EWCA (2-1) Rejects Tesla’s Request for a Declaratory Judgment that Avanci Rate Is Not FRAND

Last Thursday the Court of Appeal for England and Wales released its approved judgment in (I’m going to give the full caption here) Tesla, Inc. and Tesla Motors Limited v. InterDigital Patent Holdings, Inc., InterDigital Holdings, Inc. and Avanci LLC, [2025] EWCA Civ 193.  Others have published more timely posts than mine (see, e.g., here and here); and I am not inclined to go into the 64-page decision in all its particulars, which involves many fine points on English civil procedure and declaratory judgment law.  Instead, I will give a brief summary, and then a bit of commentary.

In effect, the decision affirms (as we would put it in the U.S.) the decision of Mr. Justice Fancourt from last July (see my post here), but by a 2-1 margin.  Lord Justice Arnold would for the most part allow the appeal, while Lord Justice Phillips and Lady Justice Whipple would not.  So InterDigital and Avanci prevail, as they did before the Patents Court last summer.

Briefly, Tesla sought among other things a declaration “of what the FRAND terms would be a for licence between Tesla and Avanci covering the Avanci 5G pool,” including 3 U.K. patents owned by InterDigital.  Under English law, courts have broad discretionary authority to make declarations that would serve a useful and legitimate purpose; in addition, “for a legal issue to be properly justiciable, it is necessary for there to be an applicable legal rule or standard” (para. 83).  Further, “to serve a foreign defendant [such as Avanci] out of the jurisdiction, the claimant has to satisfy three requirements: (1) that in relation to the foreign defendant there is a serious issue to be tried on the merits i.e. a real prospect of success (the summary judgment standard); (2) a good arguable case that the claim falls within one of the gateways for service out specified in paragraph 3.1 of Practice Direction 6B; and (3) that England and Wales is clearly or distinctly the appropriate forum for the trial of the dispute, and that in all the circumstances the court ought to exercise its discretion to permit service of the proceedings out of the jurisdiction” (para. 40). 

I’m not going to go into all of these issues, but most importantly here the judges differ over whether there is a “serious issue to be tried on Tesla’s claims against Avanci” (para. 85), which relate to Tesla’s allegation that the rate Avanci charges is non-FRAND.  The judicial disagreement stems from the fact that Avanci “acts as agent for all of the [66 SEP] licensors . . . offering a standard patent licence agreement . . .  subject to a modest set of pre-approved possible modifications” (para. 19), but it is not itself a SEP owner. Avanci itself, therefore, unlike the 66 SEP owners (including InterDigital), has never made a FRAND commitment to ETSI; and while it sets the rate it offers based on information provided by those licensors and by licensees, it determines those rates independently of the SEP owners.   Each owner, however, reserves the right to offer individual licenses, and at least one Avanci member (anonymized in the decision) has made its own bilateral deal with Tesla.  So in effect, as Lord Justice Arnold puts it, "Avanci operates in a similar manner to that of collective management organisations in the field of copyright” (para. 21). 

The key question therefore is whether, in the absence of any claim that Tesla is a third-party beneficiary of any commitment made by Avanci itself, there is a serious issue to be tried in the sense that Tesla has “a real prospect of successful claiming one or more of the declarations claimed . . . at trial” (para. 89).  Avanci argues that “it is a requirement for the grant of declaratory relief concerning a legal right that the defendant is either the owner of, or subject to, the legal right relied upon” (para. 90).  Therefore, “Avanci’s argument amounts to saying that the royalty rate of $32 per vehicle which it charges for licences under the Avanci 5G Platform does not have to be FRAND, because (i) Avanci does not itself owe any FRAND obligation and (ii) the FRAND obligations owed by the members are irrelevant because those obligations can only be enforced bilaterally against each SEP owner. It is implicit in this argument that the royalty rate can only be challenged, if at all, through the mechanism of competition law” (para. 94).  Tesla, on the other hand, argues

that the royalty rate charged by Avanci does have to be FRAND. Tesla accept that Avanci does not itself owe any FRAND obligation, but contend that that is not determinative. Tesla allege that, as a matter of commercial reality, the only licence of UK SEPs covered by the Avanci 5G Platform which can be FRAND is a global platform licence of the kind offered by Avanci as agent for the SEP owners because negotiating bilateral licences with more than 65 SEP owners is impracticable. Tesla also allege that, in reality even if not formally, most members of the Avanci 5G Platform rely upon the availability of a licence under that platform as fulfilling their FRAND obligations (para. 95).

