Summary
- On Monday, I checked my Broadcom's stock price during a market rally and it was +$20/share. I was shocked later in the day when it was down $10.
- I searched for the reason and found it was - as suspected - due to a report that Apple intends to design some Broadcom components out of its products.
- Today, I will give an historic perspective and the history between two of the arguably rockstar tech CEOs of our time: AVGO's Hock Tan and Apple's Tim Cook.
Yes, I must admit I was very happy to have checked the price of my Broadcom ( AVGO ) stock earlier this week during a market rally and find it up $20/share. Yes! I thought ... finally we're going to start seeing a rebound in some of these overly-beaten down semiconductor stocks. You can imagine my shock later in the day when the market maintained its rally yet Broadcom closed down $10/share (see graphic below). I suspected what the issue might be and started searching for the news release that was - at least in my opinion - way overdue. A few minutes later I found it: A report by Bloomberg that Apple ( AAPL ) planned to replace some Broadcom components with in-house designs. The stock's reaction was quick and violent. After all, Broadcom had 429 million shares outstanding as of the end of Q4, so a $30 intra-day swing in the stock price equates to an estimated $12.9 billion change in market cap. On some level, I suppose that might have been a rational knee-jerk reaction by investors. I say that because - after all - Apple is Broadcom's largest customer - accounting for an estimated ~20% of total revenue last year - or an estimated ~$7 billion. However, today I will take a closer look at this issue and explain why, I believe, the stock's downward move was an overreaction. Indeed, note that the stock chart below shows that Broadcom has already recovered ~50% of the drop:
The History
For those of you who aren't familiar with the history between Broadcom CEO Hock Tan and Apple CEO Tim Cook, it hit the boiling point back in late 2019/early 2020. Both CEOs are well known for being tough negotiators and both are well aware that shareholder returns are based not on sales, but on margin. As a result, in late 2019 the two CEOs locked horns during contract price negotiations for Broadcom's wireless components. Apparently they had reached an impasse. Hock Tan called Tim Cook's bluff by making statements on the December 2019 Q4 conference call that he now viewed wireless as a "noncore" business and, as such, it would be broken out from the rest of the semiconductor segment.
I remember the conference call, which I attended, because wireless was one of the reasons I originally invested in Broadcom (among many others ...) and this news shocked me.
Before we go much further here, note that Broadcom separates its wireless business into three categories:
- RF
- WiFi / Bluetooth (the "combo" chip)
- Mixed-signal custom ("exclusive to one large wireless customer")
Keep that in mind. One of the statements Hock Tan made on that conference call, for example, was as follows:
RF, which represented approximately $2.2 billion of revenues in fiscal 2019 is expected to grow high single digits given the initial ramp in 5G phones. WiFi/Bluetooth combos, which was approximately $2.2 billion in fiscal 2019, is expected to be down low single digits. The adoption of new WiFi 6 solutions at our 2 large smartphone customers will be offset by the completion of our movement away from noncore, lower-margin legacy WiFi business, which will adversely impact this product line's 2022 revenues.
Indeed, Electronic Design later reported that:
Broadcom has been reportedly readying to sell its radio frequency business, one of the pillars of its wireless chip unit. Last year, the company warned that it would separate its wireless chip business—which rolls out Bluetooth, WiFi, GPS, RF and wireless charging components used in smartphones sold by Apple and others—from its core networking business, which sells processors used in network switches and routers in data centers.
Mr. Tan also inferred that Broadcom might simply exit the wireless business because margin was so poor and it had become "non-core." That certainly got Tim Cook's attention - after all, Broadcom designs were entrenched in Apple iPhones and have a great reputation of delivering high-quality parts that work. The next thing I know, and very quietly (as per almost everything Apple related...) Broadcom shareholders discovered that the fence had been mended between the two CEOs as Hock Tan issued a simple statement on March 12, 2020:
After careful consideration, we have come to the conclusion that continuing to invest in and operate our wireless asset will create the most value for our business and for our shareholders.
So, Hock Tan won that battle with Tim Cook and the two companies signed a new two-year deal (which expires this year). That's because Cook knew that Broadcom had the best products and that there was a high level of risk in replacing wireless components in multiple iPhone devices that had to work (and get tested and qualified ...) across a variety of networks all across the world.
That said, it was clear when Apple began designing its own silicon for the Mac, for instance the M2 chip shown below, that it was highly likely Apple would not stop there:
Apple has been very successful in this regard with the Mac. And now, 9to5 Mac reports that:
To ditch Broadcom, Apple is reportedly developing a combo Wi-Fi and Bluetooth chip that could debut as soon as 2025. It’s also working on a follow up version of this chip that combines the cellular modem, Wi-Fi, and Bluetooth tech all into a single chip.
However, while some investors might think that designing a massive multi-core high-performance chip like the M2 should be much harder than designing some rather "simple" wireless chips, I would say "think again."
