AMD trades near 97 times forward earnings in one comparison, while Nvidia sits around 21 times, raising valuation-model questions

A market comparison circulating on Yahoo Finance highlights a striking difference in valuation between Advanced Micro Devices and Nvidia, using forward earnings multiples, a common metric that compares a company’s stock price to analysts’ expected earnings over the next year.

In the article, AMD is described as trading at about 97 times forward earnings, while Nvidia is described as trading at about 21 times. The author frames this as a sign that “someone’s valuation model is wrong,” suggesting the market may be underestimating Nvidia’s future fundamentals or overestimating AMD’s.

Forward earnings, often reported as a forward price-to-earnings (P/E) multiple, depends on two moving parts: the company’s future profit trajectory and the accuracy of the estimate. If either side is off, the resulting multiple can look extreme even if the underlying business is changing in line with expectations. The article’s core claim is not that one company is definitively “cheap” or “expensive,” but that the magnitude of the multiple gap is unusually large for two major semiconductor competitors.

The post also implies that the debate is likely less about accounting labels and more about assumptions. Valuation models typically incorporate expected revenue growth, gross margin expansion, operating expense discipline, and the timing of when earnings should materialize. They also embed expectations about competition and product cycles. When one company’s future earnings are expected to ramp faster than the other’s, that difference can compress or expand multiples sharply, even before any new quarter arrives.

Nvidia and AMD are both widely followed by investors, and their trading levels often become proxies for how strongly the market expects compute and AI-related spending to convert into company profits. However, the specific article does not lay out detailed segment-by-segment drivers, such as contract wins, backlog trends, or product-level unit forecasts. Instead, it centers on the headline multiple comparison and the inference that the market’s valuation math may be inconsistent.

The author’s conclusion is essentially a challenge to the underlying premise of the forward estimates: either Nvidia’s forward earnings outlook is being set too low, or AMD’s is being set too high. Because forward P/E uses analyst expectations rather than actual trailing profits, it can remain sensitive to revisions in estimates. In practice, that means a company can appear “mispriced” at a point in time if consensus expectations have not yet caught up to developments, or if they have already overreacted.