Why I’m Staying Cautious on NVIDIA—and the Stocks I Prefer Instead

Nvidia (NASDAQ:NVDA | NVDA Price Prediction) is back in a bear market, and with the $170 support level breaking down, the next stop might lie closer to $140 per share. Undoubtedly, that’d represent another 15% dip from current levels. And while shares have gone from cheap to cheaper in recent weeks, with shares now going for 33.6 times trailing price-to-earnings (P/E), there seem to be lingering doubts despite that incredible quarterly number, the $1 trillion GPU sales target through next year, the upbeat commentary delivered by Jensen Huang, and the strong showing at CES and GTC 2026.

Of course, not everything (apart from the stock, of course) has been rosy for Nvidia this year.

Nvidia’s DLSS 5 draws backlash

At the same time, though, there’s been quite a lot of harsh criticisms of DLSS 5, which some view as nothing more than an “AI slop filter” on video games.

While Jensen Huang has been quick to defend his position, saying that the technology provides generative control at the geometry level, it doesn’t seem like gamers, as a whole, are at all too impressed by the demo, either because it adds uncanny realism that “nobody asked for,” or due to the potential for the technology to “insult the art and the artist,” as a poster on Reddit put it.

You don’t have to look too far for some brutal mockery of the technology, especially on Reddit. Either way, it hasn’t been completely smooth sailing for Nvidia. And while the “honeymoon phase” for applied AI tech might be hitting a bit of a ceiling, I do think that the numbers will continue doing the talking for Nvidia.

The $1 trillion+ revenue opportunity through 2026 is a massive deal, but one that the market might be pricing in over fear of what happens in 2028, 2029, and so on.

All the while, there are more combatants entering the AI chips race with a specialty in inference. Risks remain, even if there’s no AI bubble in sight, but the biggest unknown, I think, is what happens if investors are so fatigued that nothing innovative that Nvidia delivers is able to drive up the stock anymore.

Could 20.0 times forward P/E really be the new fair value for Nvidia stock in this climate? 

I’m not sure. Either way, I think there are opportunities worth exploring that go beyond the “picks and shovels.” Arguably, the AI platform companies that are taking control of their own AI stacks, from silicon to software and everything in between, might be timelier bets as markets move on after the latest vicious valuation reset.

There’s no denying that markets were expensive going into the year, but after a correction, the broad market is looking pretty fairly priced, perhaps even undervalued if you expect AI narratives to deliver this year.

Meta Platforms: It’s a “top pick” and a deep-value play that’s ripe for buying

If they do, a name like Meta Platforms (NASDAQ:META), perhaps the ultimate AI-driven platform play, could stand tall as AI moves the needle further on engagement metrics and ad monetization. Of course, a pair of legal setbacks and ongoing chatter about the firm spending too much may cloud the many big positives going on behind the scenes. Perhaps most remarkable among the efforts is the MTIA silicon roadmap, as well as the latest partnership in place with Arm Holdings (NASDAQ:ARM) to use its AGI CPU.

Add the recent “top pick” rating over at Morgan Stanley into the equation, and it’s not a mystery as to why Meta stock may yet have positive potential from AI priced in. At this juncture, AI and the CapEx it entails seem to be a negative, even if it goes against the positives seen in that last quarter.

Morgan Stanley thinks the core of digital advertising is in great shape, and AI is just making the flywheel spin stronger. If the hyperscalers and the sell-side analysts are comfortable with the CapEx levels, perhaps so, too, should investors. Either way, Meta seems poised to go above and beyond GPUs and Nvidia over the long run as the platform titan tightens its grip on hardware. 

The bottom line

As cheap as Nvidia stock is right here at around 20.0 times price-to-earnings (P/E), Meta appears that much cheaper, currently going for 17.6 times forward P/E. As monetization and margin enhancement become the new needle movers, I think Meta might be better primed for performance. As for Nvidia, its margins and monetization are already there. And its relatively muted valuation, I think, suggests there’s no more room for growth or margins to keep moving higher.