IN RE: REVLIMID & THALOMID PURCHASER ANTITRUST LITIGATION (2026)

OPINION and ORDER

Table of Contents

I. Overview

II. Background

A. The Law

1. Drug Development

2. Paragraph IV Lawsuits

3. Paragraph IV Settlements

4. Antitrust Limits on Paragraph IV Settlements

5. Key Aspects of Disallowed Settlements

B. This Case

C. The Natco Settlement

1. The Litigation

2. The Settlement Terms and the Plaintiffs’ Theory

a) Delay

b) Pay

c) Facilitating the Pay

D. Procedural History

III. The Court's Approach

IV. The Elements

A. Delay, Justification, Size, Pay

B. A Fifth Element?

V. Delay and Justification

VI. Size

VII. Pay: Overview

VIII. Pay: Higher Per-Unit Prices

A. Method

B. Economics: Higher Prices

C. Law: Higher Prices Count

1. The Basis for the Doctrine

a) Cost to Consumers

b) Cost to the Patentholder

2. King Drug

D. Counterargument

IX. Pay: Would the Market Structure Stick?

A. An Authorized Generic?

1. Background

2. Incentives

3. The Acceleration Provision

B. Other Drugmakers?

1. Incentives

2. The 1.5x Provision

3. The “Genericization” Counterargument

4. Later-Filers

5. The Assurance Counterargument

X. Conclusion and a Loose End

* * *

Certain pharmaceutical companies were engaged in patent-infringement lawsuits against each other about a particular drug.

The suits settled, and particular hospitals and insurance companies came to believe that the settlements were collusive --- a way to keep prices high for the drug.

So the hospitals and insurance companies (along with some other entities) sued the pharmaceutical companies, claiming violations of the antitrust laws.

The pharmaceutical companies have now moved to dismiss.

As to the parts that are taken up here, the motion is denied.

* * *

I. Overview

This case is about Revlimid. Revlimid is a prescription drug used to treat certain cancers. See Amended Complaint (“Complaint”) (ECF 460) ¶ 2.¹

Celgene Corporation ² invented Revlimid.³ See id.; see also ¶¶ 169-70. Celgene sold Revlimid. See id. ¶¶ 176-77. And Celgene owned the key Revlimid patents. See id. ¶ 180.

Because of Celgene's patents, Revlimid was for a while the only drug of its kind on the market. Celgene had a monopoly. And during that stretch, Celgene could charge high, monopoly prices for Revlimid.⁴ See id. ¶¶ 78, 79, 82.

But a patent monopoly is not forever. It eventually times out.⁵ And when it does, other drugs can come online --- equivalents of the brand-name drug, called generics.

Generics are sold under different names than the original drug. But they do the same work.⁶ And they typically cost much less.⁷

But competition from a generic does not always need to wait until the brand-name drug's underlying patents run their course, and expire on their own.

Competitors can try to speed things along. They can, for example, try to get into the market before the brand-name drug's patent-expiration date --- by arguing that the patents are invalid, and should not have been given in the first place.⁸ Or that the generic would not infringe on those patents.⁹

And that is what happened here.

Potential generic competitors challenged Revlimid. See id. ¶¶ 199, 352, 365, 460, 472, 490, 503, 521, 533, 549. Litigation followed. See id. ¶¶ 201, 352, 365, 460, 472, 490, 503, 521, 533, 549. And then, each time, the litigation wrapped up in a settlement. See id. ¶¶ 284, 407, 461, 479, 493, 510, 522, 537, 550.

The question in this case is whether those settlements violated the federal and state antitrust laws.¹⁰

The settlements, it is claimed, were collusive.

They allegedly helped Celgene. It got to keep Revlimid on the market, and for long intervals out there by itself --- on a patent-monopoly basis, with no generic competition. See id. ¶¶ 4-5, 8.

And the settlements allegedly helped the generic drugmakers, by allowing them to split the market with Celgene --- which in turn gave the generic drugmakers the ability to sell generic Revlimid at especially high prices. See id. ¶¶ 5, 308. The settlements, it is said, pushed off into the future the moment when cheap generics --- generics priced the way generics usually are --- would come onto the market. See id. ¶¶ 314, 320.

* * *

In light of this, various entities sued the drugmakers.

Entities that bought Revlimid for doling out to others. Like hospitals that needed Revlimid for their patients. See Consolidated First Amended Complaint for Damages and Injunctive Relief (“Hospitals’ Complaint”) (ECF 462) ¶¶ 25-31.

And entities that had to foot the bill for others’ use of Revlimid. Insurance companies, for example, that were on the hook for Revlimid costs when a person they covered was prescribed the drug. See Complaint ¶¶ 11–29; Second Amended Complaint (“United Healthcare's Complaint”) (ECF 463) ¶¶ 11–16; Second Amended Class Action Complaint (“Plan and Trust Complaint”) (ECF 461) ¶¶ 11–14.

These entities’ basic argument: they had to bear extra costs because the drugmakers’ settlements kept Revlimid prices especially high.

Does this add up to a plausible ¹¹ antitrust claim?

That is the question this Opinion and Order works through.

To begin, move through some background first. On the law. See Part II.A. This case. See Part II.B. The Plaintiffs’ legal theory. See Part II.C. And the procedural history here. See Part II.D.

Then get going with the analysis. See Part III to Part IX. That ends on this bottom-line conclusion: as to those aspects of their claims that are taken up here, the Plaintiffs have made out plausible antitrust claims, and the Defendants’ motion to dismiss must therefore be denied. See Part X.

II. Background

A. The Law

1. Drug Development

Take as the stepping-off point a hypothetical pharmaceutical company.

Say that it sees an opportunity. An untreated medical problem, for example. The company is looking to close the gap, by developing a drug that will tackle the problem.

Researching and developing the drug will almost surely be expensive, if it works out at all.

And if the drug comes together, it will likely involve multiple underlying patents.¹²

For the patents it needs, the drug company will have to put in an application to the United States Patent and Trademark Office. See Cuozzo Speed Techs., LLC v. Lee, 579 U.S. 261, 266, 136 S.Ct. 2131, 195 L.Ed.2d 423 (2016). A patent examiner will then review the applications, and decide on the company's patent claims. See id.

Assume success. Assume that the pharmaceutical company gets all of the patents it needs.

Next up, the company will have to go to the FDA ¹³ for approval to sell the drug to people who need it. See 21 U.S.C. § 355(a). To get the ball rolling with the FDA, the company will have to send the agency an application --- and that requires laying out “the ‘number and expiration date’ of any relevant patent.” Fed. Trade Comm'n v. Actavis, Inc., 570 U.S. 136, 143, 133 S.Ct. 2223, 186 L.Ed.2d 343 (2013) (quoting 21 U.S.C. § 355(b)(1)).

Assume more success. The FDA approves the application, and the drug company gets underway --- selling the new drug it has worked to develop, likely under a brand name.

2. Paragraph IV Lawsuits

When it comes to drug development, the above is the basic set-up.

Imagine now that another pharmaceutical company is looking to come into the picture. It wants to make and market its own version of the new drug, but under a different name. As noted, that sort of drug is called a generic, and it would almost certainly be much cheaper than the equivalent brand-name drug, the one that is already out there.¹⁴

To break into the market with its generic, the second drugmaker will also have to go to the FDA for permission. See Bristol-Myers Squibb Co. v. Royce Lab'ys., Inc., 69 F.3d 1130, 1131 (Fed. Cir. 1995).

But the FDA cannot approve generic drugs that might infringe on someone else's valid, existing patents. See Caraco Pharm. Lab'ys., Ltd. v. Novo Nordisk A/S, 566 U.S. 399, 405, 132 S.Ct. 1670, 182 L.Ed.2d 678 (2012).

So it is over to the second drugmaker --- to assure the FDA that it (the FDA) can give a thumb-up to the proposed generic without violating the first drugmaker's valid patents.

How to make that assurance?

The second drugmaker can file with the FDA a “paragraph IV certification.” See id. at 407, 132 S.Ct. 1670.

The certification can say that the generic would not actually infringe on the first company's underlying patents. Or it can agree that the first company's underlying patents would be infringed by the generic drug --- but no matter, because those patents are invalid anyway. See 21 U.S.C. § 355(j)(2)(A)(vii)(IV); see also Caraco Pharm., 566 U.S. at 407, 132 S.Ct. 1670.

A paragraph IV certification will often kick up a sharp protest. Our patents are sound, the first drugmaker will typically argue --- and any new proposed drug is just an illegal copycat, an effort to use our R&D, without having to actually do the work or spend the money or take the risk.

All of this makes for pointed disagreement. And the patent laws lay out a way to get it resolved.

* * *

When a paragraph IV certification is filed, the first drugmaker can go to federal court and sue, claiming patent infringement. See Caraco Pharm., 566 U.S. at 407, 132 S.Ct. 1670; King Drug Co. of Florence v. Smithkline Beecham Corp., 791 F.3d 388, 395 (3d Cir. 2015).¹⁵

The litigation will usually be high stakes. See Watson Lab'ys., Inc. v. Forest Lab'ys., Inc., 101 F.4th 223, 229 (2d Cir. 2024); Kaiser Indus. Corp. v. Jones & Laughlin Steel Corp., 515 F.2d 964, 971 (3d Cir. 1975).

If the courts determine that the first drugmaker's patents are valid and being infringed, that company will get to hold onto the monopoly that the patent laws gave it.¹⁶ Monopoly prices will continue. Profits will stay high.

On the flipside, if the courts conclude that the first drugmaker's patents are not valid, then those patents will be undone --- and with them the patent-based monopoly the first drugmaker had. Generic competition will get underway. Prices will fall. The first drugmaker's souped-up monopoly profits will start to evaporate, and quickly.

3. Paragraph IV Settlements

Paragraph IV cases are like other lawsuits. They can be fought out to the finish. Or they can be settled. Each possibility is available.

But the balance does not split neatly down the middle, half and half. The overwhelming norm in our legal system is settlement.

In this case, for example, the settlement that is the primary focus here was signed in December 2015. See Complaint ¶ 284. Of the civil cases wrapped up in federal court around then, the great majority settled. Only about 1% ended after a trial.¹⁷

Settlement is not just a background fact, the way things tend to go. Rather, settlement is generally taken as an affirmative plus. Our law “favors settlement, particularly ․ in complex cases[.]” In re Gen. Motors Corp. Pick-Up Truck Fuel Tank Prods. Liab. Litig., 55 F.3d 768, 784 (3d Cir. 1995).¹⁸

And there is no exception to any of that here. As to paragraph IV lawsuits, the Supreme Court has “recognize[d] the value of settlements.” Actavis, 570 U.S. at 153, 133 S.Ct. 2223.

4. Antitrust Limits on Paragraph IV Settlements

Settlement is the most commonly-used path, but the Supreme Court has laid down a guardrail. Namely, settlements of paragraph IV lawsuits “can sometimes violate the antitrust laws.” Actavis, 570 U.S. at 149, 133 S.Ct. 2223.

Sketch out here some of the dividing line between what is allowed and what is not.

a. Settlements That Are Always Allowed

One category of settlements is always permitted under the antitrust laws.

To resolve a paragraph IV lawsuit, the settling drugmakers can simply agree to early market entry for the generic drugmaker --- and leave it at that, nothing more. That sort of settlement does not break the antitrust laws. See Actavis, 570 U.S. at 158, 133 S.Ct. 2223; King Drug, 791 F.3d at 405 n.23; Fed. Trade Comm'n v. AbbVie Inc., 976 F.3d 327, 359 (3d Cir. 2020); see also Mayor of Balt. v. AbbVie Inc., 42 F.4th 709, 716 (7th Cir. 2022).¹⁹

To see what this means, think of a paragraph IV litigation between a brand-name drugmaker (like Celgene) and a generic drugmaker. And say that the brand-name drugmaker's patents on the relevant drug were due to run out in 2040 --- at which point the generic drugmaker could start selling its own version of the drug.

The brand-name drugmaker and the generic drugmaker can, in 2030, enter into a settlement that gets the generic firm onto the market as of 2035, five years ahead of schedule.

This can be a real upside for the generic drugmaker --- five extra years of sales. And that can convince it to wrap things up, to settle the paragraph IV lawsuit.

If early entry (2035 versus 2040) is the whole story --- if early entry is all that the generic drugmaker gets --- then the settlement cannot violate the federal antitrust laws. See Actavis, 570 U.S. at 158, 133 S.Ct. 2223; see also King Drug, 791 F.3d at 405 n.23.

b. Settlements That May Not Be Allowed

Move away now from early-entry-alone settlements. Outside of that safe harbor, settlements of paragraph IV lawsuits can sometimes violate the antitrust laws.

Take a settlement that provides for “an early-entry date plus valuable consideration.” King Drug, 791 F.3d at 405 n.23.

To see this, start with the same imagined facts as before. But now, the generic drugmaker not only gets to come in five years early (2035 versus 2040), but also gets an added dollop of “valuable consideration,” King Drug, 791 F.3d at 405 n.23, like a big cash payment. At that point, there might be an antitrust issue, on account of the cash “plus.”

Early entry alone is categorially fine. But fold some extra sweetener into the mix --- and that can sometimes break the antitrust laws. See King Drug, 791 F.3d at 405 n.23.

* * *

In an early-entry-plus settlement, what is the possible antitrust concern?

With the added transfer of value (the “plus”), the brand-name drugmaker might be paying for “the exclusive right to sell its product [during the patent period], a right it already claims [on account of its patent] but would lose if the patent litigation were to continue and the patent were held invalid or not infringed by the generic product.” Actavis, 570 U.S. at 153-54, 133 S.Ct. 2223.

And this may hurt customers through higher prices.²⁰ As the Supreme Court has explained, if the brand-name drugmaker's “exclusive right to sell [during the patent period] produces $50 million in supracompetitive profits per year,” and there are 10 years left on the clock before patent expiration, paragraph IV litigation that “results in patent invalidation ․ could cost the [brand-name drugmaker] $500 million in lost revenues, a sum that then would flow in large part to consumers in the form of lower prices.” Id. at 154, 133 S.Ct. 2223.

But the parties to the paragraph IV litigation can work together to put that possibility out of play.

Instead of risking a courtroom loss --- and what follows from it, an implicit transfer of $500 million to consumers “in the form of lower prices” --- the brand-name drugmaker can opt to split its “supracompetitive” profits with the generic competitor it is litigating against. Id. How? By cutting a settlement deal. That avoids the court-loss possibility. And it benefits the brand-name drugmaker by keeping prices and profits up.

Consumers do not see the $500 million worth of lower drug prices. The deal nips that possibility in the bud. Some of the money that would have run to consumers via lower prices can ²¹ be used to fund the settlement. With $500 million that might not otherwise have been available to it, the brand-name drugmaker can afford a large payment to the generic drugmaker to get the settlement done. See id.

As the Supreme Court has put it, the “desire to maintain and to share patent-generated profits,” id. at 158, 133 S.Ct. 2223, helps to mark out the line between what is categorically allowed under the antitrust laws (an “early-entry date alone” settlement) and what might not be allowed (an early-entry-plus settlement --- an early-entry-date settlement that comes bundled together with something more, added value given to the generic drugmaker).

The early-entry-plus settlement may benefit the brand-name drugmaker, which protects its patent-monopoly profits from the possibility of a judgment that might have declared the underlying patents invalid. And it can also benefit the generic drugmaker, which gets on the market and gets paid.²²

But all of this may hurt consumers --- who must wait longer for “the introduction of lower cost, generic alternatives.” Id. at 153, 133 S.Ct. 2223. “The patentee and the challenger gain; the consumer loses.” Id. at 154, 133 S.Ct. 2223.

5. Key Aspects of Disallowed Settlements

As noted, settling a paragraph IV lawsuit can sometimes violate the antitrust laws.

If, first, there is delay.

In the example above, see Part II.A.4, under the 2030 settlement agreement the generic drug comes to market not in 2030, but in 2035.²³

And second, if there is also payment for the delay.

In the example, the brand-name drugmaker gets upside under the 2030 settlement --- the five-year delay to 2035, during which the generic drug stays off the market. And the brand-name drugmaker pays for that upside --- by sending a check to the generic drugmaker.

In short, there can be a possibly illegal “pay-for-delay”²⁴ when a settlement agreement has (i) a generic drugmaker delaying its entry onto market and (ii) a brand-name drugmaker paying for the delay.²⁵

* * *

What sort of “payment” can count?

More later. See Part IV and Part VIII. But for now note that there is a quantitative aspect. The payment has to be “large.” Actavis, 570 U.S. at 158, 133 S.Ct. 2223. And there is a qualitative dimension, too. The payment must be “unjustified.” Id.

As to form, a cash payment counts. See id. at 140, 145, 133 S.Ct. 2223; see also AbbVie, 976 F.3d at 356. And so can other “large transfer[s] of value from the patent holder to the alleged infringer.” King Drug, 791 F.3d at 403; accord In re Loestrin 24 Fe Antitrust Litig., 814 F.3d 538, 550 (1st Cir. 2016).

And that is what the Plaintiffs ²⁶ allege happened here --- that Celgene's Revlimid-related settlements included large non-cash transfers of value to generic drug companies. See Complaint ¶¶ 322, 353, 430, 440, 461, 479, 493, 510, 522, 537, 550. That was the pay.

And there was also delay to go with it. Settlements were struck beginning in 2015. See id. ¶¶ 206, 284. But under those settlements, generic Revlimid could only begin coming on the market in 2022.²⁷ See id. ¶¶ 8, 288.

* * *

Does this make out a plausible antirust claim?

To answer, some more background first. A more detailed overview of this litigation. See Part II.B. And a description of the key settlement agreement --- and how, per the Plaintiffs, it adds up to a violation of the antitrust laws. See Part II.C.

B. This Case

In light of the relevant settlement agreements,²⁸ various entities filed lawsuits here.

Pharmacies.²⁹ See Consolidated Amended Complaint (“Retailers’ Complaint”) (ECF 459) ¶¶ 17–20. Insurers.³⁰ See Complaint ¶¶ 11–29; United Healthcare's Complaint ¶¶ 11–17. Union health care plans and trusts.³¹ See Plan and Trust Complaint ¶¶ 11–14. And hospitals.³² See Hospitals’ Complaint ¶¶ 25–30.

From here, these entities are together called “the Plaintiffs.”

Per the Plaintiffs, the settlements amounted to unlawful collusion to keep Revlimid prices high. See Complaint ¶¶ 307, 314, 651.

So when an insurance company/Plaintiff covered a Revlimid claim, it paid extra. See id. ¶ 649; United Healthcare's Complaint ¶ 628; see also Complaint ¶¶ 650-54; United Healthcare's Complaint ¶¶ 629-33.

Or when a pharmacy/Plaintiff or a hospital/Plaintiff bought Revlimid for a patient --- it ended up paying more. See Retailers’ Complaint ¶¶ 369-72; Hospitals’ Complaint ¶¶ 669-77.

* * *

The Plaintiffs sued Celgene, plus the company that later bought Celgene.³³ And the Plaintiffs also sued five generic drugmakers that allegedly entered into Revlimid-related settlements with Celgene.³⁴ See Retailers’ Complaint ¶¶ 24–27; Complaint ¶¶ 33–40; Plan and Trust Complaint ¶¶ 18–25; Hospitals’ Complaint ¶¶ 35–42; United Healthcare's Complaint ¶¶ 20–26.³⁵

From here, the pharmaceutical companies listed out in the preceding paragraph and the associated footnotes are together called “the Defendants.”

* * *

In a nutshell: the drugmakers allegedly involved in the Revlimid settlement agreements (the Defendants) are sued here by entities that were allegedly required to bear higher Revlimid costs because of those settlements (the Plaintiffs).

C. The Natco Settlement

This Opinion and Order zeroes in on one settlement agreement in particular. The one inked by two of the Defendants, Celgene and Natco.³⁶

Go through here the run-up to the Natco settlement. And then its key terms.³⁷

1. The Litigation

In 2005, Celgene filed and the FDA approved an application to market Revlimid. See Complaint ¶ 170. The application relied on “approximately thirty patents,” id. ¶ 180, the last of which was due to expire in 2027. See id. ¶ 204. Generic competitors would likely need to wait until then to come onto the market. See id.

But Natco wanted to launch a generic version of Revlimid sooner. See id. ¶ 198. So it filed a paragraph IV certification. See ¶ 199. The certification said that Celgene's relevant patents were “invalid unenforceable, and/or not infringed by Natco's generic version of Revlimid.” Id.³⁸

In response, Celgene sued Natco. See id. ¶ 200. That was in 2010. See id. The lawsuit went on for a while, and the case was expected to be trial-ready by the first part of 2016. See id. ¶ 271.

* * *

A Natco courtroom win would have opened the way for Natco to launch a generic version of Revlimid. See generally Janssen Pharmaceutica, N.V. v. Apotex, Inc., 540 F.3d 1353, 1355-57 (Fed. Cir. 2008).

But getting to a courtroom win always costs, and here more than usual --- because paragraph IV litigation can be especially “time consuming, complex, and expensive.” Actavis, 570 U.S. at 153, 133 S.Ct. 2223.³⁹ And a courtroom victory in a complicated case is rarely a sure thing.⁴⁰

The basic equation: a big Natco payoff for winning, but an uncertain chance of getting there --- and no matter the outcome, an expensive road to go down.

2. The Settlement Terms and the Plaintiffs’ Theory

Against the above backdrop, Celgene and Natco started negotiating, see Complaint ¶¶ 274-83, and got to a final settlement agreement in 2015. See id. ¶ 284.

The settlement did three main things. Go through these here.

a) Delay

The first main plank of the 2015 settlement: it set the time when Natco would be allowed to launch its generic Revlimid. Not in 2015. But more than six years later --- in March 2022.⁴¹ See id. ¶¶ 8, 288; see also Complaint, Exhibit B (“License Agreement”) (ECF 460-2) § 1.11.

As noted, to make out a plausible antitrust claim, the Plaintiffs must allege pay-for-delay. See Part II.A.5; In re Wellbutrin XL Antitrust Litig. Indirect Purchaser Class, 868 F.3d 132, 163 (3d Cir. 2017).

And per the Plaintiffs, the time-lag baked into the Celgene/Natco settlement agreement amounted to that delay. See Complaint ¶ 320. Natco could potentially have won at trial.⁴² But under the settlement agreement, Natco's generic would come to market only in 2022. See id. ¶¶ 8, 288; see also License Agreement § 1.11.

b) Pay

The second key part of the 2015 Celgene/Natco settlement agreement: Natco could come into the market in 2022 --- but only slowly, with limits on how much generic Revlimid it could annually sell. See Complaint ¶¶ 8, 276, 288-90, 308; see also License Agreement § 1.13.⁴³

Per the agreement, these volume caps would shift over time.

For its first year on the market, Natco's sales would be restricted to 7% of Revlimid sales.⁴⁴ See Complaint ¶¶ 288-89; see also License Agreement § 1.13(a).

For the second year, Natco's generic Revlimid sales would have to come in under a 12% ceiling, and they would be limited to 20% for the year after that. See Complaint ¶¶ 288-89; see also License Agreement § 1.13(b)-(c).

And then, for the eleven months after the third year, Natco's market share would be fixed at 27.5% or less. See Complaint ¶¶ 288-89 & n.156; see also License Agreement § 1.13(d).

This would run until January 2026, at which point Natco would be able to sell its generic Revlimid freely. With no limits, as much as it wanted to. See License Agreement § 1.13(e); see also Complaint ¶¶ 1, 8, 287.

In short, Natco could not dive into the market headfirst. In its first year, it could dip in only a toe. The next year, the water would be calf-high. And so on. A bit deeper each year.

* * *

As noted, to make out an antitrust claim, the Plaintiffs must allege “pay-for-delay.” See Part II.A.5.

And per the Plaintiffs, the volume caps required by the agreement generated a “pay,” because they allowed Natco to charge especially high prices for its generic Revlimid, much more than a generic would normally go for. See Complaint ¶¶ 5, 276-77, 291, 307-08, 314, 561; see also Consolidated Plaintiffs’ Opposition to Defendants’ Joint Motion to Dismiss Plaintiffs’ Amended Complaints (“Plaintiffs’ Opposition”) (ECF 533) at 6, 23.

“[W]ith supply capped” by the Celgene/Natco settlement agreement, “[Natco] could price the quantum-limited supply of generic at near-brand-price levels,” it could “charge highly supra-competitive prices[.]” Complaint ¶ 308.

This, per the Plaintiffs, was the inducement for the delay, for Natco agreeing to come into the Revlimid market not in 2015, when the settlement was struck, but in 2022.

The “reverse payment primarily takes the form of Celgene splitting its monopoly profits with [Natco.]” Id. ¶ 324; see also Plaintiffs’ Opposition at 5 (“[Natco] could charge markedly higher prices for the generic than [it] could with competition.”); id. at 9 (arguing that, per the complaint, Natco “created a regime under which [it] ․ would be able to charge exclusivity prices throughout the almost four years (1,430 days) of volume-capped periods”).

c) Facilitating the Pay

Per the Plaintiffs, the volume caps on Natco (7%, 12%, etc.) would leave behind a market roughly divvied up between Natco and Celgene --- and that market split (93/7, 88/12, etc.) would allow Natco to charge high prices for its generic Revlimid.

