The cash session delivered almost exactly the map — and then went one step further. The framework's base case called for an open near 29,500–29,560, a push into the first lid, a pullback to the 29,373–29,453 stack that holds, and a chip gap approached but not filled. The Nasdaq-100 opened at 29,481, pushed straight through the lid, pulled back to a low of 29,366 — seven points under the stack's floor, which the close then reclaimed entirely — and finished at 29,641.56, up 1.43%, with the day's high at 29,643.53: exactly 33.76 points under the chip gap's bottom edge at 29,677. Approached, not filled. The S&P 500 spent the entire day inside its 7,714–7,745 magnet, touched 7,741.27, and closed at 7,730.99, up 0.72% — leaving 3.61 points of the gap's top unfinished.
Nvidia did what the futures promised. The stock closed up 8.7% — its largest one-day gain since April 2025 — after Wednesday night's report: revenue up 106%, next quarter guided to $108 billion, and zero China revenue in the guide. The software complex went with it: Salesforce up 22.6% on a beat and an expanded Anthropic partnership, CrowdStrike up 20.5%. The index had its best session in three weeks.
And the count finally moved — by triple the threshold. The Nasdaq-100 bullish-percent index, frozen at 51.49 for three straight sessions, jumped 9.90 points to 61.39. Ten stocks flipped to point-and-figure buy signals in one day — the largest single-session improvement in this entire window — and the gauge re-crossed the 60 regime line it broke on 21 August. Yesterday's pre-commitment was explicit: three or more net buys and the range path becomes the base case, the chip-gap fade is void, and the rollover path recedes. It printed nearly ten. All three consequences are executed below.
But the average stock did not come. On the NYSE, decliners beat advancers all day — the advance-decline count never went positive, finishing at −333 on an up-index day. The NYSE composite itself fell 0.38%. Transports fell 0.66%, giving back their one-day repair. The share of Nasdaq-100 stocks above their own 50-day average — a series that had been printing identical numbers to the bullish-percent count, a coincidence this digest noted rather than explained — resolved the coincidence by splitting from it: it fell to 48.51, below half, on the very day the count jumped ten. Ten stocks crossed price thresholds; the median stock ticked down.
The fear gauge broke its own floor, and oil re-armed the cap. VIX closed at 14.51, down 4.6% — through the 14.8 protective line the framework had drawn under its long-volatility core, which exited by rule at the close (more in the ledger). The September VIX future refuses to follow at 16.73: a 2.2-point premium — the market's price for one speech. Brent settled near $89.5, back above the $88.50 relief line one session after breaking below it: the oil rule re-arms, and the position cap returns to 30% after exactly one session at 50%. Released by two dollars on Wednesday, re-armed by one on Thursday — the framework does not round in its own favour, in either direction.
The stadium wave this digest has tracked all week resolves here, so the metaphor can retire with its work done: on Wednesday the biggest section stood up alone; on Thursday, ten more rows stood — the point-and-figure count measures exactly that, members crossing from downtrend to uptrend — while the rest of the stadium sat back down. A wave and an ebb at the same time, in the same building.
Which one governs? The framework answered that question before the session, which is the only honest time to answer it: the count is the pre-committed binary, and the count said members joined. Range becomes the base case. The structural-short licence — in force since the count broke 60 on 21 August — is revoked by the same rule that granted it. What the framework does not do is pretend the other half of the tape didn't happen. A ten-flip day on negative breadth is also what a failed re-cross looks like; the confirmation test is whether 61.39 holds. One number decides again, and it decides into a Fed chair's keynote and Monday's monthly close.