Lord Justice Arnold concludes that Tesla has “a real prospect of establishing” its position, and that if it succeeded in doing so the sought-after declaration (that a FRAND rate would be below $32) would serve a useful purpose, insofar as this “would force Avanci to reconsider its position” and would make it clear that the Avanci licensors could not rely on the $32 rate “as discharging their FRAND obligation” (para. 97).   

By contrast, Lord Justice Phillips concludes that there is no serious issue to be tried, because “[t]he jurisdiction of the courts of England and Wales to determine a FRAND licence of a portfolio of SEPs which includes foreign patents is based entirely upon the contractual undertaking of the owner of those patents to grant such licences” (para. 222).  He continues:

What the owners have not agreed to do, on any sensible interpretation of the contractual arrangements with ETSI, is to license their SEPs on a collective basis with other SEP owners, whether on “FRAND terms” or on any terms. The undertaking clearly and distinctly creates an obligation on individual owners to license the Patent Family of their declared SEPs, but it cannot be interpreted as extending to include licensing a portfolio which includes many SEPs owned by other organisations altogether. . . .

 

Does the fact that the owners have voluntarily placed their SEPs on the Avanci 5G Platform change the contractual analysis? I cannot see how it does. The fact that the owners have given undertakings to ETSI, derogating from their rights under the general law to that extent, in no way limits their freedom to exploit their rights in any legitimate way, whether on their own or jointly with others. . . . In my judgment the owners who have joined the Platform have not somehow extended the scope of their undertaking to ETSI or entered any other binding agreement to license their SEPs on a collective basis. . . .

 

A further question is what is meant by a licence on FRAND terms of the SEPs on the Avanci 5G platform. . . . Tesla is plainly contending that the FRAND rate for a licence of all the SEPs on the Platform would be a single collective rate, heavily discounted for bulk and convenience. But the owners of the SEPs have simply not agreed to license their SEPs on discounted collective terms. They may offer to do so voluntarily (as they are in fact doing through the Platform), but they have not undertaken to do so and they cannot, in my judgment, be subjected to an English court’s examination and determination of the rate Avanci is offering (paras. 228-30).

Thus, “[i]n the absence of a contractual foundation for English jurisdiction,” there is no justification for “the English court engaging in the extensive exercise of determining FRAND terms in relation to foreign SEPs as against parties who have not given any relevant contractual undertaking”; “there is no such thing as a free-standing FRAND claim” (paras. 232, 236). 

Similarly, in the view of Lady Justice Whipple:

 

Before this Court, Tesla does not suggest that it has any legally enforceable right against Avanci . . . but still it submits that there is a serious issue to be tried between Tesla and Avanci. In oral submissions, Mr Segan returned to his unamended Particulars of Claim and focused on paragraph 40 where Tesla asserts a right to a licence on FRAND terms covering InterDigital’s SEPs (that is not, I think, of itself a controversial assertion), that it is a beneficiary of the FRAND commitments of all the Avanci 5G Platform members (again, that is probably not a controversial assertion) and then this: that Tesla is “accordingly, entitled to a licence on FRAND terms covering the Avanci 5G Pool” (my emphasis). The “accordingly” is not explained and I cannot see that it follows from the first two statements. Tesla goes on, at paragraph 42 of its Particulars, to assert that “it would be impractical and onerous” for Tesla to have to negotiate bilateral licences with the individual SEP owners, and at paragraph 58 repeats the assertion that Tesla is entitled to enforce the FRAND commitment of each Avanci 5G Pool member, “by seeking a licence from and/or through Avanci covering the entirety of the Avanci 5G Pool” but again without explaining the basis for that asserted entitlement. In short, I cannot identify any basis in Tesla’s pleaded case for suggesting that the SEP owners’ FRAND obligations are to be transposed to, or read into, a licence granted by Avanci over its 5G Platform (para. 245).   

There is much more to each of the three opinions, especially Lord Justice Arnold’s, but I think the above quotes provide an adequate distillation of the crucial portions.  Because Avanci itself has not made a FRAND commitment, it would not be appropriate, in the majority’s view, for a court in the U.K. to exercise jurisdiction to determine whether the terms offered by Avanci are FRAND.   

Here are some initial thoughts:

First, just to be clear, under Unwired Planet, Tesla can request that a U.K. court determine the terms of a global FRAND license in an action against InterDigital or any other SEP owner with relevant U.K. patents.  These would all be bilateral licenses, however, not licenses to the entire Avanci portfolio.

