Nvidia (NASDAQ:NVDA | NVDA Price Prediction) shares are starting to make a run for those all-time highs again, thanks in part to a reheating of the semiconductor trade and increased enthusiasm for the Vera Rubin era. Undoubtedly, GPUs are continuing to sell, and not much is going to change about that as AI demand stays through the roof, as hyperscalers look to raise the bar, perhaps significantly, on their CapEx going into year-end and into the new year.
There’s a lot of things to worry about, but, all the while, the stock just keeps getting cheaper
Of course, investors have grown less comfortable with the state of hyperscaler ROIs of late, but, at the end of the day, it feels like big tech is going to need to keep spending to improve their chances of winning the AI race, even if it means disappointing many investors and, of course, convincing some of them to hit the sell button.
Add recent chatter about an AI bubble and growing concern about the increased risk appetite with this revolution, and it seems like those who are still up in Nvidia ought to start thinking about taking profits while they’re still ahead, even if it means selling at a price that’s supposedly far cheaper today than at any point in the past couple of years.
Like it or not, GPUs are going to keep selling, and earnings are going to keep marching higher in a way that could make shares of Nvidia look that much cheaper. But just how cheap does Nvidia have to get before the bargain hunters step in?
Given the cyclicality of semiconductors, maybe a price-to-earnings (P/E) multiple in the low-teens or perhaps even a single-digit P/E might not prove low enough for the skeptics insistent on timing the top of the cycle. In any case, Vera Rubin is a serious catalyst that might just allow the good times to keep coming for Jensen Huang and company.
The stage might be set for Nvidia’s next leg higher
With hyperscaler CapEx on an accelerating curve and Elon Musk’s Space Exploration Technologies (NASDAQ:SPCX) poised to go exclusive to Nvidia chips, it feels like the AI heavyweights are willing to go the extra mile to cut to (or at least close to) the front of the line to grab Nvidia’s latest GPUs, which are worlds better than previous generations.
Indeed, when it comes to Vera Rubin, the new era of chips is 10x more efficient. That’s not just another incremental upgrade; that’s a game-changer, but not one that could mark the imminent event for Blackwell chips.
According to Susquehanna, Nvidia is likely to ramp its GB300 line of GPUs. Indeed, there won’t be enough Vera Rubin to go around in the earlier days, so such a chip might be needed to bridge that gap as hyperscalers continue to give the great buildout their all.
With quarterly earnings less than two weeks away and a $500 billion multi-firm AI infrastructure financing consortium in the bag, it feels like going “full speed ahead” is the only way to go. Any way you look at it, it’s Nvidia reigning as king in the early days of this AI revolution. And my guess is the reign stands to last a lot longer than skeptics think.
The bottom line
Indeed, an eventual downfall may be inevitable, given the cyclical risks that, in my view, have been amplified by circular deal-making. More deal-making, more financing, and more loyalty might just be key to amplifying the next leg higher, though. And that makes Nvidia stock a really tough name to sell short. As the valuation multiples compress further in the looming Vera Rubin boom, I think going long beats going short.
Personally, I want to see where hyperscaler CapEx goes in 2027. Of course, by then, it might be too late to buy Nvidia stock at below $225 per share.
Contact [email protected] for any questions or corrections.