Walmart cracked. The company actually beat revenue estimates — but its US same-store sales growth of 2.6% missed, its full-year profit guidance came in under Wall Street's number, and management said $4 gasoline is changing how customers shop. The stock fell 9.15%, its worst day in more than four years. Because Walmart sits in the Dow, the S&P 500 and the Nasdaq-100, the damage spread everywhere: the Dow lost 704 points (−1.32%), the S&P fell 0.87%, and even TJX — which beat and raised guidance — fell 3%. When good news gets sold, the market is telling you something.
The bond market called the Treasury's bluff. On Wednesday, the Treasury announced it would more than double its buybacks of long-dated government debt, and yields fell sharply — relief. On Thursday, Treasury Secretary Bessent went further, saying buybacks "could exceed $4 billion" per operation, done partly "to show that we believe the yields don't reflect the underlying fundamentals." The bond market's answer: yields went straight back up. The 10-year closed at 4.71% — its third close above the 4.70 line, which flips the framework's rate rule from "armed" to confirmed. The 30-year sits near 5.26%, a whisker from 19-year highs, on the same day the national debt crossed $40 trillion. A rescue that gets erased in 24 hours isn't a rescue — it's evidence the sellers are bigger than the buyer.
"Economic D-Day." President Trump declared "the MOST CRUSHING ECONOMIC OPERATION EVER" against Iran — sweeping sanctions plus threats of "tremendous economic consequences" against any country that helps Tehran, with China (Iran's biggest oil customer) squarely in the crosshairs. Iran's foreign minister shrugged it off; no talks are scheduled; a second US aircraft carrier arrived in the region. Oil jumped more than 2% to nearly $94 — through the framework's $93.19 pre-alarm line and its highest level since late July. The escalation just changed shape: from a military standoff the market has priced, to an economic war with a possible China trade-conflict tail the market has not.
Yesterday we wrote that a healthy market fixes its edges first, and Wednesday had fixed the middle while the edges bled. Thursday the bleeding reached the middle. The NYSE's weekly momentum gauge broke below 250 — a floor the framework watches because breaks below it have historically preceded "staircase" declines, not gentle drifts. Both daily breadth thermometers went back below zero. And the single most important number on the board: the count of Nasdaq-100 stocks still on technical buy signals fell to 60.40 — four-tenths of a point above the 60 line that separates a correction from something structural. One more down day breaks it.
Two things still argue against panic. Credit refuses to crack — junk bonds sit ~0.4% off their highs, which historically caps how bad things get, fast. And the market is ten points from its magnets: the early-August "gaps" on both indices (price zones the market skipped on the way up, at S&P 7,629 and Nasdaq 29,109) sat 9–10 points below Thursday's lows. Markets are drawn to fill these zones, and a tag-and-bounce there is the market's most natural next move — which is exactly why the framework treats the first touch as a buying opportunity and the bounce that follows as the real test.
Why rollover is back on top: its full checklist fired in one session — banks, transports, new lows, both breadth thermometers, the weekly momentum floor, the consumer bellwether, and the index itself closing below its 50-day average for the first time this rally. Why shock jumped eight points: oil through the pre-alarm on an open-ended escalation, a failed government intervention in the bond market, and gold at records is the classic pre-shock signature — held in check only by calm credit markets and a fear gauge still at just 16. Why range collapsed: its entire thesis was "money is rotating, not leaving." Thursday, money left.
