Payrolls printed 162,000 against a 56,000 consensus — nearly triple — with private payrolls at 127,000 versus 45,000 expected and average hourly earnings running 3.1% year-over-year against a 3.0% estimate. Two days earlier ADP printed 38,000 against 47,000 and Challenger cuts jumped to 52,881 from 33,429. The two labor signals now point in opposite directions and the tape resolved the conflict in favour of the payroll print.
The repricing was selective, and that is the tell. Gold and silver absorbed the entire hit while TLT finished within two-tenths of a percent of unchanged. A market that genuinely believed the Fed had been pushed toward a hike sells duration first, not the debasement trade. VIX at 14.49, up 1.19%, confirms it — no fear was bought, only an inflation hedge was sold.
Underneath the index prints there is a split. SPY and DIA closed red while QQQ and IWM closed green, because memory and semiconductors absorbed the day's entire risk budget: SNDK, MU, INTC, AMD, MRVL, CBRS and ALAB all finished up 4.5% or more. Long-duration software went the other way — GWRE, FICO and EGAN each lost over 16%. Near-term cyclical cash flows were bought and 2030 cash flows were sold, which is exactly what a hot labor print does to equity duration.
The two heaviest dollar-volume sectors were both sellers — Energy at $1.67B and Financials at $1.52B — and the raw advance/decline line finished negative at 5,095 to 5,436. The distribution is in the deepest parts of the tape, not the tails.
More telling: two clean beats were sold. LULU beat EPS by 63% and lost 17.55% on 6.2x turnover; IOT beat by 25%, traded to 45.12 and closed at 40.01, the bottom eighth of its range, on the largest volume in its history. When beats are sold the constraint is positioning, not results — and no card on this page shows that.
Weight the tails over the median: 558 names up 4% against 421 down, and 263 new highs against 148 new lows, argue concentration rather than distribution despite the red breadth score. That read fails if SPY closes below its 21-day average at 766.01.
The 162,000 payroll print against a 56,000 estimate is the whole session. The transmission ran through real rates rather than nominal ones: GLD -0.96% and SLV -1.28% took the full repricing while TLT closed +0.17%. That asymmetry says the market cut its inflation-hedge exposure without moving its rate-path exposure, which is a narrower conclusion than the 'rate-hike bets revived' framing in the headlines.
The confirmation came from crypto, not from bonds — Bitcoin sold off on commentary that the Fed 'really should' raise this month. Cross-asset, that is a debasement-trade unwind, not a growth-scare: USO was almost unchanged at -0.15% and VIX at 14.49 never left the low-14s.
Note the composition of the beat before extrapolating it. Government payrolls swung from -50,000 to +35,000, an 85,000 delta, and labor force participation rose to 61.6% from 61.4%. A rising participation rate with an unchanged 4.1% unemployment rate means the supply side absorbed the hiring, which is the disinflationary reading of the same print — and it is the reason the long end did not sell off.
Trend quality is improving but not yet clean: the blended daily chop reading moved to 62.3 from 51.95, with the Nasdaq leg at 63.9 from 48.8. That is the trend-development zone, not an established trend. The macro read inverts if the long end starts pricing what gold already did — TLT below 82.07 on the next session.
CBRS carries the only hard corporate catalyst among the day's leaders: a 165 MW AI data centre agreement with Compute Nordic in Mikkeli, Finland. Volume confirmed it at 1.6x average on $2.85B — this is a contract, not a technical bounce, and it is why the leveraged trackers CBRG and CBRX printed +20.1% and +19.7% while the inverse CBRZ lost 20.8%.
LULU is the day's most instructive headline pair: the print beat by 63% and the analyst commentary read 'sales reset has more stretching to do' with EPS estimates cut. The event was positive and the structure is not — Stage 4B, relative strength 9. When the event and the structure disagree by that margin, the structure has been winning all year.
FICO is the honest gap in the attribution. A $24.2B-cap name gapped -16.49% on 4.56x turnover and $1.33B of flow, and the snapshot carries no earnings event and no catalyst field for it. The only reading consistent with the session is rate-sensitive mortgage-volume repricing on the payroll print, and that is a hypothesis the data here does not confirm — treat it as unattributed.
CHPT deserves a flag rather than a chase. Volume ran 31.43x average — unprecedented against its 60-day maximum — on a B. Riley note that maintained Neutral and raised the target to $8, while the stock closed at $9.95. The volume is real; the stated catalyst carries a target 20% below the closing price, so the flow is not explained by the news attached to it.
