How the Five Best Tech Stocks of the Year Made It to the Top—and Why You Should Care
It’s that time again.
Investors are beginning to worry about over-the-top valuations, especially in tech. The usual gloomsters are cheerfully reminding us that it’s never “different this time” and reversion to the mean is not less assured than death and taxes.
Okay, we buy the “it’s never different this time” argument. And some valuations have us worried. And yes, we are likely heading into a (mostly) “buy-and-sell” rather than a “buy-and-hold” market.
In a flattish market, however, the details become more important than ever. If a tech stock bears an extraordinary valuation, before we put it on sale we want to know why.
Sometimes extraordinary valuations are justified, or justified enough to sustain buy-and-hold even through some expected volatility.
Take a quick look at the five best-performing large-cap semiconductor stocks of 2024, and let’s try to figure whether their current valuations make sense.
The top five large-cap (>$50 billion) companies by performance in 2024 are:
- NVIDIA (NASDAQ: NVDA), +171%
- Broadcom (NASDAQ: AVGO), +108%
- Taiwan Semiconductor Manufacturing (NYSE: TSM), +90%
- Marvell (NASDAQ: MRVL), +83%
- SMIC (0981.HK), +61%
All five have one big thing in common. Every single one has benefited massively from the boom in artificial intelligence (AI). Will the boom last? Will the stocks?
NVIDIA (NVDA)
Nvidia stands alone both in reputation and reality. Four companies on this list were carried by the AI boom; Nvidia carried them.
Nvidia did not invent AI—that’s been a process of decades—but its GPUs (graphic processing units) made AI practical and economically useful. It commands some 92% of the AI GPU market today and has no serious competitors in that market. The impact of its invention is reflected in this five-year summary of its growth.
Not only have revenues grown by almost 10 times, the share of revenues that hit the bottom line has grown from 25% to 55%. As a result, profits have grown by more than 17 times in five years. With an extraordinary price-earnings ratio of 50 and a price-to-sales ratio of 27, NVDA is by all normal standards an incredibly expensive stock. The only way to justify its current valuation is to assume continued extraordinary growth in revenue and profits for at least the next several years.
We think that’s likely. We said above that NVIDIA has no serious competitors in the GPU market. We expect that to continue. The company is more than a decade ahead of the rest of the industry on the GPU learning curve.
Nvidia’s competition will come not from other GPU makers but from companies trying to sidestep the high costs of using elite GPUs for artificial intelligence. Lots of companies are trying to do this. Google (NASDAQ: GOOG), Meta (NASDAQ: META) Amazon (NASDAQ: AMZN) and Microsoft (NASDAQ: MSFT), all big NVIDIA customers, are trying to cut their Nvidia bill of materials by swapping in their own custom silicon. Broadcom, another big 2024 winner, is making a powerful strategic shift in this direction. And the devices these NVDA customers are making are excellent.
But guess what? Or rather guess “who” is doing its own “custom silicon” to save money on GPUs. Nvidia. Jensen Huang, a very wise man, has decided that Nvidia should compete with itself and, with good fortune, beat itself.
We said above the only justification for Nvidia’s price is to assume that its astonishing growth will continue. But what justifies that assumption? The best companies in the world experience volatility in revenues, profits, and especially their stock price. NVDA will inevitably hit nausea-inducing air pockets. Why not sell now in anticipation of the inevitable?
Some companies, though not many in a generation, change the world more than the world changes them. Nvidia is such a company.
We’re long past the point where NVDA’s growth can be ascribed to a fortunate alignment of circumstances. As one of a handful of the most well run and innovative companies in the world—arguably the best—we invest in Nvidia not to share in the GPU market but to benefit from a company and its leadership.
When a company like this comes along—a company that creates a whole new market with its technology—its reign tends to last not years but decades.
That was true of Intel (NASDAQ: INTC), which invented the microprocessor in 1971, the first practical DRAM in 1970 and dominated the markets it created for some 40 years.
From the time Qualcomm (NASDAQ: QCOM) introduced CDMA (code division multiple access), transforming mobile communications and all but inventing mobile data, to the present moment, there have only been a few weeks when buying the stock was not a good 10-year investment. More to the point, QCOM has maintained its ascendancy through several radical innovations in its own technology. QCOM investors, it turns out, almost 30 years later, were not buying CDMA, they were buying Qualcomm…
Want to read the rest?
No problem! Subscribe to the Gilder Technology Report and get the very next issue wherein we tell the rest of the tale.