About two weeks ago, BlackBerry announced that it was going to receive a refund of $814.9 million from Qualcomm (mentioned in this post), and Qualcomm confirmed that fact:
"The parties had agreed to arbitrate a contract dispute relating to one specific issue: whether Qualcomm's voluntary per unit royalty cap program applied to BlackBerry's non-refundable prepayments of royalties for sales of a specified number of subscriber units from 2010 through the end of 2015."
BlackBerry's lawyers from the Sullivan & Cromwell firm stated the basis for the payment consistently with Qualcomm's repreentations, but a bit more specifically:
"The dispute arose in 2015 following Qualcomm's agreement to cap certain royalties applied to payments made by BlackBerry pursuant to a licensing deal. Blackberry argued that it was overpaying Qualcomm."
That wording sounds even more like the "rebate" (a term rejected by Qualcomm) Apple says Qualcomm promised under a special cooperation agreement.
There still isn't any indication of the arbitration panel having made a FRAND rate determination. It all sounds like a contract dispute, including what BlackBerry wrote about this in its last annual report:
"On April 20, 2016, [BlackBerry] and Qualcomm entered into an agreement to arbitrate a dispute over the application of a royalty cap agreement related to a license agreement between the parties. The Company filed its Demand for Arbitration and Statement of Claim on May 2, 2016. Qualcomm filed its response on May 16, 2016. Proceedings are ongoing."
The term "specified number of subscriber units from 2010 through the end of 2015" in Qualcomm's press release on this month's arbitration award could mean all or some of the devices BlackBerry sold during the period in question. What I'm interested in (because I believe many readers will be curious, too) is what indication the "rebate" gives us with a view to Qualcomm's standard-essential patent (SEP) royalty demands. A couple of months ago I saw indications, by deducing and inferring information from certain public documents, that Apple may have been paying Qualcomm approximately $20 for its baseband chip and a second amount like that for patent license (a total of $40 per device for the chip and the license). The higher the rebate is on a per-unit basis, the more likely it is that Qualcomm's royalty demands are really that high (we're talking about stratospheric heights compared to what other companies are rumored to receive; for example, financial investors appear to believe that Nokia receives about $2 per device from Apple).
So let's look at publicly-available information in the light most favorable to Qualcomm: that the "royalty cap" applied to all BlackBerry smartphones sold in the years 2010-2015. Not only is that most favorable to Qualcomm but it's also a reasonable assumption.
Here are some quotes from several annual reports by BlackBerry that state unit volumes:
"The Company recognized revenue related to approximately 3.2 million BlackBerry handheld devices in fiscal 2016, compared to approximately 7.0 million BlackBerry handheld devices in fiscal 2015."
"[...] approximately 13.7 million BlackBerry handheld devices in fiscal 2014, compared to approximately 28.1 million BlackBerry handheld devices recognized in fiscal 2013."
"The Company shipped approximately 49.0 million BlackBerry handheld devices in fiscal 2012 compared to 52.3 million devices in fiscal 2011."
That's 153.3 million devices. If the arbitration award is divided by that number, the per-unit figure is $5.31. Regardless of some remaining uncertainty as to whether the royalty cap applied to all BlackBerry smartphones, that number is pseudoprecise since BlackBerry's fiscal year (March 1-February 28) overlaps with only ten months of a given calendar year. But if we round that number down to $5 per unit, then we don't imply more precision than we can deliver and we have enough of a cushion that the number should be just about right.
Theoretically, a $5 rebate could be granted on a royalty payment of $6, but more realistically the refund represents a fraction--not necessarily a small fraction, but still a fraction--of the amount that was paid. Maybe $5 per unit is exactly what BlackBerry wanted. Even arbitration can have an outcome that favors only one party, especially when merit is a binary question , though a middle ground is more common in arbitration. At any rate, I would view the $5 per-unit refund to BlackBerry as another indication of my $20 per-unit royalty estimate not having been off base.
If Qualcomm's royalty levels are indeed extremely high, it comes as no surprise that various major automative and information and communications technology companies are interested in the ongoing FTC v. Qualcomm litigation, as their open letter to President Trump shows.
It may also explain why Qualcomm doesn't want Judge Koh to treat other Qualcomm FRAND antitrust cases (particularly in the EU and in South Korea) as related cases.
But the BlackBerry story also shows that Qualcomm paints a rosy picture when it claims that the industry at large has accepted its royalty rates, with only Apple and Samsung allegedly trying to avoid paying a fair license fee. Let me quote again from that Sullivan & Cromwell PR piece: "Blackberry argued that it was overpaying Qualcomm." And who knows who else...
Finally, just a message for the professional and amateur stock traders who message me via the contact form on this blog. Please appreciate that I don't have the time to answer individual questions. Sometimes I will answer those questions on my blog. Also, I'm always grateful when someone points me to interesting publicly-available information that I may have missed. It has happened quite often. One last thing for the investment folks among you: no matter what seemingly-private email address you use or how you phrase your messages, and whether or not you refer to companies by their ticker symbol, I can easily tell your messages apart from the ones I receive from so-called (not my own terminology) "fanbois" and "phandroids." This includes a recently-received question about whether Qualcomm could obtain an injunction against Apple. I have nothing to say about that except that Qualcomm has been consistent over the years arguing that SEP holders are entitled to injunctive relief, Apple has been consistent that a FRAND commitment and injunctive relief are irreconcilable, I've been consistent in that regard since late 2010, and over the years I've seen Microsoft (which still told the FTC in 2011 that injunctive relief was available over SEPs), Google and Samsung come over to the good side of history. At this stage of the Apple-Qualcomm dispute it's premature to speculate about injunctive relief. Even Pokémon GO is more relevant to this dispute at this juncture.
Share with other professionals via LinkedIn:
