Nvidia's PEG Ratio Is Screaming Buy and the Market Isn't Listening

The 36x trailing PE captures a full fiscal year that includes quarters where Nvidia was still ramping its data centre business. By the time those older quarters roll off, the earnings base will have shifted dramatically. Consensus estimates for FY2027 put EPS growth north of 25%, which compresses the forward multiple further.

Consider the maths. At $4.90 EPS on a trailing basis and a share price around $177, you get the 36x headline number. But the forward estimate — baking in the demand pipeline from hyperscalers, sovereign AI programmes, and enterprise inference deployment — points to EPS above $8. That is a forward PE below 22x for the company that controls roughly 90% of the AI training chip market.

The semiconductor sector median forward PE sits around 24x. Nvidia, growing revenue at 65% annually, trades at a discount to the sector median. Read that sentence again.

We saw a similar dynamic in 2016-2017 when Nvidia's gaming GPU cycle was accelerating. The trailing PE looked stretched at 40x, but the forward PE was compressing rapidly. Investors who sold on the trailing number missed a 300% run over the next two years.

Operating margins at 65% are not just high — they are structurally defensible. Nvidia's CUDA ecosystem has created switching costs that no competitor has cracked. AMD's MI300X is gaining share at the margin, but enterprise and hyperscaler procurement cycles are multi-year commitments. Once a data centre architecture is built around Nvidia's stack, the cost of switching exceeds the cost of paying Nvidia's premium.

Profit margins of 55.6% on $215.9 billion in revenue produced $120.1 billion in net income. That is more net profit than the entire revenue base of most S&P 500 companies. The cash generation is so substantial that Nvidia could retire its entire market float at current prices in roughly 45 years using FCF alone — and that calculation ignores growth entirely.