Simply put, this article is a rebuttal of a recent, bearish Seeking Alpha article about Intel’s (INTC) ‘losses’ – as such, this is a bullish article. Its thesis, as the article went, is that Intel losing the Apple (AAPL) iPhone modem business, as well as Apple’s MacBook CPU business, would swing it from revenue and earnings growth, to decline.
The simplest rebuttal to the Apple items is that they are one-time events. To use an analogy from mathematics: if you have a function like y=72+5x, then this could represent Intel’s 2019 revenue and annual growth rate. Given that the Apple event removes $5 billion in revenue, the new function might look like y=67+5x. What this analogy conveys is that while the base revenue might decrease, the growth (the slope of the function) remains unchanged. In this example, if Intel adds $5 billion every year in revenue, then it will have made up for the Apple loss in one year. (And Apple said the transition would take two years, making it possible that Intel’s revenue will ‘merely’ slow down instead of decline.)
Since long-term value most often comes from (expected) future returns, I argue investors should look at the slope of the curve, not its base value.
About the ‘losses’
While it is undeniably true that these two items, the Apple modem and PC business, will be about a $5 billion headwind, it will be, when all is done, a one-time impact. Intel was not gaining additional modem customers (although 5G modems might have seen an uptick in pricing), and neither would Apple’s MacBook line contribute to additional long-term growth (in case Apple did not make the switch to using repurposed iPhone chips in its macs). To that end, Intel’s Apple business represents $5 billion in nice-to-have revenue and associated profits (and Intel had readily admitted the modem business was below corporate gross margin anyway), but not a core growth driver.
As a thought experiment, let’s say Apple had kept Intel as supplier of both its Mac chips and cellular modems and let’s say this is a $5 billion business. How much would this business generate long-term, let’s say in 2030? I personally don’t see Mac or iPhone units growing a lot over the coming decade. (The only caveat to this is that if Intel, for example, had maintained its process technology lead it had until 14nm, the situation might have been reversed, with the iPhone transitioning to either Intel x86 chips, or more likely, Apple switching from TSMC (TSM) as foundry to Intel Custom Foundry. Either way, in that scenario, Intel might indeed have continued to expand its business with Apple instead of losing it.)
So in that sense, these two headwinds, while the significance of the size of the losses may be up for debate, I would argue they might be best compared to the 2017 McAfee spin-off: a one-time revenue impact. McAfee impacted Intel revenue by $2 billion (comparable to or just below the size of the modem business). It has not stopped Intel from growing at an 8% 3-year CAGR in spite of that headwind.
So I do not foresee the Apple saga as being more than a blip in Intel’s long-term growth story. It will obviously drag down Intel’s growth in the coming two years while the transition unfolds: Intel likely will not provide growth numbers in its earnings where it leaves out the Apple business (“x% growth YoY without the Apple business”), as it provided during the McAfee transition. But for simplicity, it basically comes down to a $5 billion one-time impact, per my mathematics analogy above.
Growth and Gains
Posting $72 billion in revenue in 2019 surely looks better than $67 billion when leaving out the Apple business. (With the McAfee business, it would have been $74 billion.) But as I argued, Apple is a $5 billion business to Intel without growth. It does not impact the areas where Intel’s real growth comes from. So if one accepts that Intel is a growth company, then this should not really change its long-term outlook or valuation. To that end, I already noted Intel’s ~8% 3-year CAGR (in spite of the McAfee spin-off) as an indication that Intel is indeed a growth business.
Taking a look at Intel’s most recent (Q1) earnings further shows this. I will use its latest report to detail where its growth is coming from.
First, the PC group delivered a substantial 14% increase in revenue. Q1 is typically Intel’s weakest quarter, but these solid results were due to several tailwinds. First, the commercial Windows 10 refresh. Secondly, as I wrote at the time, COVID-19 caused an increase in demand for PCs (and data centers), and thirdly, Intel was able to realize some additional revenue as it finally caught up to its 18-month shortage coming out of Q4. This is likely to continue into Q2 to some degree as Intel said it would build up inventory to historical levels. So at least in the short term, gains are amassing in the PC segment.