But this sort of market division could be easily undone. A 7% cap on one producer (Natco) does not generate an overall 93/7 market-split if new producers come on board --- and pour large volumes of the relevant product (generic Revlimid) into the market.

And without generating an overall market division,⁴⁵ the settlement agreement's cap on Natco's sales volume would be markedly less likely to give Natco an ability to charge high prices on its generic Revlimid.⁴⁶

So this poses the question: in the wake of the settlement agreement, would non-Natco generic Revlimid swamp the market?

No, the Plaintiffs contend. Because the settlement agreement tended to prevent that outcome --- by working to stop up the two main sluices through which generic Revlimid might otherwise be expected to flood into the market.

Generic Revlimid could pour into the market from Celgene, through its sales of an “authorized” Revlimid generic. But per the Plaintiffs, the settlement agreement strongly disincentivized Celgene from going this route.⁴⁷

And generic Revlimid could stream into the market via generic drugmakers, ones other than Natco. These generic drugmakers might cut a deal with Celgene as Natco had. But the Plaintiffs contend that the settlement agreement made it unlikely that non-Natco generic drugmakers would gain meaningful market share that way.⁴⁸

* * *

To sum up.

On the Plaintiffs’ theory, the 2015 Celgene/Natco settlement agreement built in a delay, because Natco's generic Revlimid could first come onto the market only in 2022. And it included a pay, because the agreement's volume caps meant that Natco's generic Revlimid could be expected to sell for especially high prices --- provided that, in the wake of the settlement agreement, generic Revlimid from non-Natco sources would not swamp the market.

Does all of this this add up to a plausible pay-for-delay claim?

Start the analysis on that in a moment, after a last bit of background.

D. Procedural History

The first complaint here was filed in 2019. See In re Revlimid & Thalomid Purchaser Antitrust Litig., 2024 WL 2861865, at *22 (D.N.J. June 6, 2024); Complaint at 115 (ECF 1).

Two years ago, the Court, per Judge Salas, granted the Defendants’ motion to dismiss. See In re Revlimid, 2024 WL 2861865, at *113; June 6, 2024 Order (ECF 447).⁴⁹

Part of the Court's opinion took up allegations that the various Revlimid-related settlements amounted to a pay-for-delay settlement. See In re Revlimid, 2024 WL 2861865, at *6-10, *51-67. But the Plaintiffs pressed, and the Court considered, numerous other theories, too. See id. at *25-37, *40-51, *76-105.

After the Court granted the motion to dismiss, the Plaintiffs were permitted “to file an amended complaint curing the deficiencies outlined in the Court's ․ opinion.” June 6, 2024 Order at 4.

In response, the Plaintiffs filed five amended complaints. See Retailers’ Complaint ¶¶ 374-83; Complaint ¶¶ 670-1227; Plan and Trust Complaint ¶¶ 667-1315; Hospitals’ Complaint ¶¶ 678-1797; United Healthcare's Complaint ¶¶ 683-785.

These are now the operative complaints, and they are substantially different than what came before.

They concentrate only on the alleged pay-for-delay aspects of the Celgene settlements. See Memorandum of Law in Support of Defendants’ Joint Motion to Dismiss Plaintiffs’ Amended Complaints (“Defendants’ Brief”) (ECF 529) at 7.

And they rest on new arguments. See generally Plaintiffs’ Opposition at 1 (stating that the Plaintiffs’ allegations now “disclose the methods by which the [D]efendants protected the likelihood of anticompetitive harm to be inflicted”).⁵⁰

Via a joint motion, the Defendants moved to dismiss the Plaintiffs’ amended complaints in their entirety. See Defendants’ Brief at 1 & n.1, 4 n.3, 60.⁵¹

After the parties briefed their motions, the Chief Judge reassigned this case to the undersigned. See May 5, 2025 Text Order (ECF 563).

The Court then held oral argument on “issues 1-4,” which are aspects of the Defendants’ motion to dismiss that focus on the Celgene/Natco settlement agreement. See Transcript of Oral Argument (“Transcript”), In re Revlimid & Thalomid Antitrust Litig., No. 19-07532 (D.N.J. Feb. 19, 2026) (ECF 580).

That is the part of the Defendants’ overall motion to dismiss that is taken up here.

III. The Court's Approach

A quick preview.

The Court goes forward on the understanding that there are four elements of a pay-for-delay claim. See Part IV.

Of these, two are not contested by the Defendants. See Part V.

As to the remaining two, the first is size --- whether the alleged value transfer from Celgene to Natco via the settlement agreement was large enough. Yes, the Court concludes. And that holds up whether value is measured in absolute terms or relative to litigation costs saved by settling. See Part VI.

As to the last element, pay --- as a matter of both common sense and basic economics, the settlement agreement plausibly opened the way for Natco to charge especially high prices for generic Revlimid. See Part VIII.A and Part VIII.B. And to the extent it did --- that can count as a “pay” as a matter of law. See Part VIII.C.

But the settlement agreement could likely allow Natco to charge high prices only on the assumption that generic Revlimid would not flood the market from sources other than Natco. Is that assumption plausible? Yes, the Court holds. See Part IX.

What this adds up to: the Plaintiffs have pressed plausible allegations as to each of the two contested pay-for-delay elements --- and the Defendants’ motion to dismiss as to the issues (1-4) taken up here must therefore be denied. See Part X.

IV. The Elements

A. Delay, Justification, Size, Pay

The Defendants have filed a motion to dismiss. To analyze it, job one is to call out the various elements of the claim --- and then ask whether there are plausible allegations on each element. See Connelly v. Lane Constr. Corp., 809 F.3d 780, 787 (3d Cir. 2016); see also Badalamenti v. Resideo Techs., Inc., 807 F. Supp. 3d 414, 418 n.8 (D.N.J. 2025); Doyle v. Matrix Warranty Sols., Inc., 679 F. Supp. 3d 42, 44 (D.N.J. 2023).

A plaintiff who presses a pay-for-delay claim must allege four things.⁵²

First, a delay. A sufficient lapse of time between (i) the settlement agreement and (ii) the moment when the generic drugmaker can get into the market under the agreement. See In re Wellbutrin, 868 F.3d at 163.

Second, a payment. A transfer of value from the patentholder to the generic drugmaker. See AbbVie, 976 F.3d at 356; King Drug, 791 F.3d at 403; In re Lipitor, 868 F.3d at 260; In re Wellbutrin, 868 F.3d at 161; see also United Food & Com. Workers Local 1776 Participating Emps. Health & Welfare Fund v. Teikoku Pharma USA, Inc., 74 F. Supp. 3d 1052, 1065 (N.D. Cal. 2014). A cash payment. See Actavis, 570 U.S. at 145, 133 S.Ct. 2223. Or some other form of payment. See King Drug, 791 F.3d at 403; AbbVie, 976 F.3d at 356; see also In re Loestrin, 814 F.3d at 550.

Third, that the payment was “large” and fourth that it was “unjustified.” Actavis, 570 U.S. at 158, 133 S.Ct. 2223.⁵³

B. A Fifth Element?

Next up is to consider where there are plausible allegations as to each element. See Connelly, 809 F.3d at 787.

But before getting there, pause to take on a wrinkle --- the possibility that a fifth element must be alleged.

In AbbVie, the Third Circuit seemed to suggest that. It stated that an antitrust plaintiff must “allege[ ] that an agreement's anticompetitive effects outweigh its procompetitive virtues[.]” 976 F.3d at 356.

And under AbbVie, “anticompetitive effects” looks to be a genuinely independent element. Not just a shorthand term to convey the bottom-line conclusion that there are sufficient allegations as to a pay-for-delay claim's actual elements. See 976 F.3d at 357 (working through whether “it [wa]s plausible that the anticompetitive effects of [the patent-owner's] settlement with [a competitor] outweighed any procompetitive virtues of the [side] deal” --- on the merits, and separately from considering two other elements, the alleged size and lack of justification for the settlement in question).⁵⁴

* * *

But while AbbVie appears to suggest that “anticompetitive effects” is a bona fide element, in King Drug the Third Circuit may have taken a different tack.

King Drug implies that the plaintiff can make out a prima facie pay-for-delay case by proving the basic four elements ⁵⁵ --- with “anticompetitive effects” understood as just an inference that follows from those four, not as a stand-alone fifth element that must be separately put forward.⁵⁶

Is “anticompetitive effects” a fifth element that must be independently pled, as AbbVie perhaps suggests? Or is it just a kind of automatic output, the necessary implication of the actual four elements being satisfied, as King Drug maybe suggests?

* * *

Tie all of this off and put it to one side. It does not matter for now.

The Court referenced this issue at oral argument. See Transcript at 10:22-11:5, 14:23-15:1. But no party picked up the baton. No one suggested that “anticompetitive effects” must be alleged as its own, free-standing element.⁵⁷

So the Court follows the parties’ lead, and does not pursue the matter. It assumes for present purposes that there are four elements that must be made out to establish a pay-for-delay antitrust claim, the ones laid out in Part IV.A.

V. Delay and Justification

Push quickly past two of the four pay-for-delay elements. The Defendants have not put those in play.

Delay first.

The Celgene/Natco settlement agreement was inked in 2015, but under the agreement a Natco generic could not generally come on the market until 2022. See Complaint id. ¶¶ 8, 288; see also License Agreement § 1.11.

The Defendants do not argue that this delay was insufficient. So delay does not need to be taken up here.⁵⁸

Same as to justification.⁵⁹

The Defendants do not argue that the Plaintiffs failed to check this box. So this element, too, can be put aside.

* * *

The remaining two elements relate to whether the Plaintiffs have adequately alleged a payment, and whether any payment that was made was sufficiently “large.”⁶⁰

Get to the payment element in Part VII. Take up now the size element.

VI. Size

There are two ways to assess whether a payment is large enough for pay-for-delay purposes.

* * *

First, the size of a payment can be measured in relative terms, stacked up against legal costs saved. A settlement payment of $1 million that saves $1 million in anticipated litigation fees --- that is no “large” settlement.

The Supreme Court and the Third Circuit have relied on this sort of relative-size analysis. See Actavis, 570 U.S. at 156, 133 S.Ct. 2223; AbbVie, 976 F.3d at 357; Watson Lab'ys., 101 F.4th at 239; see also In re Solodyn (Minocycline Hydrochloride) Antitrust Litig., 2015 WL 5458570, at *7 (D. Mass. Sept. 16, 2015); United Food, 74 F. Supp. 3d at 1071-72; In re Opana ER Antitrust Litig., 162 F. Supp. 3d 704, 719 (N.D. Ill. 2016); In re Aggrenox Antitrust Litig., 94 F. Supp. 3d 224, 243 (D. Conn. 2015); In re Bystolic Antitrust Litig., 583 F. Supp. 3d 455, 480, 484 (S.D.N.Y. 2022); cf. In re Asacol Antitrust Litig., 233 F. Supp. 3d 247, 263-64 (D. Mass. 2017); Phillip E. Areeda & Herbert Hovenkamp, Antitrust Law: An Analysis of Antitrust Principles and Their Application ¶ 2046e2 (2025 ed.) (“The Actavis decision does not require evidence of a payment of a particular size, but only a payment in excess of reasonably anticipated litigation costs.”).⁶¹

* * *

A second approach: to ask whether a payment is large enough by looking to the payment on a stand-alone basis. Not against any particular comparison point. But in absolute terms.

The Third Circuit has used this approach, too.

In King Drug, for example. There, “the allegations were simply that [the settlement] provided the alleged infringer with ‘many millions of dollars of additional revenue,’ and that the patentee otherwise had ‘an incentive to launch its own [generic,]’ ” In re Lipitor, 868 F.3d at 255 (quoting King Drug, 791 F.3d at 410), which under the settlement it opted not to pursue. This was enough. See King Drug, 791 F.3d at 409-10. With numbers of this size on the table, a comparison to a specific baseline was unnecessary.

So too in Wellbutrin. The payment there was “worth $233 million[,] ․ an amount that would qualify as large in most any context.” 868 F.3d at 162. The payment was treated as all but self-evidently large enough. It spoke for itself.

In Lipitor, basically the same approach. That appeal involved two different drugs. See In re Lipitor, 868 F.3d at 239. As to Lipitor, the alleged payment was allegedly “worth hundreds of millions of dollars,” and therefore passed the “large” test. Id. at 253-54; see also id. at 255, 257-58. As to Effexor, the transfer of value was allegedly worth more than half a billion dollars. See id. at 247, 258, 260. Without further analysis, the Third Circuit held that the “alleged reverse payment ․ [wa]s plausibly large.” Id. at 260.

And finally AbbVie. There, the Third Circuit held that the alleged “payment was plausibly ‘large.’ ” 976 F.3d at 357. The settlement “was ‘extremely valuable.’ ” Id. It would help to clear the way to nearly $200 million in projected sales “over a four-year period.” Id.⁶² Again, a transfer of value measured in the hundreds-of-millions-of-dollars was treated as large enough --- without more explanation.

* * *

How to measure size? The relative approach or the absolute? Or maybe a combination of the two?

In some cases, these questions may matter. But not here.

The payment from Celgene to Natco ⁶³ under the settlement agreement allegedly added up to billions of dollars. See Complaint ¶ 332.⁶⁴ That is plainly large enough. In absolute terms. Or in relation to the litigation costs saved by settling.⁶⁵

In short, the Plaintiffs have adequately pled that, via the 2015 settlement agreement, Celgene made a “large” payment to Natco.

VII. Pay: Overview

Pause to see where things stand.

The Court treats the Plaintiffs as needing to plausibly allege four things --- (1) delay (by Natco in getting its generic to market) and (2) a payment (to Natco) that was (3) large and (4) unjustified. See Part IV.

Delay and justification, (1) and (4), are not on the table here. See Part V.

And there are plausible allegations as to (3). See Part VI.

That leaves (2) --- the question of whether the Plaintiffs have plausibly alleged that the Celgene/Natco settlement amounted to a payment to Natco.⁶⁶

* * *

To answer that question, the Court breaks it in two.

First, the Court assumes that, after the 2015 Celgene/Natco settlement agreement, a large volume of non-Natco generic Revlimid would not plausibly have flooded into the market.

On that assumption, Natco's volume caps (7%, 12%, etc.) would generate an overall market division (roughly 93/7, 88/12, etc.). And in that circumstance, the Court concludes, the settlement agreement would plausibly allow Natco to charge high prices for generic Revlimid. See Part VIII.B. And that can count as a “pay” as a matter of law. See Part VIII.C.

Second, the Court circles back to the earlier assumption --- and asks whether it is indeed plausible that, in the wake of the settlement agreement, a large tide of non-Natco generic Revlimid would not swamp the market.

Yes, the Court concludes. That is plausible, and Part IX explains why.

Putting all this together --- the “pay” element is plausibly satisfied.

VIII. Pay: Higher Per-Unit Prices

Get started now with the “pay” analysis.

If the assumption set out above holds,⁶⁷ how would the Celgene/Natco settlement agreement plausibly play out?

The Court concludes: Natco would plausibly be able to charge especially high prices for its generic Revlimid, not too much less than what Celgene charges for its brand-name Revlimid.

To begin to see why, take an illustration.

Say 100 people want Revlimid, and Celgene has been charging $20 for each dose of it. And say that generics usually go for around 1/2 of what a brand-name drug costs, so here generic Revlimid might be expected to run around $10. Because of the settlement agreement's volume caps, Natco would only be able to sell to seven people in its first year on the market, 2022. In that circumstance, Natco would be well-advised to charge a good deal more than $10. A slightly lower price than $20 (what Celgene has long been charging for brand-name Revlimid) would allow Natco to make the seven sales allocated to it for 2022 to 2023.⁶⁸ But there would be no reason for a price much less than $20. After all, whether Natco were to charge $19.75 or $9.75 -- - it could still sell to only seven people.⁶⁹

Lower the drug price just enough to generate sales, yes. But why lower the price further than that, if there are no more sales to get?⁷⁰

Natco's drafting close behind Celgene in terms of price would open the way to Natco charging relatively high prices for its generic. Because “[a] manufacturer of a brand-name drug protected by a patent,” like Celgene, “is able to sell the drug at monopoly prices.” Kaiser Found. Health Plan, Inc. v. Abbott Lab'ys., Inc., 552 F.3d 1033, 1036 (9th Cir. 2009).⁷¹

In short, common sense --- applied while giving the Plaintiffs the benefit of all reasonable inferences ⁷² --- shows that the settlement agreement's volume caps would plausibly allow Natco “to charge highly supra-competitive prices” for generic Revlimid. Complaint ¶ 308.

This, per the Plaintiffs, was the pay --- the inducement for the delay, for Natco agreeing to come into the Revlimid market not in 2015, but in 2022.

On these allegations, the competition that the Celgene/Natco settlement opened up did not matter very much. Competition in that two companies would be on the market instead of one. But competition without the thing that matters most about it here -- - lowered prices.⁷³

* * *

As the Plaintiffs put the logic of all this, “[u]nder fundamental principles of applied microeconomics,” Natco “could price the quantum-limited supply of generic at near-brand-price levels.” Complaint ¶ 308.

A generic drugmaker restricted to a narrow part of a market (like 7% of the Revlimid market) “knows that no matter how low it prices its generic product, it will only capture 7% of the market.” Id. Therefore, it “has no rational economic incentive to price its generic product lower.” Id.; see also id. ¶¶ 5, 292, 295, 311-12; Plaintiffs’ Opposition at 9, 27.⁷⁴

* * *

The conclusion that the settlement agreement allowed Natco to charge especially high prices for generic Revlimid is plausible and rests on common sense.⁷⁵ That may be enough.⁷⁶

But the Plaintiffs go further. As alluded to, they contend that their allegations are also backed up by “fundamental principles of applied microeconomics.” Complaint ¶ 308.⁷⁷

So turn now to the economics literature --- to get a further handle on how market conditions would plausibly evolve under the Celgene/Natco settlement agreement.

* * *

The analysis runs as follows.

First, the Court establishes that a look to economics is appropriate here --- even standard. See Part VIII.A.⁷⁸

Second, the Court shows that economics points in the same rough direction as common sense. Namely, that the settlement agreement's volume caps would plausibly open the way to Natco being able to charge high per-unit prices on generic Revlimid.⁷⁹ See Part VIII.B.

And third, the Court concludes that allowing Natco to charge high per-unit prices can count as a “pay” as a matter of law. See Part VIII.C.

A. Method

Per the Supreme Court: “[w]hether an antitrust violation exists necessarily depends on a careful analysis of market realities.” Nat'l Collegiate Athletic Ass'n v. Alston, 594 U.S. 69, 93, 141 S.Ct. 2141, 210 L.Ed.2d 314 (2021).⁸⁰

The Third Circuit has followed this lead.

Antitrust ․ requires the courts to seek the economic substance of an arrangement, not merely its form. The ‘substance’ of an arrangement often depends on the economic incentive of the parties.

Weiss v. York Hosp., 745 F.2d 786, 815 (3d Cir. 1984) (citations omitted).⁸¹

And the Third Circuit has gone this way in the pay-for-delay context --- noting that “a reverse payment's legality depends mainly on its economic substance[.]” AbbVie, 976 F.3d at 356; see also King Drug, 791 F.3d at 406 n.24. This has often meant drawing from economics-minded scholarship. See King Drug, 791 F.3d at 404-05; In re Wellbutrin, 868 F.3d at 144 n.7.⁸²

The Court takes this same approach here.

Antitrust “requires the [C]ourt[ ] to seek the economic substance of an arrangement,” Weiss, 745 F.2d at 815 (emphasis added) --- here, the plausible impact of the Celgene/Natco settlement agreement. And a look to the economics literature helps to get closer to that “substance,” id., to the “market reality” that “is the touchstone” here. Harrison Aire, Inc. v. Aerostar Int'l Inc., 423 F.3d 374, 383 (3d Cir. 2005).

* * *

To be sure, economic analysis cannot become a float out into abstraction. To be useful, economics must be the opposite of that --- a way to keep judicial analysis anchored down, grounded in a real-world sense of how markets work.

“Federal courts are not economists, and ․ should avoid an unnecessary ‘ramble through the wilds of economic theory.’ ” Host Int'l Inc. v. MarketPlace, PHL, LLC, 32 F.4th 242, 253 n.13 (3d Cir. 2022) (quoting United States v. Topco Assocs., Inc., 405 U.S. 596, 609 n.10, 92 S.Ct. 1126, 31 L.Ed.2d 515 (1972)).⁸³

And as then-Judge Breyer indicated, antitrust law “cannot precisely replicate the economists’ (sometimes conflicting) views.” Barry Wright Corp. v. ITT Grinnell Corp., 724 F.2d 227, 234 (1st Cir. 1983); see also Leegin Creative Leather Prods., Inc. v. PSKS, Inc., 551 U.S. 877, 914-15, 127 S.Ct. 2705, 168 L.Ed.2d 623 (2007) (Breyer, J., dissenting) (closely similar).

Against this backdrop, the Court steers away from unmapped “wilds,” and also from areas that are dense with crossroads and forking paths --- where “conflicting[ ] views” send debating economists off in disparate directions.

Instead, the focus here is on well-trod routes and conventional principles.

In particular, the Court turns to the bread-and-butter economics literature on the dominant firm model. See Part VII.B.

Doing so helps to systematize and stress-test the common-sense conclusion set out above, that the volume caps in the Celgene/Natco settlement agreement would plausibly allow Natco to charge especially high prices for its generic Revlimid.

B. Economics: Higher Prices

Look now to the dominant firm model.⁸⁴

The model has long been used as a way to understand the pharmaceutical industry.⁸⁵ And it has routinely been used in the antitrust context.⁸⁶

The dominant firm model envisions a market in which one firm, the dominant one, controls a big chunk of total sales. See, e.g., Oliver E. Williamson, Dominant Firms and the Monopoly Problem: Market Failure Considerations, 8 Harv. L. Rev. 1512, 1512 n.1 (1972).

Maybe 60% or so.⁸⁷ Or sometimes even less than that.⁸⁸

The dominant firm's ascendance typically stems from some special advantage. Maybe an especially strong first-mover advantage. Or lower production costs. See John M. de Figueiredo & Brian S. Silverman, Churn, Baby, Churn: Strategic Dynamics Among Dominant and Fringe Firms in a Segmented Industry, 4 Mgmt. Sci. 632, 633-34 (2007); see also United States v. Oracle Corp., 331 F. Supp. 2d 1098, 1113 (N.D. Cal. 2004).

But although the dominant firm has an advantage, it is not so dominant as to be a monopolist. See Lawrence J. White, Monopoly and Dominant Firms: Antitrust Economics and Policy Approaches, in 1 The Oxford Handbook of International Antitrust Economics 313, 320 (Roger D. Blair & D. Daniel Sokol eds., 2015).

Rather, the dominant firm coexists with one or more competitors --- which, together, serve only a relatively small slice of the overall market.

These competitors make up the “fringe,” and they produce goods that are closely similar to (or identical to) what the dominant firm is offering up. See Simran K. Kahai et al., Is the “Dominant Firm” Dominant? An Empirical Analysis of AT&T's Market Power, 39 J.L. & Econ. 499, 502 (1996); William M. Landes & Richard A. Posner, Market Power in Antitrust Cases, 94 Harv. L. Rev. 937, 944 (1981).

Fringe firms often --- but not always --- have only limited capacity, an ability to produce only some of the total volume that the market demands. See Louis Kaplow, On the Relevance of Market Power, 130 Harv. L. Rev. 1303, 1339-40 & n.68 (2017); Landes & Posner, Market Power in Antitrust Cases, 94 Harv. L. Rev. at 944 & n.15; Robert Cherry, The Dominant Firm Model Revisited, 16 Rev. Indus. Org. 89, 89 (2000) (quoting William Shepherd, The Economics of Industrial Organization 206).

* * *

What is laid out above is analogous to the set-up in this case.

Here, there are two companies making virtually the same thing. Brand-name Revlimid. And a generic version of it.

One of the companies (Celgene) is dominant by any measure --- able to service the large bulk of the market for 2022 to 2026, to the tune of something between 93% in the first year to 72.5% in the last. Well past the 40% or 50% or 60% mark that typically makes an entity “dominant” for these purposes.

And the dominant company here (Celgene) is armed with an advantage (the settlement agreement) to lock in its dominance. Natco might do great work in 2022. But no matter how much it pulls down its production costs or smooths its distribution channels --- Natco would still be required by the settlement agreement to leave 93% of the Revlimid market to Celgene.

And under the settlement agreement, Natco is plainly on the fringe. Able to service a piece of the market. But stuck with just that, and no more --- a ceiling on its capacity to go after added business.

