Many artificial intelligence (AI) stocks are flying high, and have been for multiple years. Nvidia (NVDA +4.30%) and Palantir Technologies (PLTR +1.10%) have been among the hottest investments to own: Both are up by more than 1,000% since 2023.
Nvidia provides the largest volume of the AI-accelerator chips needed for developing and powering next-gen technologies, while data analytics company Palantir helps businesses find ways to uncover efficiencies with the help of its AI-powered platform.
These and other stocks have delivered tremendous returns for patient shareholders. But five years from now, I predict many AI investors will regret not taking heed of one specific piece of Warren Buffett's wisdom, which is to be fearful when others are being greedy.
Image source: Getty Images.
Why the AI bubble could burst
There's plenty of greed in the markets these days, as evidenced by the lofty valuations some stocks are trading at. Palantir is perhaps the best example of that. While it has been declining over the past month, its market cap still hovers around $400 billion -- a remarkably steep valuation for a company that only generates around $4 billion in annual revenue.
While CEO Alex Karp may be dismissive of people who aren't bullish on his company and its stellar growth, that doesn't mean that paying any price for the stock is justifiable. Growth stock or not, picking up shares when they are trading at more than 100 times sales and 390 times earnings is an incredibly risky move financially, and leaves investors exposed to the potential for significant losses.
NASDAQ: PLTR
Key Data Points
What may trigger the bursting of this AI bubble is a continued weakening of economic conditions broadly. Right now, many tech companies are investing heavily in other related tech businesses, so demand appears to be robust. Whether these trends can truly continue is debatable, particularly in light of the warning signs flashing elsewhere in the economy.
Consumers aren't in great shape and are pulling back on discretionary spending to focus mainly on essentials. And an MIT study earlier this year found that 95% of generative AI investments weren't paying off for companies. The end result for some companies may be a combination of rising costs from unprofitable AI investments and top-line declines due to struggling consumers. That could lead those businesses to make significant cutbacks in their own expenditures to protect their profitability.
The big tech companies are so closely intertwined with one another that Alphabet CEO Sundar Pichai admitted that "no company is going to be immune" from a bubble bursting in AI. Hyperscalers are responsible for the bulk of the growth in the AI sector these days. Any hint of a slowdown in their spending could trigger a significant pullback in the markets.
Low earnings multiples could offer a false sense of security
Under these conditions, the temptation for many investors may be to pivot their portfolios toward tech stocks that currently have modest earnings multiples, based on the premise that those stocks might be relatively safe buys. Nvidia, for instance, may look comparatively cheap, as its forward price-to-earnings multiple is only 23. And its price-to-earnings-growth (PEG) ratio is well under 1.0 -- the level below which stocks are generally viewed as undervalued.
The problem, however, is that those are both forward multiples that are based on analysts' projections and outlooks -- not hard data. Right now, the projections for the tech sector remain rosy, and they assume that AI-related growth will remain strong for years.
But analysts' forecasts do change. And if all of a sudden, there are cutbacks in spending, as there were in tech back in 2022 when companies were left scrambling after realizing they had expanded too quickly and overestimated where demand was headed, those forecasts may change quickly. In the wake of those revisions, a once cheap-looking stock such as Nvidia could look far more expensive.
It's always wise to do a qualitative analysis to gauge whether the assumptions you're relying on in your analysis are sound and reliable. If they're flawed, the whole thesis you might have for investing in a stock could be invalid.
Diversifying outside of tech could be crucial
Some AI-focused investors won't scale back their exposure to tech and diversify more into other sectors. I predict that within the next five years, they'll regret not doing so. There's a big disconnect these days between the tech world and what's happening with consumers and the overall economy. When such conditions persist, they can set the market up for a correction or perhaps even a crash.
Investing in blue chip stocks from other sectors or holding exchange-traded funds that give you broad exposure outside of tech can help reduce the overall risk that an AI bubble poses to your portfolio. While the current conditions don't mean you need to get out of the market, it would be wise to adapt to it and manage your overall risk. Failing to do so could leave you with significant losses later on.
