Certainly all the investment newsletters have been a-titter over the fortunes to be made with 5G… but now Jeff Brown is upping the ante a bit with his teased bet on 6G, and that surely caught the eye of many a Gumshoe reader in recent days. So that’s where we’re targeting the Thinkolator’s work today: What’s this 6G business, and who’s Brown teasing?
Here’s a little lead-in from the ad
“5G has barely hit the market…
“And already there is something more powerful.
“A combination of technologies that takes regular 5G – and turbocharges it.
“In fact, the new technology is so superior to consumer 5G…
“I believe we are looking at what I call an “early version” of the 6G network…
“The next level up in speed, processing, and computing power.
“And right now, the biggest companies in America are paying a king’s ransom to get access.”
He even lists some of the companies who are spending big for access to this “6G” network…
Johnson and Johnson
And gives a couple examples of the huge cost of access…
“Apple is paying $30 million every month to use this early 6G network.
“Facebook forks over $11 million per month for private access.
“And Netflix spent $19 million a month during the pandemic.
“The list of groups paying for 6G access doesn’t stop in the corporate world.
“NASA, the US Department of State, the USDA…”
And he calls this “6G pioneer” the “Most important tech company in the world”…
What’s talking up there, he later reveals, is, of course, Amazon (AMZN), which has soared to dominance on the back of its Amazon Web Services division, born years ago from the brainstorm that they could take the huge and responsive global network they built to support their e-commerce empire and rent it out to all comers. AWS is now the real profit engine for Amazon, though it certainly faces real competition from other cloud computing providers, particularly Microsoft’s Azure but also a host of other smaller players.
So he does recommend Amazon, which is hard to argue with, but that’s not the “teased” investment for his 6G pitch… that’s just the free appetizer. As he puts it…
“… if you want to multiply your money even quicker…
“There’s another, backdoor way to play the rise of Amazon Web Services.Are you getting our free Daily Update
"reveal" emails? If not,
just click here...
“And it could dwarf Amazon’s gains.”
So what is that “backdoor play?”
First I should include his explanation of what “6G” is — he’s not talking about the next generation of wireless technology (it’s still on the drawing board and 5G mass adoption is going to take several years, so that’s good), he’s talking about improvements in processing power and computing speed… here’s a little bit from Brown:
“I call it 6G because when you combine 5G’s speed with an extraordinary set of brand-new technologies…
“You can launch your network into hyper speed.
“Maybe a better way to describe it is 5 ½G or 5G+.
“This is a far more powerful network than what you and I will have….
“When 5G’s latency is combined with what’s called edge computing, it makes the network capable of thousands – even millions – more applications.
“And it gives the network the ability to process exponentially larger amounts of data.
“That’s why I call it 6G.”
So what’s the “backdoor” way to play this “6G” idea?
Here are our clues…
“In order for Amazon Web Services to power all of these cloud computing, artificial intelligence, and machine learning applications, they need to have the most advanced hardware in the world.
“One company makes the key components that run Amazon Web Services.
“Without this technology, as much as 100,000 of the biggest companies in the world would go dark.
“But unlike Amazon or other high-profile Silicon Valley companies, very few Americans know this company’s name.
“Yet they have the most advanced AI and deep learning processors available on the market.”
And it’s not just Amazon…
“Microsoft’s Azure is one part of the overall Office 365 business, which generated $13.3 billion combined last quarter.
“Google Cloud is currently on track to make $10 billion this year.
“And guess who makes the advanced hardware that powers Google’s and Microsoft’s cloud operations?
“The same company critical to Amazon.
“In fact, they make 97% of the pertinent hardware for three of the four largest companies in the world.”
And, interestingly enough, we’re also told that he has pitched this stock before:
“Way back in 2016 – light years ago in tech terms – I pounded the table on this company, then barely bigger than a microcap.
“I told a small, private group of individuals, this company would be the top tech stock of 2016.
“And I was right…
“Shares gained 1,048% in just over two years.
“At the time, they were mostly a gaming company.
“But that was just the beginning…
“This was before the rise of cloud computing and artificial intelligence….
“I think this company is a screaming buy right now.”
