Nvidia Is So Cheap That It'll Be a Bargain Even If It Doubles, Here's the Math That Proves It

The $5 Trillion Elephant Nobody Wants to Talk About

Let me ask you something.

If I told you a company generated $216 billion in revenue last year, growing 65%, and printed $96.6 billion in pure free cash flow, what would you guess it's worth?

Now what if I told you that same company trades at 26 times forward earnings, sports a PEG ratio of 0.68, and is projected to nearly double its earnings per share over the next two years?

You'd probably say: "That sounds like a screamer. Where do I sign?"

Now here's the punchline: that company is Nvidia. And its market cap is roughly $5.23 trillion.

I know. Your brain just hit a wall.

Five. Trillion. Dollars. The number feels reckless. It sounds like peak bubble territory. Like something you'd see in a meme, not a brokerage account. And that reaction, that sticker-shock reflex, is exactly why Nvidia remains one of the most misunderstood bargains in the market right now.

Because here's the thing the sticker-price crowd keeps missing: valuation isn't about how big the number is. It's about how fast the engine underneath it is growing.

And Nvidia's engine? It's not slowing down. It's accelerating.

The Math That Makes Your Spreadsheet Jealous

Let's strip away the feelings and look at what the business actually produces. Because feelings are terrible investment advisors.

Wait, $5 Trillion and It's Cheap?

Think of it like this: would you rather buy a $500,000 laundromat that makes $20,000 a year, or a $5 million apartment building that generates $500,000 a year?

The absolute price tag means nothing. What matters is what you're getting for it.

Nvidia's FY2026 net income hit $120.07 billion. That's $4.77 in earnings per share. At $215.20 per share, you're paying about 45 times trailing earnings.

"Forty-five times earnings? That's not cheap!"

Hold on. That's the rearview mirror. Nobody buys a stock based on what it earned last year. You buy it for what it'll earn next year, and the year after that.

The PEG Ratio: The One Number Value Investors Dream About

This is where it gets fun.

The forward P/E, based on what analysts expect Nvidia to earn over the next 12 months, sits around 26.

Now here's the magic number: PEG ratio of 0.68.

If the PEG ratio is a new concept for you, here's the simplest explanation you'll ever get: a PEG of 1.0 means you're paying exactly fair value for a company's growth. Below 1.0, and the market is essentially undercharging you for the growth you're getting.

At 0.68, Nvidia isn't just "fairly valued." It's trading at a 32% discount to what its growth rate alone would justify.

To put that in perspective: most mature tech companies trade at PEGs of 2.0 to 3.0. Nvidia, growing 65% a year, trades at 0.68. Find me another company of this size with that ratio. I'll wait.

Free Cash Flow: The Engine That Prints Money

If earnings are the car's speedometer, free cash flow is the fuel gauge. And Nvidia's tank is overflowing.

FY2026 free cash flow: $96.58 billion, up 58.7% year over year.

That's not a typo. A company with a $5.2 trillion market cap generated nearly $100 billion in cash that it can use for buybacks, dividends, R&D, or acquisitions. Management has already committed to returning roughly 50% of free cash flow to shareholders through buybacks and dividends.

For context, Nvidia's free cash flow over the next two calendar years is projected to exceed $400 billion, a figure approaching the combined free cash flow of Apple and Microsoft.

When a company can buy back 2-3% of its shares every year while still investing billions in next-generation technology, the per-share economics start compounding in your favor without you lifting a finger.

Even If It Doubles, the Math Still Humiliates

This is the headline thesis, so let's stress-test it.

Imagine Nvidia's stock doubles from here, call it $430 per share. At that price, with FY2027 consensus EPS estimates of roughly $8.17, you'd be paying about 52 times forward earnings.

Is 52x expensive? For a company growing revenue at 60-70%? For a business whose CEO just guided to $1 trillion in cumulative Blackwell and Rubin revenue through 2027?

Zoom out one more year. Consensus FY2028 EPS sits around $10.77, and some analysts believe the real number could hit $14-16 if Blackwell and Rubin ramp as expected.

At $430 a share,  double today's price , that's under 31 times FY2028 earnings. For a company with 60%+ operating margins and a near-monopoly on the infrastructure layer of the most important technological shift since the internet.

Let me say that again, slowly: even if the stock doubles, you'd still be paying a below-average multiple for above-average growth.

That's not investing. That's theft.

The $1 Trillion Tailwind Nobody's Priced In

At Nvidia's 2026 GTC conference, CEO Jensen Huang did something he rarely does: he gave Wall Street a number.

Combined revenue from the Blackwell and Vera Rubin platforms, he said, would reach at least $1 trillion through the end of 2027.

