Weekly Market Pulse

Market map: technology leads, four sectors advance

S&P 500 ETF · SPY+1.27%Sep 25 close: $771.35

Nasdaq-100 ETF · QQQ+3.16%Sep 25 close: $744.50

Dow ETF · DIA+0.31%Sep 25 close: $517.49

Russell 2000 ETF · IWM-0.75%Sep 25 close: $281.97

XLK +3.52%XLC +1.94%XLV +1.37%XLI +0.40%XLF -1.83%XLRE -2.28%XLE -3.53%XLU -3.87%

SPY daily chartOpen on Finviz

XLK, XLC, XLV and XLI were the positive sector ETFs. Utilities lost 3.87%, energy 3.53%, real estate 2.28% and financials 1.83%. The Dow ETF’s 0.31% gain was also much smaller than QQQ’s advance. The result is a growth-led week with weak participation in several groups that would normally help confirm a broader improvement.

Software rose 1.59% for the week but declined 1.06% on Friday; semiconductors gained 0.96% that day. Even within technology, weekly leadership did not mean every segment finished with the same momentum. Reading the sector, industry and daily panels together is more informative than treating a positive Nasdaq proxy as a uniform signal.

Source: Marketstack end-of-day closes. Weekly return = September 25 close / September 18 close − 1; Friday = September 25 / September 24 − 1. These are unadjusted price returns, without dividend reinvestment. The live Finviz chart can advance beyond this edition.

Cross-asset map: bonds and metals fall as the dollar strengthens

TLT’s 2.32% decline was larger than IEF’s 0.86% loss, consistent with the greater price sensitivity of longer-duration exposure. LQD fell 1.42% and HYG 0.85%. Those observations describe fund prices; they do not isolate credit spreads from duration or justify inventing an underlying yield level. The combination nonetheless offers a useful caution against describing the equity rebound as an across-the-board easing in financial conditions.

Gold and silver funds fell 1.93% and 2.99%, respectively, while the dollar fund rose 0.81%. The moves are compatible with several explanations, including currency and discount-rate sensitivity, but this dataset cannot establish which factor dominated. A stronger dollar also does not mechanically dictate the weekly direction of every overseas equity market: EEM gained 1.42%, EFA 0.56% and FXI fell 1.05%.

Energy was internally divided. UNG gained 6.92%, USO fell 3.57% and XOP lost 4.78%. Futures exposure, roll effects and company operating leverage make these different instruments. A natural-gas fund rally is not evidence that oil producers or the broad energy sector had a strong week, and none of these ETF closing prices is a spot commodity quotation.

Sessions in review: September 21–25

Much of the week’s benchmark gain arrived on Monday: SPY rose 1.55%, QQQ 2.74% and SMH 4.09%. Wednesday then took 1.84% off small caps, a larger decline than the broad benchmarks. Friday’s 0.11% IWM gain was too small to repair that weekly damage. The path explains why a constructive closing session and a negative small-cap week can coexist.

The five daily returns compound to the weekly return; adding rounded daily percentages will not reproduce it exactly. All calculations use the same Marketstack regular-session close series, beginning September 18. The session descriptions report relative performance and do not attribute every move to a single headline.

Macro: low claims, mixed activity and a major release cluster ahead

The policy backdrop is inherited from last week

The September 16 FOMC decision raised the target range by 25 basis points to 3.75–4.00%, with a 12–0 vote. It was not a new decision during September 21–25. The statement paired solid economic activity with elevated inflation, which leaves the next data releases relevant to both growth and the policy outlook. It would be premature to turn one good equity week into evidence that this tension has been resolved. Federal Reserve statement

Claims show limited layoffs, not the whole labor market

The September 24 release reported 197,000 initial claims for the week ended September 19, down 1,000 from a revised 198,000. The four-week average declined to 202,250. These figures indicate relatively low new unemployment-insurance claims in that reporting period. They do not establish the pace of hiring, job availability or wage growth. Next week’s JOLTS and employment reports test different parts of the labor market, so a low claims figure should not be used as a substitute for them. Department of Labor release

Housing and manufacturing need a measured reading

Census reported August durable-goods orders of $338.6 billion, essentially unchanged from July. A flat headline is not enough to determine business investment momentum without the detailed components and revisions. August new-home sales were estimated at a seasonally adjusted annual rate of 684,000, up 6.4% from July. That is an annualized estimate, not the number of homes sold in one month; the release’s sampling uncertainty also argues against calling one monthly increase a confirmed housing recovery. Census economic indicators

September 30 combines new data with historical revisions

BEA schedules both August Personal Income and Outlays, including PCE inflation, and the third estimate of second-quarter GDP for September 30 at 8:30 a.m. ET. The GDP release covers Q2, not Q3. Annual updates also revise parts of the historical accounts, so the apparent trend can change because the past has been re-estimated as well as because the newest month differs. The useful comparison is the new release against its revised history, with the revision component stated explicitly. BEA schedule · Annual-update information

Earnings: Costco offers a concrete test of consumer resilience

Costco’s September 24 report is the verified consumer-results anchor for this edition. Its 16-week fourth quarter produced net sales of $93.9 billion, up 11.2%; full-year sales were $297.2 billion, up 10.1%. Total comparable sales increased 9.4%, or 6.7% after the company’s gasoline-price and foreign-exchange adjustments. The adjusted comparison matters because reported sales growth alone mixes underlying activity with those external effects. Costco fiscal 2026 results

Quarterly net income was $2.998 billion, or $6.75 per diluted share, versus $5.87 a year earlier. The company identified a $0.15 per-share nonrecurring benefit related to IEEPA tariff refunds, net of a partial reinvestment in member value. That item belongs in the explanation of earnings growth; it should not be silently treated as a recurring operating improvement. Adjusted digitally enabled comparable sales rose 19.8%, offering another operating lens alongside the warehouse business.

