A Chinese court recently ruled in Ningbo Ketian Magnet Co., Ltd. v. Hitachi Metals, Ltd. that a patent-holder’s refusal to license patents for producing rare-earth magnets to competitors constituted an abuse of dominance in violation of China’s Antimonopoly Law (AML).1

Chinese courts and regulators have previously tackled allegations of anticompetitive licensing practices of standard essential patents (SEPs).2 But SEP holders generally commit through a voluntary standard-setting process to license patents incorporated into a technical standard of fair, reasonable, and non-discriminatory (FRAND) terms to any willing licensees. In contrast, the patents at issue in Hitachi Metals were never included in any industry standards or encumbered by any FRAND commitments by the patent holders.

Nevertheless, in April, the three-judge panel of the Ningbo Intermediate Court deemed the patents “essential facilities” indispensable to competing in the production of sintered neodymium-iron-boron (NdFeB) magnets. They concluded that Hitachi abused its dominance in the licensing market for its own NdFeB patents by refusing to license.

Essential facilities, intellectual property, and the AML

The “essential facilities” doctrine establishes an exception to the prevailing principle that one competitor – even one with substantial market power – is not obliged to share its assets and resources with actual or potential rivals. The US Court of Appeals for the Seventh Circuit in MCI Communications Corp. v. AT&T Co. articulated the four elements of an essential facilities claim under the Sherman Act as: “(1) control of the essential facility by a monopolist; (2) a competitor’s inability practically or reasonably to duplicate the essential facility; (3) the denial of the use of the facility to a competitor; and (4) the feasibility of providing the facility.”3 Although the doctrine stems from US antitrust cases involving natural monopolies or regulated industries, the modern US Supreme Court has neither endorsed nor repudiated it. The doctrine has, however, firmly taken root in Europe. Under European practice, the owner of an asset may be subject to a duty to deal where (1) the asset owner holds a dominant position in an upstream market; (2) the asset to which access is being denied is indispensable to competing in a downstream market; (3) refusal to grant access eliminates or threatens to eliminate effective competition in the downstream market; and (4) the refusal has no objective justification. Technical data, SEPs, and copyrighted materials have been treated as essential facilities, and a game developer has invoked the essential facilities doctrine in litigation challenging the denial of access to smartphone app stores. However, courts and authorities have been reluctant to treat non-SEP patents as essential facilities.

The prospect that foreign-held patents might be deemed essential facilities subject to compulsory licensing has long loomed over the AML. The AML’s general rules against “abuse of dominance” list refusal to deal as an illustrative abuse.4 The AML states that its provisions shall not apply to the exercise of intellectual property rights in accordance with laws and regulations, but “shall apply to the elimination or restriction of market competition by the abuse of intellectual property rights.”5

Early drafts of the AML also included additional text explicitly incorporating the essential facilities doctrine, but these provisions were deleted from the final text that took effect in 2008.6 As attention shifted to implementation, foreign scholars, trade groups, and governments continued to warn of the risk that commercially essential patents might be deemed essential facilities subject to compulsory licensing or coercive negotiations under threat of antitrust action.7 Leading Chinese scholars also warned that treating patents as essential facilities might have unforeseen consequences in chilling innovation and investment within China– echoing foreign critiques of the essential facilities.8

Nevertheless, the State Administration of Market Regulation (SAMR) issued new IP Guidelines in 2019 applying the essential facilities doctrine to patents.9 Article 16 addresses the refusal to license IP:

Refusing the licensing of IPRs is a form of exercise of IPRs by undertakings. Under normal circumstances, undertakings are not obliged to conduct transactions with competitors or transaction counterparties. However, if an undertaking which has a dominant market position refuses to license an IPR without a justification, it may constitute the abuse of the dominant market position to exclude or restrict competition. The following factors may be considered for specific analysis: (1) The commitments made by the undertaking on the IPR licensing; (2) Whether other undertakings must obtain the licensing of the IPR for entering the relevant market; (3) The impact of refusal to license the relevant IPR on market competition and the innovations by undertakings and the degree thereof; (4) Whether the rejected party lacks the will and ability to pay reasonable licensing fees; (5) Whether the undertaking has ever made a reasonable offer to the rejected party; and (6) Whether the refusal to license the relevant IPR may damage the interests of consumers or public interest.10

The rare earth magnets dispute

Long simmering frictions over control of the core technology for producing NdFeB magnets set the stage for applying the essential facilities doctrine to patents. The lengthy 69-page opinion surveys the parties’ factual claims, legal arguments, and evidentiary submissions before turning to the court’s own findings of fact and law.

