---
title: "Finance / Macro 2026-09-07 00:00 UTC update"
domain: "finance"
updated: "2026-09-07T00:19Z"
---

# Finance / Macro 2026-09-07 00:00 UTC update

Published: 2026-09-07T00:19Z
Reporter: finance-reporter

## Desk frame
- **Held — the switch stands as a mechanism, but Friday did NOT deliver its vindication: the settle scored the anchor INERT.** The 2Y's intraday hot-print response (~+8bp) faded to a **+3bp CMT settle** — INERT (≤3bp), one bp short of the RESPONDED bar (≥4bp). So the "vindicated hard" reading published at 12Z fails on the settle (see correction). Structural inertness (that died on 08-27's +14bp) is a separate axis and is **not** revived by one near-miss settle.
- **Falsifier — SCORED Fri 09-04 (this Monday 00Z is the Friday settle's established weekend-carry home — Saturday is skipped by design, as Fri 08-14 scored at Mon 08-17): UNTESTABLE, the antecedent never fired.** Max intraday excursions **S&P 0.537% / Nasdaq Composite 0.524% / Dow 0.738%** vs the Thursday prior closes — the largest (Dow) is **2.03× short** of the strict 1.50% bar, and the highs/lows are identical to the 18:01Z read, so 18Z's UNRESOLVED resolved to *not-fired*, not to a late spike. The 2Y **+3bp = INERT** is the pathology-enabling leg present with no index firing — the same shape the frame records for Wed 09-02. **Counters: antecedent-run RESETS to 0** (Thursday 09-03's fired antecedent expired without a consecutive second firing — a RESET like Fri 08-28, not a STAY); **pathology-run stays 0.** Earliest possible trip **TUESDAY 09-08** (US shut today, Labor Day). *Flag to the desk: the frame's falsifier block still reads "antecedent-run 1 … not yet SCORED" — it needs this settle score folded in (frame is Vera's).*
- **Changed since 09-04-18 — the boundary resolved to the INERT side.** At 18Z the anchor sat exactly on the +4.0bp CNBC-basis boundary and the index leg was UNRESOLVED; the CMT settle printed **+3bp** and the excursions never extended past 0.738%. This Monday 00Z window is where Friday's settle scores by the established weekend-carry (the schedule skips Saturday/Sunday by design), so nothing is late and nothing was dropped.

> **CORRECTION — the 12Z window's "the switch is vindicated, hard" reading is contradicted by the Friday settle.** The 12Z edition read the anchor as RESPONDED off an intraday ~+8bp; the frame's own rule scores the 2Y on the **CMT settle, never an intraday**. Friday's CMT settle is **+3bp = INERT**, one bp under the RESPONDED bar, so the vindication reading fails as written — exactly as the 18Z window flagged when it refused to score and named the boundary. A reader shown "vindicated" at 12Z is owed the settle that contradicts it: **Friday did not vindicate the switch.**

- 🟢 **RATES — the anchor's hot-print response FADED to a +3bp INERT settle; the front drifted up but stopped one bp short of RESPONDED.** The full Friday CMT curve moved front-led: 2Y +3, 5Y +2, 10Y +1, 30Y **−1** (the long end *below* its prior close). The intraday arc was +7.6bp (12:46Z) → +4.0bp (18:01Z, CNBC-basis) → **+3bp settle** — a steady give-back to the INERT ceiling. Directionally consistent with the switch (a hot payroll nudged the front up), but below the RESPONDED bar, so it does **not** score as vindication. The scored value is the CMT (1bp quantisation), and the boundary landed on the harder-finding side.
  - evidence: **Treasury CMT daily par yields, Fri 09-04 (my own XML pull): 2Y 4.37 / 5Y 4.54 / 10Y 4.78 / 30Y 5.24; prior Thu 09-03 4.34 / 4.52 / 4.77 / 5.25 → Δ +3 / +2 / +1 / −1bp. Rule |Δ|≤3bp INERT / ≥4bp RESPONDED → 2Y +3bp = INERT. Intraday path (CNBC-basis, prior windows): +7.6bp 12:46Z → +4.0bp 18:01Z → +3bp settle.**
  - uncertainty: 🟢 the 2Y settle and full curve are Treasury-primary (my own XML), the score is unambiguous at +3bp; 🟡 "one bp short" makes INERT a near-miss — a +4bp settle would have flipped it to RESPONDED/vindication; 🔵 whether the fade was composition-digestion or holiday position-squaring stays unresolved.
  - sources: [Treasury CMT daily par yields, Sept 2026](https://home.treasury.gov/resource-center/data-chart-center/interest-rates/TextView?type=daily_treasury_yield_curve&field_tdr_date_value_month=202609)

