While lawyers often obsess over the “legal” terms of an agreement, it is likely that the most important contractual language from the parties’ perspective – at least the business representatives of the parties – is that describing their financial obligations to one another. Attorneys often give far too little time and attention to these financial clauses, assuming that the parties and their accountants, bankers and financial advisors will work out the “numbers” to their satisfaction. This assumption, however, is grossly inaccurate. The financial clauses of an agreement are rife with legal details that can have an inordinate impact on the deal. Thus, while attorneys need not opine as to the market value of a particular technology, they must be prepared to draft and negotiate language that gives effect to the business and financial assumptions of their clients.

This being said, there is a virtually unlimited array of financial clauses that can be deployed in a licensing agreement. Save for limited antitrust considerations (see Chapter 25) and voluntary restraints (e.g., the FRAND licensing commitments described in Chapter 20), there are few legal constraints on the form or amount of compensation that an intellectual property (IP) holder may charge for its IP. Accordingly, parties are relatively free to formulate whatever business arrangement they wish.

This chapter summarizes the most common forms of financial clauses that appear in IP and technology licensing agreements and illustrates how they are used in typical transactions.

8.1 Fixed Payments

Fixed payments are amounts that are predetermined and specified in a contract, usually without reference to the licensee’s revenue or use of the licensed rights. Such payments can be due upon contract signing (or within a short period thereafter), in which case they are called up-front payments.

8.1.1 Up-Front and Lump-Sum Payments

Up-front payments are not uncommon in many types of licensing agreements. Often, privately held biotech companies require up-front payments from their licensees in order to fund their operations prior to the development, approval and marketing of a product. These up-front payments are often made in exchange for exclusive license rights in a particular field or fields of use and are often paid in addition to running royalties on products that are eventually created under the license.

Example: Up-Front Fee

Licensee shall pay Licensor a nonrefundable, up-front fee of $1,000,000 within ten (10) business days from the Effective Date.

If an up-front payment represents the entire consideration for the license, and no further payments or royalties are due, then it is also referred to as a lump-sum payment. When you pay $1.99 for a new app at the Apple AppStore, you have paid a lump-sum fee for a nonexclusive license.

One of the best-known lump-sum license fees was that paid by Microsoft to Spyglass, Inc., the start-up that licensed the original “Mosaic” web browser from the University of Illinois. In the early 1990s, Microsoft realized that the WorldWide Web had significant commercial potential, yet Marc Andressen’s Netscape Communications seemed to have a corner on the market for web browser technology with its then-ubiquitous Netscape Navigator program. To jump-start its own entry into the web browser market, Microsoft sought rights to distribute Mosaic from Spyglass. Microsoft’s first offer was $100,000. As the CEO of Spyglass explained to the New York Times, “the first offer from Microsoft on licensing deals is always $100,000.”Footnote ¹ By December 1994, Spyglass negotiated Microsoft’s lump-sum license fee up to $2 million, which entitled Microsoft to distribute Mosaic with its Windows 95 operating system. Microsoft renamed the browser “Internet Explorer” and began to give it away for free, a business model with which Netscape simply could not compete.

Figure 8.1 Spyglass, Inc. licensed its Mosaic web browser code to Microsoft for $2 million. Mosaic became the basis for Internet Explorer.

Depending on the industry, up-front fees in license agreements can be quite large. One recent study of more than 1,000 biopharmaceutical licenses entered into between 1998 and 2018 found that average up-front fees paid were $11.5 million, with a high of $240 million (for a tumor therapeutic licensed by Exelixis to Bristol-Myers Squibb).Footnote ²

8.2 Running Royalties: the Royalty Rate

For many types of IP the most common form of payment is the royalty. A royalty is a periodic payment that is typically based on the licensee’s manufacture, use or sale of a licensed product or service, whether a new drug, a book or an action figure. These payments are often referred to as “running” royalties because they are paid over the course of the agreement term. The term “earned” royalty is commonly used to refer to royalties based on the licensee’s revenue, usually from sales of licensed products.

Royalties are typically calculated and paid periodically over the course of an agreement – monthly, quarterly, annually or over some other fixed interval. Quarterly royalty payments are the norm in patent licensing agreements, though semi-annual payments are common in literary rights agreements. In general, royalties with respect to a particular period will be due and payable within some reasonable time after the end of the period (e.g., thirty days after the end of the relevant calendar quarter). A royalty report showing the basis for royalties paid is often required to accompany each royalty payment, and in Section 8.9 we will discuss audit and other mechanisms used by licensors to check the accuracy of these reports.

Figure 8.2 Former President Barack Obama received a $20 million advance in 2017 for his memoir Dreams from My Father, the largest single-book advance on record, though some argue that Bill Clinton’s $15 million advance for his 2001 memoir My Life was larger in inflation-adjusted dollars.

Royalties are popular forms of compensation because they tend to align the interests of the licensor and licensee. That is, they typically increase in proportion to increases in the licensee’s own profits attributable to the licensed rights. Thus, if the licensee does well, so does the licensor. But while this general principle sounds straightforward, the calculation of royalties in licensing agreements can become devilishly complex, filled with room for interpretation and opportunistic behavior. This section addresses some of the basic concepts necessary to understand royalty calculations, and Section 8.3 addresses additional clauses used by parties when structuring their royalty arrangements.

8.2.2 Percentage Royalties: The Royalty Rate

For all of their advantages, per-unit royalties do not closely link the licensee’s royalty obligation to its actual revenue from licensed product sales. As a result, many parties to licensing agreements choose to specify royalties in terms of a percentage rather than a fixed rate per product.

8.2.2.1 The Basics

A percentage royalty has two key components: the royalty rate and the royalty base. The royalty base is the amount of licensee revenue that is multiplied by the royalty rate to yield the royalty owed. In other words,

Royalty ($) = rate (%) × base ($).

The royalty base is often expressed in terms of the licensee’s “net sales” of licensed products. The definition of net sales is often one of the most complex and most contentious in a licensing agreement, and is discussed in greater detail in the next subsection.

Example: Percentage Royalty

Licensee shall pay to Licensor earned royalties at the rate of x percent on Net Sales of Licensed Products.

8.2.2.2 Tiered Royalties

Like per-unit royalties, percentage royalties may be flat or tiered. An example of a volume-based tiered percentage royalty schedule is given in Table 8.2:

Table 8.2 A volume-based tiered percentage royalty schedule

Units sold during quarter	Royalty rate (%)
0–5,000	2.5
5,001–50,000	2.0
50,001–100,000	1.5
100,000+	1.0

These royalty rates are incremental quarterly sales tiers. Thus, if the licensee sells 75,000 units in the quarter, and each unit results in net sales of $100, the licensee will pay:

• Units 1–5,000: $100 × 0.025 × 5,000 = $12,500

• Units 5,001–50,000: $100 × 0.02 × 45,000 = $90,000

• Units 50,001–75,000: $100 × 0.015 × 25,000 = $37,500

• Total $140,000

But tiered royalty schedules need not be based only on sales volume. For example, different royalty rates and rate schedules may be based on:

• the geographic market in which a licensed product is sold – for example, many pharmaceutical products are priced differently in different geographic markets, with rates in low-income countries only a fraction of what they are in high-income countries;

• the type of licensed product sold – for example, the DVD patent pools charged very different royalty rates depending on whether the licensee manufactured a DVD player or a DVD disc;

• the date on which the license is granted – for example, licensors sometimes seek to incentivize early adoption of their technology by offering rates that are determined based on when the license agreement was signed.

Finally, tiered royalty schedules may combine all of these, as well as percentage and per-unit royalty rates, in any number of variations.

8.2.2.3 Royalty Rate Levels

One thing that attorneys advising clients in licensing transactions are typically not called upon to do is determine the royalty rate at which the licensed rights will be licensed. The determination of royalty rates, just as the selling price of a product or service, is typically a decision left to business and financial experts. This being said, licensing attorneys should understand the general landscape of royalty rate determination, both to assist their clients during negotiations, and because the framework for negotiating royalty rates often becomes a key factor in IP infringement litigation.Footnote ⁴

In truth, the establishment of a royalty rate in a licensing agreement depends to a large degree on custom and practice in the relevant industry, as well as the negotiation leverage of the parties.Footnote ⁵ For example, patent royalty rates in the semiconductor and electronics industries are often in the low single digits, while software licenses often carry a royalty of 40–50 percent. And even within these general categories there can be significant variation depending on the strength and desirability of the licensed IP, the size and reputation of the licensee, the type of licensed product being authorized, the exclusivity of the license and numerous other factors. For example, according to one industry source, the average royalty rate for licensing a brand or character for use in toys and games is approximately 8 percent.Footnote ⁶ Yet it has also been reported that Disney charges Hasbro between 20 and 25 percent for its exclusive toy/merchandise license of its Star Wars and Marvel Comics properties.Footnote ⁷

Figure 8.3 The Walt Disney Company owns some of the world’s most valuable brands and is reported to earn 15.5 percent of all IP licensing revenue in the United States (IBISWorld, Intell. Prop. Licensing in the US p.35 (March 2020)).

