Did Warren Buffett Dump This Artificial Intelligence (AI) Stock Too Soon? It's Up 26% Since Berkshire Hathaway Hit the Sell Button.
Sometimes, even the best investors in the world struggle to time the market.
Warren Buffett is the CEO of Berkshire Hathaway (BRK.A 0.30%) (BRK.B 0.18%). Over his 59-year tenure, Berkshire stock has delivered a compound annual return of 19.8%, almost twice the average annual return in the S&P 500 over the same period.
Therefore, it’s no surprise that Wall Street watches Buffett’s every move. He is on a selling spree this year, trimming several long-standing holdings in Berkshire’s portfolio. Snowflake (SNOW 2.38%) is one stock the conglomerate sold entirely during this year’s second quarter.
The cloud computing company faces a number of challenges, and its stock barely delivered any gains during its four-year stint in Berkshire’s portfolio. However, thanks to the company’s growing presence in the artificial intelligence (AI) industry, Snowflake stock has jumped 26% since Buffett and his team sold out a few months ago.
So, did they pull the trigger too soon? Let’s see.
Image source: Getty Images.
A growing portfolio of AI products and services
Snowflake’s Data Cloud is a revolutionary platform that allows organizations to aggregate their valuable information, where it can be analyzed to extract deeper, more actionable insights. It’s a critical tool because large organizations often spread digital workloads across multiple providers of cloud services, which leads to data fragmentation.
Snowflake’s focus on solving big-data challenges makes it an ideal candidate to build AI products and services. The company launched a new platform in October 2023 called Cortex AI, which allows businesses to access large language models (LLMs) from leading providers like Anthropic and Meta Platforms, which they can combine with their own data to create powerful AI software applications.
Cortex also comes with powerful built-in AI tools. Document AI allows businesses to easily extract data from unstructured sources like contracts or invoices, and Snowflake Copilot is a virtual assistant that can turn simple text-based prompts into computer code, to accelerate tasks like software development.
Earlier this month, the company also unveiled a new tool called Snowflake Intelligence. It allows organizations to easily create data agents, which employees can interact with using natural language. It means even nontechnical staff members can draw valuable insights from their company’s data, which creates opportunities to save money and generate more revenue at every level of the organization.
As of the fiscal 2025 third quarter (ended Oct. 31), Snowflake said over 3,200 of its 10,618 customers were using the company’s AI products. Uptake has been impressive, considering Cortex AI was launched barely one year ago.
Decelerating revenue growth, and blowout losses
Snowflake was growing at a lightning-fast pace when Berkshire invested in the company ahead of its initial public offering (IPO) in 2020. It would regularly generate triple-digit percentage increases in its quarterly revenue.
However, Snowflake has struggled with decelerating revenue growth more recently. Increased competition is one factor behind the slowdown, but it also becomes harder to grow quickly as companies get bigger.
During Q3, Snowflake generated a record $900 million in product revenue, a 29% increase from the year-ago period. While that was a solid growth rate at face value, the downward trend is hard to ignore:
To make matters worse, Snowflake actually ramped up its spending during Q3, with almost $1 billion going toward operating expenses (an all-time high). Research and development spending grew by 33% compared to the year-ago period, and marketing costs jumped 23%.
Typically, increased spending should translate into faster — not slower — revenue growth. Unfortunately, this disparity resulted in a 51% increase in Snowflake’s net loss, which came in at $324.2 million for the quarter. It was the biggest quarterly loss the company has generated in fiscal 2025 so far.
On a more positive note, Snowflake did see a whopping 55% increase in its remaining performance obligations (RPOs), which came in at a record $5.7 billion. RPOs are sort of like an order backlog. But before investors get too excited, the company only expects to recognize 50% of them as revenue in the next 12 months.
There is no clarity on when the rest might convert into revenue, and even though there are contracts in place, there is no guarantee they will convert within a time frame that will actually drive strong revenue growth. With that said, it’s good news that customers are making long-term commitments to spend money on Snowflake’s platforms.
Snowflake stock remains quite expensive
As mentioned, Berkshire invested in Snowflake ahead of its IPO in 2020. The stock came public at a price of $120, and when Berkshire sold its stake during Q2 2024, it was trading at around $135. That represents a total return of just 12% over the four-year period — considering Berkshire stock averages a return of 19.8% every year, Snowflake was a drag on its portfolio, to say the least.
Snowflake stock has jumped 26% since Berkshire sold out. But I don’t think Buffett and his team made a mistake by selling because the stock is very expensive, and decelerating revenue growth makes its current valuation very hard to justify. Snowflake’s price-to-sales (P/S) ratio is currently 16.6. While it’s steadily trending lower, it’s still far more expensive than the three largest providers of cloud services, Amazon, Microsoft, and Alphabet:
SNOW PS Ratio data by YCharts
Those three cloud giants also have a number of other successful businesses, spanning e-commerce, enterprise software, digital advertising, streaming, and more, which Snowflake doesn’t have. Not to mention that they are the three leading developers of AI products and services.
Additionally, Microsoft Azure and Google Cloud grew their revenue at a faster pace than Snowflake in recent quarters — despite the fact they both generate significantly more revenue than Snowflake.
Therefore, I don’t think Snowflake stock is a good fit for a value investor like Buffett, so Berkshire’s sale wasn’t a surprise. Even though the company has a bright future in the AI space, any investor looking to buy the stock now should probably wait for a pullback, given its current P/S ratio.
John Mackey, former CEO of Whole Foods Market, an Amazon subsidiary, is a member of The Motley Fool’s board of directors. Suzanne Frey, an executive at Alphabet, is a member of The Motley Fool’s board of directors. Randi Zuckerberg, a former director of market development and spokeswoman for Facebook and sister to Meta Platforms CEO Mark Zuckerberg, is a member of The Motley Fool’s board of directors. Anthony Di Pizio has no position in any of the stocks mentioned. The Motley Fool has positions in and recommends Alphabet, Amazon, Berkshire Hathaway, Meta Platforms, Microsoft, and Snowflake. The Motley Fool recommends the following options: long January 2026 $395 calls on Microsoft and short January 2026 $405 calls on Microsoft. The Motley Fool has a disclosure policy.