In short, the analogy to the dominant firm model largely fits.⁸⁹ In light of the settlement agreement's volume caps,⁹⁰ Celgene can be taken as a dominant firm, and Natco as working from the fringe.

* * *

How do things tend to work out when a market is structured along dominant/fringe lines?

The economics literature backs up what common sense teaches.

Namely, that fringe companies are generally “too small to affect market price through [their] output decisions.” Hideki Yamawaki, Dominant Firm Pricing and Fringe Expansion: The Case of the U.S. Iron and Steel Industry, 1907-1930, 67 Rev. Econ. & Stat. 429, 430 (1985); United States v. Von's Grocery Co., 384 U.S. 270, 287 n.12, 86 S.Ct. 1478, 16 L.Ed.2d 555 (1966) (Stewart, J., dissenting) (citing Derek C. Bok, Section 7 of the Clayton Act and the Merging of Law and Economics, 74 Harv. L. Rev. 226, 312 n.261 (1960)).

The fringe is mostly “reactive.” White, Monopoly and Dominant Firms at 321; see also Landes & Posner, Market Power in Antitrust Cases, 94 Harv. L. Rev. at 944 n.15.

Rather than looking to drive down prices, as other competitors might, the fringe largely takes the price as set by the dominant firm. See, e.g., Roger D. Blair & Victoria G. Maurer, Umbrella Pricing and Antitrust Standing: An Economic Analysis, 4 Utah L. Rev. 763, 764 (1982) (“the fringe firms set their prices under the ‘umbrella’ of the dominant firm[ ]”); George J. Stigler, The Kinky Oligopoly Demand Curve and Rigid Prices, 55 J. Pol. Econ. 432, 445 (1947) (noting that the fringe sells “at the leader's price”); Christoph Schenzler et al., The History of the Static Equilibrium Dominant Firm Price Leadership Model, 18 E. Econ. J. 171, 171 (1992) (noting that the fringe “rivals ․ behave as price-takers”); Stephen J. Rassenti & Bart J. Wilson, How Applicable is the Dominant Firm Model of Price Leadership?, 7 Experimental Econ. 271, 272 (2004) (noting that “the dominant firm chooses its price,” and thereby determines the price “of the competitive fringe,” which “tak[es] the industry price as given”).⁹¹

As Professor George Stigler laid it out:

(1) one firm sells such a large proportion of the commodity in question that the other (small) firms individually ignore any effect they may have on prices; and (2) this dominant firm behaves passively, i.e., it sets the price and sells the remainder after minor firms have sold all they wish at the ruling price.

George J. Stigler, Notes on the Theory of Duopoly, 48 J. Pol. Econ. 521, 522-23 (1940).⁹²

When it comes to price, the dominant firm and the fringe are, in short, a yoked pair. They are linked together. But the dominant firm pulls a small step ahead, setting the pace. It fixes the price. And the fringe firm essentially follows along, taking it.

* * *

On the picture of things sketched out above, the dominant firm works like a “quasi-monopolist.”⁹³

The dominant firm can all but set its price. White, Monopoly and Dominant Firms at 322 (“the dominant firm ․ causes the price to be higher and the quantity sold lower than if full competition ․ could prevail”).

That is the “monopolist” side of things.

On the other hand is the “quasi” part of the ledger. The dominant firm does not fully dominate the market, and prices reflect that to an extent. See id. at 323.⁹⁴

Taking all this together, in a market with a dominant firm, the market price may be lower than it would be in a pure monopoly. But because the dominant firm is a “quasi-monopolist,” the market price will typically remain supracompetitive. See George J. Stigler, The Dominant Firm and the Inverted Umbrella, 8 J.L. & Econ. 167, 171 (1965) (“the dominant firm will find, usually, that it is more profitable to yield up some share of the industry, for higher prices may more than offset the decline in share”).

* * *

Now look to how all of this links up with the previously-described relationship between (i) the dominant firm (Celgene) and (ii) the fringe firm (Natco).

Celgene enjoyed the ability to charge supracompetitive patent-monopoly prices on Revlimid.⁹⁵

And Natco, from the fringe, would plausibly be expected to take the dominant firm's price. To follow Celgene's price lead --- and therefore to also charge (like Celgene) a supracompetitive monopoly-type price (or something in the rough neighborhood of it).

* * *

To summarize:

The economics literature generally supports the common-sense idea ⁹⁶ that it is plausible that the volume caps in the Celgene/Natco settlement agreement ⁹⁷ would open the way to the fringe (Natco) charging especially high prices for generic Revlimid --- lower than the monopoly-type price charged by the dominant firm (Celgene) but not much lower, as would typically be the case with a generic.⁹⁸

C. Law: Higher Prices Count

As a matter of common sense and as a matter of economics, the Celgene/Natco settlement agreement plausibly provided Natco with an ability to sell its generic Revlimid at especially high per-unit prices.⁹⁹ See Part VII.B.

As a matter of law, can that sort of transfer of value satisfy the “pay” element of a pay-for-delay claim?

The Plaintiffs say “yes.” See Complaint ¶¶ 4, 6, 291, 322-24; Plaintiffs’ Opposition at 23. The Defendants see it the other way. See Defendants’ Reply Memorandum of Law in Further Support of Joint Motion to Dismiss Plaintiffs’ Amended Complaints (“Defendants’ Reply”) (ECF 538) at 2; Transcript at 15:23 (“[t]here's no payment”); Defendants’ Brief at 24.

The Plaintiffs have it right, the Court concludes.

First, their position is consistent with the basic underpinnings of the pay-for-delay doctrine.

And second, it is backed up by King Drug --- a Third Circuit decision that takes up a settlement agreement analogous to the one here.

Work through these two points here, starting just below.

1. The Basis for the Doctrine

This section looks to two of the main wellsprings of pay-for-delay law --- (i) why the doctrine is there in the first place, and (ii) how it is supposed to work.

Those suggest that a settlement agreement can count as a “pay” where, as here, it plausibly allows a generic drugmaker to charge especially high prices on the relevant generic drug. Point (i), as to why, is covered in Part VIII.C.1.a. Point (ii), as to how --- that is the subject of Part VIII.C.1.b.

a) Cost to Consumers

“Ever since Congress overwhelmingly passed and President Benjamin Harrison signed the Sherman Act in 1890, protecting consumers from monopoly prices has been the central concern of antitrust.” Apple Inc. v. Pepper, 587 U.S. 273, 288, 139 S.Ct. 1514, 203 L.Ed.2d 802 (2019) (cleaned up).¹⁰⁰

And this animated the Supreme Court's Actavis decision. Per Actavis, pay-for-delay settlements warrant antitrust scrutiny because they might “keep[ ] prices at patentee-set levels,” by dividing “monopoly return[s] ․ between the challenged patentee and the patent challenger.” Actavis, 570 U.S. at 154, 133 S.Ct. 2223.

The settling brand-name firm might be “induc[ing] the generic challenger to abandon its [patent-infringement] claim” by offering the generic drugmaker “a share of its monopoly profits that would otherwise be lost in the competitive market.” Id.

On this approach, the key trouble with a pay-for-delay settlement is that its “objective” may be to “maintain supracompetitive prices to be shared among the patentee and the challenger rather than face what might have been a competitive market --- the very anticompetitive consequence that underlies the claim of antitrust unlawfulness.” Id. at 157, 133 S.Ct. 2223.

If a settlement aims “to maintain and to share patent-generated monopoly profits, then, in the absence of some other justification, the antitrust laws are likely to forbid the arrangement.” Id. at 158, 133 S.Ct. 2223 (emphases added).

The profit-sharing implicit in a settlement matters. Not mainly for its own sake. Profit-sharing between companies is, after all, an everyday thing.

Rather, profit-sharing matters because it may harm consumers --- where it is used to leave the potentially improper patent-monopoly intact, and the higher prices associated with it.

“[S]har[ing] patent-generated monopoly profits” is a concern in part because it allows those profits to be “maintain[ed]” through high prices --- to the detriment of consumers. “[P]ayment in return for staying out of the market[,] simply keeps prices at patentee-set levels, potentially producing the full patent-related ․ monopoly return while dividing that return between the challenged patentee and the patent challenger.” Id. at 154, 133 S.Ct. 2223.

Bottom line: “protecting consumers from monopoly prices” is the “the central concern of antitrust,” Apple, 587 U.S. at 288, 139 S.Ct. 1514 (cleaned up), and pay-for-delay law is a thread in that fabric. It limits payments from brand-name drugmakers to generic drugmakers mainly as a means to an end --- to ensure that consumers are not improperly being charged high monopoly-type prices, that the settlement payment from the brand-name drugmaker to the generic drugmaker is not in a sense subsidized by supercharged profits, that would have flowed back to consumers (in the form of lower prices) but for the settlement having been struck.

* * *

This all but clinches things.

On the assumption invoked above, see footnote 67, the settlement agreement's Natco volume caps (7%, 12%, etc.) would plausibly generate a roughly 93/7, 88/12, etc. market division --- between a dominant Celgene, and Natco from the fringe.¹⁰¹

And a 93/7 market split or an 88/12 market split plausibly holds prices up at monopoly-type levels.

For the vast majority of consumers, the 93% --- whose only choice is to buy brand-name Revlimid, from Celgene.¹⁰²

And also for the other 7% of consumers --- who can look only to Natco, which can be expected to follow Celgene's lead, and keep prices high as to its corner of the market.¹⁰³

In a nutshell, a core animating concern of the pay-for-delay doctrine (as laid out in Actavis) is that settlement payments might be used to artificially keep drug prices high.

So it should count as a “pay” as a matter of law where a settlement directly works to keep prices high --- where, as here, the very way a settlement allegedly transfers value to a generic drugmaker is by giving ¹⁰⁴ it an opportunity to charge high per-unit drug prices.

b) Cost to the Patentholder

For the next point, start with a small step back.

In the run-up to the Supreme Court's 2013 Actavis decision, there were pushes and pulls in the lower-court caselaw.

Pay-for-delay settlements seemed to raise antitrust issues.¹⁰⁵

But a countervailing concern was out there, too. To decide whether a settlement payment improperly prolonged a drugmaker's patent monopoly, a court might have to engage in “an after-the-fact calculation of how ‘likely’ a patent holder was to succeed in a settled lawsuit if it had not been settled.” Federal Trade Comm'n v. Watson Pharms., Inc., 677 F.3d 1298, 1313 (11th Cir. 2012), rev'd sub nom., Actavis, 570 U.S. 136, 133 S.Ct. 2223. As the Eleventh Circuit put it, that would mean “deciding a patent case within an antitrust case about the settlement of the patent case, a turducken task.” Watson Pharms., 677 F.3d at 1315; accord, e.g., Schering-Plough Corp. v. Federal Trade Comm'n, 402 F.3d 1056, 1076 (11th Cir. 2005).

In Actavis, the Supreme Court worked to thread the needle.

It held that pay-for-delay claims could be viable. And it did so in a way that aimed to avoid the Eleventh Circuit's “turducken” problem --- by requiring courts to use the size of the alleged settlement payment as a proxy for the strength of the patent claims in the underlying infringement litigation. See Actavis, 570 U.S. at 157-58, 133 S.Ct. 2223. Size of payment was said to offer “a workable surrogate for a patent's weakness, all without forcing a court to conduct a detailed exploration of the validity of the patent itself.” Id. at 158, 133 S.Ct. 2223.

An analogy from the leading treatise shows the logic of this approach:

A landowner with a clear title does not ordinarily pay a trespasser large amounts of money to stay off her property. This has nothing to do with the number of alternative trespassers out there but results from the fact that land title records are good, are generally reliable, and thus expose the landowner to minimal risk. A very large payment from the landowner to the trespasser to stay off the land is thus an irrational act unless it reflects significant doubts about the quality of the title.

Areeda & Hovenkamp, Antitrust Law ¶ 2046e2.

A big payment to a trespasser suggests that the underlying title might be understood not to be rock-solid --- and that conclusion can be reached without having to do a title search.

So too with a large and unjustified reverse payment. It suggests that the settlement-payor, the brand-name drugmaker, had concerns about the validity of its underlying patent. And that suggestion is there for the seeing even without having to assess the patent's validity.

A big payment implies that the patent holder “ha[d] serious doubts at the patent's survival.” Actavis, 570 U.S. at 157, 133 S.Ct. 2223.¹⁰⁶ After all, why would a pharmaceutical company make a large payment for no reason? Zeroing in on size-of-payment rather than strength-of-patent made “an antitrust action ․ more feasible administratively than the Eleventh Circuit believed.” Id.

* * *

Note, now, a critical (though implicit) aspect of all this.

Namely, a payment can shed light in the way Actavis envisioned only if the payment feels costly to the payor, if it is painful in some meaningful sense to the drugmaker that holds the relevant patents.¹⁰⁷

To see the point, take an example.

Imagine two houses that sit next to each other. Say they are far removed from anything. No nearby town. No other houses in the area. And say that one house has a tree that produces around 500 apples each year --- but the homeowner has other things going on and she can only pick 400 apples. The homeowner may go ahead and let her neighbor pick the extra 100 apples and keep them for himself. That amounts to a transfer of value from the perspective of the neighbor who gets the apples. 100 apples are worth something real.

But this transfer of value does not impose any meaningful costs on the first homeowner. She could not pick those apples anyway. If they were not given away, the 100 apples would just end up on the ground, as waste. So giving up the apples is no pain-point. It imposes no felt costs on the homeowner.

In this circumstance, the homeowner's transfer of value (the 100 apples) is not a potentially “irrational act,” Areeda & Hovenkamp, Antitrust Law ¶ 2046e2 --- such that there is a need to look around for an explanation of why she gave them up. After all, the last 100 apples were useless to the homeowner anyway. So there is no reason to assume she must have quietly had a special motive for foregoing them --- something analogous to a concern that a patent is invalid (in the Actavis context) or that title is not solid (in the trespass analogy).

A transfer of value like the one described above does not cry out for explanation, because it is just sleeves off the vest. It transfers value to the recipient of value, the neighbor who gets the 100 apples. But it does not impose real costs on the giver of value, the homeowner who parts with the 100 apples. It is a benefit to the payee. But because it is not a burden to the payor --- there is no reason to cast around, trying to figure out why the payor might have done what she did.

* * *

Against the above backdrop, can the Celgene/Natco settlement agreement count as a “pay” to the extent it plausibly allowed Natco to charge especially high Revlimid prices?

Yes. Two reasons why.

First, doing so does not lead to what the Eleventh Circuit called the “turducken” problem. See Watson Pharms., 677 F.3d at 1315. It keeps antitrust law and patent law in their separate corners. It does not open the way to pay-for-delay litigation that pulls antitrust law and patent law together, that makes the merits of the current antitrust case turn on an assessment of the merits of the now-settled patent case.

And second, paying Natco by giving it a way to charge high prices for generic Revlimid --- that imposed real felt costs on Celgene.

On balance, the settlement agreement may well have been a good deal for Celgene.

But nonetheless, striking that deal would likely have hurt. For Celgene, it would have meant walking away from selling Revlimid at protected monopoly prices to 7% or 12% etc. of a $7 billion market. See Complaint ¶ 329.¹⁰⁸

That is a stiff cost, and it could only have felt like one. It is the pain of that cost that is clarifying --- because per Actavis, it calls out for an explanation. Why would Celgene give up on selling at monopoly rates to a part of the Revlimid market (7%, 12%, etc.) unless there was a solid reason to do so?

No rational drugmaker, Actavis’ logic goes, would make a good-sized cash payment without some sturdy-enough rationale --- perhaps related to concerns as to whether its patent would hold up if it did not get to a settlement.

And by that same logic, no rational drugmaker would opt, as Celgene did here, to cut itself out of monopoly returns on 7% or 12% of the multi-billion dollar market it had created --- unless it, too, had a reason for doing so. Per Actavis, concerns Celgene had about the strength of the underlying patents could well have been that reason.

2. King Drug

Where things stand.

As a matter of common sense and economics, the Celgene/Natco settlement agreement plausibly transferred value to Natco --- by allowing it to charge high prices on its generic Revlimid. See Part VIII.B.

As a matter of law, that kind of transfer of value can check the “pay” box.

First, pay-for-delay doctrine aims to address the problem of consumers improperly being charged artificially high prices because of a settlement. And that is precisely what happens when a generic drugmaker (here, Natco) is compensated for its delayed market entry precisely by being allowed to charge high prices for its generic drug. See Part VIII.C.1.a.

Second, counting as a “pay” an agreement that compensates the generic drugmaker by allowing it to charge high prices --- that would be consistent with how the pay-for-delay doctrine is supposed to work. Doing so does not spin off a “turducken” problem. See Watson Pharms., 677 F.3d at 1315; see Part VIII.C.1.b. And doing so hones in on a “pay” that imposed felt costs on Celgene. Opting to forego monopoly returns on 7% or 12% of the $7 billion Revlimid market would have been a pain-point for Celgene, and therefore calls for an explanation. See id.

The Court's as-a-matter-of-law conclusion is also supported from a third direction --- by the Third Circuit's King Drug decision.

To get there, start with some background.

* * *

A generic drugmaker that is first out of the gate to make a paragraph IV certification gets a 180-day period of exclusivity --- during which it alone can come to market with a generic version of the relevant drug. See 21 U.S.C. § 355(j)(5)(B)(iv).

During that 180-day period, the patentholder's brand-name drug can continue to be marketed. See In re Lamictal Direct Purchaser Antitrust Litig., 957 F.3d 184, 188 (3d Cir. 2020).

And the patentholder (or an entity it licenses) can also come to market with its own generic drug --- an “authorized generic.” See King Drug, 791 F.3d at 396.

In short, for 180 days the first-filing generic drugmaker gets a monopoly in the generic-drug segment of the market. But the monopoly can become a duopoly if the patent-holding drugmaker opts to jump in with a generic version of its own drug. See id. at 405.

Common sense predicts that the generic drugmaker can charge more during its 180 days if the authorized generic does not come to market.¹⁰⁹

And industry experience backs that up.¹¹⁰ One solid study is the Federal Trade Commission's from 2011. See Fed. Trade Comm'n, Authorized Generic Drugs: Short-Term Effects and Long-Term Impact (2011), https://www.ftc.gov/sites/default/files/documents/reports/authorized-generic-drugs-short-term-effects-and-long-term-impact-report-federal-trade-commission/authorized-generic-drugs-short-term-effects-and-long-term-impact-report-federal-trade-commission.pdf. It looked at more than 100 authorized-generic launches over nearly a decade, see id. at 7, and “applie[d] ․ sophisticated statistical techniques.” Id. at 2. “On average,” the report concluded, “the retail price of a typical generic drug during the 180-day exclusivity period is 86 percent of the pre-entry brand price without [authorized generic] competition and 82 percent of the pre-entry brand price when an [authorized generic] competes.” Id. at ii-iii.

The takeaway: generally higher prices for generics without an authorized generic; generally lower prices with one.

* * *

Come now to King Drug.

There, the Third Circuit held that it counts as a “pay” for the purposes of the pay-for-delay doctrine when the brand-name drugmaker and a generic drugmaker agree in a settlement that the brand-name company will not launch an authorized generic. See 791 F.3d at 403.

* * *

There is a tight-enough analogy between King Drug and this case.

* * *

First, a no-authorized-generic settlement agreement (as in King Drug) conveys value to the settling generic drugmaker by restructuring the market. It does this by requiring a main player, the authorized generic, to keep off the field --- away from the action, forbidden from competing.

The Celgene/Natco settlement agreement also plausibly restructures the market. And it does so in a roughly similar way --- by requiring key players to stay out of bounds (or at least by strongly encouraging them to stay out of bounds).

Celgene's most natural competitor --- Natco, the first-filer --- would be disabled by the settlement agreement from competing as to large swathes of the Revlimid market. 93% of it, 88% of it, etc.

Moreover, the Celgene/Natco settlement agreement plausibly made it likely that other generic drugmakers would be in essentially the same boat --- able to sell their product, if at all, only to relatively modest slices of the market.¹¹¹

And finally, the Celgene/Natco agreement plausibly made it likely that no authorized generic would be released. The same outcome as in King Drug, though arrived at using different means --- a bar on an authorized generic (in King Drug) versus strong incentives not to launch an authorized generic (in this case).¹¹²

In sum:

The King Drug settlement remade the market by keeping an obvious possible competitor (an authorized generic) from competing. Here, the Celgene/Natco settlement would plausibly remake the market in the same way. By forbidding a competitor (Natco) from competing over 93% or 88% etc. of the market. And, through incentives, by making it plausibly unlikely that other possible competitors would get into the action --- (i) an authorized generic, and (ii) other generic drugmakers.

* * *

A second aspect of the analogy: the market restructuring in King Drug and the market restructuring in this case would each have similar effects.

With no authorized generic entering the market, the King Drug generic drugmaker would likely be able to charge higher prices for its generic.¹¹³

Same effect here. With other competitors plausibly unlikely to swamp the market with generic Revlimid,¹¹⁴ the settlement agreement's volume caps would allow Natco to charge high prices on its generic Revlimid.

And in each case, King Drug and this one, the relevant transfer of value would come at the expense of consumers, and twice over. Consumers would not get the lower prices that might have been on the way --- if the litigation had run its course, and ended in a win for the generic drugmaker. And prices would also be affirmatively propped up --- because the settlements (in King Drug and here) put in place a market structure that would have the predictable effect of keeping prices especially high.¹¹⁵

Moreover, the pricing-impacts resemblance between King Drug and this case cannot be dismissed as marginal --- an overlap as to an aspect of things that does not go to the heart of the matter.

After all, a concern for settlement agreements that keep drug prices high for consumers was a main part of the Supreme Court's Actavis logic. See Part VII.C.1.a. And it is at the core of King Drug’s thinking, too.¹¹⁶

In addition, it is a central basis for the analogy between King Drug and this case. Each involves a settlement agreement that plausibly keeps drug prices high. Passively --- by delaying generic entry. And actively --- by shaping the extent to which various players can participate in the market, so that even after delayed generic entry prices stay up, to the benefit of the settling drugmakers, but at the expense of consumers.

* * *

Bottom line: the analogy between King Drug and this case is close enough.

Given this, because a no-authorized-generic settlement agreement counts as a “pay” for pay-for-delay purposes, as King Drug held, the settlement agreement here should count as a “pay,” too.

D. Counterargument

To this point, the Court has held that the settlement agreement could plausibly allow Natco to sell generic Revlimid at especially high prices, and that can count as a matter of law -- - as the “pay” that is a necessary element of a pay-for-delay claim. See Part VIII.B, Part VIII.C.

Against the second part of this --- the as-a-matter-of-law holding --- the Defendants press a counterargument.

Natco, the argument goes, would have been allowed to come into the Revlimid market with no limits on how much it could sell, because early-entry-alone settlements never violate the antitrust laws. See Part II.A.4. So why would Natco not be allowed to come into the market in a volume-capped way? After all, “ ‘a volume cap could only have taken value away from Natco’ compared to unlimited entry.” See Defendants’ Reply at 2 (quoting In re Revlimid, 2024 WL 2861865, at *61) (cleaned up). If the greater (Natco selling all the market demands) is lawful --- then the lesser (Natco being allowed to sell less than the market demands --- 7% or 12% etc.) must be lawful, too. See id. at 2; Transcript at 15:23.

* * *

But this counterargument runs aground on the complaint.

The Plaintiffs have alleged that the value to Natco of its volume-capped entry was much greater than its early-entry-alone would have been. See Complaint ¶¶ 326-28, 332. And these allegations are sufficiently detailed. See id.

So the Defendants’ greater/lesser counterargument gets no solid traction here. Because, as alleged, this is not a greater/lesser situation.

* * *

Putting that aside, could a greater/lesser counterargument potentially hit the mark here, on different factual allegations?

Maybe not. Take three possible difficulties with the counterargument.

First, it seems to run two things together.

Whether the generic drugmaker has received a large enough transfer of value (“a volume cap could only have taken value away from Natco’ compared to unlimited entry”). And whether something --- here, a generic's ability to charge higher prices --- can count as a transfer of value at all.

But why should the one (a quantitative question, about how large the payment must be) determine the other (a qualitative question, about the form of payment that can count)?

Second, pay-for-delay doctrine asks whether a payment is large enough based on its absolute size or relative to litigation costs saved. See Part VI.

The Defendants’ proposed approach would add a new factor --- an inquiry into how a payment lines up size-wise against the value of early entry.

If this were the law, there would presumably be some cases in which a payment is large enough under the established criteria (absolute size/size relative to litigation costs) --- but not large enough when measured against the value of early entry.

But no Third Circuit authority is cited for tightening up the doctrine in this way.