Sample Royalty Rates in Book Publishing

Book publishing agreements contain a range of royalty rates for different outlets and forms of publication. Below is a sample of the rates charged by a US publisher for a novel by a non-celebrity author:

• US hardcover retail sales 10–15 percent of retail price

• Export hardcover sales 6 percent of retail price

• US paperback retail sales 7.5 percent of retail price

• Export paperback sales 5 percent of retail price

• Foreign-language editions 75 percent of publisher’s royalty received

• Electronic editions 25 percent of publisher’s royalty received

• Serialization 50 percent of publisher’s royalty received

Rise and Fall of the 25 Percent Rule

• In an effort to analyze patent royalty rates more systematically, economists beginning in the 1970s postulated that a licensee should pay a royalty equivalent to 25 percent of its anticipated profit from sales of the licensed product. Thus, if the licensee’s profit margin on sales of a licensed product were 16 percent, an appropriate royalty to the licensor would be 4 percent.

  

• As explained by the rule’s leading proponent, Robert Goldscheider, the 25 percent rule is based on the “assumption [that] the licensee should retain a majority (i.e. 75 percent) of the profits, because it has undertaken substantial development, operational and commercialization risks, contributed other technology/IP and/or brought to bear its own development, operational and commercialization contributions.”Footnote ⁸ Empirical work by Goldscheider and others seemed to corroborate the use of the 25 percent rule in a variety of licensing transactions.

• Yet the 25 percent rule has its detractors, who charge that it fails to account for both the importance of the licensed patents to the product sold and the relationship of the parties. In 2011, the Federal Circuit rejected use of the 25 percent rule in reasonable royalty patent damages calculations, calling it “fundamentally flawed” and holding that “there must be a basis in fact to associate the royalty rates used in prior licenses to the particular hypothetical negotiation at issue in the case. The 25 percent rule of thumb as an abstract and largely theoretical construct fails to satisfy this fundamental requirement.”Footnote ⁹ The court also held that evidence of the reasonableness of a royalty rate based on the 25 percent rule failed to satisfy the minimum threshold for admissible “scientific, technical, or other specialized knowledge” under Federal Rule of Evidence 702, as interpreted under Daubert v. Merrell Dow, 509 U.S. 589 (1993). To illustrate the absurdity of the 25 percent rule in practice, the court hypothesized that it “would predict that the same 25%/75% royalty split would begin royalty discussions between, for example, (a) TinyCo and IBM over a strong patent portfolio of twelve patents covering various aspects of a pioneering hard drive, and (b) Kodak and Fuji over a single patent to a tiny improvement in a specialty film emulsion.” From the court’s standpoint, the 25 percent rule was dead.Footnote ¹⁰

8.2.2.4 Hybrid Royalty Rates

As we will discuss in greater detail in Section 24.3, US law does not permit a patent holder to charge a royalty with respect to a patent that is expired, or in a jurisdiction where the patent is not in force. Doing so is said to extend the temporal or geographic scope of the patent grant impermissibly, and is prohibited as patent “misuse.”

As a result, patent licensors often combine licenses of patents with licenses of know-how or trade secrets.Footnote ¹¹ Because trade secret rights have no natural expiration, a trade secret license may technically last in perpetuity.Footnote ¹² Under such a structure, one royalty is charged when and where patent claims remain in effect, and a lower royalty is charged when/where no patent claims are in effect (i.e., the royalty is consideration only for the know-how). As the Ninth Circuit held in Chromalloy Am. Corp. v. Fischmann, 716 F.2d 683 (9th Cir. 1983), when a licensed patent is found to be invalid,

while the licensor is not entitled to recover royalties as such under the patent license, compensation must be allowed to the extent that non-patent assets, such as know-how, are transferred to the licensee in the patent agreement.

But as the court went on to emphasize, it is important that the distinction between royalties for the patent rights and nonpatent rights be clearly delineated in the agreement and not blended in a single rate.

The key question, then, becomes how to allocate royalty payments between patents and know-how. One common approach is to set the know-how-only royalty at 50 percent of the patent + know-how royalty, but there is no precise formula for making this determination. One leading treatise notes that “[t]he prospect of long or perpetual royalties may create the temptation to skew the allocation in favor of the trade secret component” in such a hybrid license.Footnote ¹³ The author warns, however, that such gamesmanship (e.g., allocating 75 percent of the royalty to trade secrets when they represent only 25 percent of the value of the combined patent plus trade secret package) could backfire on the licensor:

Suppose, for example, that a reasonable economic analysis of a bundle of patent and trade secret rights would require payment of 75% of the running royalty for the patents and only 25% for the trade secrets. Suppose further that a “clever” licensor, intent upon maximizing an entitlement to long term royalty and avoiding the risk of loss of royalty due to a finding of invalidity (or, for that matter, to expiration) of the patent, decides to shape the royalty allocation so that only 25% is allocated to the patents and 75% to the technology. The licensor may then, as a patentee, encounter a problem with other prospective licensees. Those licensees, interested in the patents, may not have any interest in the secret but unpatented technology, either because they have their own or believe they can develop it more cheaply than they can purchase it from the licensor … And if the patent-only licensees insist upon most-favored licensee treatment, the prior license arrangement and its artificially low patent rate will haunt the licensor.

Another risk is that the licensor, by deflating the contractual consideration for the patent license may be creating an evidentiary pitfall for the future. Suppose the licensor must establish the value of those patents in patent infringement litigation against third parties. Say that the total “package” rate for licensed patented and trade secret technology is 5%, of which 4% would have been fairly attributable to the patent component, but the licensor has adopted the practice of charging only 1% for the patents and 4% for the trade secret technology. A third party … defendant in an infringement action, [could] argue that the licensor’s licensing practices have established the true value of the licensed patents and hence supply the “reasonable royalty” measure of damages to which the licensor is entitled if it prevails in the patent infringement action.Footnote ¹⁴

While courts have not analyzed the question of an appropriate split between patent and know-how royalties in a hybrid license, one data point of interest appears in Justice Kagan’s opinion in Kimble v. Marvel Entertainment, LLC, 576 U.S. 446 (2015) (reproduced in Section 24.3):

[P]ost-expiration royalties are allowable so long as tied to a non-patent right—even when closely related to a patent. That means, for example, that a license involving both a patent and a trade secret can set a 5% royalty during the patent period (as compensation for the two combined) and a 4% royalty afterward (as payment for the trade secret alone).

Another useful data point is the 50 percent royalty rate reduction seen in Aronson v. Quick Point Pencil Co., 440 U.S. 257 (1979), which was triggered when Aronson failed to obtain a patent on her keyring design (see Section 24.3, Note 10).

Notes and Questions

1. Fixed fees versus running royalties. What are the comparative advantages and disadvantages of fixed fees versus running royalties? When would you advise a client to seek one over the other? Are there any circumstances under which you would advise your client to reject either of these options entirely?

2. Royalty versus royalty share. In the sample list of book publishing rates reproduced above, you will see that there are two ways in which a publisher calculates royalties due to an author: as a percentage of the retail price of the book, and as a percentage of the publisher’s royalty received from a third party. What accounts for this difference in treatment? For additional thoughts on this approach, see Section 8.4 relating to sublicensing income.

3. More rules of thumb. In addition to the 25 percent rule, parties litigating the reasonableness of patent royalty rates have also drawn upon the work of Nobel laureate John Nash, whose work predicted that parties bargaining over a matter would reach agreement when they evenly split the profits attributable to the patented technology.Footnote ¹⁵ This 50–50 profit split became known as the Nash Bargaining Solution, and was frequently introduced in royalty rate cases. But in VirnetX, Inc. v. Cisco Systems, Inc., 767 F. 3d 1308, 1333 (Fed. Cir. 2014), the Federal Circuit rejected the Nash 50–50 rule of thumb as well, on grounds similar to those that it cited in the earlier Uniloc case. Though the Federal Circuit rejected both of these rules of thumb on evidentiary grounds, economists and damages experts continue to employ them when advising clients regarding royalty rates. Is this continuing reliance on these rejected rules of thumb justified? Why or why not?

