Tech Stock Thoughts #10: Qualcomm, PayPal, eBay, Pinterest, Apple
Thoughts and takeaways for notable tech earnings reports posted over the last few days.
Qualcomm (2/3) — Their (slight) December quarter revenue miss has gotten a lot of attention, but Qualcomm’s chip unit (QCT) is still in very good shape. QCT revenue was up 81% Y/Y in spite of supply constraints, aided not just by iPhone 12 modem shipments but also 157% RF chip growth (thanks to 5G phone ramps) and 40%+ automotive and networking/IoT growth. In addition, March quarter QCT guidance of $6B-$6.5B is well above a $5.62B consensus, and Qualcomm signaled that QCT would also see strong growth in the June quarter and beyond, thanks to a slew of demand drivers.
It’s the licensing unit (QTL) that’s seeing headwinds right now, due to soft Chinese smartphone demand and broader, COVID-related, phone demand weakness. Though up 18% Y/Y due to Qualcomm’s Huawei settlement, QTL revenue of $1.66B slightly missed consensus, and Qualcomm both issued a March quarter QTL guidance range of $1.25B-$1.45B (largely below consensus) and forecast June quarter QTL revenue would be similar.
Some takeaways here:
- Qualcomm’s post-earnings selloff — and for that matter, Qorvo and Cirrus Logic’s — shows how many of the chip stocks that have surged in recent months now have much smaller margins of error. Qualcomm’s report and call were far from disastrous, but with the stock up ~80% over the prior 12 months going into its print, some moderately bad news for one of its two business segments easily overshadowed good news for the other one.
- The enormous growth that QCT is delivering at a time when global smartphone shipments remain down Y/Y drive home how much of a secular tailwind 5G is for modem and RF chip suppliers (not just Qualcomm, but also MediaTek, Skyworks and Qorvo). And with 5G phones still only expected to account for about a third of 2021 smartphone shipments (albeit a much larger % of high-end shipments), and automotive/IoT 5G adoption in its early stages, there’s still a lot of headroom here.
- With Qualcomm estimating that 3G/4G/5G phone shipments were down 7% Y/Y in Q4, Apple, which reported 17% December quarter iPhone sales growth and guided for company-wide revenue growth to accelerate in the March quarter from the December quarter’s 21%, seems to be gaining meaningful share right now. This seems particularly the case in China, given Qualcomm’s comments about Chinese phone demand and the fact that Apple’s “Greater China” revenue was up 57% Y/Y last quarter.
(Qualcomm’s earnings report, results/guidance tables)
PayPal (2/3)— Their constant-currency payment volume (TPV) growth was at an impressive 36% for the second quarter in a row. Revenue was only up 23% due to take rate pressures (among them, lower eBay revenue and a mix shift to P2P and bill-pay transactions), but that’s still several points above pre-COVID levels, and Q1 guidance is for 28% growth. 16M net new accounts were added in Q4, and they guided for 50M to be added in 2021. Also guided for 2021 (possibly conservatively): high-20s TPV growth, 19% revenue growth, and $6B in FCF.
PayPal also disclosed — to the applause of both its investors and crypto enthusiasts — that half of its crypto users open its app every day, that it will start letting customers pay for transactions with their crypto balances in late Q1 (it first announced plans to support this in October), and that it’s investing in a “new crypto, blockchain and digital currencies business unit” that aims to rethink how financial transactions are done (sounds a little Diem/Libra-like).
The valuation has definitely become rich — PayPal now trades for ~50x its 2021 FCF guidance. But either way, this is a pretty interesting company to follow right now. Much like Amazon, PayPal is set to gain a lot long-term from the impact COVID has had on its transaction volumes, network effects and app/service engagement. And outside of the tech giants, it’s hard to think of a company that’s exposed to so many notable consumer tech trends. E-commerce, marketplaces, content transactions, mobile wallets, NFC, QR codes, crypto, P2P payments…PayPal has its fingers in all of these pies.
(PayPal’s earnings report, earnings slides)
eBay (2/3)— They soundly beat Q4 estimates and issued Q1 sales guidance that topped consensus by even more. GMV growth accelerated to 21% from Q3’s 15%, while revenue rose 28% and (with favorable comps helping out) is forecast to be up 35%-37% in CC in Q1. And U.S. GMV growth (25%) was particularly strong.
