Winner, 20 J. Corp. L. 451, 458 (1995) (reviewing Roberta Romano, The Genius of American Corporate Law (1993)).

That perception is no longer true. Beginning in 2000, Delaware dramatically departed from tradition, amending its statutes to enable corporations to waive a critical component of loyalty—the corporate oppor­tunities doctrine—which forbids corporate fiduciaries from appro­priating new business prospects for themselves without first offering them to the company. 9       See Del. Code Ann. tit. 8, § 122(17) (2017) (codifying this amendment); infra section I.B.2. From that moment forward, Delaware corpo­rations and managers became free to contract out of a significant portion of the duty of loyalty. 10 See infra sections I.B.2–.3. In the ensuing years, eight other states have followed Delaware’s lead, granting their own incorporated entities the statutory authority to execute corporate opportunity waivers (COWs). 11 Kan. Stat. Ann. 17-6102(17) (2014); Md. Code Ann., Corps. & Ass’ns § 2-103(15) (LexisNexis 2014); Mo. Rev. Stat. § 351.385(16) (2016); Nev. Rev. Stat. § 78.070(8) (2007); N.J. Stat. Ann. § 14A:3-1(q) (West 2016); Okla. Stat. tit. 18, § 1016(17) (2001); Tex. Bus. Orgs. Code Ann. § 2.101(21) (West 2012); Wash. Rev. Code § 23B.02.020(5)(k) (2016); see also infra section I.B.4. The Corporate Laws Committee of the American Bar Association (ABA) has also recently proposed amending the Model Business Corporation Act to permit advance waivers of corporate opportunities. 12 See Comm. on Corp. Laws, ABA Bus. Law Section, Changes in the Model Busi­ness Corporations Act—Proposed Amendments to Sections 2.02 and 8.70 (and Related Changes to Sections 1.43, 8.31, and 8.60) Permitting Advance Action to Limit or Eliminate Duties Regarding Business Opportunity, 69 Bus. Law. 717, 721–24, 727–31 (2014).

The reform movement sparked by Delaware represents a significant departure from long-settled understanding and common law tradition—a departure that concerns a basic tenet of company law. It is thus surprising that no significant study to date has empirically assessed firms’ responses to these reforms (save for anecdotal accounts suggesting that there has been little reaction 13 See, e.g., Christopher E. Austin & David I. Gottlieb, Renouncing Corporate Opportunities in Spin-offs, Carve-out IPOs and Private Equity Investments, VC Experts: Buzz, [] (last visited Jan. 30, 2017) (“Since the enactment of Section 122(17), it appears that only a small number of corporations have gone public with or adopted corporate opportunity provisions in their charters.”). ). This Article endeavors to fill this void, presenting what appears to be the first broad empirical assessment
of how public companies have responded to the statutory reforms, 14 See infra Part III. and devel­­­­oping a broader conceptual and theoretical account to predict and explain that response. 15 See infra Part II.

Based on an extensive data set of U.S. public companies’ filings with the Securities and Exchange Commission (SEC), we isolated over 10,000 unique disclosures that were plausible COWs. We then refined these data in two ways. First, we manually coded a large subset along a variety of dimensions pertaining to the existence, scope, reach, and location of a waiver. Second, we used our hand-coded sample to train a machine-learning algorithm, thereby extending our coding protocol to the entire population of candidate documents. 16 For details on our predictive coding methodology, see Gabriel Rauterberg & Eric Talley, A Machine Learning Classifier for Corporate Opportunity Waivers (Columbia Law Sch., Law & Econ. Research Paper Series, Working Paper No. 553, 2016), (on file with the Columbia Law Review). In contrast to the conventional wisdom, we find that hundreds of public corporations in our sample—and well over one thousand in the population—have disclosed or executed waivers, 17 We note that the vast majority—but not all—of the disclosures in our data set come from publicly traded companies. See infra sections III.A–.B. Our data set deems an issuer to have “executed” a COW if it discloses either (1) a representation that it has executed one; or (2) an operative provision in a legal document that purports to waive the corporate opportunities doctrine. with terms that apply broadly across both managerial ranks and cate­gorical domains. 18 See infra section III.B. This Article thus establishes a central empirical fact that is an important baseline for further discussion: Public companies have an enormous appetite for tailoring the duty of loyalty when freed to do so.

Alongside these empirical findings, this Article takes on several fundamental questions raised by widespread adoption of waivers of corporate opportunities: Why would a corporation ever choose to restrict the reach of the duty of loyalty? What form will the optimal allocation of corporate opportunities plausibly take in different companies? Under what conditions would such waivers be valuable to shareholders, even as such waivers constrain the fiduciary duties owed to share­holders? Delaware’s seventeen-year statutory experiment also provides a unique opportunity to revisit some foundational issues in corporate law with a fresh perspective. Indeed, there is a vigorous, decades-old debate that asks whether any of corporate law’s rules should be mandatory, or whether parties instead should be free to contract out of every gov­ernance requirement as they already can from most. 19 See infra notes 103–106, 206 and accompanying text (discussing the Delaware Court of Chancery’s treatment of an early attempt at contracting out of the corporate opportunities doctrine and noting a foundational question in corporate law is whether corporations will “adopt optimal corporate governance structures on their own accord”). Does enlarging the contracting space for fiduciary duties result in greater efficiencies, or does it instead result in unchecked opportunism? There is also a significant set of issues involving whether corporations actually make use of the freedom frequently given them by corporate law to replace default rules and whether, when they do so, it serves shareholders’ interests. 20 See infra section II.B.4 (highlighting the possibility that permitting corporations to tailor a nebulous standard provided by the common law may lead to a more efficient legal framework).

This Article argues that there are, in fact, several plausible economic rationales for a corporation to embrace a COW for the sake of share­holder value. Indeed, in the years leading up to Delaware’s initial reform, a growing chorus of critics argued that the exacting requirements of the duty of loyalty had begun to impede corporations’ ability to raise capital, build efficient investor bases, and secure optimal management arrange­ments. 21 See infra section I.B.2. This claim was based in part on the recognition that many then-emerging sources of capital, such as private equity, venture capital, or spin-off transactions, may subject their financial sponsors to fiduciary duties in profound conflict with either their larger business plans or with fiduciary obligations they owe to other business entities. 22 See infra notes 109–113 and accompanying text. That said, prior to the reform, the Delaware Chancery Court took a particularly dim view of the enforceability of such contractual clarifications. See Siegman v. Tri-Star Pictures, Inc., No. 9477, 1989 WL 48746, at *8 (Del. Ch. May 30, 1989) (denying defendants’ motion to dismiss on the grounds that there was “at least one plausible state of facts” in which the provision at issue “would arguably operate to eliminate or limit the directors’ liability for breach of their duty of loyalty”). Absent the contractual ability to clarify ownership rights regarding new business opportunities, it is difficult to see how such capital structures could stably persist within the standard corporate arrangement.

Consider, for example, one of the issuers in our database: Prosper Marketplace, Inc., the first and still one of the largest peer-to-peer lenders. 23 Benjamin Lo, It Ain’t Broke: The Case For Continued SEC Regulation of P2P Lending, 6 Harv. Bus. L. Rev. Online 87, 88 (2016),
uploads/2016/08/B.-Lo_Regulation-of-P2P-Lending.pdf [].
The waiver that Prosper adopted in its charter covered any member of Prosper’s board who was not also an employee. 24 See Prosper Marketplace, Inc., Amended and Restated Certificate of Incor­poration of Prosper Marketplace, Inc. (Form S-1, exh. 3.1) (Oct. 30, 2007),
1ex3d1.htm [].
The four outside directors in place at the time of the company’s public filing (and a majority of the board) worked for financial firms—three of them in venture capital—and at that point served as directors for at least fourteen other companies, including another online commercial market. 25 See Prosper Marketplace, Inc., Registration Statement (Form S-1), at 23–24 (Oct. 30, 2007), [] (discussing outside directors’ other mana­ge­­­rial roles). As a risky entrepreneurial start-up, Prosper was an ideal candidate for the ven­ture capital financing model. Yet it is difficult to see how those outside directors could avoid intractable fiduciary conflicts without first securing waivers defining the boundaries of their loyalty obligations across differ­ent companies. 26 For those who followed the well-known litigation in In re Trados Inc. S’holder Litig., 73 A.3d 17, 40–41 (Del. Ch. 2013), Prosper suggests a number of possibly instructive analogies.

That said, simply because there are plausible conditions under which COWs could enhance shareholder welfare, it does not follow that the firms actually adopting waivers satisfy those conditions. Our empirical analysis provides insights into this question as well. We find that COW adopters are, on average, reasonably established firms with moderate-to-high asset values. 27 See infra Table 6. They typically generate sizeable revenues, and they tend to deliver larger overall market returns to their capital investors by comparison to other public companies. 28 See infra Table 6. Delaware corporations are overrepresented, 29 See infra Table 5. as are firms in industries in which diversified, active investments across multiple portfolio companies are the norm (such as oil and gas). 30 See infra notes 258–259 and accompanying text. As a descriptive matter, then, it does not appear that companies that execute waivers are systematically the unscrupulous bottom feeders of the corporate ecosystem. To the con­trary, they appear—by and large—to be healthy, growing, and profitable business organizations.

In addition to these descriptive statistics, this Article further assesses whether the adoption of a waiver tends to add or dilute value on the margin by analyzing market reactions to issuers’ first public disclosure of a COW. 31 See infra section III.D. Our event-study analysis reveals that market reactions appear to be favorable, resulting in an average positive abnormal stock return hovering around one percent in the days immediately surrounding the announcement date. 32 See infra Table 7. While suggestive, the relationship is generally not statistically significant at conventional levels, underlining the need for additional future research. This positive reception is more pronounced for Delaware corporations and for those with asset values just below $1 billion. 33 See infra Table 8. Market response does not seem sensitive, in contrast, to whether the waiver also covers officers or dominant shareholders, nor does it appear to vary depending on whether the COW was adopted in a pro­posed charter amendment, a board-promulgated resolution, or some­thing else. 34 See infra Table 8. In addition, we demonstrate that firms incorporated in the nine states embracing Delaware-style reform have experienced positive abnormal stock-price returns on and around the date that such reforms became inevitable. 35 See infra Figure 6. All told, these findings suggest not only that public companies have embraced their newfound liberty to tailor the fiduciary duty of loyalty but also that they have done so in ways that shareholders welcome (or at least do not disfavor). Our findings are also pertinent to ongoing debates about shareholder activism, the appropriate role of “con­­stituency” directors, and whether the delineation of such roles should be subject to immutable rules or left up to the companies themselves. 36 See, e.g., J. Travis Laster & John Mark Zeberkiewicz, The Rights and Duties of Blockholder Directors, 70 Bus. Law. 33, 49–50 (2015) (arguing constituency directors should all be compelled to pursue long-term value for shareholders).

Several caveats warrant attention before proceeding. First, this Article’s study is limited to waiver disclosures contained in SEC filings. This data source imposes some unavoidable constraints. Most obviously, because privately held firms are far less likely to be SEC reporting entities, the ob­served disclosures skew towards publicly traded companies. In addition, the statutory provisions permitting COWs tend to grant wide latitude to companies about how and where they promulgate a waiver, and thus even for public companies, a waiver could be buried in any number of public filings. While it is important to remain mindful of these limitations, we note that a significant fraction of our data set does include private companies, including newly spun-off entities and private companies with public debt. Moreover, the protocol we designed for identifying candidate waivers casts a deliberately wide net, combing through the entire universe of SEC filings with search criteria designed deliberately to capture a large fraction of “false positive” candidates—a population we could later winnow using sampling, manual coding, and machine-learning techniques. 37 See infra section III.A.

Second, in practice, the disclosure of a COW is frequently “bundled” with other disclosures (such as an announced spin-off, carve-out, signif­icant investment, or reorganization). It is no doubt possible—and perhaps likely—that the market’s reception of a COW is really an amal­gamated reaction to the totality of contemporaneous disclosures. Failing to account for such confounding variables, one might argue, can lead to spurious conclusions about the market response to waivers in particular. Although this is an important concern, the appropriate response is not clear. Proponents of Delaware’s reform maintained that an enforceable COW is a critical precondition for the very transactions that now typically accompany it. To the extent that this claim is valid, it would be inappropriate not to analyze the COW alongside the other bundled disclosures it facilitates. Nevertheless, as a robustness check, we also exploit the time-lagged implementation of COW reforms at the state level, which started with Delaware and spread to eight other states over time. We similarly find evidence of a positive market response to passage of a COW reform among issuers incorporated inside the reforming jurisdiction. 38 See infra Figure 6.

Third, it is important to acknowledge that even if COWs help augment shareholder welfare, it need not follow that other corporate constituencies benefit as well (if at all). In particular, a recent literature in antitrust economics has begun systematically to document the rise of “horizontal” ownership structures in which a common investor holds appreciable equity across several ostensibly competing firms (often with board representation). 39 See José Azar et al., Anti-Competitive Effects of Common Ownership 12–26 (Ross Sch. of Bus., Working Paper No. 1235, 2016) [hereinafter Azar et al., Anti-Competitive Effects], (on file with the Columbia Law Review) (find­ing evidence of common-ownership structures in the airline industry and examining the effects of these ownership structures on ticket prices); see also Einer Elhauge, Horizontal Shareholding, 129 Harv. L. Rev. 1267, 1273–78 (2016) (reviewing the literature on “hori­zontal shareholding”). Such practices raise legitimate concerns that the common shareholder will help orchestrate anticompetitive coordination by the firms against their customers, employees, or trading partners. The ready availability of a COW may further exacerbate this problem, by permitting the common shareholder to choreograph collusion behind the scenes, with little fear of litigation risk from minority shareholders. Such concerns seem entirely plausible in our data: About one quarter of the waivers in the event-study sample (and about half in the full sample 40 Infra Table 4. ) extend protection to certain of the issuer’s shareholders. 41 Infra Table 8.

Finally, most statutes enabling companies to execute COWs (including Delaware’s) also subject that decision to a “back door” duty of loyalty analysis. 42 For a general discussion of this limitation (which appears in the statutory synopsis but not the statute itself), see infra section I.B.2. For example, if an interested director, officer, or dominant shareholder were to use her domination of the board to force through a self-serving COW, then courts could invalidate the prom­ulgation of the waiver itself as self-dealing. The issuers embracing waivers within our data set have done so against the backdrop of this liability exposure—one animated, ironically enough, by a lingering vestige of immutability within the duty of loyalty. Thus, one must be careful before using this study to draw inferences about the broader merits of default versus mandatory rules.

The remainder of the Article proceeds as follows. Part I revisits the broad contours of fiduciary duties, providing a brief overview of the duty of loyalty and the corporate opportunities doctrine. It also traces the evolution of COWs, from prereform waiver efforts, to their skeptical reception by Delaware’s courts, to the enactment of the Delaware reforms and related statutes, and finally to subsequent litigation address­ing waivers. Part II lays out a conceptual framework and model of efficient contracting over corporate opportunities in order to capture the contexts in which a COW would be value enhancing and the form it would plausibly take. Part III describes the empirical methodology behind our data set, offering a series of descriptive statistics about both the structure of COWs and the types of companies embracing them. Part III also reports on a series of event studies documenting positive market reactions to companies’ first disclosure of a COW. Part IV discusses the broader legal and policy implications of this analysis.

I. Revisiting the Duty of Loyalty

To provide legal background to the theoretical and empirical enterprise that follows, this Part gives an overview of the duty of loyalty and corporate opportunities doctrine (COD). Section I.A explores the broad doctrinal contours, while section I.B turns to the process by which Delaware and other states implemented statutory reforms empowering corporations to execute COWs.

A. Fiduciary Law and the Corporate Opportunities Doctrine

1. Fiduciary Duties and the Duty of Loyalty. — Corporate fiduciaries—the officers who manage a company’s daily operations, the directors who wield ultimate decisionmaking authority, and the dominant shareholders who possess swing voting power—exercise control over a vast amount of social resources on behalf of corporations’ ultimate owners, their shareholders. 43 See Singer v. Magnavox Co., 380 A.2d 969, 976–77 (Del. 1977) (noting that under Delaware law, officers, directors, and controlling shareholders are corporate fiduciaries); cf. Principles of Corp. Governance: Analysis and Recommendations §§ 3.01–.02 (Am. Law Inst. 1994) (describing the significant powers held by corporate officers and directors). Among the law’s principal tools for ensuring that corporate fiduciaries serve the interests of all of a corporation’s owners faithfully and competently are the fiduciary duties of loyalty and care.

The duty of care mandates that corporate fiduciaries exercise in­formed business judgment in their stewardship of the company, imposing liability if a fiduciary acts (or fails to act) without first being adequately informed. 44 Smith v. Van Gorkom, 488 A.2d 858, 872–73 (Del. 1985). While this duty could conceivably reach almost any major decision by corporate decisionmakers, a wide variety of judicial and private limitations cabin its scope. Alongside the famous defendant-friendly “business judgment rule,” 45 The business judgment rule is a “presumption that in making a business decision the directors of a corporation acted on an informed basis, in good faith and in the honest belief that the action taken was in the best interests of the company.” Aronson v. Lewis, 473 A.2d 805, 812 (Del. 1984). corporations are permitted to insure their directors and officers against breaches of the duty of care and to indemnify their directors for expenses incurred in connection with defending against allegations of breaches. 46 See, e.g., Del. Code Ann. tit. 8, § 145(a)–(g) (2017). Lastly, Delaware and the vast majority of other states allow parties to contract around the duty of care in various respects. For instance, since the 1980s, almost all states have permitted a corporation to adopt a charter provision limiting or eliminat­ing the personal liability of corporate directors for breaching the duty of care. 47 See id. § 102(b)(7) (empowering corporations to eliminate “the personal liability of a director to the corporation or its stockholders for monetary damages for breach of fiduciary duty as a director”); William T. Allen & Reinier Kraakman, Commentaries and Cases on the Law of Business Organization 229–30, 246 (5th ed. 2016). Public companies regularly execute such exoneration provisions. 48 See Holger Spamann, Monetary Liability for Breach of the Duty of Care?, 8 J. Legal Analysis 337, 338 (2016).

The duty of loyalty prohibits fiduciaries from benefiting improperly from financial conflicts of interest. 49 Allen & Kraakman, supra note 47, at 283 (“[I]nterested transactions are regulated first and foremost by the fiduciary duty of loyalty.”). In stark contrast to the duty of care, loyalty has traditionally been immutable. The Delaware statute that enables corporate charters to limit or eliminate directors’ monetary liability for breaches of fiduciary duty, for example, expressly excludes the duty of loyalty from its reach. 50 See tit. 8, § 102(b)(7) (specifically precluding a corporate charter from eliminating or limiting director liability “[f]or any breach of the director’s duty of loyalty to the corporation or its stockholders”). A small number of states’ corporate law may differ from Delaware in this respect. Nevada, for example, seems devoted to developing a niche as a near “liability-free” jurisdiction for managers. Michal Barzuza, Market Segmentation: The Rise of Nevada as a Liability-Free Jurisdiction, 98 Va. L. Rev. 935, 947–58 (2012). Under Nevada law, the default rule provides for no liability for a breach of the duty of loyalty, absent “intentional misconduct, fraud or a knowing violation of law.” Nev. Rev. Stat. § 78.138(7) (2003). Moreover, unlike with the duty of care, the deferential business judgment rule is also inapplicable to alleged breaches of loyalty. 51 See Continuing Creditors’ Comm. of Star Telecomm., Inc. v. Edgecomb, 385 F. Supp. 2d 449, 462 (D. Del. 2004) (“If a defendant does not breach his duty of loyalty to the company, he is permitted to rely on the business judgment rule . . . .”). Loyalty’s traditionally mandatory character is part of why commentators have widely held the duty to be “the most important duty which arises within the context of fiduciary relation­ships” 52 1 Ries, supra note 3, § 11:18; see also Frances S. Fendler, Losing Faith: Limited Liability Companies in Arkansas and the Fiduciary Duties of Loyalty and Good Faith, 31 U. Ark. Little Rock L. Rev. 245, 259 (2009) (“The duty of loyalty is preeminent in the constellation of the fiduciary duties recognized by common law.”); Thomas M. Griffin, Note, Investing Labor Union Pension Funds in Workers: How ERISA and the Common Law Trust May Benefit Labor by Economically Targeting Investment, 32 Suffolk U. L. Rev. 11, 22 (1998) (“Preeminent among the fiduciary duties is the duty of loyalty.”); supra note 3 (collecting several sources that note the importance of the duty of loyalty). as well as the subject of most fiduciary litigation. 53 See Allen & Kraakman, supra note 47, at 229.