  • Hitachi manufactures NdFeB magnets, with 15 percent of the global market for NdFeB magnets.
  • As a result of past acquisitions and R&D activities, Hitachi holds a patent portfolio of over 600 patents in multiple jurisdictions including China covering key aspects of the NdFeB magnet production process.
  • Some of these patents have been the subject of litigation, including exclusion orders from the US International Trade Commission barring imports of unlicensed products and the partial invalidation of some US patents on the NdFeB materials.
  • As of 2013, Hitachi had issued non-exclusive licenses to nine competing producers of NdFeB magnets (eight Chinese and one British) to practice certain Hitachi patents outside of Japan.
  • The plaintiff, Ningbo Ketian, engaged in research and development of magnet production technology, as well as the production and sale of NdFeB magnets.
  • In 2014, Ningbo Ketian and six other Chinese companies formed an alliance and requested a license to the NdFeB magnet patents from Hitachi Metals.
  • Hitachi indicated that it was unwilling to grant the license.
  • Nevertheless, the alliance and Hitachi executed an NDA and conducted two days of preliminary negotiations in May 2014.
  • Subsequently, Hitachi Metals requested each of the alliance companies to provide responses to technical questions about their production processes and provide samples of their NdFeB magnets.
  • The Chinese companies replied that it would only provide the requested information once Hitachi agreed to the license.
  • Hitachi Metals declined, and negotiations ended.

In December 2014, Ningbo Ketian sued Hitachi, alleging that Hitachi had abused its dominance in violation of the AML by unreasonably refusing to license its commercially essential patents.11 Seven years of jurisdictional challenges, interlocutory appeals, evidentiary hearings, and argument ensued.

Analytical framework

The court explained the elements of a refusal to deal claim under the AML.

In the anti-monopoly review of whether it constitutes a monopoly, the refusal to deal with intellectual property rights should meet all of the following requirements: first, the accused operator has a dominant position in the relevant market due to the possession of specific intellectual property rights. Other operators must obtain the permission of this operator to enter the downstream market. Secondly, the accused operator is able to license the relevant intellectual property rights under appropriate trade conditions but refuse to do so. Third, the refusal to license the relevant intellectual property rights substantially restricts or excludes competition in the downstream market and may harm consumers’ interests and public interest. Fourth, there are no reasonable grounds for refusing to license related intellectual property rights.

The court then described the essential facilities doctrine as a tool for assessing refusal to deal claims.

Generally, the principle of essential facilities must meet the following five requirements: first, the facility is indispensable for other operators to participate in competition; second, the monopolist controls the essential facility; third, the competitor cannot duplicate the same facility with reasonable efforts; fourth, the monopolist unreasonably refuses the competitor to use the necessary facility; fifth, it is possible for the monopolist to provide the necessary facility.

The Court then emphasized that intellectual property rights should only be treated as essential facilities in exceptional circumstances, with due regard to impacts on competition, innovation, consumer welfare, and the public interest.

Intellectual property rights inherently have the "bottleneck effect” due to exclusiveness and exclusivity. Looking from a broader scope, it also has the function of promoting innovation and enhancing public welfare. Hence, the principle of "essential facilities" should be applied very carefully in the field of intellectual property rights. Besides the above five requirements, whether the refusal to license will adversely affect competition or innovation in the relevant market and harm the consumer interest or the public interest should also be considered.

Substantive analysis

As a threshold issue, the court defined two distinct relevant product markets: an upstream technology licensing market for the licensing of NdFeB commercially essential patents and a downstream market for the production of NdFeB magnets.

With respect to the downstream market, the court distinguished NdFeB magnets from other rare earth magnets in terms of demand and supply substitutability. With respect to the upstream market, the court identified two categories of essential patents: patents that were technically indispensable because they could not be circumvented through alternate designs, and patents that were commercially indispensable because the cumulative cost of designing around them would be prohibitive. The relevant market for “patent licensing of essential patents for sintered NdFeB owned by Hitachi Metals” comprised both categories of indispensable patents. This conclusion relied on expert testimony, as well as investment, litigation, and promotional materials from the plaintiff, third parties, and the defendant itself emphasizing the necessity of Hitachi technology in producing reliable NdFeB magnets.

This approach to market definition directly fed the essential facilities analysis. The court found that Hitachi was dominant in the upstream market for licensing for its own essential patents, that use of these patents was indispensable to competing in the downstream market for production of NdFeB magnets, and that rival NdFeB magnet producers could not feasibly reproduce the commercially essential patents. (After all, if other magnet producers were actually able to bypass the Hitachi NdFeB patents by developing alternate non-infringing technology, it would vitiate the factual grounds for both the definition of the upstream market and the resort to the essential facilities doctrine). The fact that Hitachi already licensed the patents to third parties satisfied a fourth element, the feasibility of granting access to the essential facility.

