The Friday Wrap — July 31, 2026: a flat month and a correction, in the same four weeks
Archived exactly as it was sent. Prices, links and product details are as of July 31, 2026 and are not updated afterwards.
July ended with the index you happened to own deciding your entire month. SPY +0.03%. QQQ -6.57%. IWM -3.08%. A flat month and a correction, describing the same four weeks of the same market — 6.60 percentage points between the best and worst of three broad US indices.
Zoom out one level, though, and the ranking inverts. Year-to-date: IWM +18.79%, QQQ +12.26%, SPY +10.14%. The index that had the worst July is still ahead of the S&P for the year, and the small caps that have spent the last month underperforming lead everything. A bad month is not a broken year, and we're not going to write it as one.
The week itself was the compressed version of July. The Fed held for a fifth straight meeting with its dissents pointing toward hikes. The megacap AI-capex verdict landed and split the complex rather than lifting it. And crude gave back a large piece of a war premium it had spent the month building.
The Fed
On July 29 the FOMC held the federal funds target at 3-1/2 to 3-3/4 percent on a 9–3 vote. The direction of the dissent is the part worth keeping: Beth Hammack, Neel Kashkari and Lorie Logan each preferred to raise the range by a quarter point.
The statement's own language matters more than the hold. Activity is "expanding at a solid pace despite elevated uncertainty that owes, in part, to the conflict in the Middle East." Job gains "have kept pace with the workforce." And inflation "remains elevated relative to the Committee's 2 percent goal, in part reflecting supply shocks that have driven price increases in certain sectors, including energy."
That last clause is the week's connective tissue: the committee explicitly routed energy into its inflation read, in a month when crude ran up roughly 22%. Futures currently price two additional 25bp hikes across 2026, and the next meeting is in September.
We don't carry FOMC odds, so no probability from us and no guess at September. The reportable fact is the direction the risk is priced — upward, not downward. That is a different rate backdrop for long-duration equity than a cutting cycle would be.
The AI capex verdict
This was the week the market graded AI spending, and it graded on conversion, not ambition. Every move below is a regular-session close from our broker feed, with the session named — after-hours prints and closing moves diverged widely this week and a lot of coverage mixed them.
Reported Wednesday night, traded Thursday: - Microsoft — FQ4 revenue $90.01B vs $87.62B consensus, Azure +43% and past $100B annualized for the first time. Thursday: +15.51% (390.54 → 451.10), reported as its largest single-day gain on record. Added another +3.10% Friday. FY27 capex guided to $255–260B. - Meta — revenue $60.8B (+28%) but EPS $6.18 against roughly $7.20 expected, net income -14% to $15.85B, operating margin compressing to 31% from 43%, and the full-year capex floor raised to $130–145B. Thursday: -7.95%. Recovered +3.24% Friday.
Reported Thursday night, traded Friday: - Amazon — revenue $200.61B (+19.6%), AWS $42.2B. Friday: +15.32% (235.50 → 271.57). Read the earnings quality before the headline: reported EPS of $5.75 includes a $53.4B unrealized gain on its Anthropic stake. That is not operating income. - Apple — record revenue $109.42B (+16%), iPhone $54.25B, but Services missed at $30.74B and the company flagged component and memory shortages constraining forward growth. Friday: -7.32% (333.43 → 309.03).
Around them: AMD +13.00% Thursday then -1.89% Friday, NVDA +2.65% then +2.96%, Alphabet roughly flat both days after its own capex-driven drawdown the prior week.
The pattern is consistent. Capex paired with visible revenue conversion got paid; capex paired with margin compression or a supply bottleneck got sold. And note what that does to an index: Thursday delivered an S&P 500 +1.7% and a Nasdaq +2.8% session with the tech sector +5.2%, and QQQ still finished the week below its 50-day and the month down 6.57%. Two of the index's largest weights moving 15% in opposite directions on consecutive days is dispersion, not an advance — and dispersion is what opened that 6.60pp July gap.
Oil and Iran
Crude fell hard on the week and rose hard on the month. Brent lost more than 7% and WTI more than 6% week-over-week, finishing Friday near $90.39 and $85.44 — while both held gains of roughly 22% for July, their first monthly gain since April, after Brent approached $100 and WTI traded above $92 at the late-July highs.
What came out was geopolitical premium, and the mechanism was shipping. Twenty-five vessels transited Bab al-Mandeb despite continued Houthi attacks, and Qatar resumed LNG shipments through the Strait of Hormuz after a three-week pause that followed a tanker attack. But Hormuz is still severely restricted: only two vessels transited Friday, in a waterway that normally carries about a fifth of global energy exports. On a bilateral agreement, Wood Mackenzie's read is that "routes to a deal are far from obvious." Weak China manufacturing data and expected OPEC+ supply increases were also cited as pressuring price.
This is why our USO proxy reads the way it does — -5.5% on the week but +21.35% for July, and +4.76% against its 50-day, +28.9% against its 200-day. That monthly figure is our own, computed from our bars, and it lands within a point of the Brent and WTI moves above. The week and the trend are telling different stories and both are real. It is also, per the statement above, the input the Fed named.
The weekly frame
SPY closed at 747.03, up 1.10% on the week, back above the weekly averages it had been fighting: the 10-week SMA at 743.48 and the 8-week EMA at 741.69, price 0.48% and 0.72% above them, red-week streak zero. But the weekly MACD line (19.22) is still under its signal (19.72), histogram -0.49 — price reclaimed the averages, momentum has not confirmed. Weekly RSI 61.4, stochastics elevated at 78.0/75.4.