Interpretation: these results provide company-level evidence that value-oriented consumer spending remains substantial. They do not prove that all discretionary categories are healthy or that higher financing costs no longer matter. The next set of releases broadens the test: Micron on September 30 addresses memory and capital spending; Accenture’s October 1 call addresses enterprise demand; Nike’s October 1 report addresses another part of the consumer market. Their dates are future events in this edition, not results already reported.

The comparison across those businesses is useful precisely because their drivers differ. Strong memory demand can coexist with cautious consulting budgets or uneven apparel demand. A persuasive market narrative needs the actual revenue mix, margins, cash conversion and management guidance from each release, rather than counting how many headlines use the word “growth.” This is a selected analysis, not a complete earnings calendar.

Biotech: one approval, one advisory outcome and one imminent readout

IBB gained 2.43%, while equal-weight XBI lost 1.08%. Among the ten names on our detailed research page, GRAL rose 57.10%; SMMT fell 12.55%, SVRA 6.89% and PRAX 6.08%. These full-week returns measure price performance. They do not establish that a particular news item caused the entire move, and they do not measure the probability of a future clinical or regulatory outcome.

GRAIL: the committee has voted; FDA has not issued the decision described here

GRAIL reported a 7–2 favorable benefit-risk vote, with one abstention, for Galleri on September 23. The reported safety vote was 10–0 and effectiveness 6–4. Those different votes should not be collapsed into a unanimous endorsement of every aspect of the evidence. Advisory input can inform FDA, but the committee does not grant marketing approval. The remaining question is the agency’s review and any conditions it may impose; no final action date is invented. GRAIL advisory-committee outcome

Mirum: the September 26 target is now a completed regulatory event

FDA approved Atebrioz (zilurgisertib) on September 25 to reduce the volume of total new heterotopic ossification in FOP patients aged 12 and older. The label concerns an endpoint and population that need to be stated precisely. Approval does not by itself establish uptake, reimbursement or commercial success. The joint company release expects U.S. availability in October through Mirum Access Plus; no specific day is announced. The next research questions are access, prescribing and launch costs. MIRM’s 2.32% weekly decline does not isolate the reaction to approval. The price series ends at the regular-session close, while the joint company release was published later that evening; FDA news may have been available earlier through other channels. FDA approval and clinical basis · Joint company approval release

Kodiak: an exact date replaces the September window

Kodiak’s September 25 announcement schedules DAYBREAK results for September 28, with an 8:30 a.m. ET webcast. Zenkuda and KSI-501 are separate investigational arms against aflibercept in wet AMD. Each needs its own efficacy and safety reading; success in one arm would not automatically validate the other. Visual-acuity results, treatment burden and ocular safety will matter beyond a headline statement of non-inferiority. The scheduled presentation is confirmed, while its outcome remains unknown. Kodiak presentation notice

These developments are also reflected in the updated Top Ten Biotech research page, where trial design, finances and the limits of each catalyst receive a fuller discussion.

Verified watchlist: September 28–October 2

The week combines a binary clinical event, inflation and GDP revisions, major company reports and payrolls. Times are U.S. Eastern Daylight Time; Rome is six hours ahead on these dates. A conference-call time is not necessarily the earnings-release time. Check the linked issuer or agency notice for subsequent changes.

Wednesday’s PCE/GDP combination and Friday’s jobs report test different risks. Firmer inflation alongside resilient employment would not carry the same policy implication as softer inflation with deteriorating hiring. Likewise, annual GDP revisions may alter the historical story without representing a new change in current-quarter activity. Those distinctions are more useful than assuming every large release must push the market in the same direction.

Sources and methodology

Marketstack end-of-day data were retrieved September 26 for September 18–25. Weekly and daily returns use unadjusted regular-session closing prices on a consistent basis, with no dividend reinvestment. ETF prices are USD; they are not underlying index levels, Treasury yields or spot commodity quotes. Finviz supplies the existing live chart, whose date may move beyond this edition.

Marketstack · FOMC September 16 · DOL claims September 24 · Census activity releases · Costco results · GRAIL panel outcome · FDA Atebrioz approval · Kodiak September 28 notice · BEA calendar · BEA annual updates · BLS calendar · Micron notice · Accenture notice · Nike notice

Transparency boundary

Agency and issuer documents establish policy decisions, operating results and scheduled events; price-based interpretation is labeled separately. This is a selected weekly review, not a claim to cover every release. A company announcement may be the only primary source for its future event; that confirms the announced schedule, not the outcome. Data and dates can be revised after publication.