Today let’s take a moment to dig into another teaser pitch that the Motley Fool tends to use over and over, regardless of what the price of the hinted-at stock happens to be in any given week. I’ve been trying to be more diligent in updating my thoughts on these teases, since I know we have new readers all the time who might not have seen the stories I did about these companies years ago… and, to be fair, the companies themselves have changed a lot over the years, as have their share prices, so an updated thought or two might be helpful.

We did this a little while ago with their “All In Buy Alert” pitch for the Trade Desk (TTD), which is all about digital advertising, but this one’s got a different trend to ride on — here’s a little excerpt from the recurring ad:

“Amazon Founder Says His Company ‘Will Go Bankrupt’…

“Jeff Bezos has been investing in a fast-emerging technology that he believes will ‘improve every business.’

“This technology is reliant on a tiny manufactured component… a component that Amazon does not produce in-house.

“Some market researchers believe this technology could potentially be worth up to $19 trillion!”

So what is this business?

“Mark Cuban is on record saying it will create the world’s first trillionaire

“Elon Musk is contributing to a $1 billion investment in this technology

“Even super-investor Warren Buffett says that it’s ‘enormously disruptive’ and will have a ‘hugely beneficial social effect'”

OK, so those are all general references to “artificial intelligence” or AI, which essentially means teaching machines how to spot patterns and learn, without specific instructions from the a programmer. This is a computing-intensive process, as you might imagine, whether you’re talking about the AI systems that help to power Google’s search engines or Amazon’s package routing or the AI systems that make self-driving cars possible.

What do we learn about the specific company? Only one really company specific clue:

“What most people don’t know is… the technology that Bezos is busy betting on is reliant on a tiny manufactured component… a component that Amazon does NOT produce in-house.

“That’s because another company (less than 1/6th the size of Amazon)… is producing a version of this component so powerful that it is absolutely annihilating the competition.”

And that’s pretty much it, just “less than 1/6th the size of Amazon.” And to make matters worse, we know that this ad is re-used over and over for months at a time, seemingly with nobody checking on the current price of the stock in question, so it might not be right around 1/6th today — maybe that was last year. But it does mean it’s a very large company — “less than 1/6th” means it’s probably a lot bigger than, say, 1/10th the size of Amazon, or they’d say that. And Amazon was a $2 trillion company not long ago, and remains a $1 trillion company today, so that gives us a range, I’d narrow it down and say it has to be a company whose market cap was in the $150-400 billion neighborhood over the past several months.

It’s also almost certainly a company the Fool has recommended for a considerable period of time, at least a few years, since we’ve seen almost the same language used over and over, since before the COVID pandemic hit, and they tend to do that with “old faithful” stocks that they know they’ll keep in their active recommendations in perpetuity, not newer and less-proven ideas.

And that “tiny manufactured component” which makes AI possible would almost have to be some sort of semiconductor product. This isn’t a software or platform company if the key is something manufactured.

A little squishy? Sure, but even saying the company has to be above $100 billion shrinks the pie a lot, especially these days. And the Thinkolator, while unable to be certain given the limited clues, will place a hefty bet on the table that this is yet another re-tease of NVIDIA (NVDA) from Motley Fool Stock Advisor, which first teased this company as a hot idea almost exactly eight years ago, when the tease was that NVIDIA was “Warren Buffett’s Worst Nightmare” because they were powering self-driving cars, which would make auto insurance obsolete and put a frown on the face of the GEICO Gecko.

That was far from being true in the near term, as is often the case with such thematic teaser pitches… but what matters is not the promise the newsletter ad uses to get your attention, that’s almost always foofaraw, what matters is the company itself and its operating performance, and whether investors recognize that performance and pay a higher price for the stock as it grows.

And on that front, it has certainly been a great pick over the long term, even after being cut in half from the all-time highs of late last year it would have posted a return today of better than 3,500% if you bought it eight years ago… and I own the stock as well and have a lot of confidence in their future (sadly, I didn’t buy it eight years ago at $5, split adjusted, I paid about $30 for my first shares).