QQQ closed at 687.99, up 0.55% on the week, and is a different chart entirely: 3.3% below its 10-week SMA and 1.68% below its 8-week EMA, weekly MACD histogram -4.25, roughly an order of magnitude deeper than SPY's. IWM closed at 291.20, flat on the week (+0.01%) and down 0.49% Friday, sitting closest to the line — 0.56% under its 10-week, 0.16% under its 8-week, weekly MACD just rolled over.
Our tracked analog — two or more consecutive red weeks and below both weekly averages — is not in effect this week. For the record: 13 occurrences in our ~129-week window, one week later a median +0.51% and higher in 8 of 13, range -3.07% to +5.67%. Tiny sample, ~2.5 years, and it did not trigger.
The daily chart
Friday printed the same shape on all three — a long lower tail. SPY 62% of the range, IWM 55%, QQQ 51%. SPY's was a small-body indecision candle closing above its open; QQQ and IWM both closed below their opens despite the tails.
SPY's daily stack is compressed nearly flat: 10d 741.30, 50d 744.22, 20d 745.69, price above all three and 7.12% above the 200-day at 697.39. The 20/50 spread is +0.20% — roughly one ATR (7.28, or 0.97% of price) resolves that compression either way. Daily MACD is still negative (histogram -0.83), RSI 53.1, stochastics crossing up at 67.7/38.9. The 20-day traded range is 729.10–755.58; the 60-day is 716.58–758.45.
QQQ's stack is not compressed — it is inverted and separating: 1.86% under the 20-day, 3.74% under the 50-day, spread -1.92% and widening, ATR 1.74% of price. Its 20-day low, 661.14, is also its 60-day low. IWM compressed fastest of the three (spread +0.59%, narrowing 0.239pp), with its 50-day directly overhead at 292.26 against a 291.20 close, and the group's widest 200-day cushion at +9.92%.
Cross-asset (ETF proxies — labeled as such)
These are funds, not the underlying: USO is not WTI, UUP is not DXY, VIXY holds VIX futures and is not spot VIX. On the week — USO -5.5%, UUP -1.43%, TLT -1.2%, VIXY -4.34%, GLD -0.1%, IEF -0.09%. On the month — USO +21.35%, TLT -4.47%, VIXY -3.66%, GLD +0.86%, IEF -1.39%, UUP -0.84%.
The month is where the bond move shows up: TLT -4.47% in July against IEF -1.39%, the long end giving up more than three times the belly. Coverage attributes that to the 10-year yield reaching its highest since January 2025 — we carry no yield feed, so the yield claim is theirs; the ETF spread is ours.
One honest discrepancy: the energy coverage cited a stronger dollar among the factors pressuring crude, while our dollar proxy fell 1.43% on the week and 0.84% on the month. Different instrument, different window — we report our own tape and don't reconcile it to the narrative. And gold gained under a point in a month when crude gained twenty, sitting 9.79% below its 200-day despite an active Middle East conflict — premium leaving the haven trade even as it stayed in the barrel.
Open interest
Open interest only — our key carries no greeks, so this is not gamma and we won't call it that. OCC as-of: 2026-07-29, which lags and is not Friday's book.
July 31 expiry, within ±6% of spot: the largest call strike is 755 at 57,543 contracts, then 760 (22,180), 750 (21,834), 745 (21,230). The largest puts sit lower — 720 (56,927), 715 (55,829), 740 (50,799), 725 (46,882). The August 3 book is an order of magnitude thinner on both sides. Structurally, below 740 the next put strike with comparable size is 725, then 720; above 750 the next call strike with size is 755.
Next week
AMD reports Tuesday, August 4, after the close — confirmed by the company, and the one name in our local semis cache with a print next week. The July employment report is scheduled for Friday, August 7 and CPI for Wednesday, August 12 — the first inflation read the September FOMC will weigh, in a committee where three members already dissented toward a hike, and with the statement having named energy as a driver.
What we don't carry
Dark-pool prints, MOC imbalances, spot VIX/VIX1D/VVIX, FOMC odds, true gamma or greeks, and a Treasury yield feed. If a level in this letter depended on one of those, it wouldn't be in here.
The Coil board
Account mode is DAY. SPX is names-on and agile: SPY plus 3 of 11 sectors open (Consumer Staples, Industrials, Real Estate), 252 names fully qualified. QQQ is risk-off and in cash — above its 200-day but 3.7% under its own 50-day, with no index setup and no sector open beneath it. Macro is risk-off and defensive. Crypto's gate is closed, 0 of 2 sleeves in: BTC is 9.6% below its 200-day, and both sleeves stay in cash regardless of their own trends until a UTC daily close reopens it.
Read the board at coil.trade/scanner · own the engine for $29.
This is impersonal research and market commentary, not investment advice. Nothing here is a recommendation to buy or sell any security, and no representation is made about future performance.
Provenance — all prices, moving averages, ranges and single-stock session moves are our own, from SIP consolidated bars cross-checked against a second broker feed; open interest is OCC as of 2026-07-29. The FOMC decision, vote and statement language are quoted from the Federal Reserve's July 29 release. Earnings figures, energy and shipping reporting from CNBC, The National, The Globe and Mail, Money Morning and Saxo.
The Friday Wrap is impersonal market commentary. It is not investment advice, not a recommendation, and not tailored to anyone’s circumstances. It contains no forecasts, targets, or trades. Prices come from our own market-data bars; every external claim names its source in the letter.