A quiet portfolio shuffle by Billionaire Chris Rokos has become a loud signal in the artificial intelligence trade. By unloading Palantir and loading up on Nvidia, he effectively swapped a richly valued software story for the chipmaker at the center of the AI hardware boom, and the valuation gap between the two suggests his timing could hardly be better. With other heavyweight investors rethinking their exposure to both names, the move captures how fast the market’s AI favorites are being repriced.
At the same time, the shift highlights a broader divide between companies selling AI infrastructure and those monetizing AI through data and analytics platforms. As I see it, the Rokos trade is not just a stock pick, it is a bet on which layer of the AI stack will deliver the best risk adjusted returns in 2026.
Inside Chris Rokos’s big switch from Palantir to Nvidia
Billionaire Chris Rokos has built a reputation as a macro trader who is willing to make concentrated, high conviction bets, and his latest move fits that pattern. In the most recent quarter he exited Palantir and bought Nvidia, effectively rotating capital from a controversial data analytics name into the company that designs the chips powering the current wave of AI training and inference. The shift came as other sophisticated investors, including Michael Burry, were signaling caution on Palantir, with Burry holding more bearish exposure to the stock than to Nvidia through put options, a stance that underscored how stretched Palantir’s story had become relative to its fundamentals, according to a Quick Read on the trade.
Rokos’s pivot was not a minor rebalance but a clear expression of preference for Nvidia’s side of the AI trade. Reporting on his portfolio shows that Billionaire Chris Rokos buys Nvidia and sells Palantir in the same window, a pairing that makes the comparison between the two companies unavoidable and highlights his view that the chipmaker offers a better entry point than the software group at current prices. That decision came as investors were digesting the next generation of Nvidia’s roadmap, including Rubin Ultra and Feynman, and as expectations for AI infrastructure spending in 2026 were being reset higher, a backdrop that helps explain why Rokos chose to lean into Nvidia’s growth engine while stepping away from Palantir’s premium valuation, as detailed in coverage of how Billionaire Chris Rokos repositioned his book.
Palantir’s sky high valuation versus Nvidia’s relative discount
The core of the Rokos thesis, as I read it, is that Palantir’s share price already bakes in a best case scenario while Nvidia’s still leaves room for upside if AI demand continues to surprise. Palantir has long traded more like a narrative driven growth story than a mature software vendor, and that gap has only widened as investors have chased anything tied to AI. One analysis of the trade notes that Nvidia certainly looks far cheaper than Palantir when you line up their earnings power and growth trajectories, arguing that the chip designer’s multiple is more defensible given its dominant position in AI accelerators and the visibility of data center demand, while Palantir’s valuation remains significantly higher relative to its current profitability and contract base, a comparison that underpins the view that Nvidia certainly looks far cheaper.
Palantir’s own trading statistics reinforce that sense of froth. According to PLTR Key Statistics, the stock is associated with a Market cap figure of 400.04B, a Price Earnings ratio of 415.69, a Dividend yield marked as a dash, and shares outstanding of 35.11M, numbers that include the specific metrics 400, 415.69, 69 and 35 and that collectively paint a picture of a company priced for perfection rather than steady compounding. By contrast, Nvidia’s valuation, while hardly cheap in absolute terms, is supported by explosive revenue growth from AI chips and a product roadmap that includes Rubin Ultra and Feynman, which investors expect to extend its lead in high performance computing, a dynamic that helps explain why one detailed breakdown concluded that Billionaire Chris Rokos was making a brilliant move by swapping into the chipmaker at this stage of the cycle.
How other billionaire investors are repositioning around AI
Rokos is not the only billionaire reassessing how to play AI, and that context matters for anyone trying to gauge whether his Nvidia over Palantir call is an outlier or part of a broader pattern. Another high profile figure, described as a Billionaire With 30 Years of 30% CAGR Gains, recently dumped Nvidia, Palantir and Eli Lilly, a sweeping reduction in exposure to some of the market’s most crowded winners that suggests even long term outperformers are wary of how much optimism is already embedded in AI and adjacent trades. The fact that an investor with Years of compounding at a 30% CAGR was willing to join the list of Gains Dumped Nvidia and Palantir underscores how polarizing these names have become among professionals, as detailed in a breakdown of how a Billionaire With 30 Years of 30% CAGR Gains Dumped Nvidia, Palantir and other leaders.