Secondly, while Intel’s data center business is quite cyclical, the overall trend is up to the right, at low- to mid-teens historically. In Q4, a new growth cycle started, and this continued into Q1: $7.0 billion in revenue represented 43% YoY growth, although down modestly QoQ. The growth drivers of this segment are the cloud and Intel’s networking business. Cloud is benefiting from the accelerating digital transformation, while Intel announced in Q1 it would enter the base station segment of the networking space with the goal of reaching 40% market share with its new 10nm Snow Ridge product in 2021 (supposedly up from negligible market share at present as this marked a new segment it entered). This of course coincides with the adoption of 5G. In its April earnings call, Intel confirmed the demand for this product. The overall networking market is worth over $20 billion today, of which Intel has about $5 billion – I remember Seeking Alpha articles back in 2013 about how networking was one of the major growth opportunities to grow data center revenue.
So by all means, gains in the data center are amassing in the short term, while the long-term growth trends of cloud and 5G obviously also remain intact, including the opportunity to take a lot of market share in networking.
Thirdly, besides these core (no pun intended) and mostly CPU businesses, Intel also has some adjacent businesses. I noted in a previous article that Intel’s adjacencies altogether are already the size of Nvidia, and growing about as fast, even if they are individually quite small.
IoT actually declined modestly in Q1 due to COVID-19, after achieving $1 billion in revenue for the first time in Q3’19. Historically, though, this segment has seen low double-digit returns. Mobileye revenue, also impacted by COVID-19, was up 22%. Intel’s 3D NAND group, Intel’s most cyclical business, saw a strong 46% growth to $1.3 billion. Lastly, Intel’s FPGA group also saw a modest 7% growth to $0.52 billion.
While this covers all of Intel’s segments, there is more to it than Intel reports. For example, Intel said in January it generated $3.8 billion in AI-related revenue, indicating that this was an increase of over 30% compared to 2018. In my first Seeking Alpha publication in January, I noted that Intel’s opportunity in AI is one of its key growth drivers, with a market opportunity that could be worth tens of billions of dollars over the next five to ten years. Further, the adjacency segment of Intel’s Data Center Group (about $2 billion in 2019) has also reported strong double-digit growth lately based on the adoption of silicon photonics and Optane persistent memory.
Lastly, Intel is investing in some big bets that are not yet delivering much revenue today, but which could become Intel’s next $5 billion (“Apple”) business. While Mobileye is the smallest segment Intel reports, it has the largest opportunity. Intel itself talks about a 2030 TAM of tens of billions of dollars (collectively) for ADAS, autonomous vehicle compute silicon, and lidar/radar sensors, and a robotaxi opportunity of well over $100 billion. Additionally, Intel will launch its first discrete GPU product later this year. Intel is investing in a full portfolio to compete against all of Nvidia’s GPUs and GPU segments. Thirdly, I already mentioned the growth of silicon photonics and Optane above: both have only just begun cannibalizing respectively copper-based interconnects in the data center and DRAM (which made Samsung (OTC:SSNLF) temporarily become the world’s largest semiconductor company from 2016-17).
Summing up, as its 23% Q1 revenue growth and ~8% 3-year CAGR show, gains are steadily and surely amassing, while Intel is investing in the next big growth drivers such as Mobileye and discrete GPUs. None of these investments and growth opportunities are impacted by the Apple changes (aside possibly from missing out on selling discrete graphics products for future Macs).
Replacing the Apple business
Interestingly, the McAfee spin-off came at around the same time as Intel’s 2015-2016 acquisition of Altera (for FPGAs). Both generated around $2 billion in revenue, so in that sense the Altera business has basically replaced McAfee (although the acquisition happened before the spin-off).
This presents another tantalizing possibility: could Intel replace the $5 billion Apple business simply by making another big acquisition. One answer might be that Intel has already done so: Mobileye. I discussed in January that Mobileye might quintuple in revenue (to $5 billion) by 2023 already. This is roughly in accordance with the timeline of Apple’s transition from Intel, so in that sense Mobileye might indeed have fully replaced Apple by the time it is selling its last Macs and iPhones with Intel silicon.
