AI Stocks Aren’t in a Bubble. Look at the Numbers.

Microsoft is layering AI into its cloud offerings, making it even more competitive in the space. It’s also launching ChatGPT—and its profit outlook in the last quarter was stellar. The software giant’s stock is up more than 35% year to date. Alphabet, which is layering AI into its ad offering and launching Bard, a ChatGPT-competitor, has seen shares surge nearly 40% so far this year.

Meta and Nvidia stock have more than doubled in 2023. Meta, the parent of Facebook and Instagram, had seen shares beaten down in 2022, so they were especially cheap before the AI rally. Nvidia is expected to sell more chips in its data-center business to power AI-related cloud advancements. That’s lifting AMD stock more than 80% year to date, as this chip maker also sees a chunk of sales come from the data center. 

These gains, on the surface, look out of control. The recent outperformance of these large-cap tech stocks versus the rest of the market has been on par with that of previous bubbles, which strategists at Evercore recently pointed out. The total market value of seven of the largest U.S. tech stocks now account for just under 30% of the

S&P 500’s
market value, the highest proportion since at least 2013, according to Bank of America. Strategists at the bank call the AI rally a “baby bubble.”

Making an analogy to a bubble, regardless of size, seems tempting, but some key factors just don’t support that case.

First off, valuations are up, but they aren’t in bubble territory. The aggregate forward price/earnings multiple on the technology-heavy

Nasdaq Composite Index,
which comprises these AI stocks, is at about 27 times. That’s below the 35 times it hit before the pandemic-era bubble started to burst in 2020, and nowhere near the dot-com peak in early 2000 of over 60 times, according to RBC. In those earlier days, the Nasdaq “looked highly stretched but that’s not the case today,” Lori Cavlasina, chief U.S. equity strategist at RBC, wrote in a research report. 

Moreover, these multiples are fundamentally justified. Analysts expect the Nasdaq’s annualized earnings-per-share growth to be almost 18% for about the next three years, according to FactSet. That means the current multiple is about 1.5 times the growth rate. Simply put, this “PEG ratio,” which stands for price/earnings-to-earnings growth ratio, means investors are paying 1.5 P/E multiple points for every percentage point of earnings growth they’re getting. That’s not so high considering that the S&P 500’s PEG ratio is jut over 2. 

Multiples over the next few years could be stable as both earnings and stock prices rise, when more investors pile in. Nvidia stock, for example, could trade at just over $530 by the end of next year, about 39% above the current $381; that’s assuming shares continue to trade at the current 45 times earnings, and that 2025 earnings-per-share estimates of $11.84 remains in place. The current multiple seems justified, given that it’s not even 1 time the annualized 50% earnings-per-share growth analysts expect over roughly the next three years. That’s a lot different from the real bubble in the dot-com era when unprofitable companies traded at high valuations to estimated sales.

To be sure, AI stocks may need to take a breather. They’re already moderating near their recent peaks. They may even see a small decline in the near term, but that might be mere drops in the bucket relative to the larger uptrends these stocks are in—and buying dips could prove prescient. 

Write to Jacob Sonenshine at jacob.sonenshine@barrons.com