And a comparison of (i) value conveyed to (ii) the value of early-entry-alone --- that would be hard to make work. This is because early entry is not a fixed measuring stick, against which the worth of something else can be consistently compared. The value of early entry itself moves around. A 2030 settlement on a patent due to expire in 2040 might allow early-entry-alone in 2032 (lawful) or 2036 (also lawful). But these are not worth the same thing. A 2032 early-entry-alone means eight extra years of generic sales. A 2036 entry means four. The first is much more valuable than the second. So if the question, as the Defendants contend, is how the value conveyed by a settlement stacks up against the value of early entry alone, there would just be another question to ask. Which early entry? And as to that question, the Defendants offer up no answer.

Third, the Defendants’ counterargument, as noted, is that because a settlement (early-entry-alone) that allows for a high volume of generic drugs to be sold is lawful, then a settlement (as here) that allows for a lower volume of generic drugs to be sold must be lawful, too.

But why does that follow?

At the core of pay-for-delay law is a focus on ensuring that monopoly-type prices are not improperly kept up for consumers. See Part II.A.4. So it makes sense that a settlement (early-entry-alone) that dissipates monopoly-type pricing ¹¹⁷ always passes muster under the antitrust laws. But why would that suggest that a settlement (as here) that plausibly helps to preserve monopoly-type pricing ¹¹⁸ also passes muster?

Put differently, under the Defendants’ counterargument a settlement (the one here) that plausibly sustains monopoly-type prices should be allowed under the antitrust laws --- because a settlement (early-entry-alone) that lowers monopoly prices is allowed under the antitrust laws.

But this gets the basic logic of antitrust backwards. That antitrust law shields agreements that lower prices does not provide a reason to think that antitrust law shields agreements that keep prices up.

IX. Pay: Would the Market Structure Stick?

To this point, the Court has held that the settlement agreement would plausibly allow Natco to charge higher prices, and this can amount to a “pay.” See Part VIII.

But the analysis until now has been tentative. It has rested on an assumption that has been repeatedly noted, but not yet tested.

The assumption: that in the wake of the Celgene/Natco settlement agreement, large amounts of generic Revlimid would not flood the market, so that the volume caps imposed on Natco would generate a dominant/fringe structure in the overall Revlimid market --- with (i) Celgene servicing the large bulk of the market with its brand-name Revlimid, and (ii) Natco servicing a small and capped slice of the market with its generic Revlimid (and maybe with a modest amount of generic Revlimid supply provided here and there by other generic drugmakers ¹¹⁹ ).

Is this assumption plausible? This Part answers: yes.

* * *

Before getting there, pause for a quick sense of why the cited assumption is an important building block of the Plaintiffs’ theory.

Say, as before,¹²⁰ that there are 100 Revlimid customers out there each year, and that Celgene has for years been selling its brand-name Revlimid to them for $20 per unit. Stipulate that generics usually go for around 1/2 of what a brand-name drug costs, and that it is now 2022. Natco (per the settlement agreement) now emerges onto the market and gets to sell seven units --- its 7% market allocation for 2022, keyed to prior years’ total Revlimid sales.

In this scenario, it is plausible that Natco would be able to price its generic Revlimid at only a modest bit below $20 --- not at the sort of much lower price that a generic would generally go for. See Part VIII.

But now tweak the picture. Imagine that 100 added units of generic Revlimid come into the market.

In that circumstance, it is almost surely the case that overall Revlimid prices would fall.¹²¹ And with falling prices, Natco would no longer be able to charge the especially high prices it had before --- the a-bit-less-than-$20 price that was plausible when the large bulk of the market (93% in 2022, 88% in 2023, etc.) was serviced by brand-name Revlimid going for $20.

In short, if non-Natco generic Revlimid were to meaningfully surge into the market, the volume caps would, as always, limit how many sales Natco could make. But there would be little reason to think that the volume caps would also carry with them another effect: generating the sort of overall market division ¹²² that plausibly opens the way to Natco charging especially high per-unit prices on its generic Revlimid.

* * *

Where might such a flood come from? Two main sources --- and per the Plaintiffs, each one was plausibly dammed up by the settlement agreement.¹²³

First, Celgene could itself pour generic Revlimid into the market --- in the form of an “authorized generic.”

But the Plaintiffs allege that the settlement agreement discouraged Celgene from going that way,¹²⁴ to the point that during the volume-capped period no authorized generic would likely be launched. See Complaint ¶¶ 292, 295, 309-14.

Is this plausible? Yes, the Court concludes, for reasons set out in Part IX.A.

Second, other drugmakers could try to produce generic Revlimid. Celgene owned the key Revlimid patents. But other drugmakers could still get into the market. By convincing a court that Celgene's Revlimid patents are invalid.¹²⁵ See Teva Pharms. USA, Inc. v. Sebelius, 595 F.3d 1303, 1305 (D.C. Cir. 2010). Or by striking a deal with Celgene to enter the market.¹²⁶

But the Plaintiffs have pled ¹²⁷ that background economic incentives made it likely that relatively fewer drugmakers would sue, if any. And for those that did sue, a key provision of the Celgene/Natco settlement agreement allegedly ¹²⁸ pushed against their getting anything other than small slivers of generic Revlimid market share in any settlement they (non-Natco generic drugmakers) might strike. So generic Revlimid from other drugmakers might enter the market. But if so a trickle, not a flood. Not enough to change the basic dominant/fringe market structure here, or the underlying pricing dynamics.

Is all of this plausible? Yes, the Court again holds, as explained in Part IX.B.

* * *

In short:

The settlement agreement plausibly provided Natco with a way to charge especially high per-unit prices for generic Revlimid. See Part VIII.B.

But that conclusion depends on the assumption that generic Revlimid (from (i) Celgene or (ii) non-Natco generics) would not swamp the market in game-changing volumes.

If that assumption is plausible, as the Plaintiffs contend,¹²⁹ the volume caps would likely help to spin-off a dominant/fringe overall market structure --- which would allow Natco to charge high prices.

If the assumption is not plausible, as the Defendants contend,¹³⁰ the volume caps would likely not help to produce a dominant/fringe structure, and would therefore not give Natco an ability to charge high prices for its generic Revlimid.

The Court's conclusion: the Plaintiffs have it right. This Part explains why.

A. An Authorized Generic?

It is plausible that Celgene, via an authorized generic, would not opt to flood the market with its own generic Revlimid.

To see why, tick through some quick background. See Part IX.A.1. After that, look to Celgene's (faint) incentives to launch its own generic, see Part IX.A.2, and then zoom in on a provision of the settlement agreement that would affirmatively deter Celgene from going the authorized-generic route. See Part IX.A.3.

1. Background

An authorized generic is “chemically identical” to the patentholder's brand-name drug --- but is marketed by the patentholder “as a generic rather than under the ․ brand.” Areeda & Hovenkamp, Antitrust Law ¶ 2046f2.

“Authorized generics, like other generics, are sold at a reduced price compared to the brand name drug.” Sanofi-Aventis v. Apotex Inc., 659 F.3d 1171, 1175 (Fed. Cir. 2011).

A patentholder can launch a generic at any point. See In re Lipitor, 855 F.3d at 135-36.¹³¹

But why would it?

Say that the brand-name drug retails for $20, and a generic version would go for around $10. If given a choice, many people will reach for the cheaper one.¹³²

This means that by launching its own generic drug, the patentholder may well hurt its own bottom line.

It will cannibalize its own revenues --- because it will cost itself a (lucrative) brand-drug sale for every (less lucrative) generic-drug sale it makes. And introducing a generic version of its brand-name drug can also pull down the price of the brand-name drug.¹³³

Given all this, most of the time the drugmaker will leave things as they are --- and not release a generic version of its brand-name drug.¹³⁴

Most of the time. But not always. In two main circumstances, the drugmaker will be more likely to go ahead and market a generic version of its own drug.

The next section explains.

2. Incentives

First, a patentholder might put out an authorized generic to cut losses.

Another drugmaker may be breaking into the market with a generic. And when it does, drug prices will likely fall. So the patentholder might think: “if people are going to be buying a cheaper generic anyway, better to have them buying mine --- and with prices due to fall anyway, I won't be undercutting my brand-name product too much by launching a generic.”

A hedging/loss-cutting strategy of this sort will sometimes make sense.¹³⁵

But a loss-cutting strategy would likely be a mis-fit here.

Under the Celgene/Natco settlement agreement, Natco would be entering the Revlimid market subject to tight volume caps.

With those limits, even as generic Revlimid (made by Natco) came online, prices for the brand-name version of Revlimid (made by Celgene) would plausibly stay high --- at monopoly or near-monopoly levels. See footnote 94.

And if prices for brand-name Revlimid could be kept up even after Natco's arrival on the scene --- it would make much less sense for Celgene to launch its own generic. That would just start to bring overall prices down --- and substitute some sales of Celgene-made brand-name Revlimid for sales of cheaper Celgene-made generic Revlimid.

Moreover, there would be little reason to launch an authorized generic to prevent Natco's generic Revlimid from snatching customers away from Celgene's brand-name Revlimid. Because under the settlement agreement Natco was stuck servicing only a small wedge of the Revlimid market --- 7%, 12%, etc.¹³⁶

* * *

Move now to a second main reason why a patentholder might look to launch an authorized generic --- to brush back other generic drugmakers, who might be thinking about trying to get into the market.

When a patentholder hears footsteps like this, it may choose to fire on its own position ¹³⁷ --- and come out with an authorized generic, to convince generic drugmakers that it is not worth it for them to try to push onto the market.

A preemptive “launch of authorized generic drugs deters and delays the entry of standard generic drugs.”¹³⁸ And authorized generics are sometimes put out for this reason, even as to “drugs that had not been subject to a patent challenge.”¹³⁹

But Celgene would have no reason to put out an authorized generic to scare off Natco from trying to come into the generic market. Because under the settlement agreement Natco was already coming in --- and on terms that Celgene could well be comfortable with.¹⁴⁰

* * *

In short, for a brand-name drugmaker like Celgene, coming out with an authorized generic is often counter-productive. But it might go that way, mainly (i) to cut losses or (ii) to deter other entrants. But those rationales would plausibly apply here only in a relatively weak way, if at all. So Celgene would start off with relatively low incentives to put out an authorized Revlimid generic.¹⁴¹

3. The Acceleration Provision

Take the incentives sketched out above --- and add to them a piece of the Celgene/Natco settlement agreement that tilted the scale further in the same direction, an acceleration provision that affirmatively disincentivized Celgene from launching an authorized generic.

Per this acceleration provision, if Celgene were to bring out an authorized Revlimid generic, Natco would be able to launch its generic Revlimid early --- from “the date on which any [authorized generic] is first sold.” License Agreement § 1.11(c).¹⁴²

And at that point, Natco would be able to enter the market with none of the volume caps that would otherwise apply. See id. § 5.1(b). Instead, Natco would be permitted to sell “unlimited quantities of each dosage strength ․ beginning on the accelerated [l]aunch [d]ate.” Id.¹⁴³

This provision plausibly reduces the likelihood that Celgene would opt to launch an authorized generic.

On balance, the settlement agreement would probably have seemed like a relatively good deal for Celgene.¹⁴⁴ The ability to service the very large bulk of the Revlimid market.¹⁴⁵ And a chance to do so while charging high monopoly-type prices ¹⁴⁶ --- and on a protected basis, without any possibility of its most natural competitor (Natco, the first-filer) grabbing too many customers.

Because of the acceleration provision, a decision by Celgene to come out with an authorized generic would mean trading all of this for an immediate and uncapped Natco entry --- which would likely bring down prices across the board, and also take some sales away from Celgene.

It is certainly plausible that Celgene would not opt for that trade. Because that trade would mean exchanging high prices for lower ones.¹⁴⁷ And because that trade would mean swapping all or virtually all (100%, 93%, 88%, etc.) of the market for less of it.¹⁴⁸ See Complaint ¶ 309.

* * *

The Court's conclusion: given the combined effect of Celgene's background incentives ¹⁴⁹ and the acceleration provision ¹⁵⁰ --- it is plausible that Celgene would not opt to put out an authorized Revlimid generic.¹⁵¹

B. Other Drugmakers?

It is plausible, the Court has held, that generic Revlimid would not surge into the market via a release of an authorized generic after a settlement.

What about from the other possible source, from generic drugmakers other than Natco?

This section takes up that question.

1. Incentives

The Defendants contend that a deal between Celgene and a non-Natco drugmaker could flood the market with generic Revlimid. See Defendants’ Brief at 9-11, 18, 20-23.¹⁵²

But this argument starts off in the hole.

As alluded to above, see Part IX.A, the basic arrangement set into place by the Celgene/Natco settlement agreement would likely have seemed like a relatively ¹⁵³ good one from Celgene's perspective.

That set-up would plausibly let Celgene keep charging monopoly-type prices on its brand-name Revlimid.¹⁵⁴ It would let Celgene service the entire Revlimid market for around six post-settlement years, plus the very large bulk (93%, 88%, etc.) of the overall Revlimid market for years after that. And all of this with protection --- with no way for Celgene's obvious competitor, Natco, the first-filer, to surge up the charts (above 7% or 12%, etc.) to snatch more sales from Celgene and by doing so to bring down prices on Celgene's brand-name Revlimid.

Celgene could conceivably opt to kick over this applecart. It could choose to cut a settlement deal that gave another generic drugmaker a large slice of the Revlimid generic market. And if that slice were large enough, that could change the overall market structure --- and with it, the basic pricing dynamic. What had been a small (and price-taking) fringe (made up of just Natco) could become too big to work like a fringe. Together, the new generic drugmaker and Natco could begin collectively working like bona fide Celgene competitors --- eating into Celgene's brand-name sales (beyond 7% or 12% etc.) and pulling down across-the-board prices (including as to brand-name Revlimid).

But why would Celgene settle with another generic drugmaker in a way that would allow all that to happen?

There are possible answers.¹⁵⁵

But it is plausible that Celgene would not let the above happen --- that it would not agree to settle with another generic drugmaker on terms that would unwind the beneficial financial arrangement that it had built for itself with the Celgene/Natco settlement agreement.

And plausible is all that any of this needs to be at this point. See Ashcroft v. Iqbal, 556 U.S. 662, 678, 129 S.Ct. 1937, 173 L.Ed.2d 868 (2009) (noting that the plausibility standard is not akin to a “probability requirement”); see also Cornish-Adebiyi v. Caesars Ent., Inc., 184 F.4th 172, 182 (3d Cir. 2026) (“The plausibility standard at the pleading stage is not a probability requirement.”) (internal quotation marks omitted); Doe v. Princeton Univ., 30 F.4th 335, 344 (3d Cir. 2022).

2. The 1.5x Provision

Now add consideration of a provision in the settlement agreement that plausibly adds an extra push in the same direction as the incentives described just above.

To see how this provision --- call it the “1.5x provision” --- works, take an illustration.

Imagine that 100 units of Revlimid are sold each year, and that there is no generic yet on the market. And say that a unit of Revlimid cost nothing to make --- so that if a unit of Revlimid sells for $10, the company that sells it can expect to earn $10 in profits.

Now hypothesize that in 2022 generic drugmaker ABCD sues Celgene, and indicates that it would settle for $60. Say that Celgene is on board. It wants to get the deal done. How would it deliver $60 in value to ABCD to push things across the finish line?

One way would be for Celgene to give ABCD 6% of the 100-units-per-year Revlimid market. At $10 per unit of Revlimid, that would do it. A 6% slice of the market would deliver $60 in settlement value to ABCD.

On the flipside, a 6% settlement would also cost Celgene around $60 ¹⁵⁶ --- because the six units allocated to ABCD are six units that Celgene cannot itself sell.¹⁵⁷

Now change the picture a bit, by accounting for the 1.5x provision of the Celgene/Natco settlement agreement.

Under that provision, if Celgene were to cut a deal with another generic drugmaker that gives it a share of the Revlimid market, Natco would sometimes have to be topped up --- enough to ensure that its piece of the Revlimid market is equal to at least 1.5x of what the settling entity is getting. See Complaint ¶ 297; see also License Agreement § 5.1(a).

The Plaintiffs contend that this provision increases the likelihood that any non-Natco generic drugmaker that received an allocation of Revlimid would get only a relatively small slug.¹⁵⁸

And that makes sense.

In the example, the ABCD 6% settlement is for 2022. And that year, Natco was allowed under the Celgene/Natco settlement agreement to service 7% of Revlimid's market share.

Keeping Natco at 1.5 times the 6% that ABCD would get --- that would require bumping Natco up from 7% to 9%. And those extra two percentage points of market share would run Celgene $20 --- because each extra unit of Revlimid it might allocate to Natco is a $10 unit it cannot sell for itself.

To summarize:

Without the 1.5x provision in the Celgene/Natco agreement, Celgene could settle with ABCD by giving up 6% in market share --- and that would run Celgene $60.

With the 1.5x provision, Celgene could settle with ABCD by giving up 8% in market share (6% for ABCD plus 2% for Natco) --- and that would cost Celgene the same $60, plus $20 extra to level up Natco's market share.

Bottom line: Celgene's money would go further if Celgene opted to settle cases with other generic drugmakers using only small pieces of Revlimid market share --- too small to trip the 1.5x wire.

For 2022, for example, a settlement of 6% or above would cost Celgene a disproportional amount --- because, as set out above, Celgene would need to build in the extra expense of topping up Natco.

But for the same year, a small settlement of, say, 3% --- that would create no similar issue. 3% times 1.5 is 4.5% --- less than the 7% Natco was allocated for 2022, so no need for extra spend to bring Natco up.

Given all of this, for any settlement deal it might aim to cut, Celgene would plausibly have strong reasons not to offer anything but relatively small pieces of market share.¹⁵⁹

The 1.5x provision ¹⁶⁰ plausibly spins off powerful financial incentives for Celgene to offer up only small market-share settlements, not larger ones.

And that makes it even less likely that Celgene would cut a settlement with a generic drugmaker, and grant it a large wedge of market share.

* * *

Maybe, though, Celgene would enter into a parade of small settlements, each below the 1.5x line --- and these would collectively add up, such that generic Revlimid would flood into the market in meaningful volumes, a dominant/fringe overall market structure would not take hold and stick, and the Celgene/Natco settlement agreement would therefore no longer give Natco an ability to charge high generic prices.

The Defendants essentially argue this. Consider their argument now.

3. The “Genericization” Counterargument

Per the complaint, a dozen drugmakers were lining up to try to develop their own generic Revlimid around when Natco settled with Celgene. See Complaint ¶ 265; see also Defendants’ Brief at 20 (citing Complaint ¶ 265); Transcript at 22:23-25.

The Defendants hypothesize that settlements with each of these drugmakers --- settlements just up to where the 1.5x switch would be flipped, but not beyond --- could add up within Natco's first year on the market to 63%. See Defendants Brief at 21; Transcript at 28:21-29:2, 31:10-12; see also id. at 31:2-6, 31:21-32-6.

At that point, the argument goes, incentives would tip and it would become worth it for Celgene to launch an authorized generic. See Defendants’ Brief at 20-21 (citing Complaint ¶ 310 n.157). Within the first year of Natco's permitted launch (2022), the Revlimid market would fully “genericize,” and that would dilute away any transfer of value to Natco in the form of higher prices. See Defendants’ Brief at 22-23; see also Transcript at 75:16-19 (noting that, if Celgene settled with all 12 other generics up the maximum under the settlement agreement, “the market is totally genericized and Natco is getting no price protection at all”).

The Defendants’ legal papers include a table that depicts this scenario. See Defendants’ Brief at 21. Look to a boiled down version of it, zooming in on just the first year following Natco's March 2022 market entry date.

The chart shows that generic Revlimid would flood into the market. No overall dominant/fringe market division would come to fruition. And as a result Natco would not be able to charge especially high per-unit prices.

* * *

But this argument does not work.

Whatever else might be said about it,¹⁶¹ the Defendants’ argument heaps contingency on contingency.

Twelve generic drugmakers were allegedly out there. On the Defendants’ theory each of the 12 would have to have been ready to roll in 2022, having by then invested in the sort of well-developed and well-investigated arguments that can induce a settlement --- even though (as discussed below in Part IX.B.4), given Natco's first-filer advantage there would have been only diminished incentives for each drugmaker to make those sorts of investments.

And on the Defendants’ theory, each of the 12 generic drugmakers would need to have been not just ready in 2022 --- but ready with especially strong arguments. So strong that, in the face of them, Celgene would not only be willing to settle with each of the 12 drugmakers, but also willing to settle with each of them up to the topmost limit of the 1.5x multiplier.

This even though Celgene would have had plenty of reason to hold the line in settlement negotiations at least a few times --- a fully “genericized” market presumably being something it would badly have wanted to avoid.¹⁶²

And this even though for any given generic drugmaker in settlement negotiations, the threat of following through to the end, of litigating all the way to a courtroom victory --- that would have been relatively hollow, because a courtroom win would only have opened the way to Natco, already on the market, to immediately start operating with no volume limits.¹⁶³

Could all of this have happened? Maybe. Everything would have to break right, and everything sometimes does.

But the Defendants’ maybe-it-would-happen-that-way argument does not displace the Plaintiffs’ allegations as to how things would plausibly go. The Plaintiffs’ allegations do not need to be more than plausible for now. And they certainly do not need to cross off the list the seemingly low-likelihood alternative possibility sketched out by the Defendants, of full within-a-year “genericization” of the Revlimid market. See Bell Atlantic Corp. v. Twombly, 550 U.S. 544, 556, 127 S.Ct. 1955, 167 L.Ed.2d 929 (2007) (the plausibility requirement “does not impose a probability requirement at the pleading stage”); see also In re Ins. Broker Antitrust Litig., 618 F.3d 300, 341 n.42 (3d Cir. 2010) (explaining that there is no “require[ment] as a general matter that the plaintiff plead facts supporting an inference of defendant's liability more compelling than the opposing inference”).¹⁶⁴

4. Later-Filers

Move now to another cluster of incentives.

These further weaken the Defendants’ “genericization” counterargument, the one discussed in the preceding section. And they put more wind in the sails of the Plaintiffs’ contention that generic drugmakers other than Natco would not flood the market with generic Revlimid.

* * *

The “Hatch-Waxman [Act] provides a special incentive for a generic [drugmaker] to be the first to file” a paragraph IV certification. See Actavis, 570 U.S. at 143, 133 S.Ct. 2223; see also Scott Hemphill & Mark A. Lemley, Earning Exclusivity: Generic Drug Incentives & the Hatch-Waxman Act, 77 Antitrust L.J. 947, 947 (2011) (describing “a carrot, encouraging generic firms to challenge and invalidate bad patents ․ early and often”).¹⁶⁵

The first-filing drugmaker's “special incentive” is this --- if it can get onto the market and start selling its drug, it gets “a period of 180 days of exclusivity.” Actavis, 570 U.S. at 143, 133 S.Ct. 2223 (citing 21 U.S.C. § 355(j)(5)(B)(iv)).

During that period, the FDA cannot approve any other drugmaker's paperwork. See 21 U.S.C. § 355(j)(5)(B)(iv)(I). So while the 180-day clock ticks, the first-filing generic drugmaker will have to compete only with the patent-holding drugmaker --- its brand-name product or any authorized generic it might launch.¹⁶⁶ See Part IX.A. No other generic drugmakers can get going.¹⁶⁷

In short, the Hatch-Waxman Act's 180-day exclusivity period “effectively grants the first filer a temporary monopoly over the generic market.” In re Lamictal, 957 F.3d at 188.¹⁶⁸

This is typically a very lucrative prospect. It is usually the case ¹⁶⁹ that “the vast majority of potential profits for a generic drug manufacturer” come from these 180 days. See Actavis, 570 U.S. at 144, 133 S.Ct. 2223 (cleaned up).¹⁷⁰

In short, the 180-day exclusivity period works as a kind of “bounty.” C. Scott Hemphill, Paying for Delay: Pharmaceutical Patent Settlement as a Regulatory Design Problem, 81 N.Y.U. L. Rev. 1553, 1560 (2006). And for most generic drugmakers, it is a rich one.

But “only the first generic firm to challenge a [brand firm's] patents has any prospect of earning the bounty.” Id.

The reward is all-or-nothing. There is a gold medal for finishing first, but nothing for the runner-up. The exclusivity period created by the Hatch-Waxman Act is used up by the first-filer. And once it has been, no leftover Hatch-Waxman incentive remains behind for anyone else.

* * *

With this background in mind, come back now to this case.

Recall that as to the relevant Revlimid strengths Natco was the first-filer.

And recall the Defendants’ suggestion ¹⁷¹ that a progression of non-first-filing drugmakers might (i) sue Celgene and then (ii) extract strong settlements, one after another, up to the 1.5x line but not over it --- and in doing so would flood the market with generic Revlimid.

* * *

But point (i) is not very forceful.

Later filers have diminished incentives to actually go to the expense of kicking off a patent litigation. By choosing to litigate, they would be going down a very expensive road ¹⁷² --- and with a markedly smaller reward at the end of it for winning. “[L]ater filers generally have much less incentive to challenge a brand-name drug patent.” Hemphill, An Aggregate Approach to Antitrust, 109 Colum. L. Rev. at 635; see also Impax Lab'ys., 994 F.3d at 489 (describing “the reduced incentive a subsequent generic has to challenge the brand marker's patent”).

