Jun. 18 at 10:27 AM
$MSFT may be the clearest test of the entire AI investment cycle.
Not because demand is uncertain. Demand is already visible.
Microsoft just reported:
$82.9B in quarterly revenue, up 18%.
$54.5B in Microsoft Cloud revenue, up 29%.
Azure growth of 40%. An AI business now running above
$37B annually, growing 123% year over year.
Commercial backlog reached
$627B. The bullish case is easy to understand.
Microsoft owns the cloud infrastructure, enterprise distribution, developer ecosystem and productivity software needed to monetize AI across several layers at once.
Azure captures infrastructure spending. GitHub captures developers. Microsoft 365 Copilot captures enterprise users. The existing business helps finance the buildout. But the other side of the story matters too.
Microsoft Cloud gross margin has moved from 69% to 66% over the past five quarters, and management expects roughly 64% next quarter.
The company now expects around
$190B in calendar-year 2026 capital expenditures and still expects AI capacity to remain constrained through the end of the year.
That tells me the real Microsoft debate is no longer:
“Is AI demand real?”
It clearly is.
The question is whether Azure, Copilot and Microsoft’s first-party AI products can scale revenue faster than depreciation, infrastructure costs and component inflation scale with them.
At roughly 22.6x trailing earnings,
$MSFT is not being valued like a speculative AI startup.
But the market is still assuming that strong growth and operating discipline can coexist throughout one of the largest infrastructure buildouts in corporate history.
My view:
Microsoft has one of the strongest AI positions in the market because it owns both the infrastructure and the distribution.
But the next major re-rating may depend less on another AI announcement and more on proving the return generated by every new dollar of AI capex.
Would you rather own Microsoft during the buildout…
or wait until cloud margins stop compressing?
$MSFT $AMZN $GOOGL $NVDA