Why range takes the top by pre-commitment, not by mood: Wednesday this digest wrote that if Thursday's count moved, rollover would not stay on top — and it moved by triple the line. Rollover surrenders seven points not because its internals died (they did not: nine straight negative NYSE oscillator closes, a ninth summation decline, transports non-confirming again, fewer than half the Nasdaq-100 above their 50-day) but because the one number it was ordered to obey went against it. Those surviving internals are exactly why C keeps 27 rather than collapsing: if 61.39 gives way within days, the whole move re-reads as a one-day spike and the licence returns at higher, better levels. Why bull gains three and stops at 17: the count re-crossed, every daily lid was reclaimed in one session, and credit sits five cents off its record — but the buyer is already in. The manager survey printed 102.66 this week: more than fully invested, with leverage, into a binary. A crowd that long has no dry powder to chase with. Why shock only eases one: oil re-armed its own rule, the 10-year is drifting back toward 4.70 with the chair at the podium, the sanction name is still owed by the week-end — and the 2022 Jackson Hole precedent (next section) is the reason a 14.5 VIX into this speech is a tail worth pricing, not a comfort.
The event, precisely. The Kansas City Fed's 49th Jackson Hole Economic Policy Symposium opened Thursday under the theme "Financial Innovation: Implications for Payments and Policy." Fed Chair Kevin Warsh delivers the keynote today, Friday 28 August, at 10:00 a.m. Eastern — 10:00 p.m. Singapore time — thirty minutes after the US cash open. It is his first Jackson Hole as Chair: nominated on 30 January to replace Jerome Powell, confirmed 54–45 on 13 May after a delayed Senate process, sworn in on 22 May as the seventeenth chair of the Federal Reserve. This digest publishes before he speaks; nothing below is a reaction, all of it is a map.
The lore says Jackson Hole moves markets. The record says: usually it doesn't — and when it does, it is because a chair chose credibility over comfort. The average keynote-day S&P move across 2018–2025 is roughly minus 0.12%. One exception towers over the sample and it is the one that rhymes with today's setup.
| Year | Keynote date | Speaker & core message | S&P same day | Nasdaq same day | What followed |
|---|---|---|---|---|---|
| 2021 | Fri 27 Aug | Powell — taper coming, but rate hikes are a separate, distant test (dovish separation) | +0.88% | record close (Composite ~+1.2%) | drift higher into early Sep |
| 2022 | Fri 26 Aug | Powell — the eight-minute "pain" speech: fighting inflation "will bring some pain to households and businesses" | −3.37% | −3.94% (Composite) | −3.2% the next week; ~−13% the next month |
| 2023 | Fri 25 Aug | Powell — "navigating by the stars under cloudy skies"; prepared to hike further, proceeding carefully | +0.67% | ~+0.9% (Composite) | +2.6% the next week |
| 2024 | Fri 23 Aug | Powell — "the time has come for policy to adjust": the explicit pivot to cuts | +1.15% | +1.5% (Composite) | +0.3% the next week; Sep cut delivered |
| 2025 | Fri 22 Aug | Powell (his finale) — "shifting balance of risks may warrant adjusting" = green light for the Sep 2025 cut | +1.52% | +1.45% (Composite) | flat the next week; cut delivered |
Same-day figures verified against close reports; Nasdaq figures are the Composite where a dedicated Nasdaq-100 print could not be verified. The pattern that matters: the keynote day is usually noise; the month after carries the signal. In 2022 the same-day damage was −3.37%, but the real cost was the thirteen percent that followed over the next month — and Treasury yields, nearly unchanged on the day, rose almost a full point over the following month.
Here is what makes today different from every row in that table: 2021, 2024 and 2025 were markets waiting to hear how soon policy would ease. Today's market is pricing whether policy will tighten. The funds rate has been held at 3.50–3.75% for five straight meetings — but three FOMC members dissented in July in favour of a hike, core PCE has been stuck at 3.3% for two months (a sixth consecutive year above target), July payrolls printed −23,000, and September pricing is split in a way the prediction markets cannot reconcile: roughly 68% odds of no action on one venue, roughly 38% odds of a hike on rate futures. The direction of risk points toward tightening for the first time since 2022 — the year of the only red row in the table. That is the asymmetry under a 14.5 VIX: the market has priced the comfortable speech, not the credible one.
