For well over two years, the stock market has been roaring higher, with all three major indexes -- the Dow Jones Industrial Average, S&P 500, and Nasdaq Composite -- reaching multiple record-closing highs.
While select catalysts have played a role in lifting equities, such as stock-split euphoria, better-than-expected corporate earnings, and Donald Trump's November victory, nothing has sustained this rally quite like the emergence of artificial intelligence (AI).
Empowering software and systems with the ability to reason, act on their own, and evolve to learn new skills, gives this technology a seemingly limitless ceiling. Although the analysts at PwC pegged the global addressable market for AI at $15.7 trillion by 2030 in Sizing the Prize, the sky truly is the limit if businesses adopt AI solutions at a high rate and the technology gains mainstream utility across a wide swath of industries.
No company has been a more direct beneficiary of the rise of AI than Nvidia (NVDA +0.87%). At its peak, Nvidia stock added more than $3 trillion in market value in less than two years.
Image source: Getty Images.
Despite being placed on Wall Street's pedestal, Nvidia stock has shed 26% of its value – calculated from its all-time intra-day high of $153.13 on Jan. 7, 2025, through its closing price of $112.69 per share on March 7.
The million-dollar question is: How much further can Nvidia stock fall? Historic precedent can be a useful tool to answer this question.
Make no mistake, Nvidia is doing a lot of things right
But before making assumptions about the future, it's just as important to understand the path Nvidia took to get where it is today.
Investors can make an argument that the AI revolution doesn't occur without Nvidia's hardware. The company's graphics processing units (GPUs) are the effective brains that power split-second decision-making in high-compute data centers. Demand for Nvidia's Hopper (H100) chip and now its successor Blackwell GPU architecture has been off the scales.
NASDAQ: NVDA
Key Data Points
To build on this point, Nvidia hasn't been afraid to invest aggressively in innovation. Even with its Hopper GPU maintaining well-defined computing advantages over other AI chips, the debut of Blackwell further solidifies its hardware as the fastest. Additionally, Blackwell is considerably more energy efficient than its predecessor.
Something else that's undeniably helped Nvidia get to where it is today is AI-GPU scarcity. Even with Taiwan Semiconductor Manufacturing ramping up its monthly chip-on-wafer-on-substrate (CoWoS) capacity -- CoWoS is necessary to package the high-bandwidth memory needed for high-compute data centers -- demand for AI-GPUs has decisively overwhelmed their supply. For Nvidia, it's led to exceptional pricing power and sizable uptick in its gross margin.
Nvidia's CUDA platform is also helping to tie things together. CUDA is the toolkit developers rely on to build large language models and get the most out of their Nvidia GPUs. In other words, it's a tool that's helping to keep customers loyal to its ecosystem of products and services.
The final piece of the puzzle for Nvidia has been its draw with Wall Street's most-influential businesses. Microsoft, Meta Platforms, Amazon, and Alphabet have consistently been some of Nvidia's top customers.
Image source: Getty Images.
Nvidia is down 26% -- here's how much further it might decline
With a better understanding of how Nvidia became one of Wall Street's most-valuable businesses, let's return to the question at hand: How far could its stock fall?
Although history can't concretely predict the future with 100% accuracy, it does have an uncanny track record of rhyming, more often than not, on Wall Street. Based solely on historic precedent, Nvidia stock has plenty of downside to come.
For starters, every next-big-thing technology and innovation for the last three decades has endured an eventual bubble-bursting event. This is to say that investors have consistently overestimated the early innings adoption rate of a new technology or innovation, along with its early stage utility. All hyped innovations need time to mature, and there's nothing to suggest that artificial intelligence is going to be the exception to this historic correlation.
If investors dive into the companies building out their data center infrastructure and AI solutions, they'll find that most aren't anywhere close to optimizing this technology, nor are they generating a positive return on their AI investments. These are hallmarks of investors overestimating the adoption and utility of a next-big-thing trend, and it points to a bubble forming.
Previous bubbles of game-changing technologies, such as the internet, 3D printing, blockchain technology, and the metaverse, to name a few, saw the leading businesses of these trends lose in the neighborhood of 80% to 90% of their value on a peak-to-trough basis.
However, there's an asterisk to this historic data. The market leaders of next-big-thing trends that were hit the hardest over the last three decades often weren't diversified. Prior to the AI revolution taking shape, Nvidia was selling GPUs for PC gaming and cryptocurrency mining, and it had a well-established virtualization software segment. This is to say that these existing segments would more than likely provide a notably higher floor if the AI bubble were to burst.
Valuation -- specifically the price-to-sales (P/S) ratio -- can also provide valuable historic context.
NVDA PS Ratio data by YCharts.
With few exceptions, such as Palantir Technologies' P/S ratio of nearly 100 three weeks ago, market leaders of next-big-thing innovations have peaked at P/S ratios in the neighborhood of 30 to 40 over the last 30 years. This is true of Microsoft, Amazon, Cisco Systems, and Nvidia, the latter of which topped out at a P/S ratio of 42.39 last summer.
As of the closing bell on March 7, market leaders Alphabet, Meta Platforms, and Microsoft were commanding respective P/S ratios of 6.3, 9.9, and 11.2. Even after Nvidia's 26% pullback, it's still trading at a multiple of roughly 21.4 times sales. To simply align with other Mag-7 stocks, in terms of P/S ratio, Nvidia could easily shed a little over half of its value. This would take its share price down to around $55 from a peak of $153.13.
History also teaches investors that early stage scarcity tied to a hyped innovation doesn't last. While Nvidia has benefited immensely by charging a significantly higher price for its H100 and Blackwell GPUs, a ramp-up in production by competitors, coupled with growing internal competition (most of Nvidia's top customers by net sales are developing their own AI-GPUs), spells big trouble for its pricing power and gross margin.
Though Nvidia stock might appear attractively valued to some investors, historic precedent points to additional downside.
