Microsoft pulled off a big one.

Microsoft pulled off a big move faster than expected.
They say they are cutting CapEx numbers (though Gemini explains actual spending remains fixed)...
I thought Meta would do it, but anyway... now that someone ignited the spark, over time they will competitively trim CapEx spending, right? Prices for semiconductors and memory might not drop immediately, but wouldn't they go down in about six months?
They say the devil is in the details, but since they said they are cutting it, the impact will likely be huge.
It seems Microsoft is cutting CapEx because they fell behind Google in this quarter's competition.
Amazon will likely lag behind Google in the AI competition too, so I wonder what they will say about CapEx.

It is said that roughly 30% of hyperscaler data center construction expenditure goes toward buying HBM and DRAM memory chips (whereas 2 to 3 years ago, memory prices were not expensive, so it accounted for less than 8% of construction costs).
https://www.deloitte.com/us/en/insights/industry/technology/why-memory-chip-crunch-is-greater-than-expected.html

Even if HBM prices do not change quarterly because of long-term contracts, DRAM memory prices are reportedly rising by 25% every quarter... So even if they kept CapEx fixed, it could be seen as reducing memory purchases by that much... but since they are actually reducing CapEx, well...

Anyway, regarding the critical interest rate variable, the current macro environment suggests that a rate hike is more likely than a drop for at least a year. The BOE freezing interest rates today is good news for stocks.

Microsoft lowered its calendar 2026 Capex forecast to 175 billion (from190 billion). This was an accounting adjustment achieved by expanding data center useful lifespans to 25 years, which reassured investors looking for capital efficiency.

Microsoft’s lower reported CapEx guidance—dropping from 190 billion to175 billion for calendar year 2026—is highly significant because it is driven by an accounting adjustment rather than a reduction in physical AI infrastructure investment, effectively easing Wall Street's fears of runaway spending without compromising growth.

By reclassifying data center asset lifespans, Microsoft satisfied investor demands for spending discipline. This triggered a major relief rally, lifting MSFT stock by up to 8.7% following its earnings report.

The Optical "Cut": Accounting Reclassification
The Numbers: Reported full-year guidance shifted down from 190 billion to175 billion.
The Reality: Microsoft extended the estimated useful life of its data centers and office buildings from 15 to 25 years.
The Impact: This reclassification shifts future data center leases from "finance leases" (which count as CapEx) to "operating leases" (which do not).
The Bottom Line: Microsoft is not spending a single dollar less on servers, AI chips, or infrastructure. Physical deployment remains exactly on track.

Key Financial Implications

• Relieving "AI Fatigue" and Valuation Pressures

Prior to the announcement, the market was heavily penalizing Big Tech hyperscalers for uncontrolled AI spending. For instance, Alphabet’s stock fell 7% after it raised its CapEx forecast to $205 billion. Microsoft’s optically lower number signaled spending discipline, reassuring investors that the company is not engaging in an unconstrained cash-burn race.

• Protecting Free Cash Flow (FCF) Margins

Massive capital expenditures drastically compress short-term free cash flow. By holding the line on its underlying spending run-rate and reclassifying these assets, Microsoft signaled that it expects to remain free cash flow positive into fiscal 2027. This stands in stark contrast to competitors like Meta, whose massive CapEx expansion caused quarterly free cash flow to plummet.

• Proving High Demand Match (No Speculative Glut)

Microsoft minimized fears of a "speculative AI bubble" by matching its current build-out directly to concrete enterprise revenue.
Azure Expansion: Azure cloud revenue accelerated by 43%, proving that capacity is being monetized immediately upon coming online.
Commercial Backlog: Microsoft boasts a robust $627 billion commercial backlog, meaning its capital deployment behaves like a factory expansion for pre-ordered demand rather than a risky gamble.

That is a fair point, and comparing the two reveals exactly why the stock market reacted so differently to Alphabet (Google) and Microsoft.
While Google posted a 24% revenue surge and an 82% explosion in Cloud revenue, Google's stock crashed because it aggressively escalated its CapEx forecast, whereas Microsoft’s rally was triggered because its CapEx did not escalate further.
The contrast between Google's earnings drop on July 22 and Microsoft’s spike today comes down to three structural differences in how their numbers were delivered.

The 3 Reasons Why Google Dropped and Microsoft Rose

• The CapEx Trajectory (Surprise Hikes vs. Stability)

The market did not punish Google for its earnings; it punished Google for accelerating its cash burn unexpectedly.
Google's Surprise Hike: Google shocked the market by hiking its full-year 2026 CapEx outlook upward to a staggering 195 billion –205 billion (up from 180B–190B just three months prior).
Microsoft's Stability: Microsoft did the exact opposite. It stabilized its forecast, capping calendar 2026 CapEx at 175 billion (down from190 billion). Even though it was an accounting shift, it gave investors peace of mind that spending was peaking, not spiraling out of control.

• Free Cash Flow Shock

Google’s massive quarterly infrastructure spend of $44.9 billion (a 100% year-over-year increase) actually caused its free cash flow to turn negative for the first time since 2004. Microsoft managed to grow its revenues and maintain positive, highly robust cash flows, proving to Wall Street that its margins are better protected against heavy AI investments.

• "Phantom" Earnings vs. Core Product Strength

While Google's net income looked massive on paper, the market immediately saw through the top-line numbers:
Google's Artificial EPS Boost: Google’s headline EPS was heavily inflated by a massive, one-time $98 billion paper gain from its equity stakes in Anthropic and SpaceX. The market heavily discounts non-operating, unrealized investment gains.
Microsoft's Organic Beat: Microsoft’s earnings beat was purely operational, driven directly by enterprise demand for its software, Copilot commercial scaling, and Azure cloud infrastructure.

Key Financial Comparison (Most Recent Quarters)

Metric: Alphabet (Q2 2026) / Microsoft (Q4 FY2026) / The Market's Interpretation
Revenue Growth: +24% (Beat) / +18% (Beat) / Google grew faster, but it was already priced into its stock.
Cloud Growth: +82% / +43% / Both infrastructure segments are booming.
2026 CapEx Outlook: Raised to 195B–205B / Lowered to $175B / The Pivot Point: Google's costs are accelerating; Microsoft's are capping.
Stock Reaction: Dropped ~7% / Surged ~16.7% / Fear of margin compression vs. Relief on capital discipline.