Redemption Mechanisms in Poison Pills: Evidence on Pill Design and Law Firm Effects

A shareholder rights plan is a long and rather complicated document, but it is also highly standardized. The contract drafting generally focuses on certain definitions, such as “Acquiring Person,” “Beneficial Owner,” “Exempt Person,” and “Triggering Threshold” to tailor the triggering mechanism of the poison pill to the specific needs of the company, such as: (1) the triggering threshold; (2) whether there is a higher threshold for institutional shareholders filing a Schedule 13G; (3) the forms of shareholding (i.e., beneficial ownership of shares), which result in triggering a threshold; and (4) whether certain existing shareholders are exempted (typically shareholders who already hold more than the trigger threshold at the time the poison pill is installed).

This redemption mechanism of poison pills is typically structured pursuant to one of the following two alternatives:

• Trip-Wire Provision: If the poison pill implements a “trip wire” concept, the rights granted thereunder are triggered if, and can no longer be redeemed by the board once, the acquiror exceeds the triggering threshold set in the shareholder rights plan.
• Last-Look Provision: If the poison pill implements a “last look” concept, the board of the target company has a “last look” for a certain period—in most cases ten business days—after the poison pill has been triggered to decide whether to redeem the rights granted thereunder. If the board redeems the rights, the pill is thereby “defused,” and the bidder can continue buying shares of the target company.

The pill will invariably have two different mechanisms to effectuate the dilution, if such dilution needs to occur. The first, known as the “flip in” mechanism, provides all shareholders, other than the triggering shareholder, the right to buy shares of the target company at a deep discount (typically 50 percent of market value). Using the flip-in mechanism has two general problems: It requires an overwhelming number of common shares, and fractionalizing the preferred shares, which is the way to avoid the first problem, is not as clear-cut a solution as it used to be. We discuss each of these issues in turn.

1. Headroom in the Common Shares

First, the pill requires an enormous number of authorized but unissued shares in order to satisfy the rights issued thereunder. Among all U.S. public companies today (n=4,810), the average ratio of outstanding shares to authorized shares is 32 percent, meaning that there is only 68 percent of “headroom” at the typical U.S. public company. This headroom is what would be used to issue common shares to give the flip-in feature its dilutive effect.

Although the exercise price allegedly represents the (very) long-term value of the company, it is essentially an arbitrary number well in excess of the current trading price. For example, the Twitter exercise price was $210, compared to a then-trading price of approximately $48. Assuming (conservatively) an exercise price that is only twice the market price and a 10 percent trigger threshold, a company would need 78 percent headroom in order to effectuate full exercise of a flip-in feature using only common shares. Approximately one-third of U.S. public companies have sufficient headroom to be able to effectuate a flip-in pill using only common shares, even under these conservative assumptions. And for exercise prices that are greater than twice the current market price (which then increases the potency of the pill), the number of companies able to effectuate a flip-in exercise using only common shares goes down even further. For example, at a 5x ratio of exercise price to current market price, a company would require 90 percent headroom. Only 15 percent of U.S. public companies have this much headroom.

At least in Delaware, authorizing additional shares requires a shareholder vote, which would be virtually impossible in the timeframe required to effectuate a pill. Instead, it is widely understood that most companies would create synthetic common shares through a series of preferred stock that is currently authorized but has no issued shares (“blank check preferred stock”). The vast majority of U.S. public companies have such blank check preferred stock. In order to synthetically create the necessary number of common shares, a board of directors could “fractionalize the preferred” by mandating that each share of preferred stock had all of the rights of, e.g., 1,000 shares of common stock. The conventional wisdom among most practitioners is that fractionalizing the preferred is the way to effectuate the pill dilution for the vast majority of companies that would not have sufficient headroom in their common shares.

2. Viability of Fractionalizing the Preferred Shares

Until recently, this conventional wisdom was unchallenged. However, in 2023, AMC Entertainment Inc. (“AMC”) attempted to create synthetic common stock by fractionalizing a series of blank check preferred stock. The AMC board had no remaining headroom in the common shares, but needed equity financing to avoid bankruptcy. The board resorted to the preferred share issuance only after trying unsuccessfully (twice) to get the AMC common shareholders to approve a new share authorization.

The resulting AMC Preferred Equity (APE) shares traded at a significant discount to the AMC common shares, thereby leading to shareholder litigation which ultimately resulted in a settlement in favor of the common shares. Although the AMC preferred share issuance was in the context of an equity issuance and not a pill trigger, it raises questions about the feasibility of fractionalizing the preferred shares as a means of achieving the necessary dilution in the aftermath of a trigger event.

3. Viability of the Exchange Feature Safety Valve

Virtually all pills have a “safety valve” on the flip-in feature, known as the exchange feature (discussed above). In comparison to the flip-in feature, the exchange feature is straightforward: all shareholders, other than the triggering shareholder, receive one new share for every right currently held. From the perspective of the target’s board, exercising the exchange option is generally preferable to the flip-in feature, because the exchange avoids the huge influx of cash that the board would then need to invest or, lacking sufficient investment opportunities, re-distribute to the shareholders (which triggers tax). The exchange feature also requires less headroom than the flip-in feature, though the lack of headroom can still be a barrier for some companies even when using the exchange feature. For example, with a 10 percent trigger threshold, effectuating a pill dilution through the exchange feature requires 47 percent headroom in the common shares. A full 19 percent of U.S. public companies do not have sufficient headroom to even effectuate a pill trigger through the exchange feature. For these 19 percent of companies, even exercising the exchange feature would require issuance of the company’s blank check preferred stock.