Pared back incentives to sue mean a reduced likelihood that (as the Defendants contend) a large progression of generic drugmakers would put together a complaint and file a lawsuit.

* * *

And point (ii) is not especially strong, either --- because it stands to reason that those later-filers that actually did choose to bring a lawsuit would be a bit less likely to be able to get for themselves the best settlements.

Settlements, after all, are not given away. To build leverage for a favorable settlement, a drugmaker ¹⁷³ will often have to invest. In a powerful draft expert report, for example, for passing across the conference table to convince the patentholder that it should cut a deal now, rather than fight things out in court to the end.¹⁷⁴ But this costs. Finding, hiring, and consulting with experts (and lawyers). Gathering facts, surveying the scientific literature, doing legal research, writing things up. None of this is free.

A later-filer's ultimate upside for winning its case is much lower on the back end. So it stands to reason that it will be somewhat less likely to spend heavily on the front end --- on the sorts of things ¹⁷⁵ that can help to generate favorable settlements.

And in a similar vein, note that a drugmaker's negotiating-session threat that it will litigate all the way through to a courtroom victory if it does not get what it wants --- that threat is less persuasive when it comes from a later filer.

After all, all-the-way-to-the-end court costs are especially high.

And for a later-filer, the reward on the other side of bearing those costs would not be especially rich. For a later-filer, there is no 180-day exclusivity. Win in court, and Natco, the first-filer, would come into the market.¹⁷⁶ And not only that. Natco would come into the market with 180 exclusive days to itself.¹⁷⁷

High costs, lower benefits. This is hardly a recipe for litigating all the way to the end of a case. And that dynamic is widely understood ¹⁷⁸ ---therefore, it would presumably be relatively difficult for a later-filer to credibly threaten to take things all the way to a final courtroom verdict. And the diminished power of that threat would likely make it a bit less likely that any such threat could be effectively used as a bargaining tool for extracting a top-shelf settlement.

5. The Assurance Counterargument

It is plausible, the Court has concluded, that, in the wake of the settlement agreement, the market would not be swamped by large volumes of generic Revlimid.

First, it is plausible that Celgene would not launch an authorized generic. Because of the combined impact of Celgene's background incentives, see Part IX.A.2, and the settlement agreement's acceleration provision. See Part IX.A.3.

And second, it is plausible that generic drugmakers other than Natco would not strike deals with Celgene to pour generic Revlimid into the market. Because of Celgene's incentives not to cut those sorts of deals. See Part IX.B.1. Because the agreement's 1.5x provision would encourage Celgene to offer only small slivers of the generic Revlimid market. See Part IX.B.2. And because non-Natco generic drugmakers would be later-filers, and so (i) fewer of them would likely step up to the plate to actually litigate, and (ii) those litigations that did get brought would be somewhat less likely to generate strong settlements. See Part IX.B.4.

Because it is plausible that generic Revlimid would not flood into the market, it is plausible that the Natco volume caps would generate an overall market structure shaped along dominant/fringe lines --- in the context of which Natco would be able to charge especially high per-unit prices on its generic Revlimid.

* * *

Against this conclusion, the Defendants invoke “assurance” --- the idea that Natco needed to be assured that things would turn out as sketched out above. If there was no such assurance --- if generic Revlimid might flood into the market and accordingly might ultimately undo Natco's ability to charge especially high generic prices --- there could be no “pay,” and therefore no viable pay-for-delay claim.¹⁷⁹

This argument starts off on the back foot.

“[A]ssuredness” is not the pleading standard. See Plaintiffs’ Opposition at 26. The law “require[s] only plausibility, a standard not akin to a probability requirement.” In re Lipitor, 868 F.3d at 254 (cleaned up). And certainly not akin to an “assuredness” requirement --- which asks for even more.

“[I]t is inappropriate to apply [the] plausibility standard with extra bite in antitrust and other complex cases.” W. Penn Allegheny Health Sys., Inc. v. UPMC, 627 F.3d 85, 98 (3d Cir. 2010); see also Kerwin v. Casino, 802 F. App'x 723, 725 (3d Cir. 2020). And the Third Circuit has “rejected” application of a “heightened ․ pleading standard” in reverse payment cases. AbbVie, 976 F.3d at 354 (cleaned up).

But “assuredness” is just that. A raised bar, set higher than our law requires.

* * *

In addition, it is not crystal clear what the Defendants mean by assurance.

Maybe that the terms of the payment here --- high per-unit prices for Natco, as a by-product of the settlement agreement's volume caps --- needed to have been more explicitly spelled out.¹⁸⁰

But why? The practical meaning of the 93/7, 88/12 market split, the high per-unit Revlimid prices that it would plausibly allow Natco to charge --- none of that is especially tough to see. Let alone for the two sophisticated entities that struck the deal.

And the Sherman Act bakes in no clear-statement rule --- no obligation that the parties to an allegedly unlawful agreement explicitly lay out the terms of that agreement.¹⁸¹ So why should that be required here?

* * *

Alternatively, the Defendants’ assurance argument may be focused on the idea that the Celgene/Natco settlement agreement was only “ ‘between Celgene and Natco,’ and thus could not ‘bind any other generic manufacturer.’ ” Defendants’ Brief at 10 (quoting In re Revlimid, 2024 WL 2861865, at *60). Generic drugmakers other than Natco were not contractually compelled to stay out of the Revlimid market by the settlement agreement --- because they were not party to it.

But commercial entities A and B routinely make agreements that impact B's dealings with C --- even though C is not itself bound to the A/B agreement. Think of most-favored nation clauses. Rights of first refusal. Poison pills. Non-compete agreements. Etc. These are everyday sorts of contractual provisions. In the context of them, the impact on B's dealings with C is not treated as too speculative to count. Quite the opposite. These sorts of provisions are bargained over, often fiercely --- precisely because their impact on C is not abstract. It is real, and widely understood to be. And when these sorts of provisions are breached, courts calculate damages. They do not throw up their hands because the value of the provisions is too amorphous to pin down.

An A/B impact on C is understood as a real-enough and solid-enough thing when it comes to, say, a most-favored-IP-licensee agreement between A and B. So why should different rules apply in the antitrust context --- when A and B (Celgene and Natco) enter into a contract that plausibly impacts C (other generic drugmakers)?

And if a different rule did apply, that would “elevate[ ] form” --- which companies’ names are on the signature page --- “over substance[,]” and would allow “companies ․ [to] avoid liability for anticompetitive reverse payments simply by structuring them” in this way or that. AbbVie, 976 F.3d at 358. “In effect, Actavis would become a penalty for bad corporate lawyering instead of anticompetitive conduct.” I_d. A global settlement agreement might be a problem. But a single settlement agreement with global effects would not be.

The Celgene/Natco settlement agreement plausibly had an impact on third parties. By, for example, making it clear that even if threatened with a strong case --- Celgene would be less likely to offer up too much Revlimid market-share to a generic drugmaker, because of the incentives created by the 1.5x provision.

Antitrust law routinely reasons from the incentives that a course of action creates. See, e.g., Burlington Indus. v. Milliken & Co., 690 F.2d 380, 392-93 (4th Cir. 1982) (discussing Am. Soc'y of Mech. Eng'rs, Inc. v. Hydrolevel Corp., 456 U.S. 556, 102 S.Ct. 1935, 72 L.Ed.2d 330 (1982)); Freeman v. S.D. Ass'n of Realtors, 322 F.3d 1133, 1152 (9th Cir. 2003). There is no reason for that to be limited to the incentives that are created between contracting parties.

* * *

Another way to think of the Defendants’ “assurance” counterargument is as a substantive principle of antitrust law --- a principle that a pay-for-delay payment (here, Natco's ability to charge high prices) must be a sure thing, not just a plausibly likely thing. See, e.g., Defendants’ Reply at 3 (“under Actavis, [the] Plaintiffs must prove payment for delay”) (cleaned up); id. (“an actual payment ---not merely a possibility of one --- must be plausibly alleged”); Transcript at 25:18-25 (“the Third Circuit does not require certainty at the pleading stage as to the size of ․ payment ․ but absent a guarantee or a promise or an assurance to Natco,” the Plaintiffs’ claims must be dismissed).

But reasoning from a likely market impact is not forbidden in antitrust law. It is routine. See, e.g., Topco Associates, Inc., 405 U.S. at 602, 608, 92 S.Ct. 1126 (concluding that an agreement allocating territory to grocery stores “who do no actual business in those areas on the theory that they may wish to expand at some indefinite future time” was “a per se violation of [§] 1”); Quality Mercury, Inc. v. Ford Motor Co., 542 F.2d 466, 471 (8th Cir. 1976) (“The granting of a perpetual exclusive franchise is suspect under the Sherman Act, because changes in conditions, unforeseen when the exclusive franchise was granted, may combine to create an anticompetitive effect.”).

Indeed, in the pay for delay context, the Third Circuit has held that all that is necessary is an allegation that an agreement “unjustifiably transferred ․ an opportunity, and [that] the profits associated with the opportunity were large.” AbbVie, 976 F.3d at 359 (emphases added). Opportunity. A chance, not something that will necessarily come out a certain way.

Moreover, under the Third Circuit's King Drug decision, a no-authorized-generic provision counts as a “pay.” See 791 F.3d at 403. But the value of that sort of provision is not as locked-in as, say, a hand-over of a check. A no-authorized-generic provision is only likely to deliver value. On the assumption, for example, that when the generic launches, consumers will still be out there and still be interested in the drug.¹⁸² And on the assumption that the generic drugmaker will have the capacity to service meaningful parts of the market when the opportunity created by the no-authorized-generic provision materializes. A likelihood that value will be ultimately transferred. And, per King Drug --- that is enough.

Finally, note that the Defendants argue that the 2015 Celgene/Natco settlement agreement should be assessed ex ante, not ex post. In light of how things would have been expected to go at the time the settlement agreement was signed --- not how they actually went, looking back.¹⁸³ A going-forward look --- the ex ante approach the Defendants advocate for --- necessarily calls for a look to likelihoods and probabilities. So how can the Defendants then turn around and suggest that likelihoods and probabilities categorically come up short, that harder-edged assurances are needed?

* * *

To be sure, the Court's prior opinion spoke of the need for an “assurance.” See, e.g., In re Revlimid, 2024 WL 2861865, at *59-*60.

But by using this term the Court was not suggesting a new standard --- assurance, not plausibility. See Plaintiffs’ Opposition at 26 (“[t]he [D]efendants contort into a new pleading standard the Court's prior observation about ․ an ‘assurance’ ”).

“Assurance” was simply shorthand for the conclusion that it made no sense, on the Insurer Plaintiffs’ allegations then before the Court, to think that the volume caps alone would keep prices up.¹⁸⁴ And why did it not make sense? In part because the Court had not previously been pressed to address some of the settlement agreement provisions, see footnote 50, that have been critical here to establishing that generic Revlimid would not flood the market. Without being asked to reckon with those provisions, the Court held that it was not plausible to think that the settlement agreement's volume caps would be likely to keep generic Revlimid prices high.

The use of “assurance” in that discussion should not be overread. But that is what the Defendants’ argument does with it.

X. Conclusion and a Loose End

Wrap up here with a summary of the Court's analysis.

An antitrust pay-for-delay claim generally has four elements, see Part III, and the Court has concluded that there are plausible allegations as to each of these here.

* * *

As to delay and justification, those have not been put in play by the Defendants. See Part V.

* * *

The next element concerns the size of the payment. That box is plainly checked here.

For pay-for-delay purposes, a payment's size can be measured in relative terms --- stacked up against attorneys’ fees saved by settling. Or a payment can be assessed on its own. In absolute terms. By a kind of dead-reckoning sense of what is large enough.

But either way, whether the approach is relative or absolute --- an alleged multi-billion dollar value transfer, see Complaint ¶ 332, counts as large enough. See Part VI.

* * *

The final element is pay. Via the 2015 settlement agreement, did Celgene plausibly transfer value to Natco?

Yes, the Court has concluded.

The settlement agreement split up the market between a large player able to charge monopoly-type prices (Celgene) and a much smaller player (Natco) --- and that structure would plausibly have allowed Natco to draft just behind the leader (Celgene), and to charge especially high per-unit prices for its generic Revlimid, much more than a generic usually goes for.

Common sense shows that. See Part VIII.A. And the economics literature backs it up. See Part VIII.B.

Can the opportunity to charge especially high prices count as a “pay” as a matter of law?

Yes. That is consistent with the underlying rationale for the pay-for-delay doctrine. See Part VIII.C.1. And it is consistent with the Third Circuit's decision in King Drug. See Part VIII.C.2.

To be sure, all of this leans on the assumption that meaningful volumes of generic Revlimid would not flood into the market. That assumption is a key part of how the volume caps on Natco (7%, 12%, etc.) could plausibly generate a market that would be divided, overall, between (i) a dominant firm (Celgene) and (ii) a fringe (Natco, and maybe some other generic drugmakers supplying modest quantities of generic Revlimid).

Is this assumption plausibly correct?

Yes.

It is plausible that Celgene would not itself pour generic Revlimid into the market. Because of its overall economic incentives. See Part IX.A.2. And the acceleration provision in the settlement agreement. See Part IX.A.3.

And it is plausible that as to generic drugmakers other than Natco, Celgene would not in the end cut a deal or deals that would put those other generic drugmakers in a position to service a large chunk of the generic Revlimid market. Celgene would have financial incentives not to let that happen. See Part IX.B.1. The 1.5x provision would further deter Celgene from going that way. See Part IX.B.2. And as later-filers, generic drugmakers other than Natco would be pursuing a relatively limited upside --- which would weigh against their investing heavily in building the sorts of especially strong cases that can induce large settlements. See Part IX.B.4.

The Defendants’ counterargument, among others, is that assurance is needed here, not just plausibility. But that asks more of the Plaintiffs than our law does. See Part IX.B.5.

The bar here is not set at assurance. It is set at plausibility.

The Plaintiffs have cleared it, and so the Defendants’ motion to dismiss must be denied as to the parts of it (issues 1-4) that have been taken up here.

* * *

Before ending, a final note.

The Defendants have argued that in deciding this motion only some of the Plaintiffs’ allegations can be considered.

Only the ex ante allegations. Allegations as to the state of things as of 2015 when the Celgene/Natco settlement agreement was signed --- plus reasonable inferences as to what would happen next, as to what the settlement agreement would spin off in the post-2015 future. See Defendants’ Brief at 2, 20, 30; Defendants’ Reply at 6-8. That, per the Defendants, “is the relevant lens.” Transcript at 29:12.

But the Plaintiffs have suggested that the aperture should be opened wider. Consider ex ante allegations, yes. But also ex post allegations, they say --- as to the ways in which the Revlimid market actually developed after the 2015 agreement was struck. See Plaintiffs’ Opposition at 2; see also Transcript at 49:6-9.

There is no need to resolve this back-and-forth between the parties, at least for now.

Throughout this Opinion and Order, the Court has considered only the ex ante allegations --- and these add up to enough as to each element of the Plaintiffs’ pay-for-delay claim. So there is no reason to ask if considering more allegations, the ex post allegations, might further yeast up the Plaintiffs’ case.

Put differently, even assuming that the Defendants have it right and that the ex post allegations are out of bounds --- their motion fails.

* * *

But note: the parties’ ex post/ex ante debate may become relevant later. So sketch out its contours now.

The Plaintiffs’ ex post approach may get support from AbbVie, a Third Circuit pay-for-delay case. There, the court of appeals relied on the fact that the first-filing generic's “actual sales were much higher” than expected to conclude that the alleged reverse-payment was “plausibly ‘large.’ ” 976 F.3d at 357. This may arguably amount to an endorsement of at least some “post hoc analysis”¹⁸⁵ in reverse payment cases.

This said, AbbVie did not discuss the ex post/ex ante question.

And on the other side of the balance, in support of the Defendants’ ex-ante-only position, there seems to be a good deal of caselaw.

Pay-for-delay decisions from federal appellate courts. See Valley Drug Co. v. Geneva Pharm., 344 F.3d 1294, 1306 (11th Cir. 2003) (in a reverse payment case, zooming in on “the time the agreements are entered into”); Watson Pharms., 677 F.3d at 1308 (“a court must judge the antitrust implications of a reverse payment settlement as of the time that the settlement was executed”); Impax Lab'ys Inc. v. Fed. Trade Comm'n, 994 F.3d 484, 496 (5th Cir. 2021) (“[I]t is a basic antitrust principle that the impact of an agreement on competition is assessed as of the time it was adopted.”) (internal quotation marks omitted).¹⁸⁶

And from district courts, too. See, e.g., In re Zetia (Ezetimibe) Antitrust Litig., 2022 WL 4355149, at *24 (E.D. Va. Sept. 2, 2022) (“the court must approach that settlement from the viewpoint of the parties in May 2010, when they settled”), report and recommendation adopted, 655 F. Supp. 3d 406, 421 (E.D. Va. 2023) (“The valuation of the provision at the time of settlement is the proper inquiry.”).

Moreover, the ex ante approach seems to line up with broader principles of antitrust law. See, e.g., Polk Bros, v. Forest City Enters., 776 F.2d 185, 189 (7th Cir. 1985) (“A court must ask whether an agreement promoted enterprise and productivity at the time it was adopted.”); see also In re Cipro Cases I & II, 61 Cal.4th 116, 158, 187 Cal.Rptr.3d 632, 348 P.3d 845 (2015); Areeda & Hovenkamp, Antitrust Law ¶ 2046el.

And an ex ante approach helps to maintain the usual sharp division in our law between liability (which generally kicks in at a defined moment) and damages (which build from there). Keeping those two cleaved apart can be important. For example, it can be difficult to stay on the right side of the law if the liability question --- whether a person did something illegal at time one --- depends on how things develop at time two and three.¹⁸⁷

* * *

As indicated above, choosing between the ex ante approach and the ex post may be necessary down the road. But for now, the issue does not need to be taken on.

Because even doing only what the Defendants advocated for --- relying only on the ex ante allegations --- the Plaintiffs have pled enough.

They have put forward plausible allegations as to each element of the pay-for-delay claim with respect to the December 2015 settlement agreement between Celgene and Natco. As to Natco, the Defendants’ motion to dismiss must therefore be denied as to issues 1-4, the ones taken up here.

IT IS on this 29th day of September, 2026, SO ORDERED.

FOOTNOTES

1.   Because this case is at the motion-to-dismiss stage, the Court must give the Plaintiffs the benefit of all reasonable inferences and treat the complaint's allegations as true. See McTernan v. City of York, 577 F.3d 521, 526 (3d Cir. 2009). Which complaint counts? There are five here, filed on behalf of different parties. The complaints are closely similar, one to the next. Compare, e.g., Complaint ¶¶ 329-32, with Consolidated Amended Complaint (“Retailers’ Complaint”) (ECF 459) ¶ 310. And in their legal briefs, the parties mainly root their arguments in one of the complaints --- the one filed by certain insurance companies. See Memorandum of Law in Support of Defendants’ Joint Motion to Dismiss Plaintiffs’ Amended Complaints (“Defendants’ Brief”) (ECF 529) at 4 n.3; Consolidated Plaintiffs’ Opposition to Defendants’ Joint Motion to Dismiss Plaintiffs’ Amended Complaints (“Plaintiffs’ Opposition”) (ECF 533) at 1 n.1. The Court follows the parties’ lead --- and mainly looks to that complaint, the insurers’ complaint, at ECF 460. It is called the “complaint” in the text and the “Complaint” in the citations.

2.   In 2019, Bristol-Myers Squibb Company bought Celgene. See Complaint ¶ 32.

3.   “Revlimid” is the consumer-facing brand name Celgene gave the drug in question. Where necessary, the Court differentiates between brand-name Revlimid and the generic version of Revlimid. (The underlying drug, in its branded and generic forms, can also be referred to by the name it takes in the scientific literature --- lenalidomide. See Complaint ¶ 2. But that term might complicate things here, so the Court does not use it.)

4.   Patents are “an exception to the general rule against monopolies[.]” Walker Process Equip., Inc. v. Food, Machinery & Chem. Corp., 382 U.S. 172, 177, 86 S.Ct. 347, 15 L.Ed.2d 247 (1965) (internal quotations omitted); see also United States v. Line Material Co., 333 U.S. 287, 309-10, 68 S.Ct. 550, 92 L.Ed. 701 (1948); Standard Oil Co. (Ind.) v. United States, 283 U.S. 163, 169 n.2, 51 S.Ct. 421, 75 L.Ed. 926 (1931). The possibility of supercharged monopoly returns during the patent period is intended to work as a kind of magnet --- pulling companies forward through years of pricey research and development, attracted by the high profits during the patent period that await if things break right. In this sense, monopoly prices during the patent period aim to spur the “Progress of Science and useful Arts.” U.S. Const. art. I, § 8, cl. 8 (Patent Clause); see generally Sears, Roebuck & Co. v. Stiffel Co., 376 U.S. 225, 229, 84 S.Ct. 784, 11 L.Ed.2d 661 (1964); see also Helsinn Healthcare S.A. v. Teva Pharms. USA, Inc., 586 U.S. 123, 129, 139 S.Ct. 628, 202 L.Ed.2d 551 (2019); Bonito Boats, Inc. v. Thunder Craft Boats, Inc., 489 U.S. 141, 146, 109 S.Ct. 971, 103 L.Ed.2d 118 (1989).

5.   See 35 U.S.C. § 154(a)(2).

6.   A generic is “chemically equivalent to the approved brand-name drug.” Mut. Pharm. Co. v. Bartlett, 570 U.S. 472, 477, 133 S.Ct. 2466, 186 L.Ed.2d 607 (2013). It has “the same ‘active ingredient’ or ‘active ingredients,’ ‘route of administration,’ ‘dosage form,’ and ‘strength’ as its brand-name counterpart.” Id. (citing 21 U.S.C. § 355(j)(2)(A)(ii)-(iii)).

7.   See Michael A. Carrier, Innovation for the 21st Century 355 (2009); see also IMS Health Inc. v. Sorrell, 630 F.3d 263, 268 (2d Cir. 2010). One reason generics are cheaper --- their price does not need to account for the steep R&D costs it would have taken to develop the drug from scratch. See generally United States v. Generix Drug Corp., 460 U.S. 453, 455 n.1, 103 S.Ct. 1298, 75 L.Ed.2d 198 (1983); Rena M. Conti & Marta E. Wosińska, The Economics of Generic Drug Shortages: The Limits of Competition, 39 J. Econ. Persps. 79, 79 (2025); Sharat Ganapati & Rebecca McKibbin, Markups and Fixed Costs in Generic and Off-Patent Pharmaceutical Markets 4-5 (Nat'l Bureau of Econ. Rsch., Working Paper No. 29206, 2021); Aylin Sertkaya et al., Cost of Generic Drug Development and Approval 16-21 (E. Rsch. Grp., Task No. 0360.00.006, Dec. 31, 2021), https://aspe.hhs.gov/sites/default/files/documents/66e13df52a6d314ba4e10c8e55b68b4b/cost-of-generic-drugs-erg.pdf; Gregory Dolin, Do Patent Challenges Reduce Consumer Welfare?, 83 U. Chi. L. Rev. 256, 267 (2017).

8.   As to why a patent might be invalid, see, for example, 6 Chisum on Patents § 19.01 & n.2 (2026 ed.) and 6 Moy's Walker on Patents § 17:15 (4th ed. 2025).

9.   See 21 U.S.C. § 355(j)(2)(A)(vii)(IV).

10.   The Plaintiffs have pressed antitrust claims under both federal and state law, see Complaint ¶¶ 670-778, and the Defendants have moved to dismiss all of them. But the Defendants’ arguments focus only on federal antitrust law. See Defendants’ Brief at iii-v (table of cases in the Defendants’ brief listing out 37 federal cases --- and no state-law cases); Defendants’ Reply Memorandum of Law in Further Support of Joint Motion to Dismiss Plaintiffs’ Amended Complaints (“Defendants’ Reply”) (ECF 538) at iii-v (same). Given this, the Court assumes for now that federal and state antitrust law are not meaningfully different here. So the federal and state claims sink or swim together. If the federal antitrust claims survive, then the state ones do, too. If the federal claims fail, then the state claims also cannot work.

11.   “Plausible” because that is this the standard on a motion to dismiss. See Ashcroft v. Iqbal, 556 U.S. 662, 129 S.Ct. 1937, 173 L.Ed.2d 868 (2009); Bell Atlantic Corp. v. Twombly, 550 U.S. 544, 127 S.Ct. 1955, 167 L.Ed.2d 929 (2007).

12.   See, e.g., Lisa Larrimore Ouellette, How Many Patents Does it Take to Make a Durg? Follow-On Pharmaceutical Patents and University Licensing, 17 Mich. Telecomm. & Tech. L. Rev. 299, 300, 314-17 (2010).