Who is speaking matters as much as what. Warsh is a former Fed governor (2006–2011), a long-standing QE skeptic who wants a smaller balance sheet, fewer words from the Fed, and rules over discretion. Since May he has run two meetings his own way: a markedly shorter statement, an abbreviated press conference, five internal task forces (communications, balance sheet, data, productivity, inflation framework), and a refusal to submit his own dot. He has lately been quoting Volcker's definition of price stability — inflation low enough that households barely notice it — and told reporters his draft was "like a blank sheet of paper." The prediction-market word odds for the speech: "inflation" 90%, "price stability" 84%, "AI" 78%, "balance sheet" 74%, "independence" 71% — and only 23% for "rate hike." One more pressure point: Treasury Secretary Bessent has been buying long bonds ($4bn+ buyback operations whose effect lasted two days) while the 30-year sits at 5.19% — a Treasury leaning on the long end while the Fed's chair preaches discipline is precisely the credibility question the gold market (4,647, record zone) is already voting on.
| Scenario | What it sounds like | Market read | The framework's levels |
|---|---|---|---|
| 1 · The middle path (consensus central case) | Sticky inflation named plainly, no September pre-commitment, task-force progress, a nod to data dependence — enough clarity to avoid a volatility surge, nothing tradable pre-committed | The 2.2-point VIX futures premium starts dying; the tape returns to its own structure (the chip gap above, the new gap below) | NDX chops 29,450–29,700; SPX 7,700–7,760; VIX crushes toward 14; the basis short (after 16:00 ET) is the trade |
| 2 · The Volcker moment (the 2022 rhyme; the priced-in-nowhere tail) | A blunt commitment that rates rise if core PCE does not fall — the line one major house says is required to regain market confidence — or a hawkish balance-sheet preview: faster runoff, a smaller Fed | The 2022 row of the table, scaled to today's positioning: managers at 102.66% exposure, retail already bearish, VIX at 14.5 — the exits are narrow | 10-yr through 4.72% re-fuels the rate rule; NDX fills the new gap 29,366→29,297, then the stack; VIX through 16.9; gold's first move is a flush toward the 4,540–4,560 re-entry gate before the debasement bid reasserts |
| 3 · The professor's speech (framework-only, AI-optimist) | Big ideas over guidance: productivity, AI disinflation, payments innovation per the symposium theme, institutional reform — September left entirely open | Read as dovish by omission; the market keeps its rate, loses its fear premium, and the chip gap becomes unfinished business | NDX fills 29,677→29,907; SPX through 7,763 toward the 7,799 record; 10-yr under 4.55% would disarm the rate rule entirely and add fuel |
The framework's stance is the same as it was for Nvidia: bracket the binary, don't predict it. The trigger pair is published and mechanical — a 10-year close above 4.72% is the hawkish confirmation (rates-banks structure re-builds, gold pair flattens); below 4.55% is the dovish one (rate rule disarms). Everything else is commentary. And one discipline note peculiar to today: the speech lands thirty minutes after the open. Every directional position in the book gets a 9:55 a.m. decision rule — no leg inherited from Thursday meets the chair by accident.
The simple version, as requested by the day: where each instrument is, where the framework's entries sit, where profits are taken, and the line that says the idea was wrong. Every number here reappears with its full context in the detailed tables and trade cards below.
| Instrument | Now | Buy / entry zone (long side) | Sell / trim zone (upper levels) | Take-profit refs | Line in the sand |
|---|---|---|---|---|---|
| S&P 500 (US500 CFD) | 7,731 close · 7,723 futures | 7,690–7,714 first floor (Thu low + gap bottom, post-Warsh hold) · 7,629–7,600 resting structural rungs · 7,558 50-day | 7,745–7,763 magnet top + 13-Aug shelf (fade only on a post-speech rejection) | Longs: 7,741 / 7,763 / 7,799 record. Shorts: 7,714 / 7,701 / 7,690 | Longs above 7,515 (below the 50/69-day cluster). Shorts void over 7,765 |
| Nasdaq 100 (USTECH CFD) | 29,642 close · 29,582 futures | 29,366–29,446 the reclaimed stack + 20-day (post-Warsh hold) · 29,023–29,105 deep floor (W 20-avg + gap) · 28,760–28,800 the Monday structure | 29,677–29,972 chip gap (no pre-speech fades; breakout through 29,690 flips it to a long trigger) | Longs: 29,677 / 29,907 / 29,972. Shorts: 29,366 / 29,297 / 29,105 | Bull structures dead under 29,250; bear brackets void over 29,470 re-close |
| Gold (XAUUSD) | ~4,647 spot (record zone) [web] | 4,540–4,560 re-entry gate on a hawkish flush (30-min reversal hold + DXY under 98.8) · deeper: 4,505 abandon line | 4,650–4,659 record shelf — no fresh longs INTO it; breakout only on a 30-min close above 4,660 | Flush-buy: 4,594 / 4,650. Breakout: 4,700, then trail | 4,505 (flush leg) / 4,620 (breakout leg) · any gold long flat if DXY takes 99.5 |
Gold levels quoted from wire pricing [web] — no chart feed in today's capture; the framework's gold rules are unchanged from their pre-commitment two sessions ago: nothing before the keynote passes, one leg maximum after it.