4. Implications for Pill Design

The viability of effectuating a pill trigger through common shares, preferred shares, or an exchange provision is not the focus of this article, so we note the issue only to observe the uncertainty with respect to many aspects of pill mechanics. For example, with respect to the “flip over” feature that every pill has as well, a then-sitting Vice-Chancellor has expressed skepticism about how it would work in practice:

Looking back on the Household fight, [Delaware Vice Chancellor Steve] Lamb was still wondering how it could be that the directors of one company can be allowed [to] send their own shareholders to plunder the stock of another company at a discounted price that they unilaterally determine in advance. He rolled his eyes and smiled: “I just don’t know.”

A pill has been triggered only once in modern times, when Versata crossed 5 percent ownership at Selectica, which, at such time, had a $20 million market capitalization. The Selectica pill, installed on February 4, 2003, originally had a 15 percent trigger and was structured as a trip-wire pill. However, on November 17, 2008, Selectica amended its pill by reducing the trigger threshold to 4.99 percent. In connection with this amendment, Selectica also broadened the definition of “Exempt Persons,” which, in short, gave Selectica’s board full discretion to exempt shareholders from the 4.99 percent trigger. The effect of this provision was to serve as a “safety valve” that is similar to a “last look” feature (notwithstanding the fact that the redemption provision of the Selectica pill included a trip-wire feature).

Selectica stopped trading in its stock for nearly a month (from January 6 to February 4, 2009) while the Selectica board and its advisors considered its options. Solomon (2009) observed that “perhaps the Selectica board had second thoughts or settlement negotiations between the parties were occurring” during this one month period. David Katz, a partner at Wachtell, Lipton, Rosen & Katz, observed that the Selectica pill’s de facto last look “can put boards under a great deal of pressure and may potentially lessen the deterrent effect of a rights plan.” In the end, the Selectica board chose not to exempt Versata from the 4.99 percent trigger—by exercising its discretion that allowed for the de facto last look—instead diluting Versata down from 6.7 percent to a 3.3 percent ownership interest (representing a $680,000 loss to Versata).

Putting aside the relatively minor pill trigger at Selectica, a pill has never been meaningfully triggered in the United States. This means that practitioners have not been forced to think through the mechanics of what actually would happen in the event of a pill trigger. Our conversations with attorneys who install pills for boards indicate that they rarely present the mechanics of what would actually happen in the event of a pill trigger (or even calculate for themselves what would happen), perhaps in part because the pill is often installed in exigent circumstances, where business considerations (e.g., motivations of the buyer, possibility of other bidders, etc.) are more important board-level considerations than the mechanics of how the pill dilution would work. This “uncharted territory” provides important context for the empirical analysis we present in the next Section.

B. Prior Literature

1. Pill Design

While there is a vast prior literature on poison pills, this literature focuses primarily on the Delaware doctrine of pills, as well as the policy implications of endorsing pills. Subramanian (2007) is the first to identify the difference between “last look” and “trip wire” provisions in poison pills. He observes that the PeopleSoft pill had a “last look” feature, and describes the negotiation implications of this last look:

In negotiation analytic terms, the [PeopleSoft] pill does not effectively commit to financial Armageddon in the event that Oracle [the hostile bidder] buys more than 20% of PeopleSoft. By giving PeopleSoft a “last look” to avoid disaster, the pill in fact weakens PeopleSoft’s bargaining hand. Analogies to nuclear weapons are obvious here: whoever has the last chance to pull back is likely to take it due to the severity of the instrument, which in turns gives bargaining power to the party that can irrevocably commit. By giving itself a last look, PeopleSoft made it less likely that Oracle would have suffered the dilutive effect of the pill.

To our knowledge no further academic literature observes the difference between “last look” and “trip wire” provisions in poison pills.

Among practitioners, commentary on last-look/trip-wire pills is also sparse—perhaps because most law firms (particularly top tier law firms) might view their boilerplate pill documents as being proprietary intellectual property. A notable exception is Morrison & Foerster’s “Poison Pill Deep Dive” series, which observes:

There is some debate as to whether including a last look provision in a rights plan is good for the company. . . . On the one hand, it seems sensible to put this decision in the hands of the board rather than a third party. A triggered rights plan will significantly affect the company and its capital structure, and events significantly affecting the company should be decided by the board. On the other, giving the board the final call on whether the dilutive effects occur may weaken the rights plan’s deterrent value. This happens because, during the 10-day window after the plan has been triggered, the board will be under considerable pressure in deciding whether to redeem the rights. The pressure comes from the fact that the board’s decision must be made consistent with the board’s fiduciary duties, based on current knowledge of the company’s situation, including the “threat” posed by the particular acquiror, and the potentially significant effects of the triggered plan on the company.

Pressure may also come from the company’s other stockholders. Although diluting the acquiror gives the other stockholders the chance to increase their ownership percentages at a deeply discounted value, or even for free (if the board opts to implement an exchange rather than to allow exercise of the rights), it will likely take the acquiror’s proposed deal off the table. Moreover, at the point when a plan is triggered, at least in the takeover context, a significant number of the company’s stockholders will be arbitragers who had bought company stock with the expectation that a deal would happen and will likely want the board to redeem the rights and let the acquisition proceed, with less interest in the company’s long-term prospects.

