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 cO. Therefore, if each owner complies
with its FRAND commitment, it earns on net (r
– cO), 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 cI. Each
implementer therefore incurs on net (r + cI) 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 rpx, and each owner would earn rp per implementer. From each owner’s standpoint, the pool offers
a better deal as long as rp > (r – cO). 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 + cI) > rp. For the pool rate to be acceptable to both
sides, then,
(r + cI) > rp > (r
– cO).
For purposes of illustration,
let’s plug some (admittedly arbitrary) numbers in. Suppose that
r = 15, cO = 5, and cI = 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 > rp > 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 – cO)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
+ cI)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.