| Level | What it is | Role right now |
|---|---|---|
| 7,700–7,708 | Thursday's gap · 5-day average 7,698 | First resistance — a five-layer lid; the fade zone for shorts |
| 7,683 | 14-day average | The rally's short-term trendline — now overhead |
| 7,641 | Thursday's close · old record high 7,640 | The decision line — price is sitting ON it |
| 7,631 | 20-day average · Bollinger mid | Closed exactly on it — hold = range, lose = gap opens |
| 7,629–7,600 | Early-August gap | THE magnet — 10 points below Thursday's low; the framework's buy ladder |
| 7,533 | 50-day average | The correction line — the rally is intact above it |
| 7,092 | 200-day average | The bull/bear line — 7.7% below |
| Level | What it is | Role right now |
|---|---|---|
| 29,306–29,415 | Broken 50-day avg · weekly 5-day · 69-day avg | The fade cluster — bounces into this stack are the short entry |
| 29,379–29,426 | Thursday's gap | The sixth lid layer — filling it on real news would be the first repair step |
| 29,118 | Thursday's low | First support |
| 29,109 | Early-August gap top | THE magnet — 9 points below the low; tag-and-bounce = the bull's best case |
| 29,087 | 20-day average · Bollinger mid | Part of a 31-point floor cluster with the gap and the low |
| 28,777 | Full gap fill · weekly 20-avg 28,811 | A "healthy retest" — but only while credit holds |
| 26,810–26,780 | 200-day avg + quarterly trendline | The framework's highest-conviction long-term buy zone — rising ~0.5%/month |
The core structural thesis — a three-peak weakening of long-term momentum on the Nasdaq-100 across 2018, 2021, and 2025-26, matching the pattern last seen in 1999 — had a good week. The monthly momentum reading eased again to 69.24, now 8.64 points under the 77.88 cancellation line, trading below the 70 threshold for a third straight session. Price making new highs in June while this gauge made a lower third peak is the warning; price now breaking its 50-day average while the gauge fades is the warning starting to resolve. Only a future monthly close with the momentum gauge back above 77.88 cancels the thesis. Nothing this week moved toward cancellation.
Today is monthly options expiration plus flash business surveys — expect pinning and noise. But next week stacks the month's two biggest events 36 hours apart: Nvidia reports Wednesday evening (Aug 26), and new Fed Chair Kevin Warsh delivers his first-ever Jackson Hole keynote Friday morning (Aug 28) — the same morning as the Fed's preferred inflation gauge. Warsh has deliberately stopped telegraphing the Fed's intentions, which makes this speech carry real information for a market that just lost its rate anchor. Every tactical position below is sized with that Friday in mind.
Framework-derived structures with explicit entries, stops, targets and abandon-conditions. Sized for a leveraged CFD book under the 30% tactical cap currently imposed by the oil rule. These are educational illustrations of how the framework expresses its read — not recommendations.
| Horizon | Base case | What changes it |
|---|---|---|
| Today (opex Friday) | Early tag of the magnets (29,109 / 7,629), a reflex bounce into the expiration pin, close inside the decision bands | Strong business surveys lift the tape into the fade zones; a $95 oil settle turns the magnets into trapdoors |
| Next week | Defensive drift 28,777–29,677 into Nvidia (Wed) and Warsh (Fri); the breadth binary likely breaks on any down day | Nvidia blowout + dovish Warsh = the overhead gaps finally fill; hawkish Warsh + $95 oil = the staircase accelerates to 28,564 |
| One month | August's monthly close locks the third momentum peak; battleground 28,280–30,196, action in the lower half; Sept Fed meeting (~⅓ hike odds priced) | Cancellation requires momentum >77.88 — currently 8.64 points away and moving the other direction |
| Six months | 26,800–30,800 range with downward drift; the rising 200-day/quarterly band at 26,810–26,780 is the destination of any staircase — and the framework's highest-conviction buy zone | A peace dividend plus a credible Warsh framework reopens 32,000+ |
| Year-end | 27,000–29,000 — lower highs per the divergence, the deep band defended on first test | Bull tail 32,000–33,500 · shock tail 24,500–26,000 on stacked crises |
What worked: the volatility position paid (+7.5% day, first scale at the door); the rates-banks short is +437 with half banked at its first target; the gold entry filled and is +42; the Walmart bracket's bear leg triggered on the gap-down and paid; and the fade-rungs above were correctly never chased — price fell 213 points without offering the entry. What didn't: the rotation pair was exited at a scratch — its thesis (broad market absorbing chip weakness) broke the day after entry, and the framework closed it rather than argue with the tape. What was fixed: the stale breadth feed (three identical prints) updated and caught down to 51.49 — matching every other gauge, and validating the decision to exclude it while stale.