As a result, if a rights plan with a last look provision is triggered, the acquiror knows that the board may be incentivized to negotiate and that there is a possibility that the board will decide to redeem the rights—thereby reducing the rights plan’s deterrent value.

In the end, whether or not to include a last look provision is an important issue for the board to consider, and it is key that the board be advised as to the positives and negatives of including the provision.

2. Law Firm Effects

There is also an academic literature—again sparse—on law firm effects in corporate law. Coates (2001) provides empirical evidence that West Coast law firms (in particular, Wilson Sonsini, Goodrich & Rosati) are far less likely to put in takeover defenses (such as a staggered board) at the initial public offering, relative to East Coast (primarily New York City) law firms. Coates observes that the variation in takeover defenses at the IPO stage of a company can be traced back to “failure in the market for legal services. In short: blame the lawyers.” Subramanian (2005) similarly finds that law firms that have significant M&A experience are more likely to use cutting-edge techniques for freezing out minority shareholders than other law firms, thereby yielding lower freeze-out pricing for their buy-side clients.

The American Lawyer suggested broader implications of these findings: “In corporate law, of course, most of the elite firms are in New York, and charge accordingly. But are either their prestige or their prices merited? Guhan Subramanian provides evidence that they are.” The Deal further expounded: “So there really is a cutting edge, and New York firms seem to be on it, the Harvardians [Coates & Subramanian] suggest. And the edge cuts pretty slowly through the bar in the rest of the country.”

Most relevant for present purposes, The Deal further described the implications of cutting-edge techniques for the dissemination of knowledge within a profession: “[W]e were struck by other questions the [Subramanian (2005)] piece implies. How does knowledge spread within a profession? Why are some people and firms better than others at acquiring—and, more importantly, employing knowledge?” Research in the medical profession, for example, indicates that specialized surgeons who do the same operation repeatedly make fewer mistakes than generalist surgeons. In general, other professions (such as medicine) seem to have more systematic processes for dissemination of cutting-edge tools and techniques than does the practice of law.

III. Empirical Evidence on Pill Design

A. Data Sources and Sample

For the purpose of our study, we constructed a new dataset of all shareholder rights plans in the SEC’s EDGAR database that met the following three criteria:

• Issuer incorporated in the United States: Given that the legal framework for takeover defenses in general, and the implementation of poison pills in particular, depends on the applicable corporate law, the analysis in this article focuses on poison pills implemented by companies incorporated in the United States. Accordingly, even though some companies incorporated in foreign jurisdictions implement U.S.-style poison pills, only poison pills implemented by companies incorporated in the U.S. are included in the data set of this analysis.
• No 382-Pills: Originally, poison pills were designed as a defense mechanism by companies against coercive two-tier tender offers. After the Delaware courts endorsed poison pills in the context of hostile takeovers, companies began implementing poison pills to protect their net operating loss carryforwards (“NOLs”) as well. The focus of NOL pills was the prevention of an “ownership change” that results in significant limitations of the use of NOLs as set forth in Section 382 of the Internal Revenue Code of 1986 (“IRC”). As these 382- or NOL-pills serve a different purpose and have much lower thresholds, any such 382- or NOL-pills were not included in the data set of this analysis.
• Pills implemented and/or amended in the period between January 1, 2020 and March 31, 2023: In order to have a data set that is representative of the current prevalence—and consequently the preference of legal advisers—regarding the implementation of a redemption mechanism with a trip-wire or last-look feature, we included only poison pills that were implemented and/or amended in the period from (and including) January 1, 2020 to (and including) March 31, 2023.

B. Methodology

1. Sample Description

For each poison pill that met these criteria, we examined the SEC filings by the relevant company (primarily 8-K and 8-A12B), as well as press releases published in connection with the introduction of the pill (or the overall transaction that the poison pill was a part of). To ensure the completeness of the data set so collected, we cross-checked data gathered from the SEC’s EDGAR database against FactSet’s corporate activism database (“FactSet”). The FactSet database did not contain any additional poison pill, thereby confirming the completeness of the EDGAR database.

These search criteria identified 135 poison pills. With one exception, all companies filed a complete copy of the shareholder rights plan (or the amendment thereto) with the SEC. Four companies implemented two identical poison pills within the observation period of this study. Given that the two poison pills implemented by these four companies do not have separate informative value, we only included the most recent poison pill in the data set. The final data set includes 130 poison pills (the “Pill Sample”).

2. Identifying Trip-wire Design

For each pill in the Pill Sample, we determined whether the pill was a last-look pill or a trip-wire pill. The typical language implementing the trip-wire concept is the following:

Redemption. (a) The Board of Directors of the Company may, at its option, at any time prior to such time as any Person becomes an Acquiring Person, redeem all but not less than all the then outstanding Rights at a redemption price of $0.01 per Right, appropriately adjusted to reflect any stock split, stock dividend or similar transaction occurring after the date hereof (such redemption price being hereinafter referred to as the “Redemption Price”) . . . .

The key identifier of the trip-wire concept is the wording “at any time prior to such time as any Person becomes an Acquiring Person.” Shareholder rights plans define “Acquiring Person” as a person that holds common shares in an amount equal to or exceeding the threshold that triggers the poison pill. Alternatively, shareholder rights plans use a defined term (e.g., “Shares Acquisition Date,” “Flip-in Date,” etc.) for the date on which a shareholder becomes an “Acquiring Person” (or announces that it has acquired shares in an amount that reaches or exceeds the triggering threshold). In both cases, the wording has the effect of terminating the board’s right to redeem the rights issued thereunder when a shareholder reaches or exceeds the triggering threshold. This is the key feature of the trip-wire provision: Exceeding the triggering percentage by the shareholder automatically results in triggering the poison pill, with the board having no option to avoid the dilution.