4. Not too high. Is it always in the licensor’s interest to charge a royalty that is as high as possible? Consider the following assessment:

Perhaps counterintuitively, maximizing the royalty rate may not always be in the best interests of the licensor. If the royalty rate is exceptionally high, it may serve as a disincentive to the licensee because the profit associated with the commercial product will be negatively affected by the royalty. An exceptionally high royalty rate may also incentivize a licensee to develop technology that works around the IP defined in the license agreement.Footnote ¹⁶

Do you agree? How might you advise your licensor client to approach the delicate issue of royalty rate determination?

Problem 8.1

Consider this clause from a license agreement between the University of Texas and IDEXX Laboratories for a veterinary (canine) diagnostic test reagent:

1. 5.1.b [Upon University’s receipt of a notice of allowance of the Patent, Licensee will pay University] a running royalty as follows:

   1. i. Four percent (4.0%) of Net Sales for all Licensed Products Sold to detect Lyme disease alone.

   2. ii. One percent (1.0%) of Net Sales of all License Products Sold to detect Lyme disease in combination with one other veterinary diagnostic test or service (for example, but not limited to, a canine heartworm diagnostic test or service) …

   3. iii. Two and one-half percent (2.5%) of Net Sales for all Licensed Products Sold as a product or service to detect Lyme disease in combination with one or more veterinary diagnostic products or services to detect tick-borne disease(s).

IDEXX sells a combined canine diagnostic test that detects Lyme disease, heartworm and at least one other tick-borne disease. What royalty rate should IDEXX pay with respect to this test? Is there an ambiguity in the agreement? How would you resolve it?Footnote ¹⁷

Problem 8.2

Suppose that you represent Sy Scientific, an inventor who has just filed a patent covering a new polymer-rubber compound that has amazing tensile properties that it retains even at ultra-high temperatures. Sy anticipates applications of this substance, which he has named “slubber,” in markets from space exploration to nuclear power to cooking utensils. Sy has also developed a novel technique for determining the length of time that a batch of slubber must be heated in order to give it optimal tensile properties. How would you go about advising Sy to develop a royalty program for slubber?

8.2.3 Percentage Royalties: Royalty Base

As noted above, the amount that the licensee must pay when a percentage royalty is charged depends both on the royalty rate, discussed above, and the royalty base.

8.2.3.1 Net Sales

In an agreement, the royalty base is often expressed in terms of the licensee’s “net sales” of licensed products or services. Definitions of net sales are often highly contested and heavily negotiated. A simplified version might read as follows.

Example: Net Sales (Licensee-Favorable)

“Net Sales” means the actual amounts received by Licensee with respect to sales of Licensed Products.

As with many clauses involving payment, there are numerous variations on the themes set out above. For example, the licensee will generally favor defining Net Sales with reference to “actual amounts received by Licensee,” as this is a reflection of the licensee’s actual revenue from the licensed products. The licensor, on the other hand, may prefer a definition based on “amounts invoiced by Licensee,” as this reflects the amounts that should have been paid for the licensed products, and is not dependent on the licensee’s collection efforts or the compliance of the licensee’s customers.

Likewise, it is not uncommon to define net sales with reference to amounts received or invoiced by “Licensee and its Affiliates,” particularly if the licensee will be selling or distributing licensed products through its own foreign subsidiaries. Without this addition, net sales could be construed to mean the intercompany transfer prices that the licensee receives from its affiliates, which could be far below market rates.

Example: Net Sales (Licensor-Favorable)

“Net Sales” means the actual amounts invoiced by Licensee and its Affiliates with respect to sales of Licensed Products.

8.2.3.2 Licensed Products

Mathematically speaking, an infinite number of royalty rate and base combinations will yield the same per-unit royalty in any given situation. For example, a royalty rate of 1 percent charged on a $100 base yields a royalty payment of $1.00, as does a rate of 10 percent on a $10 base and a rate of 0.01 percent on a base of $10,000.

Yet, for many reasons, it is important to get the royalty base “right,” particularly when royalty rates are based on general benchmarks in the industry. Thus, if an apparel manufacturer licenses a popular cartoon character for use on children’s pajamas at a royalty rate of 5 percent, and then prints the character only on the pajama tops, must it also pay the 5 percent royalty on the matching pajama bottoms that do not display the character? What about a smartphone manufacturer that licenses a patented film that eliminates scratches on a smartphone screen – must the manufacturer pay the agreed royalty only on the price of the film, or on the entire smartphone? This detail is critical to define, either in the definition of net sales or, preferably, the definition of licensed products as to which net sales are calculated. Consider the following case in which this issue was raised.

Allen Archery, Inc. v. Precision Shooting Equipment, Inc.

865 F.2d 896 (7th Cir. 1989)

WOOD, JR., CIRCUIT JUDGE

This is a contract case arising out of a patent license granted by plaintiff-appellee Allen Archery, Inc. (“Allen”) to defendant-appellant Precision Shooting Equipment (“Precision”) and defendant Paul E. Shepley. The parties dispute the amount of royalties due Allen for the use by Precision of Allen’s invention.

Factual Background

On December 30, 1969, United States patent No. 3,486,495 entitled “Archery Bow with Draw Force Multiplying Attachments” was issued to Holless W. Allen. The patent was assigned to plaintiff-appellant Allen Archery, Inc. in 1973 and it expired in 1986. The patent relates to an archery bow known commonly to archers and the archery industry as the “compound bow.”

The longbow or straight bow has been in existence for centuries and consists of a single piece of material with a single bowstring attached to the ends of the limbs. Another traditional bow, the recurve bow, is similar to the longbow, but its limbs curve forward at the tips where the bowstring is attached. The crossbow is a weapon having a short bow known as a “prod.” The prod is mounted crosswise at the end of a stock.

The compound bow system covered by Allen’s patent employs rotatable pulleys or cams and multiple-line lacing of the bowstring or cable to create compound leverage. The important advantage of the compound bow, as opposed to more conventional bows, is that the compound bow casts an arrow at greater speed with increased striking power while reducing the amount of force needed to draw the bow … A compound bow comprises a handle section and a pair of limbs secured to the handle section. An eccentric wheel or cam is mounted on the end of each limb. A bowstring is trained around the wheels to present a central stretch and two end stretches. The central stretch includes a nocking point for receiving the nock or slotted tail of an arrow. The pulley wheels may be round or oval-shaped and are referred to as eccentrics since they are mounted off center in either case.

The compound bow quickly became popular in archery circles. Within eight years of obtaining a patent, Allen had licensed virtually the entire archery industry. When the compound bow first appeared, all of the compound bows built under the licenses were modifications of the longbow or straight bow. Not until 1982 was a crossbow developed that used a compound bow prod.

Figure 8.4 H.W. Allen (1909–1979) with a prototype of his compound bow invention.

Pursuant to an agreement dated July 1, 1973, Shepley became a licensee under the patent. Precision … is a sublicensee under the patent pursuant to a sublicensing agreement with Shepley dated November 1, 1975.

[The] disputes between Allen and Precision center on their differing interpretations of the licensing agreement. The agreement basically gives Precision a license to manufacture, use, and sell “bows embodying the inventions covered” by the compound bow patent held by Allen. Precision had to pay royalties to Allen during the life of the agreement on each bow sold and on replacement parts. The royalty schedule provided that Precision pay a royalty of 5 1/2 percent of the net selling price on the first 31,000 bows sold during a one-year period and a 5 percent royalty on any other bows sold during that same one-year period. The agreement stipulates that:

Licensor agrees that royalties are not to be paid on accessories such as stabilizers and sights and their mountings, bow quivers and fish reels, which are invoiced, billed or sold as separate items from the complete basic operable bow.

Precision began manufacturing crossbows embodying the compound bow principle in 1982. Allen contends that the “complete basic operable bow” described in the licensing agreement includes both the stock and the prod of the crossbow. The Crossfire, Foxfire, and Spitfire crossbows, all manufactured and sold by Precision, utilize the compound principle. Precision argues that the “complete basic operable bow” is provided by the prod alone and that Allen is entitled to no royalties on the value of the stock. Precision notes that when a crossbow is shipped, the prod is not attached to the stock. Precision claims that the prod of the crossbow could be used as an operable bow. Precision values the prod alone at $75, an amount arrived at through a comparison with its regular bow line and also based on the manufacturing cost of the product.

