Retail investors are clearly bullish on artificial intelligence (AI) stocks as the new year begins. In The Motley Fool's 2026 AI Investor Outlook Report, 90% of investors polled said they will buy or hold AI stocks this year. However, that's a decidedly different perspective on the stock market than legendary investor Warren Buffett.
Buffett spent his last year running Berkshire Hathaway (BRK.A +0.71%) (BRK.B +0.74%), trimming positions and stockpiling cash. Buffett has generally eschewed technology stocks, although Apple has become his largest position over the years. However, he has spent most of the past few years significantly reducing that stake while not buying much of anything. In fact, over the past 12 quarters, Berkshire has sold more stock than it has bought every quarter.
More recently, that hesitancy to buy has extended to his own stock, as Berkshire has not repurchased any of its own shares since Q2 of 2024. Given its lack of stock buybacks and equity purchases, Berkshire ended Q3 with a record $381.7 billion of cash on its balance sheet.
Image source: Getty Images.
While Berkshire did add a position in Alphabet in Q3, these actions are clearly being made by someone who thinks the market is significantly overvalued. And there are certainly some valuation metrics that back up Buffett's position. One of Buffett's favorite indicators as to whether the market is overvalued is taking the total U.S. stock market capitalization (market cap) and dividing it by the country's gross domestic product (GDP). In fact, this is commonly referred to as the Buffett Indicator. It currently sits at 222%, which considers the stock market considerably overvalued. A ratio between 111% and 135% is viewed as fairly valued.
There are other metrics that also point to the market being overvalued. The S&P 500's (^GSPC +0.01%) trailing price-to-sales (P/S) ratio recently hit an all-time level of 3.3 times, more than double its long-term median of 1.6 times. The Shiller P/E (CAPE) ratio, which measures the current value of the S&P 500 index and divides it by the index's 10-year inflation-adjusted earnings, is also at extreme levels, right around 40 times. That's only the second time this metric has climbed above 40, and the other time was before the dot-com bubble burst.
Should investors worry about a market crash?
It is worth noting that most of the metrics mentioned above are trailing metrics, not future metrics. Investors generally invest in stocks based on their perception of future earnings and cash flow growth, not what they've done in the past.
Meanwhile, it's also notable that the S&P 500 and the market as a whole have really changed over the years. Much of the U.S. market's market cap is now in the hands of large tech companies. In 2003, technology stocks only represented a 14% weighting in the S&P 500, while today it is close to 40% if you include Amazon and Tesla as tech stocks. These are strong companies with solid balance sheets that generate a lot of cash flow and are growing quickly. Many of these top tech companies have been riding long-term secular trends, and they don't have the same type of cyclicality as most other industries. As such, comparing the valuation of today's market with the markets of the past that were much more heavily weighted toward financial and industrial stocks just does not make much sense. Meanwhile, Buffett himself has admitted his struggles with analyzing tech stocks and has missed out on a lot of gains in the sector.
Now, whether the market is overvalued and is going to crash this year or sometime soon after is largely going to depend on whether AI infrastructure spending is more secular or cyclical in nature. AI infrastructure spending is driving a lot of growth in the market right now, and this growth can continue well into the future if there are a lot of winners and companies are getting a return on their AI investments. However, the semiconductor industry has been one area of technology that has tended to be cyclical in nature, where spending does ebb and flow.
Right now, I think AI is still in the early innings, and because of that, it is a place where investors want to invest over the next several years. At some point years down the road, AI infrastructure spending will likely become more cyclical in nature. We are just not there right now, and I view the current trend as being more secular.
That said, now is not the time to be afraid of the market being overvalued, given its current makeup. AI is a once-in-a-generation technological shift, and you want to be invested in the companies leading this charge. I'd expect another solid performance from the market in 2026.
