Berkshire Hathaway still hasn’t picked up a single share of the hot AI chipmaker.
Nvidia‘s (NVDA 0.98%) stock has rallied more than 26,000% over the past decade. From fiscal 2014 to fiscal 2024 (which ended this January), the chipmaker’s revenue grew at a compound annual growth rate (CAGR) of 31% as its earnings per share (EPS) rose at a CAGR of 50%.
That impressive growth was driven by the rapid expansion of its data center business, which sells high-end graphics processing units (GPUs) for machine learning and artificial intelligence (AI) tasks, and its steady sales of gaming GPUs, which could also be used to mine certain cryptocurrencies. It also expanded its automotive chip business and ramped up its shipments of Tegra system-on-chips (SoCs) for other devices.
Image source: Getty Images.
Nvidia generated dazzling market-beating gains, but one famous investor who missed that historic rally was Warren Buffett, who didn’t buy a single share of the chipmaker for Berkshire Hathaway (BRK.A -0.43%) (BRK.B -0.39%). Let’s see why the Oracle of Omaha still avoids one of the market’s hottest tech stocks.
Nvidia checks a few boxes for a Buffett investment
Nvidia satisfies some of Buffett’s well-known criteria for picking stocks. Buffett favors companies with wide moats, and Nvidia is clearly the 800-pound gorilla of the discrete GPU market. It held 88% of the desktop GPU market in the first quarter of 2024 according to JPR, while TechInsights estimates it controls about 98% of the data center GPU market today.
Buffett also likes companies with a gross margin of over 40%, which implies it has plenty of pricing power and competitive advantages. Nvidia’s gross margin expanded from 59.2% in fiscal 2023 to 73.8% in fiscal 2024, and rose another 13.8 percentage points year over year to 78.4% in the first quarter of fiscal 2025. That rapid expansion was driven by its unmatched pricing power in the data center GPU market as more companies upgraded their servers to support more AI applications.
Buffett also wants a company to spend less than 30% of its gross profits on research and development (R&D), since it means it isn’t aggressively investing in new projects to maintain its competitive advantage. Nvidia only spent 20% of its gross profit on its R&D expenses in fiscal 2024, and that ratio dropped to just 13% in the first quarter of fiscal 2025.
Buffett believes that companies should spend less than 30% of their gross profits on their sales, general, and administrative (SG&A) expenses, since companies with established moats generally don’t need to spend too much cash to attract new customers. For Nvidia, that ratio came in at just 6% in fiscal 2024 and 4% in the first quarter of fiscal 2025.
Lastly, Buffett looks for companies that can maintain a net profit margin of at least 20%. Nvidia reported a net margin of 49% in fiscal 2024 and 57% in the first quarter of fiscal 2025.
So why hasn’t Buffett invested in Nvidia?
It’s easy to see why so many investors are still bullish on Nvidia. However, Buffett famously avoided tech stocks throughout most of his investing career, saying that he preferred evergreen businesses that generated predictable long-term returns. Moreover, Buffett once told investors to “never invest in a business you cannot understand” — and Nvidia’s chipmaking business can be difficult to fully analyze without an engineering degree.
Buffett finally started buying more tech stocks over the past decade, but he mainly invested in mature blue chip companies like IBM, Oracle, HP, and Apple. The cloud data warehousing company Snowflake was the high-growth tech standout in his portfolio.
Buffett notably reduced his exposure to the tech sector over the past few years. He sold all of his IBM shares in 2017, exited Oracle in 2019, and recently liquidated all of his shares in HP and Snowflake. He even sold half of Berkshire’s massive stake in Apple.
Instead of flocking toward high-growth tech stocks like many retail investors, Buffett has been hoarding a lot of cash and short-term Treasuries. Berkshire ended its latest quarter with a record $277 billion in cash and equivalents — which suggests the conglomerate is getting ready for some buying opportunities in a market pullback. That cautious approach isn’t too surprising, since the S&P 500 is hovering near its all-time highs and looks expensive at 21 times forward earnings.
Buying Nvidia’s stock wouldn’t match that strategy. Its stock isn’t cheap at 47 times forward earnings and 25 times this year’s sales, and Wall Street’s expectations are hitting sky-high levels as the AI market expands, so any hint of a slowdown could crush Nvidia’s valuations. Those aren’t the stable and predictable returns Berkshire Hathaway usually looks for.
Should you follow Warren Buffett’s example?
Buffett clearly favors predictable cash-rich companies in the insurance, banking, railroad, utility, and consumer staples sectors. If you want to follow that conservative investing strategy, you might want to avoid Nvidia. But if you believe the nascent AI market still has room to grow over the next few decades, then you might still want a little exposure to this red-hot stock — even if it doesn’t quite fit Buffett’s criteria for a dependable long-term investment.
Leo Sun has positions in Apple, Berkshire Hathaway, and Nintendo. The Motley Fool has positions in and recommends Apple, Berkshire Hathaway, HP, Nvidia, Oracle, and Snowflake. The Motley Fool recommends International Business Machines and Nintendo. The Motley Fool has a disclosure policy.