Amazon's AWS Blowout and Apple's Guidance Miss Split the Tape as Oil Tests the Fed's Patience

July ended the way it lived: violently. The S&P 500 climbed 0.7% to 7,489.72, the Dow added 0.5% to 52,485.03, and the Nasdaq composite jumped 1% to 25,373.85 on Friday^[1]. But the index-level calm masked a violent rotation underneath. Amazon surged 15.3% to $271.58^[2] while Apple plunged 7.4% to $308.91^[2] — a 22-point spread between two of the world’s largest companies on the same day. The Russell 2000 fell 0.5%^[1], confirming that the AI-driven surge is sidestepping small caps entirely.

Amazon: The AI Capex Thesis Gets Validation

Amazon’s Q2 results were the cleanest positive signal of the week. Net sales rose 20% to $200.6 billion, operating income jumped 43% to $27.5 billion, and AWS revenue grew 37% — its fastest pace in 18 quarters — to a $169 billion annualized run rate^[3]. CEO Andy Jassy framed the acceleration as AI and core services driving each other’s growth^[3].

But the capex number tells its own story: Amazon now expects capital spending to reach $220 billion this year^[3], a figure that reflects both the scale of the AI buildout and rising memory-chip costs. The trajectory here matters. If AWS growth continues to accelerate above 35%, the capex is self-funding — operating income at $27.5 billion already covers a meaningful portion. If growth decelerates back toward 30%, the $220 billion becomes a margin question. I’d put the odds of AWS sustaining above 35% growth through year-end at roughly 60/40, with the 40% case hinging on whether enterprise AI workloads continue scaling at their current pace.

Apple: The Memory Flood Warning

Apple’s quarter was a paradox — a beat that the market treated as a miss. Revenue hit a June-quarter record of $109.4 billion, up 16% year over year, with iPhone sales surging 22%^[4]. But guidance for the current period came in weak, citing “supply constraints”^[4], and China plus Services fell short of expectations^[4].

The more striking detail was Tim Cook reportedly signing off his final earnings call with a warning about a “100-year flood” in memory chip pricing^[5]. That warning was not abstract. Micron fell 5.9% to $823.03^[2], and the materials sector ETF (XLB) dropped 2.3%^[6] — the worst-performing sector ETF on the day. If memory prices are indeed flooding the cost side of the semiconductor supply chain, Apple’s margin compression is a leading indicator, not a one-off.

Oil, Bonds, and the Fed’s Hawkish Three

While equities rotated, the bond market was doing something more structural. The 10-year Treasury yield rose to 4.73% on Friday, its highest level since January 2025^[7] — and according to one analysis, a 19-year high^[7]. The FRED series confirms the move: 4.67% as of the July 29 FOMC meeting^[8], up from 4.61% just one trading day earlier^[8]. The VIX climbed to 20.66^[8], up from 16.64 the prior week^[8] — a 24% jump in five sessions that reads as genuine risk repricing, not noise.

The driver is oil. Brent crude surged above $100 per barrel in late July^[5] after Houthi attacks on Saudi tankers in the Red Sea combined with Trump administration threats of military retaliation against Iran^[5]. On Friday, prices rose further after Iran said it had stopped ships in the Strait of Hormuz^[5]. Oil rallied 22% in July^[5], and the energy sector ETF (XLE) gained 1.0% on the day^[6].

The Fed met July 28-29 and voted 9-3 to hold the federal funds rate at 3.5-3.75%^[9]. The three dissenters — Cleveland’s Hammack, Minneapolis’s Kashkari, and Dallas’s Logan — all argued for a hike^[9]. Chair Warsh avoided forward guidance but said he would not be constrained by market pricing^[10]. Markets interpreted the press conference as a “hawkish tilt” setting up for a September hike^[11].

This is the tension that cannot be resolved with base rates alone. CPI inflation is 3.46%^[12], well above the 2% target, and oil at $100 makes a July CPI re-acceleration plausible. But consumer sentiment sits at 49.5^[12], down 18% year over year — a level historically associated with growth slowdowns, not overheating. Real GDP is growing at 2.1%^[12], and unemployment is 4.2%^[12]. The FRED analog search flags mid-2006 as the closest historical match^[12] — a period where the Fed kept hiking into a slowing economy. The parallel is not comforting.

The Two-Speed Tape

All ETF moves as of 16:00 ET close, July 31, 2026^[6].

The pattern is clear: Amazon-adjacent sectors (discretionary, communications) rallied. Apple-adjacent sectors (technology, materials) sold off. Rate-sensitive sectors (real estate, utilities) weakened as the 10Y climbed. The market is not moving as one — it is sorting winners from losers in real time.

What to Watch Next

• Jobs report (first week of August): Payroll and business survey data are the next major data points^[11]. A hot print cements the September hike case; a soft one gives Warsh room to stay on hold.
• Oil and Hormuz: Any escalation in the Strait of Hormuz standoff pushes Brent higher and inflation risk into the September FOMC. De-escalation is the single biggest downside risk to the inflation narrative.
• AWS growth trajectory: Amazon’s next quarter will test whether 37% AWS growth is a sustained inflection or a one-quarter peak. The $220 billion capex commitment makes this the most expensive inflection point to get wrong.
• Memory-chip pricing: Tim Cook’s “100-year flood” warning and Micron’s 5.9% decline are the same story. If memory costs keep rising, Apple’s margin guidance and the entire semiconductor supply chain face a cost-driven squeeze that the market has not fully priced.
• September FOMC: The 9-3 vote and Warsh’s avoidance of forward guidance leave the policy path deliberately ambiguous. The market’s job is to read the data, not the Fed — which is exactly how Warsh said he wants it^[10].