In re Edgio, Inc. Stockholders Litigation, decided by VC Zurn this week, presented the unsettled question whether Corwin cleansing would apply to post-closing shareholder actions seeking injunctive relief for defensive/entrenchment measures.  Interestingly, she held it would not.

The set up: A company was in distress and its stock price was tumbling.  It became afraid that it might be targeted by activist investors.  It also had the opportunity for a very favorable deal: It could buy a business unit owned by Apollo for a large number of newly-issued shares.  The contract specified that Apollo would get one-third of the board seats, but would be required to vote its shares in favor of existing board members.  It would also be prohibited from selling for two years, and after that, it would not be able to sell to “any investor on the most recently published ‘SharkWatch 50’ list for twelve months.”

This was not a transaction that required shareholder approval under Delaware law, but the share issuance did require shareholder approval under NASDAQ rules, and the shareholders voted overwhelmingly in favor.

Post-closing, shareholders brought an action seeking to enjoin the entrenchment measures – not the deal itself – arguing that they violated Unocal.  Defendants, naturally, argued that the whole thing had been cleansed under Corwin by means of the shareholder vote.

VC Zurn began by noting that Corwin itself distinguished between injunctive relief and damages relief; in that case, the Delaware Supreme Court held that “Unocal and Revlon are primarily designed to give stockholders and the Court of Chancery the tool of injunctive relief to address important M & A decisions in real time, before closing. They were not tools designed with post-closing money damages claims in mind.”

In light of that language, VC Zurn held:

Corwin explains that conduct supporting a post-closing claim for damages can be cleansed by stockholders who were satisfied with the economic value they received in a transaction. Inequities in a transaction’s price or process are compensable by monetary damages, and therefore able to be cleansed by stockholders satisfied with the consideration they already received.  But Unocal scrutiny is inspired by concerns that directors may act to “thwart the essence of corporate democracy by disenfranchising shareholders,” which prototypically causes irreparable injury.  Because a dollar value cannot be affixed to the harm caused by unjustifiably entrenching actions, it cannot be said that a stockholder can consider wrongfully entrenching actions as part of the “economic merits” of a transaction.

Similar language about Corwin’s scope, she noted, was used in Morrison v. Berry, 191 A.3d 268 (Del. 2018).

But then she had to reconcile three different pre-Corwin Delaware precedents.

The first, In re Santa Fe Pacific Corporation Shareholder Litigation, 669 A.2d 59 (Del. 1995), involved a pre-closing, post-vote challenge to a merger agreement that contained various lockups that had prevented the company from striking a deal with a competing bidder.  In that context, the court held that the shareholder vote had not restored business judgment review, reasoning:

Permitting the vote of a majority of stockholders on a merger to remove from judicial scrutiny unilateral Board action in a contest for corporate control would frustrate the purposes underlying Revlon and Unocal. Board action which coerces stockholders to accede to a transaction to which they otherwise would not agree is problematic. …Thus, enhanced judicial scrutiny of Board action is designed to assure that stockholders vote or decide to tender in an atmosphere free from undue coercion…

In voting to approve the Santa Fe–Burlington merger, the Santa Fe stockholders were not asked to ratify the Board’s unilateral decision to erect defensive measures against the Union Pacific offer. The stockholders were merely offered a choice between the Burlington Merger and doing nothing. The Santa Fe stockholders did not vote in favor of the precise measures under challenge in the complaint. Here, the defensive measures had allegedly already worked their effect before the stockholders had a chance to vote.  In voting on the merger, the Santa Fe stockholders did not specifically vote in favor of the Rights Plan, the Joint Tender or the Termination Fee.

Since the stockholders of Santa Fe merely voted in favor of the merger and not the defensive measures, we decline to find ratification in this instance.

But, in two other decisions, Stroud v. Grace, 606 A.2d 75 (Del. 1992) and Williams v. Geier, 671 A.2d 1368 (Del. 1996), corporations had adopted various changes to their articles of incorporation that had the effect of entrenching existing management.  When these changes were subsequently challenged by dissenting minority shareholders, the Delaware Supreme Court held that the shareholder votes in favor ratified the amendments.