HNGE — the only confirmed volatility-contraction base in the scan, all eight template criteria passed, pivot at 95.57 with volume drying to 0.88x. Invalidation: close below 85.75.
SNDK — relative strength 99, Stage 2A, inside-day break on $26.5B with clear runway to 4.0R. Invalidation: close below 1586.87.
MU — $34.5B of dollar volume is deeper than any other name on the page and the trend structure is intact above a rising 21-day at 943.56. Invalidation: close below 960.60.
BE — the reason Industrials outdrew Technology on dollar volume; relative strength 97 holding a first touch of the 10-day at 223.95. Invalidation: close below 230.60.
| Ticker | Price | Trigger | Target | Stage |
|---|---|---|---|---|
| $HNGE | 91.95 | 95.57 | 115.21 | Stage 2A |
| $SNDK | 1737.53 | 1740.00 | 2046.27 | Stage 2A |
| $MU | 1013.51 | 1017.77 | 1132.11 | Stage 2A |
| $CBRS | 209.75 | 216.90 | 258.04 | Base |
| $BE | 252.77 | 253.30 | 298.69 | Stage 2A |
| $TYRA | 28.65 | 29.56 | 33.21 | Stage 3B |
| $MRVL | 222.74 | 220.90 | 251.78 | Stage 2A |
Leading: Technology +0.70%, Industrials +0.41%
Lagging: Utilities -0.02%, Materials -0.34%, Real Estate -0.72%, Consumer Staples -0.74%, Financials -0.85%, Energy -0.91%, Health Care -0.99%, Communication Services -1.18%, Consumer Discretionary -1.34%
Two sectors green out of eleven is not leadership, it is survivorship. The more useful number is dollar volume: Industrials moved $1.02B against Technology's $955M, so the sector that led on percentage was not the sector that took the most capital — and the two heaviest flows of the day, Energy at $1.67B and Financials at $1.52B, were both on the sell side.
The Technology figure of +0.70% badly understates what happened inside it. SNDK +11.8%, MU +5.93%, CBRS +10.18%, ALAB +9.96% and MRVL +6.58% against GWRE -20.14% and FICO -16.45% nets out to a sector bar that describes neither trade. Memory and AI-compute were bought and enterprise software was sold, inside one sector line.
Industrials leading on dollar volume is the same trade in a different wrapper: BE at +6.57% on $3.59B is AI-power infrastructure, not classic cyclical industrial demand. That means the two green sectors are one position, and they fail together.
Consumer Discretionary at -1.34% is almost entirely LULU's $3.73B. Strip that name out and the sector is unremarkable — which is the invalidation for reading the tape as a consumer-demand problem. The sector read fails if Financials turn green while semiconductors stall: that would mean the rate repricing was being bought rather than sold, and this session's leadership would rotate out.
Every name on this list moved during the 9:30–4:00 session, not after it. LULU, GWRE, EGAN and NX were reactions to Thursday-evening prints that gapped at the open — gaps of -19.4%, -17.41%, -22.64% and +9.97% respectively — and traded out from there.
No distinct after-hours mover is identifiable in this data. The earnings calendar lists only FIZZ for today, with an estimate of $0.592 and no actual reported, and the post-close news flow is limited to a TOP headquarters relocation, an IRSA quarterly print and an OpenAI availability note tagged to MSFT — none of which produced a price move in the snapshot. Friday after-the-bell releases are structurally rare, and this session followed that pattern.
The list itself divides cleanly by attribution. Confirmed: AOUT +43.42% on 15.98x turnover with 34% of its 12.6M float traded, NX on 5.47x, BBCP on 4.44x, TYRA on 3.75x. Unconfirmed: TRBG gapped -34.97% and closed -38.09% on 0.85x turnover — a 38% move on below-average volume is a liquidity event, not a repricing, and XTND printed +32.21% at exactly 1.0x volume.
The leveraged trackers all resolve to their underlyings: AEHG +25.85% to AEHR +13.19%, CBRG +20.06% and CBRZ -20.79% to CBRS +10.18%, and LULG -35.26% to LULU -17.55%. No tracker moved without its underlying present, which means there is no hidden name in this list.
The grade distribution is the story: BBCP at 96, CBRS at 79 and TYRA at 76 are the only A-grades, and they share no theme — concrete pumping, AI compute and biotech. The A-grades did not come from the sector that led.
Relative strength splits the list in two. MU, SNDK, INTC, AMD, MRVL, SNXX, BE and KORU all carry readings of 94 or better with Stage 2A or 3B structures — these are extensions of an existing trend. CBRS at 21 and AOUT at 58 are first moves off broken or flat bases, which is a different holding period and a different failure mode.