The parties also dispute the definition of accessories that are exempted from the royalty obligation. Allen contends that the overdraw mechanism which is standard equipment on the Mach II model bow is not an accessory. An overdraw is a device that enables a bow to shoot a shorter-than-normal arrow. The overdraw uses a ledge mounted on the side of the bow handle to support the tip of the arrow. The tip of the arrow can then be supported behind the handle of the bow and does not need to project forward from the front of the bow as is usually the case. Precision claims that the overdraw is an accessory since the Mach II could be a “complete basic operable bow” without it. Allen also asserts that the special camouflage paint applied to some bows is part of the basic bow since it is invoiced and billed as part of the basic bow price. Precision states that the special paint is in reality sold separately and the royalty obligation does not apply.

Discussion

[The issues] revolve around the proper construction of the licensing agreement as it relates to crossbows and “accessories.” Richards v. Liquid Controls Corp., 26 Ill. App. 3d 111, 325 N.E.2d 775 (1975), sets out the method for construing contracts in Illinois:

The primary objective is to give effect to the intention of the parties. This is to be determined solely from the language used in the executed agreement when there is not ambiguity, but a strict construction which reaches a different result from that intended by the parties should not be adopted. Previous agreements, negotiations and circumstances may be considered in finding the meaning of the words used and when there is an ambiguity, or when the language used is susceptible of more than one meaning, extrinsic evidence is admissible to show the meaning of the words used.

Precision asserts that the language of the licensing agreement is ambiguous and it disputes the meaning of such terms as “complete basic operable bow,” “bows embodying the inventions covered by said patent,” and “accessories … invoiced, billed, or sold as separate items.” The district court found that the language used in the agreement was not ambiguous. The court stated that there was no evidence before it concerning prior negotiations or agreements that could be considered in determining the meaning of the terms.

Examining the issue of how crossbows fit under the term “complete basic operable bow,” the district court found that Precision was obligated to pay royalties on the full sale price of any crossbow embodying the Allen compound bow invention. The court rejected Precision’s contention that the prod and the stock could be separated for the purposes of computing royalties. Finding that the agreement was meant to cover all bows embodying the inventions, the court stated that the stock of a crossbow was an integral part of the bow, not a mere accessory.

A review of the record indicates that the district court was correct in its determination that the entire sale price of a crossbow was subject to the royalty payment set forth in the licensing agreement. Precision, in paying royalties to Allen on its compound crossbow sales, had only paid royalties on the supposed value of the prod minus the stock. It unilaterally set this price at $75 per compound crossbow. Precision sold a large number of these compound crossbows, far more than its sales of conventional or noncompound crossbows. Each of these compound bows sold is marked with the Allen patent number.

It is strained logic to argue that because the prod can be separated from the stock, it can qualify as a “complete basic operable bow.” In attempting to define those terms, we must find the “ordinary and usual connotation attributable to those words.” While it may be possible for some people to fire an arrow from the prod minus the stock, it is clear that an ordinary user would not consider the prod on its own to be a complete bow. The owner’s manual shipped with every crossbow clearly explains how the prod is to be mounted on the stock and drawn and shot only after it has been secured to the stock; nowhere is it suggested that the prod is operable as a separate unit. It is of no consequence that Precision did not manufacture crossbows at the time it obtained the license, since the agreement covers all bows embodying the patented principles, including those bows not yet designed or built. The district court’s conclusion that royalties must be paid on the full sale price of the crossbows was correct.

Precision … argues that since the licensing agreement does not specifically detail how royalties should be apportioned for crossbows, overdraw mechanisms, and paint, the court should not include nonpatented elements in the determination of royalty obligations. Precision points to the case of Velsicol Chemical Corp. v. Hooker Chemical Corp., 230 F. Supp. 998 (N.D. Ill. 1964), for the proposition that royalties should be computed in proportion to the use of the patented device. However, the Velsicol case involves a situation that in many ways is unique to the chemical industry. In Velsicol, the defendant’s patented end product used the plaintiff’s patented chlorendic as an ingredient. Unlike the present case, Velsicol concerns products where both the component and the end product are patented by the respective parties. The proportionate use argument does not apply here where there is no issue of relative contribution between the parties.

[We also] consider the question of whether the overdraw mechanism on the Mach II bow and the camouflage paint found on some bows qualify as accessories under the agreement. The district court found that they were not accessories within the meaning of the agreement and ordered that royalty payments be made on the full sale price of the bow including the overdraw and the camouflage paint. The district court emphasized the testimony of a Precision manager who stated that no accessories are attached to a bow when a bow is shipped. Working from that premise, the court determined that, since the overdraw mechanism and the camouflage paint were both attached to the bow when shipped, they could not be accessories.

While this proposition at first may seem to be an over-simplification, by looking closely at the licensing agreement, we conclude that the district court’s finding that items attached to a bow cannot be accessories is correct. The licensing agreement provides that accessories will be excluded from royalty calculations if they are invoiced separately. Precision did not choose to invoice the paint jobs or the overdraw devices separately. It included them, as well as the crossbow stocks, in the invoice prices of the respective bows. This leads to the conclusion that they are part of the complete bow. Also, the agreement names items that it considers accessories, such as “stabilizers, sights, and their mountings, bow quivers, and fish reels.” These items are all clearly separable from the “complete basic operable bow.” Camouflage paint and overdraw mechanisms are not separable items and should not be considered as accessories.

Figure 8.5 The court in Allen Archery held that a crossbow’s stock, overdraw mechanism and camouflage paint were not “accessories” excluded from the royalty calculation, even if they were not covered by the patent claims and could be purchased separately.

Notes and Questions

1. EMVR versus SSPPU. The dispute in Allen Archery required the court to interpret the scope of the royalty base defined in the licensing agreement between Allen and Shepley. A similar analysis is often conducted in patent infringement cases in which no agreement exists between the patent holder and the infringer. In these cases, if infringement is proven, the court must determine both an appropriate royalty rate and a royalty base. When patents cover a single component of a multi-component product, such as the crossbow in Allen Archery, a smartphone or a computer, the court must decide whether the royalty base should be the price of the component covered by the patent or the larger product in which the component is used.

The court considered this issue in Cornell University v. Hewlett-Packard Co., 609 F. Supp. 2d 279 (N.D.N.Y. 2009). In that case, Cornell obtained a patent covering one component of an instruction buffer that could be embodied in a computer processor chip used in computer servers and workstations. Cornell sued Hewlett-Packard for infringement and was awarded damages at a royalty rate of 0.8 percent. The court then had to determine the base to which this royalty should be applied: H-P’s sales of computer servers and workstations, CPU “bricks” including processors, coolant, external memory and power converters, or processors alone. Cornell argued that it was entitled to royalties on the $23 billion that H-P would have made if it had sold the infringing processors as bricks. The court disagreed, holding that Cornell was entitled to royalties only on a base that represented the “smallest salable patent practicing unit” or SSPPU. Though H-P primarily sold computer servers and workstations, it had in the past sold individual processors. Thus, the processor, and not the CPU brick, was the SSPPU that formed the base on which Cornell was entitled to royalties (still a substantial $8 billion sum).

The court in Cornell also held that in order for the holder of a patent covering a component of an end product to receive royalties on the basis of the sale price of an end product (the so-called “entire market value rule” or EMVR), three conditions must be met: (1) the infringing component must be the basis for customer demand for the entire product; (2) individual infringing and noninfringing components must be sold together to form a functional unit; and (3) individual infringing and noninfringing components must be analogous to a single functioning unit, not sold together for mere business advantage.

2. Process royalties. Not all patents relate to products that can be sold. Some patents cover processes or methods for performing services, for manufacturing goods, for improving efficiency, and for many other purposes. In these cases, the royalty base is often expressed in terms of the revenue that the licensee earns from using the patented process. For example, a license of patents covering an automated system for operating a customer service call center might bear royalties based on a percentage of the licensee’s call center revenue, and a patented process for improving the efficiency of an assembly line might bear royalties based on the licensee’s sales of products manufactured on that line. While such arrangements are not uncommon, they require significant analysis and negotiation in order to give the licensor a fair share of revenue derived from its licensed process without capturing value arising from the licensee’s own know-how, techniques and other licensed technology.