Investors should consider that designing a mostly digital high-performance processing chip like the M2 is actually not that difficult. There are established software and validation suites, and the core designs can be optimized and then replicated across the multi-core chip. Sure, there were performance challenges, bugs, and no doubt the chip was iterated - likely more than once (I don't know this for a fact, but as a design engineer myself, it is more likely than not given the level of complexity). But at the end of the day, digital design is simply easier than analog - which can sometimes be as much of an art as it is a science.
Investors also should consider that Apple started with replacing the processing chips and is just now getting around to the wireless designs (note that the announcement said they would start the Apple chips in 2024 and 2025). One reason might be that the M1/M2 processing chips were the biggest bang for the buck. The other perspective is that they had a separate team working on the wireless designs and they're taking longer.
One reason is that wireless designs can be "spooky." I say that because any small lump of capacitance or an etch whose inductance turns out to be greater (or less) than expected can wreak havoc on analog-based designs. High speed signal reflections... frequency rejection... potential oscillation problems... there are a host of issues in the analog world that can affect component integration and functionality. If the new parts are not pin-for-pin compatible with the old parts, a new printed circuit board ("PCB") design is required. In the high speed analog world, the PCB is actually viewed as a component - wherein even the capacitance between the power and ground planes matters, not to mention that the length and impedance of the individual signal traces matter as well.
In addition to these hardware challenges, these wireless chips typically have a ton of related software and firmware that must be maintained and tested - again, across multiple iPhone devices running on multiple networks around the world. As my followers know, I have reported that one of Broadcom's big strengths is that the company has, in my opinion, the best high-speed networking development platform on the planet (and that includes both hardware and software). This is why Broadcom is consistently one step (and at times, two steps ...) ahead of the competition.
My point is, replacing wireless chips - which by their nature are heavily analog-based designs - is risky and not as easy as it seems. Yet Tim Cook really has no choice... he knows revenue growth is going to be more difficult going forward, so to keep shareholders happy with high-returns, he needs to increase margin by reducing costs. He wants Apple to control the most important semiconductors in the iPhone - and who can blame him, I would have the exact same strategy. Note that the announcement referenced earlier also reported that Apple is also designing its own cellular-modem chip to replace Qualcomm ( QCOM ) parts. This all makes great sense to me. But it does come with associated risks.
Indeed, 9to5Mac agrees with me on this point:
All of this is a major undertaking for Apple ... These are the types of changes that require years of testing in the real world, so it will be interesting to see how Apple rolls everything out.
To design out Broadcom is likely one big reason why, back in 2019, Cook decided to spend $1 billion to buy Intel's modem business . However, can Intel's design engineers - and the other engineer Apple has hired in this space since - really stand-up to Broadcom's design engineers and the Broadcom hardware/software development platform? That's a valid question - after all, there is a reason that Broadcom won the wireless slots it owns today in the high-end Apple iPhones, right? The reason is that AVGO delivers high-quality components that work. And it has been on the forefront of networking designs (again, HW and SW) for many years now.
Let's Say Apple Is Successful
Yet Broadcom investors should assume that Apple will be successful in its effort to design out AVGO parts. That said, I'm not totally convinced this isn't merely a negotiating ploy on Tim Cook's part. But let's just assume Apple pulls it off: What is the impact? First of all, Hock Tan is arguably one of the best technology CEOs in the business. That being the case, Hock has likely been planning for Apple to design its own parts for years, or at least since the last contract negotiations two years ago.
Secondly, as most Broadcom investors already know, for at least the past two years Broadcom had sold out its entire semiconductor capacity before the year even started. In addition, Broadcom implemented a "no cancellation" policy considering - you order it, you take delivery and pay. Period. After all, AVGO has customers waiting in line and therefore if a customer cancels without payment, that would be taking bread from the mouths of shareholders. So, the first consideration is that Broadcom has very strong demand across a wide variety of semiconductor solutions and Hock Tan has proven he (and the team) are masters at managing the backlog and customer base (he calls the process "scrubbing" the backlog).
Just to reiterate that point, note that on the Q4 conference call , Hock Tan reported:
In summary, we’re guiding consolidated Q1 revenue of $8.9 billion, up 16% year-on-year. While we're fully booked for fiscal 2023, in this environment, we're not providing you guidance for the year.
So that makes at least three years in a row that Broadcom has been fully booked at the start of its fiscal year. Going forward, one bright spot that may be able to substantially fill capacity currently used for Apple might be the hyperscaler cloud sector, where companies like Amazon (AMZN), Google (GOOG) (GOOGL), and Microsoft (MSFT) continue to expand. And of course Broadcom has owned the high-end high-speed networking space for years.
