Nvidia as a 'Value Stock'? Running the Numbers on a $4.8T Company

Nvidia is up about 6% year-to-date — dead last in a group where Intel is +151%, Marvell +85%, Micron +68%, AMD +61%. So a few people are seriously calling Nvidia "a value play." That phrase made me laugh out loud, because the company is worth $4.8 trillion. But intuition isn't proof, so I ran the numbers.

The fundamentals are absurdly good

Start with the business.

• Market cap: $4.8T
• Gross margin: 71% (Intel is at 41.5%)
• Net margin: 55% — after taxes and overhead
• TTM operating cash flow: $97B
• TTM net income: $120B
• Very little debt

Note the contrast: Intel's 41.5% is before taxes; Nvidia's 55% is after. That's the gap.

8 Pillars: everything passes except valuation

On the Everything Money framework, Nvidia checks every box except one — valuation. And that one box drives the price decision.

One caveat: trailing-1Y FCF is roughly 2.5x the 5-year average. Looking at 5-year P/FCF or P/E here is somewhat misleading because the business is in a vertical phase. Usually 1Y and 5Y averages sit close; Nvidia's don't. You have to remember that when reading the multiples.

Bull vs. bear

Bull

• Blackwell is the most advanced AI chip shipping today
• Demand continues to outrun supply
• Hyperscaler AI capex from Amazon, Meta, Microsoft, Google shows no sign of slowing

Bear

• DeepSeek revived the question of whether AI really needs this many GPUs
• The ROI on hyperscaler capex hasn't shown up yet — and if it doesn't, spending slows
• A $4.8T price already prices in clean future execution

Both cases are legitimate. So the price has to decide.

My 10-year scenarios

I ran these intentionally conservative — assuming AMD takes some share and margins normalize.

Scenario	Revenue growth	Net margin	Exit P/E
Bear	10%	30%	20
Base	15%	37.5%	24
Bull	25%	45%	28

Even the bull case's 45% margin is below the current 55%. The framing is: "some share loss, some margin compression — and even then…"

Outputs: low $83, mid $175, high $513. Current price $195.

At the midpoint, you're roughly getting a 9–10% forward return. That's S&P 500 territory. There's no reason to take single-name risk to capture index returns.

Cross-check with consensus

Analysts model:

• Revenue: $217B → $684B over the next several years
• EPS: $4.80 → $13.27

At $13.27 × 25 P/E that's ~$331 — roughly 70% upside from $195. But remember analysts carry a structural bias: they can't drift too far from price, so consensus is partly anchored to today's quote.

My take

There is almost nothing wrong with Nvidia as a company. Margins, growth, market position — all exceptional. But "great company" doesn't guarantee "great price."

We've seen this pattern before — EV adoption, cannabis legalization, the metaverse. Each had a kernel of truth. Each rewarded investors who entered at the right price and punished the ones who didn't.

To justify calling a $4.8T name a "value stock," you'd want clear margin of safety. The current $195 vs. midpoint $175 isn't a margin — it's a slight premium. I'd call this fairly to slightly fully priced. If you just want index-like returns, you don't need to take single-name risk to get them.

Related: Intel's 151% rally — already priced in?, AMD, Micron, Marvell vs. fair value