So what’s the stock? Thinkolator says we’re again being teased about NVIDIA (NVDA), the pioneering GPU company which Brown has indeed touted before, and whose high-end chipsets designed for better graphics processing have not only revolutionized video gaming, but have turned out to be fantastically well-suited to the huge processing demands of “artificial intelligence” work. Brown did call it the #1 tech stock in 2016 and I think also 2017, as I recall (before shifting to Skyworks last year), and it has been a good ride for quite some time now (for me, too, I must confess, and for many others — NVDA was everyone’s favorite for a few years, and was among the top performing stocks in the S&P for some of that time, though 2018 and 2019 were not as pleasant because the cryptocurrency boom and bust led to a surprise inventory glut).
NVIDIA has recovered nicely from that cryptocurrency glut, when they benefitted hugely from crypto miners using their gaming chipsets for high-power mining rigs but didn’t realize the extent, and were shocked with the inventory crisis they had when cryptocurrencies collapsed — revenues have turned up again in recent quarters, and should continue to rise as new products are introduced. Here’s what the revenue (orange) and the share price (blue) look like over the past five years:
One of the main reasons that I bought and still hold NVIDIA is that they’ve got the inside track when it comes to AI processors — anyone developing AI projects over the past decade or so was almost certainly using NVIDIA’s chips and probably their operating systems, and I’ve always expected that to build on itself with a bit of a network effect from their “first mover” work. Getting the early innovators and experts onto your platform can create a bit of a “moat,” particularly if competitors aren’t building anything that’s dramatically better enough that it forces you to learn a new system (see Intuitive Surgical (ISRG), for example).
I don’t know whether the 97% market share is still a fair thing to claim, but it was reported last year that NVIDIA had 97% of the cloud AI GPU market share, meaning those who use GPUs for AI processing through a big cloud provider like AWS or Google instead of hosting the GPUs on their own servers or in-house.
And NVIDIA is at a pretty key inflection point here — it’s priced at a rich premium again (forward PE is up to 45 or so), and they report earnings in a week and will introduce the next iteration of their GeForce gaming GPUs in a splashy event on September 1 (CEO Jensen Huang channels Steve Jobs a bit whenever you let him onto a stage). Earnings estimates have come up a bit in hte past few months, but certainly the stock price has outpaced those estimate increases (and the stock, at $450, is about 10% above the average analyst target… so the earnings report and any forecast NVIDIA gives will be hugely important to the share price next week).
More on NVDA in a minute, but I do want to first find the third stock Brown is recommending — he teases it thusly:
“Data centers require powerful processors to store all of that data.
“One company is producing the bulk of these processors for AWS.
“In the fourth quarter of 2019, sales were over $2 billion…
“A 49% increase from the year before.
“As 5G and edge computing drastically increase the amount of data being collected and stored, these processors will become even more popular.”
OK, so there’s a third stock — here’s how Brown sums up his ad:
“I believe there are three ways to play the rise of AWS and advanced computing:
1) Buy Amazon
2) Buy the company whose hardware powers AWS technoloeis
3) Buy the company whose processors sever AWS’s rapidly-growing data centers”
And that last one, Thinkolator sez, is NVIDIA’s arch enemy, AMD (AMD). Though it seems that AMD has so far failed to surpass NVDA in the GPU business, which was the initial focus of investors, they have begun to take a bite out of Intel (INTC) in the server processor business and regain their years-ago prominence in the data center.
I confess to having been caught off guard by AMD, it was so terrible and such an also-ran in each of its businesses for so long that their improvement really jumped up and bit me on the butt a little bit… and I still didn’t believe it, having seen their weakness and their near-death years before that. But they are the real deal now, churning out 7nm server chips and taking meaningful market share at a time when Intel has been embarrassed by its own inability to get their 7nm architecture to work.
Here’s what I kept looking back at when I considered AMD in recent years, just to give you an idea of where my perspective was coming from — this is AMD’s share price (blue), revenue (orange) and earnings (red) from 1985 to 2015:
NVIDIA wasn’t public for all of that time, but here’s their history, for comparison’s sake:
And here’s Intel, to whom AMD was hardly even a splinter in the toe for most of that time:
There’s a really interesting piece here from March about AMD’s resurrection, and their ambitions in taking on both Intel in server processors and NVIDIA in GPU/AI processors. And morem recently, just yesterday, there was a good Fool article about the gains AMD is making in AWS with its Epyc server chips.