The court then turned to the competitive effects of the refusal to license.

First, the court made a compound determination that (a) the defendant had refused to license the patents on reasonable terms in response to the plaintiff’s “sincere” request for a license; (b) the refusal had “substantially improperly prevented the plaintiff from entering the relevant technology market, thereby substantially restricting or excluding the plaintiff from participating in the downstream market competition as a large-scale competitor”; and (c) such conduct “resulted in insufficient downstream market competition.” The court acknowledged the existence of “unauthorized” producers of NdFeB magnets operating without licenses from Hitachi, but found such unlicensed producers could not be considered an “effective and stable” source of supply due to the legal risks facing such producers and their downstream customers and thus did not constitute evidence of sufficient competition in the downstream market.

Second, the court determined that the defendant’s request for technical information and samples from the plaintiff and other prospective licensees in the course of negotiations was itself an “improper use of market dominance in an attempt to obtain technical information from competitors.” Having observed earlier in the decision that such information “obviously” included trade secrets, the court found that the “the plaintiff’s relevant technology lists requested by the defendant during the licensing negotiations include innovative technologies that can be directly transferred to the relevant applications. The part of the innovative technology involved is likely to be directly used in the relevant market competition of patent licensing.”

Third, the court pointed to the defendant’s statement in the course of negotiations that the plaintiff’s contemporaneous initiation of US litigation to invalidate certain patents demonstrated a lack of “sincerity.” The court equated this negotiation position with the imposition of a “no challenge” clause conditioning a license on contractual commitments not to challenge the underlying patents.

Fourth, the court found that the defendant failed to establish any legitimate justification for denying the requested license. The court suggested that such justifications might have included showing that plaintiff “did not propose appropriate transaction conditions,” that the license would “clearly detract from the legitimate business interests” of the defendant, that the refusal to license would have “an obvious causal relationship with promoting innovation and improving efficiency.” The court added that the “defendant also did not prove that its “refusal to deal” behavior has a positive impact on innovation and efficiency, such as by demonstrating: “an obvious causation between the behaviour and promoting innovation and efficiency;” that its conduct would have “less of an impact of restricting or eliminating market competition” among “reasonable business choices and compared with other behaviours that promote innovation and improve efficiency;” that the conduct “would not exclude or severely restrict market competition;” that the refusal would not “seriously obstruct the innovation of other businesses,” and that “the benefits promoting innovation and efficiency interests can be shared and enjoyed by consumers.”

On this basis, the court concluded that Hitachi’s refusal to license constituted abuse of dominance in violation of the AML.

The court ordered Hitachi to pay RMB4.9 million (~ USD767,000) in damages. Moreover, although the court did not formally order Hitachi to license its patents to the plaintiff, the decision implies that continued refusal in unchanged circumstances would further violate the AML.

Implications for patent holders

The Hitachi Metals opinion marks a path that other plaintiffs might follow by threatening or lodging AML litigation in the course of licensing negotiations for commercially essential patents. For both Chinese and foreign patent holders seeking to stave off such claims, it is prudent to consider both the stated reasoning of the opinion and the potential import of things left unsaid.

Patent infringement and patent validity

Most essential facilities cases in other jurisdictions have involved firms with clear legal title to an asset either revoking existing arrangements granting rivals access or refusing to ever allow access in the first place. The facts around Hitachi Metals, in contrast, appear to involve an infringement by the plaintiff and allegations of bad faith registration against the defendant.

On one hand, the opinion implies that the plaintiff was already infringing Hitachi’s patents. This conclusion flows from the court’s express findings that the plaintiff was already producing NdFeB magnets, that it had no license from the defendant, and that Hitachi’s patents were indispensable to produce NdFeB magnets; these premises imply that the plaintiff was already infringing Hitachi’s patents. Indeed, the court’s competitive effects analysis emphasized that Hitachi’s refusal to license its patents constrained the competitive opportunities of the plaintiff and other unlicensed Chinese producers not by withholding access to the critical technology (which was, by definition, disclosed through the relevant patents) but by restraining them from supplying export markets and domestic customers concerned by the legal risks of sourcing unlicensed products. In this vein, the defendant flagged the contradiction between the plaintiff’s public statements that it produced NdFeB magnets with proprietary technology without infringing the Hitachi patents and the plaintiff’s core contention that the Hitachi patents were indispensable.