Thirdly, most Broadcom investors also know that the company has been diversifying into software for years now. Indeed, in the Q4 report last month, the Infrastructure Software Segment revenue was $1.84 billion, 21% of total revenue. The point is, Broadcom's high-margin software business has nicely diversified Broadcom away from its semiconductor solutions business. (And this is before any consideration of the pending VMware acquisition).
However, Bernstein analyst Stacy Rasgon estimates that Broadcom would lose in the neighborhood of 25% of its Apple business - or an estimated $1.50 to $2 a share in EPS in 2025 should Apple proceed as reported (that estimate appears to be based on replacement of the WiFi/Bluetooth combo chip only). However, note that the estimated impact is only a 3%-4% of Rasgon's 2025 EPS estimate of $48.11 a share.
That analysis goes on to point out that of the 21 analysts who cover Broadcom, six have an average 2025 EPS estimate of $45.42 a share while 19 have buy-grade ratings. At Broadcom's current price ($578), the $45.42 EPS estimate implies a two-year go-forward P/E of only 12.7x.
That's arguably a low valuation for a company that generated a whopping $16.3 billion of FCF in FY22 (that equates to an estimated $38/share of FCF). Note that AVGO remains one of the best (if not the best) dividend growth stocks in 2022, with a 12% increase in the quarterly dividend to $18.40/share on an annual basis. That's good enough for a current 3.2% yield.
Indeed, since 2016, AVGO has grown its dividend at a CAGR of 38%:
If you know of a better dividend growth stock in the entire S&P 500, please let me know.
Meantime, Broadcom's current forward P/E (per Seeking Alpha) is only 14.1x. Again, that's a pretty low valuation level for a company that grew revenue by 21% last year and ended FY22 with $12.4 billion in cash (or an estimated 28.90/share in cash). AVGO's FCF profile and strong cash position obviously bodes well for excellent dividend growth going forward.
VMWare
I view the VMware ( VMW ) acquisition as a win either way it turns out. If the acquisition is turned down by regulators, that means AVGO doesn't have to stress its balance sheet and will - as a result - likely return even more FCF directly to shareholders via dividend increases and accelerated share buybacks.
If the VMW transaction does goes through, Hock Tan's brilliant and very successful track-record of integrating acquisitions and making them much more efficient and value added will likely continue. I say that because VMware's SaaS-based data center virtualization software fits in very nicely with Broadcom's existing infrastructure software business. That being the case, VMW's offerings can be leveraged over AVGO's existing customer base. Longer term, VMware is ideally positioned to benefit from the trend of hybrid cloud-computing and its software to achieve commonality between its customers' private and public cloud domains will likely become quite popular and successful.
That being the case, I'm pretty ambivalent about the VMware acquisition, although in the long term - given Hock Tan's successful integration track record - I think the deal would likely be quite beneficial for AVGO shareholders should it move forward.
Summary and Conclusions
The bottom line here is that the stock market reaction that whipped out an estimated $12.9 billion in Broadcom market cap was arguably an overreaction. As pointed out earlier, cooler heads have prevailed and AVGO stock has already recovered ~50% of the drop.
Meantime, I'm not convinced that the reports aren't simply a ploy by Apple CEO Tim Cook to position himself and his company prior to contract negotiations with Hock Tan. It's a CEO shakedown over margin, and Hock Tan won the last round. It will be very interesting to see if Tim Cook and Apple actually takes on the risks and pulls-off their wireless part replacement strategy and makes the switch, or, if Hock Tan concedes a bit of margin to Apple in an attempt to keep the peace and hold onto the business for another two years. At this point, no one knows, so stay tuned.
However, either way it turns out, and similar to my ambivalence about the pending VMware acquisition, Broadcom is so well positioned it is going to be just fine either way. Indeed, and next to Google, AVGO is one of my favorite companies in the beaten down technology sector - and that remains the case even after the Apple "switch" reports.
AVGO didn't miss a beat in 2022, and it won't miss a beat in 2023 - recession or no recession. That's because Broadcom is fully booked for FY23 and it has an iron-clad no cancellation policy. It has the best high-speed networking development platform on the planet. And you can bet that Tim Cook knows it. In addition, AVGO generates tons of free cash flow and is one of the best dividend growth stocks in the entire S&P 500 (likely the best). And despite all that, Broadcom trades with a forward P/E of only 14x and yields 3.2%. Put it altogether and Broadcom is a BUY here.
I'll end with a five-year chart of Broadcom's total returns vs. Apple, the Nasdaq-100 ( QQQ ) and the S&P 500 as represented by the ( VOO ) ETF and make the observation that the collaboration between Apple and Broadcom, and between Hock Tan and Tim Cook, two of the arguably best technology company CEOs, has - obviously - led to tremendous success for both companies:
Clearly, Broadcom is in a class of its own.
For further details see:
Broadcom: CEO Shakedown - Hock Tan Vs. Tim Cook