Thursday, March 30, 2017

Qualcomm allegedly kept Samsung out of the wireless chipset market: unconvincing denial

When routinely checking for Twitter news about Qualcomm's antitrust issues, I found a job ad for an antitrust counsel at Qualcomm. Seriously, if you're an antitrust lawyer looking for a job, this might be one of the most interesting places to be in the months and years ahead. While some secondary issues such as a case brought over oversight duties go away from time to time, various regulators on multiple continents are currently doing everything to provide job security for San Diego-based antitrust attorneys:

Three months after a Korean antitrust ruling against Qualcomm, which was appreciated by industry groups on both sides of the Atlantic, another antitrust front in Qualcomm's multi-front, cross-jurisdictional fight against regulators and device makers has become known: the Korea Fair Trade Commission's (KFTC) concerns are apparently not limited to Qualcomm's licensing terms in general but also Qualcomm's practice (since 1993) of allegedly preventing Samsung from selling its Exynos wireless baseband chips, with CDMA (code division multiple access) technology, to other companies. I read about it on AndroidAuthority, which quotes Qualcomm's denial:

"Qualcomm has never stood in the way of Samsung selling chips to third parties, and nothing in our agreements has ever prevented Samsung from doing so. Any statement to the contrary is false."

At first sight, that denial appears to be complete and clear, but at a closer look it doesn't convince me. Apart from the fact that Qualcomm obviously could never admit to totally anticompetitive behavior (restriction of competition), in this case going back to an agreement signed in 1993 and failed negotiations a few years ago, the denial merely says that Samsung could somehow have sold chips to third parties, but not that Samsung could have sold, for example, CDMA-capable chips to third parties.

As AndroidAuthority notes, Qualcomm sued a Chinese Samsung customer (Meizu), which built some devices incorporating Samsung's Exynos chipset, and I agree with AndroidAuthority that "we have to wonder why the Korean giant only sells its mobile SoCs to one small company in China" (in light of Samsung's large customer base for other types of chipsets).

One of the things I learned from Apple's complaint against Qualcomm was that Qualcomm withheld "rebates" claiming, among other things, that Apple had persuaded Samsung to complain about Qualcomm's conduct to South Korea's antitrustauthority. Now that a competition enforcer has concluded that Qualcomm anticompetitively kept Samsung (for the most part) out of the baseband chipset market, it takes more than an incredible stretch of the imagination to believe that Samsung needed to be persuaded by Apple. By the way, Apple and Samsung can soon celebrate the 6th anniversary of the first Apple v. Samsung patent infringement action (the first California case, which among other types of intellectual property rights also involves design patents and still hasn't been settled or definitively decided). I'm mentioning this because it additionally--though the fact that Samsung was apparently harmed by Qualcomm in two respects (as a device maker and as a supplier of components) is the strongest point in this context--makes it hard to believe that Apple basically talked Samsung into taking action against Qualcomm. Apple and Samsung are rivals in the marketplace, they're adversaries in the courtroom, and while I like both companies' products and admire both companies in different ways, I've also criticized both of them at different times (since Samsung withdrew its standard-essential patent assertions but had--and still has--to fight against patentee overcompensation, I've largely agreed with Samsung in recent years, but before that happened, I was mostly on Apple's side and throughout all those years I usually agreed with Apple to the extent that it was a defendant).

Just this week it became known that Samsung will ship its next flagship Android phone, the S8, in two variants, one incorporating Samsung's own Exynos chipsets and another one with Qualcomm's Snapdragon chip. I read on Twitter that the Exynos version of the S8 is going to be sold in certain markets but not in the U.S., where Qualcomm probably has a lot more leverage based on its CDMA patent portfolio. That is a pity if certain benchmarks, which appear to show a major advantage for the Exynos variant, are true.

The plot is thickening with respect to Qualcomm's two mutually-reinforcing monopolies, and while Apple's antitrust cases against Qualcomm in three jurisdictions are at this stage the best source of information with respect to Qualcomm's practices, Samsung has even been affected in two roles (as a device maker and as a chipset maker), so the longer this takes, the more we'll likely learn about how Qualcomm acquired and held onto its monopoly power (see another AndroidAuthority article: "A lack of alternatives to Qualcomm is hurting the ecosystem").