Faced with this somewhat conflicting set of cases, VC Zurn held:

Stroud and Williams are inconsistent with my reading of Corwin, and, unlike Santa Fe, neither was acknowledged in relevant part by the Corwin Court. Corwin cites Stroud as an example of a vote required by statute or charter that affected the standard of review, indicating the Supreme Court thought the Stroud vote properly had that effect.  Corwin also cites Williams, albeit for the pedestrian principle that a vote will not have a cleansing effect if it is coerced and for the general principle that an informed statutory vote is “the highest and best form of corporate democracy.” Corwin did not explicitly resolve the apparent tension between Santa Fe on one hand, and Stroud and Williams on the other.

I believe I am duty-bound to follow the most recent and specific Delaware Supreme Court authority. Notwithstanding Stroud and Williams, I read Corwin’s plain text as reiterated in Morrison v. Berry, together with Santa Fe’s instructions that Corwin implicitly preserved, to take a claim to enjoin defensive measures under Unocal enhanced scrutiny out of Corwin’s reach. I read that to compel the conclusion that a claim for injunctive relief under Unocal enhanced scrutiny is not susceptible to restoration of the business judgment rule under Corwin.

After that, it was just a question of actually applying Unocal to this particular deal.  And since there was evidence that these moves were intended as a takeover defense – and since the defendants did not even argue that they would pass the Unocal test (only that Unocal did not apply) – she sustained the complaint and ordered the parties to confer further.

So.

It’s actually fairly difficult to imagine that shareholder votes can’t cleanse entrenchment measures.  For example, proxy advisors like ISS generally oppose poison pills that last for over a year unless they are put to a shareholder vote; see also The Williams Cos. Stockholder Litigation, 2021 WL 754593 (Del. Ch. Feb. 26, 2021) (explaining that both ISS and Fidelity had recommended the board put a pill up for a shareholder vote).  Which makes perfect sense: Poison pills, as originally conceived, operated to solve a shareholder collective action problem – two-tier tender offers, for example, put shareholders in a prisoners’ dilemma that requires a cooperative response. 

Further, Corwin did distinguish between injunctive and damages relief in the Unocal context, ‘tis true, but it also distinguished between pre-closing and post-closing.

Moreover, VC Zurn’s pronouncement that “a dollar value cannot be affixed to the harm caused by unjustifiably entrenching actions” is curious; I mean, leave out the “unjustifiably” piece and it seems clear that control rights can be priced.  After all, in Brookfield Asset Mgmt., Inc. v. Rosson, 261 A.3d 1251 (Del. 2021), the Delaware Supreme Court, quoting Paramount Communications, Inc. v. QVC Network, Inc., 637 A.2d 34 (Del. 1994), held:

“the acquisition of majority status and the consequent privilege of exerting the powers of majority ownership come at a price,” and that price “is usually a control premium which recognizes not only the value of a control block of shares, but also compensates the minority stockholders for their resulting loss of voting power.”

That said, there is a very real distinction between Stroud and Geier on the one hand, and Santa Fe on the other, and it comes from the language of Santa Fe itself.  In that case, shareholders were presented with a deal – a full package of items, including a buyout at a premium; there was no opportunity to vote separately on the deal protections.  By contrast, in Stroud and Geier, shareholders were presented with an opportunity for a clean up-or-down vote on the entrenchment measures alone.

What immediately springs to mind, then, is the possibility – again, suggested in Santa Fe – that the shareholder vote in that case was not in fact free and fair, but was coerced.  Shareholders had to accept the deal protections if they wanted the premium, and there was certainly no way to know whether another deal would come through if they rejected the package outright.

That is what’s known as a “bundling” problem, i.e., when shareholders are not permitted to vote separately on particular items, and it’s actually something the SEC (theoretically) regulates, although of course the SEC hasn’t changed the requirements of Delaware law.

