If I could hold only one AI (artificial intelligence) stock for the next 10 years, I wouldn’t reach for the market’s hottest name. A decade is long enough for at least one brutal downturn and a wave of competition nobody sees coming.
What survives all of that, arguably, isn’t momentum. It’s business quality — a durable competitive advantage, heavy cash generation, and demand that outlasts the boom phase.
My pick is Microsoft (MSFT +0.15%).
That answer may sound boring next to chipmakers posting triple-digit growth rates. But boring is part of the appeal. Here’s why I’d be comfortable holding the software giant for the next decade.
Image source: Getty Images.
Why Microsoft fits the 10-year test
Microsoft‘s competitive advantage is entrenchment. Its software runs the workday at most large businesses (Office, Windows, Teams, and the developer and security tools around them), and its Azure cloud platform is where many of those same customers already store data and run applications. Switching away is expensive and disruptive, so customers rarely do it.
A durable advantage in enterprise software isn’t necessarily a better product. It’s a customer base that has organized itself around yours.
That entrenchment is now pulling AI demand toward the company. In Microsoft’s fiscal third quarter (the period ended March 31, 2026), its “Azure and other cloud services” revenue grew 40% year over year.
And that’s no one-quarter spike. Azure grew 40% in fiscal Q1 and 39% in fiscal Q2, so the business has held steady at about 40% growth for three straight quarters — remarkable consistency for a business of this scale. Enterprise adoption of AI is arguably still early, too. And when businesses do adopt, Microsoft is often the first vendor they try, because it’s the vendor they already pay.
The forward-looking signal is even stronger. Microsoft’s commercial remaining performance obligations (contracted work it hasn’t yet recognized as revenue) roughly doubled year over year to $627 billion in fiscal Q3. And CEO Satya Nadella said on the company’s fiscal third-quarter earnings call that Microsoft’s AI business surpassed an annual revenue run rate of $37 billion, up 123% year over year.
Then there’s the cash. Fiscal third-quarter revenue rose 18% year over year to $82.9 billion, net income climbed 23% to $31.8 billion, and earnings per share grew 23% to $4.27. That profit engine lets Microsoft fund one of the largest AI build-outs in the world from its own operations while still paying a dividend.
Today’s Change
(0.15%) $0.58
Current Price
$384.94
Key Data Points
Market Cap
Day’s Range
$381.53 – $391.91
52wk Range
$349.20 – $555.45
Volume
796.8K
Avg Vol
40.1M
Gross Margin
68.31%
Dividend Yield
0.92%
The price of admission
So why is a stock this good trading about 30% below its 52-week high of $555.45?
In a word, spending.
Chief financial officer Amy Hood said on the fiscal third-quarter call that Microsoft expects to invest about $190 billion in capital expenditures in calendar year 2026, including about $25 billion from the impact of higher component pricing. That plan is likely the biggest reason the stock still sits well below its high: Investors want proof that the build-out’s returns will justify its cost. And as those investments hit the income statement as depreciation in the coming years, they could weigh on margins.
That concern is worth taking seriously. But it’s also why the valuation finally looks reasonable. At about $385 per share as of this writing, Microsoft trades at about 23 times earnings and about 20 times forward earnings. Nvidia, the obvious alternative, trades at about 23 times forward earnings versus Microsoft’s roughly 20. That isn’t a huge gap, but Nvidia’s earnings are far more exposed to swings in chip demand. For a decade-long holding, I’d rather own the company that buys the chips, rents them out through Azure, and keeps the customer relationship either way.
Could Microsoft stumble? Of course.
Competition in cloud computing is fierce, with Amazon and Alphabet investing just as aggressively. And if AI demand pauses, a $190 billion spending spree would look painful for a few years. Of course, Microsoft’s growth could also simply slow as its numbers get bigger, making today’s premium harder to defend.
But a 10-year holding isn’t a bet on the next few quarters. It’s a bet on who will still own the customer relationships and the infrastructure behind them 10 years from now. On those criteria, I think Microsoft is the safest way to own the AI era — and at today’s price, I think it’s a buy-and-hold for the next decade.
