TL;DR
The Fed's September hike resolved the biggest open question from last time — unanimously, hawkishly, and with an open-ended path that prices in more hikes through 2027. Oil's risk premium is deflating as Saudi supply adapts and diplomacy re-emerges, which is exactly what let AI and chip names lead the post-FOMC rebound. Meanwhile, the two "mechanical" catalysts everyone watched — the Nasdaq-100 rebalance and the memory supercycle — both delivered, but in ways that punish front-runners and reward reading the contracting structure.
The Warsh Fed Hikes, and Refuses to Tell You Where It Stops
The Federal Reserve's first hike since 2023 came with a framing that matters more than the move itself: policy is officially still stimulative, and the number of remaining hikes is open-ended.
"Today, the FOMC decided that this standard has not been satisfied." — Kevin Warsh's Fed Delivers First Hike Since 2023
(via CNBC)
The September 16 FOMC voted 12-0 to raise the funds rate 25bp to 3.75%–4.00%, with 16 of 18 participants expecting another hike this year and first cuts not penciled in until 2028. The phrase that repriced everything was Warsh's characterization of the hike as removing merely "a dose of accommodation" — Evercore ISI's Krishna Guha called it "a deliberate choice" that "raises the possibility of a more open-ended approach to the number of hikes" (Kevin Warsh's Fed Delivers First Hike). Goldman Sachs added an October hike to its forecast, BofA expects October plus December, and CME FedWatch put October odds near 58% by Friday. For a self-directed investor, the practical consequence is that rate-cut hopes are off the table for 2026–27, and valuation pressure now runs through the long end — the 10-year briefly topped 5% and the 30-year mortgage hit 7.19%.
What to watch: Whether the FOMC delivers a second consecutive hike at its late-October meeting, with the vote split and equity reaction as the tell.
Oil's Risk Premium Is Leaking, Not Collapsing
The energy shock that drove the Fed's hawkish turn is easing — but through supply adaptation and diplomacy, not because the conflict ended.
"Middle East oil transit 'remains more resilient than the market had previously feared.'" — Oil's Risk Premium Deflates
+3 (via TradingKey)
Brent fell a fourth consecutive session to $101.97 — the longest losing streak since June — after Saudi Arabia rerouted Red Sea exports through Hormuz, pushing its Hormuz shipments to roughly 2.9M bpd from about 700k in August. Qatar is mediating renewed U.S.–Iran talks, and the market is betting on diplomatic signals out of the UN General Assembly. This matters directly for equities: easing oil eases the energy-inflation pressure the Fed explicitly cited, which is one reason chips led the post-FOMC rebound. But this remains an active conflict with Houthi attacks continuing — the de-escalation is flow-based, and a diplomatic collapse would re-inflate both the oil premium and the Fed's path.
What to watch: A decisive daily close below $100 Brent, which would open the $94–$98 range and further soften the Fed's inflation narrative.
Index Mechanics Are Certain; Price Direction Is Not
The most-watched passive inflow event of the quarter arrived on schedule — and the stock still slipped, confirming that the "front-run the index cross" trade is now crowded and decaying.
"Significant front-running is already in motion." — SpaceX (SPCX) Passes the Nasdaq-100 Double-Weight Test
(via Investing.com)
The September 18 closing cross more than doubled SpaceX's Nasdaq-100 weight from ~1.28% to ~2.82% — above JPMorgan's 2.25% estimate — with Morgan Stanley pegging passive buying at ~$15.5B across ~$1.7T benchmarked to the index. Yet SPCX closed the cross day down 1.36% at $152.71, a partial repeat of July 7, when the stock fell nearly 6% on inclusion day as pre-positioned sellers used the price-insensitive bid to exit. That resolution closes the thread from last cycle's digest: mechanical flows are being absorbed, and from here the drivers are fundamental — the largest supply event on the calendar is more than 2.3 billion restricted shares exiting lockup in late October and mid-November, coinciding with Q3 earnings, against a trailing P/S near 106x with the 10-year near 5%.
What to watch: The late-October/mid-November lockup wave hitting the tape alongside Q3 earnings — the biggest single supply test the stock has faced.
The Memory Supercycle Is the Loudest AI ROI Signal — and It's Now Debt-Funded
The AI complex survived its Fed-week stress test, and the most concrete evidence the buildout is real comes from an unlikely place: memory pricing and contract structure.
"Memory scarcity today is being driven in large part by the AI buildout itself." — AI Capex Returns: Post-FOMC Rebound, Memory Supercycle Signals
(via BigGo Finance)
The signals are everywhere: Nvidia's CFO describes "extreme pricing conditions in memory" headed even higher into next year; SK Hynix doesn't expect supply-demand balance until 2030 at the earliest and approved ~$38.3B of capacity investment whose first new cleanrooms don't open until December 2028; and Micron has shifted to 16 take-or-pay agreements carrying ~$100B in minimum contracted revenue, backed by ~$22B in customer deposits (Micron Into the $50B Quarter+1). Micron reports September 30, guided to $50B ± $1B revenue and ~86% gross margin — yet trades at roughly 6.6x forward earnings versus ~21x for the S&P 500, and is ~22% off June's high. The debate is whether contracted revenue has converted a boom-bust commodity into something closer to a utility with a multi-year floor. The risk mirror: SoftBank just launched over $11B of bonds to fund its third $10B OpenAI tranche — leverage entering the AI funding stack is a classic late-cycle marker (AI Capex Returns
).
What to watch: Micron's September 30 print against the $50B/$31 guide, plus hyperscaler capex commentary and HBM4 pricing durability into 2027–28.
What surprised us
- AI stocks rallied after the hike. The day following a unanimous rate hike, the Nasdaq gained 1.69% and the Philadelphia Semi Index jumped 3.1% — AMD +6.4%, Micron +5.5% — as yields and oil dipped. The sensitivity chain (AI equities ← 10-year ← oil ← Fed path) is now explicit, and it cuts both ways (AI Capex Returns
- The index-inclusion trade has no edge left. Both of SpaceX's forced passive-buying events — July 7 and September 18 — saw the stock fall on the day, as front-runners used the price-insensitive bid to exit. When everyone knows the mechanical trade, the mechanics stop moving the price (SpaceX Passes the Double-Weight Test
- A buyer is the best evidence of a shortage. Nvidia guiding its own gross margin down to a 71–72% bottom on memory input costs is, paradoxically, the strongest confirmation that memory tightness is real — the world's most important AI company is absorbing the squeeze (Micron Into the $50B Quarter
- SpaceX surged on Fed day. While the indexes fell on September 16, SPCX rose 5.15% on Starship Flight 14 scheduling and confirmed Golden Dome and Starshield defense contracts — idiosyncratic catalysts trumping macro for once (SpaceX Passes the Double-Weight Test