The duty of loyalty requires fiduciaries to “exercise their authority in a good-faith attempt to advance corporate purposes.” 54 Id.; cf. Smith v. Van Gorkom, 488 A.2d 858, 872–73 (Del. 1985) (contrasting the duty of loyalty with the duty of care). While there is an affirmative dimension to this duty, its normal role is to bar self-interested action by officers or directors, which involves a conflict of interest with the corporation itself. 55 Allen & Kraakman, supra note 47, at 283. Perhaps the most colorful summary of the duty of loyalty is still the seminal opinion of Meinhard v. Salmon—a widely adopted and cited case 56 See, e.g., NLRB v. Amax Coal Co., 453 U.S. 322, 329–30 (1981) (citing Meinhard as to the “uncompromising rigidity” of fiduciary obligation); Guth v. Loft, Inc., 5 A.2d 503, 515 (Del. 1939) (citing Meinhard in characterizing fiduciary standards as requiring more than “the morals of the market place”); In re Galasso, 978 N.E.2d 1254, 1257 (N.Y. 2012) (citing Meinhard for the content of fiduciary obligation); In re Estate of Wallens, 877 N.E.2d 960, 962 (N.Y. 2007) (citing Meinhard for the proposition that fiduciary obligation involves “the punctilio of an honor the most sensitive”). involving a contested business opportunity of an unincorporated partnership. 57 164 N.E. 545, 546–47 (N.Y. 1928). Defendant Walter Salmon and plaintiff Morton Meinhard entered a joint venture to lease and improve a Manhattan hotel, which ultimately proved quite successful. 58 Id. at 545–46. As a result of the venture, Salmon (the active manager) came to owe fiduciary duties to Meinhard (a largely silent partner). 59 Id. at 546. Near the end of the lease, the hotel’s new owner offered Salmon a successor lease to manage and improve a vastly larger set of related properties. 60 Id. After the new lease had been signed, Meinhard, who was previously uninformed about the lease offer, discovered it and demanded a pro-rata share. 61 Id. A referee found that Meinhard was owed such an interest in the expanded venture in a fraction roughly proportionate to his interest in the first one. 62 Id.

In a majority opinion authored by Chief Judge Benjamin Cardozo, the high court of New York affirmed. 63 See id. at 549. The opinion was quickly recog­nized as one for the ages. Judge Cardozo famously declared that fiduciary ties demand a duty of “finest” and “undivided” loyalty, far “stricter than the morals of the market place,” concluding instead that “[n]ot honesty alone, but the punctilio of an honor the most sensitive, is then the standard of behavior.” 64 Id. at 546 (emphasis punctiliously added). By adopting “[u]ncompromising rigidity” in their approach to the duty of loyalty, the court articulated an aspiration that fiduciaries’ conduct would be “kept at a level higher than that trodden by the crowd.” 65 Id. (emphasis added). The broad language of Meinhard and similar cases becomes concrete in the various forms of managerial conduct regulated by the duty of loyalty, including transactions between a corporation and its directors, self-dealing, control transactions, and executive compensation. That said, this Article’s focus will be trained on—as was Chief Judge Cardozo’s—the allocation of new business opportunities between the firm and the fiduciary.

  1. The Corporate Opportunities Doctrine. — The COD is a key component of the duty of loyalty. 66 In re Cumberland Farms, Inc., 249 B.R. 341, 349 (Bankr. D. Mass. 2000) (“Perhaps the most common instance of breach of a director’s duty of loyalty is misappropriation of a corporate opportunity.”), aff’d sub nom. Haseotes v. Cumberland Farms, Inc., 257 B.R. 691 (D. Mass. 2001), aff’d sub nom. In re Cumberland Farms, Inc., 284 F.3d 216 (1st Cir. 2002); see also Myron T. Steele, Judicial Scrutiny of Fiduciary Duties in Delaware Limited Partnerships and Limited Liability Companies, 32 Del. J. Corp. L. 1, 10 (2007) (“Corporate opportunity ‘takings’ were, until [section 122(17)], a fundamental focus in the development of the constraints imposed by the common law on fiduciaries under the heading of the duty of loyalty.”). The basic idea of the COD is that corporate fiduciaries may not appropriate for themselves a new business opportunity that belongs to the corporation, unless they first present it to the corporation and receive authorization to pursue it personally. 67 See Guth v. Loft, 5 A.2d 503, 511 (Del. 1939) (noting the circumstances under which a corporate officer or director may not “seize” a corporate opportunity); 3 William Meade Fletcher, Fletcher Cyclopedia of the Law of Corporations § 862.05 (perm. ed., rev. vol. 2010) (“A director is free to take a business opportunity for himself or herself once the corporation has properly rejected the opportunity.”). As a result of the considerable amount of litigation arising from the COD and its complexity, it has over time developed its own labyrinth of rules, subcategories, standards, and tests.

The Delaware cases of Guth v. Loft 68 5 A.2d 503. and Broz v. Cellular Information Systems, Inc. 69 673 A.2d 148 (Del. 1996). provide defining benchmarks for the modern COD (both inside and outside Delaware). The doctrine states that an officer or director of a corporation usurps a business opportunity if:

(1) the corporation is financially able to [undertake] the oppor­tunity; (2) the opportunity is within the corporation’s line of business; (3) the corporation has an interest or expectancy in the opportunity; and (4) by [pursuing] the opportunity [per­son­ally], the corporate fiduciary will thereby be placed in a position inimicable [sic] to his duties to the corporation. 70 Id. at 155; see also Guth, 5 A.2d at 511. The same test continues to be employed consistently. See, e.g., In re Riverstone Nat’l, Inc. Stockholder Litig., No. 9796-VCG, 2016 WL 4045411, at *8–9 (Del. Ch. July 28, 2016) (applying the test to hold that the plaintiff adequately pled facts alleging that the directors of a cash-out merger target were not disinterested, since the directors’ personal exposure to viable corporate opportunity claims would be “obliterated” under the terms of the acquisition).

Courts engage in a fact-intensive inquiry to determine whether a given fiduciary’s pursuit of a business opportunity was impermissible under this multifactor test. 71 See, e.g., Broz, 673 A.2d at 151.

A simpler way to think of the doctrine, however, is as a sequential inquiry. The first major question is whether a given business prospect constitutes a bona fide “corporate opportunity,” thereby posing a genuine conflict of interest. 72 See Eric Talley, Turning Servile Opportunities to Gold: A Strategic Analysis of the Corporate Opportunities Doctrine, 108 Yale L.J. 277, 288–89 (1998) (presenting a summary algorithm for analyzing COD cases). A number of fairly involved tests are employed by courts in determining whether a given business prospect is a corporate opportunity. 73 The three major tests are the “line-of-business” test, see, e.g., Guth, 5 A.2d at 511, the “interest-or-expectancy” test, see, e.g., Lagarde v. Anniston Lime & Stone Co., 28 So. 199, 201 (Ala. 1900), and the “fairness” test, see, e.g., Durfee v. Durfee & Canning, Inc., 80 N.E.2d 522, 529 (Mass. 1948). Additionally, some courts have adopted a test created by the American Law Institute. See, e.g., Ne. Harbor Golf Club, Inc. v. Harris, 661 A.2d 1146, 1150–52 (Me. 1995); Principles of Corp. Governance: Analysis and Recommendations §§ 5.05, 5.12 (Am. Law Inst. 1994). If the prospect is not a corporate opportunity, then a fiduciary does no wrong by simply pursuing it herself without informing the corporation. 74 See Talley, supra note 72, at 287–88. If the prospect is a corporate opportunity, though, then the appropriate course of action is to offer it to the corporation. 75 Id. If the corporation properly rejects that opportunity—paradigmatically, by a majority vote of disinterested directors—then the fiduciary again does no harm by pursuing it, while if the corporation does not reject it, the fiduciary is barred from pursuit of the prospect. 76 Id.

There are also several affirmative defenses that materially shape the contours of the COD. Thus, some courts have found fiduciaries not to have usurped corporate opportunities because they encountered the opportunity in their personal capacity; 77 Cf. Weiss v. Kay Jewelry Stores, Inc., 470 F.2d 1259, 1271 (D.C. Cir. 1972) (recognizing the fact that an “opportunity . . . came to [the defendant] in his individual capacity in a non-corporate matter lying outside the field of his duties as a director” as undermining the corporation’s corporate opportunity claim). because the corporation impliedly rejected the opportunity; 78 See, e.g., Lussier v. Mau-Van Dev., Inc., 667 P.2d 804, 813 (Haw. Ct. App. 1983) (finding “implied consent by the shareholders”). and most importantly, because the corporation was not able to pursue a given business pros­pect, perhaps because the corporation’s financial condition precluded it from doing so. 79 See, e.g., Broz v. Cellular Info. Sys., Inc., 673 A.2d 148, 155 (Del. 1996). If even this stylized description of the COD sounds involved, that is largely a reflection of the doctrine itself. The law’s attempt to regulate fiduciaries’ independent pursuit of business opportu­nities has produced a doctrine of startling complexity and unpredictability. 80 See Talley, supra note 72, at 279 n.2 (“[T]here has been much confusion about the specific extent of [the fiduciary] duty when . . . it is contended that a fiduciary takes for herself a corporate opportunity.” (alteration in original) (internal quotation marks omitted) (quoting Ne. Harbor Golf Club, Inc. v. Harris, 661 A.2d 1146, 1148–49 (Me. 1945))); see also Miller v. Miller, 222 N.W.2d 71, 79 (Minn. 1974) (“We have searched the case law and commentary in vain for an all-inclusive or ‘critical’ test or standard by which a wrongful appropriation can be determined and are persuaded that the doctrine is not capable of precise definition.”); Robert Charles Clark, Corporate Law §7.6.2, at 244–45 (1986) (“The traditional tests are extremely ambiguous and uncertain in their application.”).

Not only is the COD a heavily litigated 81 A Westlaw search found 8,679 cases involving the COD.  Westlaw, http:// (filter search by “All States” and “All Federal”; then search in search bar for “corporate /s opportunit!”) (last visited Mar. 13, 2017). For one example arising in the bankruptcy context, see In re Cumberland Farms, Inc., 249 B.R. 341, 349–54 (Bankr. D. Mass. 2000), aff’d sub nom. Haseotes v. Cumberland Farms, Inc., 257 B.R. 691 (D. Mass. 2001), aff’d sub nom. In re Cumberland Farms, Inc., 284 F.3d 216 (1st Cir. 2002).
and murky subspecies of the duty of loyalty, but the form of misconduct it targets—the usurpation of business opportunities—is also one of the most pernicious types of agency costs. When the issue is one of who “owns” rights to a business opportunity, then the interests of a corporation and its agent are not merely misaligned; they may be completely at odds with each other. This difficulty is compounded by the fact that the COD addresses not only behavior that can reduce the value of a firm but also behavior in which a fiduciary may usurp from its principal a new venture whose value significantly exceeds that of the corporation—as Meinhard v. Salmon 82 164 N.E. 545 (N.Y. 1928). illustrates. In Meinhard, the initial lease had been for $55,000, while the new one was for $350,000 to $475,000; the initial building improvements had been for $200,000, but were now to cost $3,000,000. 83 Id. at 546–47.

B. Evolution of the COW: A Brief History of Endeavors to Contract Out of the Corporate Opportunities Doctrine

As noted in the previous section, the COD has always permitted boards of directors to “reject” a corporate opportunity ex post—after it has emerged and has been properly presented to the company by a fiduciary interested in pursuing it personally. 84 See supra note 76 and accompanying text. Just as with other forms of independent board dispensation, however, this authorization power historically did not apply to the prospective (or ex ante) waiver as to opportunities or projects that had yet to arise. In fact, the notion that parties might contract around (or out of) the duty of loyalty in advance was traditionally considered anathema to foundational commitments of corporate law, of which the duty of loyalty is one of few mandatory components. 85 See, e.g., John C. Coffee, Jr., The Mandatory/Enabling Balance in Corporate Law: An Essay on the Judicial Role, 89 Colum. L. Rev. 1618, 1649 (1989) (“American courts clearly have been hostile to most attempts to exculpate with respect to the duty of loyalty.”); Jeffrey N. Gordon, The Mandatory Structure of Corporate Law, 89 Colum. L. Rev. 1549, 1593 (1989) (“[O]pting out of fiduciary duties [is] particularly troublesome and ultimately wrong-headed, especially for elements of the duty of loyalty.”). For the most part, that is still how much of fiduciary law operates (at least in Delaware corporations). 86 Sutherland v. Sutherland, No. 2399-VCL, 2009 WL 857468, at *4 (Del. Ch. Mar. 23, 2009) (“While . . . a provision [limiting the fiduciary duty of loyalty] is permissible under the Delaware Limited Liability Company Act and the Delaware Revised Uniform Limited Partnership Act, where freedom of contract is the guiding and overriding principle, it is expressly forbidden by the [Delaware corporate statute].”). There is some variation among states, to be sure, but the nonwaivability of the duty of loyalty appears relatively constant. See, e.g., Nev. Rev. Stat. § 78.138(7) (2003) (imposing a floor of liability on directors for breach of fiduciary duty accompanied by intentional misconduct, fraud, or a knowing violation of law); In re Amerco Derivative Litig., 252 P.3d 681, 704 (Nev. 2011) (reversing dismissal below and finding plaintiffs adequately pled a breached duty of loyalty for corporate opportunity appropriation under Nevada law). But the late-twentieth century bore witness to several early attempts to chisel the edges of the status quo, at least insofar as it pertained to COWs. The judicial response to those early attempts, in turn, arguably catalyzed the subsequent statutory innovations.

1. Primordial COWs. — Prior to the amendment of Delaware’s Code in 2000, no specific statutory authority empowered companies explicitly to contract out of the COD in advance. 87 Indeed, as Siegman v. Tri-Star Pictures, Inc. noted—and we discuss shortly—the law of Delaware positively prohibited such attempts to contract out of the COD. See No. 9477, 1989 WL 48746, at *7–8 (Del. Ch. May 30, 1989). A company motivated to do so in a legally legitimate fashion had few, if any, options, save a “nuclear” one: A corporation could theoretically—in many states—attempt to cabin the breadth of the doctrine by narrowing the purpose articulated in its charter to specified lines of business, effectively using that scope limitation to cabin the reach of all corporate activity, including the COD. 88 Cf. Zenichi Shishido, Conflicts of Interest and Fiduciary Duties in the Operation of a Joint Venture, 39 Hastings L.J. 63, 94–95 (1987) (noting that a court could find no misappropriation of a corporate opportunity when that opportunity was “ultra vires”—outside the scope of the corporation’s purposes). Such measures, however, invite a host of other ultra vires challenges to corporate decisionmaking 89 See Michael A. Schaeftler, Clearing Away the Debris of the Ultra Vires Doctrine—A Comparative Examination of U.S., European, and Israeli Law, 16 Law & Pol’y Int’l Bus. 71, 80, 86 (1984) (examining the potential consequences of a narrow purpose clause, including ultra vires suits and limits on corporate executives’ authority). —obstacles that have caused limited-purpose provisions to be disfavored and exceedingly rare in modern times. 90 Id. at 88 (“[A] majority of U.S. corporations have purpose clauses which either encompass every conceivable lawful business activity by means of boiler-plate language, or state simply that the corporation may engage in any lawful activity.”). Indeed, conventional corporate charter wisdom has long advocated extremely broad purpose provisions, authorizing the corporation to engage in “any lawful act or activity” for which corpora­tions may be organized under the applicable corporate statute. 91 See 1A Fletcher, supra note 67, § 91 (discussing the near ubiquity of general purpose corporate statutes, including in the Model Business Corporations Act); see also 1 Marvin Hyman, Corporation Forms § 1:13, Westlaw (database updated Mar. 2017) (“Present practice, which is permitted by most state statutes, is to simply include a sentence providing that the corporation can engage in any other activity permitted by law.”); id. § 1:16 (offering as standard form language for New York corporations the power “[t]o engage in any lawful act or activity for which corporations may be organized under the Business Corporation Law”).

A more tailored form of carve out, on the other hand, had speculative legal validity: Prior to 2000, Delaware statutes did not explicitly permit (or even appear to contemplate) contracting out of the COD. 92 See supra note 87 and accompanying text. On the other hand, neither did Delaware law unambiguously prohibit the practice. Towards the end of the twentieth century, several corporations began to experiment with such provisions—experiments that inevitably attracted legal challenges. Perhaps the best-known example was the 1989 Delaware Chancery Court decision in Siegman v. Tri-Star Pictures, Inc. 93 No. 9477, 1989 WL 48746 (Del. Ch. May 5, 1989); see also Senate Bill 363: Original Synopsis, Del. Gen. Assembly,
10399 [] (last visited Jan. 31, 2017) (“The subsection is intended to eliminate uncertainty regarding the power of a corporation to renounce corporate opportunities in advance raised in Siegman v. Tri-Star Pictures, Inc. . . . .”).
Siegman was a putative class action challenging a proposed combination between Tri-Star Pictures, Inc. (“Tri-Star”), 94 When it formed in 1985, Tri-Star was the first major new movie studio since RKO was formed in 1927. Jennifer Holt, ‪Empires of Entertainment: Media Industries and the Politics of Deregulation, 1980-1996, at 45 (2011). Coca-Cola, and Time. Under the terms of the contemplated transaction, Tri-Star acquired the entertainment assets of Coca-Cola, and Coca-Cola received a large number of shares of newly issued Tri-Star common stock. 95 Siegman, 1989 WL 48746, at *1–2. Substantial portions of the opinion (denying three motions to dismiss) are devoted to procedural issues, justiciability issues, unrelated charter provisions, and other issues that are beyond our remit for current purposes. See id. at *3–6, *9–12.

The plaintiffs challenged the validity of several proposed amend­ments to Tri-Star’s certificate of incorporation (executed as part of the combination). One such amendment purported to eliminate liability for Tri-Star’s directors for breach of the duty of loyalty under specified circumstances involving the appropriation of corporate opportunities. 96 Id. at *7–8. Another amendment provided that neither Coca-Cola nor Time, as signi­ficant block stockholders of Tri-Star, would be liable for any breach of fiduciary duty stemming from having pursued a corporate opportunity belonging to Tri-Star. 97 Id. The business combination was approved by both Tri-Star and Coca-Cola, and the proposed amendments were subse­quently adopted by shareholders. 98 Id. at *2.