13.   The Food and Drug Administration, a federal agency.

14.   For part of why generics are cheaper, see footnote 7.

15.   The patent laws treat the paragraph IV filing as itself an act of patent infringement. See 35 U.S.C. § 271(e)(2)(A). So there can be an infringement lawsuit before any allegedly infringing sales have been made.

16.   A valid patent gives the patentholder a lawful monopoly. See Actavis, 570 U.S. at 147, 133 S.Ct. 2223 (“ ‘[a] valid patent excludes all except its owner from the use of protected process or product’ ”) (quoting Line Material Co., 333 U.S. at 308, 68 S.Ct. 550); see also Albrecht v. Kinsella, 119 F.2d 1003, 1004 (7th Cir. 1941) (“[t]he owner of a valid patent has a legal monopoly”); footnote 4.

17.   See Admin. Off. of the U.S. Cts., Table C-4. U.S. District Courts --- Civil Cases Terminated by Nature of Suit and Action Taken During the 12-Month Period Ending March 31, 2016, at 1 (2016), https://www.uscourts.gov/sites/default/files/data_tables/fjcs_c4_0331.2016.pdf. During this period, less than 2% of patent cases ended after trial. See id.

18.   See also, e.g., Williams v. First Nat'l Bank, 216 U.S. 582, 595, 30 S.Ct. 441, 54 L.Ed. 625 (1910) (“[c]ompromises of disputed claims are favored by the courts”); Fed. R. Evid. 408 advisory committee notes (noting “the public policy favoring the ․ settlement of disputes”).

19.   Other types of settlements are categorically allowed, too. See Actavis, 570 U.S. at 152, 133 S.Ct. 2223; see also In re Lipitor Antitrust Litig., 868 F.3d 231, 250 (3d Cir. 2017). Those are not relevant here.

20.   Higher consumer prices are a classic antitrust concern. See Nat'l Collegiate Athletic Ass'n v. Alston, 594 U.S. 69, 81, 141 S.Ct. 2141, 210 L.Ed.2d 314 (2021); Ohio v. Am. Express Co., 585 U.S. 529, 541, 138 S.Ct. 2274, 201 L.Ed.2d 678 (2018).

21.   At least notionally.

22.   See generally Phillip E. Areeda & Herbert Hovenkamp, Antitrust Law: An Analysis of Antitrust Principles and Their Application ¶ 2046a (2025 ed.) (“a patentee and a patent infringement defendant have powerful economic reasons for wishing to cartelize the market covered by the patent rather than to permit competition to emerge”); Carrier, Innovation for the 21st Century at 347 (describing “the [settling] parties’ aligned incentives”).

23.   See In re Wellbutrin XL Antitrust Litig. Indirect Purchaser Class, 868 F.3d 132, 163 (3d Cir. 2017) (explaining that where there “[i]s no delay ․ the analysis in Actavis does not apply”). Note that the cases toggle back and forth. Sometimes using the word “delay.” And sometimes “early.” From the perspective of a 2030 settlement, coming to market in 2035 amounts to a “delay.” It is five years late. But switch the baseline and look at things from 2040, when the patents were set to expire on their own. From there, coming to market in 2035 is not five years late, but five years “early.” Two different words for the same 2035 entry. “Delay,” with a negative flavor. “Early,” with a positive one. But no matter. Nothing turns on whether a 2035 entry is called “delayed” (implicitly taking 2030 as the reference point), or “early” (taking 2040 as the standard).

24.   This is the term used by the Third Circuit. See King Drug, 791 F.3d at 405; In re Wellbutrin, 868 F.3d at 161-62; AbbVie, 976 F.3d at 359; see also C. Scott Hemphill, Paying for Delay: Pharmaceutical Patent Settlement as a Regulatory Design Problem, 81 N.Y.U. L. Rev. 1553, 1557 (2006) (using this term).

25.   A pay-for-delay settlement is sometimes “called a ‘reverse payment’ settlement[.]” Actavis, 570 U.S. at 141, 133 S.Ct. 2223. “Reverse” because these settlements seem to flip the normal order of things. When a typical case settles, the defendant pays the plaintiff. But in a settlement of a paragraph IV lawsuit, the payment runs the other way --- in “reverse,” from the plaintiff (the brand-name drugmaker) to the defendant (the generic drugmaker). See Actavis, 570 U.S. at 141, 133 S.Ct. 2223; In re Lipitor, 868 F.3d at 249-50. This seems backwards. But too much should not be made of it. The Hatch-Waxman Act created a mechanism for early teeing-up of disputes as to paragraph IV certifications, so they could be resolved in court before the generic drugmaker actually starts selling anything. The Act chose the brand-name drugmaker as the party to kick off the lawsuit, by serving as the plaintiff. See 21 U.S.C. § 355(j)(5)(B)(iii). But the statute could readily have assigned the generic drugmaker the plaintiff role. And if the Act had opted to go that route, the defendant (the brand-name drugmaker) would have been the one to make any settlement payment --- and there would be nothing about the direction of the settlement payment that would seem to be in “reverse,” against the typical grain. In an ordinary tort suit, who-sues-who is intuitive. The injured party (the person hit by the car) is the plaintiff. She sues the injuring party (the driver), and he becomes the defendant. In that context, it would be puzzling if the plaintiff were to pay a defendant. Why would a person knocked over by a car fork over money to the person that hit her? But when it comes to a paragraph IV suit, it is not obvious from the get-go who the injured party is. Maybe the brand-name drugmaker is the injured party. It worked to get a patent and now has to fend off a meritless challenge. Or maybe the generic drugmaker is injured. It cannot sell its drug, because it is being boxed out by an invalid patent. To see which of these is right --- to see who is the injured party in the first place --- the dispute-resolution process needs to get going. And to start it off, someone needs to be labeled the injured party, the plaintiff. That the Hatch-Waxman Act chose the brand-name drugmaker for this role does not suggest that any given brand-name drugmaker is, in fact, injured --- such that any settlement payment from a brand-name drugmaker is therefore in “reverse,” something intrinsically odd, something that should provoke some skepticism. In a paragraph IV lawsuit, a class of entities, brand-name drugmakers, are statutorily assigned the plaintiff's role even though it is not clear before the lawsuit gets off the ground whether they are injured. So the fact that a settlement payment runs from the plaintiff to the defendant in a paragraph IV case may not be especially telling --- as it would be, say, in a car-accident case, in which the party structure seems more natural, and it would therefore be hard to understand why the (injuring) defendant is paying the (injured) plaintiff.

26.   For who the Plaintiffs are here, see Part II.B.

27.   At least under the default terms of the settlements. Some provisions allowed for entry before 2022. See, e.g., Part IX.A.3.

28.   Celgene entered into 10 Revlimid-related settlements, each with a different generic drugmaker. See Complaint ¶¶ 284, 353, 407, 461, 479, 493, 510, 522, 537, 550.

29.   Walgreen Co.; Walgreens Specialty Pharmacy, LLC, LLC d/b/a Alliance Rx Walgreens Pharmacy; Kroger Specialty Pharmacy Inc.; and CVS Pharmacy, Inc.

30.   Cigna Corp.; Humana Inc.; Blue Cross Blue Shield Association; Health Care Service Corporation; Blue Cross Blue Shield of Florida, Inc.; Molina Healthcare Inc.; and United Healthcare Services, Inc. These entities (aside from United Healthcare Services, Inc.) are sometimes called “the Insurer Plaintiffs.”

31.   Jacksonville Police Officers and Fire Fighters Health Insurance Trust; Carpenters and Joiners Welfare Fund; Teamsters Local 237 Welfare Fund and Teamsters Local 237 Retirees’ Benefit Fund; and Teamsters Western Region and New Jersey Health Care Fund.

32.   Mayo Clinic; Lifepoint Corporate Services, General Partnership; and Intermountain Health Care Inc.

33.   Bristol-Myers Squibb Company. See footnote 2.

34.   Natco Pharma Limited; Teva Pharmaceuticals USA Inc.; AbbVie Inc.; Dr. Reddy's Laboratories Inc.; and Dr. Reddy's Laboratories Ltd. AbbVie, in a separate motion, disputes the extent to which it should be lumped in with these firms. See Memorandum of Law in Support of AbbVie Inc.’s Motion to Dismiss or Strike Plaintiffs’ Claims with Prejudice (ECF 530-1) at 1. That motion raises arguments not taken up here.

35.   AbbVie is a defendant in only three of the five complaints. See Complaint ¶ 36; Plan and Trust Complaint ¶ 21; Hospitals’ Complaint ¶ 38; see also United Healthcare's Complaint ¶ 22 n.7. And note that one of the Plaintiffs’ complaints omits Bristol-Myers Squibb Co. See Plaintiff United Healthcare Services, Inc's Memorandum of Law in Opposition to Defendant Celgene Corporation's Motion to Dismiss (Issue E) (ECF 537) at 1 n.1 (“[w]hile listed as a movant, Bristol-Myers Squibb Company is not named as a defendant”).

36.   Three things here. First, in addition to Celgene itself, certain other entities signed the agreement on the Celgene side of the house. See Complaint ¶ 285; see also Complaint, Exhibit B (“License Agreement”) (ECF 460-2). Second, “Natco” is Natco Pharma Limited, a generic drugmaker and one of the Defendants. See footnote 34. Along with Natco, there were three other entities that signed the settlement agreement on its side of the line: Arrow International Limited; Watson Laboratories, Inc.; and Allergan plc. See Complaint ¶ 285; Complaint, Exhibit A (ECF 460-1); Complaint, Exhibit D (ECF 460-4). Each of these allegedly had a relationship to Natco. See, e.g., Complaint ¶ 285 (describing Arrow as Natco's “exclusive U.S. licensee” for its generic Revlimid). Third, the settling parties “executed four interdependent [settlement] documents.” See Plaintiffs’ Opposition at 6. These are treated here as what they plainly were --- a single, overarching settlement agreement.

37.   The parties have keyed in on the Celgene/Natco settlement. See Defendants’ Brief at 8-36 (discussing “issues 1-4”). This said, there are some arguments in the legal papers that address Celgene's settlements with non-Natco entities. See Defendants’ Brief at 36-44 (discussing “issues 5-7”). Those are not taken up in this Opinion and Order. And there are arguments in the papers that could implicate the Celgene/Natco settlement but that are not taken up here either. See id. at 44-49, 59-60 (discussing “issues 8-9”). This Opinion and Order resolves and discusses only what the parties have called issues 1-4, each of which is directly related to the Celgene/Natco settlement agreement.

38.   Natco was the first drugmaker to file an application to launch a generic as to the relevant Revlimid strengths. See Complaint ¶ 240. The importance of this is discussed as it becomes relevant, mainly in Part IX.B.4.

39.   As to the especially steep costs of patent litigation, see, for example, Blonder-Tongue Lab'ys., Inc. v. Univ. of Ill. Found., 402 U.S. 313, 334, 91 S.Ct. 1434, 28 L.Ed.2d 788 (1971); View Eng'g, Inc. v. Robotic Vision Sys., Inc., 208 F.3d 981, 986 (Fed Cir. 2000); Am. Home Prods. Corp. v. Lockwood Mfg. Co., 483 F.2d 1120, 1122-23 (6th Cir. 1973); George J. Meyer Mfg. Co. v. Miller Mfg. Co., 24 F.2d 505, 507 (7th Cir. 1928); see also Am. Intell. Prop. L. Ass'n, Report of the Economic Survey 2023, at 62 (2023), https://fundamentalpatlit.com/wp-content/uploads/2024/02/AIPLA.pdf; Stephen Yelderman, Do Patent Challenges Increase Competition?, 83 U. Chi. L. Rev. 1943, 1966 (2016).

40.   The value to a plaintiff of its case depends on the odds that it will win in the end. See, e.g., Farrow v. U.S. Specialty Ins. Co., 755 F. Supp. 3d 550, 557 (D.N.J. 2024); see also Hacker v. Elec. Last Mile Sols. Inc., 722 F. Supp. 3d 480, 498-501 (D.N.J. 2024).

41.   Under the default terms of the settlement. Cf., e.g., Part IX.A.3.

42.   Set for 2016. See Complaint ¶ 643.

43.   These are the default terms of the settlement.

44.   With sales “calculated based on annual sales level of the single highest-level sales year from the prior five years of Revlimid sales.” Complaint ¶ 290; see also License Agreement § 1.13.

45.   Of 93/7 in the example, or something roughly analogous to it.

46.   The logic of this is unpacked in Part IX.

47.   More on this in Part IX.A.

48.   More in Part IX.B.

49.   For a procedural history up to that point, see In re Revlimid, 2024 WL 2861865, at *22–25.

50.   The core of the Plaintiffs’ argument is that the volume caps on Natco helped to divide the overall Revlimid market between Celgene and Natco --- which gave Natco an ability to charge high prices on its generic Revlimid. See Part II.C.2.b. But as noted, that theory can likely only work if generic Revlimid would not be expected to flood the market from non-Natco sources. Before, the Insurer Plaintiffs gave Judge Salas very little on that point. For example, the idea that non-Natco generic Revlimid would not swamp the market heavily depends on provisions of the settlement agreement that would tend to dissuade Celgene from (i) launching an authorized generic, see Part IX.A, or (ii) offering good-sized wedges of market share to other generic drugmakers. See Part IX.B. But these were not pressed by the Insurer Plaintiffs earlier in the case. See Plaintiffs’ Opposition at 2-3 (as to the first); Defendants’ Brief at 16 n.13 (as to the second). (To be sure, there was a passing reference to (ii) in a complaint that was not the subject of Judge Salas’ prior decision. See Amended Class Action Complaint (ECF 139) ¶ 133. But that did not form the basis for the “two-pronged in-kind payment,” In re Revlimid, 2024 WL 2861865, at *54, that led the Court to grant the motion to dismiss the Insurer Plaintiffs’ complaint.)

51.   The Defendants have also filed three motions to dismiss on other grounds. One raises arguments particular to AbbVie. See Memorandum of Law in Support of AbbVie Inc.’s Motion to Dismiss or Strike Plaintiffs’ Claims with Prejudice (ECF 530–1) at 1. Another locks in on the preclusive effect of a prior settlement. See Memorandum of Law in Support of Defendants Celgene Corporation and Bristol-Myers Squibb Company's Motion to Dismiss (ECF 531-1) at 1-2. And a final motion addresses claims under Minnesota law. See Memorandum of Law in Support of Defendants Celgene Corporation and Bristol-Myers Squibb Company's Motion to Dismiss (ECF 532-1) at 1. These three motions will be addressed separately. Not here.

52.   Assuming there was a “contract, combination, or ․ conspiracy.” 15 U.S.C. § 1. No one contests that here.

53.   See also King Drug, 791 F.3d at 403 (noting the need to evaluate whether a reverse payment is “an unexplained large transfer of value from the patent holder to the alleged infringer”); In re Wellbutrin, 868 F.3d at 162; In re Lipitor, 868 F.3d at 251; AbbVie, 976 F.3d at 357; accord, e.g., Watson Lab'ys., 101 F.4th at 238 (a reverse payment must be “both (1) ‘large’ and (2) ‘unjustified’ or unexplainable”); Impax Lab'ys., Inc. v. Fed. Trade Comm'n, 994 F.3d 484, 494 (5th Cir. 2021); United Food, 74 F. Supp. 3d at 1071-72; In re Aggrenox Antitrust Litig., 94 F. Supp. 3d 224, 241-42 (D. Conn. 2015); In re Bystolic Antitrust Litig., 583 F. Supp. 3d 455, 480 (S.D.N.Y. 2022); In re Opana ER Antitrust Litig., 162 F. Supp. 3d 704, 718-19 (N.D. Ill. 2016); see generally Michael A. Carrier & Edward Bank, The Missing Caselaw of Reverse-Payment Settlements, 98 St. John's L. Rev. 959, 991-1003 (2025); Kristen O'Shaughnessy et al., A Decade of FTC v. Actavis: The Reverse Payment Framework is Older, But Are Courts Wiser in Applying It?, 86 Antitrust L.J. 473, 492 (2024).

54.   One possible reason to require a plaintiff to allege “anticompetitive effects”: because if the motion-to-dismiss hurdle is cleared, at the next stage in the litigation the plaintiff will then have to satisfy the rule of reason. Analysis under the rule of reason requires some back-and-forth burden-shifting --- which gets started with the plaintiff trying to make out a prima facie case. See Am. Express Co., 585 U.S. at 541, 138 S.Ct. 2274. And as part of putting forward a prima facie case, an antitrust plaintiff must generally show “anticompetitive effects.” See id. at 542, 138 S.Ct. 2274; see also Impax Lab'ys, 994 F.3d at 492. So if the plaintiff will later --- for its prima facie case --- need to make an anticompetitive effects showing, it may make sense to require from it an earlier anticompetitive effects allegation, in its complaint. After all, it is often the case in civil litigation that what a plaintiff must later prove it must earlier allege --- from the start, out of the gate, in its pleading. But if this is part of the underlying logic here, note that in Title VII, another area where burden-shifting is required, the key aspects of a prima facie case do not need to be pled in the complaint. See, e.g., Connelly, 809 F.3d at 788; see also Swierkiewicz v. Sorema N.A., 534 U.S. 506, 510, 122 S.Ct. 992, 152 L.Ed.2d 1 (2002). Moreover, requiring that something must be pled may make less sense when the underlying facts are unlikely to be available before discovery. Cf. United States v. N.Y., New Haven & Hartford R.R. Co., 355 U.S. 253, 256 n.5, 78 S.Ct. 212, 2 L.Ed.2d 247 (1957). And information as to anticompetitive effects may be tough for a plaintiff to put hands on without discovery.

55.   Delay, plus a payment that is large and unjustified.

56.   See King Drug, 791 F.3d at 410 (concluding, on the basis of allegations about the other elements, that “at the pleading stage [the] plaintiffs have sufficiently alleged that any procompetitive aspects ․ were outweighed by the anticompetitive harm from the ․ agreement”).

57.   Indeed, the Defendants took things in the opposite direction. They asserted that “under the Third Circuit opinions,” a reverse-payment plaintiff “do[es] not get to the secondary question of what competitive effects do or do not flow from that agreement until [it] ha[s] plausibly ple[d] a payment.” Id. at 10:17-20; see also id. at 38:2-6. On this view, a payment is an element --- a gate to be passed through by plausible pleading. And “anticompetitive effects” is on the other side. A “second question” to eventually “get to” --- presumably in the rule-of-reason analysis, as part of the Plaintiffs’ prima facie case.

58.   The Plaintiffs assume that the delay period starts with a hypothesized Natco trial-court win, at the trial that was scheduled to go in 2016. See Complaint ¶ 643. But that might not be the right way to think about things. Cf. C. Scott Hemphill & Mark A. Lemley, Earning Exclusivity: Generic Drug Incentives and the Hatch-Waxman Act, 77 Antitrust L.J. 947, 963-64 (2011) (explaining that a generic product launch typically requires winning in the trial court --- and also on appeal); 21 U.S.C. § 355(j)(5)(D)(i)(I)(bb)(AA). No matter for now, though. The Defendants have not made a delay argument. Another point. Under the settlement agreement, Natco's timeline for launching its generic Revlimid could be pulled forward if certain contingencies materialized. Should that impact how delay is calculated? This, too, can be ignored for now. Same reason why. The Defendants do not move to dismiss on too-little-delay grounds.

59.   A payment might be justified for reasons that are internal to the lawsuit. Maybe, for example, the payment “amount[s] to no more than a rough approximation of the litigation expenses saved through the settlement.” Actavis, 570 U.S. at 156, 133 S.Ct. 2223. Or a payment can be justified for reasons that have little to do with the lawsuit. See id.; AbbVie, 976 F.3d at 357 (concluding that the payment in question “was ․ plausibly unjustified” in part because it could not “be explained as an independent business deal”) (cleaned up).

60.   The Defendants stated at oral argument that their motion to dismiss is not “focused on the large aspect.” Transcript at 11:10-11; see also Defendants’ Reply at 5. But although their “focus[ ]” is “on whether there was a payment at all,” Transcript at 11:11-12, the Defendants did not concede that if there was a payment here it was sufficiently “large.” See id. at 11:13-14 (noting that, “if there was a payment,” the Court would need to address “size”).

61.   On this approach, the “size” and the “justification” elements may blend into each other. A settlement is not large if it is roughly the same as litigation costs saved. And a settlement is justified if it is roughly the same as litigation costs saved. Cf. King Drug, 791 F.3d at 411 (seeming to treat litigation costs as relevant to the justification analysis); In re Lipitor, 868 F.3d at 256 (similar).

62.   Elsewhere in the decision, the court of appeals noted that the branded version of the drug in question “generated billions of dollars in sales,” AbbVie, 976 F.3d at 338; see also id. at 341, and that “billions of dollars” in those sales would be “protected by deferring competition.” Id. at 345.

63.   In the form of providing Natco an ability to charge high per-unit prices for generic Revlimid.

64.   To calculate this, the Plaintiffs compared what Natco “likely would have received under competitive conditions,” Complaint ¶ 325, with what it anticipated it would earn under the settlement after coming into the market in 2022. See id. ¶ 329. Moreover, “while these estimates do not include an estimate of the growth of the Revlimid market after March 2022, incorporating such an estimate would only increase the size of the payment.” Id. ¶ 333. There is no special rigor or precision needed in this area. See In re Lipitor, 868 F.3d at 255; AbbVie, 976 F.3d at 354, 356-57; see also, e.g., In re Loestrin, 814 F.3d at 552; In re Opana, 162 F. Supp. 3d at 719; In re Aggrenox, 94 F. Supp. 3d at 244-45. And the Plaintiffs’ allegations here are more precise than the “many millions of dollars of additional revenue” that was “sufficient” in King Drug, 791 F.3d at 409-10 (cleaned up), or the “hundreds of millions of dollars” that cleared the bar in In re Lipitor. 868 F.3d at 253.

65.   And as to litigation costs, the Plaintiffs have included an allegation. See Complaint ¶ 335 (“the payment is large” in that it is “much greater than the avoided litigation costs”).

66.   The Plaintiffs say yes. See Complaint ¶¶ 4, 6, 291, 322-24; Plaintiffs’ Opposition at 23. The Defendants say no. See Defendants’ Reply at 2; Transcript at 15:23 (“[t]here's no payment”).

67.   That the market would not be flooded with meaningful volumes of generic Revlimid supplied by entities other than Natco in the wake of the settlement agreement. (As to “meaningful,” see footnote 101.)

68.   Celgene would have been on the Revlimid market for years before Natco could make its first 2022 sale. So Celgene's brand-name Revlimid would have certain first-mover-type advantages. It would already be familiar to some prescribing doctors, for example. So Natco would almost surely have to charge less than Celgene to win customers for its generic. In direct head-to-head competition, at exactly the same price, customers might well otherwise reach for the more familiar brand-name Revlimid.

69.   Antitrust analysis is premised on the idea that firms act rationally. See Matsushita Elec. Indus. Co. v. Zenith Radio Corp., 475 U.S. 574, 596-97, 106 S.Ct. 1348, 89 L.Ed.2d 538 (1986) (“if petitioners had no rational economic motive to conspire, and if their conduct is consistent with other, equally plausible explanations, the conduct does not give rise to an inference of [an antitrust] conspiracy”); see also Chase Mfg., Inc. v. Johns Manville Corp., 84 F.4th 1157, 1168 (10th Cir. 2023) (“In antitrust cases, we ․ presume that businesses act rationally in conducting their affairs.”); Alan's of Atlanta, Inc. v. Minolta Corp., 903 F.2d 1414, 1430 n.22 (11th Cir. 1990) (“In Matsushita the Supreme Court warned courts ․ to be skeptical when presented antitrust allegations that clash with established economic theory. Such skepticism is well warranted because established economic theory indicates how rational entities generally act, and most antitrust cases involve the acts of rational entities.”); Adaptive Power Sols., LLC v. Hughes Missile Sys., Co., 141 F.3d 947, 952 (9th Cir. 1998) (“Antitrust claims must make economic sense.”).

70.   A counterargument might look like this: in the face of volume caps, Natco might potentially opt not to charge a very high price --- but rather a very low price. Charging $9.75 per unit instead of $19.75, for example. If that draws in new customers --- 1,000 people looking for Revlimid, instead of 100 --- watering down prices to expand the size of the overall market could work out for Natco. After all, 7% of 100 customers paying $19.75 per unit would bring in around $138 in revenue. But 7% of 1,000 customers paying $9.75 per unit spins off about $682 in revenue. There are any number of things that might be said about a counterargument along these sorts of lines. (For example, as to the elasticity of demand for Revlimid --- there are limited numbers of people, whatever the price might be, that need to seek out a cancer drug like Revlimid.) But the Defendants say nothing about the points referenced in this footnote, so the Court leaves it here.

71.   Celgene would itself be plausibly unlikely to reduce its prices. Why lower monopoly-type prices to keep sales away from a competitor (Natco) that can only aspire to service a small fraction (7%, 12%, etc.) of the market?