Two “end runs” around the trip wire must be cut off in order for the trip wire to be effective. First, pills typically contain an amendment provision that allows the target’s board to amend the pill without shareholder approval. If such amendment provision were to allow changes to the pill even after it were triggered, this could serve as a loophole to get around a trip-wire mechanism in the redemption provision. Specifically, the board could provide for a de-facto-redemption by (among other things) amending the plan to exempt the acquiror from the definition of “Acquiring Person,” or amending the plan to make the exercise of the rights economically unattractive. Accordingly, a trip-wire concept is effective only if the irrevocable commitment to the dilution is reflected in the amendment provision as well. This can easily be achieved by cutting off the ability to amend the poison pill once it is triggered.

A second possible end run involves the inadvertent trigger exception. The definition of “Acquiring Person” in pills typically contains some form of a carve-out for the scenario which a person inadvertently triggered the pill (thereby inadvertently becoming an “Acquiring Person”). There is significant variation in the wording of these carve-outs, with respect to both the discretion of the board and the obligations of a shareholder that supposedly inadvertently triggered the pill. If the wording of this carve-out does not sufficiently limit the board’s options, there is a risk that this serves as a loophole that allows a board to, or to be pressured to, back out from the threatened dilution by declaring that an acquiror only “inadvertently” triggered the pill. Hence, the respective language should provide that the board can declare this carve-out applicable only if (i) the board in good faith determines that the person inadvertently triggered the poison pill (i.e., no intention of acquiring or exerting control); and (ii) such person has already divested or divests as promptly as practicable a sufficient number of shares to no longer qualify as an “Acquiring Person.” Again, as with the amendment provision, a trip-wire concept is effective only if the irrevocable commitment to the dilution is reflected in the inadvertent trigger exception as well.

3. Identifying Last-look Design

By contrast, the typical wording of a redemption mechanism of a poison pill implementing a “last look” is as follows:

The Board may, at its option, at any time prior to the earlier of (i) the Close of Business on the tenth (10th) Business Day following the Shares Acquisition Date (or, if the tenth (10th) Business Day following the Shares Acquisition Date occurs before the Record Date, the Close of Business on the Record Date) and (ii) the Final Expiration Date (the “Redemption Period”), direct the Company to, and if directed the Company shall, redeem all but not less than all of the then outstanding Rights at a redemption price of $0.001 per Right, as such amount may be appropriately adjusted to reflect any stock split, stock dividend, or similar transaction occurring after the first public announcement by the Company of the adoption of this Agreement (such redemption price, as adjusted, being hereinafter referred to as the “Redemption Price”) . . . .

Pursuant to this provision, the board’s right to redeem the rights issued under the shareholder rights plan terminates only 10 business days after the date on which a shareholder publicly announces that it has triggered the poison pill. This has the desired effect of granting the board a “last look” and the ability to avoid the dilution resulting from the triggering of the poison pill by redeeming it. As described above, pills that permit an end run around the trip wire—through either the amendment provision or the inadvertent trigger exception—are also coded as last-look pills.

C. Overall Results

Across the overall Pill Sample, 70 out of 130 (or 53.8 percent) followed a trip-wire concept, while 60 (or 46.2 percent) implemented a last-look feature. Figure 1 shows the distribution of trip-wire versus last-look design by the year the pill was installed:

Figure 1 shows no significant time trend by year, indicating little to no “learning” going on within the sample period.

There is considerable variance in the design of this “last look” that the board of directors may have. This finding is not surprising, given that the very concept of the last look is precisely to avoid the hard bifurcation intended by a trip-wire provision. The standard formulation provides a last look for ten business days between the date on which the poison pill was triggered and the date as of which the rights become exercisable. Three specific structuring alternatives could be observed:

• (1) Very short “last look”: Two shareholder rights plans provided for a very short last look until the close of business on the date the poison pill was triggered by a shareholder reaching or exceeding the triggering threshold.
• (2) “Last look” shorter than period until “Distribution Date”: One shareholder rights plan provided for a last look only during the first five days of the ten days until the “Distribution Date” (i.e., the date on which rights become separated from the shares and exercisable).
• (3) “Last look” even after “Distribution Date”: Two shareholder rights plans provided for a last look not only during the period until the “Distribution Date,” but until the fifth day after the “Distribution Date.”

While the reasons for designing a last look according to alternatives (1) and (2) are perhaps understandable, structuring alternative (3) is more puzzling. Affording the board with a right to redeem for the first five days during which such rights are exercisable by shareholders results in an overlap of exercisability and redeemability of the rights. The idea behind this structuring alternative might be to incentivize shareholders to exercise their rights as quickly as possible, given that the board could still decide to redeem the rights—that will, in this case, be deeply in the money—and thereby severely limit the pay-off of such shareholders. However, such a threat does not seem necessary. All shareholder rights plans—regardless of whether they implement a trip-wire or last-look concept—provide that the board of directors may, at any time after the trigger event, exchange the rights outstanding for common shares at a ratio of one-for-one. From an economic perspective of an individual shareholder, a one-for-one exchange is always worse than the exercise of the rights that are deeply-in-the-money. Therefore, there is already an economic incentive to exercise the rights quickly, which makes the overlap of exercisability and redeemability of the rights difficult to explain. In addition, such overlap also entails the risk of unequal treatment of shareholders should the board decide, or feel pressured to decide, to redeem the rights after some shareholder already exercised their rights vis-à-vis the rights agent.

