Billionaire Ken Griffin Sold 79% of Citadel's Stake in Nvidia and Is Piling Into Another Artificial Intelligence (AI) Stock With a Competitive Moat

A little over two months ago, on Aug. 14, investors received what can arguably be described as the most important data release of the third quarter -- and I'm not talking about any inflation report.

No later than 45 calendar days following the end to a quarter, institutional investors with at least $100 million in assets under management (AUM) are required to file Form 13F with the Securities and Exchange Commission. This filing tells investors which stocks, exchange-traded funds (ETFs), and occasionally options, Wall Street's top money managers purchased and sold in the latest quarter (in this case, the June-ended quarter).

As you might imagine, professional and everyday investors are particularly interested in seeing how Wall Street's smartest asset managers are approaching artificial intelligence (AI) stocks. AI has been the hottest trend over the last two years, with the analysts at PwC expecting this technology to provide a $15.7 trillion boost to the global economy by 2030.

Interestingly, 13Fs from the second quarter show that billionaire investors, including Citadel's Ken Griffin, have a mixed view of the companies powering the AI revolution.

Since its inception in 1990, Citadel's hedge fund has been more successful generating investment gains than any other hedge fund. It's why investors wisely pay attention to what Griffin and his team are buying and selling.

During the June-ended quarter, Griffin oversaw the disposition of a sizable percentage of his fund's stake in market-leading AI stock Nvidia (NVDA -2.81%), but chose pile into another AI company that appears to have a virtually insurmountable moat.

Griffin's Citadel shows the bulk of its Nvidia shares to the door -- and with good reason

Let me preface this discussion by pointing out that Citadel is an active hedge fund with countless positions, and it often hedges its common stock holdings with put and/or call options. With this being said, perhaps no sell-side activity stood out more during the second quarter than Griffin and his crew shedding 79% of their fund's stake in Nvidia.

Since 2023 began, Nvidia's stock has gained 844%, as of the closing bell on Oct. 18, 2024, and tacked on close to $3 trillion in market cap. These are never-before-seen gains from a market-leading business, which suggests that profit-taking played at least some role in Citadel reducing its sizable position. But there's a reasonable chance this selling activity is about more than benign profit-taking.

To start with, competition is coming at Nvidia from all angles. Rival chipmakers are developing and ramping production of AI-graphics processing units (GPUs) designed to complete with Nvidia's ultra-popular H100 GPU and next-generation Blackwell GPU architecture. Although Nvidia's hardware shouldn't have any trouble maintaining a computing advantage in AI-accelerated data centers, competing chips are cheaper and may be easier for businesses to get their hands on.

To build on this point, competition from within can be even more dangerous. All four of Nvidia's top customers by net sales are developing AI-GPUs for use in their data centers. Nvidia's hardware maintaining computing advantages may not be enough when compared to the cost advantages of internally developed chips. This suggests Nvidia could lose out on precious data center real estate down the road.

Ken Griffin and his team at Citadel may also be taking their cues from Nvidia's management team. While CEO Jensen Huang has been incredibly optimistic about his company's future and the successor Blackwell chip, it's coming up on four years since the last time any company insider purchased shares of their stock on the open market. A barrage of insider selling paints a picture that implies Nvidia's stock has gotten ahead of itself.

Lastly, Griffin may be letting history act as a guide. In spite of the buzz surrounding the rise of AI, no game-changing innovation or technology has avoided an early stage bubble in 30 years. Professional and retail investors consistently overestimate the adoption and utility of new technologies, which eventually leads to disappointment and a bubble-bursting event. Since no stock has benefited more than Nvidia from the AI revolution, it would, presumably, be hit the hardest if the AI bubble bursts.

But while billionaire Ken Griffin was meaningfully cutting back his fund's stake in Nvidia, he was buying into another AI company, with well-defined competitive advantages, hand over fist.

Griffin increased his fund's stake in this unique AI stock by more than 1,100%!

The high-flying artificial intelligence stock that caught the attention of Citadel's brightest investors, including Griffin, is none other than cloud-based data-mining specialist Palantir Technologies (PLTR -0.81%). Citadel's hedge fund increased its stake in the company by roughly 1,140% during the second quarter.

The best answer to "why Palantir?" is because it's irreplaceable at scale.

Palantir's two core operating segments are Gotham and Foundry. Gotham is an AI-driven platform that helps government entities collect data and plan various missions. This includes the autonomous tasking of satellites and drones.

Meanwhile, Foundry is Palantir's AI and machine learning-inspired platform that's geared toward businesses. It's tasked with making sense of large sums of data and helps businesses streamline their operations. While there are companies that handle bits and pieces of what Gotham and/or Foundry bring to the table, nothing else is close to the scale of Palantir. Wall Street tends to reward uniqueness, irreplaceability, and sustainable moats.

Much of Palantir's growth has been fueled by Gotham. Many of the contracts Palantir secures from government agencies last multiple years, leading to highly predictable operating cash flow. But Gotham's long-term growth potential is also reasonably limited given that Palantir will only allow the U.S. and its allies access.

Palantir's future is very much reliant on the success of Foundry. Although it's still very early in its expansion, the company's commercial segment appears to be finding its footing. Enterprise customer count jumped 55% to 467 from the prior-year period, as of June 30, with sales from this segment climbing 33%. Considering the addressable market for Foundry has no glass ceiling like Gotham, it has the potential to become a considerably larger profit driver in the latter-half of this decade.

The one notable drawback with Palantir is its valuation. While Citadel got in well ahead of its recent run-up, it's a tough task trying to validate Palantir's forward price-to-earnings (P/E) ratio of 100! Though it looks to have all the tools and intangibles needed for success over the long run, Palantir's stock may find the short-term sledding difficult.