The 3 Biggest Bargain "Magnificent Seven" Stocks to Buy Today

This looks less like a fresh fundamental re-rating and more like the market rediscovering duration: the group’s valuation reset is mostly a discount-rate story, not a collapse in business quality. That matters because the cheapest names are the ones with the most operating leverage to even modest estimate upgrades over the next 1-2 quarters; if AI capex holds, the “cheap” label can disappear quickly without the stocks needing heroic growth.

NVDA is the cleanest expression of that setup because the market is effectively pricing a future where hyperscaler spend normalizes far earlier than management teams imply. The second-order risk is not competition from any single chip vendor; it is customer concentration and capex fatigue, which would hit the stock through forward revisions before it shows up in reported revenue. In contrast, META’s AI spend is more likely to pay for itself inside the ad engine before it becomes a balance-sheet problem, making it a better near-term margin expansion story than a pure AI infrastructure story.

MSFT is the defensive compounder in the group: the real risk is not “AI disruption” to Office, but whether AI monetization remains incremental enough to offset rising infrastructure intensity in Azure. The market is underweighting how sticky enterprise workflows are, but it is also assuming Copilot can keep lifting ARPU without forcing meaningful pricing resistance. The contrarian view is that the crowd is focusing on cheap multiples while ignoring that these are still long-duration assets; if 10Y yields back up, the rerating thesis breaks first, and if AI spending pauses, NVDA is the fastest to de-rate.