Share with other professionals via LinkedIn:

Monday, March 27, 2017

Guest post: Nokia’s aggressive patent suits and recent share buybacks; are they related?

The following post was written by Peter W. Rudder, a graduate student at the University of Sydney Business School, who contacted me with some observations and potential conclusions regarding Nokia's share repurchase program and its earth-spanning patent litigation against Apple. I found Peter's analysis insightful, and the fact that Nokia put unusual pressure on Apple through litigation in numerous countries right at the start of a dispute (as opposed to escalating over time, which is the more common approach and also what Nokia did against HTC) could be attributed to some of what Peter has noticed.

Generally speaking, a share repurchase in a situation in which a certain percentage of a company's profitability is being renegotiated either means positive leverage (Nokia's stock would later be worth more that way) if the outcome is good or it can also make things worse (if the stock price goes and stays below where it was at the time of a buyback).

Now the actual guest post:

Nokia's aggressive patent suits and recent share buybacks; are they related?

Back in December Nokia launched several patent suits against long-time adversary, Apple. This was an event that drew much media attention to Nokia, with many calling it a return to their patent trolling days. What didn’t draw nearly as much attention however, was the commencement of a massive share repurchase program that began just over a month before the announcement of the litigation

The announcement came during Nokia's June, 2016 Annual General Meeting in which the board was authorised to repurchase a maximum of 575 million shares, with the authorisation set to expire in December, 2017. Large corporations repurchase shares all the time, most commonly to return capital to existing shareholders or to alter capital structure and almost always when it believes its shares are undervalued. Nokia has made no allusion that it believes its shares are undervalued, stating that the repurchases are only part of its ongoing capital structure optimisation program, which it has invested a total of €7 billion in since October 2015, mainly for de-leveraging purposes as a consequence of the Alcatel-Lucent acquisition.

While it is fair to assume Nokia did not believe its shares were undervalued at the time, the evidence tends to suggest otherwise. The actual repurchase of shares did not commence until November 16, 2016 when Nokia shares were at a 3-year low. This low can mainly be attributed to Nokia's poor Q3’16 earnings results announced in October 2016, compounded with a later announcement from CEO Rajeev Suri on November 15, offering an additional 2% decrease in Networks sales to the existing 2017 guidance figures. The market reaction was harsh, with Nokia's share price falling 18% after the Q3’16 announcements, and many writing off Nokia, saying poor global network sales would stifle any "catalyst for rebound". Despite the prophesised doom and gloom, from November 16, 2016 to March 17, 2017 (the date of the last recorded repurchase) Nokia purchased 93.1 million (€416.6 million in value) of its own shares and saw a 32% increase in its share price.

Even with poor Networks sales and a gloomy outlook from the market, Nokia went ahead with its ambition repurchase program, seemingly unfazed; why? Analysts had already priced in these revenue figures into their models, and concluded that Nokia's share price wouldn’t recover until there was tangible evidence that Networks sales could improve. From this, one could conclude that either Nokia believed it could quickly deliver these results to the market, or that it had an ace up its sleeve that could deliver value to shareholders another way. Sure enough, just over a month later, Nokia announced a slew of patent suits against Apple.

With the launch of these suits, analysts have estimated that Nokia has lost around €150 million in royalty revenues per year from Apple. On the surface, it seems like Nokia is taking an extraordinarily large risk to renegotiate a seemingly small amount of revenue, after all, €150 million is less than a percent of the almost €24 billion Nokia generated in net sales in FY2016. So why go to so much trouble over such little revenue?

Since the disputed patents are mostly to do with smartphone and other consumer electronics devices, all the lost revenue has impacted on Nokia’s Technologies segment. The Technologies segment, once the core of Nokia’s business, now plays a relatively small part, making up only 4.6% of revenues in Q4’16. When it comes to profitability however, the Technologies segment becomes a much more important part of the business, making up 14.3% of positive EBITDA and an EBITDA margin of 52.4% in Q4’16 (report). When compared to the backbone of Nokia’s business, the Networks segment, which has a much lower EBITDA margin of 16%, we can see why Nokia has taken this risk, as each Euro of revenue earned in the Technologies segment contributes to earnings over three times more than it would in the Networks segment. The reason for this is mainly that much of the revenues generated in the Technologies segment come from royalties which are very low risk, high margin cash flows. With an operating profit of €2.2 billion in FY2016, the €150 million represents approximately a 7% loss to Nokia's bottom line, which is a significant hit to its profitability.