On the other hand, the opinion acknowledges the plaintiff’s claims that Hitachi had abused the patent system by filing a barrage of process patents covering technologies and processes purportedly developed and disclosed by Chinese companies. The plaintiff objected that Hitachi had filed too many process patents for any rival reasonably to challenge them all. The opinion also noted two US patents for the basic composition of NdFeB magnets had been partially invalidated in the US, while other early patents had expired.

Neither the validity of the defendant’s patents nor the potential liability of the plaintiff for infringement were before the court. The opinion contains no finding that Hitachi’s remaining patents were invalid, and no conclusion that the plaintiff was infringing. Nevertheless, these circumstances may have impacted other aspects of the court’s analysis.

Negotiations with infringers

In theory, the elements of an essential facilities claim might still be established, even if the plaintiff already misappropriates access to the essential facility. In light of the Hitachi Metals opinion, certain strategies for confronting patent infringement might increase these risks. The court characterized the request for technical information and samples as a bid to compel disclosure of trade secrets. The opinion does not discuss the possibility that Hitachi was seeking information for assessing the extent to which the alliance members were already infringing Hitachi’s patents (which might then impact licensing terms). The court similarly characterized Hitachi’s position that plaintiff’s initiation of patent litigation suggested bad faith in the license negotiations was tantamount to a “no-challenge clause.” The opinion does not distinguish a patent holder’s insistence on a no-challenge clause to forestall future litigation from a patent holder’s decision to resolve an infringers’ challenges through already pending litigation (arguably a more definitive resolution) before negotiating licensing terms.

Objective justifications for patent royalties

The court’s finding that Hitachi failed to establish a legitimate justification for refusal to license may set an unsurmountable hurdle for many patent holders. Modern patent regimes rest on the premise that assuring inventors of the legal right to exclude others from practicing their useful invention for a finite period while profiting from its implementation by the patentee or its licensees is necessary to incentivize investment in innovation and the published dissemination of new technology. IP rules may be recalibrated to optimize the trade-off between restricting stimulating competition and incentivizing innovation, but the dynamic efficiencies of patent systems are intended to benefit consumers. However, in a static view of any discrete patent licensing decision in isolation (disregarding counting the inherent risks and costs of invention), sharing technology with rivals would presumably reduce the rival’s costs (as compared to independent development of alternate technologies) and drive down prices. Thus, it would be difficult to demonstrate the consumer welfare benefits of any particular denial of a patent license without accounting for the overall benefits of the patent system. The defendant appears to have advanced these arguments to no avail.12

Risks of limited licensing

The court cited Hitachi’s prior decision to license its patents to a limited number of competitors as demonstrating the existence of separate upstream licensing and downstream product markets, as demonstrating the feasibility for the patent holder to license its technology, and as providing a reference royalty rate. Patent holders may factor these risks into future decisions of whether to license their critical patents to any third parties at all, and thus avoid creating an upstream market in their own IP.

Deciphering public policy concerns

Many national intellectual property regimes and international IP treaties allow for compulsory licensing of patent rights on public policy grounds in certain circumstances, regardless of the patent owners’ market power or conduct. Indeed, the global COVID-19 pandemic highlighted debate about waivers of IP rights amidst public emergency. But compulsory licensing of intellectual property on industrial policy and national security grounds is fraught with diplomatic and domestic policy risks. The plaintiff explicitly invoked national interests, emphasizing industrial policies prioritizing the development of China’s rare earth technology sector. Although the court mentioned public interest considerations when describing the general statutory framework, the opinion includes no specific factual findings regarding the significance of rare earth technology to Chinese industrial policy or security policy and does not mention such public policy concerns in applying the essential facilities test. This silence is conspicuous, as China’s rare earths sector remains a focal point of trade and investment tensions.

But this silence also invites speculation about the likely outcome on a different record. What if the defendant had simply engaged in a categorical, non-discriminatory, unconditional refusal to license its commercially essential patents to any third parties with no requests for disclosure of technical data or samples and or suggestion of a no-challenge licensing commitment? The essential facilities doctrine as developed in the US and Europe requires no misconduct by the owner of the essential facility; the duty to deal arises instead from the commercial indispensability of the asset. Absent any suspicion of misconduct or bad faith by the defendant, would the facts of Hitachi Metals still have supported a finding of abuse based solely on the ramifications for China’s domestic consumer welfare and public interest?

Ultimately, China’s transition from dependence on foreign-controlled IP to ascendance as an engine of global innovation might temper Chinese courts’ readiness to treat patents as essential facilities. Following Hitachi Metals, foreign holders of commercially vital IP rights should be circumspect in assessing the risks of an essential facilities claim under the AML and devising strategies to mitigate these risks.