The gravamen of the complaint centered on the director provision, asserting that the COW purported to eliminate or limit liability in a way that was simply impermissible under Delaware law. Specifically, plaintiffs argued that section 102(b)(7) of the Delaware General Corporation Law (DGCL) did not permit elimination or restriction of directors’ liability for breach of the duty of loyalty. 99 See id. at *7. The statute provides express limitations on exoneration provisions and excludes any waiver “[f]or any breach of the director’s duty of loyalty to the corporation or its stockholders.” 100 Del. Code Ann. tit. 8, § 102(b)(7) (2017). Effectively, the plaintiffs asserted that the COW amendment purported to do exactly this: reduce the directors’ liability exposure for a particular type of duty of loyalty breach (an appropriation of corporate oppor­tunities), a move that the plaintiffs argued transgressed the immu­table boundaries of section 102(b)(7). 101 See Siegman, 1989 WL 48746, at *7 (“Plaintiff contends that Article Sixth is invalid as a matter of law, because it eliminates or restricts the directors’ liability to the corporation or its shareholders for breaches of their fiduciary duty of loyalty, in violation of 8 Del.C. § 102(b)(7).”). The defendants countered that the provi­sion was valid and enforceable under Delaware law, and they moved to dismiss. 102 See id. at *4, *7 (noting defendants moved to dismiss on the grounds that the provision was “valid as a matter of law” and describing their arguments).

Then-Vice Chancellor Jacobs sided with the plaintiffs, noting that the appropriate judicial standard for a motion to dismiss in this context “requires that the motion must be denied if under any plausible construction or operation,” the COW “arguably would contravene” Delaware law. 103 See id. at *7–8. Employing this analytical approach, Vice Chancellor Jacobs determined that at least one plausible set of facts would—under the articulated terms of the charter provision—eliminate or limit the liability of Tri-Star directors for breach of their fiduciary duty of loyalty. 104 See id. at *8. Indeed, Vice Chancellor Jacobs envisioned a very general scenario as violating section 102(b)(7): A director appropriates for herself or her other employer a business opportunity that rightly belongs to Tri-Star but had not been offered to that director in her capacity as a Tri-Star director or in writing. 105 See id. Finding such a result to be impermissible under the limits established by section 102(b)(7), Vice Chancellor Jacobs denied the defendants’ motion to dismiss. 106 See id. As the Chancery Court put it, “Article Sixth would eliminate or limit the liability of Tri-Star directors for breach of their fiduciary duty of loyalty—a result proscribed by § 102(b)(7).” Id. It bears observing that Jacobs was more sympathetic to a different contractual provision that purported to limit or eliminate directorial liability to the fullest extent permitted by the DGCL in the event it were amended in the future. See id. at *8–9. Here, Vice Chancellor Jacobs found that neither the statute nor its underlying policy forbids such prospective planning in legislative enact­ments or amendments. See id.

2. Legalizing COWs in the Delaware General Corporation Law. — Siegman substantially put to rest the question of how and when corporate opportunities were waivable. 107 There were precious few post-Siegman cases that squarely dealt with advance COWs until after the amendment of the DGCL in 2000 (even though citations to the case made occasional appearances in practitioner journals and law reviews). See, e.g., Theodore D. Moskowitz & Walter A. Effross, 23 Seton Hall L. Rev. 897, 916 & n.141 (1993) (citing Siegman); infra note 108 and accompanying text (collecting cases that cite Siegman). Thus, much of the practical wake of Siegman was the shadow it appears to have cast over corporate conduct for the ensuing decade, an inference bolstered by the legislative history of the ensuing reform. See infra note 125 and accom­panying text. Under any fair reading of the opinion, the duty of loyalty was simply not contractible, be it through a corporate governance provision, a board resolution, or through a contractual term. In the ensuing decade, the Siegman opinion stayed in full force and was rarely discussed by subsequent judicial opinions (inside or outside Delaware). 108 The opinions that did cite Siegman generally did so for issues unrelated to the COD. See, e.g., Katz v. Pels, 774 F. Supp. 121, 129 (S.D.N.Y. 1991); Ryan v. Aetna Life Ins., 765 F. Supp. 133, 138 (S.D.N.Y. 1991); In re Tri-Star Pictures, Inc., Litig., 634 A.2d 319, 325 (Del. 1993); Harbor Fin. Partners v. Huizenga, 751 A.2d 879, 888 (Del. Ch. 1999); Levine v. Smith, No. 8833, 1989 WL 150784, at *6 n.5 (Del. Ch. Nov. 27, 1989), aff’d, 591 A.2d 194 (Del. 1991); In re Tri-Star Pictures, Inc., Litig., No. 9477, 1989 WL 112740, at *1 (Del. Ch. Sept. 26, 1989).

By the end of the twentieth century, however, market dynamics began biting at the jurisprudential heels of the Siegman approach. The dot-com era of the 1990s ushered in a wave of novel market-mediated corporate structures, including spin-offs, partial IPOs, venture capital, private equity, and equity carve-outs. 109 See, e.g., Karen M. Hogan & Gerard T. Olson, The Pricing of Equity Carve-Outs During the 1990s, 27 J. Fin. Res. 521, 528 (2004) (showing growth of equity carve-outs in the early 1990s); Steven N. Kaplan & Per Strömberg, Financial Contracting Theory Meets the Real World: An Empirical Analysis of Venture Capital Contracts, 70 Rev. Econ. Stud. 281, 286–95 (2003) (discussing the novel features of venture capital financing contracts); Steven N. Kaplan & Per Strömberg, Leveraged Buyouts and Private Equity, 23 J. Econ. Persp. 121, 126 (2009). Many of these innovations resulted in extended families of corporate affiliates with partially overlapping ownership, partially overlapping board membership, and partially overlapping lines of business. 110 See Hogan & Olson, supra note 109, at 529, 530 tbl.2, 535 tbl.5 (reporting descriptive statistics related to “share overhang”—shared control of the parent and spun-off subsidiary—and its predictive effect on excess returns). Such structures, in turn, placed considerable stress on the canonical “undivided-loyalty” model of corporate opportunities. Any time a fiduciary’s duties extended to multiple affiliated entities (as was increasingly frequent), she faced an unwinnable Kobayashi-Maru scenario 111 See, e.g., Janet D. Stemwedel, The Philosophy of Star Trek: The Kobayashi Maru, No-Win Scenarios, And Ethical Leadership, Forbes (Aug. 23, 2015, 10:18 AM), (on file with the Columbia Law Review). of carving up what was judicially indivisible: her loyalty. Consider, for example, the conundrum of allocating corporate opportunities between a parent and its partially owned subsidiary, both operating in a similar industry and sharing common board members and officers. How might those overlapping fiduciaries (or for that matter, the parent, as the dominant shareholder of the subsidiary) comply with their simultaneous duties of “undivided loyalty” between the two firms? How should they go about allocating corporate opportunities?

These questions are profound and probably unanswerable. Scholars have long recognized that the undivided-loyalty model is simply not well adapted for fiduciaries shared by two companies. 112 See, e.g., Clark, supra note 80, § 7.3, at 231–34 (summarizing case law and characterizing two seemingly inconsistent “riddles” posed by alternative accepted approaches to COD cases). In fact, the canonical approach may be the least attractive from the parties’ perspectives, since a time-honored prescription for conduct of “dual agents” under fiduciary law is to disclose the corporate opportunity to both interested corpo­rations, effectively encouraging the two to compete with one another for the new business prospect. 113 See, e.g., Energy Res. Corp. v. Porter, 438 N.E.2d 391, 394 (Mass. App. Ct. 1982) (concluding a third party’s purported “refusal to deal” with a disfavored corporation was not credible unless first “adequately tested” by disclosing to the disfavored corporation so as “to verify the unwillingness to deal”); Meinhard v. Salmon, 164 N.E. 545, 547 (N.Y. 1928) (“The trouble about [defendant’s] conduct is that he excluded his coadventurer from any chance to compete, from any chance to enjoy the opportunity for benefit that had come to him alone by virtue of his agency. This chance, if nothing more, he was under a duty to concede.”). While such competition is no doubt attractive to counterparties in the business opportunity (offering signif­icant transactional surplus), it seems an unlikely governance feature for augmenting the combined welfare of the fiduciary and the two bene­ficiaries of her duties.

Nevertheless, by the turn of the century, the corporate structures described above had become increasingly common. Two high-stakes cases during that era—Thorpe v. CERBCO, Inc. 114 676 A.2d 436 (Del. 1996). and In re Digex, Inc. 115 789 A.2d 1176 (Del. Ch. 2000). —help underscore the resulting challenges. Both cases focused centrally on corporate opportunities claims made by shareholders of a controlled subsidiary, asserting that the parent had usurped a corporate opportunity related to an acquisition of the subsidiary. 116 See Thorpe, 676 A.2d at 438; Digex, 789 A.2d at 1178. In both cases, the plaintiffs alleged that the controller had commandeered takeover negotiations with a third-party buyer, redirecting the buyer’s interest towards the parent and away from the subsidiary, thereby fleecing the subsidiary’s minority shareholders of their impending control premium. 117 See Thorpe, 676 A.2d at 439; Digex, 789 A.2d at 1178. In both cases, the corporate opportunities claims either narrowly lost or were substantially narrowed, under the theory that the prospective acquisition deal was not a corporate opportunity since the parent possessed the power (and the right) to use its voting shares to veto any proposed transaction at the subsidiary level. 118 See Thorpe, 676 A.2d at 443; Digex, 789 A.2d at 1190. In Thorpe, the Chancery Court found (and the supreme court concurred) that a lack of disclosure by the defendants to the plaintiff substantiated a formal breach of the COD. However, the court found that because the controllers had the votes to block any alternative transaction, plaintiffs were not entitled to transactional damages and instead were limited to reliance damages associated with initial negotiations with the third-party acquirer. See Thorpe, 676 A.2d at 445. Nevertheless, both opinions recog­nized the generic and intractable challenges posed by corporate opportunities claims in cases involving ownership–board–industry overlap. 119 See Thorpe, 676 A.2d at 442; Digex, 789 A.2d at 1193. The cases served up a sobering reminder of the uncom­fortable indeterminacy of corporate opportunity claims when firms have overlapping dominant ownership or boards. They also made apparent that there might be some value in allowing parties to prearrange how they would divide property rights over corporate opportunities. Thus was the stage set for a legal-reform push at the dawn of the twenty-first century. 120 See Lewis S. Black, Jr. & Frederick H. Alexander, Analysis of the 2000 Amendments to the Delaware General Corporation Law, in Contests for Corporate Control: Current Offensive & Defensive Strategies in M&A Transactions 729, 732–34 (2002) (providing contemporaneous commentary on section 122(17)’s enactment and noting the sale by a parent of minority interests in a subsidiary as an important business context affected by the statute’s adoption).

The wait was not long. In the summer of 2000, the Delaware Assembly amended the state’s statutes to add subsection (17) to section 122 of the DGCL, explicitly permitting COWs. 121 72 Del. Laws 619 (2000). The codified location is a curious one. Section 122 is a general and longstanding provision whose purpose is to articulate a variety of powers possessed by Delaware corporations. 122 See Del. Code Ann. tit. 8, § 122 (2017). Many of these powers are fundamental (such as the right to hold property, the right to enter contracts, and the right to sue and be sued). 123 See id.; cf. 12 U.S.C. § 24 (2012) (codifying many similar fundamental powers for national banking associations). The new subsection (17), however, was different and somewhat sui generis, constituting the sole provision in the section that concerns fiduciary duties. The subsec­tion provides that a Delaware corporation has the power to:

(17) Renounce, in its certificate of incorporation or by action of its board of directors, any interest or expectancy of the corpo­ration in, or in being offered an opportunity to participate in, specified business opportunities or specified classes or cate­gories of business opportunities that are presented to the corporation or one or more of its officers, directors or stockholders. 124 tit. 8, § 122(17).

The legislative synopsis accompanying the amendment further states:

The subsection is intended to eliminate uncertainty regarding the power of a corporation to renounce corporate oppor­tunities in advance raised in Siegman v. Tri-Star Pictures, Inc. It permits the corporation to determine in advance whether a specified business opportunity or class or category of business opportunities is a corporate opportunity of the corporation rather than to address such opportunities as they arise. The subsection does not change the level of judicial scrutiny that will apply to the renunciation of an interest or expectancy of the corporation in a business opportunity, which will be determined based on the common law of fiduciary duty, including the duty of loyalty. 125 Del. Gen. Assembly, supra note 93 (citation omitted); see also 72 Del. Laws at 619.

Several aspects of the amendment and its synopsis warrant elaboration. First, and most obviously, the synopsis makes clear that the amendment was meant to repudiate the then-decade-old Siegman approach. Indeed, the amendment specifically permits enforceable COWs under Delaware law, a position that—both before and after Siegman—most had con­sidered untenable.

Second, the new section explicitly applies symmetrically to all corporate fiduciaries, including officers, directors, and dominant or controlling shareholders. By contrast, the statutory provision permitting duty of care waivers is far narrower, applying only to monetary damages claims against corporate directors; 126 See tit. 8, § 102(b)(7) (providing the certificate of incorporation can include “[a] provision eliminating or limiting the personal liability of a director to the corporation or its stockholders for monetary damages for breach of fiduciary duty as a director”). it does not extend to either injunc­tive relief of any kind or claims for monetary damages lodged against offi­cers or dominant shareholders. This asymmetry has been consis­tently recog­nized by the Delaware courts (even if periodically scorned by commentators). 127 Gantler v. Stephens, 965 A.2d 695, 709 nn.36–37 (Del. 2009). Although several commentators—including a former Delaware Vice Chancellor—have criticized this asymmetry, it continues to persist. See, e.g., Stephen P. Lamb & Joseph Christensen, Duty Follows Function: Two Approaches to Curing the Mismatch Between the Fiduciary Duties and Potential Personal Liability of Corporate Officers, 26 Notre Dame J.L. Ethics & Pub. Pol’y 45, 46 (2012) (stating section 102(b)(7)’s contribution to corporate law is “incomplete” because “the Delaware Supreme Court held that officers owe the same fiduciary duties as directors . . . but cannot be exculpated for the same class of fiduciary breaches as directors”); see also Andrew D. Appleby & Matthew D. Montaigne, Three’s Company: Stone v. Ritter and the Improper Characterization of Good Faith in the Fiduciary Duty “Triad,” 62 Ark. L. Rev. 431, 469 (2009) (“While Gantler leaves officers extremely vulnerable to Van Gorkom-like liability, the current ‘anti-executive social-political climate’ may preclude the courts or legislature from extending officers any section 102(b)(7) protection.” (quoting Usha Rodrigues, Delaware v. The Feds on Officer Regulation, Conglomerate (Feb. 4, 2009), [])); Dennis R. Honabach, Smith v. Van Gorkom: Managerial Liability and Exculpatory Clauses—A Proposal to Fill the Gap of the Missing Officer Protection, 45 Washburn L.J. 307, 307–08 (2006) (advocating extension of protection to officers); Meghan Glaspy, Note, Delaware’s Gantler Decision: A Solution to Corporate Corruption?, 12 Geo. J.L. & Pub. Pol’y 289, 307 (2014) (arguing “there is no . . . reason to hold officers to a higher standard” given that “officers and directors have the same fiduciary duties”).

Third, consider the level of specificity required for a corporate opportunities waiver to be effective under the statute. On its face, the amendment requires a COW to be worded with some particularity, identifying “specified business opportunities or specified classes or categories of business opportunities” that would be subject to waiver. 128 tit. 8, § 122(17). This construction appears in tension with an expansive, “blanket” waiver, in which a corporation, say, disclaims all corporate opportunities, or even a broad waiver purporting to disclaim all opportunities except a specified set of carved out exceptions. In contractual parlance, the wording of the section appears, at least facially, to suggest a type of sticky default rule that does not invite corporations to invert or “flip” the default, either in toto or in substantial part. 129 As discussed below, however, a portion of the legislative synopsis casts some doubt on the requirement of specificity for an effective COW. See infra text accompanying note 146.

Fourth, consider the means and location for executing an effective COW. The text of the section permits a waiver to be included in a corporation’s charter, but it also allows a waiver to simply be adopted by action of the board of directors. Bearing in mind that a charter amendment cannot generally be promulgated exclusively by action of the board once stock is sold, 130 See tit. 8, § 242(b)(1) (requiring a shareholder vote for most charter amendments). the portion that authorizes an action of the board opens up tremendous latitude, permitting COWs to be couched in a contract approved by the board, a board-promulgated bylaw (if the board has such power under the charter), 131 Section 109 of the DGCL allows the certificate of incorporation to grant the board concurrent power to pass, amend, and repeal corporate bylaws. See id. § 109(a). a board resolution, or any other declarative action promulgated by the board. The fact that a COW may be executed outside the charter context is notable, since the other principal means by which corporations may waive fiduciary duties (the duty of care) specifically requires a charter provision. 132 Id. § 102(b)(7). Indeed, it was the prohibition of waiving the duty of loyalty under section 102(b)(7) that formed the basis of the Siegman opinion in the first place. 133 See supra notes 99–104 and accompanying text. Section 122(17) is thus in some conspicuous ways broader than conventional waivers of the duty of care.

Finally, while section 122(17) grants increased latitude for enforcing valid COWs, the legislative synopsis notes that the initial adoption of a COW is subject to traditional fiduciary principles—just as the renunciation of an opportunity when it arises. Thus, should a financially conflicted board decide to adopt a COW without first seeking to cleanse the deci­sion through conventional means (such as a vote of disinterested direc­tors or shareholders), 134 See tit. 8, § 144. then the very act of executing the COW could be challenged under Delaware’s stringent entire fairness standard (and pos­sibly invalidated). 135 Cf. Emerald Partners v. Berlin, 787 A.2d 85, 97 (Del. 2001) (describing the enhanced entire fairness standard of scrutiny that applies to certain conflicted transactions and requires “fair dealing and fair price”).

3. Subsequent Delaware Litigation. — The footprint of Delaware’s statutory reform in case law and commentary has been surprisingly faint. In the decade and a half since section 122(17) was promulgated, the statute appears to have been invoked in only a single Delaware case and in only a few secondary sources. 136 A small number of other cases discuss waivers fleetingly but with no reference to the DGCL. A Westlaw search for COWs after 2000 yielded eight cases and thirty-one secondary sources. Westlaw, (filter search by “All States” and “All Federal”; then search in search bar for “‘corporate opportunit!’ /10 waiv! and DA(aft2000)”) (last visited Mar. 13, 2017). By contrast, the enabling statute for duty of care waivers has been invoked in at least 269 cases and 358 secondary sources over that same period. Id. (filter search by “All States” and “All Federal”; then search in search bar for “(‘8 Del. C. s 102(b)(7)’) or (‘waive!’ /10 ‘duty #of care’) and DA(aft2000)”) (last visited Mar. 13, 2017).