AMD’s share price is up almost 50% in a month, partly because of their very strong quarterly report a couple weeks ago, so it’s not exactly an easy time to jump aboard when some of the growth stocks are seemingly fizzling a little bit… but it’s a very impressive growth story. Right now, they’re growing revenue at 26%, with a 23% return on equity (ROE), and they’re trading at about 12X sales, with a forward PE of 47 — and with earnings that are expected to triple from 2019 to 2022 as revenue roughly doubles. Expensive, but perhaps worth it.
NVIDIA, by contrast, has a bit stronger growth (39%) recently, a higher ROE of 29% thanks to much higher margins, but they trade at 23X sales — and growth expectations are a little lower, with revenue almost doubling from 2019 to 2023 (they use different fiscal years) and earnings expected to double in that time. They trade at about 44X next year’s earnings, so on that comparison you’d probably want to edge a bit to AMD — they are riskier, and they don’t have that embedded GPU user base in AI that NVDA does so I’d say NVDA still has an advantage, but in the server chipset business AMD is clearly biting big chunks off of Intel’s hide right now, and savoring every bite… and processors are a huge business, far larger than GPUs, so you can imagine some very long-term growth if they can take even more market share from Intel.
Intel (INTC), if you’re curious, is now smaller than NVIDIA ($205 billion market cap to $270 billion) — I confess I never thought that day would come (and AMD is catching up, too, just cresting $100 billion). And if you want to further consider how much the market obsesses about growth, Intel, once again decried as a has-been dinosaur, is cheap again, with revenue growth still coming in at 20% last quarter and ROE at 30%, but they trade at 2.6X sales because everyone thinks the growth is over, forever. Analysts expect three-year revenue growth for Intel to be about 5% (total, not per year), and that earnings will only be about 10% above current levels in 2022. So it shouldn’t be surprising, I guess, that Intel is trading now at a forward PE of 10… though frankly, it’s getting pretty tempting now, if only because it’s hard to imagine Intel really losing as badly as the stock market tells us they’re losing. Intel’s net profit of $24 billion over the past year still makes NVIDIA (at $3.3 billion) and AMD ($609 million) look downright puny.
Kind of like IBM, for that matter, another stodgy old dinosaur whose growth dreams may have gone away, but who still puts together tens of billions of dollars in EBITDA every year, buys back stock, and keeps on chugging along with a strong dividend (now 2.7% for Intel, 5% for IBM… which, unlike Intel, is actually shrinking).
So that’s the sum-up of Brown’s teased ideas, at least — buy Amazon, buy NVIDA, buy AMD. That would have worked out almost unimaginably well over the past few years… but those are certainly all very expensive stocks now, pricing in continued rapid growth and market share gains from AMD and continued dominance from NVDA and Amazon. That might be what happens, and it might even be the highest-probability forecast give the strength of all three, but if you’re someone who likes to buy with a “margin of safety,” well, the market doesn’t give you that in these kinds of stocks right now. The market is telling you that these companies are going to be even-more-powerful market-dominating giants five years from now, and you have to pay a premium price for that anticipated dominance if you want to own a piece today. Fair, perhaps, but that doesn’t make it easy to buy.
The factor that has jumped out as interesting for me lately is Arm Holdings — about two weeks ago the rumors started to firm up that NVIDIA was the leading company in negotiations to acquire UK chip design giant Arm from troubled Softbank (SFTBY), which bought Arm Holdings for $32 billion four years ago. If that happens, NVIDIA suddenly becomes the most feared company in the tech world, as Arm’s basic chip architecture designs are used by everybody from Apple to Qualcomm, but, partly because of that widespread fear of this combination, it’s hard to see a purchase being approved by regulators… even if it’s fun to imagine.
That would give NVIDIA a clear runway into other chip markets outside of the GPU business that they still dominate (though Intel and AMD are fighting them there), including into CPU chips with products that could perhaps take the competition back to Intel and AMD. The UK government might well fight such a move, too, since Arm Holdings is one of their real homegrown tech success stories and the only real UK tech giant… it’s probably more likely that Arm goes back to the public markets in a spinoff or IPO, but you never know. That link above is to a Bloomberg story, the take from the UK’s FT is worth reading as well… I haven’t seen any meaningful updates in the last ten days or so.
Personally, I have owned Amazon and NVIDIA for many years… and I’ve been slowly adding to Amazon because, absent strong regulatory action, it’s hard to see anyone beating Amazon in a meaningful way as they are increasingly woven into all of our lives… but I haven’t bought any NVIDA in a long time (in fact, I stopped out of a third of my posi