The timing of the share repurchases and the announcement of litigation against Apple soon after, seems to suggest that Nokia is confident in its ability to quickly resolve the patent disputes and renegotiate its contracts on more favourable terms. This could possibly entail larger yearly royalty payments and a large sum paid as compensation for the alleged infringement, similar to the resolution of the Ericsson v Apple case in 2015 (Financial Times article; paywall).

The market tends to agree, although a little more conservatively. Analysts from JP Morgan stated that the €150 million in lost revenues should be returned by no later than the end of FY 2017, and are not factoring in the potential for greater royalty payments or a lump sum paid as compensation either. Surely, you’d think, by taking this risk, Nokia is aiming for a much better result than just the return of €150 million. The fact that the potential upside is not priced in to many analyst’s models tends to suggest that they believe future licensing revenues will not be significantly different from the past, and could mean they’re undervaluing Nokia somewhat. The continuing share repurchase and the ongoing litigation, which they are reportedly willing to spend €100 million per year to resolve, seems to support that Nokia believes the potential upside will far exceed the downside, and are expecting a successful result which will have a significant positive impact on their earnings.

Considering the recent share repurchases, the fact that Nokia is willing to spend €100 million per year on this litigation, the success it has had in settling patent suits in the past, and the broadening of Nokia's patents thanks to the recent Alcatel-Lucent acquisition, strongly suggests that Nokia is very confident that it will triumph over Apple. But are they being too confident? As Florian has written previously, Nokia may possibly not receive the outcome it evidently wishes to seek. Given Apple's vast resources, incredibly powerful market position and subsequent leverage over Nokia (see: Apple pulling Withings products from Apple stores), Apple has the potential to draw out this case for years. The longer the case goes on, the longer Nokia are without a significant portion of their earnings, and combined with a shaky outlook for its Networks segment, it could see a significant loss to its share price if shareholders are not delivered relatively quick results. Time is of the essence, so could Nokia's confidence be a risk to its investors? That remains to be seen, but what we do know is that there are certainly interesting times ahead for Nokia.

Share with other professionals via LinkedIn:

Thursday, March 16, 2017

Samsung files petition for Supreme Court review of 2nd Apple case weeks ahead of deadline

Stalling is something else: even though the Chief Justice of the United States had granted Samsung an extension until March 29 for a petition for writ of certiorari (request for Supreme Court review) relating to the second California Apple v. Samsung case , it made its filing on March 10, almost three weeks ahead of the deadline:

17-03-10 Samsung Cert Petition 2nd Apple Case by Florian Mueller on Scribd

Timing is often an interesting indication of a party's priorities. Over these past seven years of Apple v. Android lawsuits (it all started with HTC in March 2010), Android companies--HTC more than anyone else--have often shown the behavior of stallers, at least when they were (as Samsung is here) on the defending end of a litigation (obviously not when they were asserting standard-essential patents themselves). Even parties that don't intend to stall in the slightest (such as Oracle when enforcing its copyrights against Google) typically wait until the end of a filing deadline. It provides them with an opportunity to wait for further relevant developments (case law, public statements by key persons and entities, etc.). So I really am surprised here. Further remedies-related proceedings in that case are ongoing in district court, and a case management conference has just been postponed to next month. With a view to that conference, the Supreme Court is unlikely to make any decision either way in the meantime.

Maybe Samsung believes Apple is going to bring a motion for contempt in connection with an injunction and believes that a more advanced state of its Supreme Court petition will be helpful when seeking stays. It could also be the opposite: with the most important one of the patents-in-suit ('647, often called "quick links") having expired, Samsung might not fear anything and, instead, be pursuing this Supreme Court appeal mostly because of the fundamental principles at stake: overarching issues that affect Samsung in other cases, and not just Samsung, but even Apple would benefit from some of Samsung's proposed statutory interpretations here whenever and wherever its shoe is on the other foot.