Finally, two shareholder rights plans—both drafted by Sidley Austin LLP—de facto eliminated the exercisability of the rights under the flip-in option by way of the last-look provision: Both of these pills included a clause in the redemption provision that prevented the shareholders from exercising their rights with respect to the flip-in option until the expiration of the rights plan, thereby effectively extending the last look until the expiration date. A trigger event under either of these rights plans has the effect that the board can decide, until the expiration of the plan (which could be as long as ten years later), to redeem the pill or exercise the exchange option. No shareholder has the right to exercise the rights even after a trigger event, which then puts pressure on the board to quickly decide whether to redeem the poison pill or exercise the exchange option. The lack of commitment to the threat of dilution inherent in this structuring of the redemption mechanism is detrimental to the intended effect of the poison pill and should be avoided.

D. Results by Law Firm

We now examine the law firms that installed the pills in the Pill Sample. Because adopting a pill is a significant M&A event, virtually all companies rely heavily on outside counsel, typically with M&A expertise, to provide the pill documents. However, the company does not need to disclose the name of the law firm that put in the pill. Discussions with practitioners indicate that many of the top tier law firms will not want to be named in the pill documents. The reason is that these law firms consider their boilerplate pill documents to be part of their intellectual property. If other firms could readily identify the pill documents of a top-tier law firm, they could free ride off the top-tier firm’s expertise in pill design.

Consistent with this practical reality of transactional practice, we find that less than half (58 of 130) of the shareholder rights plans in the data set identified the outside legal adviser in the notice section of the rights plan. Interestingly, last-look pills disclosed such legal adviser in 60 percent of cases (36 out of 60) while trip-wire pills disclosed such legal adviser in only 31 percent of cases (22 out of 70). We return to this point in our examination of law firm effects below.

In order to identify the legal adviser for as many pills as possible, we examined press releases published by the company in connection with the introduction of the relevant pill (or the overall transaction that the pill was a part of) and other SEC filings. Based on this examination, we were able to determine, with reasonable certainty, the legal adviser in an additional fifteen rights plans, which did not identify such legal adviser in the notice section of the relevant agreement.

Most law firms seem to have a clear “house view” on structuring the redemption mechanism, as they generally follow either the trip-wire or the last-look concept. Only a few law firms implement both of these structuring alternatives in their poison pills:

• Only trip-wire provisions (number of poison pills in parenthesis): Arnold & Porter Kaye Scholer LLP (1), Cadwalader, Wickersham & Taft LLP (1), Cleary Gottlieb Steen & Hamilton LLP (1), Davis Polk & Wardwell LLP (2), Faegre Drinker Biddle & Reath LLP [together with Sidley Austin LLP] (1), Gibson, Dunn & Crutcher LLP (1), Goodwin Procter LLP (1), Gracin & Marlow, LLP [subsequently joined forces with Blank Rome LLP] (1), Kirkland & Ellis LLP (6), Paul, Weiss, Rifkind, Wharton & Garrison LLP (1), Ropes & Gray LLP (1), Stearns Weaver Miller Weissler Alhadeff & Sitterson, P.A. (2), Wachtell, Lipton, Rosen & Katz (5), Winston & Strawn LLP (1), and Winthrop & Weinstine, P.A. (1).
• Only last-look provisions (number of poison pills in parenthesis): Baker Botts LLP (1), Dechert LLP (1), Harter Secrest & Emery LLP (1), Morgan, Lewis & Bockius LLP (7), Proskauer Rose LLP (4), Shearman & Sterling LLP (2), Sichenzia Ross Ference LLP (1), Skadden, Arps, Slate, Meagher & Flom LLP (1), Thompson & Knight LLP (1), Vinson & Elkins L.L.P. (7), Willkie Farr & Gallagher LLP (1), Wilson Sonsini Goodrich & Rosati (5).
• Mix (number of poison pills with trip-wire provisions and last-look provisions in parenthesis): Haynes and Boone, LLP (2 trip-wire, 1 last-look), Mintz, Levin, Cohn, Ferris, Glovsky and Popeo, P.C. (1 trip-wire, 1 last-look), Morrison & Foerster LLP (1 trip-wire [D.C. office], 2 last-look [San Francisco office]), Sidley Austin LLP (2 trip-wire, 5 last-look).

Among the “mix” firms, Morrison & Forester is particularly noteworthy, because, as described in Part II.B.1, it is only firm that has published commentary on the pros and cons of trip-wire versus last-look pills. We find that “MoFo” put in one trip-wire pill (D.C. office) and two last-look pills (San Francisco office). Coates (2001) similarly finds a difference between east coast law firms and west coast law firms, with east coast firms (he argues) adopting better takeover defenses at the IPO stage.

Putting aside the few firms, such as MoFo, that take a mixed approach, it is perhaps unsurprising that most law firms have a house view on pill design, in view of the ubiquitous nature of boilerplate documents in corporate law firms. In addition, and unlike many other kinds of transactional documents, pills often have to be implemented quickly, which creates an even greater need to rely on “off the shelf” documents. The data strongly reflects this dynamic. However, the existence of a strong house view on pill design should cast doubt on arguments by practitioners that pills are designed to reflect the nuances of the particular bidder or target company. Our evidence indicates that they clearly are not.

