Amazon's Cloud Blowout Splits Big Tech's AI Capex Story

The week ending July 31, 2026 drew a line through Big Tech. On one side: Amazon, Microsoft, and Alphabet, each posting cloud growth that validated the hundreds of billions poured into AI infrastructure. On the other: Apple, whose chip-supply warning cratered the stock 7.4%, and Meta, whose AI spend swallowed nearly all free cash flow. The S&P 500 closed Friday at $747.03 (SPY), up 0.72% on the session and roughly 1% for the week^[1]. The Nasdaq 100 (QQQ) gained 0.65% to $687.99, while the Russell 2000 (IWM) slipped 0.48% to $291.20 — small caps did not get the memo^[1].

The headline indexes were green, but the composition told a sharper story. The S&P 500 rose about 1.03% for the week, the Nasdaq Composite climbed 1.77%, the Dow added 0.53%, and the Russell 2000 fell 0.57%^[2]. Wednesday’s FOMC meeting marked the week’s low close across all four indexes; from there, earnings-powered rallies in the hyperscalers pulled the tape higher over Thursday and Friday^[2].

The hyperscaler verdict: cloud revenue is catching up to capex

Amazon delivered the week’s standout print. AWS revenue grew 37% year over year — its fastest growth in 18 quarters — reaching a $169 billion annualized run rate^[3]. Net sales for the company rose 20% to $200.6 billion, and operating income jumped 43% to $27.5 billion^[3]. CEO Andy Jassy framed the AWS acceleration as AI and core services reinforcing each other, and the market agreed: AMZN closed Friday at $271.58, up 15.32% on the day — its largest single-session gain since 2012^[4][3]. The company also raised its 2026 capital expenditure guidance to $220 billion, citing higher memory costs, and investors rewarded rather than punished the spend^[3].

Microsoft’s report reinforced the same thesis. Revenue reached $90 billion, up 18%, with operating income of $40.6 billion and net income up 31% on a GAAP basis^[5]. Azure strength drove the beat, and the company issued healthy forward guidance with more cloud momentum expected^[5]. MSFT closed at $464.72, up 3.02%^[4]. Unlike peers whose capex alarms investors, Microsoft appeared to have an answer for the return-on-investment question^[5].

Alphabet, reporting a week earlier on July 22, set the template. Revenue grew 24% to $119.8 billion, with Google Cloud surging 82% to $24.8 billion^[6]. That was its 12th consecutive quarter of double-digit revenue growth^[6]. GOOGL closed Friday at $356.13, up 6.73%^[4].

The common thread: companies that can point to accelerating cloud revenue — where AI workloads translate into billed compute — are getting credit for their infrastructure spending. The market is distinguishing between capex that generates visible top-line growth and capex that does not.

The cost side: Apple’s supply crunch and Meta’s cash-flow squeeze

Apple’s quarter exposed a different set of pressures. The company posted record June-quarter revenue of $109.4 billion, up 16% year over year, with iPhone sales rising 22%^[7]. EPS of $2.02 beat estimates and included a favorable impact from tariff refunds^[7]. But management issued below-consensus guidance for the current period, citing “supply constraints” — specifically chip shortages and rising memory costs that forced Apple to raise prices in June^[7]. The stock slid more than 6% in extended trading on July 30 and closed Friday at $308.91, down 7.35%^[4][7].

The supply-chain problem is not hypothetical. Apple has been struggling to secure enough processors all year, and the earnings call made clear that Mac shortages will worsen^[7]. For a company that has been a tariff-refund beneficiary, the guidance cut raised a question the bull case had avoided: whether hardware-driven growth can survive a memory-cost cycle that shows no sign of easing.

Meta told a related but distinct story. The company beat on revenue but missed on earnings as costs surged 55%, and profit fell 14%^[8]. Free cash flow narrowed to the point where the AI investment bill consumed nearly all of it^[8]. Meta narrowed its capex guidance range but kept the top end intact^[8]. The stock initially dropped more than 8% in after-hours trading on July 29 before recovering through the week to close at $556.71, up 3.28% on Friday^[4][8].

The Meta recovery is worth pausing on. What would have to be true for the market to reward a company whose profits are falling and whose free cash flow is evaporating? The answer is that investors are pricing the option value of AI-driven enterprise revenue — the same thesis Amazon and Microsoft have already validated in their cloud numbers. Meta has not yet demonstrated that monetization at scale, but the week’s recovery suggests the market is willing to wait, given the hyperscaler precedent.

The Fed: a divided hold

The FOMC voted 9-3 on July 29 to hold the federal funds rate in a 3.5%–3.75% range, the fifth consecutive meeting without a change^[9]. Three regional presidents — Beth Hammack (Cleveland), Neel Kashkari (Minneapolis), and Lorie Logan (Dallas) — dissented in favor of a hike, an unusually large split that underscores the committee’s internal tension^[9].

The macro backdrop explains the discord. CPI inflation stands at 3.46% year over year, above the Fed’s 2% target for more than five years^[10]. The 10-year Treasury yield sits at 4.68%, up 24 basis points month over month^[10]. Real GDP growth is 2.1% year over year — solid but not booming^[10]. Unemployment is 4.2%, and consumer sentiment sits at 49.5, down 18.45% year over year^[10]. The high-yield credit spread is 2.84%, tight by historical standards and not signaling imminent stress^[10].

The VIX closed the period at 20.66, up 29% year over year but down from midweek levels, consistent with the volatility-receding narrative as earnings clarity arrived^[10][2].

Chairman Kevin Warsh has argued for giving markets fewer forward signals, a stance that leaves the data — not the communication — as the primary driver of rate expectations^[9]. Fed-funds futures show a 64% probability of a rate hike in September, according to one report^[2]. That probability is not high enough to call the next move settled, and the August 7 payrolls report is the next data point that will reshape it.

The divergence checklist

What to watch next

July payrolls (Friday, August 7). June’s report was weak — just 57,000 jobs added against expectations for 115,000, with the unemployment rate’s dip driven by people leaving the labor force^[11]. Economists expect July payrolls to rebound to about 120,000 and the unemployment rate to tick up to 4.3%^[11]. A print near or below June’s level would reinforce the slowdown narrative and complicate the Fed’s inflation-fighting posture. A strong rebound would give the hawks more ammunition.

The AI capex bifurcation. The market has drawn a clear line between hyperscalers with cloud revenue proof points and companies whose AI spending is not yet offset by monetization. Next week’s remaining earnings — and any commentary on forward cloud demand — will test whether the split holds or narrows.

Small-cap participation. The Russell 2000’s weekly decline while the S&P and Nasdaq rose is a breadth concern worth monitoring. If the rally remains concentrated in mega-cap tech, the index-level gains mask a narrower advance than the headlines suggest.

Yield direction. The 10-year Treasury at 4.68% with a positive 10-2s curve of 47 basis points reflects a regime that is neither recession-pricing nor aggressively hawkish^[10]. Whether yields push higher — which would pressure equity multiples — depends heavily on the payrolls print and the next CPI reading.

The base-rate read is that markets entering a period of decelerating but positive growth, with inflation above target and a divided central bank, tend to stay range-bound with sector-level dispersion dominating index-level direction. This week was a clean example of that pattern. Whether next week’s jobs data disrupts or confirms it is the question the tape is waiting to answer.