The petition as a whole does look very principled. I've never seen a litigant of this nature and stature--no matter which party--who would have managed to be 100% consistent and principled, but of all the motions, petitions and other procedural steps taken by Android companies defending against Apple's (or, in other cases, Microsoft's or Oracle's) patent infringement assertions, I really can't remember a more principled initiative. Obviously, a petitioner's intentions aren't considered by the Supreme Court when deciding on certworthiness, but while the Supreme Court will just focus on the questions presented and their implications, I've been following the entire Apple v. Samsung dispute for almost six years, so I am trying to understand what the parties are trying to achieve. Their last filing with Judge Koh in San Jose said there was no progress regarding a settlement. But neither party has brought a new case against the other in years; instead, various pending lawsuits were withdrawn, with only two U.S. district court cases still awaiting final resolution.

What's ambitious about Samsung's petition is that it raises three questions for review, covering the big three patent litigation questions:

  • validity (here, obviousness),

  • remedies (here, injunctive relief, which is always a more important issue than damages unless damages would really be devastating), and

  • infringement (here, whether all elements of the relevant "quick links" claim were infringed).

If the Supreme Court granted all three, it would be the most comprehensive patent case ever before the top U.S. court, and the implications of a decision could, collectively, go beyond Alice. How did Samsung's petitions fare in the past? The one regarding design patents was a slam dunk. I believed in it 100% from the start, at least in the "article of manufacture" theory, with respect to which cert was granted while a different theory wasn't evaluated. Last year, Samsung brought a little-noticed (I, too, had failed to notice before it was "game over") injunction-related petition that went nowhere, maybe because it wasn't deemed ripe for review. But when evaluating Samsung's track record with cert petitions involving Apple, "1 out of 3" would be the wrong conclusion since one has to weight the importance of the issues and the fact that Samsung only needed to prevail on one of its design patent damages theories, which it did except that there still is some uncertainty as to what the ultimate outcome would be.

The three questions raised have unique strengths-weaknesses profiles from a certworthiness point of view (just talking about certworthiness, not merits):

  • The injunction part is where the petition says something that may get the Supreme Court, especially justices who either were involved with the famous eBay v. MercExchange appeal or care about the related principles anyway, very interested. Samsung argues that the Federal Circuit would basically (and this is my choice of words) gut eBay. I bet Apple will argue (as it did in the past) that a "causal nexus" between infringement and irreparable harm is none of the four eBay factors, while Samsung argues that it is needed. Justice Kennedy's eBay concurrence is nowadays, by far and away, the most influential concurrence in a patent case, and what he wrote in 2006 is probably the closest authority to its own position that Samsung could point to. But the strongest "argument" for getting the Supreme Court interested (which has nothing to do with the merits) is cited at the bottom of page 2 and the top of page 3 of Samsung's petition:

    "As to the injunction decision, its author stated at oral argument, 'I think eBay was wrongly decided .... I think patentees should get injunctions.'"

    The decision's author is Circuit Judge Moore. That statement might persuade the Supreme Court that this case is indeed about eBay reloaded, 11 years after. Samsung also quotes from Chief Judge Prost's dissent, which is quite persuasive, too. What makes Judge Moore's statement so powerful is that even a Supreme Court Justice who doesn't necessarily believe a reasonably strict "causal nexus" requirement is dictated by eBay (or even someone who disagrees with eBay altogether) might find that attitude so dismissive of the highest U.S. court's decision that the Supreme Court would want to take a look. Samsung's cert question quote the two words of the Federal Circuit's majority opinion that sound most eBay-incompatible: "some connection" (between an infringing feature and asserted irreparable harm)

  • As far as the merits are concerned, Samsung's petition exudes maximum confidence with respect to the infringement-of-all-claim-elements part: they say that even if the Supreme Court didn't want to hear this case, the "quick links" infringement judgement "should be summarily reversed or vacated."

    This is the part that would be economically most impactful (about 80% of the $120 million verdict at issue), yet Samsung raised it only as the last of three cert questions. Samsung portrays its position here as what one might call a "no-brainer" that won't be difficult or time-consuming to decide.