The 2009 Delaware Chancery Court opinion in Wayne County Employees’ Retirement System v. Corti 137 No. 3534-CC, 2009 WL 2219260, at *1 (Del. Ch. July 24, 2009), aff’d, 996 A.2d 795 (Del. 2010). appears to be the only Delaware opinion to date to engage the statutory framework for COWs explicitly (although cursorily). Like Siegman, Corti was a purported shareholder class action challenging a business combination involving a waiver of corporate opportunities. 138 See id. at *17. The combination called for Vivendi S.A. (Vivendi) to transfer its subsidiary, Vivendi Games, Inc., to Activision, Inc. in return for newly issued shares of Activision and a post-closing tender offer by Vivendi for up to half of Activision’s remaining shares. 139 Id. at *1. Together, Vivendi’s acquisition of shares through the business combi­nation and back-end tender would result in Vivendi acquiring a majority of Activision voting stock, which it then renamed Activision Blizzard. 140 Id. The charter of the surviving corporation (Activision Blizzard) included a broadly-worded COW. 141 The charter provision read:
“In the event that a director or officer of the Corporation who is also a director, officer or employee of Vivendi acquires knowledge of a potential transaction or matter which may be a corporate opportunity for both the Corporation and Vivendi (a ‘Mutual Corporate Opportunity‘), such director or officer shall to the fullest extent permitted by law have fully satisfied and fulfilled his fiduciary duty with respect to such Mutual Corporate Opportunity, and the Corporation to the fullest extent permitted by law waives and renunciates [sic] any claim that such Mutual Corporate Opportunity constituted a corporate opportunity that should have been presented to the Corporation, if such director or officer acts in a manner consistent with the following policy: a Mutual Corporate Opportunity offered to any person who is an officer or director of the Corporation, and who is also an officer, director or employee of Vivendi, shall belong to Vivendi, unless such Mutual Corporate Opportunity was expressly offered to such person in his or her capacity as a director or officer of the Corporation (an ‘Activision Opportunity‘), in which case such Activision Opportunity shall not be pursued by Vivendi. In the event Vivendi decides to pursue any Mutual Corporate Opportunity (other than an Activision Opportunity), then, subject to any contractual restrictions on Vivendi with respect to confidentiality, Vivendi shall provide prompt written notice to the Corporation of such decision.”
Id. at *17. The Activision Blizzard COW’s text helps motivate some of our decisions in coding the COW database detailed and summarized in the next section.

Plaintiff—a shareholder of the target Activision—alleged, inter alia, that the COW was invalid under Delaware law because it contravened section 122(17)’s limitations through its sweeping language. The provision, plaintiff contended, failed to specify explicitly which corporate opportunities (or classes or categories thereof) the corpora­tion was renouncing as the statutory text arguably requires. 142 Id. at *1; cf. Del. Code Ann. tit. 8, § 122(17) (2017) (referring only to “specified business opportunities or specified classes or categories of business opportunities”). Rather, the provision utilized the opposite, holus-bolus grammatical construction, categorically sweeping away all liability exposure with the exception of opportunities that were expressly offered to Activision fiduciaries in their capacity as such. 143 Corti, 2009 WL 2219260, at *1. In short, plaintiff averred, the Activision Blizzard waiver fell outside the bounds authorized by section 122(17).

Chancellor Chandler remained uncowed, denying plaintiff’s motion for a preliminary injunction of the contemplated transaction and hold­ing that “[t]he mere existence of [the broadly worded COW] does not threaten plaintiff with harm that justifies expending judicial resources to render a declaratory judgment on the issue of whether the corporate opportunities allegedly renounced by [the COW] are sufficiently ‘specified.’” 144 Id. at *18. Any plausible harm to plaintiff due to the wording of the waiver, Chancellor Chandler concluded, “is too remote and speculative to justify rendering a declaratory judgment, and plaintiff is not entitled to a declaratory judgment merely because it is able to conjure up hypothetical situations in which the challenged provisions may be applied contrary to Delaware law.” 145 Id. at *19. Ultimately, Chancellor Chandler took no position on the question of whether such hypothetical situations might actually arise down the road, in which case the Activision-Blizzard COW might be invalidated under the statute. But any such claim would have to wait for an actual disputed business opportunity.

It is unclear how the court might have wrestled with the Corti COW had the case presented an actual contested business opportunity. While the text of section 122(17) could arguably favor plaintiff’s preferred narrow construction, the legislative synopsis suggests that plaintiff’s posi­tion faces an uphill battle. The synopsis offers several characteriza­tions of how corporations’ power to renounce business opportunities in advance might be used:

[C]ategories of business opportunities may be specified by any manner of defining or delineating business opportunities or the corporation’s or any other party’s entitlement thereto or interest therein, including, without limitation, by line or type of business, identity of the originator of the business opportunity, identity of the party or parties to or having an interest in the business opportunity, identity of the recipient of the business opportunity, periods of time or geographical location. 146 Del. Gen. Assembly, supra note 93; cf. 72 Del. Laws 619 (2000) (referring to “specified business opportunities or specified classes or categories of business oppor­tunities”).

After Corti, there appear to be no other opinions endeavoring to interpret section 122(17). Although parties in a few post-Corti cases have advanced theories touching on the applicability or scope of a purported waiver, none of these opinions has discussed the section explicitly, and each court either struck down the waiver argument on other grounds or avoided it so as to shed little additional light on how courts will likely apply section 122(17) in future Delaware cases. 147 One notable 2012 case that takes up the enforceability of a COW most directly is Dweck v. Nasser, No. 1353-VCL, 2012 WL 161590 (Del. Ch. Jan. 18, 2012). In Dweck, the court found that the CEO and a minority shareholder of a Delaware corporation had appropriated a corporate opportunity in forming several competing children’s retail businesses. See id. at *1. The defendant asserted that her conduct was permitted under a waiver, but Vice Chancellor Laster disagreed, noting that the plaintiff and controlling shareholder had never executed the purported waiver. See id. at *15. And, while the defendant and the plaintiff had evidently executed a COW in the governance documents of a separate company they had formed, Vice Chancellor Laster held that such a provision was not binding on the fiduciaries of the instant corporation. See id. at *15–17. The court never reached the issue of judicial construction of the COW against the statutory language in section 122(17).

The paucity of published jurisprudence around COWs does not imply, however, that all important legal questions pertaining to waivers have been answered—far from it. One particularly intriguing question pertains to the so-called “duty of good faith,” which entered the scene years after section 122(17)’s promulgation. 148 See In re Walt Disney Co. Derivative Litig., 906 A.2d 27, 66 (Del. 2006) (announcing the duty of good faith). By all accounts, good faith is thought to be immutable 149 Stone ex rel. AmSouth Bancorporation v. Ritter, 911 A.2d 362, 367 (Del. 2006) (noting a “section 102(b)(7) provision . . . can exculpate directors from monetary liability for a breach of the duty of care, but not for conduct that is not in good faith or a breach of the duty of loyalty”). and a subspecies of the duty of loyalty (rather than a freestanding fiduciary obligation). 150 Id. at 370 (“[T]he requirement to act in good faith ‘is a subsidiary element[,]’ i.e., a condition, ‘of the fundamental duty of loyalty.’” (quoting Guttman v. Huang, 823 A.2d 492, 506 n.34 (Del. Ch. 2003))). Moreover, under Delaware jurisprudence, the good faith duty proscribes conduct that harms the corporation and is motivated by a fiduciary’s “subjective bad faith.” 151 Disney, 906 A.2d at 66–67. Another focus of the good faith duty is deliberate inaction by a corporate fiduciary in the face of a known duty to act. Id. It seems entirely possible to characterize corporate opportunity diversion as embodying precisely this type of deliberate bad faith, in which the corporation is “harmed” when the fiduciary willfully deprives it of a lucrative business opportunity. How might such an allegation proceed if the corporation had also executed a COW that covered the fiduciary? On the one hand, the waiver’s power derives directly from the statute; but on the other, the statute never specifically carves out good faith claims. While one could clearly make sound arguments in either direction, a definitive answer must await clarification by the Delaware courts—or legislature—as to whether section 122(17) also unwittingly contractualized part of the duty of good faith.

  1. Non-Delaware COWs. — Delaware was the clear pioneer in authorizing COWs through statute. However, not long after the Delaware reforms took root, other states began to follow suit. Table 1 offers an overview of the states that have followed Delaware to date in amending their statutory frameworks explicitly to allow COWs. Note from the Table that the other states promulgating waiver statutes represent more of a trickle than they do a flood. Promulgating states moved in relatively evenly spaced intervals, with no two states adopting waiver statutes in the same calendar year. It is also worth noting that outside of Delaware, only a few adopting states (Maryland and Nevada) are considered “bellwether” states for incorporation. 152 See Guhan Subramanian, The Influence of Antitakeover Statutes on Incorporation Choice: Evidence on the “Race” Debate and Antitakeover Overreaching, 150 U. Pa. L. Rev. 1795, 1856 (2002) (discussing Maryland and Nevada’s leading roles in the market for out-of-state incorporation). Notably absent from the list are New York and California—two of the largest jurisdictional homes to incorporated entities outside Delaware. 153 See id. at 1815 fig.2 (showing New York and California are two of the largest states for incorporations).

Table 1: Statutory Authority for Corporate Opportunity Waivers by State 154 For citations to these statutory provisions, see infra Appendix at Table A.1.


The structures of the follow-on statutes are similar—but not always identical—to Delaware’s. Oklahoma, Kansas, New Jersey, and Maryland each have provisions that closely track section 122(17) in all respects. 155 Compare Del. Code Ann. tit. 8, § 122(17) (2017), with Kan. Stat. Ann. § 17-6102(17) (2014), Md. Code Ann., Corps. & Ass’ns § 2-103(15) (LexisNexis 2014), N.J. Stat. Ann. § 14A:3-1(q) (West 2016), and Okla. Stat. tit. 18, § 1016(17) (2001). However, there are also variations providing more or less latitude. For exam­ple, Maryland and Washington omit the express modifier “speci­fied” in describing the scope of waivable classes or categories of business opportunities. 156 See Md. Code Ann., Corps. & Ass’ns § 2-103(15); Wash. Rev. Code § 23B.02.020
(5)(k) (2015).
In addition, Missouri and Washington extend the scope of permissible waivers to cover others beyond officers, directors, and dominant shareholders. 157 See Mo. Rev. Stat. § 351.385(16) (2016); Wash. Rev. Code § 23B.02.020(5)(k). In Missouri, in fact, a COW may also cover any agent or employee, 158 See Mo. Rev. Stat. § 351.385(16). and Washington allows COWs to cover “any other person” beyond the usual suspects. 159 See Wash. Rev. Code § 23B.02.020(5)(k). One must not make too much of these seemingly more expansive provisions, particularly in light of the fact that nonmanagerial employees, independent contractors, and other nonfiduciaries generally owe weaker (if any) fiduciary duties to the corporation as principal. See Mattel, Inc. v. MGA Entm’t, Inc., 782 F. Supp. 2d 911, 991 (C.D. Cal. 2011) (finding none of several employment agree­ments imposed fiduciary obligations because “[a]n employer’s trust and confidence in its employee must precede the acceptance of a fiduciary obligation”). Moreover, such employees may be protected by other statutory mandates that help ensure their rights to compete with the principal. See, e.g., Cal. Bus. & Prof. Code § 16600 (West 2008) (prohibiting noncompete restrictions as a matter of public policy for California employees, regardless of state of incorporation). While a version of the corporate opportunities doctrine certainly applies to such actors, for them it is a far more forgiving legal proscription. At the same time, Washington’s statute is also more restrictive in at least two ways. First, the waiver must be part of a charter provision and cannot be adopted by action of the board. 160 See Wash. Rev. Code § 23B.02.020(5)(k) (listing COWs among provisions that “[t]he articles of incorporation may contain”). Second, insofar as a COW reaches officers and dominant shareholders, the statute requires specific board approval (effectively a “reaffirmation”) of the waiver when a business opportunity arises that the officer or dominant shareholder wishes to pursue. 161 See id.

Given the paucity of developed case law even in Delaware, the leadership position that Delaware generally has in establishing precedents, and the fact that the other promulgating states moved later to introduce their amendments, it would be reasonable to expect the case law to be even less developed outside Delaware. Our investigation reaffirms this conjecture: Beyond Corti, we were unable to find any reported cases interpreting the statutory provisions described in Table 1.

The lack of attention COWs have received in judicial opinions and commentary might lead one to believe that the authorizing statutes have simply not created much interest among eligible firms and that corpo­rations have by and large declined the invitation to waive or truncate the application of the COD as to their fiduciaries. As Part III will show, however, this is hardly the case. In fact, corporations have demonstrated an enormous appetite for contracting out of the COD. Before Part III explores these findings, Part II lays down an analytical framework to explain why this appetite may exist.

II. Theory: Efficient Contracting over Corporate Opportunities

The acknowledged unpredictability of the corporate opportunities doc­trine has long generated interest among commentators in developing a coherent account of the doctrine that could serve as a normative lodestar for courts. 162 Harvey Gelb, The Corporate Opportunity Doctrine—Recent Cases and the Elusive Goal of Clarity, 31 U. Rich. L. Rev. 371, 372 (1997) (characterizing courts’ “yearning . . . for clarity in the corporate opportunity area” as “evident and under­standable”); see also Victor Brudney & Robert Charles Clark, A New Look at Corporate Opportunities, 94 Harv. L. Rev. 998, 998 (1981) (attempting to provide “relatively clear rules” to guide courts in their application of the COD). Among these accounts are contractarian approaches that seek to fashion a COD based on precepts of efficient contract design between a corporation and its fiduciaries—that is, how would they plausibly bargain over their prospective rights to corporate opportunities ex ante, given their capital constraints, information constraints, and relative bargaining power? 163 See, e.g., Talley, supra note 72, at 310–36. While not unique among conceptual approaches, the contractarian account can be of considerable value to efficiency-minded courts attempting to adjudicate the thorny equities around disputed claims of ownership of new business prospects.

A significant limitation of the contractarian project as applied to the COD, however, was that the law traditionally prohibited parties from using contracts to alter the relevant fiduciary obligations. 164 See supra section I.B.1 (describing early difficulties associated with COWs); see also supra note 8 and accompanying text (noting commentators have described the duty of loyalty as “immutable”). Thus, if the contractarian account was to have any sway, it would be through the immutable precepts of the doctrine as announced and understood by courts. Since Delaware’s 2000 reforms, of course, jurisdictions have begun to permit parties to contractualize corporate opportunities. 165 See supra sections I.B.2–.4 (discussing several states’ legislative reforms that enable corporations to opt out of the COD). With this change, contractarianism acquired far greater significance, both as a normative guidepost for doctrine and as a vehicle for predicting—as a positive matter—the incidence of COWs as well as their qualitative characteristics. Section II.A briefly outlines a contractarian account, adopting and extending a theoretical framework originally developed in prior work. 166 For a more fulsome exposition, see generally Talley, supra note 72. Recall that Delaware amended its statute in 2000. See supra notes 120–126 and accompanying text. Section II.B suggests why it may frequently be efficient for corporations to allocate corporate opportunities ex ante, rather than merely at the time of board presentation—ex post—which historically was the sole tool available.

Before proceeding, however, it is worth reiterating a lawyerly disclaimer from the introduction: Conceptual arguments for how COWs could be efficient should not be understood to evince a conviction that they are efficient. That conclusion simply does not follow. Indeed, the agency cost paradigm remains a powerful frame for corporate law, 167 See, e.g., Ronald J. Gilson & Jeffrey N. Gordon, The Agency Costs of Agency Capitalism: Activist Investors and the Reevaluation of Governance Rights, 113 Colum. L. Rev. 863, 869–71 (2013) (describing the role of agency costs in corporate governance). The canonical work on agency costs is Michael C. Jensen & William H. Meckling, Theory of the Firm: Managerial Behavior, Agency Costs and Ownership Structure, 3 J. Fin. Econ. 305 (1976). and it is entirely possible that the main reason for the adoption of COWs is managerial opportunism. Indeed, some empirical work on LLCs suggests that entities that contract around fiduciary defaults often do so at their own peril and with wealth-destructive consequences. 168 See, e.g., Mohsen Manesh, Contractual Freedom Under Delaware Alternative Entity Law: Evidence from Publicly Traded LPs and LLCs, 37 J. Corp. L. 555, 574–78, 583–89 (2012) (providing evidence of widespread waiver and exculpation of fiduciary duties in LPs and LLCs and analyzing the impact of these waivers and exculpations); Peter Molk, How Do LLC Owners Contract Around Default Statutory Protections?, 42 J. Corp. L. (forthcoming 2017) (manuscript at 52–54), (on file with the Columbia Law Review) (finding “very little evidence that supports of the efficiency bargaining envisioned by LLC contractarian enthusiasts” and characterizing the results as “worrisome” for “vulnerable LLC owners”). Instead, this Part’s aim is far more modest: It aspires to show that there exist plausible cases in which waivers can augment shareholder value. Ultimately, the normative desirability of COWs cannot be settled on theoretical grounds and requires empirical investigation (to which Part III turns).

A. Framework

To begin, assume a simple framework involving only two parties: (1) a single principal (P, or the “firm”) and (2) a single agent (A), who ostensibly works on P’s behalf. 169 Section II.B discusses other scenarios. In the corporate opportunities context, P will generally refer to the focal corporate entity, while A might represent an officer, director, or dominant shareholder of P. One of A’s key economic functions (and a source of value creation) comes from evaluating new business “projects” that are presented to the firm and taking on those projects that are sufficiently well suited to the firm’s capabilities (generating a net profit in the process).

Assume (for simplicity) that each new project yields revenues of $100 (in expectation). Different projects, however, require different mixes of skills and specialties. Accordingly, the net revenues that P can capture from each project vary heterogeneously across projects. Some projects, for example, may be directly in the firm’s “sweet spot” and maximally profitable, while others are so far afield from the firm’s area of specialty that they are wholly unprofitable. We use the Greek letter theta (q) to represent the specific “type” of skill or specialty requirements entailed in a project presented to the firm. To fix ideas, suppose (arbitrarily) that the value of q describes the percentage composition of verbal (relative to technical) skills that the project requires. When q = 0%, the project is entirely technical in nature, while when q = 100%, it is entirely verbal. Projects with intermediate values of q entail a proportional mixture of technical and verbal requirements. Suppose that moreover ex ante, the next project’s requirements are probabilistically distributed evenly (or “uniformly”) between 0% and 100%.

Project heterogeneity is crucial to how corporate opportunities should be allocated, because the project’s net profitability turns on how closely aligned it is with P’s (or A’s) skill sets. Suppose (for concrete­ness) 170 Although we hypothesize specific numbers and linear functional forms for purposes of discussion, the framework easily generalizes to other formats. See Talley, supra note 72, at 310–35, 344–49 (applying and extending a similar framework). that P specializes in completing projects that are 50% verbal in nature (i.e., projects of type q = 50%) and can complete those projects at a relatively low cost, assumed (arbitrarily) to be $20, for expected net revenues of $80. Should the offered project differ from P’s area of expertise, P could still conceivably accept it, but only by bearing an adaptation cost away from its sweet spot to do so. Namely, it costs P an additional $4 for each percentage increment it moves away from its specialty. For instance, to take on a project with a 40% concentration of verbal tasks (q = 40%), P would need to bear a cost of $60 (i.e., $20 in fixed costs plus an additional $40 representing the cost of moving 10 percentage points from its specialty at $4 per point). In mathematical terms, P’s cost of taking on a given type of project can be summarized by the cost function CP(q) = 20 + 4 · |q − 50|. Figure 1 contains a diagram representing the firm’s cost and profit structure given these parameters. This cost structure (pictured in green) is intended to capture the intuition behind a firm’s “line of business”: P operates most profitably when the project coincides with its sweet spot of expertise (i.e., where q = 50%), but P can adapt its production techniques to take on projects that are further afield. This adaptation, however, comes at a cost, and this cost increases with the “distance” between the firm’s specialty and the requirements of the project. Given the numerical figures posited above, the firm would never take on any negative-value project where q < 30% or q > 70%. In an economic sense, then, the firm’s effective “line of business” is that set of projects the firm would have an economic incentive to pursue. In Figure 1, this would correspond to any project whose requirements fell on the region defined by 30% ≤ q ≤ 70% (a region shaded in green).

