AMZN Stock Is Quietly Hitting Record Highs — Here’s What Wall Street Isn’t Telling You

Early in 2026, Amazon’s stock appeared to be a question mark for the first time in a long time. The price of shares fell below $200. The story of the Magnificent Seven seemed more unreliable. Hedges were added by analysts who had been mindlessly typing “buy” for years. Then it turned around, almost silently. This week, AMZN surpassed $249. Just over $258 is the 52-week high. The stock has increased by nearly 26% over the last three weeks, and the tone of the investor notes has changed from anxious to more impressed.

The price itself isn’t what’s intriguing. It’s the force behind it. The obvious narrative, which includes advertising, AWS, and e-commerce, only goes so far. A $200 billion capital expenditure plan, a $11.6 billion satellite acquisition that no one anticipated two months ago, and a CEO whose yearly letter succeeded in reassuring people without overselling them—something Jeff Bezos had not consistently done in his later years—are all part of the larger picture.

Just the $200 billion figure is worth pondering for a moment. That is almost twice as much as what Amazon spent on capital expenditures in 2024. It exceeds the entire federal budget of some mid-sized nations. This week, Amazon revealed that it would be investing an additional $12 billion in two new data center campuses in Mississippi, bringing its total investment in that state to $25 billion. The plants will be located next to power infrastructure that Amazon is directly funding, which is, if nothing else, a very contemporary business practice. These days, the tech giants construct their own electrical grids. Robber barons laying rail has an almost 19th-century feel to it, but the iron is silicon and the rail is fiber.

According to Alex Carchidi of Motley Fool, the bull case has three legs. E-commerce is subtly evolving from a volume story to a margin story; last quarter, North American operating income increased by 24% despite a mere 10% increase in revenue. Investors in Walmart and Costco would exchange their organs for that level of operating leverage. AWS recently reported its fastest growth in thirteen quarters, at 24% year over year, after trailing Microsoft Azure for a few years. Additionally, the revenue run-rate for the custom chip business, which includes Trainium and Graviton, is currently $20 billion, with internal use pushing the effective figure closer to $50 billion. You can learn more about Amazon’s AI posture from the Anthropic data center, which opened late last year, than from any press release because it runs solely on Amazon’s own silicon.

Then there is the Globalstar deal, which was announced on April 14 with surprisingly little fanfare for a deal of this magnitude. With an additional $6.2 billion in stock and $4.6 billion in cash, Amazon now owns the LEO satellite provider that powers emergency satellite messaging on all iPhone models 14 and later. Apple continues to be a partner. Amazon’s Leo broadband network incorporates the satellites. Amazon intends to launch its own direct-to-device satellite system by 2028, at which point it will directly compete with Starlink for a market that Elon Musk has been controlling for the better part of ten years. The move’s lack of subtlety is difficult to ignore. To catch Musk, Amazon didn’t want to start from scratch with a satellite network. The shortcut was purchased.

The bear case is still there. Lately, it has simply been drowned out. This year, Amazon will have negative free cash flow, which is an odd statement to make about one of the most profitable companies in history. The amount of debt is increasing. The customer base that Amazon depends on is still being squeezed by tariffs and gas prices, and consumer spending is still erratic. Furthermore, the claim that Google’s TPUs can be surpassed by AWS’s Trainium chips is, at best, only partially supported. Anyone dismissing these legitimate worries is most likely trying to sell something. Additionally, regardless of how convincingly the cloud story is accelerating, AMZN doesn’t appear cheap on a P/E basis at 35 times trailing earnings.

However, the more intriguing perspective, which Amy Liu proposed in a piece that went viral this week, is that P/E might be completely inappropriate for a business that is in the midst of its biggest reinvestment cycle to date. In 2025, operating cash flow increased by 20% to $139.5 billion. The price-to-operating-cash-flow ratio is close to 19 times, which is far less concerning than what the headline multiple implies. The current spending binge will eventually cause depreciation to slow. Presumably, revenue from the current investments—the Globalstar satellites, the Anthropic partnership, and the Mississippi data centers—will begin to appear on the earnings line. When it occurs, the question is whether investors will still be interested.

It’s difficult to avoid the impression that AMZN stock is in one of those stages where the story is actually bigger than the numbers can support at any given time. The company is making three or four large bets at the same time, none of which will be resolved for years and any one of which could support its market capitalization on its own. It’s obvious that Andy Jassy is now playing offense after initially appearing happy to just keep the machine Bezos left him intact. It’s still unclear if the expenditures will be profitable. The fact that this stock will be much more interesting to watch over the next 12 months than it has been over the previous 12 is less uncertain.