The liquidity range spans four orders of magnitude: MU at $34.5B and SNDK at $26.5B against BBCP at $9.1M. BBCP has the highest score on the page and the thinnest tape on it, and that combination is where scores mislead — a 96 grade on $9.1M of flow cannot be sized like a 68 on $26.5B.
Cross-scanner confirmation is thin today. Only CBRS appears with both a graded setup and a disclosed corporate catalyst; the rest are technical. The list invalidates as a group if SPY loses 766.01, since eleven of the twenty are semiconductor-complex names.
The August wage line printed hot — 3.1% year-on-year against 3.0% expected, with July revised up from 0.1% to 0.2% — while the headline payroll count is still unpopulated in the calendar feed. The market spent the week building the opposite case: ADP at +38K, Challenger cuts at 52,881, services employment at 47.8. Weak labor quantity with firm labor price is the stagflationary corner, and the response says so: S&P +0.02%, Nasdaq +0.48% at the print.
Underneath, the bid is narrow and it is not in AI logic. AVGO reported a 3.11% beat, traded $21.2B, and still sits 2.29% below its 10-day average with a relative rank of 38. The actual leadership is semiconductor equipment and memory — AXTI at rank 99, ICHR at 97, AEHR at 97 — every one of them moving 4-5.5% on below-average volume and every one more than 45% off its high. That is a bounce being ranked as leadership.
The earnings tape is punishing beats. LULU beat by 63% and is down 20.46%; OXM beat both lines and is down 17.06%; GWRE beat by 6.11% and is down 15.46%. CURV missed on revenue and is up 7.63%. When four beats are sold and a miss is bought, the market is trading guidance and positioning, not results.
Payrolls at 162,000 against a 56,000 estimate is the whole session. The front end repriced hawkish, but TLT is +0.47% with the long end bid - the market is buying a Fed that tightens into strength, and XLF at -0.96% on a hot growth print is the curve-flattening confirmation rather than a contradiction.
The damage is confined to real-rate assets - GLD -0.82%, SLV -1.21%, miners near -2.5% off the GDXU tracker, crypto proxies near -5.5% - while VIX falls to 14.14. That is a discount-rate move, not risk aversion, and it is why QQQ can hold +0.19% against DIA -0.52%. Semiconductors absorbed everything: SOXL +9.51% implies the index up roughly 3.17%, led by memory and storage in SNDK, WDC and MU, with TSLA at -5.52% the largest megacap casualty on the other side.
The caution is width, not direction. Advancers trail decliners 3,933 to 5,094 at 43.6% green, yet new highs beat new lows 258 to 103 - leaders extending while the median stock is sold. Daily CHOP has risen from 51.9 to 61.5 and only 15 of 90 scanned names clear a 40 score or 1.5x relative volume, with nothing in the semiconductor complex scoring above 48.
Midday reads as a rotation, not a direction. The payroll beat at 162,000 against 56,000 briefly flipped rate-hike odds above 50%, yet TLT is up 0.43% and VIX is down to 14.0 - the bond market is trading the 3.1% wage deceleration rather than the headline count, and that is what lets QQQ hold +0.16% while SPY sits at -0.32% and DIA at -0.50%.
The money is in memory, and the catalyst is disclosed rather than technical. SNDK +9.28%, MU +4.43%, WDC +4.28% and SKHY +4.57% are all moving on one input: TrendForce puts the memory cost of a 256-gigabyte iPhone 18 Pro up nearly 400% year over year. KORU +9.57% carries the same trade offshore. What undercuts it is participation - SNDK produced 9.28% on 0.69 times average volume, and AMAT, the equipment layer, managed 4.53% on 0.36.
The confirmed flow is entirely on the sell side. The ten largest decliners carry $4.53 billion of dollar volume against $235 million for the ten largest gainers. LULU -17.84% and GWRE -20.56% both beat on earnings and both were sold, EGAN makes it three, and TSLA at -6.47% does the rest of the damage to Consumer Discretionary. Breadth is flat at 4,979 to 4,995 while new highs still lead new lows 255 to 115 - concentration, not deterioration, but only just.
Two hours into the afternoon the session has resolved into a single trade with a single flaw. MU and SNDK have absorbed $45.3B of turnover between them — seven times the combined flow through every sector ETF — and both are moving on less than average relative volume. A 10.4% advance in a $228B company on 0.92x turnover is a thin-tape drift, not accumulation, and it is the reason the XLK bar reads only +0.45% while the memory layer runs 4-10%.