72.   See Iqbal, 556 U.S. at 678, 129 S.Ct. 1937; Hikma Pharma. USA Inc. v. Amarin Pharma, Inc., 608 U.S. 580, 589-90, 146 S.Ct. 1391, 225 L.Ed.2d 159 (2026).

73.   Competition, to be sure, has other important virtues. Not just “undercutting” on price, but also, for example, “improving” products. F.A. Hayek, Individualism and Economic Order 96 (1948). But that is not relevant here. The brand-name Revlimid sold by Celgene and the generic Revlimid sold by Natco were chemically equivalent, see footnote 6, and were intended to remain that way. So competition mainly matters here for pricing purposes, not for others.

74.   The “reverse payment [here] primarily takes the form of Celgene splitting its monopoly profits with [Natco.]” Complaint ¶ 324; see also Plaintiffs’ Opposition at 5 (“[Natco] could charge markedly higher prices for the generic than [it] could with competition.”); id. at 9 (arguing that, per the complaint, Natco “created a regime under which [it] ․ would be able to charge exclusivity prices throughout the almost four years (1,430 days) of volume-capped periods”).

75.   Applied with an eye to the favorable inferences the Plaintiffs are entitled to for now. See footnote 72.

76.   “While ․ antitrust law is strongly premised on economic theory, common sense still has a role.” Champagne Metals v. Ken-Mac Metals, Inc., 458 F.3d 1073, 1087 n.14 (10th Cir. 2006); see generally Diamond Alt. Energy, LLC v. Ent't Prot. Agency, 606 U.S. 100, 116, 145 S.Ct. 2121, 222 L.Ed.2d 370 (2025) (drawing upon “commonsense economic realities”).

77.   The Plaintiffs invoke “applied microeconomics.” But they leave it at that. Nothing more specific.

78.   This work needs doing because the Defendants contend that there is no role for economics here. See Defendants’ Brief at 14-15 & 14 n.11.

79.   On the assumption referenced in footnote 67.

80.   See also, e.g., Cont'l T.V., Inc. v. GTE Sylvania Inc., 433 U.S. 36, 59, 97 S.Ct. 2549, 53 L.Ed.2d 568 (1977) (suggesting that antitrust law should be guided by “demonstrable economic effect”); Eastman Kodak Co. v. Image Tech. Servs., Inc., 504 U.S. 451, 466–67, 112 S.Ct. 2072, 119 L.Ed.2d 265 (1992); Copperweld Corp. v. Indep. Tube Corp., 467 U.S. 752, 774, 104 S.Ct. 2731, 81 L.Ed.2d 628 (1984); United States v. Concentrated Phosphate Exp. Ass'n, 393 U.S. 199, 208, 89 S.Ct. 361, 21 L.Ed.2d 344 (1968).

81.   See, e.g., United States v. Dentsply Int'l, Inc., 399 F.3d 181, 189 (3d Cir. 2005); Avaya, Inc., RP v. Telecom Labs, Inc., 838 F.3d 354, 404 (3d Cir. 2016); Deutscher Tennis Bund v. ATP Tour, Inc., 610 F.3d 820, 835 (3d Cir. 2010).

82.   The Supreme Court leaned on academic analysis in formulating the pay-for-delay doctrine. See Actavis, 570 U.S. at 142, 144, 151, 154, 133 S.Ct. 2223; see also id. at 169-70, 133 S.Ct. 2223 (Roberts, C.J., dissenting). And courts routinely look to the economics literature in working through antitrust issues. See, e.g., Leegin Creative Leather Prods., Inc. v. PSKS, Inc., 551 U.S. 877, 889-90, 127 S.Ct. 2705, 168 L.Ed.2d 623 (2007); Cal. Dental Ass'n v. Fed. Trade Comm'n, 526 U.S. 756, 771-72, 119 S.Ct. 1604, 143 L.Ed.2d 935 (1999); Cornish-Adebiyi v. Caesars Ent., Inc., 2026 WL 2182291, at *1 n.5 (3d Cir. July 29, 2026); It's My Party, Inc. v. Live Nation, Inc., 811 F.3d 676, 689 (4th Cir. 2016); United States v. Microsoft Corp., 253 F.3d 34, 49 (D.C. Cir. 2001).

83.   Economics can be “a powerful, neutral tool.” But courts are on their steadiest ground in using economics that does not turn on “difficult mathematics or highly technical analysis but ․ just informed common sense about how markets work[.]” Herbert Hovenkamp, The Looming Crisis in Antitrust Economics, 101 B.U. L. Rev. 489, 544 (2021).

84.   For some of the key general discussions of the model, see, for example, William G. Shepherd, The Economics of Industrial Organization: Analysis, Markets, Policies 206-08 (Leah Jewell, ed., 4th ed. 1997); Alice P. White, The Dominant Firm Structure: Theoretical Myth or Empirical Reality?, 48 S. Econ. J. 427, 427 (1981); Oliver E. Williamson, Dominant Firms and the Monopoly Problem: Market Failure Considerations, 8 Harv. L. Rev. 1512, 1512-14 (1972).

85.   See, e.g., Henry G. Grabowski & John M. Vernon, Brand Loyalty, Entry, and Price Competition in Pharmaceuticals after the 1984 Drug Act, 35 J.L. & Econ. 331, 341 (1992); cf. Patrick J. Welch, Simple Price Restrictive Patent Licensing and Simple Dominant Firm Leadership: Do the Similarities Outweigh the Differences?, 60 J. Pat. Off. Soc'y 230, 236 (1978).

86.   For example, a merger may generate antitrust concerns if it has the potential to push a market into a dominant-firm structure and in doing so to weaken the price-lowering effect of an existing band of smaller competitors. See, e.g., Fed. Trade Comm'n v. Staples, Inc., 190 F. Supp. 3d 100, 128 (D.D.C. 2016); United States v. Oracle Corp., 331 F. Supp. 2d 1098, 1113 (N.D. Cal. 2004); see also Areeda & Hovenkamp, Antitrust Law ¶ 911.

87.   See, e.g., Williamson, Dominant Firms and the Monopoly Problem, 8 Harv. L. Rev. at 1512 n.1.

88.   See, e.g., John M. de Figueiredo & Brian S. Silverman, Churn, Baby, Churn: Strategic Dynamics Among Dominant and Fringe Firms in a Segmented Industry, 4 Mgmt. Sci. 632, 634 (2007) (“Empirical studies of dominant firms have identified dominance primarily by a firm's market share, with a share of 40% or 50% used as the typical threshold for dominance.”); cf. Hideki Yamawaki, Dominant Firm Pricing and Fringe Expansion: The Case of the U.S. Iron and Steel Industry, 1907-1930, 67 Rev. Econ. & Stat. 429, 429, 437 (1985) (finding that, between 1907 and 1930, “U.S. Steel behaved as a dominant firm” with 50.1 percent of “industry ingot capacity”).

89.   “Largely fits” because the dominant firm model was formulated to describe the knock-on effects when economic considerations split the market. Here, it is an agreement that divvies things up. This is a distinction, but not one that matters. There is no reason to think that the dominant firm model's basic insights as to pricing (of which more in a moment) no longer apply when it is a contract (and not market forces) that leaves one entity as dominant and the other as fringe. A firm can occupy the fringe because of contractually-imposed supply limits, as in the settlement agreement's volume caps on Natco. Or a firm can occupy the fringe because of supply limits that flow from other kinds of sources. Like a lack of capital that prevents building a new factory. Or an inability to put hands on more of a resource that is needed for the relevant production process. Whatever the cause of its fringe status, the fringe firm (Natco) will plausibly be in a price-taking mode (more on that soon) --- and that is the key point here.

90.   And on the assumption at footnote 67.

91.   Because of this dynamic, the dominant firm is often described as the “price leader.” See Raymond J. Deneckere & Dan Kovenock, Price Leadership, 59 Rev. Econ. Stud. 143, 143 (1992); see also Yoshiyasu Ono, Price Leadership: A Theoretical Analysis, 49 Economica 11, 11 (1982); Schenzler et al., The History of the Static Equilibrium, 18 E. Econ. J. at 171; Attila Tasnádi, On Forchheimer's Model of Dominant Firm Price Leadership, 84 Econ. Letters 275, 275 (2004); Leonardo Salazar & Jorge Dresdner, Market Integration and Price Leadership: The U.S. Atlantic Salmon Market, 25 Aquaculture Econ. & Mgmt. 245, 250 (2020); cf. In re Baby Food Antitrust Litig., 166 F.3d 112, 132 (3d Cir. 1999) (noting allegations “that Gerber, the dominant company in the industry [was a] price leader”).

92.   See also, e.g., Qwest Corp. v. Fed. Commc'ns Comm'n, 689 F.3d 1214, 1232 n.9 (10th Cir. 2012); Roscoe B. Starek III & Stephen Stockum, What Makes Mergers Anticompetitive? “Unilateral Effects” Analysis Under the 1992 Merger Guidelines, 63 Antitrust L.J. 801, 808-09 (1995); Roger D. Blair & William H. Page, “Speculative” Antitrust Damages, 70 Wash. L. Rev. 423, 442 (1995) (“Economic theory shows that a rational fringe competitor will match the price increase of a dominant firm ․ in its market; if it did not do so, it would fail to maximize its profits.”); William H. Page, The Scope of Liability for Antitrust Violations, 37 Stan. L. Rev. 1445, 1474-75 (1985).

93.   White, Monopoly and Dominant Firms at 323; see also Stigler, Notes on the Theory of Duopoly, 48 J. Pol. Econ. at 522 (describing the dominant firm model “as partial monopoly”); Nils-Henrik M. von der Fehr, Leader, or Just Dominant? The Dominant-Firm Model Revisited 2 (Dep't of Econ., Univ. of Oslo, Working Paper, Memorandum No. 15/2010) (noting that the dominant firm model is “also called partial monopoly and monopoly with fringe”).

94.   Two especially critical points here. First, as to the part of the market where it is the only source of supply, because the fringe firm is not producing enough to service that part --- the dominant firm can essentially act as a true monopolist. See, e.g., Rassenti & Wilson, How Applicable is the Dominant Firm Model of Price Leadership?, 7 Experimental Econ. at 272 (discussing the dominant firm's ability to act as a monopolist over the residual demand curve --- the total market demand minus what the fringe supplies); see also White, Monopoly and Dominant Firms at 320-23; Areeda & Hovenkamp, Antitrust Law ¶ 521b; Lawrence J. White, Market Power and Market Definition in Monopolization Cases: A Paradigm is Missing, U.S. Dep't of Just., https://www.justice.gov/archives/atr/market-power-and-market-definition-monopolization-cases-paradigm-missing-0#img17 (last accessed Sept. 23, 2026). Second, even where it faces a residual demand curve, if the dominant firm sets its prices too high it can ultimately undermine itself. For example, fringe firms can see the especially high prices the dominant firm sets and be induced to invest in greater production capacity --- to pave the way to their eventually grabbing more (high-margin) business. See Stephen Martin, Industrial Economics: Economic Analysis and Public Policy 76-78 (1994). This sort of possibility can give the dominant firm reason to set prices lower, to dissuade others from one day jumping in. See id. at 68-94; cf. Part IX.A.2. But it is plausible that this sort of price-disciplining effect would have no meaningful role here. Even if (i) Natco, the fringe firm, were to see an opportunity because Celgene sets brand-name Revlimid prices very high on the 93% or 88% of the market that it alone controls, and (ii) Natco might therefore otherwise be incentivized to invest in greater capacity --- the settlement agreement would nonetheless prevent Natco from bringing more supply to the market. See, e.g., Martin, Industrial Economics at 75, 82 (discussing “blockaded entry”).

95.   See King Drug, 791 F.3d at 400-01 (“a valid patent gives its holder the right to ․ charge prices of its choosing, including supracompetitive prices”); Mayor of Balt., 42 F.4th at 712 (“valid patents authorize their owners to exclude competition and charge monopoly prices”); United States v. Studiengesellschaft Kohle, m.b.H., 670 F.2d 1122, 1128 (D.C. Cir. 1981) (“charging of supracompetitive prices [is] at the core of the patentee's rights”).

96.   Described at the beginning of Part VIII.

97.   On the assumption alluded to in footnote 67.

98.   Courts have sometimes expressed skepticism in this area. See, e.g., United States v. Int'l Harvester Co., 274 U.S. 693, 708-09, 47 S.Ct. 748, 71 L.Ed. 1302 (1927) (“[T]he fact that competitors may see proper, in the exercise of their own judgment, to follow the prices of another manufacturer, does not establish any suppression of competition or show any sinister domination.”); United States v. U.S. Steel Corp., 251 U.S. 417, 449, 40 S.Ct. 293, 64 L.Ed. 343 (1920) (“The government does not hesitate to present contradictions ․ In one, competitors ․ are represented as oppressed by the superior power of the corporation; in the other, they are represented as ascending to opulence by imitating that power's prices, ․ and yet confederated action is not asserted.”). But this skepticism is not grounded in doubt about the correctness of the idea that the dominant firm generally sets the “ruling price.” Stigler, Notes on the Theory of Duopoly, 48 J. Pol. Econ. at 523. If anything, the opposite. The above-quoted judicial opinions are premised on confidence that the dominant-firm model can accurately reflect how prices generally get set. The dominant-firm model predicts that when it comes to prices, B (on the fringe) largely follows A (the dominant firm) --- naturally, as a matter of economic self-interest. And precisely because that is taken as true-to-life, the fact that B follows A on prices does not, without more, call for an explanation --- let alone suggest that B and A have engaged in consciously coordinated action. It just proves that the dominant firm model is in play --- and that it works. Not that there has been a hidden B/A agreement to work together on “confederated action.” U.S. Steel Corp., 251 U.S. at 449, 40 S.Ct. 293. Here, the dominant firm model is not invoked to suggest that Celgene and Natco had an agreement. They plainly did. Their settlement agreement. Rather, the dominant firm model has been put on the table here for a different purpose --- to assess the common-sense idea that price-taking by Natco from Celgene is a plausible outcome of the arrangements put in place by their settlement agreement.

99.   This, per the Plaintiffs’ allegations, was worth billions of dollars. See Complaint ¶ 332.

100.   See also, e.g., Brooke Grp. Ltd. v. Brown & Williamson Tobacco Corp., 509 U.S. 209, 221, 113 S.Ct. 2578, 125 L.Ed.2d 168 (1993) (noting “the antitrust laws’ traditional concern for consumer welfare and price competition”); Associated Gen. Contractors of Cal., Inc. v. Cal. State Council of Carpenters, 459 U.S. 519, 538, 103 S.Ct. 897, 74 L.Ed.2d 723 (1983) (“the Sherman Act was enacted to assure customers the benefits of price competition”).

101.   The basic dominant/fringe structure could plausibly be maintained even if non-Natco generic drugmakers were able to service modest slivers of the market. See generally footnotes 87-88. 93/7 is the overall 2022 market division that would likely come about if Celgene and Natco were the only generic Revlimid providers on the market. But imagine a 2022 overall market division of, say, 85/15 --- with the added 8% of generic Revlimid supplied by a non-Natco generic drugmaker. It is plausible that this 85/15 set-up would yield a dominant/fringe pricing dynamic (price-leader and price-taker) closely analogous to the pricing dynamic that a 93/7 overall market division would produce. It is because of this similarity in terms of result that this Opinion and Order is focused throughout on (i) the possibility of the market being flooded with meaningful volumes of generic Revlimid --- but not on (ii) the possibility of a modest bit of extra generic Revlimid coming to market here and there. The former, (i), could readily undo a dominant/fringe structure. The latter, (ii), plausibly would not. And it is the existence of a dominant/fringe structure, and the price-leader/price-taker dynamic that goes with it, that is critical here --- regardless of whether that pricing dynamic flows from a 93/7 (dominant/fringe) market division or from an 85/15 (also dominant/fringe) market division. The key thing is the bottom-line existence of the price-leader/price-taker dynamic --- because that is what allows Natco to plausibly be able to charge high per-unit prices for generic Revlimid.

102.   Footnote 94 is critical for this point.

103.   See Part VIII.B.

104.   Plausibly giving it, that is.

105.   See, e.g., In re K-Dur Antitrust Litig., 686 F.3d 197, 218 (3d Cir. 2012) (“[a] payment ․ to the challenging generic firm may suggest strongly the anticompetitive intent of the parties entering the agreement”) (cleaned up), vacated, 570 U.S. 913, 133 S.Ct. 2849, 186 L.Ed.2d 904 (2013); In re Cardizem CD Antitrust Litig., 332 F.3d 896, 908 (6th Cir. 2003) (“it is one thing to take advantage of a monopoly that naturally arises from a patent, but another thing altogether to bolster the patent's effectiveness in inhibiting competitors by paying the only potential competitor ․ to stay out of the market”).

106.   See also Areeda & Hovenkamp, Antitrust Law ¶ 2046d2 (“An unreasonably large settlement payment indicates that the parties predicted a high chance of patent failure and were using the settlement plus the exclusionary protections of Hatch-Waxman to make the patent invulnerable to attack.”)

107.   Cf. King Drug, 791 F.3d at 405 (noting that, in a reverse payment settlement, “ ‘[t]he source of the benefit to the claimed infringer is something costly to the patentee’ ”) (quoting Aaron Edlin et al., Activating Actavis, 28 Antitrust 16, 22 n.22 (2013)).

108.   Celgene could have opted to try to service those 7% or 12% slices of the market. But it would have been doing so in direct competition with a generic. So to compete for that business, Celgene would almost surely have needed to pull down its prices, and a good deal.

109.   See, e.g., Jesus Seade, On the Effects of Entry, 48 Econometrica 479, 479 (1980) (describing the “conventional wisdom”).

110.   See, e.g., Keith M. Drake et al., Trends in Authorized Generic Drug Launches And Their Effects on Competition In Oral-Solid Drug Markets In The US, 2016-23, 44 Health Affs. 745, 752 (“authorized generic launches were associated with lower initial generic prices”); Ryan Conrad & Randall Lutter, Generic Competition and Drug Prices: New Evidence Linking Greater Generic Competition and Lower Generic Drug Prices, Food & Drug Admin. 2 (2019) (“generic drug prices after initial generic entry decline with additional competition”); see also Steven Tenn & Brett W. Wendling, Entry Threats & Pricing in the Generic Drug Industry, 96 Rev. Econ. & Stat. 214, 221 (2014) (noting that “additional competitors lead to lower prices”); Atanu Sasha et al., Generic Competition in the US Pharmaceutical Industry, 13 Int'l J. of the Econ. of Bus. 15, 29 (2006) (“each additional generic manufacturer induces nearly a 2.3 [percent] monthly decrease in the value of the generic-to-brand price ratio”); David Reiffen & Michael R. Ward, Generic Drug Industry Dynamics, 87 Rev. Econ. & Stat. 37, 37 (2005) (“we find that generic drug prices fall with an increase in the number of competitors”).

111.   This point is developed in Part IX.B.

112.   More on this in Part IX.A.

113.   As noted above, generic prices will generally be higher without an authorized generic than with one. The 2011 Federal Trade Commission study, for example, establishes that.

114.   This is the assumption made above, see footnote 67, and analyzed below. See Part IX.

115.   This second effect is why some commentators think that a no-authorized generic settlement provision “is more harmful to competition than a cash settlement of the same magnitude.” Aaron Edlin et al., The Actavis Inference: Theory and Practice, 67 Rutgers U. L. Rev. 585, 597 (2015). Such a provision “keep[s] up prices at consumers’ expense.” Id. at 598; accord, e.g., Michael Carrier, Eight Reasons Why “No-Authorized Generic” Promises Constitute Payment, 67 Rutgers U. L. Rev. 697, 716-20 (2015); Areeda & Hovenkamp, Antitrust Law ¶ 2046f2.

116.   Per King Drug, “reverse payments are problematic because of their potential to negatively impact consumer welfare by preventing the risk of competition.” 791 F.3d at 403-04; see also id. at 405 (reasoning that the settlement under consideration had the effect of “transfer[ing] the profits the patentee would have made ․ to the settling generic”); id. at 409-10 (concluding that the agreement was plausibly “an inducement ---- valuable [to both firms] ․ --- to ensure a longer period of supracompetitive monopoly profits based on a patent at risk of being found invalid or not infringed,” and that on that basis, the plaintiffs’ allegations were “sufficient to state a rule-of-reason claim”); see also, e.g., In re Lipitor, 868 F.3d at 260 (holding that the plaintiffs “plausibly allege[d] a large reverse payment, with [the brand firm's] ․ agreement allow[ing] [the generic firm] to maintain a supra-competitive generic price as the only generic manufacturer on the market, and to earn substantially higher profits than it otherwise would have earned”) (cleaned up).

117.   From generic entry forward.

118.   Again, from generic entry forward.

119.   See footnote 101.

120.   See Part VII.

121.   More supply almost always brings down prices. And consistent with this, the more drugmakers that bring their generic drugs to market, the more prices tend to fall. See, e.g., Conrad & Lutter, Generic Competition and Drug Prices, at 2; Tenn & Wendling, Entry Threats & Pricing in the Generic Drug Industry, 96 Rev. Econ. & Stat. at 222 tbl. 3; Reiffen & Ward, Generic Drug Industry Dynamics, 87 J. Econ. & Stat. at 33 fig. 1. The Defendants emphasize roughly this point in their legal papers. See Defendants’ Brief at 9-13, 16-23; Defendants’ Reply at 2-6, 9-14. (Indeed, if it appeared that generic Revlimid was poised to soon begin swamping the market, Celgene might opt to use its price-leader position to bring down prices --- in part to try to deter that generic Revlimid from ever coming in. See Martin, Industrial Economics at 76-85; see also Part IX.A.2.)

122.   Where, for example, a 7% volume cap spins off a 93/7 overall market split, with perhaps some modest amount of extra volume for non-Natco producers --- modest enough that together, with Natco, the various generic-drug producers still amount to a fringe, in the sense that they continue to mainly take their price from the dominant firm. Cf. footnote 101.

123.   See generally Plaintiffs’ Opposition at 3 (“the new allegations [in the complaint],” beyond those in the prior motion to dismiss, “detail multiple provisions to protect the likelihood of high generic prices”). The Defendants say these dams would not work well enough. The Plaintiffs argue they would. Compare Defendants’ Brief at 16-23, 29-30, 32-36, Defendants’ Reply at 6; Transcript at 23:14-18 with Plaintiffs’ Opposition at 2-4, 6-7, 25-32, 41-42.

124.   Or licensing a third party to go that way.

125.   Or that their drug would not infringe on Celgene's Revlimid patents. See 21 U.S.C. § 355(j)(2)(A)(vii)(IV).

126.   As Natco had.

127.   See Complaint ¶¶ 123, 125, 319.

128.   See Complaint ¶¶ 297, 317.

129.   See Plaintiffs’ Opposition at 25-27.

130.   See Defendants’ Brief at 9-11; Defendants’ Reply at 6; Transcript at 23:14-18.

131.   And it can also always license another company to make a generic version of the drug on its behalf. See, e.g., Sanofi-Aventis, 659 F.3d at 1174.

132.   As equivalents, the drugs are mainly substitute goods. So buyers will often go for the less pricey one. See Phillip E. Graves & Robert L. Sexton, Cross Price Elasticity and Income Elasticity of Demand: Are Your Students Confused?, 54 Am. Economist 107, 107 (2009); Henry W. Kinnucan & Yuliang Miao, Distributional Impacts of Generic Advertising on Related Commodity Markets, 82 Am. J. Agric. Econ. 672, 672 (2000).

133.   An authorized generic generally results in “lower prices.” Sanofi-Aventis, 659 F.3d at 1175; see also King Drug, 791 F.3d at 404 n.21; Mylan Pharms., Inc. v. U.S. Food & Drug Admin., 454 F.3d 270, 273 (4th Cir. 2006); Drake et al., Trends in Authorized Generic Drug Launches, 44 Health Affs. at 751; Fed. Trade Comm'n, Authorized Generic Drugs: Short-Term Effects and Long-Term Impact, at 40.

134.   See, e.g., Annabelle C. Fowler et al., Authorized Generics in the US: Prevalence, Characteristics, And Timing, 2010-19, 42 Health Affs. 1037, 1074 (2023) (describing “an authorized generic incidence of around 6 percent [for] 2010-19”).

135.   See, e.g., In re Niaspan Antitrust Litig., 67 F.4th 118, 124 (3d Cir. 2023) (via an authorized generic, the brand-name drugmaker can “recover some of the sales and profits it would otherwise lose” to a generic); Mylan Pharms., 454 F.3d at 273 (“[b]y selling an authorized generic,” a brand-name firm “prevents [its generic competitor] from winning all of the customers who want to switch from the branded drug to a cheaper generic form”); Fed. Trade Comm'n, Authorized Generic Drugs: Short-Term Effects and Long-Term Impact, at 12 (explaining that authorized generics can serve “as a way for brand-name companies to stem their losses to generic competition”); see also id. at 26; Fowler et al., Authorized Generics in the US, 42 Health Affs. at 1076.