But it is also, one must note, a lot more expensive than it was in 2014, as it has ridden a surge in data center investment and AI progress, as well as continuing huge demand for their high-end gear for gaming, to become one of the more popular stocks in the market. NVIDIA is now the largest US-based semiconductor company, and carries a much higher valuation than semiconductor companies usually enjoy, given the fairly cyclical nature of that business over the decades.

Why is that still justifiable as an investment? To me, it’s their head start in AI software and other high-end graphics systems that makes the difference, including NVIDIA’s support for creating “metaverse” style immersive real-time 3D graphics.

The brilliance of NVIDIA’s Jensen Huang is that he saw the applicability of his company’s GPU designs for doing parallel computing-intensive work beyond just providing high-speed graphics for gamers and design professionals, pushing those GPU designs into use for AI machine learning and autonomous car “thinking”, but what might protect their “moat” in this business is that they were early — they developed the software systems that so many AI projects use and so many developers learned on, and may well do the same for the metaverse, and if they keep those developers happy they’ll probably stay in the leading position, which will mean programs will continue to be developed that run best on NVIDIA’s leading hardware. When most of the innovative companies are building things on your systems, as was arguably the case for the past decade of AI work, then the next wave of students and entrepreneurs also learns on those systems, and it builds on itself.

There is competition, for sure — Intel (INTC) has aims to develop its own GPU chipsets and compete with NVIDIA, and INTC is a LOT cheaper, and unlikely to be growing its earnings much over the next couple years… and Advanced Micro Devices (AMD) has been nipping at NVIDIA’s heels with their own GPUs for years and is, as far as I can tell, very competitive on that front. AMD is also competing with Intel in the CPU and server chip markets, so they have the uneasy situation of competing with the two largest US chip companies, but they’re doing a lot better than I ever imagined they would, and their stock is also cheaper than NVIDIA’s by many measures.

Here’s some of what I wrote to the Irregulars after NVIDIA reported its last quarter, about ten days ago:

Friday File, May 27: NVIDIA (NVDA) reported this week, with again a pretty spectacular quarter of growth, particularly in their sales to data centers — which this quarter became their biggest business, finally surpassing the gaming market that has been their primary driver for a decade or more. They also cut $500 million off of their projected sales for the current quarter because of Russia and the continuing lockdowns in China, so their forecast of $7.94-8.26 billion in revenue in the second quarter came in about $300-400 million short of the analyst estimates.

I don’t know how you could possibly have any certainty over when those challenges might clear up, but I remain quite confident that NVIDIA is doing very well… and it is getting to be more rationally valued these days. Not cheap, we haven’t been able to associate the word “cheap” with NVIDA for many years, but reasonable…

After NVIDIA’s 2021 earnings report, I penciled in $165 as my preferred “buy below” price. We have to look for buying discipline where we can find it, I think, because buying with some kind of discipline is what makes us move relatively slowly, and what makes it easier to have confidence about companies when the share price moves against us during weak times. For me, with NVDA, I’ll again fall back on my PEG ratio to provide a buying range for a company that I think still has substantial long-term potential — I’m willing to bet that NVDA can sustain their earnings growth at 15-20% for the next several years, driven in part by their strength in AI software, so that means I can slot this in as a “growth at a reasonable price” idea if I pay a forward PE ratio of between 30-40.

40X forecasted earnings for this coming year would be about $220 with the current analyst estimates ($5.45 in EPS for FY 2023, which started on Feb. 1), which is much higher than my past NVDA buy targets. 30X forward earnings, a multiple you could probably pay more comfortably for a company as large and strong as NVDA, even if it is at heart a chipmaker and big chipmakers almost never stay at premium valuations like that for very long, would be about $165. If you want to be more aggressive, using the estimates for the following year of $6.52 would put those numbers at $260 and $195, but I’d prefer to be a little more cautious. There’s not much indication that the market is going to go back to the bubble valuations of last Fall and reward that kind of aggression (at the peak back in November, NVDA was valued at about 75X forward earnings estimates, and with a trailing PE of 100 — essentially unheard of for a large cap chip company).