We now separate the law firms according to M&A expertise. Specifically, we divided the sample according to whether the law firm was included by Chambers as Band 1, 2, or 3 in Corporate/M&A. Kirkland & Ellis LLP (Band 1) and Paul, Weiss, Rifkind, Wharton & Garrison LLP (Band 2) could be identified from the notice sections of shareholder rights plans in the Pill Sample. Based on the analysis of further disclosures, however, Band 1-ranked law firms Davis Polk & Wardwell LLP, Skadden, Arps, Slate, Meagher & Flom LLP, and Wachtell, Lipton, Rosen & Katz as well as Band 3-ranked law firms Cleary Gottlieb Steen & Hamilton LLP and Gibson, Dunn & Crutcher LLP could be matched with pills in the pill sample. This evidence is broadly consistent with the conclusion that top-tier law firms are less willing to be readily identified in the pill documents, potentially because non-disclosure helps them protect their intellectual property in pill design.

Figure 2 divides the sample according to the top tier M&A firms and all other firms:

Figure 2 shows a pronounced law firm effect: 94 percent of pills put in by top-tier M&A firms have a trip-wire provision, while only 29 percent of pills put in by other law firms have a trip wire. When the law firm is unknown, a trip-wire feature is used in 67 percent of pills. For reasons described above, we believe that top-tier firms are over-represented in the sample of pills where the law firm is unknown; and top-tier firms are far more likely to put in trip-wire pills.

One theory that might explain this stark difference between top-tier M&A firms and other law firms is that target companies engaging counsel might not consider the implementation of poison pills as an M&A transaction—as, in most cases, they precisely want to prevent an M&A deal. As a result, the target’s selection of the law firm advising on the implementation of a poison pill might very well be based on other factors: Given that law firms advising a company on an IPO typically provide such company’s board of directors with a handbook on the key rules that will apply to such company once it becomes public—and might also conduct a “teach in” session to the board of directors on these topics—which would also include takeover defenses such as poison pills, it seems reasonable to assume that a company might just reach out to its IPO counsel (in particular, if only little time has passed since the IPO). Alternatively, a company might just reach out to its standing outside counsel for general corporate matters and ask them to quickly implement a poison pill.

There is a statistically significant (at 95 percent confidence) difference in the size of the companies that are advised by top-tier M&A firms versus other firms in our sample. The mean (median) market capitalization of companies advised by top-tier firms is $3.9 billion ($1.9 billion), and the mean (median) market capitalization of companies advised by other firms is $1.4 billion ($0.2 billion). This makes the difference documented in Figure 2 even more significant, as a practical matter, because the consequences of triggering the pill dilution at a large-cap firm would generally be more severe than triggering the pill dilution at a smaller-cap firm. In effect, the top-tier M&A firms are more willing to play “chicken” with potential acquirors or activists even though the stakes are much higher from doing so.

IV. Discussion

At the highest level, our evidence indicates that whether a poison pill of a company implements a trip-wire or a last-look feature in the redemption provision of a poison pill seems to depend not on the specific circumstances of the company and/or the takeover that the poison pill is protecting against, but rather whether the law firm selected by such company includes a trip-wire or a last-look provision in their pill boilerplate documents. Moreover, we find that top tier M&A firms are far more likely to have a trip wire “house view,” while most other firms have a “last look” house view.

Negotiation theory supports the trip-wire approach to pills, and (by extension) the decision by top-tier firms for putting in trip-wire pills. Schelling (1958) explains:

The distinctive character of a threat is that one asserts that he will do, in a contingency, what he would manifestly prefer not to do if the contingency occurred. The motive behind the threat is to coerce or to deter, to constrain the other player’s choice of action. It works by altering the other player’s expectation of how the threatening player would react; it is an attempt to alter the threatener’s own incentives as they are seen by the party threatened.

This lack of motivation to follow through is critical, especially if the threatened action has serious consequences: If the other party disregards the threat and takes the action that such threat was supposed to deter, there is likely a significant discrepancy between the incentive that the threatening party tried to convey and its actual incentive structure at this stage. Suddenly, the key factor in this “game” is not who made the threat but rather who has the last chance to prevent the threatened action from occurring, as such party has significant incentives to pull back. As Schelling points out:

In case a threat is made and fails to deter, there is a second stage prior to fulfillment in which both parties have an interest in undoing the commitment. The purpose of the threat is gone, its deterrence value is zero, and only the commitment exists to motivate fulfillment.

The negotiation literature, then, clearly prefers the trip-wire concept so as to not allow for a last look of the target’s board. Only a trip-wire provision allows the board to irrevocably commit to the threat of dilution made by implementing the poison pill. This puts the onus for triggering the poison pill entirely on the acquiring person. Given that issuers have to file their shareholder rights plans with the SEC (and the plans are therefore public), the message of such a move to the capital market is very clear: “In accordance with our fiduciary duties, we want to protect the interests of the shareholders and will thus only allow a takeover in a negotiated transaction. We will not hold any negotiations at gunpoint after the poison pill has been triggered and while the clock for redemption thereof is ticking. Any person who disregards this will automatically be diluted if they reach or exceed the triggering threshold; there is nothing we can do to stop it. You have been warned.”

