Is Nvidia Cheap Right Now?

The bar was enormous. Nvidia cleared it. The stock still slipped.

That’s the setup heading into Thursday morning – NVDA trading around $223, roughly 1% lower in the after-hours session after printing $81.6 billion in Q1 revenue, guiding to $91 billion for Q2, announcing an $80 billion buyback, and hiking the dividend 25x in a single move. Every number beat. The reaction was a shrug. Which, depending on how you look at it, is either concerning or completely irrelevant to the actual question: is this stock cheap right now?

Here’s where I’m at on that.

On a trailing basis, NVDA sits at roughly 45.6x earnings. That’s 34% above the broader technology sector average of around 34x. Expensive on its face – and if you stopped there, you’d close the tab and move on. But trailing earnings are already stale the moment Nvidia prints a quarter growing 85% year-over-year. Using trailing numbers here is like pricing a car based on last year’s mileage. The rearview mirror doesn’t tell you much when the vehicle is accelerating.

The forward picture is a different conversation entirely. Forward P/E comes in at approximately 26.9x – and that number is sitting 25.5% below the semiconductor industry median of roughly 36x. The most dominant chip platform on the planet, guiding to $91 billion next quarter while assuming zero China revenue, is trading at a discount to its own industry peers on a forward basis. That’s not a sentence most people would have expected to write about Nvidia at a $5.4 trillion market cap.

And those forward estimates are pre-revision. Wall Street will be marking up numbers after tonight’s print. A $91 billion guide with a 75% gross margin doesn’t leave room for analysts to stay cautious. So the 26.9x forward multiple may compress further as earnings estimates move higher over the next few weeks.

The PEG ratio is the data point worth holding. At 0.67, Nvidia’s price-to-earnings-to-growth ratio sits well below 1.0 – the level that has historically suggested a stock is undervalued relative to its own growth rate. At 85% revenue growth and an accelerating guide, a sub-1.0 PEG is not a trivial signal. It’s the kind of number that shows up when earnings are growing faster than the market has priced in.

Slight tangent, but it matters: the 10-year average P/E for NVDA is approximately 66x. The stock is currently trading 31% below that historical average – at a time when it’s generating $96.7 billion in trailing free cash flow, carrying $62.6 billion in cash against $11.4 billion in debt, and posting a return on invested capital of 126%. Investors have paid far more for this company in prior cycles when the business was a fraction of this size.

Now the counterargument, because it exists and it’s not nothing. The average consensus price target sits around $275, which implies roughly 23% upside from current levels – but the stock just hit an all-time high of $236.54 on May 14. At that level, the forward multiple was already compressing fast. And with a market cap north of $5.4 trillion, the law of large numbers starts to create real friction. Jensen Huang has talked about $1 trillion in combined revenue from Blackwell and its successor Vera Rubin across 2026 and 2027. If that plays out, tonight’s valuation discussion looks different in hindsight. If it doesn’t, 26x forward earnings on a slowing grower is a different story.

Cheap in the traditional sense? No. Cheaper than it looks given the growth trajectory? The data makes a reasonable case for yes. The real debate isn’t the trailing multiple – it’s whether Nvidia can keep the guide moving upward faster than the stock price. After tonight, that question got harder to answer bearishly.