Then there is Billionaire Stanley Druckenmiller, another closely watched money manager whose trades often ripple across Wall Street. Recent disclosures show that Billionaire Stanley Druckenmiller Sold Nvidia and Palantir and Piled Into One of Wall Street’s Hottest Drug Stocks Ahe, a dramatic pivot that swapped AI exposure for a high momentum pharmaceutical name. Despite acknowledging the sustainable moats around Nvidia’s chip franchise and Palantir’s government and commercial data platforms, Druckenmiller chose to harvest gains and redeploy capital into a different pocket of the market, a reminder that even investors who believe in the long term AI story are sensitive to entry price and near term risk, as captured in coverage of how Billionaire Stanley Druckenmiller Sold Nvidia and Palantir and Piled Into One of Wall Street’s Hottest Drug Stocks Ahe.
Why Nvidia looks like the stronger AI bet for 2026
Against that backdrop of profit taking and rotation, I see the Rokos move as a more nuanced expression of AI conviction rather than a simple risk off trade. Instead of abandoning the theme, he is concentrating it in the company that many analysts view as the linchpin of AI infrastructure, and recent comparative work backs up that preference. One detailed analysis argues that Nvidia and Palantir have similar growth stories in the sense that both are tied to the expansion of AI workloads, but concludes that Nvidia’s valuation is much more attractive than Palantir’s and that, overall, the chipmaker offers a better risk reward proposition than its software counterpart for 2026, a judgment that aligns closely with Rokos’s decision to favor Nvidia and Palantir in that order.
Part of the appeal is that Nvidia sits at the intersection of several powerful trends at once. Cloud providers like Amazon Web Services and Microsoft Azure are racing to expand AI capacity, enterprise software vendors from Salesforce to ServiceNow are embedding generative models into their products, and consumer applications such as ChatGPT and Midjourney are driving demand for ever more powerful accelerators, all of which funnel spending toward Nvidia’s GPUs and networking gear. Palantir, by contrast, depends on winning and expanding specific government and commercial contracts for its platforms, a slower and lumpier process that can justify a premium multiple only up to a point. When I weigh those dynamics, I see why a billionaire investor looking out over the next year might decide that Nvidia’s combination of scale, product roadmap and still reasonable relative valuation makes it the cleaner way to stay long AI while sidestepping some of the exuberance embedded in Palantir.
What individual investors should watch next
For everyday investors, the lesson is not to blindly copy any billionaire’s trades but to understand the logic behind them and the risks they are managing. Rokos’s decision to ditch Palantir in favor of Nvidia highlights how important it is to scrutinize valuation, business quality and competitive positioning, even in sectors with powerful secular tailwinds like AI. It also shows that sophisticated players are willing to differentiate between layers of the AI stack, favoring the picks and shovels providers of compute over software names whose pricing already assumes flawless execution. To track how those judgments evolve, I find it useful to monitor real time market data and fundamentals through tools that aggregate quotes, charts and financials, while keeping in mind the standard disclosures that services such as Google Finance provide about the limitations and delays in their feeds.
Ultimately, the fact that investors as different as Billionaire Chris Rokos, Michael Burry, the Billionaire With 30 Years of 30% CAGR Gains and Billionaire Stanley Druckenmiller are all actively rebalancing their Nvidia and Palantir exposure tells me that the easy money phase of the AI trade is over. From here, returns are likely to depend less on owning anything with an AI label and more on choosing the right companies at the right prices, and on being willing to change course when the numbers no longer justify the narrative. In that environment, the Rokos pivot from Palantir to Nvidia looks less like a one off headline and more like an early example of how the smartest money is trying to stay ahead of the next phase of the AI cycle.
More From TheDailyOverview
Grant Mercer covers market dynamics, business trends, and the economic forces driving growth across industries. His analysis connects macro movements with real-world implications for investors, entrepreneurs, and professionals. Through his work at The Daily Overview, Grant helps readers understand how markets function and where opportunities may emerge.