| Level | What it is | Role right now |
|---|---|---|
| 29,677–29,972 | 18-Aug chip gap | Still the magnet — untouched a seventh session; Thursday's high stopped 33.76 under the bottom edge. The bracket's bull trigger (29,690) fires through it; 29,907 (the 23.6% retracement) sits inside it |
| 29,657 | 60-min Bollinger top | Twenty points under the gap — the last technical lid before the magnet |
| 29,550–29,564 | 30-min 200-avg · 30-min 5-avg · banked half-target 29,554 | The overnight shelf the futures are sitting on; first support above the day session's structure |
| 29,446–29,458 | 20-day avg (= daily Bollinger mid, rose 77) · 10-day avg | The old lid is now the first floor — price closed 196 above the 20-day for the first time in six sessions; a post-speech hold here is the long entry |
| 29,366–29,297 | Thursday's gap (#32, confirmed 29,296.77–29,366.42) | The magnet underfoot — the NVDA gap the cash session confirmed; the bear bracket trigger (29,340) sits just inside its top; its S&P twin closed by 84 cents and was never logged |
| 29,275 | 50-day average | Reclaimed by 367 in one session after five closes below — the swing that summarizes the day |
| 29,042–29,105 | Weekly 20-avg · 60-min 200-avg · 25-Aug gap | The deep floor stack — the bear bracket's flat target zone |
| 28,760–28,800 | 4-Aug gap remnant (99 points to full fill at 28,777) | The Monday structure: only a hawkish flush reaches it (−2.8%), and it is the highest-conviction long zone above the R3 band |
| 26,906–26,922 | 200-day · quarterly 5-EMA | The R3 band narrowed to 16 points and inverted — the quarterly average now sits above the 200-day, both rising. The framework's structural buy zone, 9% below |
| Level | What it is | Role right now |
|---|---|---|
| 7,745–7,763 | Magnet top (3.61 unfinished) · 13-Aug shelf | The fade cluster — Thursday ate the magnet end-to-end and left 3.61 points; a post-speech rejection here is the tactical short; a close through 7,765 voids it |
| 7,799 | The record (14 Aug) | 68 points above — the professor's-speech target |
| 7,714 | Gap bottom (18-Aug) | First reference under the price; shorts' first target |
| 7,701 · 7,698 · 7,695 | 20-day (= Bollinger mid, retaken by 30) · 10-day · 5-day | Three averages in six points — the first floor; a post-speech hold of 7,690–7,714 is the long entry |
| 7,690 | Thursday's low | The day's truth: the dip under 7,700 was bought immediately |
| 7,629–7,600 | 4-Aug gap · resting structural rungs (7,629 / 7,604) | Deliberately left working through the keynote — a hawkish flush into this magnet with credit at records is what the ladder exists for; stop 7,515 |
| 7,558 · 7,538 | 50-day · 69-day | The cluster above the ladder's stop; 23.6% retracement at 7,521 just below |
| 7,114 | 200-day | 8.0% below — distant; the year's rising spine |
The monthly momentum gauge — the heart of the three-peak thesis — moved back above 70 on Thursday: 70.10, up 0.83, after seven sessions underneath. Monday's close prints the August bar and the verdict this digest has counted down to all week. The pre-commitment from two sessions ago stands word for word: a print at roughly 70 is "on the line, not through it — that changes nothing," and the framework will say so on Tuesday whichever side of 70 the bar lands. What would change everything is 77.88 — the cancellation line — and it is 7.78 points away, which one session cannot cover. The structure (three peaks across 2018, 2021, 2025-26; the 1999 monthly precedent; the lower high at 30,196 and the top at 30,780) is intact. What Thursday changed is the tactical licence: with the count back above 60, the framework holds the thesis without a structural short expressing it — a distinction between what the framework believes and what it is allowed to trade. The licence returns on a count re-close under 60; the thesis dies only at a monthly close through 77.88.