Not long ago, in Sciabacucchi v. Liberty Broadband Corp., 2017 WL 2352152 (Del. Ch. May 31, 2017), VC Glasscock held that a shareholder vote in favor of a share issuance to an existing blockholder was coerced – and thus not subject to Corwin – because it was “bundled” with certain transactions that would be beneficial to shareholders.  As VC Glasscock put it, the issuance was plausibly “extrinsic, and tacked to the Acquisitions to strong-arm a favorable vote.”

One could say something similar about the facts of Edgio: a favorable transaction was conditioned on acceptance of the entrenching measures, and therefore the shareholder vote was not free and fair, as Corwin requires.

But that argument perhaps proves too much.  Is it “bundling” if there is a premium-generating deal that comes with the loss of control rights, or is that merely a bargained-for exchange?  Commenters have repeatedly attacked Corwin itself for creating a “bundling” problem, in that shareholders voting on a deal are deemed to also be voting to absolve fiduciaries of any liability, but the Delaware Supreme Court obviously didn’t have a problem with that.

In other words, what is needed is some kind of theory as to what counts as an “extrinsic” condition versus what counts as intrinsic to the deal.  (Obligatory plug: I previously likened this issue to the unconstitutional conditions doctrine in my essay, The Three Faces of Control.)

One other point I’ll make on this, though.  Corwin itself, as well as Kahn v. M&F Worldwide, 88 A.3d 635 (Del. 2014), are implicitly predicated on the recognition that today’s shareholder base is concentrated and institutionalized, and therefore presumably capable of protecting its own interests; indeed, many of the Chancery court decisions that paved the way said so explicitly.  See, e.g., In re MFW S’holders Litig., 67 A.3d 496 (Del. Ch. 2013); In re Cox Commc’ns, Inc. S’holders Litig., 879 A.2d 604 (Del. Ch. 2005); In re Netsmart Techs. S’holder Litig., 924 A.2d 171 (Del. Ch. 2007).  Which is why it is striking that Delaware is proposing to amend its law to make it easier to pass certain charter amendments (pertaining to increasing or decreasing shares outstanding), in recognition of the meme stock era and the re-retailization of the shareholder base.   It remains to be seen whether Delaware takes any steps to soften its standards of review as well.

In Amalgamated Bank v. Yahoo!, 132 A.3d 752 (Del. Ch. 2016), VC Laster allowed a corporation to condition production of documents sought under Section 220 on the stipulation that if the plaintiff filed a lawsuit that relied on any of the documents, the entire production set would be deemed incorporated by reference into the complaint.  VC Laster considered this to be a reasonable stipulation to protect the defendant against the plaintiffs’ strategic cherry picking of documents in order to create a misleading impression of the facts.

Since then, such stipulations have become common, and the practice has evolved for defendants to certify that they have produced all responsive documents to the plaintiffs in order to get the benefit of incorporation-by-reference.  See, e.g., In re Vaxart S’holder Litig., 2022 WL 1837452 (Del. Ch. June 3, 2022).  The predictable result is that motions to dismiss are accompanied by ever-more-lengthy lists of exhibits.  Delaware Chancery has repeatedly warned that these “incorporation by reference” stipulations do not change the standard of review, and that it has the option either to disregard excess submissions or use them as a basis for transforming the dismissal motion into a motion for summary judgment.  See, e.g., In re Match Group Deriv. Litig., 2022 WL 3970159 (Del. Ch. Sept. 1, 2022).

In Oklahoma Firefighters Pension & Ret. Sys. v. Amazon.com, Inc., 2022 WL 1760618 (Del. Ch. June 1, 2022), VC Will permitted a defendant to redact documents produced in response to Section 220 demands to remove material that was nonresponsive to the plaintiffs’ request.  As VC Will put it, “In my view, redactions to material unrelated to the subject matter of a demand are proper because Section 220 only entitles a stockholder to information essential to accomplishing its stated purposes for inspection. The redactions appropriately cabin Section 220 inspections to their intended scope.”

From VC Laster’s opinion this week in Ontario Provincial Council of Carpenters’ Pension Trust Fund et al. v. Walton, 2023 WL 3093500 (Del. Ch. Apr. 26, 2023), one gets the distinct impression that he believes these two developments – singly and in combination – have been abused by defendants, and he’s fighting back. 

