By Teresa Rivas
Name a company that's gained nearly 4,500% over the past year but trades at only 14 times forward earnings. That may sound like a trick question, but the answer is Sandisk, and it's not alone in sporting an apparent mismatch between its outperformance and its multiple.
When investors think of potential value traps, companies like Nvidia don't often come to mind. Yet relatively modest multiples on big artificial intelligence winners show that even AI enthusiasm has its limits.
Sandisk is an extreme example, but others aren't far behind. Sevens Report President Tom Essaye points to Nvidia, Broadcom, and Micron Technology as well, which have soared 44%, 51% and 770%, respectively, in the past year, but sport multiples of 21 times, 24, and 10 times. All of them are cheaper than the S&P 500, at 21.5 times forward earnings.
The only way to square that circle is to assume the market is discounting future growth; in other words, it doesn't think recent explosive earnings are sustainable.
On the one hand, that could provide investors with some comfort. After all, sky-high valuations for tech companies were a hallmark of the dot-com bubble. On the other hand, low valuations demonstrate that the market isn't fully convinced the AI gravy train will go on forever -- or even as long as bulls hope. Essaye notes that the sustainability worries are still valid, highlighted by Oracle's recent report.
At this point, investors are well aware of how spending from one tech company is directly correlated to other AI players, so if one part of the chain breaks down, that becomes a problem for many other companies (and for investors everywhere, considering the top 10 companies in the S&P 500 are all big tech firms). They've also been informed that trillions of dollars will be spent by hyperscalers like Google parent Alphabet, Facebook parent Meta Platforms, Amazon.com, and Microsoft on the AI build out.
Yet even the deepest pockets can run dry -- hence tech companies raising capital to keep investing in AI. Oracle was one of the first to do this, and its earnings last week revived old concerns.
"Using simple math, it appears that Oracle will have a close to 100% sales/capex ratio in 2027," writes Essaye. "To keep things simple, that means Oracle will spend all of its revenue on capex, the vast majority of which will go into AI infrastructure. That means that Oracle will almost certainly have negative free cash flow and that is only sustainable for so long, even for a company like Oracle."
AI adoption is another potential hiccup for AI stocks as a whole. If the technology isn't adopted as rapidly (or as profitably) as predicted, that could cause companies to cut back. The interconnected nature of the AI economy means the cancellation of a handful of data centers would send ripples up and down the supply chain, from semiconductor producers to memory makers. The market as a whole would take a hit too, given how dependent on tech growth the S&P 500 and its recent gains are.
Here is where the dot-com bubble comes back into play, Essaye notes, because the current AI buildout could wind up being similar to the widespread effort to build fiber access to the internet to homes across the country in the late 1990s. That demand turned out to be unsustainable, as connecting people to the internet wasn't as profitable as initially hoped.
"Practically, I don't think this means anyone needs to reduce tech exposure today," Essaye concludes. "But this situation (i.e., are earnings gains sustainable?) is something that needs to be watched, so ensuring one isn't too overweight tech and has proper balance remains important."
With everything Y2K back in fashion, it's not surprising some investors are on the lookout for a dot com bubble redux.
Write to Teresa Rivas at teresa.rivas@barrons.com
This content was created by Barron's, which is operated by Dow Jones & Co. Barron's is published independently from Dow Jones Newswires and The Wall Street Journal.
(END) Dow Jones Newswires
June 17, 2026 14:25 ET (18:25 GMT)
Copyright (c) 2026 Dow Jones & Company, Inc.
Comments