So I’ll continue to slot NVDA in as “worth a nibble” below $220, and as a more tempting buy around $165. I’m certainly not selling, and I didn’t try to time the top and sell last Fall, because I think this is a fantastically run company with a lot of potential, and I think their leading position in AI software is probably under-appreciated as a “moat” for their technology platform and their GPU chips. Still, at current valuations, I’d have to favor Advanced Micro Devices (AMD) by a little bit — it’s not as strong a company, but it’s growing faster, and it’s cheaper. (I own some AMD options, but not the shares, to be clear).

The average multiple for stocks has obviously fallen since I first highlighted that range of 30-40X earnings as the zone where I’m interested in buying NVDA — the forward PE for the S&P 500 has fallen from 23 to about 16 in the past six months, and if everything gets cheaper still, NVDA will very likely follow. The average PE ratio for the S&P 500 bottomed out at about 10 in the last real sustained bear market, in 2008, and analysts could always cut their forecasts for NVDA and other growth-focused companies. If you want to be more conservative and use actual past earnings, then 30X trailing (2021) revenues is $133 now, if you’re holding out hope for a lower price, but do remember that even that is no guaranteed “bottom” if the market resumes its bearishness. A trailing PE of 30X earnings is still a historical anomaly when it comes to semiconductor stocks, the big ones have really only hit those kinds of numbers during pretty manic growth markets. I think it’s a reasonable valuation for buying exceptional companies, and I think NVIDIA is different because of its leadership in AI, but that doesn’t mean this valuation is in any way a “floor” under which NVDA shares can’t fall.

The stock had a negative after-hours reaction to the reduced guidance a couple weeks ago, pretty much a knee-jerk reaction, but by the time the market got going later in the day, with green mostly across the board, it had been forgotten as a one-time challenge and NVIDIA was rising again. So the upshot is that I missed the brief opportunity to add below my $165 level, and would have been happy to add a little more to my position if I had been prepared with an order in place in that brief window, but that’s OK — it’s still pretty reasonable to add little nibbles to a position here if you don’t mind feeling foolish the next time the market has a broader drop, and we may well see the stock fall through that price again if we head into a more prolonged bear market, or if China blockades Taiwan and most of NVIDIA’s chip production gets thrown into limbo (AMD and NVIDIA are both “fabless,” like many chip design companies, both rely heavily on Taiwan Semiconductor (TSM) to do their chip fabrication), so there are plenty of things that could further delay the point when it feels smart to have bought NVDA. I have pretty strong confidence about the next decade for this company, given their continuing leadership, so I can wait.

It remains, however, your money at stake — so you get to make the call. See great times ahead for NVIDIA as Artificial Intelligence processing becomes more ubiquitous? Think it’s too expensive already and destined to fall to the competition? Something inbetween? Let us know with a comment below… and thanks for reading!

P.S. When we’ve covered some previous generic “A.I. is going to be huge” pitches from the Motley Fool Stock Advisor, they have also included recommendations for some of the other giant companies who are working in Artificial Intelligence, typically including Alphabet (GOOG) and formerly-known-as-facebook Meta Platforms (FB, soon to be META). No particular hint about those in this ad, but I’m sure the Fool is still recommending both companies as well, and they are, of course, active in pushing A.I. forward, even if they’re perhaps, because of their diversification and heavy reliance on advertising revenue, not as heavily weighted to that trend as NVIDIA.

Disclosure: Of the companies mentioned above, I own shares of and/or call options on NVIDIA, Advanced Micro Devices, Alphabet and Berkshire Hathaway. I will not trade in any covered stock for at least three days after publication, per Stock Gumshoe’s trading rules.