In response to this clear prescription from the negotiation literature, some practitioners have suggested to us that the Delaware courts require a last look in order for a target board to comply with its fiduciary duties. We believe that this argument is incorrect. In the Delaware Supreme Court’s seminal opinion in Moran v. Household International, Inc., which endorsed the poison pill, the Court held:

[T]he Rights Plan is not absolute. When the Household Board of Directors is faced with a tender offer and a request to redeem the Rights, they will not be able to arbitrarily reject the offer. They will be held to the same fiduciary standards any other board of directors would be held to in originally approving the Rights Plan . . . .

The Board does not now have unfettered discretion in refusing to redeem the Rights. The Board has no more discretion in refusing to redeem the Rights than it does in enacting any defensive mechanism.

However, nothing in the Moran opinion indicates that the board’s fiduciary duties require that the decision to redeem the pill must only occur after the pill has been triggered. Rather, the Moran court specifically addressed the circumstance that the board is presented with a request to redeem a poison pill to allow the tender offer to succeed without triggering it. Implicitly, the Moran court thus confirmed that there is no legal requirement to have a “last look” in order to comply with fiduciary duties.

To the extent the point is not clear from the opinion itself, the factual description in the opinion makes clear that the rights plan at issue in Moran itself contained a trip-wire provision. Given that the Moran court approved a poison pill with a trip-wire provision, which further restricts the board’s right to redeem the poison pill than under a last-look provision, both structuring alternatives are compliant with the board’s fiduciary duties. And, for reasons described above, the trip-wire structure is more effective in achieving the pill’s intended effect—namely, deterrence—than the last-look structure.

The last-look pill also has a potential defect for target companies that permit action by written consent and have a unitary board. For these companies, the bidder could possibly bust through the pill, replace the board by written consent, and redeem the pill before the ten-day window expired. According to one practitioner: “Being in the group of firms that do not recommend the window and always use the tripwire, it wasn’t a material concern for us, but for others [this possibility] was certainly a discussion point.”

We close this part with an anecdote. One practitioner—who was the co-leader of the M&A practice at one of the top tier firms in our sample—observed to us that many years ago he was the keeper of his firm’s boilerplate pill documents. Approximately twenty years ago, in the early 2000s, he and his colleagues realized the benefits of the trip-wire feature over the last-look feature, and amended their boilerplate documents accordingly. His impression was that most New York City firms made the same adjustment around the same time. However, in a footnote to his firm’s boilerplate document, he noted the possibility of a “last look” feature as a possible variant that might be beneficial in certain situations. A few years later, unable to think of any such situation, he deleted the footnote.

V. Application to Twitter/Musk

The Twitter poison pill would have imposed approximately $2.4 billion in dilution against Elon Musk if the Twitter board had used the flip-in feature, and $2.5 billion in dilution if the board used the exchange feature. Figure 3 provides the details of this analysis:

Figure 3 Twitter “Pill Math”

As noted in the Introduction, the Twitter pill had a last-look feature, which gave the Twitter board ten business days to consider whether to redeem the pill before the dilution would occur. Wilson Sonsini Goodrich & Rosati (“Wilson Sonsini”) installed Twitter’s pill, and Wilson Sonsini is not included, by Chambers, in Band 1–3 in Corporate/M&A. Therefore, the evidence presented in Part IV. indicates that Wilson Sonsini’s pill reflected overall market practices: 71 percent of pills put in by firms that are not the top tier firms were last-look pills.

Consider, counterfactually, the possibility that Musk deliberately triggered the Twitter pill. His downside risk is $2.5 billion. The Twitter board’s risk in triggering the pill would be either $136 billion of cash on the balance sheet (if they used the flip-in feature) or diluting him down only to 8.1 percent (if they used the exchange feature). In addition, Twitter did not have a sufficient headroom of authorized shares to fully source the flip-in feature with common shares, which might have limited its options. Irrespective of which option they would chose, the Twitter board would of course have fiduciary duties to all shareholders, including Musk, during that ten-business-day period of decision-making. Our prediction is that Twitter’s board would cave and negotiate a deal, though not from a position of strength because a ten-business-day clock would be ticking. Musk could put even more pressure on the Twitter board by crossing the 15 percent trigger threshold, and offering shareholders an additional $5.0 billion in total consideration (i.e., twice what they would get from the dilution) if the Twitter board eliminated the pill.

Knowing all this in advance, and willing to take risks, Musk is more likely to trigger the pill; in effect, daring Twitter’s board to allow the dilution to occur. His worst case is a $2.5 billion loss and the (more likely) best case is a negotiated deal from a position of strength. The decision tree analysis points very much in his favor.

The unaffected market capitalization of Twitter was approximately $30 billion in 2022, and an additional $5 billion would put the total purchase price at approximately $35 billion. Instead, Musk acquired Twitter for approximately $44 billion in total value. Our hypothetical of a deliberate pill trigger could have potentially saved Musk $9 billion in his ultimate acquisition of the company, albeit at higher risk than a conventional M&A deal.

During the pendency of the deal, a senior partner from Wilson Sonsini justified this last-look provision as follows:

It should be up to the board to delay the triggering. Among other things, there may be situations where circumstances change and the board decides that implementing the pill is not the right choice. It creates a last safety stop before doomsday where the board can step in. And the pill still works if the board does nothing. I don’t think it is a defect for a board to have flexibility, particularly in a volatile and unpredictable situation. Any such flexibility remains consistent with a principal purpose of a pill, which is to put the board back in control.