The opposite side has the volume. LULU is down 17.3% on 5.3x turnover after a 63.1% earnings beat, GWRE down 21.3% after beating both lines, EGAN down 17.2% after a beat. Three companies exceeded estimates and lost a fifth of their value; IOT cleared by 25% and was paid 2%. The market is not buying realized quarters this week, and the flow that should have gone to those names went upstream to memory companies that reported nothing.
The macro is doing less work than it appears. Payrolls at 162K against 56K estimate looks hawkish and the metals traded it that way — GLD -1.12%, SLV -1.54% — but participation rose to 61.6% and wage growth decelerated to 3.1% year-over-year, so TLT is up 0.25% and XLF is down 0.57%. Banks lagging on a hot jobs print is the strongest evidence the hiking interpretation is wrong.
Payrolls at 162,000 against a 56,000 estimate drove the entire session, but the repricing was narrower than the headline suggests. Gold and silver absorbed it — GLD -0.96%, SLV -1.28% — while TLT finished +0.17% and VIX stayed at 14.49. The market sold its inflation hedge without moving its rate-path exposure, and that is a debasement-trade unwind rather than a hike being priced.
The index split follows from equity duration. SPY -0.48% and DIA -0.60% closed red while QQQ +0.06% and IWM +0.19% held green, because near-term cyclical cash flows were bought — SNDK +11.8%, CBRS +10.18% on a 165 MW Finland data centre contract, ALAB +9.96%, MRVL +6.58%, MU +5.93% — and long-duration software was sold: GWRE -20.14%, EGAN -17.86%, FICO -16.45%.
The internals contradict each other and that is the useful part. The advance/decline line finished negative at 5,095 to 5,436, yet 558 names gained 4% against 421 that lost 4%, and new highs beat new lows 263 to 148. The median stock was sold and the tails were bought — concentration, not distribution. The warning sits elsewhere: LULU beat EPS by 63% and lost 17.55% on $3.73B, and IOT beat by 25% and closed at the bottom eighth of its range on record volume.
CBRS — a real 165 MW contract and an A-grade at 79, but relative strength of 21 makes it a first bounce. Trigger 216.90; stands down below 196.33.
TYRA — the only double-digit gainer bought intraday rather than gapped. Trigger 29.56 on a Stage 3B structure that argues against it.
BBCP — top score on the page at 96, and $9.1M of dollar volume against MU's $34.5B. Trigger is already passed; wait for a pullback that trades, not one that simply drifts on no volume.
NX — +22.23% earnings pivot, but 15% above its 21-day average at 19.90. Trigger 23.23 only after the extension resolves.
HNGE pivot at 95.57 is 3.46% away and the base is 3.3% off the 52-week high; the trigger and the breakout to new highs are effectively the same event.
CHPT — 31.43x average volume, unprecedented against its 60-day maximum, on a note that maintained Neutral with a raised target of $8 while the stock closed at 9.95. The flow is 20% above the catalyst that supposedly caused it, and price is 56% above the 21-day average.
IOT — beat by 25%, printed 45.12, closed 40.01. Close strength 0.12 and below VWAP on the largest volume in the name's history.
TRBG — -38.09% on 0.85x turnover. A move that size on below-average volume has no participation behind it and no reliable level to trade against.
ALAB — +9.96% but closed at 311 against a 321.65 high and finished below VWAP, giving back a third of the day's range into the bell.
SOXL — +9.57% and still 3.9% under the 21-day average at 121.69, in Stage 3B, with a 9.45% ATR compounding decay on any sideways resolution.
Watch whether the long end follows gold. TLT held +0.17% while GLD fell 0.96%; if TLT breaks 82.07, the hawkish repricing has finally reached duration and the software selloff extends.
Next week's inflation data against ISM Services Prices at 72.6 and wages at 3.1% YoY — the two price indices above 71 are what sustain the hawkish path. A services-price print below 70 unwinds it.
Whether beats keep getting sold. LULU at +63% surprise and IOT at +25% both closed sharply off their highs; a third instance next week makes it a positioning regime rather than two idiosyncratic guides.
Confirmation or rejection of the memory bid. MU, SNDK and CBRS need follow-through on comparable volume; a lower-volume second day marks today as a one-session rotation.
FICO's -16.49% gap remains unattributed in this data. If a disclosure emerges, it reprices the whole rate-sensitive financial-technology group; if none does, it was flow.
Breadth resolution — the 558-up-4% versus 421-down-4% split against a negative A/D line cannot persist. Whichever side converges determines whether this is concentration or narrowing.
This brief is 24 hours old
Today's setups and tape reading are live now for subscribers.
Get today's brief