    As a software developer, the problem I see with the way the Federal Circuit interpreted the patent here against a previous claim construction is that there's a huge number of client-server software patents out there and if (maybe not all, but still a number of) client-server patents could also be asserted successfully against single pieces of software (here, the client side alone), it would expose to developers to far greater risks. If I were in Apple's shoes, I would probably place particular emphasis on my resistance to this part of the petition because, even if Samsung succeeded on anything else, the net effect would be that roughly 80'% of the original verdict would be affirmed that way (with the rest potentially still going well for Apple), so Apple's PR message could be "most (if not all) of what we won got upheld." But Apple, just like Samsung with its petition, may set priorities based on key principles, and considering how hard Apple fought over the years, the injunction question is probably going to be even more meaningful to it, even if the most important one of the three patents-in-suit in this particular case has already expired.

  • The strongest part of Samsung's argument for cert regarding (non-)obviousness is that it's the most litigated issue in connection with patents but the three judges of the Federal Circuit's panel, who got overruled by an en banc majority, all wrote dissenting opinions that warn against the consequences of the majority decision.

    The patents at issue in this context cover particular aspects of autocomplete and slide-to-unlock functionalities. So Samsung's first cert question relates to the two patents that are substantially less important from a damages point of view than the "quick links" patent.

There is an unofficial fourth issue that Samsung raises and it relates to the proceedings in the Federal Circuit. Samsung points to Professor Chisum's ("Chisum on Patents") and other legal experts' criticism of how things were handled procedurally, with an en banc decision overruling a panel without a hearing and even without further briefing. That part is relevant in connection with the merits questions (validity and infringement), but not to the injunction case, which was a separate appeal. Maybe Samsung felt that a formal cert question about Federal Circuit interna wouldn't be likely to get the Supreme Court's attention, so the procedural part is raised only as a means of undermining the crediblity of the en banc majority decision.

In the design patent damages case, the cert question that the amicus briefs submitted in support of Samsung focused on was also the one that succeeded (it simply was the most interesting question). It will be interesting to see what any amici supporting Samsung will focus on. If past amicus brief activity in different patent cases is any indication, the standard for injunctive relief may very well be the #1 issue for amici. However, if different amici focus on different ones of Samsung's cert questions, then we may see even more amicus brief activity in total here than we did in the design patents case.

The most interesting de facto amicus briefs may already have been filed: the dissents by Chief Judge Prost in the injunction case and by all three panel members, including Chief Judge Prost, in the merits case. Outside of the Samsung group, no one may be more interested in this cert petition succeeding (at least in part) than Chief Judge Prost, whose dissents were very passionate and persuasive in both cases. Samsung quotes her a lot, including among other things her position that the second Apple v. Samsung case "is not a close case" for an injunction.

Share with other professionals via LinkedIn:

Monday, March 6, 2017

The EU Commission's Plan B in the Apple-Ireland "state aid" case: make Apple pay $1.2 billion

Several readers asked for a more specific number after I wrote last month that the European Commission's secondary line of reasoning in the Apple-Ireland "state aid" decision of last August (PDF) would come down to approximately one billion euros (not 13 billion).

The secondary line of reasoning is outlined in paragraph 355 of the EC decision. The gist of it is that the Commission looked at the profitability of other distribution companies and concluded that 3% on sales is an industry-standard profit margin. The numbers themselves are stated in paragraph 97 of the EC decision, with Table 1 (the one relating to ASI, meaning Apple Sales International) being economically relevant and the numbers for AOE (Apple Operations Europe) being pretty much negligible. So, let's apply the standard Irish corporate tax rate of 12.5% (a) to ASI's pre-tax profit as under the Commission's primary line of reasoning ("Plan A profit" column in my table below) and (b) to a hypothetical distributor profit of 3% of ASI's sales ("Plan B profit" column in my table below).

As in paragraph 97 of the EC decision, the numbers below are stated in millions of dollars. Wherever the Commission stated a range, I based my calculation on the middle of the range. Please note that the sum at the end may deviate slightly from the sum of the numbers above it simply due to rounding. Any such differences are negligible.