In Walton, the plaintiffs sued the Walmart board for violations of its Caremark duties with respect to opioid prescriptions, and relied on the heavily-redacted documents that Walmart produced in a prior 220 action.  Walmart, of course, submitted its own documents in support of its motion to dismiss.  Given the procedural posture (reasonable inferences drawn in the plaintiffs’ favor), VC Laster held that the certification of completeness, coupled with an incorporation-by-reference provision, allowed for the inference that “if the record lacks documentation relating to a particular event, and if it is reasonable to expect that documentation would exist if the event took place, … [then] the event did not occur.” Id. at *2.  With respect to redactions, which he often found “dubious” because “a partial-sentence redaction … depends on the premise that the author incoherently injected an unrelated topic into an otherwise responsive sentence,” id. at 3 – he either gave them a plaintiff-friendly inference or also treated them as evidence of the absence of discussion.  The net effect was a near weaponization of the lack of record evidence on various topics in order to conclude that the Walmart board failed to conduct proper oversight of the company. 

Prior cases have drawn inferences from the lack of 220 documents on a subject, e.g., In re Boeing Co. Deriv. Litig., 2021 WL 4059934 (Del. Ch. Sept. 7, 2021); Teamsters Local 443 Health Servs. & Ins. Plan v. Chou, 2020 WL 5028065 (Del. Ch. Aug. 24, 2020), but I’ve never seen anything quite like this.  To wit:

Walmart did not produce a final budget for the Health and Wellness Division as part of its Section 220 document production, entitling the plaintiffs to an inference that a budget sufficient to fund the projects necessary to comply with the DEA Settlement did not exist.

 ….

The memo identified two significant challenges. The first challenge, identified in a single sentence, was redacted. The redaction was marked “NR/ACP/AWP,” for non-responsive, attorney-client privilege, attorney work product. Because the single sentence redaction appears in an otherwise responsive paragraph, the redaction is dubious, and with three possibilities provided, the basis for it is unclear. At the pleading stage, the plaintiffs are entitled to an inference that the redacted text referenced a compliance failure that Walmart was not addressing.

….

An appendix to the slide deck identified a list of items that Walmart had completed during fiscal years 2011 and 2012. …Significant portions of the appendix are redacted. … The list of completed activities conspicuously omits significant items identified in the DEA Settlement, such as testing for doctor shopping, flagging requests for early refills, or checking for altered or forged prescriptions.

Read together, the 2012 Memo and accompanying materials support a pleading-stage inference that the Health and Wellness Division had created a summary of what a nice compliance program would look like, then never did the work to implement one.

He goes further.  He recognizes that he cannot draw negative inferences from redactions specifically for attorney-client privilege, but he also holds that a conscientious board, appropriately overseeing the company, would be expected to have at least some discussions of legal compliance that did not involve privileged conversations, and thus the absence of evidence of nonprivileged discussions would also be treated as evidence of absence.  As he put it:

Although this decision does not draw inferences from any of the passages or documents for which Walmart has asserted privilege, it does draw inferences from an absence of non-privileged documents containing discussions or decisions about the business issues necessarily involved in (i) taking the steps necessary to comply with the DEA Settlement and the Controlled Substances Act, (ii) responding to red flags of noncompliance, and (iii) assessing the effectiveness of the compliance efforts. Legal advice undoubtedly is an input into those discussions and decisions, but if directors and officers are doing their jobs, then there will be non-privileged discussions and decisions about what are inherently and ultimately business decisions (which the business judgment rule generally will protect). Walmart represented that its Section 220 production was complete, so when there are no indications of non-privileged discussions, the plaintiffs are entitled to an inference that the discussions and decisions did not occur.

The result was:

Walmart produced a copy of the meeting minutes, which comprise seven pages. All of the substantive portions of the minutes were redacted for non-responsiveness and attorney-client privilege with the exception of the following sentence: “Ms. Harris then provided an update to the Committee on the overall status of Health and Wellness Compliance projects.” … There are no non-privileged documents reflecting the Employee Compliance Committee making any business decisions or taking any action. At the pleading stage, the absence of evidence about action by the Employee Compliance Committee supports a plaintiff-friendly inference that the Employee Compliance Committee failed to take action to promote compliance with the DEA Settlement.