But how sustainable is all of this? eBay’s GMV growth was negative pre-COVID, and it’s clear that the company hasn’t just gotten a boost from COVID’s impact on e-commerce in general, but from enormous demand/price jumps for collectibles, vintage goods and various other long-tail items. Some of this demand stems from consumers taking up new hobbies since last March, but some of it is the product of speculative frenzies that can give EV SPACs a run for their money. In hindsight, I should’ve known eBay was due for a big quarter when I took a look last month at what some of the sports cards I once collected are now going for.
Though the 12-year-old version of me won’t like hearing it, I think odds are good that demand/prices for such items will cool off meaningfully as reopenings happen and things like vacations and concerts re-emerge for many people as alternatives to collecting baseball cards and record albums, both for leisure time and disposable-income spend. Meanwhile, the issues that have led eBay to cede so much e-commerce share to Amazon, Etsy and others over the years will remain.
(eBay’s earnings report, earnings slides)
KLA (2/3) — They slightly beat December quarter estimates and guided the March quarter above consensus. More importantly, they forecast their quarterly sales would remain around March quarter levels in 2021 (that puts calendar Q2-Q4 revenue above consensus as well), and they forecast 2021 chip WFE spend would grow by a mid-teens % (plus or minus 2%) from an estimated 2020 level of $59B-$60B. That more or less fits with what Lam Research guided for last week.
Also much like Lam, KLA forecast that they’d see strong demand growth among DRAM and foundry/logic clients, with EUV adoption and other process changes contributing to the former, and modest growth for their NAND equipment sales (not great news for equipment suppliers, but good news for NAND vendors). They also forecast China’s WFE spend would be “flat to up.”
Overall, this was a pretty solid report and call, and it definitely doesn’t put into question expectations that — against a backdrop of strong chip demand, a giant TSMC capex budget and steadily-rising capital-intensity for leading-edge processes — 2021 will be a pretty good year for most chip equipment suppliers. But KLA’s stock is now up ~65% over the last 12 months and trading for 20x its FY22 EPS consensus. Might be better to wait for a pullback, given that KLA does still operate in a cyclical industry (albeit one where secular chip demand and capex trends have begun softening the impact of down-cycles some).
(KLA’s earnings report, shareholder letter)
Nokia/Ericsson (2/4 and 1/29) — A week ago, I noted that Juniper reported seeing weak spending among tier-1 carriers, with the weakness especially pronounced in North America as AT&T/Verizon prioritized spectrum purchases over equipment capex. This week, Nokia also reported seeking weak North American demand, albeit while giving different reasons (price pressure, along with expected share loss as some of its 4G contracts don’t convert into 5G contracts).
Of course, Nokia isn’t seeing a lot of growth in other regions either: The company guided for revenue to drop to €20.6B-€21.8B in 2021 from 2020’s €21.9B, with a “significant decline” for its mobile infrastructure business only partly offset by growth in other businesses. And that follows a 2020 during which revenue fell 6% in euros and 4% in constant currency.
There are some company-specific issues at play here — rival Ericsson saw revenue grow 13% in CC in Q4, with 20% growth for its Networks (hardware/software) segment. But Ericsson also said (citing research from Dell’Oro) that its “planning assumptions” include an expectation for just 3% 2021 growth for the mobile radio access network (RAN) market, and its top-line growth is expected to decelerate to the mid-single digits during the back half of 2021.
Between them, Juniper, Nokia and Ericsson’s reports kind of provide the lay of the land for what it’s like right now to be a major hardware OEM that gets a large portion of its revenue from telecom carriers. If you execute well, you perhaps see moderate growth, and if you don’t, there’s a good chance your sales will drop.
The elephant in the drawing room: Unlike the hyperscalers, the telcos have been seeing little or no revenue growth, and 5G really hasn’t done much to change that. Until this does change (I’m not holding my breath), it’s hard to count on a Nokia, Ericsson or Juniper to see more than modest growth in its carrier-related sales for more than 2 or 3 quarters before a mean-reversion occurs.
(Nokia’s Q4 report, earnings slides) (Ericsson’s Q4 report, earnings slides)
Pinterest (2/4) — This is a perfect storm of sorts: COVID has led tens of millions of people to both spend more time online and take up new hobbies/pursuits (or rediscover old ones) that can be done from home; goods purchases in general and e-commerce activity in particular have spiked; the online ad market is white-hot; and Pinterest’s ad business has reached an inflection point, as investments in product and video ads pay off and Pinterest more generally exploits how its platform lends itself to search, video, e-commerce and top-of-funnel product discovery.
