The week that was

SPY finished July 27–31 up 1.1% at 747.03, and three of the four weekly recaps in circulation were describing the prior, down week. That single confusion explains most of the contradictory headlines that circulated over the weekend, so this recap separates the two weeks explicitly, using our own stored daily bars as the measuring stick.

The week's shape, from our bars: SPY was flat Monday and Tuesday (+0.02%, +0.24%), dropped 1.54% on Wednesday, then closed the week strong — up 1.68% Thursday and 0.72% Friday to finish at 747.03, a 1.10% gain on the week. QQQ followed the same path at a smaller amplitude, closing Friday at 687.99 for a 0.55% weekly gain. The week ended with buying, not selling: the only red session was Wednesday.

Atrani Capital's update — the one sourced recap that actually covers July 27–31 — matches this tape: a 1.1% gain for the S&P 500 and a 1.6% advance for the Nasdaq Composite, with the combined market capitalization of the Magnificent Seven up approximately 4.0%, or roughly $880 billion [R3]. Our bars corroborate both the index figure exactly and the concentration story directionally: within that cohort the week was violently mixed — Microsoft and Amazon rose sharply while Apple and Meta fell — and Atrani's observation that the Magnificent Seven's dollar gain exceeded the entire S&P 500's roughly $700 billion advance is consistent with a +1.1% week on the index.

The other three recaps describe the week before. Sterling Capital's −0.6% S&P reading [R1] matches the July 20–24 week in our bars (−0.59%) almost exactly; Citizens Bank's summary of a lower S&P and Dow with the Nasdaq down more than 2% [R2] likewise fits that prior week, in which QQQ fell 1.60%; and Graphite Financial Group's recap says in its own title that it covers the week ended July 26 [R4]. Weekly recaps published on a Monday conventionally describe the week just ended — a detail that matters when reading market summaries dated near a week boundary. An earlier version of this post presented all four as competing reads of the same window and attributed the disagreement to methodology; that was wrong, and the correction note below records it.

Level context, restated against our bars: SPY entered the week at 738.93 (an index-equivalent near 7,410) — not the ≈7,458 a widely shared recap claimed, which corresponds to the prior week's entry [R5]. The 52-week high context in that same recap does check out: SPY's 52-week high of 760.40 puts the index-equivalent ceiling near 7,620, so the 7,600 area it flagged as resistance was a fair description of the standing high-water mark. As of Friday's 747.03 close, SPY sat about 1.8% below that peak; QQQ, at 687.99 against its 52-week high of 748.65 [R7], sat 8.1% below its own.

The longer QQQ statistics hold up under checking. The five-year cumulative return of +94.45% as of July 30 [R6] means a $10,000 position five years prior would have grown to roughly $19,445. Year-to-date total return through August 3 stood at approximately +14.23% for QQQ against +11.70% for SPY [R10], and the 28.08% trailing 52-week change [R9] — measured to early August — is consistent with our stored closes. These figures come from sources measuring slightly different windows and bases (price versus total return), which is why they differ from a naive close-over-close calculation; each one reconciles once its window is applied.

Our tape

SPY signal-day executions, open to close: 36 winners in 53 trades — a 67.9245% win rate, measured on our stored bars since January 2023 and served by the same engine behind the daytrade panel. The rules, the bases, and the loss case are documented in how our SPY daytrade signals work.

The week now underway

This post publishes midweek, so the first sessions of August 3–7 are already tape rather than forecast: SPY rose 1.42% on Monday to 757.67 and 1.80% on Tuesday to 771.33 — through the 7,600-equivalent area discussed above and past the prior 52-week high. The remaining scheduled items for the week are the labor-market releases, which stay the most watched macro prints given the Federal Reserve's stated data dependency [R2], and the tail of earnings season across technology and consumer names. What that produces is for next week's recap to measure.

Sources

[R1] Sterling Capital — Weekly Market Recap 7-27-26 (covers the week ended July 24; link dead at verification time) — https://sterlingcapital.com/cdn/Weekly-Market-Recap-7-27-26.pdf

[R2] Citizens Bank — Weekly Market & Economic Recap July 27, 2026 (covers the week ended July 24) — https://www.citizensbank.com/dam/ced55476-ae49-49d0-ba99-b3fa0106034d/weekly-market-recap-original-file.pdf

[R3] Atrani Capital — Week #31 Market Update for July 27–31, 2026 — https://atranicapital.substack.com/p/week-31-market-update-for-july-27

[R4] Graphite Financial Group — US Market Recap for Week Ended 26 July 2026 — https://www.facebook.com/GraphiteFinancialGroup/posts/-us-market-recap-for-week-ended-26-july-2026-equitieswall-streets-major-averages/122285695088146511/

[R5] Traderverse — Weekly Market Recap: July 27–31, 2026 — https://traderverse.io/blog/weekly-market-recap-july-27-31-2026

[R6] Invesco — QQQ ETF Performance (as of 07/30/2026) — https://www.invesco.com/qqq-etf/en/performance.html

[R7] Macrotrends — Invesco QQQ 27-Year Stock Price History — https://www.macrotrends.net/stocks/charts/QQQ/invesco-qqq/stock-price-history

[R8] Yahoo Finance — Invesco QQQ Trust Historical Prices & Data — https://finance.yahoo.com/quote/QQQ/history/

[R9] Investing.com — QQQ ETF Stock Price History — https://www.investing.com/etfs/powershares-qqqq-historical-data

[R10] Total Real Returns — QQQ vs. SPY Total Return through 2026-08-03 — https://totalrealreturns.com/n/QQQ,SPY

Correction (2026-08-06). The original version of this post presented four weekly recaps as conflicting accounts of July 27–31 and attributed the conflict to methodology differences. Verification against our stored daily bars showed three of the four ([R1], [R2], [R4]) describe the week ended July 24, whose −0.6% S&P return our bars confirm; the week this post covers closed up 1.10%. The original also stated the S&P entered the week at ≈7,457.69 (that was the prior week's entry; the actual entry was ≈7,410) and described a "1% Friday selloff" (Friday was +0.10%; the 1.2% drop was the prior Thursday). Every figure in the current version is verified against our stored bars, the live engine output, or the cited source directly.

Derived, delayed data. Impersonal research, not investment advice. We publish our own measured aggregates and cite public sources.