 ….

Walmart’s privilege log contains entries suggesting that the Employee Compliance Committee met on twenty other occasions from 2016 through 2020. Walmart withheld all of the relevant meeting minutes, noting on its privilege log only that the discussions involved “controlled substances,” “opioids,” or Walmart’s “Health & Wellness compliance program.” There are no indications that the committee had any business discussions, made any business decisions, or took any type of action. If Walmart’s assertions of privilege are to be believed, then as the opioid epidemic raged, Walmart’s senior compliance employees did nothing except receive and consider legal advice. They knew about the problem and took no action whatsoever. Although that seems highly unlikely, that is the record that Walmart created through its highly redacted Section 220 production.

The fact that so many meetings took place supports an inference that the officers and employees on the Employee Compliance Committee closely monitored Walmart’s compliance with its obligations under the Controlled Substances Act. At the same time, the allegations in the complaint, together with other documents in the record, support an inference that Walmart was failing to comply with its obligations as a dispenser of prescription opioids and, in particular, was undermining its pharmacists’ ability to fulfill the Refusal-To-Fill Obligation. The court therefore must infer that the Employee Compliance Committee knew about Walmart’s failure to fulfill its obligations as a dispenser of prescription opioids. The absence of any indication that the Employee Compliance Committee did anything except gather to receive and discuss legal advice, supports a pleading-stage inference that the members of the committee consciously ignored Walmart’s compliance failures.

….

The Board also met in November 2017, and the minutes of that meeting span sixty-eight pages. Ex. 61. Only three sentences of substantive text survived the redaction tool. The first reads: “Timothy P. Flynn, Chair of the Audit Committee, then provided the Audit Committee report.” The following partially redacted text appears on the next page: “He concluded his report by stating that the Audit Committee had received updates from management regarding various other matters including … [REDACTED] … enhanced processes and training for pharmacists regarding filling prescriptions of controlled substances.” The redactions were marked for non-responsiveness, attorney-client privilege, and attorney work product. The unredacted text provides no basis to infer that the Board or Audit Committee had any business-oriented discussion about compliance issues or made any business decisions about compliance issues.

And my personal favorite:

Walmart redacted the entire director education presentation on the basis of the attorney-client privilege and work product doctrine. Because of those redactions, the only possible inference is that during a meeting specifically called to address the opioid crisis, Walmart’s directors and senior executives and unidentified members of the Walton family did not discuss any business issues, consider any business initiatives, or make any business decisions. All they did was receive and consider legal advice. Although that is hard to believe, Walmart’s redactions necessarily lead to that inference.

In more sorrow than anger, he writes:

Reinemund served on the Board during the full term of the DEA Settlement. …He therefore also faces a substantial risk of liability, and there is reason to doubt whether Reinemund could consider a demand.

The court reaches this conclusion reluctantly, because Reinemund is a person of stature who has had an impressive career. He graduated from the United States Naval Academy and served in the Marine Corps, rising to the rank of Captain. After leaving the military, Reinemund enjoyed success in the business world, culminating in the position of Chairman and CEO of PepsiCo from 2001 to 2003. From 2008 to 2014, he served as Dean of the Wake Forest University Business School. In addition to serving as a director at Walmart, he has served on the boards of other major public companies.

Why would an outside director like Reinemund ignore red flags about noncompliance with the DEA Settlement, much less make a conscious decision not to achieve compliance with the DEA Settlement? The answers likely lie in the redacted portions of the documents in Walmart’s Section 220 production. 

See also id. at *41 (“As with Reinemund, it is hard to believe that an outside director like Flynn would ignore red flags about noncompliance with the DEA Settlement, much less make a decision not to comply with it. Once again, the answers likely lie in the redacted portions of the documents in Walmart’s Section 220 production. Unfortunately, because of Walmart’s compulsive redacting of documents, the pleading-stage record supports an inference that Flynn knew that Walmart was not in compliance with the DEA Settlement, knew that Walmart could not achieve compliance by the time the DEA Settlement terminated, and did nothing to bring the company into compliance.”).