We believe this reasoning is incorrect, both generally and with respect to the specific situation of Twitter. The last “safety stop” is a fiduciary moment for the board, where they must consider all shareholders, including Elon Musk. While the decision to put in a pill affects all shareholders equally, at least formalistically (because all shareholders receive the same rights), once a pill has been triggered any decision to allow the pill dilution to occur knowingly preferences certain shareholders over others. This decision would almost certainly be litigated by a triggering shareholder.

By way of analogy, sell-side transactional planners knowingly tie hands—both buyers’ and sellers’—in auctions, through the use of “don’t ask, don’t waive” (“DADW”) standstill provisions. Consider a standard auction process where bidders sign standstill agreements in order to gain access to due diligence and make an offer for the company. In the absence of a DADW provision, a losing bidder could approach the board between signing and closing with an offer to pay more than the ostensible winner from the auction. The sell-side board would then have a fiduciary obligation to accept this higher offer. Foreseeing all of this in advance, no bidder would put their best offer on the table in the initial auction, because they would all know that they would have another opportunity to bid once the winner was announced. In order to induce bidders to place full value on the table in the auction itself, DADW provisions prohibit bidders from asking for a waiver from the standstill (“don’t ask”) and further prohibiting the target board from considering such a request (“don’t waive”) after the auction is over.

Negotiation theory indicates that the hands-tying feature of DADW provisions forces bidders to put their best offer on the table in the initial auction, knowing that they will not get an opportunity to bid again. Trip-wire pills take the same hands-tying approach: tying hands in the ex post scenario (inability to redeem the pill; inability to grant a standstill waiver) induces the intended action ex ante (negotiate with the board; put full value on the table in the auction itself). In contrast, last-look poison pills, like the Twitter pill, take the exact opposite approach of inviting the ex post fiduciary moment; and as a result they completely miss the beneficial effects of hands-tying in advance.

The more specific point relates to Elon Musk. Recall that the point of the Twitter pill was to force Musk to negotiate with the Twitter board. By putting in a last-look pill, the Twitter board sent an exceedingly weak signal to Musk that they would actually allow the dilution to occur. That is, by “creat[ing] a last safety stop before doomsday where the board can step in,” it is almost certain that the board would have taken the safety stop. In effect the Twitter board sent a lukewarm signal to Musk: “If you cross 15%, we will think for 10 business days about what happens next.” And in the ten-day window before financial Armageddon, between the risk-loving Musk and the risk-averse Twitter board, it is not difficult to predict who would win. As one practitioner pointed out to this, it would have been very much “on brand” for Musk to trigger the Twitter pill, effectively daring the Twitter board to allow the dilution to occur. Finally, while halting trading for a month might not a problem at a small-cap company like Selectica, it would be unacceptable at a mid- or large-cap company like Twitter.

VI. Conclusion

Based on an analysis of negotiation theory and legal doctrine, this article shows that structuring the redemption mechanism of a poison pill by following the trip-wire concept has clear benefits compared to last-look provisions. A trip-wire provision enhances the leverage and bargaining power of the target board due to the clear, credible and irrevocable commitment to the threatened dilution of a potential acquiror. In contrast, the last-look provision in a pill weakens the target board’s hand, as it exposes the board to pressure tactics by both a bidder and the target’s shareholders. This conclusion is supported by the analysis of the Twitter/Musk case and the dynamics that could have resulted from a deliberate triggering of the pill by Elon Musk.

However, the examination of 130 poison pills implemented and/or amended between January 1, 2020, and March 31, 2023, does not square with our theoretical and doctrinal analysis. Rather than observing a dominance of trip-wire provisions, or even a trend toward trip-wires over time, the data reveals a stable 54 percent/46 percent-split between trip-wire and last-look features. However, when we cut the data by law firm expertise, we find that 94 percent of “top tier” M&A firms (defined as ranked Band 1, 2 or 3 in Corporate/M&A by Chambers) use the trip-wire concept. These law firm effects are consistent with Coates (2001) and Subramanian (2005), which similarly find slow or non-existent dissemination of best-practice corporate law tools from the top-tier firms to other law firms in the area of corporate law.

The fact that virtually all law firms have a “house view” on pills, without tailoring the pill to the specific context, seems troubling. A shareholder rights plan, like any other complex transaction agreement, requires careful drafting and should not be treated as a cut-and-paste boilerplate instrument. Rather than simply trusting the template (which might understandably be tempting, as pills are often implemented under time pressure) pill drafters should be mindful of the underlying economic concerns and the implications of the structuring alternatives on the negotiation dynamics between the target and the unsolicited shareholder. In view of our theoretical and doctrinal analysis, the trip-wire feature should be the default option, in order to ensure the maximum effectiveness and intended deterrent effect of the pill.

Finally, given that—due to the observed “law firm effect”—the target company, by the selection of its legal adviser, de facto (pre)determines whether its poison pill will include a trip-wire or a last-look provision, the results of our empirical analysis highlight the importance of retaining a law firm specializing in Corporate/M&A and takeover defense, rather than simply relying on the general outside counsel that is typically entrusted with a broad range of matters. Put simply, expertise matters.

An earlier version of this article, titled “Redemption Mechanism in Poison Pills: A Study with Special Consideration of Twitter’s Poison Pill” (solo authored by Baum), won the 2023 Victor Brudney Prize, awarded by Harvard Law School’s Program on Corporate Governance annually to the best student paper on a topic related to corporate governance.

We thank the participants in the Harvard Law School corporate law working group for helpful comments and discussion.