Oil re-took the line it broke. Brent settled near $89.5, back above $88.50 one session after settling below it — so the rule that released the cap to 50% on Thursday re-arms it to 30% today. One session of relief, surrendered by roughly a dollar. The bid: Tehran cooled the Iran–Oman revenue-sharing headline ("does not guarantee an immediate reopening" of Hormuz — the strait remains effectively closed), and Ukrainian strikes on Russian refineries added a second layer. The $95 settle — step three, a 20% cap — is 5.5 away, and the October contract expires Monday, the same day as the monthly equity close. The sanction name promised "by week-end" is still unnamed with the week-end here: that headline cancels every resting bracket in the book on sight.
The 10-year is walking back toward the line to listen. 4.67% at Thursday's close — a fourth session under 4.70, but drifting up (4.63, 4.65, 4.67), three basis points from re-fuelling the rate rule with the chair at the podium. The 30-year at 5.19% and a Treasury still running $4bn buyback operations frame Warsh's credibility problem better than any preview: the fiscal authority is leaning on the long end while the monetary authority preaches discipline. The trigger pair for today is unchanged and mechanical: 4.72% re-fuels, 4.55% disarms.
The dollar and gold are the jury. DXY at 99.09 sits 0.41 under the 99.5 circuit-breaker; gold at roughly 4,647 holds the record zone it broke into on Monday. The debasement trade — long gold against the paper the buybacks support — is the market's standing vote on the question Warsh must answer today. A chair who convinces the bond market gets a gold flush toward 4,540–4,560 (where the framework's re-entry gate has been waiting, unchanged, for three sessions); a chair who doesn't gets a breakout through 4,660 with the dollar through its breaker.
The cash session opens at 9:30 New York time; Warsh speaks at 10:00; a payrolls benchmark revision (expected upward) lands the same day. Nothing directional inherited from Thursday meets the speech by accident: the bracket's remaining half has a 9:55 decision rule, the fade waits for a post-speech rejection, the deep rungs alone stay working because their premise (a flush into a magnet with credit at records) is the event risk. Monday brings the August monthly close — the verdict print — and the October Brent expiry together. The sentiment backdrop into all of it is the widest disagreement of the window, on fresh prints: managers at 102.66% exposure, retail at bulls 32.9 / bears 44.4. Somebody is wrong by Tuesday.
Sized under the re-armed 30% cap, on a leveraged CFD book. The organizing principle is the one that paid on Wednesday night: bracket the binary, don't predict it — plus one discipline specific to today: the speech lands thirty minutes into the session, so every structure carries a 9:55 decision rule or an explicit post-speech gate. Win rates are honest estimates for calibrated readers, not promises.