Second, while I am by no means an expert on U.K. procedural law, I’m inclined to agree with Lord Justice Phillips and Lady Justice Whipple that, if Tesla is not a third-party beneficiary of a FRAND commitment made by Avanci itself, the declaration requested here would be going a step too far, by conferring authority on the U.K. courts to establish a pool rate for the whole world.  If, however, that is correct—and if no other forum would be both willing and appropriate for this task either—that may suggest a gap in the regulatory system.  The U.S. Department of Justice’s July 28, 2020 Business Review Letter concerning Avanci, which is cited in Lord Justice Arnold’s opinion (para. 25), noted that “Avanci represents that its current rates for the 4G Platform are FRAND . . . and that Avanci intends its 5G rates also to be FRAND” (p.20).  To the extent, however, that that was a relevant consideration in the DOJ’s conclusion that the Avanci platform was unlikely to harm competition, how exactly is it supposed to be enforced?  Moreover, what exactly does it mean for a pool to charge a FRAND rate?

Consider the following stylized example.  Assume that there are x owners of SEPs relevant to some standard, each of whom owns a portfolio of identical value, for which the FRAND licensing rate per owner would be r.  (I know, FRAND is a range, not a point, but this is a stylized example.)  Assume further that if each owner were to engage in bilateral negotiations with an implementer, it would incur identical transaction costs of c_O.  Therefore, if each owner complies with its FRAND commitment, it earns on net (r – c_O), per implementer.  Assume in addition that there are y implementers, and that if each implementer engages in bilateral negotiations with an owner it incurs transaction costs of c_I.  Each implementer therefore incurs on net (r + c_I) to obtain a bilateral license from any given owner.   Now assume that there is a patent pool that can license all of the SEPs in one package, and in doing so reduce everyone’s transaction costs to zero.  For each pool license, the implementer would pay r_px, and each owner would earn r_p per implementer.  From each owner’s standpoint, the pool offers a better deal as long as r_p > (r – c_O).  From each implementer’s perspective, the pool offers a better deal as long as the pool rate per owner is less than the amount the implementer would incur in connection with a bilateral license, that is, as long as (r + c_I) > r_p.  For the pool rate to be acceptable to both sides, then,

(r + c_I) > r_p > (r – c_O).

For purposes of illustration, let’s plug some (admittedly arbitrary) numbers in.  Suppose that r = 15, c_O = 5, and c_I = 6; and that there are ten owners (x = 10).  Under bilateral negotiations, each implementer pays each owner $15, for a total royalty burden of $150; but each implementer also incurs $6 of transaction costs for each negotiation, or aggregate transaction costs of $60, so it actually incurs $210 in total.  Each owner earns on net $10 per implementer after transaction costs.  With the pool, as long as 21 > r_p > 10, owners and implementers are both better off.  At one extreme, if the pool were to discount the per-owner FRAND rate by $5, owners would be no worse off (still realizing on net $10 each from each implementer), and implementers quite lot better off (incurring only $100 in total to obtain a license from all ten owners, instead of $150 + $60 = $210).  But at the other extreme, if the pool were to charge $21 per owner, implementers would be no worse off (still incurring on net $210 to obtain a license from everyone), and owners would be a lot better off (each earning $21 per implementer as opposed to $15).  If they split the difference in half, the pool rate per owner would be $15.50, and implementers would be somewhat better off (each incurring $155 instead of $210) and owners also would be somewhat better off (each earning on net $15.50 instead of $10, per implementer).  But anything within the range would be mutually acceptable to the parties.

So what would a FRAND pool rate be?  Would it be rx (that is, $150), the FRAND rate with bilateral negotiations multiplied by the number of owners?  Or would it be (r – c_O)x (that is, $100, passing along all of the owners’ avoided transaction costs to implementers)?  Or anything in between those two?  Or could it be anything just shy of (r + c_I)x (that is, anything under $210, passing along the implementers’ avoided transaction costs to owners, but still on net benefiting implementers)?  Note that even if this last option is not FRAND, it nevertheless would be mutually acceptable to the parties.  But if it isn't FRAND, and if its non-FRANDness were effectively unreviewable, because the pool was not a party to a FRAND commitment, that seems like a problem.  Owners might be able to effectively evade their FRAND commitments by setting up pools, on the (perhaps tacit) understanding that the pool would charge an above-FRAND rate.  If this were to happen, antitrust intervention might be an option, but it may be very difficult to prove a violation.