YearASI SalesPlan A profitPlan B profit
20031 68216550
20042 22326867
20054 068725122
20065 6261 180169
20076 9511 844209
200810 3783 127311
200915 4045 662462
201028 68012 140860
201147 28122 1341 418
201263 25035 2501 898
201362 75026 7501 883
201467 77524 7502 033
total316 068133 9959 482
12.5%:16 7491 185

The Plan A total of $16.7 billion is more than the €13 billion (i.e., approximately $14 billion) that the Commission said would be the starting point for additional taxes to be paid by Apple. But there are explanations for that. Some numbers must be deducted even under the Commission's Plan A, and currency fluctuations explain the rest. ASI's numbers are stated in US dollars but Ireland is a eurozone country. So, basically the EU Commission wants to apply Ireland's 12.5% standard corporate tax rate to ASI's (and AOE's) pretax profits in the relevant period, and that's why an amount of approximately $14 billion has been mentioned over and over.

Under Plan B, however, that recovery amount goes down to approximately $1.2 billion, or less than a tenth of the number that has been making headline news since late August.

Why does the Commission even outline that Plan B in its decision? In paragraph 356, the Commission "contests" that anything other than its Plan A is correct, yet uses it as a fallback line of reasoning in order to argue that Ireland didn't tax ASI on the basis of "a reliable approximation of a market-based outcome in line with the arm's length principle." It may not be the only way in which that EU Commission decision is unusual, but it is the most conspicuous one.

It would not have been unusual at all for the Commission to present two or three legal theories that arrive at materially the same result. If the EC had presented one theory based on Irish statutory law, one based on Irish case law and one based on EU law ("acquis communautaire"), and if all had pointed in the same direction, then that would have made the case stronger, not weaker. I've seen parties make multiple arguments that support a claim, and I've seen judges present more than one theory just to bulletproof their rulings before an appeal. But in all those cases, two or more theories support the same conclusion. Not so here. The Commission's Plan A (tax ASI's pretax profits at a rate of 12.5% even though Irish tax law and even the Commission itself recognize that companies essentially run outside of Ireland need not be taxed there at all) and its Plan B (apply the 12.5% rate to a hypothetical profit amounting to 3% of ASI's sales) share at least one fatal deficiency: the "arm's length principle" as discussed in great detail by the EC in its decision is neither part of Irish statutory law nor Irish case law on the taxation of such entities and the treatment of inter-company charges within a global group. I already talked about that fact last month.

So the fact that there are two theories--with a huge discrepancy (more than a factor of 10) between the results-doesn't make the case as a whole more solid. What must the Commission's approach then be attributed to?

No matter how one looks at it, the fact that there is a Plan A amounting to $14 billion and also a Plan B amounting to less than 10% of that (and even a Plan C, but there is such a lack of specificity at least in the redacted version of the decision and presumably even in the unredacted one that it's impossible to analyze) shows that the Commission is very unsure of what it's doing here.

Let's think of a fictitious parallel. There's one person, Mr. A, demanding money from another person, Mr. B. A tells B that the amount owed is 10 grand, but even if one applied a different theory, it would still be 1 grand, so in A's opinion there can be no opinion that A is right in some way.

The answer most likely lies in politics. The Commission isn't going to collect any of that money itself; it can only (and this is obviously subject to judicial review) order Ireland to collect something from Apple. The Commission wants the amount to be huge, but the Commission would still claim victory as long as any noteworthy amount (and a billion dollars is a lot as long as one doesn't know or consider that the Commission said $14 billion was roughly the right figure) ended up being paid. They would basically say: "Maybe we missed the correct number by a factor of more than 10, but there can be no doubt that we really had to do something in order to right a wrong!" In other words, they'd deny that they wasted taxpayers' money on an investigation of a non-issue.

It may not be necessary for me to reiterate this because I've taken a consistent position on this matter in several posts by now, but I don't even see a reasonably convincing basis on which Ireland would have to collect $1.2 billion. The correct outcome would be for the CJEU to tell the EC that this is all bogus. I just wanted to provide some quantitative analysis in order to complement my previous post on this subject.

Share with other professionals via LinkedIn: