Description

Thesis

We are taking another bite out of Apple (pun intended) in a completely different landscape from our last pitch about 2 years ago, but with what we feel is, once again, a compelling price given the backdrop. It is still apparent that many investors are unwilling to underwrite the legal risks associated with Apple and the FTC, and even less so after Kerrisdale’s negative piece. We think the dire licensing verdict presented by Kerrisdale has a very low probability of systemically hurting Qualcomm’s licensing business to the extent predicted. Furthermore, if one is of the view that QCOM eventually gets some of the Apple business back, the stock is cheap. We take a common sense view that Judge Koh and the FTC would prefer to avoid getting intertwined in altering U.S. and global patent law, especially since Qualcomm’s licensing structures have been in place since the pre-CDMA era. Clearly, 130,000 patents and over $43bn spent on R&D in the last decade is worth something. We believe this intellectual property has value and is why Qualcomm is so well positioned to lead 5G worldwide. For the U.S. government to systematically target Qualcomm would seem to defeat much of the global narrative of the U.S. pushing Qualcomm 5G technology over Huawei technology. While the current valuation might not scream ’cheap’ at first glance, if one accounts for the lost Apple business, elevated legal costs, inflated R&D (due to 5G), and the accrued Apple payments to Qualcomm (~$8bn) in aggregate the stock becomes quite compelling. Finally, in the context of the recent Broadcom takeover attempt ($121bn) of Qualcomm, we think the bid underpins the value of the IP and the overall intrinsic value of Qualcomm. While on a standalone basis Qualcomm may receive a discounted valuation, it is quite obvious, based on Broadcom’s acquisition track record they saw significant synergies as well as a stable licensing business. We also doubt Broadcom would attempt to take on these potential legal liabilities from both the FTC and Apple if Broadcom management viewed it as a systemic issue to the licensing business. In fact, Broadcom executives explicitly stated they wanted the Qualcomm business alone (also thereby agreeing to possibly spin-off QCT for regulatory reasons) and even urged a break-up of the NXP deal.

Apple v. Qualcomm

Apple started a global coordinated campaign against Qualcomm just prior to the termination of its exclusivity agreement with Qualcomm ending in 2016 (year end). It is also apparent that Apple’s complaints to the FTC and other government regulatory bodies sparked a legal battle worldwide. While the Chinese and Korean regulatory bodies have seemed to come to terms with Qualcomm’s royalty rates and business model, within the U.S. legal debate is still very current.

With that in mind, Apple continues to claim Qualcomm should be charging royalties based on the value of the chip rather than the wholesale value of the phone, even though it is capped at $400 per phone (5% license fee; $20 per device). So the argument that Qualcomm continues to take a larger and larger share of the pie without equal contributions to the technology, seems irrelevant with the $400 cap. Based on numerous global licensing deals, it is also obvious that many OEMs licensing at the phone level is much more simplistic and a preferred method rather than a multitude of patents at the chip level. In fact, the FTC even admitted at the end of the Qualcomm case that if companies decided the best way to license chips was at the phone level, the FTC would not object.

Interestingly enough, while Qualcomm and Apple have been squabbling over royalty rates and modem chips, Samsung has quietly slipped into a significant technological lead by licensing and using one of Qualcomm’s newest 5G chips (Snapdragon 855; currently > 100 design wins); Samsung will release the 5G enabled phone in 1H 2019 on Verizon’s network. The upcoming Samsung phone will be one of the first major challenges to Apple where Samsung will have such a distinct technical lead over Apple. Bloomberg recently cited that Apple doesn’t plan on releasing a 5G phone until at least 2020-- awaiting Intel’s 5G (likely inferior given issues with the XMM 8060 modem) XMM 8160 multi-mode modem chip, which is expected to be released in 2H 2019, and commercially available in 2020.

FTC v. Qualcomm

Qualcomm is the holder of standard essential patents (SEPs) crucial to widely adopted cellular standards and has committed to comply with the TIA and ATIS IPR policies. The FTC has accused Qualcomm of refusing to license its SEPs to competing modern chip suppliers, but Qualcomm has argued that these policies only require Qualcomm to license its SEPs to applicants that supply complete devices (OEMs) like mobile handsets-- not applicants (competitors) that supply individual chips.

The FTC has portrayed QCOM as a monopoly that extracts “extortion-level” monopolistic royalties from OEMs. While this coordinated attack has been costly to Qualcomm, we believe the recent legal challenges to Qualcomm’s licensing business is surmountable, especially in light of how this business has been run for more than two decades. While it is clear that early on Qualcomm had monopolistic traits, in the last few years, it could not be any more the opposite. QCOM’s market share has shrunk significantly, as Dr. Chipty stated from the FTC trial, “From 2014 to 2017 in the premium LTE modem chip market, Qualcomm has lost 50 share points.” Furthermore, Chipty said, “In 2018 100% of premium handsets from Apple and Huawei used chips from other companies and only 35% of Samsung’s premium handsets use QCOM chips.”  Financials help corroborate this story: gross margins have come down steadily over the past ten years, and not just the past few years with regards to losing Apple, Huawei, and other Chinese OEM business.

*QCOM Gross margins ‘09-’18

The case from the start has been a seemingly unending stream of questionable events, as summarized by Patrick Moorhead from Forbes.

“The FTC chairwoman announced her resignation on January 13th, the FTC filed suit on the 17th, a “surprise” class action lawsuit was filed against Qualcomm on the 18th related to the FTC suit 24 hours afterward, and Apple filed its suit against Qualcomm on the 20th in San Jose (not Washington D.C), the same day President Trump was inaugurated.  All of this happened in one week, which is mind-blowing.”

Given all the information we have, it would seem highly plausible that the only reason the FTC acted was at the request of Apple. For the most profitable public company in the world to start making claims that Qualcomm is bilking, is laughable in itself-- especially In the context of what has happened to Qualcomm’s market share, margins, and market positioning over the past few years.

5G Catalyst: Millimeter-wave or Low Band

Initially with 5G, there were some technical challenges to the widespread deployment of 5G networks. Some of the concerns surrounded millimeter-wave (mmWave) technologies maintaining connectivity without direct line-of-sight (LOS)-- meaning signals were blocked or partially obstructed, which reduced signal energy and overall performance. To fix these issues, Qualcomm worked closely with 3rd Generation Partnership Project (3GPP) on developing its 5G NR standard cost effectively for both base stations and mobile handsets utilizing mmWave technology. Qualcomm’s most recent solution for the LOS issues has been innovative 3D beamforming and multiple-input, multiple-output (MIMO) antenna techniques-- which means that custom algorithms look for reflected energy when a mmWave line-of-sight (LOS) signal path is blocked by a building or other obstruction, and combine the signal energy from alternative signal paths (including off buildings etc.) into the optimal received signal energy.

Samsung and Verizon are both implementing millimeter-wave (mmWave) as core technologies in their respective 5G network upgrades. Some investors argue that mmWave products don’t have the range needed for large scale adoption, but recent developments have shown remarkable improvements, with distances greater than a half mile (both line of sight and indirect) without significant degradation in bandwidth/latency.

A recent AT&T’s press release states “…We’ve been encouraged by the performance of mmWave in our 5G trials and found that it performs better than expected and is successful in delivering ultra-high wireless speeds under a variety of conditions….”

The pros of mmWave is that it has much higher bandwidth rates and much lower latency than low band sub 6Ghz frequencies. If implementing autonomous driving, IoT, or other mission critical applications you need ultra low latency rates found in higher 24Ghz+ frequencies. The bottom line is that low band frequencies just don’t meet this cutting edge criteria. For example, autonomous vehicles require 1-millisecond latency to safely steer through traffic and maintain awareness of the traffic around them by means of vehicle-to-everything (V2X) communications. These technologies and other upcoming technologies require ultra high speed bandwidth and extremely low latency that currently is only found in mmWave technologies. Longer-term, it is clear as these technologies advance, mmWave technology (which Qualcomm has a significant technological lead in) will be very critical to enabling these advances.

To demonstrate some of the issues with low band spectrum, T-mobile recently gave indications of performance improvements from low band sub 6Ghz implementation. In a presentation, T-mobile executive Karri Kuoppamaki stated that low-frequency 5G implementations should have only about 25 to 50 percent better performance over 4G LTE. If that is the case, it would be just a small fraction of the incremental improvement from mmWave technologies, especially relative to the higher Ghz frequencies. Moreover, latency intervals at lower bands are significantly higher than at mmWave Ghz spectrum.

“The amount of bandwidth available at mmWave frequencies is enormous compared to the amount of frequency spectrum used by 4G and previous wireless network technologies.” Jack Browne-- Microwaves & RF

Valuation

Qualcomm looks expensive on almost any trailing metric since the Apple dispute. QCOM’s earnings have been depressed by significant increases in legal expenses associated with the FTC, Apple, and a class action lawsuit. To illustrate the impact of the costly legal expenses to QCOM, management guided second quarter litigation spend to be up incrementally between 6-8% sequentially (combined R&D and SG&A lines), roughly $600mn annualized or a hit to EPS of roughly 50 cents. While Qualcomm does not specifically break out legal expenses, if one includes existing legal expenses on top of the incremental additions, we think it easily tops $1bn per year. In addition, Qualcomm’s R&D budget is at an all-time high due to the upcoming 5G cycle. And lastly, Qualcomm’s current business with Apple is essentially non-existent, and we estimate that Apple owes roughly $8bn in back payments to Qualcomm through contract manufacturers. All these factors taken in aggregate have significantly obfuscated the “real earnings” picture.

*QCOM: Price to sales ‘09-’18

QCOM: Diluted shares outstanding

QCOM: Forward blended P/E

We believe the core earnings for Qualcomm without essentially any Apple business is around $3.00+ per share. Our example above outlays what we think is the flexibility in the cost structure while taking into account a 30% haircut to royalties from Apple. If Qualcomm gets absolutely no future Apple business (which we think is unlikely) then perhaps one could make the case for the stock being expensive. However, under most scenarios we get sub 12x earnings by making some adjustments to a high cash generative business that should be growing topline for the next few years in the high-single-digits plus.

Worldwide Legal Challenges and Comments from the FTC vs. Qualcomm Case

Courts in both China and Germany have found Apple to be infringing on Qualcomm patents. In Germany, Apple had to pull iPhone 7 and 8 phones out of the country following a ruling from the district court in Munich, which stated that those two specific phones were infringing on QCOM’s intellectual property related to power savings in smartphones. In China, a court ordered a sales ban of some older Apple iPhones similar to the German court ruling over patent infringement. Qualcomm is seeking to now block Chinese sales of the newer iPhones, Apple’s XS and XR. While these rulings on older phones are currently of minimal impact to Apple, they do exemplify how Apple has abused Qualcomm’s IP to its advantage.

Some compelling counter arguments to the FTC’s case (theory) against Qualcomm from court testimonies:

1. Makan Delrahim (DOJ) bashed the FTC’s theory in this case: The Justice Department’s top antitrust enforcer knocked the theory behind the Federal Trade Commission’s antitrust complaint against Qualcomm -- a case currently on trial in California federal court -- saying, “disputes over patent-licensing rates shouldn’t be decided by antitrust law.”

2. Prof. Aviv Nevo of Wharton testified that he saw no real world proof Qualcomm had SupraFRAND rates. “Rates were set before Qualcomm even sold CDMA chips. No changes after it started selling chips.”

3. Nevo looked at the rates Qualcomm charged for licensing during times of alleged market power and in other times. FTC's theory said rates would be higher during times QCOM was strong; however, that's not what Nevo found based on his tests and the data.

Nevo noted that, “Qualcomm royalty rates are the same no matter which chip it actually uses in a mobile device (because licensing terms are set before, last for 10-15 years or longer, while chip prices are set every year). So the royalty rate doesn't affect the choice of the chip.”

4. “If Qualcomm doesn't license at the device level anymore, things could get complicated fast,” Nevo said Notes that wouldn't all swap to licensing at the chip level but would be multiple tiers b/c some tech would apply to overall phone, not just the processor.

5. Interdigital exec Ranae McElvaine said, “the company licenses at the device, not the component, level.” Establishing that this element of its business was not unique.

6. In video testimony from Seungho Ahn, the highest-ranking Samsung licensing executive,  Ahn noted that, “Qualcomm never threatened to cut off chip supply and Samsung wasn't coerced to reach the amendments to the licensing agreement.”

Similarities from the Nokia case:

Nokia commentary (2007):

“Nokia has said it has paid more than $1 billion to Qualcomm in royalties since the early 1990s. The Finnish company began arguing that the rates it was paying were too high -- particularly since Nokia has acquired many patents of its own that could be used to offset those of Qualcomm. Among other legal tactics, Nokia joined with five other big electronics companies in filing antitrust complaints with regulators in Europe.”

“Qualcomm and Nokia have been squabbling in court since the 2001 patent-licensing agreement between them expired in April 2007, causing Nokia to stop making royalty payments to Qualcomm.“ --Don Clark, WSJ

--FTC case a “midnight filing”

“While the FTC versus Qualcomm case could have been green-lighted anytime during Obama’s eight-year tenure, the FTC decided to do it the very last week of the former president’s two terms right before the change in power to the Trump administration. In fact, on January 13, 2017, FTC Chairwoman Edith Ramirez announced her resignation and the FTC filed suit January 17, based on the recommendation of only two of five FTC commissioners. President Trump was inaugurated three days later on January 21st.  Was this an incredible coincidence to hold the vote with only three of five commissioners at the end of Obama’s term with a resigning chairwoman or was the timing coincident with when the votes could assure a case? It sure seems to me, given that with a new FTC crew was coming in, it would have been up to them. How many big decisions does a CEO of a company or any government leader make on the way out of the company or organization, except for things like pardons and commutations? It sure seems like this was snuck in, last minute before Trump came in and few were at the FTC.
FTC Commissioner Maureen Ohlhausen’s dissenting opinion

As I said above, only three of five commissioners were present for the FTC vote to sue Qualcomm as two were already on their way out of the administration. Even with only three still working, it wasn’t a unanimous vote. Two commissioners supported the action, while one dissented, Commissioner Maureen K. Ohlhausen (the Interim Chair). You can find her dissenting statement here on the FTC website.””

Summary

Reasonable judgement leads us to believe that a punitive judgement against Qualcomm from Judge Koh would be counterproductive and would be appealed on multiple levels. The entire political origin of the FTC case is very suspect, as outlined by Forbes’ Patrick Moorhead. To think that the U.S. would significantly impair Qualcomm’s licensing business, makes little sense given the global competitive backdrop, as well as the U.S.’ stance on Huawei. A severe judgement would stifle innovation, lower R&D, and limit the lead of U.S. wireless technology (especially in 5G) over Mediatek, Huawei, and others. It would be quite unprecedented to have the FTC attempt to modify a licensing practice which has been in place for over a decade and thereby altering technology IP value both domestically and abroad-- especially when the Justice Department’s top enforcer, Makan Delrahim, questioned the legitimacy of the entire case, saying “Disputes over patent-licensing rates shouldn’t be decided by antitrust law.”

Catalyst

A favorable judgement in FTC vs. Qualcomm, or a less severe verdict than the stock is pricing in.

A positive judgement in Apple/QCOM case.

5G growth inflection 

A further cost realignment of licensing and chip business