Figure 1: Representation of Principal’s Cost Structure and Line of Business

For present purposes, A’s key role at the firm is to attract and receive information about new potential business projects that arrive and that may be of interest to the firm. If merely identifying new prospects was all A could do, then there would be no need for the corporate opportunities doctrine. The problem—in both real life and our model—is that A’s ability to spot and attract new business prospects will often coincide with a private ability to take on projects herself, outside of the productive infrastructure of P. Such a capacity for A is easy to imagine, such as when A is a controlling corporate shareholder of P or the principal of a financial firm that also invests in competitors.

Thus, A may have her own interest in pursuing the project on the outside (either individually or through a firm in which she has an interest). In particular, suppose A faces a fixed cost of x to take on any new project (a cost figure that need not correspond with P’s fixed cost of $20). Suppose further that A’s sweet spot within her area of specialty is denoted by z (which also need not correspond to the firm’s sweet spot of 50%). Finally, suppose that A’s marginal cost of adaptation is b dollars for each percentage point difference between the project and A’s own specialty (where the value of b similarly need not correspond to the firm’s marginal cost of adaptation of $4 per percentage point). All told, then, A’s cost of taking on a project of type q can be summarized by the cost function CA(q) = x + b · |q − z|.

Standard economic intuition suggests that, so long as the charac­teristics of the offered project (reflected in q) are observable by both sides or readily verifiable by a third-party adjudicator, then an optimal contract should allocate control over the project to the lowest cost producer as between P and A. 171 See, e.g., Talley, supra note 72, at 322–26. Indeed, this division uniquely maximizes the total joint surplus that is available to the principal and agent collectively. Under such a division of authority, the parties can use an assortment of side payments (such as wages, licensing fees, or transfer pricing) to divide the contingent revenues from the undertaken project in any fashion they wish. 172 Prior work has already demonstrated this proposition formally. See Talley, supra note 72, at 357–58 (presenting “Proposition 1”). Talley also analyzes the optimal contracting problem when the characteristics of the project (q) are observable only to the agent. See id. at 326–36. In such a circumstance, the optimal contract effectively gives the agent a “call option” to appropriate the corporate opportunity individually in return for paying an exercise price whose value is set by the various cost configurations of the principal and the agent. See id. Because we see little evidence of such contractual structures in our data, we suppress that analysis here and concentrate on the complete information case, which appears to have ready instantiations in observed data.

Moreover, and central to present purposes, it is possible to use a COW—if properly crafted—to achieve the optimal allocation of authority over prospective projects. Within this Article’s framework, the particular form of the optimal COW depends on the relative configurations of P’s and A’s adaptation costs. Figure 2 illustrates four archetypal confi­gurations, each of which gives rise to a different type of optimal waiver; the scope of the optimal COW is illustrated with the green bands overlaying P’s line of business (also in green). Each configuration assumes P has the same specialization and adaptation costs as stated above, while A’s cost structure varies (with each configuration containing different values of x, z, and b). Figure 2a assumes that A’s specialty is at x = 40%, that she faces a fixed cost of z = $80 to take on the project, and that her marginal cost of adaption is the same as P’s, at b = $4. Because of A’s high fixed costs, in this configuration it turns out that there is no project that A would ever be more efficient at taking on than P, and thus it would never be optimal for P to waive any opportunities falling within its line of business. Figure 2b maintains A’s specialty at x = 40% but assumes a far lower fixed cost (of z = $20) and a higher marginal cost of adaptation (of $15 per 1% adaptation). Here, the optimal COW will tend to permit A to appropriate a specific class of projects clustered near her specialty, but it will reserve for P rights as to all other projects. Such a provision would waive “only as to” a specified class or margin of projects closely matching A’s sweet spot. 173 In the figure, this class corresponds to projects falling on the interval 36.3% ≤ q ≤ 42.1%. Figure 2c continues to maintain A’s specialty at x = 40% but assumes a fixed cost of $70 and a marginal adaptation cost of $0.60. For this configuration, A’s fixed cost is sufficiently high that P remains the most efficient producer only near its own sweet spot; however, A’s low adaptation cost gives A a comparative advantage for projects that are near the periphery of P’s line of business. This configuration would be consistent with a waiving “all but” a specified margin of projects near P’s specialty. 174 In the figure, the waiver would allow the agent to pursue projects falling on the disjoint intervals q ≤ 37.1% and q ≥ 66.5%. Finally, Figure 2d assumes an agent specialty very close to P’s (of x = 48%), a lower fixed cost (of z = $10), and the same marginal adaptation cost. For this configuration, A is always the lowest cost producer, and an optimal COW would effectively give her free rein to pursue any and all projects.

Figure 2: Optimal COWs (Green Bands) for Various Principal–Agent Cost Configurations

Figure 2a: Optimal to Retain Status Quo (No COW)


Figure 2b: Optimal COW Waives “Only as to” Specified Classes of Projects


Figure 2c: Optimal COW Waives “All but” Specified Classes of Projects


Figure 2d: Optimal COW Waives “All” COs “to the Fullest Extent Allowed” by Law


The foregoing qualitative analysis delivers several insights about the situations in which one would be likely to see different waiver configurations. First are those instances—presumably very common—in which an agent would never be more efficient at pursuing a project than her corporate principal, as depicted in Figure 2a. Absent significant agency costs, such corporations will compose a universe that falls almost entirely outside our data set, simply because they will generally (and efficiently) eschew waivers. Our coding system would thus not detect corporations that mention the renunciation of business opportunities in order to specifically declare that they are not doing so. At any rate, because nonwaiver of corporate opportunities is the legal default, we expect to observe few such disclosures. 175 On the other hand, if the agency problem is sufficiently severe that the agent can induce the firm to cede new prospects to her (inefficiently), that cost should be observable in reductions to company value. We test this proposition empirically in Part III.

Second, in situations in which the agent has a lower fixed cost than her principal, the optimal allocation of new opportunities will enable the agent to pursue specific types of opportunities closely related to the agent’s specialty, as depicted in Figure 2b. For example, imagine a director who also manages a specialty enterprise on the side. For specific projects related closely to that side enterprise, the director may be the lowest cost producer of a new project, and an efficiency-enhancing waiver would permit the director to pursue such projects.

Third, there are those configurations in which we might expect the fiduciary agent to more easily adapt to pursue projects outside of her specific specialty than the principal, as depicted in Figure 2c. If the agent is more versatile than the principal, the most efficient allocation of corporate opportunities will retain a core set for the principal centered around its specialization but will allocate the remaining opportunities to the agent—an “all but” scope for the waiver.

Two general business scenarios are plausible candidates for includ­ing agents more versatile than their principals. The first are financial fiduciaries. Venture capitalists, private equity firms, and hedge funds are all capital investors that are well known for making major minority or controlling investments in companies as well as for placing their own principals and employees onto the boards of the companies in which they invest. 176 See Christopher K. Aidun & Ernest Ceberio, Current Trends in Venture Capital Fundraising: 2002, Cyberspace Law. (Glasser Legalworks, Little Falls, N.J.), Apr. 2002, at 5–6 (describing the issues facing directors who are also venture-capital investors); Chris E. Abbinante & Jessica B. Fairchild, Obtaining Advance Waivers in PE Transactions, Law360 (Mar. 18, 2010, 11:21 am), [] (discussing the issues facing private-equity investors who sit on multiple boards); Austin & Gottlieb, supra note 13 (“In the private equity or financial investor context, funds that make multiple investments in the same or similar industries may want to avoid any undue restrictions imposed by the duty of loyalty . . . .”); Stephan H. Coonrod & Annamarie C. Larson, Washington’s New Provisions on Advance Waivers of Corporate Opportunities: Opening the Road for Investors, K&L Gates (Apr. 28, 2015), [] (“Venture capital and private equity firms commonly finance multiple investments in the same area of activity and require a seat on the board of directors as a condition to their investment.”). Both the directors these firms place on a company’s board as well as the firms themselves (if their ownership interest is significant enough) can become the sponsored firm’s fiduciaries. 177 See Aidun & Ceberio, supra note 176, at 5 (“A director of an emerging company who is also a director, officer[,] or employee of a venture investor . . . may be faced with conflicting interests and loyalties in carrying out his or her fiduciary duties to each corporation and its stockholders.”); see also Kahn v. Lynch Commc’n Sys., Inc., 638 A.2d 1110, 1113–14 (Del. 1994) (describing the circumstances under which a controlling shareholder owes fiduciary duties). As a result, these fiduciaries would be obliged under the COD’s typical strictures to present any potentially relevant new business prospects to the sponsored entity. Yet, because of the range of other companies in which these financial firms often invest—and in which they might be interested in investing—the financial fiduciary (or her employer) will often be more versatile than the principal. Second, partial corporate spin-offs may create situations in which parent companies remain major shareholders in a subsidiary—and thus owe it fiduciary duties—but in which the parent company remains the lower-cost adapter to newly arising business opportunities. 178 Austin & Gottlieb, supra note 13 (“In the case of a spin-off or carve-out, the parent company . . . may want to preserve its flexibility to pursue potential business opportunities that might also be of interest to the subsidiary without running afoul of its fiduciary duties.”).

Finally, and most strangely, is a configuration in which all new business opportunities are waived, as depicted in Figure 2d. This scenario is slightly bizarre because if the agent is the lowest cost pursuer of all new business prospects, it raises the question of why the principal is even in business. If the agent is the more efficient party to pursue all future business, then perhaps the agent should simply buy out the principal’s business. One context in which it might still make sense for a principal to exist as a going concern but waive all business prospects is when the principal is close to capacity on current projects but lacks the ability to scale up. The current COD already treats such limitations of the principal as sufficient to defeat a cause of action, but a broad waiver could serve to clarify and ensure these limitations as dispositive for a factfinder.

B. Ex Ante Versus Ex Post Waivers

The traditional corporate opportunities doctrine has always per­mitted a corporation’s board to “reject” a business opportunity presented to it by a corporate fiduciary, leaving the fiduciary free to pursue it herself. 179 See supra text accompanying note 76. So a reader may reasonably ask how precisely the ex ante waivers authorized by section 122(17) and its ilk differ from what could be called the ex post waivers already countenanced at common law. After all, the ex post allocation of corporate opportunities to fiduciaries was permissible before section 122(17) and remains so afterward. 180 See supra text accompanying note 76. The key feature of the traditional doctrine was that all corporate opportunities were initially allocated to the corporation. 181 See supra text accompanying notes 75–76. That allocation was mandatory in advance, but it became a waivable default when a fiduciary actually presented a given opportunity to the board. 182 See supra text accompanying notes 75–76.

Advance waivers will have an important efficiency role to play precisely when it is less costly to allocate ex ante what could in principle be bargained over ex post. One might think that the common law’s preclusion of ex ante waivers was sensible, simply because parties will always prefer paying fiduciaries with money and retaining opportunities (renegotiating as needed later on). Corporate entities might seem uniformly to be the more diversified party and better suited to bear the risks of new business prospects. The modern corporate context enor­mously complicates this picture, however. If the corporate fiduciary is a larger, controlling shareholder or a director who belongs to (and whose duties implicate) a private equity firm, then the fiduciary may well be more highly diversified than the company adopting a COW (especially if that company faces a high cost of insolvency). If opportunities are more efficiently allocated to fiduciaries, and both parties know this, then any additional transaction cost ex post may justify an appropriate ex ante allocation, formerly forbidden by law. None of this is to say that a default allocation of all business opportunities to a corporation’s fiduciaries is the efficient one—only that freeing corporations to replace the default allocation may sometimes produce beneficial results.

Indeed, as the remainder of this section shows, waivers under section 122(17) differ importantly in several respects, which together can make ex ante waivers both less costly to adopt and which could result in different—and more efficient—allocations of corporate opportunities than the allocation resulting from ex post waivers.

1. Transaction Costs. — To a large extent, the distinction between ex ante and ex post waivers hinges on evaluating the frictions of transaction costs. In an idealized Coasean world in which transaction costs are zero, 183 See generally R.H. Coase, The Problem of Social Cost, 3 J.L. & Econ. 1, 2–15 (1960). the timing of a waiver may not matter much: Whenever it is efficient for a corporation to renounce an opportunity in its fiduciary’s favor, it will do so, whether ex ante or ex post. In such a setting, as with many other Coasean environments, the legal rule becomes irrelevant. 184 Id. at 8 (“[T]he ultimate result (which maximises the value of production) is independent of the legal position if the pricing system is assumed to work without cost.”). Indeed, even if transaction costs were positive but always the same across time and context, then the choice of legal default would not have great significance. In the actual circumstances of corporate life, however, there are several reasons to expect the presence of transaction costs to sometimes make bargaining between the board and corporate fiduciaries more problematic when done on an ad hoc basis than through an ex ante waiver.

Most obviously, ex ante waivers allow a corporation to commit credibly to ceding certain corporate opportunities (along with other contractual quid pro quos) at the moment of contracting. 185 There is a vast literature on credible commitments and their important role in commerce. See, e.g., Ronald J. Gilson & Alan Schwartz, Corporate Control and Credible Commitment, 43 Int’l Rev. L. & Econ. 119 (2015). For instance, if the fiduciary is a controlling shareholder, it may be providing financing to a corporation years before it expects the corporation to encounter the valuable new business prospects it desires the corporation to renounce. This would be difficult to do exclusively with ex post waivers as the board could simply refuse down the road to permit a fiduciary to pursue a promised opportunity, rendering its prior promise valueless. This could matter significantly if what a given fiduciary—say, a controlling share­holder—sought from an investment was in important part new business opportunities it would generate. Further, without an ex ante waiver, the controlling shareholder would have a strong incentive to retain a 100% controlling position in order to ensure opportunities were allocated to it. With an advance waiver, however, a fiduciary can freely diversify when it is efficient for it to do so.

Advance waivers can thus significantly enhance the bargaining space—in the form of new business prospects—that a corporation has available to transfer in exchange for cheaper financing (or the service of an outstanding officer, the expert input of a venture capitalist, or the rolodex of an outside director). 186 It is worth noting that while a statutory reform like section 122(17) may allow for cheaper financing, or easier recruitment of managerial talent, it is also likely to affect how surplus is allocated between corporations adopting waivers and their counterparties, whether that counterparty be a private equity or venture capital firm or a potential new director. This is a central function of contract: to ensure that parties can credibly commit to follow through on their commercial undertakings, enabling transactions whose performances are separated in time. 187 See, e.g., Oliver E. Williamson, Credible Commitments: Using Hostages to Support Exchange, 73 Am. Econ. Rev. 519 (1983). Without the possibility of ex ante commitments, financing that a fiduciary would have provided in exchange for promised opportunities may not be exchanged, resulting in foregone gains from trade.

Another limitation on ex post waivers is that the context of fidu­ciaries’ involvement in a corporation is a classic site for the fiduciary’s development of highly specialized skills and abilities. A director or officer typically invests years of work in a company and will usually acquire knowledge and abilities that are valuable at the company but are of less or no use elsewhere. A well-known hold-up problem can thus emerge. For example, the corporation may know that a fiduciary’s service to it is worth, say, $50,000, but only $30,000 to any other company. One can further assume—plausibly, in the context of at least some early-stage companies 188 Cf. Simone M. Sepe, Intruders in the Boardroom: The Case of Constituency Directors, 91 Wash. U. L. Rev. 309, 328–29 (2013) (analyzing the contractual and financing consequences of various limitations faced by start-ups). —that corporate opportunities are one of the most valuable forms of compensation a corporation has to offer. The framework developed in section II.A likewise focused on situations in which the fiduciary could, in terms of comparative advantage, simply be the more efficient party to pursue a new business opportunity. In the absence of enforceable advance waivers, the corporation may promise to provide the fiduciary with a variety of opportunities as compensation but then fail to renounce them as anticipated. 189 See, e.g., Benjamin Klein et al., Vertical Integration, Appropriable Rents, and the Competitive Contracting Process, 21 J.L. & Econ. 297, 302–03 (1978) (discussing contractual enforcement as a potentially efficient response to parties opportunistically reneging on promises). If the fiduciary expects this, then she will have significantly dampened incentives to perform the specialized tasks and invest in firm-specific knowledge vis-á-vis what is optimal. As a result, both the corporation and its fiduciaries stand to benefit from an enforceable contract for allocating corporate opportunities among them in advance. Nonetheless, it is worth noting that the hold-up problem can obviously run both ways—fiduciaries may hold up the corporation, just as the corporation may hold up its fiduciaries. There seems to be no a priori reason to view one party to be more opportunistic than the other, 190 There are some instances in which opportunism by the corporation is significantly more likely. Consider a joint venture (JV) entity formed by two major rival corporations for a narrow and specific collaborative purpose. Typically, both of those rivals will owe the JV fiduciary duties. See Sarath Sanga, The Contract Frontier: A Study of the Modern Joint Venture 10–16 (Jan. 4, 2015) (unpublished manuscript),
sanga_modern_jv.pdf []. In this circumstance, there may be far more scope for the entity to be weaponized by one corporation and used to hold up its rival (via fiduciary duty claims) than for the rivals to hold up the JV.
but the prospect of subsequent hold-up by the firm can be a significant driver for an ex ante waiver.

2. Board Dynamics. — Advance commitments can carry a host of other benefits as well. From a psychological vantage point, the members of a board may find it easier to commit to relinquish prospective business opportunities upfront, in say a charter or contract provision, than in a later ad hoc process when the opportunity is directly presented and the loss of prospective profits is palpable.

Depending on when the adoption of an ex ante waiver occurs, it can also allow a board to avoid a conflicted vote, and potentially, to be insulated from any possible breach of duty of loyalty allegation. 191 See supra section I.A (describing the constraints imposed by the duty of loyalty); supra section I.B.2 (explaining the legal effect of COWs). For instance, if a COW is adopted in a firm’s IPO charter, then a prospective corporate fiduciary can bargain with the corporation at arm’s length and the pre-IPO board can hold a vote on whether to include the COW without any member needing to recuse herself.

3. Liquidity Constraints. — Another potential driver of advance waiver activity is the influence of liquidity constraints, either at the firm or the fiduciary level. In conventional economic frameworks, agents (the fiduciaries) are wealth and capital constrained and thus unable to “prepay” for the nonmonetary contractual rights that they value most highly. 192 See, e.g., Jensen & Meckling, supra note 167, at 321 (assuming the manager has limited “personal wealth” and thus can provide financing only to a certain level). When such rights consist of the allocation of corporate oppor­tunities, the parties may be unable to reach an agreement allocating those projects to the fiduciary even when the fiduciary is the most efficient producer. In this case, of course, the default rule (reserving rights for the corporation) may well be optimal, and thus there would be little need for an ex ante waiver. 193 This point is developed more fully in Talley, supra note 72, at 357–60.

In other instances, however—particularly involving startups—the corporation is the party that is cash poor and capital constrained, hindering its endeavors to attract managers, directors, and potential investors. 194 See Sepe, supra note 188, at 328–29. Although the entrepreneur could offer to remunerate these actors with equity, doing so has a dilutive effect on founders, common shareholders, and employees—an effect that can erode internal incentives to maximize value. 195 See, e.g., Jason M. Gordon & David Orozco, Trust and Control: The Value Effect of Venture Capital Term Sheet Provisions as Risk Allocation Tools, 4 Mich. Bus. & Entrepreneurial L. Rev. 195, 196 (2015) (discussing the dilutive effects of new equity issuances on founders and other equity holders). One possible alternative in such situations is to offer in-kind modes of compensation, including the allo­cation of property rights over future business opportunities. Symmetric to the discussion above, in-kind compensation of this sort may efficiently require the corporation to waive rights on certain projects in which it is marginally the more efficient producer, thereby providing the fiduciary a platform from which to profit from the new opportunity itself (or alternatively to resell it to the corporation at a less capital-constrained juncture). In this case, executing a seemingly broad waiver may still create value for shareholders if it is combined with attracting significant managerial talent.

4. Rules Versus Standards. — Ex ante waivers may also offer distinct benefits because they authorize private parties, through contract, to replace the immensely complex and nebulous standard of the common law with a rule crafted by the parties themselves (or at least a more refined standard). 196 See generally Louis Kaplow, Rules Versus Standards: An Economic Analysis, 42 Duke L.J. 557 (1992) (offering a canonical economic exposition of the rule–standard distinction). Even if the parties would ultimately end up with the same allocation of corporate opportunities ex post as ex ante, the ability to substantially clarify that allocation through advance contracting can substantially reduce the scope for litigation costs stemming from either opportunism or honest mistake. 197 Cf. id. at 570 (noting law enforcement is “costly, with the cost being greater if a standard governs because the adjudication will also require giving content to the standard”). To be sure, few contractual provisions define perfectly clear boundaries, and the COWs we observe do not even attempt to do so. Nonetheless, COWs appear to generally replace a highly indefinite standard in one of fiduciary law’s most convoluted areas with significantly better-defined directives. 198 See, e.g., GoDaddy Inc., Amended and Restated Certificate of Incorporation (Form 8-K, exh. 3.1), at 9–11 (Apr. 6, 2015),
data/1609711/000119312515120133/d903539dex31.htm []; HCA Holdings, Inc., Amended and Restated Certificate of Incorporation 10–11 (Mar. 8, 2011),
com/files/doc_library/file/CertificateofIncorporation.pdf [].
This may be able to deliver substantial cost savings, which can be divided by the parties.

In addition, the clarity of advance waivers can help establish clear ground rules that serve as the backbone for a long-term commercial relationship. Consider, for example, the 2007 COW disclosed by NetSuite Inc., related to its significant block shareholder and board member Lawrence Ellison. Perhaps as a condition to his involvement with the company, Ellison received a broad COW from NetSuite, which NetSuite’s board of directors approved and disclosed in an SEC filing. 199 See NetSuite Inc., Annual Report (Form 10-K), at 24 (Mar. 13, 2009), []. After describing the common law standard, NetSuite replaced it with a bright-line rule:

[A]s a majority stockholder, Mr. Ellison might, in certain circumstances, have had a duty to present to the corporation matters that come to him that are within our line of business or would be deemed of interest to us. Under the waiver, we have renounced any such duty, and Mr. Ellison will not need to present any such opportunities to us . . . . 200 NetSuite Inc., Prospectus (Form 424B4), at 109 (Dec. 19, 2007), [] (emphasis added).

Although this waiver is clearly quite broad, it may also have been necessary to bring on Ellison as a significant investor. More importantly, this waiver plausibly helped pave the way for a long-term relationship between Ellison, Oracle, and NetSuite, one that arguably facilitated Oracle’s $9.3 billion acquisition of NetSuite. 201 See Quentin Hardy & Leslie Picker, Oracle’s $9.3 Billion Deal for NetSuite Will Bolster Its Cloud Offerings, N.Y. Times: Dealbook (July 28, 2016), (on file with the Columbia Law Review) (noting the “relationship [between Ellison and NetSuite’s founder] came full circle as Oracle agreed to acquire NetSuite for $9.3 billion” years after Ellison backed NetSuite’s inception); Press Release, Oracle, Oracle Completes Tender Offer Acquisition of NetSuite (Nov. 5, 2016), [] (announcing the completion of this acquisition).

As noted earlier, the discussion above is meant to demonstrate only that there exist several plausible economic contexts where COWs—even broad ones—are value enhancing for both the beneficiary and the waiving corporation. Such contexts, of course, must uneasily co-exist with omnipresent agency cost problems, in which corporate fiduciaries may use their influence at the company to line their own pockets at shareholders’ expense. To the extent that the former efficiency-related factors outweigh agency cost considerations, Delaware’s reform program has been beneficial for capital investors. To the extent the opposite holds, however, the policy landscape flips. Ultimately, this question is indeterminate within the realm of economic theory alone, and an empirical analysis is warranted. It is to that analysis that Part III turns.

III. Empirical Analysis

The historically unprecedented power that many corporations now have to opt out of the COD seems to have been largely overlooked in both case law and law and finance scholarship, generating little systematic analysis of any form. 202 See, e.g., Gilson & Schwartz, supra note 185, at 120 (noting section 122(17) as a rare exception to the mandatory character of the duty of loyalty). This neglect is surprising, and it seems undeserved for several reasons, ranging from the technical to the policy related.

On the more technical side, the duty of loyalty is a long-hallowed “sacred cow” of fiduciary principles, traditionally unyielding to private, contractual end-runs. 203 See supra notes 49–53 and accompanying text. Any systematic shift that concerns when and how corporate entities may embrace such waivers is, at the very least, important to document, given its departure with longstanding legal tradition. Moreover, a cottage industry of corporate governance metrics for incorporated entities has taken root in the last decade. 204 See Sanjai Bhagat, Brian Bolton & Roberta Romano, The Promise and Peril of Corporate Governance Indices, 108 Colum. L. Rev. 1803, 1809–14 (2008) (providing an overview and assessment of some of these early trends). These metrics seek to aggregate and assess when firms opt out of (or otherwise contract around) various corporate law default rules, such as through classified boards, poison pills, duty of care waivers, blank-check stock, and the like. The results can be useful for both scholars and industry groups, such as proxy firms that often address corporate governance votes for institutional clients. Sophisticated producers and consumers of such metrics should be interested in these new mechanisms that permit parties to contract out of the duty of loyalty in order to provide a complete picture of governance arrangements.

More substantively, COWs bear on central “big picture” debates in corporate law. First, how corporations respond to this new power to tailor can importantly inform scholarly debates regarding whether corporations actually use their capacious legal freedoms to tailor default terms. 205 See Henry Hansmann, Corporation and Contract, 8 Am. L. & Econ. Rev. 1, 4 (2006) (“[T]he extraordinary freedom that is now available to the drafters of corporate charters is exploited in remarkably small degree.”). The question lurking in the background—and one of corporate law’s most profound quandaries—is whether we can expect (or trust) corporations, when freed to do so by law, to adopt optimal corporate governance structures on their own accord. 206 See Michael Klausner, Fact and Fiction in Corporate Law and Governance, 65 Stan. L. Rev. 1325, 1327–29 (2013) (characterizing the issue of “whether market imper­fections impede the establishment of optimal governance arrangements” as “an empirical question” that cannot be settled solely on the basis of theory); see also Zohar Goshen & Richard Squire, Principal Costs: A New Theory for Corporate Law and Governance, 117 Colum. L. Rev. 767, 828 (2017) (recommending a presumption in favor of default rules “unless there is a specific market failure”).

Even more generally, the empirical experience of COWs goes straight to the heart of a broader normative question: Are COWs socially beneficial? This question arises at the intersection of two distinct but important strains in corporate theory, for the very transaction-cost arguments that motivated the amendments permitting COWs 207 See supra section II.B.1. are in significant tension with well-known arguments justifying loyalty as a core check on managerial opportunism and agency costs. 208 See supra notes 2–7 and accompanying text. Understanding how COWs have played out in the post-deregulatory environment may shed light on other areas where scholars vigorously disagree as to whether fiduciary duties should be mandatory. 209 Compare supra note 85 and accompanying text (noting views that support mandatory rules), with Roberta Romano, Answering the Wrong Question: The Tenuous Case for Mandatory Corporate Laws, 89 Colum. L. Rev. 1599, 1603–15 (1989) (questioning the utility of mandatory rules).

The scant literature on COWs may be—at least in part—an artifact of the practical difficulty of data collection. Most conventional corporate governance provisions are located in relatively discrete and easy-to-isolate sources (such as the charter or shareholder-approved bylaws). As docu­mented in section I.B.2, however, COWs are generally not required to be executed according to any particular regimen, and thus they can be scattered across myriad corporate documents, including charters, bylaws, contracts, board resolutions, and the like. Consequently, searching for waivers requires combing through many (if not all) disclosures made by public companies to securities regulators.

This Part describes and reports on an empirical analysis of a large sample of COWs disclosed in public corporate filings, which we extracted and coded. Section III.B shows that COWs are common and widespread among public companies within our sample. Moreover, we employ machine-learning techniques to extrapolate our classification procedure outside of our sample, from which we can infer (with significant confidence) that well over one thousand public companies to date have made use of their newfound freedom to contractually allocate corporate opportunities, and they continue to do so at an increasing rate. 210 See infra Figure 3. Waivers appear across industrial sectors, although they are overrep­resented in some of them. 211 See infra notes 258–259 and accompanying text.

In addition, this Part employs a variety of empirical methods to examine the normative question of whether corporations have used the freedom to adopt COWs to create or destroy value. First, section III.C reveals that the companies embracing COWs generally tend not to exhibit indicia of questionable investment value or managerial practices. Rather, using standard measures of firm value, COW adopters appear to have markers of value creation and effective management. 212 See infra Table 6. Second, section III.D reports the results of an event study that measures the value created by adopting a COW, using the stock market’s reaction to their public disclosure as a proxy. Perhaps surprisingly, for those firms for which we could measure returns on stock-price reaction to an announced COW, the market appears to receive such news positively, generating four-day cumulative abnormal returns of between 1% and 1.5%. 213 Infra Table 7. In addition, as a robustness check, we make use of the lagged implementation of reforms across the nine reforming states. We similarly find evidence of a positive market response to passage of a COW reform among issuers incorporated inside the reforming jurisdiction. 214 See infra Figure 6. These findings go some way to allay fears that corporations adopting COWs are systematically those with worse agency costs.

A. Data

To investigate the empirical incidence (and effects) of COW adop­tion, we constructed an original data set based on disclosures by publicly traded corporations. Typically, if a public company adopts a waiver of corporate opportunities, the adoption is likely to be disclosed in some form of document that the corporation will file with the SEC. We began by utilizing a “mirror” of the SEC’s EDGAR database, which replicates over twenty-one million issuer filings available from 1995 through March 2016. 215 See EDGAR: Company Filings, SEC,
companysearch.html [] (last visited Jan. 31, 2017) (noting “access to more than 21 million filings”). 1995 was the first year for which the database had reliably global coverage. See Electronic Filing and the EDGAR System: A Regulatory Overview, SEC (Oct. 3, 2006), [] (last modified Nov. 16, 2006) (noting “full implemen­tation of mandated electronic filing” began after mid-1994); Help – History of EDGAR, EDGAR Online, [] (last visited Jan. 31, 2017).
Because most states’ statutes authorizing COWs permit their execution across a wide spectrum of contracts, resolutions, and corporate governance documents, 216 See supra section I.B.4. we did not constrain our examination of EDGAR filings to any specific filing “form” (such as 10-Ks, 8-Ks, or 14-As). To this mirrored data set we applied a Boolean keyword search 217 The exact search took the form: “[(corporate OR commercial OR business) followed by (opportunit!)] within the same sentence as [(waiv! OR renounc! OR disclaim)]”. The “!” marks here serve as root expanders to pick up any stems, alternative endings, or punctuation marks around the relevant word. —similar to what one might use in a Westlaw or LexisNexis query—designed to identify candidate instances of a disclosed COW based on a filing’s similarity with the search query. 218 For every matching candidate document, we extracted the responsive sentence (or sentences) as well as three preceding sentences and three succeeding sentences around each detected excerpt (or “snippet”). This process resulted in 10,682 distinct candidate disclosures. 219 Gabriel Rauterberg & Eric Talley, Corporate Opportunity Waivers Data Set (on file with the Columbia Law Review) [hereinafter Data Set]. This data set will be available on the Columbia Law Review website,, starting in September 2017. Our keyword search was designed to be quite general, but this generality also assured that it was relatively overinclusive.

From this large candidate data set, we randomly selected one thousand excerpts for manual coding with the help of a small group of trained research assistants. We designed a coding procedure that employed forty-one distinct variables. 220 Table A.2 in the Appendix contains the coding rubric. We engaged in a significant training and cross-validation program with the research assistants who coded excerpts, designing significant “overlap” in coded terms, so as to detect (and correct) inter-coder inconsistencies. Our rubric focused on the three fundamental questions that pertain to any waiver:

  1. How is a company adopting the waiver? The first seven variables inquire as to how a filing company is adopting a COW, whether in its charter, bylaws, board resolution, or some other kind of disclosure.
  2. Who is covered by the waiver (e.g., a single officer or director, all officers and directors, a controlling shareholder, and so forth)? The next thirty variables inquire into who is covered by the COW. This sequence includes both who is covered at the filing entity (fourteen variables), and, if another entity is mentioned, then who is covered within the man­agement or ownership structure of that other entity (sixteen variables).
  3. What is the extent to which a company is waiving the COD as to those covered? The last four variables inquire into whether the scope of the waiver covers all business opportunities, a specified list of such opportunities, all opportunities except for some specified subclass, or a wildcard term waiving corporate opportunities “to the fullest extent allowed by law” (or a substantive semantic equivalent).

As noted above, the keyword search we used to identify candidate COWs was deliberately overinclusive, flagging a series of documents that were sure to include some “false positives”—snippets that satisfied the keyword criterion but were judged by coders not to reflect COWs.

After cleaning the hand-coded sample data, we deployed several measures to enhance the richness and reliability of the data. First, we linked our database to a variety of other publicly available data sets that are widely used in law and finance scholarship (Compustat and CRSP 221 CRSP/Compustat Merged Database, CRSP,
research-products/crspcompustat-merged-database [] (last visited Mar. 6, 2017).
) to gain insight about the financial and governance characteristics of issuers adopting COWs. Second, we used the hand-coded data to train a machine-learning (ML) classifier, adapting a technique previously developed by one of us to classify the remainder of the snippets algorith­mically. 222 See Eric Talley & Drew O’Kane, The Measure of a MAC: A Machine-Learning Protocol for Analyzing Force Majeure Clauses in M&A Agreements, 168 J. Institutional & Theoretical Econ. 181 (2012). This process not only enabled us to look beyond our sample of COW disclosures with significant confidence, 223 The ML classifier proved extremely accurate, correctly categorizing COWs 93.5% of the time within our hand-coded sample. The classifier lost little of its mojo even when extended outside the sample: In Monte Carlo simulations, it correctly categorized simulated out-of-sample COWs at a mean rate of 92.5% (with standard deviation of 1.50%). For details, see Rauterberg & Talley, supra note 16. but it also facilitated an alternative means to audit the hand-coded data for errors or inconsistencies. 224 We inspected all snippets for which the human-coded and machine-classified data disagreed as to the presence of a COW (n = 65). In around 80% of these cases, we concurred with the humans. We recoded the remaining 20% where we agreed with the machine, and then we completely retrained the ML classifier.

B. Descriptive Statistics of Waivers

Within the randomly selected sample of 1,000 candidate SEC filings, our manual coding enterprise yielded 628 responsive documents from 427 unique issuers that contained a bona fide COW disclosure. 225 Data Set, supra note 219. Of these, 237 were operative provisions in which a filing company enacted a COW, and 391 were disclosures discussing such an operative provision elsewhere. 226 Id. (The remaining candidate documents were deemed nonresponsive—reflecting a 38.5% “false positive” rate from our keyword search. 227 Id. ) Extrapolating proportionally to the full population of 10,682 filings, one would predict upwards of 6,700 responsive disclosures across the population of snippets. And indeed, the ML classifier bears this prediction out, detecting 6,859 COW disclosures from 1,592 unique issuers (averaging around 326 disclosures per year over two decades). 228 Id. While the ML classifier tags between ten and eleven times as many COW disclosures as the sample (as expected), it identifies only between three and four times as many unique issuers. Id. This is also unsurprising on reflection: As sample size grows, it becomes increasingly likely that multiple disclosures by single issuers are picked up. These figures alone embody a key finding of this study: Public companies have shown a significant appetite for contracting out of the fiduciary duty of loyalty.

That appetite, however, was slow to arrive. Figure 3 illustrates the yearly counts of COW disclosures in both our hand-coded sample (left axis; blue line) and in the ML-classified population (right axis; red line). Note that the incidence of disclosed COWs was initially quite low—though not zero—both before and immediately after Delaware’s statutory amendment in 2000. 229 See infra Figure 3. The nonzero incidence of disclosed COWs prior to 2000 is not altogether surprising, given that DGCL section 122(17) was itself a legislative rejoinder to a judicial response to a disclosed waiver. 230 See supra section I.B.2 (cataloging various states’ enabling statutes). The initial reticence of issuers to embrace the COW from 2000 to 2003, 231 See infra Figure 3. in contrast, is somewhat more notable, and it attracted the attention of commentators at the time. 232 See Austin & Gottlieb, supra note 13. It is plausible that this initial tepid response reflected the post-dot-com hangover that suppressed U.S. financial markets. 233 To Fly, to Fall, to Fly Again, Economist (July 25, 2015),
news/briefing/21659722-tech-boom-may-get-bumpy-it-will-not-end-repeat-dotcom-crash-fly [] (discussing the negative effects of the bursting of the dot-com bubble on venture capital).
Whatever the reason, by 2004, the herd of COWs began to propagate with impunity, growing roughly exponentially since (with a slight buckle during the financial crisis). 234 See infra Figure 3. By the turn of the decade, COW disclosures had become commonplace—surpassing 1,000 per year by 2014. 235 See infra Figure 3. The apparent decline in 2016 is due to the fact that counts for the complete 2016 year are not available as of this writing. There is significant heterogeneity in the forms of disclosure, reflecting the open-ended waiver procedure authorized by most states’ enabling statutes. 236 See supra section I.B.4. Just under half of the COW disclosures come from IPO-related filings, approximately 10% appear in each of “current report” (8-K) filings and routine annual or quarterly filings (10-K or 10-Q), and around 7% appear in proxy materials. 237 Data Set, supra note 219. These figures are nearly identical in both the hand- and machine-coded data. 238 Id.

Figure 3: Total COW Disclosures by Year (Hand Coded Versus ML Classified)

Another important issue surrounding COWs is where they can be found within firms’ governance documents and other written instru­ments. Table 2 summarizes the location of manually coded waivers across various document types. 239 Tables 2 through 4 concentrate on our hand-coded sample, leaving for another day the calibration of an ML classifier for these nested subsidiary provisions. Note that total counts of provisions tend to exceed the aggregate number of COW disclosures, since the sub­­cat­egories are not mutually exclusive. Percentages are based on the “Operative Provision” categories.

Table 2: Location of COW in Corporate Governance Documents


As Table 2 shows, over three quarters of the COWs in the sample appear to be located in a corporate charter, evincing a degree of commitment to the waiver that is difficult to unwind absent a shareholder vote. That said, nearly 20% of the remainder waivers appear to be spread across a variety of contractual instruments, which are usually—in contrast—susceptible to renegotiation and restructuring by the board alone, without shareholder involvement. 240 Supra Table 2.

As previously noted, the wording of most states’ statutes provides appreciable freedom for the drafters to waive broad or narrow categories of corporate opportunities. 241 See supra sections I.B.2–.4. Given that the default rule in corporate law is no waiver, 242 See supra section I.A.2. a broader express waiver signals a greater departure from traditional fiduciary principles. Table 3 summarizes the scope of COWs in the sample.