3. Reach-through royalties. If a patent covers a research method or tool, it may not be practical for the patent holder to charge a royalty on the “sale” of products covered by that patent. Indeed, there may be no products sold at all. Rather, the research tool may be used to discover new compounds or drug targets, to locate subterranean oil reserves or to predict stock market movements. In each of these cases, use of the tool could result in the discovery or development of something hugely profitable – a new drug, an oil reservoir or a market windfall. On what basis should the patent holder charge a user for the use of the patented research tool? Traditionally, the developer of such a research tool – say, a microscope, chemical reagent or DNA sequencing technique – would charge a one-time fee for the use of the tool, either through the sale of a product such as a microscope or reagent, or as an up-front license fee for the use of a patented method. Some developers of research tools, however, have sought to collect royalties not only on the sale or use of their patented tool, but on the licensee’s revenue derived from products discovered or developed using the research tool.Footnote ¹⁸ These royalties on downstream products are referred to as “reach-through” royalties.

Reach-through royalties have been highly controversial. The National Institutes of Health discourages the use of reach-through royalties with respect to federally funded discoveries, stating that:

[NIH grant] Recipients are expected to ensure that unique research resources arising from NIH-funded research are made available to the scientific research community … If the materials are patented or licensed to an exclusive provider, … royalty reach-through, or product reach-through rights back to the provider are inappropriate.Footnote ¹⁹

In addition, many corporations disfavor the use of reach-through royalties by their potential licensors.Footnote ²⁰ What justification can you see for charging reach-through royalties? Why might a licensee object to the payment of such royalties?Footnote ²¹

4. Including unlicensed products in the royalty base. In Automatic Radio Co. v. Hazeltine Research, Inc., 339 U.S. 827 (1950), Automatic Radio licensed a portfolio of more than 500 radio broadcasting patents from Hazeltine (an early patent assertion entity) in exchange for a royalty based on Automatic Radio’s total sales of radio broadcasting receivers, whether or not they practiced the licensed patents. Four years into the agreement, Automatic Radio objected to paying the minimum royalty required under the license agreement, arguing that none of its products infringed the patents and that requiring it to pay a royalty amounted to patent misuse (discussed in greater detail in Section 24.4). The Court disagreed, holding that the royalty base established under the agreement was “a convenient mode of operation designed by the parties to avoid the necessity of determining whether each type of petitioner’s product embodies any of the numerous Hazeltine patents.”Footnote ²² How does the Court’s reasoning in Automatic Radio compare to that of Allen Archery, in which the licensor was permitted to include the price of unpatented crossbow components in its royalty base?

5. Allocation and copyright. Royalty base issues also arise in copyright licenses. Consider a class action complaint filed by various textbook authors against McGraw Hill. The publication agreement between McGraw Hill and each author requires McGraw Hill to pay the author a percentage of the “selling price” of the book, whether in print or electronic form, less customary deductions. Beginning in 2009, McGraw Hill transitioned many of its textbooks to an electronic platform called Connect. At first, McGraw Hill paid the required royalty based on its sale price of each electronic book on Connect. But in 2020, McGraw Hill announced that it would pay royalties based only on “revenue attributed to the ebook component.” In other words, it would allocate revenue from selling an ebook between the book and the Connect platform itself. The result was a reduction in author royalties of 25–35 percent. What result? See Flynn v. McGraw Hill LLC, Case 1:21-cv-00614-LGS (S.D.N.Y., filed Jan. 22, 2021).

8.2.3.3 Exclusions from Net Sales

As most percentage-based running royalties are expressed in terms of the licensee’s “net sales” of licensed products, the definition of net sales is often negotiated heavily. In addition to the scoping described above, there are a range of exclusions from net sales that often appear in licensing agreements. These generally permit the licensee to exclude from its royalty calculation “pass through” costs that, though billed to customers, do not contribute to the licensee’s bottom line. An example of these follows.

Example: Net Sales (with Exclusions)

“Net Sales” means the actual amounts [invoiced/received] by Licensee and its Affiliates with respect to sales of Licensed Products, less (a) shipping, packaging, delivery and freight insurance costs to the extent separately stated on the invoice; (b) standard quantity discounts, rebates and credits for returned goods; (c) applicable taxes and other duties assessed directly on sales of the Licensed Products; (d) bad debt; and (e) amounts received for training, technical assistance, maintenance, service and support.

Shipping and Packaging.

Fees received by the licensee for shipping, packaging and delivery are often excluded from net sales because these fees are usually paid by the licensee to third-party fulfillment and shipping firms without markup and do not represent actual revenue to the licensee. If the licensee handles its own fulfillment, then this provision may be less appropriate.

Discounts.

This exclusion is important if net sales are based on the amounts that the licensee invoices for licensed products. If the licensee invoices the customer $100 for a product but then extends a $10 discount to the customer, so that the licensee only receives $90, then it is fair for the licensee’s royalty to be based on the invoiced price less the discount. Even so, the licensor may insist that net sales be reduced only for discounts that are “standard” and which relate to the quantity of products ordered, as the licensee should not be permitted to reduce its royalty base.

Rebates.

In some cases the licensee will receive payment for a licensed product and then refund part of that payment to its customer in the form of a rebate. As the licensee does not retain those funds, it will often seek to exclude rebates paid from net sales. The licensor’s response to this request could be that rebates should be considered promotional expenses paid by the licensee, similar to advertising, which are not appropriate reductions to the royalty due.

Returns.

If the licensee sells 100 products but 5 are returned by customers for a refund, then the licensee may argue that its royalty obligation should only be with respect to the 95 products that were not returned. The licensor may respond that its royalties should not be reduced on account of the licensee’s poor product quality. Depending on the industry, this exclusion can be heavily negotiated and may allow deductions from net sales up to an expected return rate (say 5 percent) or may simply adjust the royalty rate to take that return rate into account.

Taxes.

If the licensee collects sales, use or value-added taxes from its customers in connection with the sale of licensed products or services (as it is often legally required to do), and then remits those amounts to the appropriate taxing authority, then it is customary to allow the licensee to deduct those amounts from the definition of net sales.

Bad Debt.

Though the licensee may invoice a customer for a product, there is no assurance that the licensee will be paid. Thus, depending on the industry, the licensee may be allowed a credit for average bad debt (possibly in the 1–2 percent range) on invoiced amounts.

Licensee Services.

If the licensee charges its customers for services (training, maintenance, etc.) as part of its sale of licensed products, then the licensee will often argue that such amounts (which are not paid with respect to the licensed products themselves) should not be used to calculate the royalty payable to the licensor. Licensors may disagree, arguing that such add-on services are enabled only because the licensed products are being sold, and may wish to prevent the licensee from shifting large amounts of revenue from the licensed product price to these services simply to reduce the royalty owed.

Notes and Questions

1. In-kind compensation. The typical percentage royalty arrangement is based on the licensee’s net sales of licensed products or services. But what if some or all of the licensee’s compensation is in the form of noncash consideration? Some noncash consideration – equity securities, marketed products, commodity services and even advertising space/time – is relatively easy to value and a net sales definition can be adjusted to reflect the cash-equivalent market value of such consideration on the date paid. Other noncash compensation – IP licenses, noncompetition covenants, technical assistance – may be more difficult to value, and the licensor will seek to ensure that the licensee is not circumventing its royalty obligations by accepting unreported noncash compensation in exchange for licensed products or services. By the same token, agreements normally contain a range of obligations in addition to payment (confidentiality, indemnification, etc.), and it would be unreasonable for a licensor to insist that each of these obligations be converted to a cash value for the purposes of royalty calculation. If you were a licensor, how might you seek to prevent your licensee from avoiding its royalty obligations through accepting noncash consideration?

8.3 Running Royalties: Adjustments and Limitations

Section 8.2 discusses the basic framework for defining and calculating running royalties. In this section we will discuss some additional provisions that are used in licensing agreements to modify and limit running royalties.

8.4 Sublicensing Income

Running royalties are usually based on the licensee’s (and its affiliates’) revenue from the sale of licensed products. But what if the licensee does not itself sell licensed products, but instead sublicenses its rights to another party who distributes or sells those products? Such arrangements are common in a number of industries, including biotechnology, branded goods and literary works. If a sublicensee is in the picture, what should the licensee be required to pay the licensor? The answer to this question can have significant financial implications for the parties.

There are three general options that can be used to allocate sublicensing income between the licensor and the licensee:

1. 1. Include sublicensee’s revenue in licensee’s net sales: “net sales” on which the licensee’s royalty is based can include revenue received by the licensee, its affiliates and their sublicensees.

2. 2. Include licensee’s sublicensing income in licensee’s net sales: “net sales” on which the licensee’s royalty is based can include all amounts that the licensee receives from its sublicensees, including sublicensing fees, royalties and milestone payments.

3. 3. Share licensee’s sublicensing income: the licensee can pay the licensor a specified percentage of all amounts that the licensee receives from its sublicensees, including sublicensing fees, royalties and milestone payments, which is at a different (and usually higher) rate than the running royalties that the licensee pays on its own net sales (e.g., 50 percent).Footnote ²⁹

The financial effects of these different payment structures can be illustrated by the below example.

1. a. The brand owner grants a US manufacturer (USM) the worldwide right to use a particular brand on apparel at a royalty rate of 25 percent.

2. b. The USM grants a Korean manufacturer (KM) the right to use the brand on apparel in South Korea.

3. c. The USM, knowing that profit margins on Korean branded apparel are very high, charges KM a royalty of 40 percent.

4. d. During a particular quarter, the USM earns $100,000 from sales of branded shirts, and KM earns $500,000.

Figure 8.6 Sales illustrating the importance of carefully allocating sublicensing income.

The USM’s royalty obligation to the brand owner for its US sales is 25 percent of $100,000 or $25,000.

The KM’s royalty obligation to the USM is 40 percent of $500,000 or $200,000.

What, then, must the USM pay the brand owner with respect to KM’s sales? This depends on the payment structure agreed by the owner and the USM.

If they chose Option 1, in which the USM’s net sales are deemed to include the KM’s sales revenue, then the KM’s entire $500,000 revenue is counted in the USM’s net sales, and the USM must pay 25 percent to the brand owner, a royalty of $125,000.

If they chose Option 2, in which the royalty income received by the USM from the KM is included in the USM’s net sales, then the USM must pay the owner 25 percent of the $200,000 royalty paid by the KM to the USM, or a royalty of $50,000.

If they chose Option 3, assuming that the USM and the owner have agreed to split the USM’s sublicensing revenue 50–50, then the USM must pay the owner 50 percent of the $200,000 royalty paid by the KM to the USM, or $100,000.

This example demonstrates the significant financial effect that the treatment of sublicensing income can have. But the effect can be even more stark. Suppose that Korean margins on apparel are much lower than they are in the United States, and the KM can only pay the USM a royalty of 10 percent. In this case, the KM pays the USM a royalty of $50,000 on its revenue of $500,000, and the USM’s payment to the owner is:

• Option 1: 25 percent × $500,000 = $125,000

• Option 2: 25 percent × $50,000 = $12,500

• Option 3: 50 percent × $50,000 = $25,000

Note that under Option 1, the USM earns only $50,000 from the KM but pays $125,000 to the owner, resulting in a net loss to the USM of $75,000. Under this scenario, the individual who drafted the owner–USM license agreement would likely be out of a job.

In reality, Option 3 is the most common method for handling sublicensing income, with a split negotiated at, above or below the 50–50 level. However, attorneys should be vigilant to ensure that definitions of net sales do not inadvertently include sublicensing income in a manner that would distort the financial deal reached by the parties.

8.5 Milestone Payments

In Section 7.3.1 we discussed “milestone” or “diligence” obligations of exclusive licensees. In this section we will cover the financial obligations (i.e., payments by the licensor) that arise in connection with the licensee’s achievement of successive milestones. Just as with up-front payments and royalties, there is no uniform methodology for determining the size of milestone payments. To some degree, these payments can be dictated by the licensee’s anticipated cash needs as its commercialization program for the licensed IP progresses. For example, as a drug candidate advances along the development pathway, the scope and cost of human clinical trials increases dramatically.

Despite this variation, one thing that can generally be said about milestone payments is that the achievement of successive milestones usually triggers increasingly large payments. The following example illustrates this principle.

One study of more than 1,000 biopharma licensing deals signed between 1998 and 2018 found that average total milestone payments were approximately $31 million, with a high of $800 million (in a 2001 deal between Eli Lilly/ImClone and Bristol-Myers Squibb for the tumor drug Erbitux).Footnote ³⁰

8.6 Equity Compensation

When a licensee is a start-up company without substantial financial resources, a licensor may accept shares of the licensee’s capital stock as full or partial compensation for a license. While this arrangement is most common in university spinout licenses (see Chapter 14), it occurs elsewhere as well.

Figure 8.7 Stanford University is reported to have earned $336 million by selling shares of Google stock in the company’s 2004 IPO and subsequent offerings.

The issuance of stock involves corporate and securities laws that are beyond the scope of this book. However, even noncorporate attorneys should be familiar with the basic contractual terms surrounding the issuance of equity securities in licensing agreements.Footnote ³⁶

US universities generally seek equity compensation from start-up licensees in the range of 5–10 percent of the company shares. UK universities are known to seek higher equity shares, in the range of 50 percent.

But what does this percentage actually mean? Usually it refers to a percentage of the total outstanding company stock, including both common and preferred stock,Footnote ³⁷ as of the effective date of the agreement. Unexercised options and warrants to acquire shares of the licensee’s stock are usually not included in this calculation.

For example, suppose that the licensor wants equity compensation equal to 5 percent of the licensee’s equity. Suppose that the licensee has a total of 50,000 shares of common stock issued to its founders, and 10,000 shares of preferred stock, which converts to common stock at a ratio of 1:5. The total outstanding shares, on an as-converted basis, at the effective date is thus 100,000. The licensor’s share will be 5 percent of the total, taking into account the issuance of the licensor’s shares. Thus, the licensor will receive 5,263 shares, as this equals 5 percent of 105,263.Footnote ³⁸

The term “fully diluted” in the above example refers to so-called “anti-dilution” provisions that are often contained in preferred stock terms. In short, these provisions result in the issuance of more shares of preferred stock if additional stock is issued to someone else. So the issuance of stock to the licensor itself could trigger an anti-dilution adjustment for the preferred stockholders, which would result in their having more stock, which would result in the licensor’s share having to increase to reach the required level, and so on. The calculation can be done, but it is a bit complex.

These days, university licensors can ask for a range of additional equity-based protections and rights, including anti-dilution, the right to participate in future stock issuances, board observer rights and the like. The provisions are beyond the scope of most typical licensing agreements and generally require the involvement of attorneys familiar with capital markets and equity issuance laws.

8.7 Cost Reimbursement

When universities and small companies license patents, they often require that the licensee reimburse them for patent prosecution costs incurred prior to the execution of the agreement. If a license is exclusive, then the licensee often covers the entirety of these costs. If the license is co-exclusive, or if it is exclusive only in a particular field, then the cost is often split among licensees. Nonexclusive licensees typically do not reimburse the licensor for prosecution costs (or, if they do, that cost is built into their nonexclusive licensing fees).

The level of patent prosecution costs will vary depending on the complexity of the technology, the developmental stage of the technology, the stage of prosecution (e.g., provisional application, utility application, examination, issuance, post-grant opposition), how many applications and patents have been filed, whether foreign protection has been sought and whether competitors have, or are likely to, opposed the patent(s) at the Patent Trial and Appeals Board (PTAB). For “mature” patents, maintenance fees may also have been paid to the PTO. These costs, when aggregated across jurisdictions, can range from as little as $10,000 to several hundred thousand dollars or more per patent.

Of course, prosecution activity often continues after a license agreement is signed, and maintenance fees will continue to become due with respect to issued patents and trademarks.Footnote ³⁹ In the United States, Europe and other countries, proceedings of various types (inter partes review, oppositions, etc.) can be initiated at patent offices to invalidate issued patents. Because these proceedings are semi-administrative in nature, and are not part of court-based litigation, they are sometimes treated as part of the patent prosecution process. This being said, the costs of these proceedings, while substantially lower than litigation, far surpass typical patent prosecution charges. As a result, parties should be careful about allocating the costs of these proceedings.

If the licensee has agreed to assume responsibility for prosecution matters (see Section 9.5), then the licensee will usually cover ongoing prosecution and maintenance costs. If the licensor retains this responsibility (e.g., if it has granted several exclusive licenses in different fields and has not granted prosecution responsibility to any one licensee), then the licensor may seek periodic reimbursement of at least a portion of its prosecution and maintenance costs. In some cases, these costs may be split evenly among all licensees.