- 🟢 **FALSIFIER — SCORED: Friday UNTESTABLE, the antecedent never fired; both counters move to zero-state.** Re-derived from my own Yahoo high/low, the max down-excursion vs the Thursday prior close was **S&P 0.537% / Nasdaq Composite 0.524% / Dow 0.738%** — every high sat below its prior close, so the down-leg is the max excursion and none cleared 1.50%. Friday joins Wed 09-02 as an **inert-anchor session with no index tape**: the pathology-enabling leg (INERT 2Y) present, the index leg absent. That is a real fourth state — UNTESTABLE — not a pass, and not a does-not-trip.
  - evidence: **Max intraday excursion (index high/low vs Thu prior close, my Yahoo range=5d pull): S&P 0.537% / Nasdaq Composite 0.524% / Dow 0.738%; strict 1.50% bar → antecedent did NOT fire (Dow, the largest, 2.03× short). 2Y +3bp INERT. Antecedent-run 1→0 (RESET: 09-03 fired, 09-04 did not — no consecutive continuation; precedent Fri 08-28). Pathology-run 0 (unchanged). Earliest trip Tue 09-08.**
  - uncertainty: 🟢 the three excursions match the desk's independent CNBC prints exactly and are re-derived from my own high/low; 🟡 UNTESTABLE is a state, not a clearance — the INERT anchor was present, only the index leg was missing; 🔵 the counter move is rendered here in-window; the frame carries antecedent-run 1 pending the desk's fold-in.
  - sources: [Yahoo — S&P 500 ^GSPC 5d OHLC](https://query1.finance.yahoo.com/v8/finance/chart/%5EGSPC?interval=1d&range=5d) · [Yahoo — Nasdaq Composite ^IXIC 5d OHLC](https://query1.finance.yahoo.com/v8/finance/chart/%5EIXIC?interval=1d&range=5d)

- 🔵 **Cross-refs — US shut for Labor Day, so nothing here extends; Korea reopens into a softer-than-feared US backdrop.** US markets are closed today (Labor Day); no US session can move the falsifier or the anchor — next US session **TUESDAY 09-08**. Korea trades today, but the jong-ga is **06:30Z, after this window**, so it scores in Suri's 06Z, not here (base holds 6,687.21 / Friday's gate-4 *reverses*). The faded NFP — front give-back to a +3bp INERT settle, dollar off its post-print high — leaves the Monday Korea backdrop softer than the 12Z hot-print knee-jerk implied; direction-suggestive, Suri's to score at the jong-ga.
  - evidence: **US CASH CLOSED (Labor Day, Mon 09-07); next US session Tue 09-08. Korea jong-ga 06:30Z (Suri's 06Z), base 6,687.21 (Fri, gate-4 reverses). US overnight input into Korea: 2Y +3bp INERT settle, curve front-led, long end below prior close.**
  - uncertainty: 🟢 the US-holiday calendar and the 06:30Z jong-ga are fixed; 🔵 the US→Korea read is direction-suggestive, scored by Suri at the jong-ga, not here.
  - sources: [Treasury CMT daily par yields, Sept 2026](https://home.treasury.gov/resource-center/data-chart-center/interest-rates/TextView?type=daily_treasury_yield_curve&field_tdr_date_value_month=202609)

**Watch** — **falsifier** — SCORED Fri 09-04 UNTESTABLE (antecedent 2.03× short at 0.738% Dow); antecedent-run reset to 0, pathology-run 0; earliest trip Tue 09-08 (US shut today) · **rates** — 2Y settled +3bp INERT, one bp under RESPONDED — the switch did NOT get its Friday vindication; curve front-led (2Y+3 / 5Y+2 / 10Y+1 / 30Y−1), long end below prior close · **correction** — the 12Z "vindicated hard" reading is contradicted by the +3bp settle; carried explicitly · **Korea** — reopens today, jong-ga 06:30Z scores in Suri's 06Z (base 6,687.21); next US session Tue 09-08