Table 3: Scope of Waiver


The breadth of the language in the waivers is notable. About 36% of COW disclosures include at least one waiver that purports to waive “all” corporate opportunities. 243 Supra Table 3. Just under 48% include at least one waiver purporting to waive “all but” a reserved class or category of opportu­nities. 244 Supra Table 3. In our estimation, these clearly represent notably broad waivers. In contrast, only just under 17% of COW disclosures waive the doctrine in a more modest fashion (following the actual wording of the statute), limiting the disclaimer to a specified class or category of opportunities. 245 Supra Table 3.

Another important scope consideration is how the coded sample allocates waivers across different types of corporate actors. Recall that this group can include corporate directors, officers, and dominant share­holders, 246 See supra sections I.B.2.–.4. all of whom owe a loyalty obligation to the corporation. 247 See supra section I.A. Table 4 summarizes the target of the coded provisions by the role of the actor—officer, director, or shareholder. (The aggregate counts listed significantly exceed the total number of detected waivers, since waivers routinely apply to multiple classes of actors and sometimes differentiate between subclasses of actors.)

Table 4: Corporate Fiduciaries Covered by COW

248 This category blends: all officers, any officers, or enumerated officers covered by the COW. 249 This category blends: all directors, any directors, or enumerated directors covered by the COW. 250 This category blends: all shareholders, any shareholders, or enumerated shareholders covered by the COW.

Table 4 shows that corporate directors are the most frequent beneficiaries of COWs, covered by nearly three quarters of COWs in our sample—an observation consistent with the aim of the reform to ameliorate conflicts that come from overlapping directorships among venture capital and private equity investors. 251 See supra section I.B.2 (describing the impetus behind the 2000 Delaware reforms). Perhaps for similar reasons, shareholders are commonly beneficiaries of a waiver as well, covered just under 50% of the time. 252 Supra Table 4. Notably, however, corporate officers are also routinely included in COW disclosures, with a waiver applying to them nearly half of the time, 253 Supra Table 4. a factor that is somewhat striking given the full-time nature of an officer’s duties to the firm. 254 See Clark, supra note 80, § 7.6, at 243 (recommending a categorical prohibition for full-time executives of public companies).

C. Issuer Characteristics of COW Adopters

The exercise of harvesting, coding, and documenting COW disclo­sures is one of the key contributions of this Article. Nevertheless, it is possible to say somewhat more by linking the disclosing issuers to several other databases providing industry and financial data. A particularly interesting industry source is the Compustat database, which contains information related to corporate governance and financial performance as disclosed in issuers’ annual and quarterly financial reports. 255 CRSP, supra note 221. We therefore undertook to link the COW database to Compustat (for the period 1994 through 2016). Although this matching process resulted (as it often does) in some loss of data, we were able to match a significant fraction of COW disclosers with their financial information: We ultimately matched approximately 377 issuers in the hand-coded data sample and 1,765 issuers in the machine-classified population. 256 Data Set, supra note 219. Since reliable financial filings occur on an annual basis, the merged data are best analyzed with an “issuer-year” unit of analysis, so that any issuer making multiple COW announcements in a single year would be collapsed into a single observation for that year. This process resulted in collapsing just over 90 of the COW disclosures in the hand-coded sample and around 3,300 in the ML-classified population. Id.

Using these matched firms, it is possible to consider a host of firm-level indicia. Preliminary analysis suggests the industry representation of disclosed COWs appears to track that of Compustat firms, with a few notable areas of industry overrepresentation. 257 See id. In particular, oil and gas issuers represent 9.71% of COWs as compared with 3.95% of Compustat firms. 258 Id. Similarly, Business Services issuers represent 13.12% of the pilot sample but only 9.04% of the Compustat universe. 259 Id. The overrep­resentation in both industries plausibly reflects the popularity in these industries of ownership structures characterized by multiple investments across portfolio companies. (Additional research and data collection will be needed to test this hypothesis.)

Because corporate law (and the permissibility of COWs) is an artifact of state law, 260 See 1 Fletcher, supra note 67, § 2.50 (noting that corporations “deriv[e] their existence and authority to act from the state”); supra section I.A.2 (discussing fiduciary duties, including the COD, under state law). adopters’ incorporation jurisdiction is of obvious interest. Perhaps not surprisingly, the public firms embracing COWs are heavily overrepresented by Delaware. Indeed, as Table 5 illustrates, in both the hand-coded and machine-classified data sets, about nine out of ten waiver disclosures come from Delaware corporations, far more than the population-wide proportion of U.S. public companies incorporated in Delaware (which is just north of 50%).

Table 5: COWs in Sample, by State of Incorporation

The significant overrepresentation of Delaware may be due to a variety of factors, including the fact that Delaware was the earliest mover, 261 See supra section I.B.2. the significant network externalities among the Delaware bench and bar, 262 See Michael Klausner, Corporations, Corporate Law, and Networks of Contracts, 81 Va. L. Rev. 757, 843–47 (1995) (explaining Delaware’s general success as a product of network externalities). the (possibly) larger comparative size of Delaware-incorp­orated adopters, and the potential remaining invalidity of COWs in many other states.

One benefit of linking the data on disclosed COWs to other financial databases is that doing so sheds considerable light on both what types of companies embrace waivers and what such adoptions portend for company value. Our empirical analysis provides several interesting insights about the relationship of COW adoption and financial measures related to profitability and value creation. Table 6 illustrates some of these relationships.

Table 6: COW Disclosers Versus Compustat Financial Profiles (1994–2016)

Table 6a: All Entities


Table 6b: Delaware Corporations


Several interesting findings emerge from these descriptive statistics. As Table 6a illustrates, COW adopters tend to be smaller on average—not larger—than the mean Compustat issuer on many balance sheet metrics, such as total assets, total liabilities, and long-term debt. Yet the median COW adopter, in contrast, tends to be somewhat larger than the population median, 263 See supra Table 6. suggesting that COWs are embraced by a narrower cross-section of issuers that are relatively large but not at the highest (or lowest) end of the distribution. In addition, disclosing firms appear to exhibit substantially stronger median income-statement metrics, reve­nues, capital expenditures, and earnings, with roughly commensurate means. 264 See supra Table 6. One frequently used mea­suring stick of value creation within finance is the annual return that the issuer makes on its overall assets (ROA) 265 See, e.g., Lucian A. Bebchuk et al., The Long-Term Effects of Hedge Fund Activism, 115 Colum. L. Rev. 1085, 1102 (2015) (noting ROA “has been significantly used by financial economists as a metric for operating performance”); David F. Larcker et al., Boardroom Centrality and Firm Performance, 55 J. Acct. & Econ. 225, 227 (2013) (using ROA to measure firm performance). —a valuation metric that is largely independent of debt–equity structures. Here, COW disclosers within the two samples tend to outperform the comparator Compustat group (along both mean and median dimensions) in generating returns for outside capital investors. Table 6b reports the same set of valuation metrics but with both samples limited to Delaware incorporated firms. Most of the patterns persist here too (with some variations). 266 See supra Table 6.

* * *

The descriptive statistics summarized in this section paint a picture that holds clear relevance to larger policy debates about both COWs specifically and the contractibility of fiduciary duties more generally. In particular, we see no evidence—based on these data—that COWs are systematically embraced by under-achieving firms that use the waiver as a pretext for inefficiently diverting value away from capital investors. To the contrary, these data suggest that waivers are adopted by relatively healthy com­panies with robust cash-flow potential and an established record of delivering attractive returns to their capital investors. This profile is consistent with the shareholder-value-enhancing account of COWs devel­oped in Part II, in which COWs can serve as important complements to building value-enhancing corporate structures by clarifying the bound­aries of what is sometimes considered an incurably opaque and recondite area of corporate fiduciary law.

D. Market Reaction to COW Adoptions

Because the results discussed in the previous section are based on summary observational data, they are ill suited to support causal inferences about the welfare effects of COW adoption. Many alternative stories could explain the same relationships. For example, it may simply be easier for opportunistic managers to adopt COWs in healthy companies because shareholders might scrutinize board action less. That said, our study’s focus on publicly traded companies permits us to use alternative approaches to develop some insight into causal relationships. One such approach is to examine the extent to which COWs contribute to (or detract from) firm value by measuring the reaction of the capital markets. To the extent that stock prices of public companies tend, on average, to reflect publicly available information, 267 This is the most popular articulation of the phenomenon sometimes referred to as the semi-strong form of the efficient capital market hypothesis (ECMH). Although the ECMH has drawn some critics over the years, most economists tend to continue to subscribe to its general precepts. See Steven L. Jones & Jeffry M. Netter, Efficient Capital Markets, in Concise Encyclopedia of Economics 138, 139–41 (David R. Henderson ed., 2d ed. 2008) (outlining the debate over the ECMH and ultimately concluding the ECMH is “still useful” partially because “the response of stock prices to new information reasonably approximates the change in the intrinsic value of equity”). The U.S. Supreme Court has recently reaffirmed its own general acceptance of the ECMH in the context of securities market litigation. See Halliburton Co. v. Erica P. John Fund, Inc., 134 S. Ct. 2398, 2410 (2014). one instructive means for assessing COWs is to consider how market prices react to the first public disclosure an issuer makes about the existence of (or plans for) a waiver. Our data set is amenable to such an inquiry: Indeed, we are able to identify the first disclosure date with all of Compustat-matched firms reported in the previous subsection, which in turn allows us to match these firms to their securities market prices as reported in the CRSP database. We measure market reaction to the disclosure of news using the familiar event-study approach. 268 For an overview of event-study methodology, see generally Charles J. Corrado, Event Studies: A Methodology Review, 51 Acct. & Fin. 207 (2011). An important feature of a securities-market event study is a designated methodology for predicting the “expected return” of the stock. See id. at 209–10. This step is conventionally accomplished by adverting to an asset-pricing model in finance—which predicts a stock’s return as a function of overall market conditions. See id. Many alternative asset-pricing model variations exist within the literature, but perhaps the most prominent of them is the capital asset pricing model (CAPM). See Ivo Welch, Corporate Finance 242 (3d ed. 2014) (noting the CAPM is “dominant”). In what follows, we deploy the CAPM as a baseline, but we check the robustness of our results against a variety of alternative pricing models.

In order to implement an event-study analysis with sample data, we matched the first available disclosure date for each unique COW with securities-market pricing data for each disclosing issuer. This matching process led to the loss of approximately 100 observations from the 377 Compustat-matched firms. 269 Data Set, supra note 219. To estimate reliably the parameters (αi and βi) for each of the remaining stocks using the underlying asset-pricing model, we additionally required each matched issuer to have sufficient predisclosure days of trading (in this case thirty-five days) at least one month prior to the disclosed COW. This criterion—while consistent with standard practices for avoiding specification error 270 See Eugene F. Fama et al., The Adjustment of Stock Prices to New Information, 10 Int’l Econ. Rev. 1, 4–5 (1969) (describing the exclusions necessary to avoid specifi­cation error). —imposed a significant data limitation: Because many COWs are disclosed as part of either new or newly spun-off companies, many of our COW-disclosing issuers had insufficient trading days preceding the disclosure. In the end, we were able to identify just over eighty distinct issuers that had sufficient predisclosure pricing information to perform an event study. 271 Data Set, supra note 219.

We considered three different event windows, each beginning one day prior to the recorded date of the COW disclosure (to allow for some predisclosure leaks) and ending either one, two, or three days postdis­closure. Figure 5 graphs the mean cumulative abnormal returns (CARs) of the disclosing issuers (pictured with the solid line) and the 95% confidence interval around that mean (pictured with the two dotted lines above and below). We estimate expected returns using the capital asset pricing model (CAPM).

Figure 5: Mean Cumulative Abnormal Return of COW Disclosers 272 The solid line represents CARs, and the dashed line represents 95% confidence intervals.

Figure 5 suggests a statistically imprecise—but economically significant—positive market reception to the disclosure of a COW. For all event windows chosen, COW disclosures predict a positive market reaction, with CARs hovering between 0.5% and 1.3%. 273 Supra Figure 5. The noise associated with these estimates is nontrivial, but for the four- and five-day event windows, the effect is borderline statistically significant at the standard 95% confi­dence level (two-tailed test). 274 Supra Figure 5.

We interrogated the robustness of these results in two ways. First, we varied the nature of the underlying asset-pricing model, intro­ducing (1) a simpler market-adjusted abnormal return measure (the equivalent of setting βi = 1 for all firms); (2) the Fama–French three-factor model, which combines the equity risk premium from CAPM with premia on a large versus small portfolio and a high book-to-market versus low book-to-market portfolio; 275 See Eugene F. Fama & Kenneth R. French, Common Risk Factors in the Returns on Stocks and Bonds, 33 J. Fin. Econ. 3, 7–10 (1993). and (3) the Carhart four-factor model, which extends the three-factor model by adding a momentum factor. 276 See Mark M. Carhart, On Persistence in Mutual Fund Performance, 52 J. Fin. 57, 61–62 (1997). Table 7 repro­duces the results. 277 The event studies here on COW disclosures are based on entities that filed the disclosing document. However, entities other than the issuer adopting a COW may file or join the filing of that disclosure. As a result, we cannot guarantee that the corporation bound by a COW is the relevant filer in each instance (although they will almost always be a party affected by the COW). In future work, we intend to address this limitation. For now, this limitation could introduce attenuation bias reducing the statistical and economic magnitudes of our results. As a result, our event study may be a lower bound for a result based exclusively on COW-adopting entities.

Table 7: Cumulative Abnormal Returns on First COW Disclosure 278 Figures without parentheses are cumulative abnormal returns; figures in parentheses are nonparametric generalized Z-statistics for the sign of CAR (up–down). *** denotes statistical significance at the p < 0.01 level; ** denotes statistical significance at the p < 0.05 level; * denotes statistical significance at the p < 0.10 level (one-tailed test).

As with Figure 5, abnormal returns continue to be estimated with appreciable noise, and many of the estimated effects are statistically insignificant at conventional thresholds. 279 See supra Table 7. Nevertheless, none of the estimation approaches yields negative abnormal returns for the announcement of a waiver, and for many specifications, the average CAR was economically significant—hovering around 1%. 280 Supra Table 7. Note further that the four-day event window spanning days (−1, +2) persistently yields the CAR estimates with the largest magnitudes, and it is statistically significant in all cases. 281 See supra Table 7. This is consistent with the news of a COW disclosure taking some time to penetrate the market. However, the weakening of this effect in the (−1, +3) window—an attenuation that largely flattens out for longer event windows—suggests that some of the initial average response dampens. 282 See supra Table 7. We view these results as evidence that the adoption of a COW does not appear to predict loss of market value: If anything, the opposite is true.

Even if a COW disclosure predicts a mildly positive market response, however, it does not follow that this response is visited evenly on all types of firms. As noted in the previous section, issuers within our sample differ along a variety of dimensions, including incorporation jurisdiction, size, and the type and nature of the waiver at issue. 283 See supra section III.C. For some types of firms (or some types of waivers), the market response is bound to be negative. Although our event study sample is limited in size, it allows us to push on some of these issues further. We therefore broke down our event study along a variety of groupings of the data: Delaware versus non-Delaware incorporation, location of the waiver, who is covered by the waiver, breadth of the waiver, and size of the issuer (as measured by assets). Along each of these dimensions, we measured the CARs associated with a CAPM model and an event window of (−1, +2). Table 8 reports these results.

Table 8: Cumulative Abnormal Returns by Group 284 In each case, we calculate mean CAR using the CAPM and a (−1, +2) event window around the first COW disclosure. Figures without parentheses are mean cumulative abnormal returns; figures in parentheses are nonparametric generalized Z-statistics for the sign of CAR (up–down). *** denotes statistical significance at the p < 0.01 level; ** denotes statistical significance at the p < 0.05 level; * denotes statistical significance at the p < 0.10 level (one-tailed test). For the table, the CAPM is used as the baseline asset-pricing model with the event window of (−1, +2) around first COW disclosure. A “broad” provision specifies either “all” corporate opportunities are waived or corporate opportunities are waived “to the fullest extent allowed” by law. Issuers are considered “large” if they have total assets in excess of $1 billion; otherwise, they are “small.”

Table 8 provides several interesting insights. Note that the Delaware-incorporated firms’ COWs appear to enjoy a uniquely positive market reception, with a mean CAR of 1.67% that is strongly significant. 285 Supra Table 8. Non-Delaware issuers, in contrast, experience a milder (and statistically insignificant) upward abnormal return of approximately thirty basis points. 286 Supra Table 8. Somewhat surprisingly, locating a waiver in a charter amendment provision (meaning it would be voted on and potentially priced by shareholders 287 State law generally requires a shareholder vote as a precondition to a charter amendment. See, e.g., Del. Code Ann. tit. 8, § 242(b)(2017). ) does not seem to improve market reception. To the contrary, COWs executed outside of the charter seem to be met with an especially positive mean abnormal return. 288 See supra Table 8. A particularly interesting aspect of Table 8 concerns how and whether coverage of a corporate officer or dominant shareholder interacts with market reception. One might—on first principles—be especially skeptical about waivers that protect officers and corporate shareholders on the theory that those are the constituencies most susceptible to problematic conflicts of interest. Table 8, however, suggests that COWs covering officers or shareholders do not trigger negative reactions in the capital markets.

Additionally, note that the breadth of the provision does not appear to explain much in predicting market response. 289 See supra Table 8. However, firm size does: Smaller firms (those with under $1 billion in assets) tend to benefit the most from COWs. 290 See supra Table 8. Indeed, they benefit to a significant degree—with a four-day CAR of nearly 2.3%. 291 Supra Table 8. This makes sense in light of theoretical justifications for allowing COWs. Small firms seem more likely to exist in streamlined entrepreneurial environments, drawing on the expertise of many corporate fiduciaries who have many prospective conflicts as the firm grows. Large established firms, in contrast, seem more likely to have significant scope of operations, with greater resources available to attract managerial talent.

A significant concern in analyzing the first disclosure of a COW is the “bundled” nature of most disclosures. As detailed earlier, the vast majority of waiver disclosures are made alongside many other disclosures (such as in a 10-K, a 10-Q, or a public-offering-related filing). 292 See supra note 237 and accompanying text. And even when the COW disclosure is made in a more sui generis filing (such as an 8-K), the disclosure tends to accompany other news (such as an announced restructuring, significant investor, or equity carve out). 293 Data Set, supra note 219. One could argue that such information bundling represents a prohibitive statistical confound, which could easily lead to a spurious misattribution of any observed abnormal returns.

While this critique raises legitimate concerns, it is not entirely clear what the appropriate response should be in the specific context of COWs. As detailed in Part I, the principal impetus behind the reform movement led by Delaware was the argument that COWs are a critical ingredient of value-enhancing corporate structures and transactions—the same ones to which COWs are now routinely attached. 294 See supra section I.B.2 (describing the factors leading to the 2000 amendments); infra note 327 and accompanying text (noting COW adoptions by firms with private-equity ownership). In other words, advocates of the reform made a strong (and to Delaware, ultimately convincing) argument that many types of innovative investment and financing structures would not be possible without the availability of a waiver. To the extent this account is plausible, a COW constitutes a “but-for” cause of the transaction accompanying it. The bundled nature of waiver disclosures, then, may be less of a bug than a feature—a manifestation of the very economic benefits COWs were designed to unleash.

