Valuation caution returns

The mood around big technology stocks changed in one week from “how big can this get” to “what exactly are you paying for.” Apple was trading around 32 times earnings in late March 2026, above its own 10-year average of 25 and above the Standard and Poor’s 500 index near 24. (parameter.io) Apple is not being questioned because the business broke. Apple reported $143.8 billion in fiscal first-quarter 2026 revenue on January 29, 2026, up 16% year over year, and said operating cash flow was nearly $54 billion in the quarter. (apple.com) That is what makes the debate sharper. When a company already posts record revenue, record earnings per share of $2.84, and returns almost $32 billion to shareholders in one quarter, investors stop arguing about survival and start arguing about how much future perfection is already in the stock price. (apple.com) NVIDIA is the same story with bigger numbers and higher expectations. NVIDIA reported fiscal 2026 revenue of $215.9 billion on February 25, 2026, up 65% from the prior year, after fourth-quarter revenue hit $68.1 billion and data center revenue reached $62.3 billion. (nvidia.com) Jensen Huang then raised the temperature again at the March 2026 Graphics Technology Conference by talking about roughly $1 trillion in combined Blackwell and Vera Rubin sales through 2027. That kind of number makes investors picture a runway so long that valuation stops feeling urgent. (finance.yahoo.com) Then the timeline got messier. TrendForce said on April 8, 2026 that Vera Rubin shipment share could fall to 22% from 29% because of High Bandwidth Memory 4 validation, network changes from ConnectX-8 to ConnectX-9, higher power demands, and more advanced liquid cooling requirements. (trendforce.com) That delay did not kill demand. TrendForce said Blackwell could rise to more than 70% of NVIDIA’s high-end graphics processing unit shipments in 2026, up from 61%, because customers still need chips now and Blackwell is the more mature platform. (trendforce.com) But that is exactly why valuation discipline is back. If Blackwell carries more of 2026 while Rubin slips, investors have to separate “the business is still strong” from “the next leap arrives on the original schedule,” and those are not the same thing. (trendforce.com) The new test for both Apple and NVIDIA is less about story quality and more about proof. Apple has to keep turning its premium brand and services base into repeatable cash flow, and NVIDIA has to keep turning artificial intelligence demand into shipped systems, steady margins, and on-time platform transitions. (apple.com) (nvidia.com) (trendforce.com) That is the shift now showing up in analyst notes and investing commentary. Great companies are still being treated like great companies, but a great company at a stretched price is no longer being treated like a bargain just because the story is exciting. (parameter.io) (msn.com)