That’s the headline, but it’s not all that’s of interest here.

I earlier blogged about Caremark and the distinction between taking legal risks and business risks.  In Walton, VC Laster articulates the distinction between legal risk and business risk thusly:

• In one hypothetical scenario, the lawyers say: “Although there is some room for doubt and hence some risk that our regulator may disagree, we believe the company is complying with its legal obligations and will remain in compliance if you make the business decision to pursue this project.”

• In the other hypothetical scenario, the lawyers say: “The company is not currently in compliance with its legal obligations and faces the risk of enforcement action, and if you make the business decision to pursue this project, the company is likely to remain out of compliance and to continue to face the risk of an enforcement action. But the regulators are so understaffed and overworked that the likelihood of an enforcement action is quite low, and we can probably settle anything that comes at minimal cost and with no admission of wrongdoing.”

In the former case, the directors can make a business judgment to pursue the project. In the latter case, the decision to pursue the project would constitute a conscious decision to violate the law, the business judgment rule would not apply, and the directors would be acting in bad faith.

Critically, I note that his assessment of good faith in these scenarios permits reliance on legal advice.  See also id. at *40.  That’s interesting, because – as I blogged when discussing the case of The Williams Companies v. Energy Transfer Equity LP, 159 A.3d 264 (Del. 2017) – there’s been some disagreement at the Delaware Supreme Court over approaches to legal advice.  In Williams, then-Chief Justice Strine in dissent recognized that even “good faith” legal advice is malleable, and may be adjusted depending on the desire of the client.  The majority, by contrast, treated “good faith” legal advice as something akin to an unwavering North Star, unresponsive to the client’s needs.  As recently as Bandera Master Fund LP v. Boardwalk Pipeline Partners, LP, 2021 WL 5267734 (Del. Ch. Dec. 19, 2022), Laster seemed to take more of Strine’s view of the matter, but his opinion was reversed on appeal in Boardwalk Pipeline Partners, LP v. Bandera Master Fund LP, 288 A.3d 1083 (Del. 2022), with a concurrence specifically warning that a great deal of deference ought to be given to good faith legal opinions.

The upshot of which is, going forward, I worry about how easy it will be for boards to satisfy Caremark obligations, despite red flags of illegal conduct, by relying good-faith-but-under-duress legal opinions.  And it is particularly ironic to see such a sentiment expressed in an opinion that does not, in the main, evince much confidence in counsel’s good faith advocacy.

As Joan already flagged, Delaware litigation has broken out over whether TripAdviser can reincorporate to Nevada.  The Complaint in that action argues that Delaware courts should enjoin the move and prevent the corporation from moving to Nevada.  The New York Post has also picked up on the reincorporation trend.  In my view, the conditions under which a corporation may exit a jurisdiction and reincorporate elsewhere is a valid question of its current state’s corporate law.  While Delaware may want to impose appropriate conditions on some departures, I doubt it wants to become the Hotel California of corporate law–where you can check in any time you want, but you can never leave.

There are undoubtedly material differences between Delaware and Nevada law.  As I covered here at the time, Nevada’s statutory business judgment rule imposes more limits on liability that Delaware law.  Nevada’s Supreme Court has explained that establishing liability under Nevada law “plainly requires the plaintiff to both rebut the business judgment rule’s presumption of good faith and show a breach of fiduciary duty involving intentional misconduct, fraud, or a knowing violation of the law.”  

Nevada’s law may be better for some corporations than Delaware’s law.  Duke’s Ofer Eldar found that “Nevada corporate law does not harm shareholder value for firms that self-select into Nevada, particularly small firms with low levels of institutional shareholding and high levels of insider ownership, and it may in fact enhance the value of these firms.”

Ultimately, as a stakeholder in Nevada corporate law, the last thing I want is for our state to become a cesspool for fraud and misconduct.  Nevada’s protective business judgment rule strikes a different balance than Delaware, but it’s not going to let people get away with rank misconduct.