These two companies design chips for the same customers, chase the same AI budgets, and get talked about in the same breath. On Investily's report card they are not close. Nvidia scores 96 out of 100. AMD scores 85. That gap is worth understanding on its own — but the part that should stop you is what happens when you put their valuations next to their returns. The better business is also, right now, the cheaper stock.
The scores: 96 vs 85
Investily grades every stock on four pillars — Growth, Returns, Ratios, and Valuation — built from reported financials rather than narrative. AMD's 85 is a genuinely good score; it is not a broken company, and it beats most of the market. But the two report cards fail in different places, and where a company loses points tells you more than the headline number:
- Nvidia: no weak pillar. Growth, returns, and balance sheet all near the top of the range.
- AMD: strong growth, clean balance sheet, and a returns pillar that drags the whole card down.
Growth: AMD wins a line, and it's the wrong one
Over the trailing year AMD grew operating income 183.2% against Nvidia's 108.0%. Taken alone that looks like the upset. It isn't. AMD is growing operating income off a base that was close to zero — its three-year operating income growth doesn't even compute cleanly, because the starting point was a loss. Percentage growth from near-nothing is arithmetic, not momentum.
The line that actually separates them is revenue. Nvidia grew revenue 85.3% over the trailing year against AMD's 38.1%, and it did that on a revenue base roughly seven times larger. Growing faster than a smaller rival, from a bigger starting point, is the harder trick — and it is the one that shows up in the score.
Returns: 82.4% against 3.5%
This is where the comparison stops being close. Return on capital employed measures what a company earns on the money tied up in the business. Over the trailing year:
- Nvidia ROCE: 82.4% — AMD: 3.5%
- Nvidia cash-flow ROCE: 61.8% — AMD: 6.2%
- Nvidia net margin: 52.4% — AMD: 9.6%
Nvidia converts more than half of every revenue dollar into profit and earns over 80 cents a year on every dollar of capital employed. AMD earns roughly three and a half. Those are not two points on a spectrum. They describe two different kinds of business that happen to sell into the same market.
Why AMD's returns collapsed
AMD's returns weren't always this thin, and the cause is visible in the balance sheet rather than the income statement. Over the past five years AMD's book value compounded at 47.9% a year while revenue compounded at 18.2%. Capital grew roughly two and a half times faster than sales. ROCE is earnings measured against that capital, so when the denominator inflates faster than the business, returns fall even if profits are rising.
The honest read is that this is recovering, not deteriorating. AMD's ROCE was negative on a three-year view, sits at 3.5% over the trailing year, and is running at 5.5% over the last six months. The direction is right. The level is still nowhere near what its multiple assumes.
Valuation: the cheaper stock is the better business
Here is the part most comparisons never reach. Multiples move with the market every day, so the number worth holding on to is not either company's P/E — it is the gap between them. On the snapshot below, AMD trades at more than four times Nvidia's trailing earnings multiple. On next year's consensus the gap narrows but survives at close to three times. Weighted against growth, AMD's PEG sits around four times Nvidia's.
Both stocks ride the same AI trade, so they tend to re-rate together — which is exactly why the ratio between them is far more durable than either figure on its own. For reference, the snapshot those ratios come from puts Nvidia near 27x trailing earnings and AMD near 117x. Those two numbers will be wrong by the time you read this. The relationship is the part that holds.
So the company with five times the net margin, more than twenty times the return on capital, and faster revenue growth is priced at a small fraction of its rival's multiple. That is unusual enough to deserve suspicion, and the honest explanation is that the market is not pricing today — it is pricing the odds that Nvidia's current returns are a peak and AMD's are a trough.
The bottom line
AMD is the story stock here, and stories are what triple-digit multiples are made of. The buy case rests on that returns pillar climbing toward Nvidia's — and the six-month trend says the climb has started. The risk is that it has a very long way to go and the price already assumes it arrives.
Nvidia's risk is the mirror image. Nothing in its report card is weak, which means every surprise from here is more likely to be a bad one. A multiple that low against 82% returns on capital isn't the market calling Nvidia cheap — it's the market betting those returns don't last. Both cases turn on the same measure — what each company earns on the capital tied up in it — which is why it carries more weight on the report card than almost anything else.