That said, one alternative lens through which to analyze the extent to which COWs contribute to shareholder value is by exploiting the fact that since Delaware’s reform in 2000, eight additional states have adopted statutes permitting COWs. 295 See supra section I.B.4. From any given incorporated company’s perspective, the timing of such reforms is plausibly exoge­nous. As a robustness check, then, we investigated the market response to the passage of COW-enabling statutes across each of the nine states detailed in Table 1. To mark the relevant “event date” in each state, we concentrated not on the statute’s effective date (as detailed in Table 1) but rather on the date on which any remaining uncertainty about enactment’s inevitability was resolved: the date the state’s governor signed the legislative bill. The signing date is also frequently a newsworthy event (and is typically accompanied by contemporaneous signing statements, legislative synopses, or press releases), 296 See, e.g., Permitting Advance Action Regarding Business Opportunities Under the Business Corporation Act, S. 5031, 64th Leg., Reg. Sess. (Wash. 2015); Stephan H. Coonrod & Annamarie C. Larson, Washington’s New Provisions on Advance Waivers of Corporate Opportunities: Opening the Road for Investors, K&L Gates (Apr. 28, 2015), []. helping ensure that it was generally known among the relevant investor communities. Around this event date, we made a similar inquiry to those made in the COW-disclosure event studies, pertaining to whether there was a measurable abnormal return in and around the signing date among issuers incorporated in the reforming state. 297 Specifically, we measured mean CAR starting two days before the event date and up through ten days after the event date. A graphical representation of the results appears in Figure 6.

Figure 6: Mean Cumulative Abnormal Return Around Bill Signing Date for Firms Incorporated in State 298    [1]. The solid line represents mean cumulative abnormal returns, and the dashed line represents 95% confidence intervals.

Figure 6 shows that market responses in the days immediately surrounding the various signing dates appear once again to be positive, hovering somewhere in the 0.4% to 0.8% range up to a week out. Not surprisingly, the economic magnitude of the detected effect is more modest than in the case of COW disclosures, 299 Compare supra Figure 5 (showing mean CAR hovering between 0.5% and 1.3%), with supra Figure 6 (reporting mean CAR between 0.4% and 0.8%). since the enabling bill’s enactment merely gives issuers the real option of adopting a waiver at a later date (presumably as part of a larger financial transaction or restructuring). It is also worth noting that the positive CARs observed in the first week following signing dissipate (and effectively disappear) for longer periods. 300 See supra Figure 6. Nevertheless, it is still a fair conclusion from this preliminary assessment that if anything, securities markets reacted mildly favorably to the enabling act’s execution in those states embracing reform. 301 In the interests of full disclosure, we considered two alternative focal events: the introduction of the enabling bill and the date of its passage in the last legislative house before gubernatorial execution. In both cases, cumulative abnormal returns tended to fluctuate around zero and were statistically insignificant. It is also worth noting that this event study is at best merely suggestive. Corporate law reforms, somewhat analogous to corporations’ disclosures, typically bundle together multiple legislative enactments, such that we cannot disaggregate the markets’ response to the COW enactment specifically. We do not attempt to resolve that limitation here, but leave it to further research.

IV. Implications

The foregoing analysis is no doubt just the tip of the iceberg when it comes to understanding the incidence, drivers, and effects of COWs. Indeed, a key impetus behind this project was a desire to understand more about the results of the statutory experiment started over a decade and a half ago in the Delaware State Legislature. The data set we have developed here will not only inform future academics and researchers, but it also may provide helpful feedback to those interested in shaping and fine-tuning the substance of fiduciary law in the future. That said, our analysis of COWs within public companies affords us a few potential policy insights. This Part explores four potential implications. Section IV.A discusses the relevance of this Article’s findings for evaluating the success of COWs, while section IV.B considers whether it is appropriate to generalize these findings to the duty of loyalty more generally. Next, section IV.C turns to the question of who should design corporate governance arrangements, and section IV.D examines the potential effects of COWs on social welfare, focusing on the possibility of anticom­petitive effects.

A. Evaluating COWs

Many outside observers and investors would plausibly react with alarm when reading a broadly worded COW such as the following: “BE IT RESOLVED: That the Company waives Mr. Ellison’s obligation . . . to present Corporate Opportunities to the Company.” 302 NetSuite Inc., Board Resolutions Approving Corporate Opportunity Waiver (Form S-1/A, exh. 4.5) (Dec. 17, 2007),
1117106/000119312507266011/dex45.htm [].
The breadth of such a provision immediately conjures up images of a thicket of dysfunctional agency costs traditionally believed to plague the diffuse ownership structures of public corporations. Indeed, we initially shared this same instinct. However, this Article’s empirical and conceptual analysis suggests that such fears may actually be mistaken—or at least overblown.

From a purely descriptive perspective, based on the evidence analyzed in section III.C, it appears that COW adoptions tend not to reflect an opportunistic free-for-all among corporate fiduciaries. We find little evidence, for example, that COWs are typically embraced by underperforming firms at which opportunism and agency costs are rife. 303 See supra notes 264–266 and accompanying text. Adopting firms appear instead to be moderate in size, with appreciable growth potential, robust revenue patterns, and relatively strong market returns. 304 See supra Table 6.

Section III.D’s event-study analysis, which is better situated to account for causal inference, suggests a similar conclusion: If anything, securities markets generally appear to welcome the disclosure of COWs, generating weak positive abnormal returns on announcement, although these effects are often not statistically significant. 305 See supra section III.D; supra Table 7. This positive response appears concentrated in moderately sized firms and in Delaware corporations. 306 See supra section III.D; supra Table 8. The underlying scope, breadth, and location of a COW may well matter, but it does not appear to be as critical a statistical factor as some might have thought. 307 See supra section III.D; supra Table 8. Indeed, if anything, the results tend to cut against plausible concerns about specific types of waivers. For instance, one might reasonably expect waivers covering officers or dominant shareholders—or those that are conspicuously broad (such as Mr. Ellison’s)—to be the most value destroying. The market reaction, however, suggests that investors do not recoil at such disclosures. 308 See supra section III.D; supra Table 8.

B. Contractualizing the Duty of Loyalty?

One obvious question raised by Delaware’s statutory experiment is whether it has any implications for debates about contractualizing the duty of loyalty in general and turning corporate law’s few mandatory rules into defaults. Even if the normative picture is muddy, for instance, a reader might be tempted to infer based on our findings that corporations would have a similar appetite for contractual tailoring in other duty of loyalty contexts as they have shown with respect to the COD.

There are two reasons for caution here, however. First, the legal terrain of the duty of loyalty is varied, and its dimensions differ in important ways. 309 See supra section I.A. For instance, the COD may simply be much more complex, indeterminate, and prone to litigation than other forms of conduct proscribed by the duty of loyalty, such as self-interested trans­actions between fiduciaries and the corporation. If this is the case, then opting out of the COD might have proved popular, but the same firms might induce dissimilar reactions in waiving other aspects of loyalty.

Second, and relatedly, if one of the principal reasons firms contract out of the COD is due to the litigation risks the doctrine generates, 310 See supra notes 196–198 and accompanying text. then an alternative explanation for the prevalence of COWs would be their function in reducing the likelihood of litigation, rather than the potential corporate governance benefits discussed above. Both of these caveats provide interesting avenues for future research that this Article leaves for another day. Until they are answered, caution must be exercised in generalizing from section 122(17) to other areas of the duty of loyalty.

C. Who Should Design Corporate Law and Governance?

If the master problem of corporate law is designing optimal governance arrangements to resolve the principal–agent problem, then one of the great “meta” questions in the background is precisely who should determine those arrangements and when. A number of positive and normative issues are implicated here, including the quality of the governance structures offered by law and regulation; whether, when those structures are defaults, corporations tailor their governance around them and whether that governance is superior; and as a result, whom the law should empower to have the last word on crucial governance issues, such as the very loyalty of senior management.

These questions are too fundamental to be answered by any single study or legal issue. Nonetheless, the corporate response to statutory enactments liberalizing COWs—and the securities-market response to that corporate response—have fascinating implications for these debates. At least as regards this part of the duty of loyalty, one descriptive fact seems unassailable: When freed to do so, a wide range of corporations eagerly embrace the power to contract out of aspects of the duty of loyalty. 311 See supra section III.B. Further, such actions tend to meet with a generally positive market response (or at least do not incur a market penalty). 312 See supra section III.D.

More broadly, our study may hold relevance for other domains of the duty of loyalty in which some commentators have long advocated for greater contractual freedom. For example, several law and economics scholars have questioned whether certain insider trading prohibitions should become more “default-like” in nature, with issuers possessing the right to allocate insider trading rights as a form of incentive compen­sation to managers, employees, and other corporate fiduci­aries. 313 See e.g., Ian Ayres & Joe Bankman, Substitutes for Insider Trading, 54 Stan. L. Rev. 235, 267–75 (2001) (proposing “a default prohibition against insider trading by the firm (or its non-managerial delegate)” with the possibility of opting out through a firm’s articles of incorporation); David D. Haddock & Jonathan R. Macey, A Coasian Model of Insider Trading, 80 Nw. U. L. Rev. 1449, 1458–64 (1987); Henry G. Manne, Insider Trading and Property Rights in New Information, 4 Cato J. 933, 938 (1985). For a fascinating picture of corporate policies restricting insiders’ trading beyond what is required by law, see Laura Nyantung Beny & Anita Anand, Private Regulation of Insider Trading in the Shadow of Lax Public Enforcement: Evidence from Canadian Firms, 3 Harv. Bus. L. Rev. 215 (2013). While certain regulatory carve-outs already permit informed insiders to buy and sell securities (such as the safe harbor for executives trading in an automatic 10b5-1 plan 314 See 17 C.F.R. § 240.10b5-1 (2017). ), such provisions are far narrower, more form­ulaic, and arguably more hazardous 315 See, e.g., Alan D. Jagolinzer, SEC Rule 10b5-1 and Insiders’ Strategic Trade, 55 Mgmt. Sci. 224, 228–35 (2009) (documenting the gaming of the administrative rules governing 10b5-1 plans by insiders). than a broader grant of authority to entities to allocate their insider trading rights, just as corporations now do with corporate opportunities. In fact, the availability of COWs may already enable contractualizing insider trading rights in part: Courts once held the opportunity to make strategic stock transactions in other companies’ stock to be a plausible corporate opportunity. 316 See, e.g., In re eBay, Inc. S’holders Litig., No. C.A. 19988-NC, 2004 WL 253521, at *4–5 (Del. Ch. Feb. 11, 2004). A sufficiently generous waiver could permit a fiduciary to enter such transactions (even using confidential corporate information) without abrogating her fiduciary duties, the breach of which is a predicate element for Rule 10b-5 liability under the “misappropriation” theory of insider trading. 317 See United States v. O’Hagan, 521 U.S. 642, 652–53 (1997).         That said, COWs generally do not (and cannot) insulate corporate insiders from liability under the “traditional” insider-trading theory when they purchase or sell their own company’s stock on the basis of material nonpublic information. 318 See Dirks v. SEC, 463 U.S. 646, 655 (1983).

It is important to acknowledge that the availability of potential gains from making loyalty more contractible need not imply that it is socially optimal to give fiduciaries unimpeded power to do so. Indeed, as noted in section I.B.2, the statutory reforms enabling COWs made nowhere near such sweeping pronouncements. Rather, they allow expanded contractual freedom to enter a waiver only under a specific condition: that the process by which the waiver is promulgated must itself be free from the taint of conflicts of interest. 319 See supra notes 134–135 and accompanying text. Like any other transaction, 320 See supra section I.A.1. the action of the board (or controlling shareholders) receives protection under the business judgment rule only to the extent that it is independent and free of influence from materially conflicted parties. This caveat suggests that even the COW reforms did not completely dispatch loyalty’s mandatory character from the COD. A more abstract notion of immutable loyalty may still play a contributing role, albeit now from a substantially greater distance.

D. Shareholder Versus Social Welfare

Even to the extent COWs enhance shareholder value, there is no guarantee that they similarly serve broader social welfare goals. While healthy marketplace competition may generally promote both the welfare of shareholders and society, the two can diverge under a variety of familiar circumstances. One such circumstance is underscored by a newly resurgent literature examining the potentially anticompetitive effects of common ownership of natural competitors within the same industry. A series of recent papers document the dramatic rise in large ownership stakes by a small series of major institutional investors, such as BlackRock, Fidelity, State Street, and Vanguard, which commentators have widely assumed to be passive investors that exercise no influence over the firms in which they invest. 321 Miguel Anton et al., Common Ownership, Competition, and Top Management Incentives (Ross Sch. of Bus., Working Paper No. 1328, 2016),
abstract=2802332 (on file with the Columbia Law Review) (studying the relationship between common ownership and compensation arrangements for top management); Azar et al., Anti-Competitive Effects, supra note 39 (outlining the theoretical case for the anticompetitive effects of common ownership and developing an empirical study of those effects in the airline industry); José Azar et al., Ultimate Ownership and Bank Competition (July 23, 2016) (unpublished manuscript), (on file with the Columbia Law Review) [hereinafter Azar et al., Bank Competition] (studying common ownership’s anticompetitive effects in banking); see also Elhauge, supra note 39, at 1301–16 (developing a legal framework for pursuing some forms of diversified common shareholding as antitrust violations).
For instance, BlackRock is now one of the ten largest shareholders of 70% of the 2,000 largest U.S. public firms, with Vanguard approaching a similar scale of widespread and significant shareholding. 322 Anton et al., supra note 321, at 23. As a result, a small number of institutional investors commonly have the largest ownership stakes in multiple (if not all) of the largest firms competing within a single industry.

Recent papers by José Azar, Isabel Tecu, and Professors Sahil Rina and Martin Schmalz examine the potential anticompetitive effects of shared ownership among competitors on those firms’ performance. Their findings are striking: Common ownership correlates with higher ticket prices in the airline industry, 323 Azar et al., Anti-Competitive Effects, supra note 39, at 17–31. higher prices for services in banking, 324 Azar et al., Bank Competition, supra note 321, at 17–31. and pay for executives that is based less on a firm’s own performance and more on its rivals’ performance across a wide sample of U.S. publicly traded firms. 325 Anton et al., supra note 321, at 26–37. A limited set of exogenous events affecting common ownership enables the authors to suggest a causal interpreta­tion of these relationships. 326 See, e.g., Azar et al., Anti-Competitive Effects, supra note 39, at 21–26. The bottom line of this and other recent research is suggestive and worrying: Two fundamental desiderata of modern finance—diversified shareholding and effective corporate governance—may be deeply in tension with the maximization of social welfare.

This line of research may also bear ominously on how one interprets the positive reaction to COWs documented above. Shareholders may be embracing COWs but at society’s expense, precisely because such provisions facilitate anticompetitive coordination among commonly owned, same-industry firms. Indeed, anecdotally, we can observe that many of the subjects of COWs in our sample are large private equity and venture capital firms likely to have ownership stakes in competitors within the same or related industries. 327 See, e.g., GoDaddy Inc., Registration Statement (Form S-1), at 35 (Nov. 25, 2015), [] (promulgating a waiver as to private equity giants KKR and Silver Lake); HCA, Inc., Prospectus Supplement (Form 424B3), at 7, S-97 (Mar. 3, 2014),
000119312514079396/d678917d424b3.htm [] (promulgating a waiver as to KKR and Bain Capital).
While these firms are quite different from passive, highly diversified institutional investors, their horizontal ownership claims may plausibly be related to anticompetitive effects as well. Indeed, many possible channels exist for COWs to serve as effective mechanisms for dampening competition. Adopting a COW could act as an informational device by which competing firms signal to one another how they will carve up geographical areas or restrict quantities to reduce competition. Alternatively, COWs could simply occur among firms with overlapping directorates or dominant shareholders to directly reduce the extent to which fiduciaries are obliged to present new business prospects to firms to compete over. In either circumstance, the normative complexion of COWs could look quite different.

Future research could adapt the approach undertaken here to study whether COWs play a supporting role in suppressing competition. One plausible avenue for this inquiry would be to examine whether common ownership at either the firm or industry level predicts (or is predicted by) increased incidence of waiver adoption. 328 Pre-existing work in antitrust scholarship constructed a modification of the Herfindahl–Hirschman index (HHI)—the HHI is frequently used by regulators to determine market concentration—called the MHHI, which includes measures of common ownership across competitors within the same industry. See Timothy F. Bresnahan & Steven C. Salop, Quantifying the Competitive Effects of Production Joint Ventures, 4 Int’l J. Indus. Org. 155, 158–61 (1986); Daniel P. O’Brien & Steven C. Salop, Competitive Effects of Partial Ownership: Financial Interest and Corporate Control, 67 Antitrust L.J. 559, 594–98 (2000); see also Anton et al., supra note 321 (adopting and deploying MHHI in a study of the effects of common ownership on executive compensation structures). The MHHI∆ (MHHI–HHI) captures the contribution of common ownership in particular as separated from market concentration. Additional event studies, too, could interact market concentration measures with stock price reactions to COW adoption, including whether an adopting firm’s principal competitors experience positive abnormal returns around the date of adoption. Locating an appropriate exogenous event that increased common ownership could assist in pursuing causal identifi­cation. This Article leaves such endeavors to pursue another day.


At the turn of this century, Delaware ignited an unprecedented, multistate experiment in empowering corporations to waive the COD—an integral part of the fiduciary duty of loyalty. Some sixteen years later, this Article has presented what we believe is the first systematic analysis of how corporations responded to this wave of statutory reforms, as well as related market reactions. Our empirical analysis suggests that public companies have shown a significant appetite for enacting waivers and that their newfound contractual freedom has not been received negatively among investors. This inquiry is interesting in its own right, but it also shares a nexus with some of corporate law’s most important and vexing questions. Descriptively, do corporations actively opt out of corporate law’s default rules when freed to do so? And when they do displace default rules, do such efforts add value or act as a new conduit for managerial opportunism and agency costs? Only when we have a good sense of the answers to both of these questions can we make progress on a third, which is among corporate law’s most indispensable: Are market forces sufficient to ensure optimal corporate governance for corporations? This Article suggests that there may be substantial scope for loosening at least some of the ossified strictures of the duty of loyalty, permitting corporations greater freedom to tailor their governance arrangements so as to best suit their needs and capabilities.

More generally, this Article also reveals an important lesson for the complementary roles that theory, practice, and empiricism can play in legal scholarship. Agency cost theories can provide helpful frames for thinking about the normative stakes involved in analyzing fiduciary waivers, but theory alone is often indeterminate. Practical experiences can help generate insightful anecdotes and stories about purported “best practices,” but they too can easily lead to errant conclusions. Lastly, empirical investigation can provide precise evidence about observable phenomena, but if such evidence is left untethered to underlying theory and intuition, it remains unclear why such evidence is either relevant or interesting. This Article has attempted to make contributions along all three dimensions, gleaning in the process a rich collection of insights that can usefully inform policy debates. Our analysis explored the course that Delaware fiduciary law has begun to chart into the twenty-first century. It is a course in which stakes remain endemically high, and foundational debates constantly recur. And perhaps consequently, it is a site at which empirical inquiry seems long overdue.


Table A.1: State COW Statutes

Table A.2: Coding Rubric

All variables are coded “1” for present, “0” for absent, and “-99” for undecided, other than the ten “Notes” categories, which are demarcated with “[text]” and are optional categories that coders can fill in with detail from the text of the excerpt. For excerpts that were “false positives” (that is, irrelevant documents mistakenly included in our population by our search algorithm), coders simply entered 0’s for each row.