8.8 Most-Favored Clauses

“Most-favored” licensee clauses find their roots in the world of international statecraft, in which the most favorable trade status that can be afforded to another country is that of a “most-favored nation” or MFN. Most-favored clauses are not uncommon in licensing agreements, and often retain the label “MFN” even when used in this private law context.

Most-favored clauses protect the licensee against competitive disadvantage arising from the licensor’s later grant of more favorable contractual terms to a competitor of the licensee. But with this type of clause, more than many others, the devil is in the details. Two examples of MFN clauses are provided here.

The first question to ask when drafting (and negotiating) an MFN clause is how broad its scope should be in terms of agreement coverage. That is, what types of later agreements will need to be compared to the agreement with MFN treatment to determine whether their terms are more favorable? Should a patent license agreement be compared only to other patent license agreements? Or should other types of agreements, such as merger agreements, supply agreements and settlement agreements, also be subject to MFN comparison? This issue, which parties often fail to address in their drafting, is the subject of the Kohle case excerpted below.

The second issue of this nature concerns which future licensees and fields of use are subject to comparison under an MFN clause. That is, should the first licensee be entitled to terms as favorable as those granted by the licensor to entities of all descriptions or only entities that can reasonably be viewed as competing with the first licensee (it is generally accepted that MFN clauses do not apply to intercompany transactions between a licensor and its affiliated companies)? For example, suppose that a patent covers a method for rapidly recharging a lithium-ion battery, and it is licensed to an electric vehicle manufacturer at a flat rate of $7.50 per car. If the licensee has MFN protection, should that protection extend to licenses that the patent holder grants to manufacturers of smartphones at $1.00 per phone? Considering that the price of a smartphone is far less than that of a car, it might seem unreasonable to compare these two licenses. Likewise, a pharmaceutical manufacturer that is licensed to sell a patented drug in the United States should probably not be automatically entitled to the same rates as a manufacturer distributing the drug in the developing world. Finally, a trademark licensor might be reluctant to grant a small, specialty business – say, a producer of hand-crafted porcelain dolls – MFN protection against lower rates that it extends to a large multinational toy manufacturer that will produce far larger quantities of licensed goods at lower price points.

Once these initial scoping questions are decided, the parties must agree which contractual terms are subject to MFN treatment. Suppose that the first licensee pays a running royalty rate of 5 percent to manufacture and sell widgets covered by the licensor’s patents. If the licensor grants a license to a second licensee at a royalty rate of 3 percent, the first licensee’s MFN clause would be triggered. But what if the second licensee pays a large up-front fee in order to secure this lower royalty rate (like prepaying “points” on a mortgage in order to secure a lower monthly interest rate)? Should the first licensee be entitled to the benefit of the 3 percent rate if it made no up-front payment? Or should it be given the option to make a similar up-front payment in order to gain the advantage of the lower running royalty rate? Likewise, what if one licensee purchases equity of the licensor? Should the first licensee be required to make such a purchase in order to enjoy the lower royalty rates enjoyed by the second?

With respect to the comparison of financial terms, some MFN clauses contain a materiality or substantiality qualifier. Licensors will argue that an MFN adjustment should not be triggered based on trivial differences among licenses (e.g., slightly different interest rates for late payments, payment terms or foreign exchange rates). But once such a qualifier is introduced, there will always be an issue of what constitutes a “material” difference. When large amounts are at stake, the parties are well advised to be as specific as possible in this regard, perhaps specifying that any difference in royalty rates or total compensation of more than x percent will trigger an MFN adjustment.

Most MFN protection is limited to protection against more favorable financial terms, as there are hundreds of other contractual provisions – notice periods, warranties, indemnities, etc. – that will vary from agreement to agreement. If a later agreement gives a second licensee forty-five days to cure a breach rather than thirty days, should the first licensee’s MFN clause give it the benefit of that longer cure period? What if the second license also has a less favorable confidentiality clause? Must the first licensee accept the bad terms of the second agreement in addition to the good? And what if some terms in the second license are entirely inapplicable to the first license – how would the electric vehicle manufacturer’s license for battery charging technology be adjusted if a smartphone manufacturer received a large milestone payment upon approval by the Federal Communications Commission? The above example contains some possible limitations on the type, size and field of use of later licenses that are subject to MFN treatment, but additional language may be necessary, depending on the specifics of the parties’ transaction.

Once these terms are decided, the process for implementing MFN treatment must be specified in some detail. This necessarily includes a notification by the licensor of the more favorable terms, a period during which the licensee may consider them, and some mechanism for the licensee to gain the benefit of the more favorable terms. In some cases, an agreement may specify that more favorable terms are automatically extended to the licensee. However, if the licensee would be required to make an up-front payment or the licensor would be required to refund amounts previously paid by the licensee, the parties should have a reasonable period of time in which to calculate and effect such reconciliation. This being said, some MFN clauses (such as example 2 above) specifically exclude any refund of prior amounts paid by the licensee.

Finally, the retroactive effect of an MFN adjustment must be considered. One approach is to make the more favorable terms apply retroactively to the date on which they were first granted to the second licensee. This eliminates any advantage that the licensor may gain by delaying its notification to the licensee. However, retroactive adjustments can have significant accounting and financial implications, which should encourage the licensor to notify the licensee as promptly as possible of the more favorable terms.

Studiengesellschaft Kohle m.b.H. v. Hercules Inc.

105 F.3d 629 (Fed. Cir. 1997)

MAYER, JUSTICE

In 1986, Studiengesellschaft Kohle m.b.H. (SGK) sued Hercules, Inc.; Himont U.S.A., Inc.; and Himont, Inc. (collectively “Hercules”) for patent infringement. Hercules counterclaimed, alleging that SGK had breached the most favored licensee provision of their license agreement by failing to offer Hercules a license with the same terms it offered other licensees. But for the breach, Hercules argued, it would have been licensed under the patents at issue during the period in question, thereby insulating it from infringement. The district court agreed and entered judgment for Hercules. Because SGK has not established that the court made any clearly erroneous findings of fact or error of law, we affirm.

Background

SGK is the licensing arm of the Max Planck Institute for Coal Research in Germany. In the early 1950s, SGK invented a catalyst that could be used to make plastics, such as polyethylene and polypropylene. In 1954, SGK and Hercules entered a “polyolefin contract” (the “1954 contract”) granting Hercules a nonexclusive license under SGK’s “Patent Applications and Patents Issued Thereon.” Although the United States had not issued SGK any patents at that time, the contract contemplated that Hercules would be licensed under any SGK patent issued in the future in the plastics field. The contract included a most favored licensee provision, set forth in pertinent part:

If a license shall hereafter be granted by [SGK] to any other licensee in the United States or Canada to practice the Process or to use and sell the products of the Process under [SGK’s] inventions, Patent Applications or Patents or any of them, then [SGK] shall notify Hercules promptly of the terms of such other license and if so requested by Hercules, shall make available to Hercules a copy of such other license and Hercules shall be entitled, upon demand if made three (3) months after receiving the aforementioned notice, to the benefit of any lower royalty rate or rates for its operations hereunder in the country or countries (US and Canada) in which such rates are effective, as of and after the date such more favorable rate or rates became effective under such other license but only for so long as and to the same extent and subject to the same conditions that such … lower royalty rate or rates shall be available to such other licensee; provided, however, that Hercules shall not be entitled to such more favorable rate or rates without accepting any less favorable terms that may have accompanied such more favorable rate or rates.

The contract also contained a termination clause, which granted SGK the right to terminate the agreement and the licenses upon sixty days written notice if Hercules failed to make royalty payments when due. However, Hercules had the right to cure its default by paying SGK “all sums then due under [the] Agreement,” in which case the licenses would remain in full force and effect.

The parties amended the contract [in 1972] by granting Hercules “a fully paid-up” license through December 3, 1980, the date the ’115 patent expired, under SGK’s “U.S. Patent rights with respect to polypropylene … up to a limit of six hundred million pounds (600,000,000) per year sales.” For sales exceeding that amount, Hercules was obligated to pay SGK royalties of one percent of its “Net Sales Price.” As to SGK’s patents expiring after December 3, 1980, Hercules possessed the right, upon request, to obtain “a license on terms no worse than the most favored other paying licensee of [SGK].” SGK concedes that this provision granted Hercules the “right to the most favored paying licensee’s terms regardless of whether those terms had been granted before or after 1972.” The amendment also provided that the terms and conditions of the 1954 contract remained in “full force and effect except as modified by, or inconsistent with, this amendment.” SGK concedes that “the notice provision, indeed the whole [most-favored licensee] clause, ‘survived the 1972 Agreement.’”

