Billionaire David Tepper is something of an investing legend. The philanthropist and hedge fund manager runs Appaloosa Management, the hedge fund he founded, with $6.45 billion in assets under management. He cemented his name in the annals of Wall Street by predicting the housing market collapse and subsequent financial crisis in 2008. He bucked the trend, piling into bank stocks on the heels of the Great Recession and "banking" $7 billion in profits in the process, becoming the most successful hedge fund manager of the era.
Some say Tepper is "arguably the greatest hedge fund manager of his generation." He "has consistently outperformed his industry peers and the broader global markets since inception," according to Tepper's Carnegie Mellon University bio.
With that as a backdrop, it's notable that Tepper sold 100% of Appaloosa's stake in Broadcom (AVGO -1.33%) in the second quarter, ended June 30, and is piling into this artificial intelligence (AI) chipmaker instead.
Image source: Getty Images.
A cornerstone of the internet
Broadcom offers a range of technology solutions that have broad reach across the internet. Its semiconductor, software, and security products are staples in the mobile, broadband, cable, and data center industries. In fact, Broadcom says "99% of all internet traffic crosses through some type of Broadcom technology."
However, it's the advent of AI that has been the company's biggest growth driver over the past few years. For example, in the second quarter, ended May 4, Broadcom delivered revenue of $15 billion, up 20% year over year, resulting in adjusted earnings per share (EPS) of $1.58, a 44% jump. Management noted that strong demand for AI fueled its surging growth, as AI revenue of $4.4 billion represented a 46% increase, extending a year-over-year growth streak that reaches back nine consecutive quarters. Within this category, custom AI accelerators grew by double digits, while AI networking solutions soared 70%.
Tepper probably saw an opportunity when Broadcom stock slumped 28% in the first quarter, weighed down by broader economic concerns and tariff uncertainty. However, in Q2, Broadcom stock surged 64% as investors looked past those worries and focused on the future of AI. With everything going so well, why would Tepper sell? The move was probably related to the stock's valuation, as Broadcom ended the quarter with a forward price-to-earnings (P/E) ratio of 41.
Given its performance, the stock is certainly worthy of a premium, but Tepper probably saw a more compelling opportunity in Nvidia (NVDA -0.16%). The hedge fund manager increased Appaloosa's stake in the chipmaker by 483%, amounting to 1.75 million shares and roughly 4.3% of its portfolio.
Change of heart?
After being a persistent seller of Nvidia over the past year, unloading 93% of his shares, Tepper appears to have had a change of heart.
After all, Nvidia pioneered the graphics processing units (GPUs) that are the gold standard for AI. Most AI processing occurs in the data center, and Nvidia holds a dominant 92% share of the data center GPU market, according to IoT Analytics.
If there was any question about the demand for AI chips, Nvidia helped dispel those concerns with its recent results. In its 2026 first quarter, the company generated revenue of $44 billion, an increase of 69% year over year and 12% sequentially. Adjusted for a one-time charge related to a moratorium on its H20 chip sales to China, EPS of $0.96 marked a 62% increase.
Management believes these robust sales will continue. For the second quarter, Nvidia is guiding for revenue of $45 billion, which would represent year-over-year growth of about 50%. That's on top of 122% revenue growth in the prior-year quarter, making it all the more impressive.
We don't know exactly when during the second quarter Tepper added to his stake, but a review of the stock chart might provide a clue. In early April, Nvidia stock plunged to its lowest price in nearly a year, and its valuation became downright intriguing. The stock was selling for less than 21 times forward earnings, marking its lowest valuation in more than a year.
It turns out that Tepper probably didn't know something that Wall Street doesn't. He simply saw an opportunity that was simply too good to resist, and he pounced. Should investors follow suit?
Nvidia has since recovered from its early April swoon, yet I would argue that the opportunity remains compelling. The company's AI-centric Blackwell processors are flying off the shelves, and the U.S. government has lifted its moratorium on sales of Nvidia's H20 chips to China. Furthermore, at 31 times next year's expected sales, Nvidia, while not as cheap as it was earlier this year, is a great picks-and-shovels way to play the AI revolution.
