Warren Buffett is the chairman and CEO of the Berkshire Hathaway (NYSE: BRK.A)(NYSE: BRK.B) investment company. Berkshire wholly owns several well-known businesses, including GEICO Insurance and Dairy Queen, but it also holds 47 publicly traded stocks and securities.

Under Buffett's 59-year tenure, Berkshire has delivered a compound annual return of 19.8% for its investors. To put it another way, an investment of $1,000 in Berkshire stock in 1965 would be worth a whopping $42.5 million today. The same investment in the S&P 500 would have grown to just $328,500.

The conglomerate's portfolio includes a host of high-quality stocks like Apple, Coca-Cola, and American Express, but in 2020, it acquired a small stake in cloud computing company Snowflake (NYSE: SNOW). In my opinion, Snowflake wasn't a good fit for Berkshire's portfolio, nor did it jibe with Buffett's time-tested investment strategy.

However, according to Berkshire's 13-F filing for the second quarter of 2024 (ended June 30), the conglomerate sold off all of its Snowflake stock. Here's why I'm not surprised.

Snowflake is pivoting toward artificial intelligence

Snowflake specializes in the development of data clouds that help organizations break down compartmentalized data into a unified well of information that they can then extract maximum value from and put to use in their day-to-day operations. Since data is the well that every artificial intelligence (AI) model draws from to generate needed answers, Snowflake is in a great position to build products in that space.

Last year, the company launched Cortex AI, a platform designed to enhance the capabilities of the data cloud and help organizations leverage their data to build AI applications. To accelerate their progress, Snowflake customers can access ready-made large language models (LLMs), like Mistral Large and Meta Platforms' Llama 3, on Cortex.

The platform also comes with a host of useful AI tools. Document AI allows businesses to quickly extract valuable data from unstructured sources like contracts and invoices, and Cortex Search understands natural language, so developers can rapidly retrieve data with a simple prompt.

Cortex AI also comes with an AI-powered virtual assistant called Copilot, which is capable of understanding the context behind an organization's data so it can offer code suggestions and even convert natural language into code, which can save developers a lot of time.

At the end of Snowflake's fiscal 2025 second quarter (ended July 31), the company said that around 2,500 of its 10,249 customers were using its AI products and services weekly, which is pretty good uptake considering most of them were only launched in the past year.

Snowflake's revenue growth is decelerating while its losses are growing

Snowflake generated $829.3 million in product revenue during Q2, a 30% increase from the year-ago period. While the result was better than management's forecast, that growth rate still marked a deceleration sequentially and year over year. Unfortunately, slowing revenue growth has been a consistent theme for Snowflake over the last few years.

That poses a problem because Snowflake is investing heavily in building new products to compete in the AI race. The company reported record-high operating expenses of $936 million in Q2, representing a 26% increase from the year-ago period. Research and development spending alone came in at $437.6 million, up almost 40%.

That resulted in a net loss of $316.9 million, a 40% jump from the same quarter last year. In other words, Snowflake is burning a lot of money at the bottom line, and it doesn't even have stable -- let alone accelerating -- revenue growth to show for it.

The spending is unlikely to slow anytime soon because the company continued to hire more employees in Q2. Its headcount was at a record 7,630 at the end of the quarter, up 14% from a year ago.

On the positive side, Snowflake reported $5.2 billion in remaining performance obligations (RPOs) at the end of Q2, a 48% increase. RPOs typically reflect the order backlog from customers who sign long-term contracts. The company expects to convert half of its RPOs into revenue within 12 months, but it doesn't disclose how long it will take to convert all of them -- so while it's possible this will lead to accelerated revenue growth sometime in the future, it isn't a guarantee.

Why Snowflake wasn't a fit for Buffett's portfolio

Warren Buffett looks for several attributes in a company when he's deciding whether to invest. They include steady growth, robust profitability, a strong management team, and shareholder-friendly programs like dividend schemes and stock buybacks. When he finds a company he likes, he also wants to pay a fair price.

Snowflake continues to grow, and it even has a stock buyback program. However, the company's increasing losses will likely prevent it from introducing a dividend or expanding its buybacks in the future. Plus, while Snowflake has enough cash and equivalents on hand ($3.2 billion) to sustain its losses in the near term, those losses will eventually hamper the company's ability to invest in growth initiatives like marketing and research and development.

So achieving profitability will be critical at some point -- but the potential cost cuts required to get there could lead to even slower revenue growth.

Then there is Snowflake's valuation. Since the company doesn't generate earnings, we can value it using the price-to-sales (P/S) ratio -- market capitalization divided by annual revenue. Based on Snowflake's market cap of $39.5 billion and its trailing-12-month product revenue of $3.1 billion, its stock trades at a P/S ratio of 12.9.

That is quite expensive. Cloud industry leaders like Microsoft, Amazon, and Alphabet are cheaper, despite Microsoft Azure and Google Cloud growing revenue at a similar pace to Snowflake. All three tech giants have diverse portfolios of other growing businesses, which Snowflake doesn't have.

Berkshire bought Snowflake stock around the time of its IPO in 2020, so it likely paid close to $120 per share. It was a small position that accounted for less than 0.5% of the conglomerate's $312 billion portfolio. We don't know which investment manager at Berkshire made the official decision to buy, but it was likely one of Buffett's lieutenants.

The second-quarter low point in Snowflake stock was $124.21, so considering the conglomerate's entry point, Berkshire probably didn't lose any money when it sold. The return, however, was lackluster given the four-year holding period.

Should you invest $1,000 in Snowflake right now?

Before you buy stock in Snowflake, consider this:

The Motley Fool Stock Advisor analyst team just identified what they believe are the 10 best stocks for investors to buy now… and Snowflake wasn’t one of them. The 10 stocks that made the cut could produce monster returns in the coming years.

Consider when Nvidia made this list on April 15, 2005... if you invested $1,000 at the time of our recommendation, you’d have $792,725!*

Stock Advisor provides investors with an easy-to-follow blueprint for success, including guidance on building a portfolio, regular updates from analysts, and two new stock picks each month. The Stock Advisor service has more than quadrupled the return of S&P 500 since 2002*.

See the 10 stocks ?

*Stock Advisor returns as of August 26, 2024

Suzanne Frey, an executive at Alphabet, is a member of The Motley Fool's board of directors. American Express is an advertising partner of The Ascent, a Motley Fool company. 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. John Mackey, former CEO of Whole Foods Market, an Amazon subsidiary, 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, Apple, 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.

Meet the Artificial Intelligence (AI) Stock Warren Buffett Just Sold Completely -- I'm Not Surprised was originally published by The Motley Fool