The AI ​​market is hot right now, but here’s why it could slow down in 2025

The excitement around artificial intelligence (AI) has led companies to spend millions of dollars on new projects and upgrade their existing infrastructure in the hope that they can take advantage of new growth opportunities. Chip maker Nvidia (NVDA 1.40%) tripled its sales due to incredible AI-fueled demand for its next-generation chips. Palantir Technologies (PLTR 2.75%) has also seen an acceleration in its growth rate as the data analytics company sees a surge in demand for its new AI-based platform.

Tech stocks, as a whole, have done well this year. The Vanguard Information Technology ETF has grown by 30% in the last 12 months, which is much higher than S&P 500gains of about 21%. But the danger is that the hype could die out. Companies could cut back on spending, especially if a recession hits. Then the stocks that flew high because of AI could come back to reality. Here’s why a slowdown could happen as early as next year.

Companies may end up abandoning generative AI projects

When there’s a flurry of spending activity, there’s often a correction that happens afterward, when companies stop to decide whether that spending is really justifiable. A recent example of this was the pandemic-fueled spending that took place in 2020 and 2021. Many companies continued to hire due to strong consumer demand. But it won’t be long after that layoffs occur as many companies, especially in the tech space, realize they overestimated demand and hired too many people.

A similar situation could happen with generative AI. Research and consulting firm Gartner projects that by the end of next year, companies could abandon 30% of generative AI projects. The company indicates not only high expectations, but also how intense the spending can be. Distinguished VP Analyst Rita Sallam says, “After the hype of last year, executives are eager to see returns on GenAI investments, but organizations are struggling to demonstrate and realize value. As the scope of initiatives widens, the financial burden of developing and implementing GenAI models is increasingly felt.”

That could be bad news for AI stock, which is trading at inflated valuations

Many stocks have soared to skyrocketing valuations due to expectations that there will be more growth due to AI. And while AI spending may grow over the long term, the pace at which it has been may simply be unsustainable. Nvidia, for example, now generates almost as much revenue per quarter as it previously did on an annual basis, before all the AI ​​hype. For the period ending April 28, its revenue totaled more than $26 billion. In fiscal year 2023 (which ended on January 29, 2023), its annual revenue was just under $27 billion.

Nvidia stock is expensive, and the main counter to its lofty $2.7 trillion valuation is that it’s cheap when you factor in its future growth. However, the problem is that those future growth prospects may be too optimistic. If you assume that the business can continue to grow at a rate of 50% or more per year, then you might conclude that the stock is cheap.

If there’s a recession, however, or if companies start to think twice about their generative AI projects and realize they’re not paying off as profitably as expected, executives can hit the brakes and significantly slow investment in AI, including new chips. Looking at Nvidia’s valuation, investors don’t seem to be considering this possibility, which is why it can be a bit of a risky stock to own, as there is almost no margin of safety with this type of investment.

Palantir is another example. The CEO of the company typically generates the growth opportunities for his business. While its growth rate has accelerated, investors are also paying a hefty premium for the stock based on its future prospects. With more than 170x earnings, it’s an even more expensive stock than Nvidia, depending on how easy its bottom line is.

These are just a few examples, but there are many other AI stocks that could struggle if there is a slowdown in AI spending.

Should You Sell AI Stock?

Not all AI stocks are bad buys. But investors should be careful not to ignore valuations and simply assume that, over the long term, these businesses will continue to achieve incredibly high growth rates. This is based on an extremely optimistic scenario which, should it fall, could result in a significant sell-off.

Valuations are important when considering stocks, and if you’re paying an extremely high multiple for a stock, you should be aware of the risks you’re taking on in doing so and be aware of the assumptions you’re making to justify a large premium. on. In some cases, a high rating may be justified, but in many others, it is not.

Nvidia, for example, is a leading chip company and could continue to do well in the long term, but if there are cracks in its growth rate next year, then 2025 could be a tough year for the stock.