As a more concrete suggestion, and relevant in the current times, Intel has talked about some of its efforts and contributions against COVID-19. Perhaps a healthcare-related acquisition such as Livongo Health (LVGO) could be a differentiated addition to Intel’s portfolio. While far from Intel’s core business, it might expand Intel’s IoT business and would probably fit my view of Intel as a data-centric conglomerate. I have also previously suggested that Intel could acquire NIO (NIO). Thirdly, perhaps Intel could acquire a cloud company and start its own cloud business to compete against the likes of Amazon (NASDAQ:AMZN), Microsoft (MSFT) and Google (GOOG) (NASDAQ:GOOG). This isn’t unreasonable, given that Amazon is now doing the reverse (competing with its supplier Intel) by developing its own Graviton chips.
Seeking Alpha recently published another article, coming at it from a slightly different angle, claiming that Arm will lead to further market share losses for Intel. In that case, the disappearing Apple business might have cascading impacts, which could lead to more (indirect) losses that the one-time impact I portrayed the Apple event as.
I have submitted a separate piece to Seeking Alpha with my analysis and take on the Arm vs. x86 debate, so feel free to watch out for that. Its conclusion is that I don’t see an inherent advantage in the Arm architecture, and I see the Apple transition more as in line with the company’s general efforts to bring its silicon engineering in-house and unify its software ecosystem, as it has done with its iPhone SoCs (including the CPU, GPU and in the future its 5G modem). In the PC space, I don’t see such a trend happening with the traditional Windows OEMs, for similar reasons that companies (increasingly) use the solutions of SaaS companies instead of developing their own software. Instead, QUALCOMM (QCOM) will be a competitor to Intel just like AMD. Additionally, the PC isn’t really a growth driver for Intel anyway (aside from its new discrete graphics business).
Secondly, I noted “if one accepts that Intel is a growth company, then this should not really change its long-term outlook or valuation.”
Whilst I tried to prove that Intel is a growth company, the implication of this statement is that an article with the title ‘Intel: Losses Are Mounting,’ if it would then present a case (if it is substantiated) of how AMD is eating into Intel’s PC and data center market share, might present a much more convincing argument against my bullish thesis, as AMD could reduce Intel’s (long-term) growth rate: while the Apple business is simply one that will disappear, if AMD adds $10 billion in revenue, then this is $10 billion in revenue that just as well could have been earned by Intel. So from that angle, Intel needs to develop the world’s best silicon and solutions to compete against competitors such as AMD and others. If it does that, then it can reasonably benefit from the growing market trends such as cloud computing and others, instead of letting others take that revenue. Or as in Apple’s case, convince others that they don’t have to develop their own silicon.
The bearish article I responded to sees revenue declines looming for Intel due to losing Apple as a customer. Here, I presented a different view of these events. While, obviously, this presents a headwind to Intel’s earnings over the next few years, people have already estimated the size of this headwind at about $5 billion. This is a known amount. And while it is a substantial business that will disappear (for example, larger than Intel’s current IoT business even after a decade of building out), it is best to see this as a one-time impact.
It does not change any of Intel’s real growth drivers, whether it is the cloud, networking as it moves to 5G, IoT (which should be able to reach the $2 billion quarter mark faster than the first $1 billion quarter, due to the laws of compounding), 3D NAND, Mobileye or FPGAs. As discussed, these segments are all growing, COVID-19 issues notwithstanding. Or whether it concerns Intel’s upcoming growth drivers such as silicon photonics, Optane persistent memory, Mobileye’s investments in autonomous vehicles, robotaxis and sensors, or discrete GPUs.
So in this view, taking a non-growing piece (whether it is $5 billion or previously the $2 billion McAfee hit) out of a growth company should not concern the long-term investor much.
Intel itself sees a $300 billion TAM it is pursuing and has guided to $85 billion in revenue by 2022-2023. That guidance came in May last year, after the modem exit, and by that time Intel was very likely already aware (even if just via the rumors) of Apple’s Mac developments. Such guidance in spite of these headwinds indicates that Intel management is about as bullish about the business as any investor (and this article).
Disclosure: I am/we are long NIO. I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.