Revenue growth accelerated to 76% from Q3’s 58%, MAU growth clocked in at 37% (same as Q3) and Pinterest, which has been easily topping its guidance in recent quarters, guided for low-70% Q1 growth. The company also disclosed product searches rose ~20x in 2020 and that the number of advertisers using its shopping ads were up 6x Y/Y in Q4. And with Pinterest’s U.S. and especially international quarterly ARPUs still fairly low, there’s a lot of room to grow even if one assumes MAU growth cools as reopenings happen.
With all that said — speaking as someone who was bullish on Pinterest before its IPO and kept the faith as the stock remained stuck in the teens and 20s in 2019 — Pinterest’s current valuation (~20x its 2021 revenue consensus) makes me a little nervous. Particularly since Pinterest did caution (much like Facebook) that it could see second-half growth headwinds both from reopenings (they could potentially impact both engagement and goods/e-commerce purchases) and Apple’s privacy policy changes, even as it remained upbeat about the ongoing payoff from its large product and ad-related investments.
This is still far from the craziest valuation among high-growth consumer Internet plays, especially given Pinterest’s ARPU headroom. But the risk/reward looks different than it did 12 or even 6 months ago.
(Of course, in a market like this one, a good growth story could very well allow high multiplies to inflate even more in the near-term. And Pinterest does have a pretty good story to sell.)
(Pinterest’s earnings report, shareholder letter, earnings slides)
NXP/Microchip (2/1 and 2/4) — Plenty of chip developers have said they have good demand visibility through Q1 and Q2. NXP and Microchip — each of which are seeing demand far outpace supply thanks to both massive automotive orders and strength in various other end-markets — went a step further and effectively forecast the whole of 2021 will be strong. And given industry news flow, it’s easy to see why they feel confident about this.
Along with providing a strong Q1 outlook (no surprise), NXP said it sees “a pretty robust demand environment all through the year,” while arguing that a lot of this growth represents a catch-up following weak demand in 2019 and early 2020. Microchip, whose quarterly guidance was also pretty good, said it doesn’t expect supply to catch up to demand in 2021, and that the imbalance “could possibly go into 2022” in spite of efforts to add more capacity.
That’s not all: CEO Steve Sanghi forecast that “a significant portion” of Microchip’s capacity for 2H21 and 1H22 will be booked through a new program that gives customers priority access to supply if they commit to placing 12 months of orders that are “non-cancellable and non-reschedulable.” And Sanghi made it clear that offering such a program (never mind seeing such uptake for it) is a pretty unprecedented event. Quote:
So the Preferred Supply Program is brand new. I have not ever implemented it in my 42 years of career, and I have never seen anybody else do that, too. The program is largely a response to the current environment because [the] bookings level is just so strong and people are booking parts out in time. The industry seems to be 30%-plus short [of] really what the capacity requirement is. And many of our customers have been asking, what can [Microchip] do if they give us longer-term demand, longer-term orders?
Such supply/demand imbalances make it inevitable that some customers (maybe a lot) will double-order in an attempt to secure capacity, and while Microchip and its peers insist they’re doing their best to prevent double-ordering, figuring out when it’s happening is often easier said than done. It’s also likely that such an ordering environment will spark inventory corrections if/when demand cools, and historically, markets have begun pricing in the end of a chip industry up-cycle before corporate earnings/guidance made the end quite clear. One only has to look back to 2018, when chip stocks began selling off in the spring/summer, for an example of this.
That said, NXP, Microchip and other microcontroller/analog chip suppliers (TI, STMicro, ON Semi, etc.) are providing a lot of reasons right now to think that it’ll still be a while before the tide turns, even if some end-markets cool off during the back half of the year.
(NXP’s earnings report) (Microchip’s earnings report)
Elsewhere in Tech
GM and Ford Go All-In on EVs — Ford now plans to invest $22B in EVs through 2025 (inclusive of $7B already spent), while GM promises to offer 30 EVs by the mid-2020s and end sales of ICE-powered cars by 2035. Major European and Asian are also unveiling big EV investment plans left and right.
