Owning NVIDIA (NASDAQ:NVDA | NVDA Price Prediction) for the next decade is a defensible decision for a retirement portfolio that wants one piece of the AI buildout and never wants to think about it again, and the path from a $5.51 trillion market cap today to $10 trillion by 2030 runs through cash flows that already exist on the income statement.
The durability case
CUDA is the moat. Customers do not swap GPU platforms the way they swap laptop brands. They retrain entire engineering organizations around a software stack, and NVIDIA’s has roughly two decades of head start. The hardware cadence reinforces it. Blackwell Ultra is shipping in volume, and the Vera Rubin platform brings six new chips delivering up to a 10x reduction in inference token cost versus Blackwell. Multi-year customer commitments anchor the demand. Meta has committed to millions of Blackwell and Rubin GPUs, OpenAI to 10+ gigawatts, Anthropic to an initial 1 gigawatt, CoreWeave to 5 gigawatts of AI factories by 2030. Sovereign deals stack on top, including South Korea’s 250,000+ GPU commitment and projects in the UK, Germany, Saudi Arabia, and the UAE. That reads like a utility’s contract book.
The compounding case
NVIDIA pays a $0.01 quarterly dividend, which means the income story sits in the buyback. The company returned $41.1 billion to shareholders in FY2026 and still has $58.5 billion remaining under repurchase authorization. The funding comes from $96.575 billion in full-year free cash flow, growing 58.7% YoY. Forward P/E sits at 27x, which is what a mature compounder trades at. If Q1 FY2027 guidance of ~$78.0 billion in revenue tracks and the share count keeps shrinking, the arithmetic to $10 trillion does not demand a multiple rerating. It demands NVIDIA to keep doing exactly what it is doing.
Nvidia’s valuation can indeed balloon to $10 trillion if the cloud buildout continues to accelerate for two to three more years. If this is paired up with the euphoric rally returning, it could take much faster than you imagine. The median semiconductor business trades at over 36 times forward earnings. And for a company like Nvidia, trading at even 60x forward earnings is not significantly above fair value due to the hypergrowth.
I see a $10 trillion valuation by 2028, less than two years from now.
Why it “survives” the next downturn
The balance sheet is fortress-grade. Shareholders’ equity reached $157.3 billion at the end of Q4 FY2026, up 98.28% YoY, against $49.51 billion in total liabilities. Trailing operating margin ran at 65%. Return on equity was 101.5%. A company that converts roughly half of revenue into net income, and most of that into free cash flow, has the cushion to absorb hyperscaler CapEx pauses, gaming supply hiccups, and the inevitable wobble at TSMC. Cycles happen. Cash fortresses outlast them.
That said, I put quotation marks around survives because simply surviving does not mean it wont leave a bad impact on your portfolio. If cloud investments slow down, you can expect a 50-60% slide from here. And most importantly, you can expect the buying opportunity of a lifetime, because AI is here to stay.
Where the thesis underperforms
If hyperscaler AI spending genuinely pauses, perhaps because a year of inference ROI disappoints, NVDA will fall harder than the index. Beta is 2.244. Recent Reddit chatter flagging a 30% weekend drop in B200 Blackwell GPU rental prices feeds exactly that fear, and the Q1 FY2027 guide explicitly excludes any Data Center compute revenue from China. Neither moves the forever case. A bad year for hyperscaler CapEx does not retire CUDA, does not unwind the OpenAI gigawatt commitment, and does not change the multi-decade arc of compute demand.
The franchise rewards long-horizon ownership rather than tactical positioning.
Top Gaining Stocks
Top Losing Stocks