Not "could reach." Not "might approach." At least $1 trillion.

That forecast doesn't include Hopper GPU revenue (mostly recognized in 2025), nor does it include Rubin Ultra revenue (likely arriving in late 2027). It doesn't include networking revenue, which grew 263% inside the data center segment. It doesn't include the $30 billion in sovereign AI revenue generated in fiscal 2026 alone.

And here's the kicker: Bank of America recently raised its AI data center systems TAM estimate to $1.7 trillion by 2030, with a 45% compound annual growth rate.

Nvidia commands roughly 86% of the AI accelerator market. Even if that share drifts down to 70-75% over the next five years, and it might, the absolute revenue pie is growing so fast that the per-share earnings trajectory barely flinches.

"But What About Competition?", The Moat That Won't Drain

I hear this objection constantly: "Amazon has Trainium. Google has TPUs. AMD is coming. The moat is weakening."

Let me give you the counterargument that nobody's talking about at dinner parties.

Nvidia's CUDA ecosystem has over 4 million developers locked in. Swapping out Nvidia hardware means rewriting millions of lines of code, retraining engineering teams, and sacrificing years of performance optimization.

As one analysis put it: each layer of Nvidia's stack creates independent switching costs that multiply together. Displacing Nvidia requires winning at every layer simultaneously, hardware, software, networking, and inference optimization.

Could a competitor eventually chip away at this? Of course. Markets evolve. But we're talking about years, not quarters, of meaningful erosion. And in those years, Nvidia will have generated hundreds of billions in free cash flow and bought back a meaningful chunk of its own stock.

The moat isn't invincible. But it's far deeper than most people appreciate.

How Nvidia Stacks Up (The Table That Makes AMD Investors Nervous)

Context is everything. Here's how Nvidia's valuation compares to peers:

Sources: Yahoo Finance, Stock Analysis, Trefis, CMoney.

AMD trades at nearly six times Nvidia's forward P/E multiple despite growing at half the rate. Intel's forward P/E is 156 because its earnings base has collapsed, it's not cheap, it's broken.

This isn't a contest. It's a category error.

There's a simple reason Nvidia trades at a discount: the market is pricing in a slowdown that hasn't arrived, and may not arrive for years. Every quarter that Nvidia beats expectations while the multiple stays compressed, the value gap widens.

The Elephant in the Room, 3 Risks Worth Watching

I'm not here to sell you a fairy tale. Every investment has risks, and Nvidia's are real. Here are the three that keep me up at night, and why I'm still buying.

Risk #1: Two Customers = 36% of Revenue

According to Nvidia's own filings, just two customers accounted for 36% of FY2026 revenue. If one of those hyperscalers, likely Microsoft or Amazon, decides to shift more spending to in-house chips, the revenue hit could be material.

Why I'm not panicking: Those same hyperscalers are projected to spend $650-700 billion in CapEx this year, and Nvidia remains the only supplier that can deliver AI compute at scale. Custom silicon (Trainium, TPU) is growing, but it's additive to the market, not entirely replacing Nvidia's stack.

Risk #2: The China Card Is Still Off the Table

Export restrictions have decimated Nvidia's Chinese business. Once commanding a dominant share, Nvidia's position in China's AI chip market has shrunk to roughly 55%, and some projections see it falling further.

Why I'm not panicking: CEO Jensen Huang has stated that Nvidia's Q1 FY2027 guidance of roughly $78.6 billion in revenue already excludes virtually all China data center GPU sales. China, at this point, is an option , not a requirement for the bull case.

Risk #3: What If They Stop Spending?

The biggest bear argument: AI CapEx is a bubble. When hyperscalers realize they've overbuilt, spending collapses, and Nvidia's growth stalls.

Why I'm not panicking: The AI ecosystem is still compute-constrained, not compute-saturated. Every major cloud vendor is capacity-limited, they can't deploy AI services fast enough to meet demand. Until that dynamic flips, the spending party continues.

The Cheapest AI Giant Nobody Believed Would Stay Cheap

Here's what I keep coming back to.

Nvidia closed Friday at $215.20. At that price, you're paying 26 times forward earnings for a company growing revenue at 65%, generating 60%+ operating margins, sitting on a $1 trillion order pipeline, and returning billions to shareholders through buybacks.

The market has been "waiting for the slowdown" for two years now. Every quarter, the slowdown doesn't show up. Every quarter, the earnings base grows faster than the stock price, and the multiple compresses further.

One day, the market will stop waiting. And when it does, the people who bought at 26 times forward earnings will look like geniuses.

Not because they timed the bottom. But because they did the math.