As I argued last month, the amount of investment and competition the EV market is poised to see over the next several years — courtesy of everyone from startups to publicly-traded EV makers to incumbent automakers — makes a lot of the valuations being granted to EV pure-plays look especially nuts. On the other hand, it’s clearly good news both for consumers and for automotive chip suppliers with meaningful EV-related exposure, such as TI, NXP, ON Semi and (though its valuation has gotten stretched as well) Cree.
Fresh Reports About Apple’s Car and AR/VR Headset Efforts Emerge— CNBC reported that Apple is nearing a deal with Hyundai-Kia to make Apple-branded electric cars at a Georgia plant. A couple days later, Bloomberg reported that the talks were on pause, with Apple upset at Hyundai for sharing the news. More interestingly, a source told CNBC that the first Apple cars (tentatively due in 2024) will be fully driverless and “focused on the last mile” — a comment that could mean the vehicles will target applications such as goods delivery and driverless taxi services within geofenced areas.
Separately, a couple weeks after Bloomberg reported that Apple is prepping a high-end AR/VR headset that could launch in 2022, The Information reported that the headset will have “more than a dozen cameras for tracking hand movements and showing video of the real world to people wearing it,” as well as dual 8K displays and eye-tracking tech that allows lower-quality graphics to be rendered in the areas that a user’s eyes aren’t focused on. It also noted that Apple “internally discussed pricing the product around $3,000” last year, and reported once more that a pair of lightweight AR glasses are due in 2023.
The upshot from these reports: It doesn’t look as if Apple is aiming for either its first cars or its first AR/VR headset to be high-volume products. Rather, it wants to use the products to showcase its long-term vision for the kinds of experiences it aims for AR/VR headsets and autonomous electric cars to deliver — and by doing so, set the stage for eventually launching higher-volume products that can also deliver such experiences, as AR/VR and autonomous driving tech matures in various ways.
(To reiterate: In the case of the car efforts, I strongly suspect that Apple’s interest in driverless vehicles isn’t just related to the convenience that autonomous cars can provide consumers, but how vehicles that feature no steering wheel or pedals, and which don’t have traditional rear/side-visibility requirements, allow for a dramatic rethink of how the interior of a car can be designed, and of the kinds of computing and entertainment experiences a vehicle can deliver.)
Microsoft Says Xbox Series X Shortages Could Last Until June — For comparison, Microsoft previously said that Series X/S shortages could last until April. Not much to add here, other than that this is another sign that the gaming hardware market remains white-hot during what’s normally its seasonally weakest quarter. Good news not just for Microsoft, but for Sony, Nvidia and AMD, as well as memory suppliers such as Micron and Western Digital.
Microsoft Launches a Platform for Connecting Remote Workers — The platform, known as Viva and set to roll out over the course of 2021, aims to help remote corporate workers stay informed about company events and connected with their colleagues. Maybe the most interesting thing for me here is how tightly Viva is integrated with Teams. Increasingly, a virtuous cycle seems to be forming between Teams adoption benefiting from the platform’s integration with other Microsoft apps/services, and vice versa. Salesforce, which last year launched its Work.com platform for assisting companies with remote work and reopenings, seems to be wagering that it can create a similar virtuous cycle with Slack.
Specs for AMD’s Milan Server CPUs Apparently Leak — The leak, which comes via the Dell Canada website, suggests the most powerful Milan CPU (the 64-core Epyc 7763) will have a base clock that’s 200MHz above that of AMD’s most powerful Rome CPU (the 64-core Epyc 7762). Combine this with the Zen 3 microarchitecture’s IPC gains, and AMD’s flagship Milan CPU could on average be 20%-25% faster than its flagship Rome CPU, albeit while having a higher thermal envelope (280W vs. 225W).
The leak also indicates AMD will launch 32-core Epyc CPUs with 256MB of L3 cache for the first time (previously, only 48 and 64-core CPUs had that much cache). And whereas AMD’s 7F-series CPUs — they feature high clock speeds and target enterprise/HPC servers — currently top out at 24 cores, it looks like the Milan lineup will include a 32-core 7F-series part.
Milan is set to officially launch in March, and based on what’s been disclosed and reported so far, it should on the whole stack up pretty well against Intel’s soon-to-launch Ice Lake Xeon server CPU line.
Disclosure: Long Western Digital and Salesforce