On November 14, 1978, SGK was issued U.S. Patent No. 4,125,698 (’698 patent) for the “Polymerization of Ethylenically Unsaturated Hydrocarbons.” The parties agree that under the 1972 amendment Hercules was licensed under the ’698 patent, without any additional payment, through December 3, 1980. It is also undisputed that this patent is covered by the 1954 agreement, as amended.

In March 1979, SGK sent Hercules a letter terminating the 1954 contract and the licenses granted under it “for failure to account and make royalty payments” when due. In accordance with the agreement, the letter stated that the termination would become effective in sixty days unless the “breach” had been corrected and the payments made. Hercules paid SGK $339,032 within the sixty-day period, which SGK accepted. Although SGK possessed the right to question any royalty statement made by Hercules, and to have a certified public accountant audit Hercules’ books to verify or determine royalties paid or payable, it did not do so.

On May 1, 1980, more than seven months before the expiration of Hercules’ “paid-up” license, SGK granted Amoco Chemicals Corporation (Amoco) a nonexclusive “paid-up” license to make, use, and sell products covered by SGK’s polypropylene patents in the United States. In exchange, Amoco paid SGK $1.2 million. SGK does not dispute that the ’698 patent is covered by this license or that it failed to apprise Hercules of the license at the time it was granted. Hercules first learned of Amoco’s license in 1987, after SGK commenced this action. It demanded an equivalent license retroactive to December 3, 1980. SGK refused, contending that (1) Amoco was not a “paying licensee,” as contemplated by the 1972 amendment; (2) Hercules’ request was too late; and (3) Amoco’s license was granted as part of a settlement agreement.

Figure 8.8 The SGK–Hercules agreement concerned patent rights in polypropylene, a plastic used to make a wide range of products from Tic-Tac containers to furniture.

On December 3, 1986, SGK filed suit in the United States District Court for the District of Delaware, charging Hercules with infringement of the ’698 patent. Hercules counterclaimed, alleging that the 1954 license, as amended, required SGK to notify it of the Amoco agreement in 1980, the terms of which it was entitled to obtain via the most favored licensee provision of the 1954 contract, as amended. Hercules argued that it would have exercised its right to obtain a license on Amoco’s terms had SGK not breached that provision. It claimed, therefore, that it was entitled to such license, retroactive to December 3, 1980, upon paying SGK $1.2 million. The court agreed and entered judgment for Hercules. This appeal followed.

Discussion

SGK concedes that the notice provision was effective but argues that it was only obligated to provide Hercules with notice of any license with terms more favorable than Hercules’ license. In 1972, Hercules obtained a “paid-up” license under SGK’s patents through December 3, 1980. In 1978, the ’698 patent issued. Hercules was licensed under that patent, without additional cost, by virtue of the 1972 license. Because Hercules obtained a “free” license under the ’698 patent for the first 600 million pounds, no terms could be more favorable, according to SGK. So, it had no duty to apprise Hercules of the Amoco license.

SGK’s interpretation does violence to the plain language of the 1954 contract. The notice clause did not condition SGK’s obligation to inform Hercules of other licenses on whether such licenses were more favorable. It required SGK to notify Hercules promptly of the terms of a license granted “to any other licensee.” Under SGK’s construction, the power to determine whether another license was more favorable resided not with Hercules, but with SGK. That simply was not what the agreement provided. It is true that the 1954 contract granted Hercules the right, upon demand, to the benefit of any “more favorable rate or rates.” However, that clause signified nothing more than the commercial reality that Hercules would opt only for a license whose terms it thought were more favorable than its own. It did not divest Hercules of the right to decide which terms were more favorable. Indeed, such a decision will not always be apparent when one considers the myriad combinations of royalty payments, lump-sum payments, and technology transfers a license can effect. Consequently, the court was correct that SGK’s failure to provide notice constituted a breach of the license agreement.

SGK next says that it had no obligation to grant Hercules a license with terms equivalent to those in the Amoco license because Amoco was not a “paying licensee” within the meaning of the 1972 amendment. Again, we turn to the plain language of the license and interpret it anew. The 1972 amendment provided that for any of SGK’s patents expiring after December 3, 1980, including the ’698 patent, SGK would “grant Hercules, upon request, a license on terms no worse than the most favored other paying licensee of [SGK].” SGK contends that Amoco was not a “paying licensee” because it made just one lump-sum payment and no royalty payments; only licensees that make ongoing royalty payments are “paying licensee[s].”

In construing the term “paying licensee,” we must give the words their ordinary meaning unless a contrary intent appears. The ordinary meaning of the term “paying licensee” is one who gives money for a license. See Webster’s II New Riverside University Dictionary 863 (1984) (defining “pay” as “[t]o give money to in return for goods or services rendered”). SGK has not established that the parties intended that the term should mean something else. We see no distinction between one who makes an up-front, lump-sum payment and one who makes continuing royalty payments. Indeed, such a distinction would be doubly doubtful because a “paid-up” license presumably includes potential future royalty payments discounted to their net present value.

SGK also argues that the $1.2 million payment was in settlement of litigation; Amoco was not intended to be a “paying licensee.” But the court found that Amoco paid SGK $1.2 million for a paid-up license for unlimited production under, inter alia, the ’698 patent. SGK has not shown how this finding is clearly erroneous: Amoco was a “paying licensee.”

Even were we to accept SGK’s interpretation as reasonable, however, the provision would be ambiguous because Hercules’ construction is also reasonable. Under such circumstances, and in the absence of any extrinsic evidence clearly establishing the parties’ intent, we construe the term “paying licensee” against the drafter of the language – SGK – under the doctrine of contra proferentem. So, Hercules’ interpretation would still prevail.

According to SGK, even if Hercules is entitled to terms equivalent to those in the Amoco license, it exercised its option too late to be effective. This argument fails because the only requirement in the 1954 contract or its amendments that limits the time in which Hercules must request a license is that it be within three months of receiving the required notice. Because SGK failed to notify Hercules of the Amoco license, that time limitation never began. The court found that Hercules first became aware of the Amoco license in 1987 through discovery in this case. Hercules demanded an equivalent license on or about March 16, 1987, so even if constructive notice could trigger the three-month limitation, Hercules met it.

SGK also contends that the court erred in concluding that Hercules was entitled to a license retroactive to December 3, 1980. It argues that for six years Hercules intentionally manufactured products covered by the ’698 patent, which it thought was invalid, without a license. Only after this court ruled that the patent had not been proven invalid, did Hercules become interested in obtaining a license. It requested a license retroactive to the date its allegedly infringing activities began, thereby insulating itself from any infringement claim. SGK argues that “nothing in Hercules’ option provides for such a right.”

To be sure, neither we nor the parties can know with certainty whether Hercules would have exercised its right to a license on Amoco’s terms in 1980, had it received the required notice. To that extent the prospect of absolving six years of alleged infringement via a retroactive license is troubling. But the uncertainty was caused by SGK’s breach, the consequences of which it must bear. The 1954 contract expressly and unambiguously provides Hercules with the right to obtain the terms of another license “effective, as of and after the date such more favorable rate or rates became effective under such other license.” The agreement must stand as written. Hercules is entitled to the terms of the Amoco license effective May 1980, when the Amoco license became effective.

8.9 Audit Clauses

In many licensing agreements, the licensee’s payments are based on information solely in the licensee’s possession: its revenue and sales figures, its achievement of certain technical and commercial milestones and the like. As a result, the licensee is usually required to submit periodic reports to the licensor informing it of the facts underlying the payments due during the period. These are often referred to as royalty reports or statements.

In most agreements, the licensor has the right to “audit” the licensee’s records in order to verify the information stated in its royalty reports. Many of the cases involving royalty disputes originated with a royalty audit. Such audit provisions are complex to negotiate, however, as the information that they seek is often confidential to the licensee, and of significant commercial and competitive value. Below are two examples of financial audit clauses, illustrating provisions that are favorable to the licensor and the licensee, respectively.

As you can see, audit provisions can vary substantially based on which party drafts them. Below are some of the more contentious issues that are usually negotiated in such clauses: