Nvidia’s 10-for-1 stock split comes with a caveat. The story goes that artificial intelligence (AI) stock may do so next.

Nvidia (NVDA 1.75%) completed a 10-for-1 stock split after the market closed on Friday, June 7th. Like most stock splits, this one followed a significant share price appreciation. Shares are up 225% over the past year and 850% since November 2022, when the launch of ChatGPT sparked the artificial intelligence (AI) gold rush.

Nvidia remains well-positioned to benefit as more businesses invest in AI. Indeed, it is arguably the best pure-play stock AI. However, stock splits have historically been bad news for Nvidia shareholders. The company’s value has fallen by an average of 23% over the 12-month period following past splits.

Here’s what investors need to know.

Historically, stock splits have been bad news for Nvidia shareholders

Excluding the most recent one, Nvidia has completed five stock splits as a public company, and the stock has fallen steadily since. The chart below details when each stock split occurred and shows how the stock performed over the next six months, 12 months and 24 months.

Data source: YCharts.

As shown above, after the last five stock splits, Nvidia returned an average of 8% over the next six months. But shares fell an average of 23% during the 12-month period following the split, and the stock was still down an average of 3% after 24 months. In short, the story goes that Nvidia could suffer a severe downturn in the not-too-distant future.

Of course, past performance is never a guarantee of future results, and that saying is especially true here. Four of the last five stock splits occurred near bearish markets. Specifically, to S&P 500 it fell 49% between March 2000 and October 2002 due to the dot-com bubble, and fell 57% between October 2007 and March 2009 due to the global financial crisis.

Despite these market crashes, there was still a line for patient investors. Nvidia stock ultimately recovered after all five stock splits and produced phenomenal long-term returns. The chart below shows the size of those returns as of June 12, 2024.

Ultimately, whether Nvidia is a good or bad investment depends on two things: (1) how quickly the semiconductor company can grow earnings per share, and (2) how much investors are willing to pay for those earnings.

Nvidia has a sustainable competitive advantage

Nvidia’s graphics processing units (GPUs) have a near-monopoly in accelerated computing, a discipline that uses specialized hardware and software to accelerate complex data center workloads such as artificial intelligence (AI) and data analytics.

Nvidia products consistently set performance records in AI training and inference in the MLPerf benchmarks, standardized tests that provide unbiased evaluations of AI systems. Additionally, the company holds more than 90% market share in data center GPUs and more than 80% market share in AI chips.

Nvidia brushes aside these strengths and cites growing competition as cause for alarm. Specifically, they point to semiconductor companies such as AMD and cloud providers like Amazon, MicrosoftAND Alphabet, all of which are building chips to replace Nvidia GPUs. But this shallow argument ignores the formidable economic gap Nvidia has in its overall strategy.

CEO Jensen Huang recently told analysts, “We literally build the entire data center.” It means that Nvidia is not just a chip maker, but rather a complete computer company. Nvidia complements its GPUs with central processing units (CPUs) and networking hardware that is purpose-built for artificial intelligence. The company also offers subscription software and cloud services that support AI workflows in multiple end markets, from manufacturing and logistics to customer service and healthcare.

The heart of Nvidia’s computing platform is CUDA, a parallel programming language that allows GPUs (originally designed for graphics processing) to function as data center accelerators. The CUDA ecosystem consists of hundreds of frameworks and software libraries that simplify the development of complex applications. This gives Nvidia a tremendous advantage because no other chip maker has a comparable ecosystem of supporting software.

To quote The morning star analyst Brian Colello, “CUDA is proprietary to Nvidia and runs only on its GPUs, and we believe this integration of hardware plus software has created high switching costs for customers in AI, contributing to the wide of Nvidia.”

Nvidia stock looks a bit pricey at its current valuation

Going forward, Wall Street expects Nvidia to grow earnings per share by 31.7% annually over the next three to five years. If that number is divided by the current price-to-earnings multiple of 75.8, the ratio is a price-to-earnings-growth (PEG) ratio of 2.4. This multiple is a discount to the three-year average of 3.2, but is still relatively expensive on an absolute basis.

That said, the consensus forecast for earnings leaves room for upside. According to Grand View Research, spending on AI hardware, software and services is projected to grow at 36.6% annually through 2030. Nvidia can certainly match that pace, and possibly surpass it. To that end, his current valuation may seem quite reasonable or even cheap in hindsight.

Ultimately, investors have a somewhat difficult decision here, but I think Joseph Moore Morgan Stanley has the right idea. “[W]”I think the background warrants AI exposure even amid the extreme enthusiasm — and Nvidia remains the clearest way to get that exposure,” he wrote in a recent note to clients.

Suzanne Frey, an executive at Alphabet, is a member of The Motley Fool’s board of directors. John Mackey, former CEO of Whole Foods Market, a subsidiary of Amazon, is a member of The Motley Fool’s board of directors. Trevor Jennewine has positions in Amazon and Nvidia. The Motley Fool has positions in and recommends Advanced Micro Devices, Alphabet, Amazon, Microsoft and Nvidia. The Motley Fool recommends the following options: long January 2026 $395 Microsoft calls and short January 2026 $405 Microsoft calls. The Motley Fool has a disclosure policy.