| Horizon | Base case | What changes it |
|---|---|---|
| Today (Fri) | Pin, speech, decide: a quiet open near 29,550–29,620, the keynote lands middle-path — sticky inflation named, September left open, no volatility surge — and the tape chops 29,450–29,700, closing inside the box. S&P 7,700–7,760; VIX crushes toward 14 as the premium dies; the 10-year stays between its 4.55/4.72 trigger pair | A professor's speech fills the chip gap to 29,907+ and sends the S&P toward 7,799. A Volcker moment prints the 2022 rhyme: the 10-year through 4.72%, the new gap fills 29,366→29,297, VIX through 16.9, and the one-glance sheet's lower zones come live in an afternoon |
| This week → Monday | The verdict print decides: Monday closes the August monthly bar with the momentum gauge at roughly 70 — on the line, not through it, which by standing pre-commitment changes nothing. The count holds above 60, range consolidates as modal, and the October Brent expiry passes without a $95 print | The count gives back 60 within days — a ten-flip spike on negative breadth failing is the C-path's re-entry, at higher levels. Or the count adds and holds: above 70 re-opens the bull case properly |
| One month | The first hike-risk September since 2022: FOMC 15–16 Sep with a live ~30–40% hike tail; the battleground 28,564–29,972 holds; every rally reads as peak-3 territory until the monthly gauge prints through 77.88 | Bull parlay (all four required): a Hormuz corridor date + a framework-only Warsh + the count above 70 + Brent under $83. Bear stack: a hike commitment + 10-yr above 4.72% twice + credit cracking (JNK under 94) + a $95 settle |
| Six months | 26,700–30,500 range with downward drift while the monthly divergence stands; the Warsh axis adds structural rates volatility — a hike-risk Fed against a buyback Treasury keeps the debasement trade alive, with gold the scoreboard; the R3 band (26,906–26,922, narrowed to 16 points) is the framework's structural buy zone if the staircase completes | A-tail: reopening date + disinflation resuming + a credible Warsh framework re-rates through 30,780. D-tail: a China-bank secondary sanction + a 95→101 Brent transit visits 24,500–26,000 |
| Year-end | Conditional on the Aug/Sep monthly closes: base is C-resolution then recovery — a Q4 low near the R3 band, year-end 26,500–28,500 recovering toward 29,000; only a monthly close with the gauge above 77.88 invalidates the three-peak structure | A-tail: 32,000–33,500 (invalidation regime). D-tail: 24,500–26,000 on stacked shocks |
What we got wrong, at full price. The structural short — held through the gap by its own rule — was stopped on the final 30-minute bar of the day: the close printed 29,641.56 against the 29,640 stop, a fire by 1.56 points, for −350 points on a 1% position, the largest single loss of this window. Two things are true and neither is an excuse: the stop was deliberately placed outside the magnet (the doctrine held), and the licence revocation at 61.39 would have forced the same exit at the same close anyway. The framework does not round in its own favour — 1.56 points over the line is over the line. Second: the long-volatility core exited by its own 14.8 protective rule at Thursday's close — the night before the binary it was bought to insure. The rule that protected three weeks of P&L surrendered the event exposure with twelve hours to go; the irony is logged, and the answer is the half-size wing (Option 3), a different structure with its own clock, not a quiet re-buy of a stopped position. Third: the base case called the count "+1 to +3" — it printed +9.90. Direction right, magnitude wrong by three times; the pre-committed threshold made the miss tradeable anyway, which is what thresholds are for.
What worked. The Thursday map, level by level: open 29,481 against a called 29,500–29,560 (19 points shy); the push into the lid; the stack pullback that held by everything but seven points; the chip gap approached and not filled — Thursday's high missed the gap's bottom by 33.76. The bracket's first half banked +124 at 29,554. The S&P traded its called 7,700–7,745 magnet end-to-end. The oil rule ran its full release-and-re-arm cycle in 48 hours, mechanically, in both directions. What was marginal: the VIX range (called 14.6–15.3, closed 14.51 — nine cents through the floor, and those nine cents fired the protective); the oil band (called $85–88, settled ~$89.5 — broken, and the break re-armed the cap). What was null: the S&P fade never confirmed (no rejection close — the condition kept it out of a rising tape); the 7,687 rung missed its fill by 2.89 points and retired by rule; the chip-gap fade was voided by the count before price ever reached it.
Data notes, kept honest. The Thursday NVDA gap on the Nasdaq is confirmed and logged as gap #32 (29,296.77–29,366.42); its S&P twin closed by 84 cents on Thursday's low and was never logged — precision cuts both ways. The NDX high-low feed returned a valid print (+1.98) after three flagged sessions — flag closed. The NYSE summation chart arrived as a weekly render (197.95, a ninth decline); the daily basis is unavailable this capture and is noted, not inferred. Brent's settle is quoted from wires [web], not an exchange print. Gold has no chart feed today; levels are wire-quoted [web]. The sanction institution "to be named by week-end" remains unnamed with the week-end here — that is now a same-day binary alongside the speech.