136.   Celgene could potentially want to come out with an authorized generic to cut losses vis-a-vis another generic drugmaker, one other than Natco. But as will be seen below, see Part IX.B, it is plausible that generic drugmakers that might enter the Revlimid market would themselves only get narrow segments of market share --- in which case Celgene could plausibly keep selling its brand-name Revlimid at monopoly-type prices to the very large bulk of the market even after those other generic drugmakers entered the market. That would seem to be a good arrangement for Celgene. Why would Celgene undo it by putting out its own generic?

137.   See Douglas H. Ginsburg & Joshua D. Wright, Dynamic Analysis and the Limits of Antitrust Institutions, 78 Antitrust L.J. 1, 3 n.8 (2012) (collecting sources); see generally B. Douglas Bernheim, Strategic Deterrence of Sequential Entry into an Industry, 15 RAND J. Econ. 1, 3-4 (1984) (“[t]o deter entry, incumbents” sometimes pursue “deterrence strategies which lower ․ their own profitability”).

138.   Lu Yao & Mengde Liu, Strategic Behavior and Entry Deterrence by Branded Drug Firms: The Case of Authorized Generic Drugs, 26 Eur. J. Health Econ. 561, 573 (2024); see also Rubaiyat Alam & Rena M. Conti, Entry Delays and Fighting Brands: Evidence from Generics and Authorized Generics 1 (June 18, 2025) (https://rubaiyat-alam.github.io/website/pharma_ag_rubaiyat_conti.pdf); Bryan A. Liang, The Anticompetitive Nature of Brand-Name Firm Introduction of Generics Before Patent Expiration, 41 Antitrust Bull. 599, 599-602 (1996).

139.   Fed. Trade Comm'n, Authorized Generic Drugs: Short-Term Effects and Long-Term Impact at 27; see also Drake et al., Trends in Authorized Generic Drug Launches, 44 Health Affs. at 752.

140.   The preemptive-strike rationale might apply with some force --- because Celgene might potentially release an authorized generic to deter a drugmaker other than Natco. But this not a strong consideration, for reasons alluded to in footnote 136.

141.   The Court previously noted that Natco's “volume limited license ․ may have disincentivized Celgene from launching an authorized generic.” In re Revlimid, 2024 WL 2861865, at *64. In this section, the Court has developed and firmed up its “disincentivized” conclusion. And in the next section, the Court adds consideration of another disincentive. Not one that flows from market logic. But a contractual disincentive, built into the settlement agreement --- and one that was not pressed by the Insurer Plaintiffs when the prior motion to dismiss was taken up. See Plaintiffs’ Opposition at 1-4.

142.   Acceleration clauses along roughly these lines are a relatively common feature of settlement agreements in this area. See Carrier & Bank, The Missing Caselaw of Reverse-Payment Settlements, 98 St. John's L. Rev. at 1003; O'Shaughnessy et al., A Decade of FTC v. Actavis, 86 Antitrust L.J. at 485.

143.   Sometimes a patentholder does not put out an authorized generic itself, but gives another company a license to produce one. See Edlin et al., The Actavis Inference, 67 Rutgers U. L. Rev. at 595; Oliver Gassman et al., Out-Licensing in Pharmaceutical Research and Development, in Value Creation in the Pharmaceutical Industry: The Critical Path to Innovation 363, 367-71 (Alexander Schuhmacher et al., eds., 2016); Keith M. Drake & Thomas McGuire, Using Stock Price Movements to Estimate the Harm from Collusive Drug Patent Litigation Settlements 16 (Nat'l Bureau of Econ. Rsch., Working Paper No. 33,196, 2025). The Celgene/Natco settlement agreement also included various provisions that affirmatively disincentivized Celgene from allowing a licensee to put out an authorized generic on its (Celgene's) behalf. See License Agreement §§ 1.11(a), 5.1(a), 5.1(d).

144.   “Relatively” because the agreement would surely be less good for Celgene than serving all of the Revlimid market, as its patents allowed. In striking the deal it did, Celgene gave up something valuable --- a piece of the market. But nonetheless, the arrangement Celgene got for itself through the settlement agreement plainly had many large upsides for it, as set out in the text.

145.   Or more precisely, the full Revlimid market until 2022, followed by most of the Revlimid market until 2026 --- 93%, 88%, etc.

146.   See footnote 94.

147.   When an authorized generic is on the market up against a first-filer, prices generally fall. This is common sense. See Kevin A. Hassett & Robert J. Shapiro, The Impact of Authorized Generic Pharmaceuticals on the Introduction of Other Generic Pharmaceuticals, 3 (2007), https://www.sonecon.com/wp-content/uploads/2022/01/050207_authorizedgenerics.pdf. And it is backed up by experience. See, e.g., Drake et al., Trends in Authorized Generic Drug Launches, 44 Health Affs. at 751; Fed. Trade Comm'n, Authorized Generic Drugs: An Interim Report, at 2; see also Fed. Trade Comm'n, Authorized Generic Drugs: Short-Term Effects and Long-Term Impact, at ii.

148.   When an authorized generic and a first-filing generic (like Natco) are together on the market, the first-filer generally wins a meaningful amount of business. See Fed. Trade Comm'n, Authorized Generic Drugs: Short-Term Effects and Long-Term Impact, at 106; Drake et al., Trends in Authorized Generic Drug Launches, 44 Health Affs. at 751.

149.   See Part IX.A.2.

150.   See Part IX.A.3.

151.   The acceleration provision was, at best, a “de facto no authorized generic” provision. It did not forbid Celgene from putting out an authorized generic. But it angled things away from that possibility. Whether a “de facto no authorized generic” provision can itself be chalked up as a pay-for-delay “pay” --- that question has divided the courts. Compare In re Xyrem (Sodium Oxybate) Antitrust Litigation, 555 F. Supp. 3d 829, 856-59 (N.D. Cal. 2021), and FWK Holdings LLC v. Shire PLC, 2017 WL 11449668, at *7-9 (D. Mass. Oct. 10, 2017), with In re Actos End Payor Antitrust Lit., 2015 WL 5610752, at *18 (S.D.N.Y. Sept. 22, 2015), aff'd in part, vacated in part, 848 F.3d 89 (2d Cir. 2017). And this Court has expressed certain views, too. See In re Revlimid, 2024 WL 2861865, at *62-64. But none of this needs to be waded into here. The Plaintiffs’ argument is not that a de facto no authorized generic provision itself counts as a payment to Natco. It is, rather, that the referenced provision --- combined with what is discussed below in Part IX.B --- would plausibly prevent non-Natco Revlimid from swamping the market. And in that circumstance, the settlement agreement's Natco volume caps would work an overall dominant/fringe market division --- and allow Natco (from the fringe) to charge high per-unit prices on generic Revlimid. On this argument, a de facto authorized generic is not itself a payment. Rather, it is one of the background conditions that allows something else (the volume caps) to generate a payment (the ability to charge especially high generic prices).

152.   Celgene was the Revlimid patentholder, so generic drugmakers could come into the market only (i) with Celgene's permission or (ii) by convincing a court to let it in. The Defendants emphasize the first possibility. And with good reason. During a relevant time period, around 98% of patent litigations ended in settlement. See footnote 17. (And 98% probably understates how often generic drugmakers come into the market based on a deal with the patentholder. Because the 98% figure does not tally up the times drugmakers strike an agreement for the generic to get on the market without a lawsuit first getting underway.)

153.   See footnote 144.

154.   See footnote 94.

155.   For example, if the relevant Revlimid patents were extraordinarily weak, maybe Celgene would have had to say yes to virtually any offer from a generic drugmaker.

156.   In opportunity cost.

157.   Stipulate that Celgene's Revlimid and ABCD's Revlimid would each go for $10 on the market.

158.   See Complaint ¶ 317 (“the agreement assured that the potential impact of later generic activities ․ would not hinder the monetary value of the profit-split”); id. (“The [1.5(x)] ․ provision[ ] assured [Natco] ․ the best position, limiting both what Celgene could in fact allocate to later generics and, as a practical matter, limiting the aggregate of all such later market-sharing arrangements”); id. ¶ 7 (“[T]he parties needed to plan how to address potential later would-be generic entrants. The final agreement reflected a plan to wrap later generics into the overall market-allocation plan ․ by sharing with the later generics smaller shares of Celgene's monopoly profits[.]”); id. ¶ 283 (alleging that Natco “knew that the structure of the agreement would likely result in later-filing generics receiving volume-capped allocations”); see also id. ¶ 297; Plaintiffs’ Opposition at 29-30.

159.   And to convey any shortfall value needed for a settlement in other ways. Cash. Other business opportunities controlled by Celgene. Etc.

160.   Recall, this provision was not previously argued to the Court by the Insurer Plaintiffs. See footnote 50.

161.   The point of the Defendants’ chart is to show that the Plaintiffs would have no plausible pay-for-delay claim. But the chart suggests that the settlement agreement would have generated a 6-month exclusivity period for Natco --- in which it alone was on the market, without competition from, among others, an authorized generic. And note the Third Circuit has held that a settlement agreement that blocks an authorized generic for six months is enough to state a pay-for-delay claim. See King Drug, 791 F.3d at 397, 403-04; In re Wellbutrin, 868 F.3d at 159-63, 159 n.40.

162.   See generally, e.g., Conrad & Lutter, Generic Competition and Drug Prices, U.S. Food & Drug Admin. at 3 (“With six or more competitors, generic prices ․ show price reductions of more than 95 [percent] compared to brand prices.”). As the Plaintiffs put it, there is little reason to think that Celgene would “give away as much of its monopoly as possible.” Plaintiffs’ Opposition at 31 (emphasis omitted).

163.   More on this in Part IX.B.4, just below.

164.   See also, e.g., In re Lipitor, 868 F.3d at 254 (rejecting application of “a heightened pleading standard contrary to Bell Atlantic Corp. v. Twombly, 550 U.S. 544, 127 S.Ct. 1955, 167 L.Ed.2d 929 (2007) and Ashcroft v. Iqbal, 556 U.S. 662, 129 S.Ct. 1937, 173 L.Ed.2d 868 (2009)” in reverse payment cases); Schuchardt v. President, 839 F.3d 336, 347 (3d Cir. 2016).

165.   For background on paragraph IV certifications, see Part II.A and Part II.B.

166.   In this sense, the “exclusivity” period is not truly exclusive. See C. Scott Hemphill, An Aggregate Approach to Antitrust: Using New Data and Rulemaking to Preserve Drug Competition, 109 Colum. L. Rev. 629, 634 & n.14 (2009).

167.   Except in circumstances that are not relevant here. If, for example, another generic drugmaker goes forward “at risk.” See In re Wellbutrin, 868 F.3d at 168 n.59; see also Cipla Ltd. v. Amgen Inc., 778 F. App'x 135, 137 n.2 (3d Cir. 2019); Drake et al., No Free Launch: At-Risk Entry by Generic Drug Firms 6 (Nat'l Bureau of Econ. Rsch., Working Paper No. 29,131, 2021). Or if multiple generic firms filed on the same (first) day. See King Drug, 791 F.3d at 396 n.10.

168.   The first-filing generic drugmaker gets its “temporary monopoly” even if it gets into the market by settling with the brand-name firm after challenging it in court. See In re Wellbutrin, 868 F.3d at 144 n.7. Or even if a court later concludes, in a suit between the brand-name firm and another generic drugmaker, that the patents are invalid. See Michael Carrier, Payment After Actavis, 100 Iowa L. Rev. 7, 40 (2014) (citing 21 U.S.C. § 355(j)(5)(B)(iv)(I)); see also Hemphill & Lemley, Earning Exclusivity, 77 Antitrust L.J. at 963-64.

169.   Though allegedly not here. See Complaint ¶ 330. Because of the volume caps, profits, it is said, would run to Natco on a slower and more long-term basis.

170.   Accord, e.g., In re Modafinil Antitrust Litig., 837 F.3d 238, 244 (3d Cir. 2016); Hemphill, Paying for Delay, 81 N.Y.U. L. Rev. at 1579; Carrier, Payment After Actavis, 100 Iowa L. Rev. at 38-39; Daniel F. Coughlin & Rochelle A. Dede, Hatch-Waxman Game-Playing from a Generic Manufacturer Perspective: From Ticlid to Pravachol, Apotex Has Difficulty Telling Who's On First, 25 Biotech L. Rep. 525, 525-26 (2006). Beyond six months of high prices, the 180-day exclusivity period also gives the first-filing generic a potentially valuable first-mover advantage. See Hemphill & Lemley, Earning Exclusivity, 77 Antitrust L.J. at 953; Farasat A.S. Bokhari et al., Entry Limiting Agreements: First-Mover Advantage, Authorized Generics, and Pay-for-Delay Deals, 29 J. Econ. & Mgmt. Strategy 516, 519 (2020); Ali Shajarizadeh et al., Newton's First Law as Applied to Pharmacies: Why Entry Order Matters for Generics, 2 Int'l J. Econ. Bus. 201, 202, 204-06 (2015); see generally Aidan Hollis, The Importance of Being First: Evidence from Canadian Generic Pharmaceuticals, 11 Health Econ. 723, 724-25, 732 (2002) (in the context of the Canadian market, cataloging reasons why “the first generic into a market obtains a lasting advantage in market share”); Yu Yu & Sachin Gupta, Pioneering Advantage in Generic Drug Competition, 8 Int'l J. Pharm. & Healthcare Mktg. 126, 129 (2014) (noting that these reasons “provide potential insights for [also] studying the US market”).

171.   See Part IX.B.3.

172.   See footnote 39 (describing the expense of patent litigation).

173.   Like many other litigants.

174.   See, e.g., Farrow, 755 F. Supp. 3d at 557 (“parties to a lawsuit bargain toward settlement in the shadow of their expectations as to how the end-of-the-road trial is likely to turn out”).

175.   In the example, an especially solid expert report.

176.   See footnote 168.

177.   The Hatch-Waxman Act gives the first-filer (here, Natco) 180 days of exclusivity. See Actavis, 570 U.S. at 143, 133 S.Ct. 2223 (citing 21 U.S.C. § 355(j)(5)(B)(iv)). But it might be thought that what the statute gives, the settlement agreement takes. Because if another drugmaker wins a lawsuit in, say, mid-2023, Natco would be held back by the volume caps it had agreed to --- and unable to take full advantage of its 180-day statutory exclusivity period. Cf. Carrier, Payment After Actavis, 100 Iowa L. Rev. at 40; Hemphill, An Aggregate Approach to Antitrust, 109 Colum. L. Rev. at 661. But the Celgene/Natco agreement largely accounts for this. In the event of a “a Final Court Decision holding that all of the then-asserted, unexpired ․ [p]atents are invalid, unenforceable, or not infringed by a [g]eneric,” Natco would be able to enter the market that very day. License Agreement § 1.11(b). And that entry would not be volume-cap constrained. The “[l]icensed [q]uantities” would “no longer apply.” Id. § 5.1(e). Instead, Natco could sell “unlimited quantities of each dosage strength of the Natco [generic p]roduct beginning on the accelerated [l]aunch [d]ate.” Id; see Complaint ¶¶ 119-20 (discussing aspects of these provisions); see also id. ¶ 123 (“[w]hen a settling generic bargains for a poison pill that is triggered upon another generic having successfully shown its generic does not infringe, the financial desirability of the de facto exclusivity is greatly diminished”).

178.   For a highly public (and oft-cited) example, see Protecting Consumer Access to Generic Drugs Act of 2007: Hearing on H.R. 1902 Before the Subcomm. on Com., Trade, and Consumer Prot. of the H. Comm. on Energy & Com., 110th Cong. 61-62 (2007) (statement of Bernard Sherman, Chief Exec. Officer, Apotex, Inc.), https://www.govinfo.gov/content/pkg/CHRG-110hhrg38992/pdf/CHRG-110hhrg38992.pdf (testimony from the CEO of a generic drugmaker, stating that “[first-filing] exclusivity [under the Hatch-Waxman Act] was given as a reward because litigation is expensive,” and indicating, based on his company's experience, that generic drugmakers “can't afford to litigate to bring about market entry if there is no reward at the end of the day”). Note in this vein that two leading scholars have studied a clause that commonly crops up in these sorts of settlement agreements. See Keith M. Drake & Thomas G. McGuire, Generic Entry Before the Agreed-Upon Date in Pharmaceutical Patent Settlements, 16 J. Competition L. & Econ. 188 (2020). The clause's content is not important here. It is enough to say that the clause is triggered only when a later-filing drugmaker wins a patent-infringement suit. See id. at 189 & n.2. The scholars concluded that in the 54 cases that they studied --- the clause was not triggered. See id. at 194. That implies that in the 54 cases there were zero courtroom wins for later-filers. Later-filers have diminished incentives to litigate to a definitive, end-of-the-line courtroom victory. And the scholarship cited here suggests that later-filers are not, in fact, getting to those victories. This can hardly be obscure to industry participants.

179.   This argument suffuses the Defendants’ papers. See Defendants’ Brief at 1-2 (“To state a claim, [the] Plaintiffs must plausibly plead that, ‘at the time Celgene settled with Natco’ in 2015, Celgene made a ‘promise[ ]’ (or ‘assurance’) that Natco would not ‘face any real competition.’ ”) (quoting In re Revlimid, 2024 WL 2861865, at *59); id. at 2 (“[S]uch a promise is the ‘only way that Natco could have been assured.’ ”) (quoting In re Revlimid, 2024 WL 2861865, at *59); id. (“[the] Plaintiffs’ failure to plead the missing assurance is dispositive”); id., at 9 (“[T]he Plaintiffs’ reverse-payment theory states no claim absent a plausible allegation that Celgene ‘promised’ Natco it would face no generic competition during its license period.”) (citation omitted); id. at 12 (“[T]o state a claim that a volume-limited license constituted a reverse payment, [the] Plaintiffs must allege that the Natco settlement provided Natco an assurance that it was not going to face any real competition during the license period․ [V]olume limits do not themselves provide that assurance.”) (cleaned up); id. at 13-14 (“[The Plaintiffs’] theories ․ are no substitute for plausible factual allegations that Celgene contractually assured Natco it would face no competition.”); id. at 16 (“Section 5.1(a) does not provide the missing assurance.”) (cleaned up); id. at 17 (“[The] Plaintiffs’ addition of the word ‘assurance’ ․ is nothing more than a legal conclusion.”); id. (“likely is not an assurance that Natco knew that it was not going to face any real competition”) (cleaned up); id. at 18 (“[The] Plaintiffs’ concession means § 5.1(a) could not have provided any assurance that Natco would not face competition, and so renders [the] Plaintiffs’ theory facially implausible.”); id. at 20 (“It is not as if [the] Plaintiffs allege some unwritten agreement with Natco; to the contrary, by pleading that the 2015 settlement was an ‘integrated agreement’ containing all ‘promises,’ they forswear any theory that any assurance was conveyed by any other means.”) (citing Complaint ¶ 301); id. at 24 (“[The Plaintiffs] fail to plausibly plead the missing assurance from Celgene to Natco.”); Defendants’ Reply at 1 (“[The] Plaintiffs ․ admit they did not plead ․ that Celgene gave Natco an assurance of no competition.”); id. at 2 (“[The] Plaintiffs fail to plead [an] assurance ․ insisting they need not allege an assurance.”) (cleaned up); id. at 4 (“To turn volume limits into a payment, [the] Plaintiffs must plead Natco was assured it could sell its limited volume at such high prices that it would exceed any potential value from unlimited sales.”) (third emphasis added); id. (“[the] Plaintiffs cannot plausibly allege a payment that was made in reverse from Celgene to Natco under Actavis absent an assurance”) (cleaned up); id. at 5 (“without an assurance, there was no payment in the first instance”); id. at 7 (“[The] Plaintiffs have no substitute for an assurance.”) (cleaned up); id. (“What [the] Plaintiffs still fail to allege plausibly is how, at the time of settlement, Natco was assured of scarcity of other generics. Their failure to do so is fatal.”); id. at 8 (“[The] Plaintiffs’ conspicuous omissions of allegations that Natco ․ forecasted this outcome underscores that they certainly cannot plead Natco was assured it.”). And it was also raised at oral argument. See, e.g., Transcript at 20:2-26:1.

180.   See, e.g., Defendants’ Brief at 2 (“that assurance is nowhere in the Natco settlement”) (emphasis added); id. at 8 (“the explicit terms of the Natco settlement foreclose any such [assurance] argument”); id. at 14 (arguing that the Plaintiffs’ allegations “are no substitute for plausible factual allegations that Celgene contractually assured Natco it would face no competition”); Defendant's Reply at 10 (“[the Plaintiffs] affirmatively plead the Natco settlement was an ‘integrated agreement’ containing all ‘promises’ in the four corners of the document, such that any assurance would have to have been conveyed inside the settlement”); Transcript at 8:2-3 (“the promise to Natco that it would not face generic competition had to be in the settlement itself”); id. at 10:5-7 (“you should only be looking at [what] Celgene promised to do inside of the settlement”); see also Defendant's Brief at 9 (“[A]s before, the terms of the Natco settlement speak for themselves; no term in the agreement, individually or collectively, precludes the possibility of competition during the license period. [The] Plaintiffs’ repeated failure to plead that assurance requires dismissal.”); id. (“the entirety of what Celgene promised Natco is reflected in the settlement”); id. at 12 (“the plain terms of the Natco settlement belie the possibility, much less plausibility, of such a promise”).

181.   “[E]xplicit agreement is not ․ necessary [.]” United States v. Gen. Motors Co., 384 U.S. 127, 142, 86 S.Ct. 1321, 16 L.Ed.2d 415 (1966); accord, e.g., Edward J. Sweeney & Sons, Inc. v. Texaco, Inc., 637 F.2d 105, 111 (3d Cir. 1980) (“proof of an express agreement is not required”); see also, e.g., Lifewatch Servs. Inc. v. Highmark Inc., 902 F.3d 323, 333 (3d Cir. 2018); Havens v. Mobex Network Servs., LLC, 820 F.3d 80, 91 (3d Cir. 2016); In re Chocolate Confectionary Antitrust Litig., 801 F.3d 383, 396 (3d Cir. 2015); Alvord-Polk, Inc. v. F. Schumacher & Co., 37 F.3d 996, 1000 (3d Cir. 1994); Terry's Floor Fashions, Inc. v. Burlington Indus., Inc., 763 F.2d 604, 611 (4th Cir. 1985); Comfort Trane Air Conditioning Co. v. Trane Co., 592 F.2d 1373, 1383 (5th Cir. 1979); Wis. Liquor Co. v. Park & Tilford Distillers Corp., 267 F.2d 928, 931 (7th Cir. 1959).

182.   Might a better one for treating the same medical problem have come out in the meanwhile?

183.   See Defendants’ Brief at 2 (“later-in-time events are irrelevant to what Natco knew in 2015”); id. at 20 (rejecting an ex post argument as “makeweight” because “the relevant question is what Natco could have known ‘at the time Celgene settled with Natco’ ”) (quoting In re Revlimid, 2024 WL 2861865, at *59, *62); id. at 30 (“post hoc analysis is improper, because the relevant question is [what] Natco knew at the time of settlement”); Defendants’ Reply at 6 (“what matters legally is not how the market ․ may have developed years later, but rather what Natco could have known ‘at the time Celgene settled with Natco’ ”) (quoting Defendants’ Brief at 20); id. at 7 (“[The] Plaintiffs ․ fail to allege plausibly ․ how, at the time of the settlement, Natco was assured of the scarcity of other generics. The failure to do so is fatal.”); id. at 8 (“whether [shortages and high prices] later arise does not retroactively subject a settlement to antitrust attack”); Transcript at 29:8-11 (“what Natco would have understood were possibilities as of 2015 ․ is the relevant lens”).

184.   See In re Revlimid, 2024 WL 2861865, at *60 (“even if Celgene's volume limited settlement agreements with later generics had the effect of ensuring that no other generics ․ would be incentivized to compete with Natco on price, no such assurance was transferred to Natco by way of its own volume-limited agreement”) (emphasis added); id. at *59 (concluding that it was “not plausible that the volume limited nature of [Natco's] license, standing on its own, could have ensured that Natco could sell its generic ․ ‘at multiples higher’ than it could have if Natco had not been volumed capped”) (emphasis added).

185.   Defendants’ Brief at 30.

186.   Note that the cited authorities refer to analysis at the rule-of-reason stage. Not on a motion to dismiss.

187.   “[I]t is an accepted proposition, logically and legally, that subsequent events may shed light upon, and be relevant in determining, what transpired at an earlier time.” United States v. Sutton, 970 F.2d 1001, 1007 (1st Cir. 1992). This can sometimes allow after-the-fact events to be considered even in the context of an otherwise ex ante approach. But that sort of line can be hard to draw and harder to hold. As to how it might play out here --- the Court expresses no views.

Michael E. Farbiarz, United States District Judge