Once the darling of the AI boom, Nvidia now trades at its cheapest valuation in years, a sign that investor enthusiasm is cooling even as the company’s revenues keep climbing.
Nvidia’s stock has long been one of the most closely watched in the world, a barometer of confidence in artificial intelligence and the companies building it.
But a sharp drop in the company’s price-to-earnings ratio tells a different story from its record-breaking revenue figures.
The gap between the two is widening, and investors are starting to ask why.
That ratio now sits around 34, roughly 37% below its ten-year historical average and at levels not seen since around 2019.
That compression is striking for a company that reported $57 billion in quarterly revenue, a 62% rise year-on-year.
The numbers themselves are not the problem. The problem is what investors believe comes next.
The AI Boom’s Awkward Second Act
When Nvidia became a household name in 2023, it was because the world suddenly needed its chips to power a new generation of artificial intelligence.
Every major technology company Google, Microsoft, Meta, Amazon was spending billions on AI infrastructure, and Nvidia’s graphics processing units (GPUs) were the engine of choice.
The stock rose more than 1,300% between the end of 2022 and late 2025.
But that kind of growth carries its own weight. When a stock rises that fast, investors price in years of continued outperformance.
Any sign that the pace might slow even slightly can send the share price tumbling.
“The debate is about sustainability and slope, not whether AI demand exists.”
Nvidia’s shares fell more than 9% from their October 2025 record high by the start of 2026, losing around $460 billion in market value.
The issue is not that AI spending has stopped, it has not. It is that the explosive confidence of 2023 and 2024 has given way to harder questions about when, and how much, those investments will actually pay off.
Export Bans Cut Off a Huge Market
One of the most concrete blows to Nvidia’s outlook has come from Washington. The US government has steadily tightened its restrictions on selling advanced chips to China.
In April 2025, it banned the export of Nvidia’s H20 chip,a product designed specifically to comply with earlier restrictions indefinitely.
The move triggered a $5.5 billion charge on Nvidia’s books for chips it could no longer sell.
The consequences go beyond a single write-down. China was once a major source of revenue for Nvidia, and those sales are now effectively gone.
Nvidia’s share of the Chinese AI chip market has fallen from roughly 95% to nearly zero. The void is being filled by domestic Chinese companies in particular Huawei, whose Ascend processors are gaining ground rapidly.
Further tightening came in early 2026, with reports that the US was working on new rules covering nearly all of Nvidia’s and AMD’s most advanced AI accelerators globally, not just China.
Competition Is Catching Up
For most of the AI boom, Nvidia had the field largely to itself. Its CUDA software platform created a powerful lock-in effect.
Switching to a rival chip meant rewriting code from scratch, and most companies simply did not bother.
AMD is gaining ground with its Instinct series of AI chips, and some of Nvidia’s biggest customers including Google, Amazon, and Microsoft are developing their own in-house chips to reduce dependence on external suppliers.
Google’s TPU chips, in particular, have attracted attention as a serious alternative for large-scale AI workloads.
None of this has toppled Nvidia from its position. But it has introduced a level of uncertainty that the market had previously chosen to ignore.
Higher Interest Rates Add Pressure
The US Federal Reserve held interest rates at between 3.5% and 3.75% at its January 2026 meeting, maintaining a “higher for longer” stance that has weighed on growth stocks broadly.
When interest rates are high, investors demand more from future earnings before they will pay a premium today.
That makes stocks like Nvidia where much of the value lies in future growth particularly vulnerable to a repricing.
This is a pattern visible across the semiconductor sector and much of the broader technology market.
Strong Revenue, Cautious Market
Data centre revenue reached $51 billion in Nvidia’s third fiscal quarter of 2026, up 66% year-on-year.
The company guided for $65 billion in revenue in the following quarter. Those are not the numbers of a struggling company.
But once a stock has been priced for perfection, even impressive growth can disappoint.
What the market is reassessing now is not whether Nvidia is a good business most analysts still think it is but whether it is worth the extraordinary premium it commanded at its peak.
Of 65 analysts tracked by major research services, 58 still hold a buy recommendation on the stock.
Price targets range widely, from $100 at the pessimistic end to $270 at the optimistic end, reflecting genuine uncertainty about where Nvidia heads from here.
