Wednesday Brief · August 5, 2026 ★ Elite Members

Four Indices. Four Failed Gates. One Pass On The Whole Sheet.

Every index in this scan failed its entry filter this morning, the Dow by 6.39% against a 1.5% band, which is the widest miss recorded in this brief all year. That empties the index side completely and pushes the entire session into single names. SoFi is the only instrument anywhere on the sheet with a passing entry gate, and it also carries the only active institutional convergence and the only live signal tier, on a track record that does not exist yet. ARK Innovation holds the tightest coil at 71% and a deeper record behind it. Bond-equity correlation jumped from 0.36 to 0.51 overnight, so nothing in a long book is hedged today. The session turns on one SoFi price and it is in The One Number.

Date August 5, 2026
Session Wednesday · Captured Pre-Market
Access Elite Members Only
Market Update
Records Paid For With Numbers

Tuesday was the strongest session of the year and two earnings reports paid for it. Palantir rose roughly 29% after second-quarter revenue grew 93% to $1.94 billion and management lifted full-year guidance to $8.15 billion. Caterpillar added 5.6% on the first quarter in its history above $20 billion of sales. The S&P closed 1.79% higher at a record 7,737, the Nasdaq added 2.59%, and the Dow gained more than 900 points to finish above 54,000 for the first time. Monday's record came from a headline Tehran denied within hours. This one came from audited numbers, which is a far better foundation and a far worse entry.

RegimeInflationary Pump
Composite+2 / 9
LayersLqd ✗ · Eq ✓ · Rsk ✓
VIX1 · Low Fear (17)
Macro Bonus+1
LeaderRisk
Bd-Eq CorrStress 0.51
Transition→ Stable

INFLATIONARY PUMP holds at Composite +2/9 with the macro bonus still a +1 tailwind, so equities and risk assets are both switched on, bonds are not, and a long book is no longer fighting the regime. The structural caveat is a single number and it got materially worse overnight. Bond-equity correlation printed 0.51 against a 0.35 stress line, up from 0.36 yesterday, which is the highest reading in this brief in weeks and means bonds and equities are now moving together hard enough that a bond position hedges nothing. Single-name feeds read DISTRIBUTION at Composite +1/9 with a −1 macro bonus and a normal 0.25 correlation, so the indices and the single names below are being scored against two different backdrops.

Yesterday's primary is dead, and it died the way it was written. Tuesday's brief ranked a Nasdaq fade first with the invalidation set at a 4H close above 29,050. The index closed above it, cleared the descending anchor that had capped every rally since mid-July, and ran to 29,757. The fade is off and the structural read that went with it is off too. What replaced it is worse for entries, not better: the Nasdaq now sits 2.83% above its own anchor and has joined the other three indices outside the gate. Two data notes before using anything below. Index readings come from cash CFD and futures feeds captured at 05:56 UTC+2 on bars that had not closed. Anchor prices are derived from the terminal's deviation percentage rather than read directly, so treat the anchor as approximate and the deviation percentage as the number that governs.
Breadth · The Two Feeds Now Disagree

Breadth earns its strip today because the two measurements have split. The IVT panel riding on the index feeds flipped to CONFIRMING at +8/8 with the divergence flag Clean, McClellan at +79.5 and 68.2% of names above their 50-day. The separate universe advance-decline scan, run five hours earlier, still prints a formal divergence warning. Both cannot be describing the same market.

MetricUniverse scanChange vs yesterday
Divergence flagSTILL ACTIVEUnchanged. Price rising, participation lagging.
A/D net+10 on 42 advancingFlipped positive from −4.
A/D ratio1.31Up from 0.90. Back above 1.0.
McClellan−113.9Still negative, improved from −129.6.
Summation+11,851Broadly flat. Long-term breadth intact.
% above 50-day56.8%Back above half, from 47.3%.
% above 200-day59.5%Up from 58.1%.

Every internal moved the right way. More names advanced than declined for the first time in three sessions, over half the universe is back above its 50-day, and the ratio is comfortably above 1.0. But the flag has not cleared and McClellan is still below zero, so the improvement is one day old and not yet a trend. The practical read is narrow: breadth is no longer arguing against the rally, which removes a reason to fade the indices, and it does nothing at all about the fact that every one of them is too far above institutional cost basis to buy. Better participation does not fix location.

The Board · Every Tradeable Setup, Ranked
#TickerDirEntryEdge (WR · Conf · Comp) · Clock
1SOFILong🔒 MembersDev PASS · Conv ACTIVE 80% · Signal 2 live · Conf 4 (3F) · Expanded 21% · No resolved sample
The only passing entry gate on the entire sheet, indices and single names together. It is also the only instrument here with an active institutional convergence at 80% strength, the only one showing a live signal tier rather than a blank, and the only one confirmed across three separate timeframes rather than one or two. Price is sitting almost exactly on institutional cost basis, close enough that the deviation rounds to under half a percent, and a bullish tier-2 signal has fired right at that level after a strong bullish divergence off the flush low in late July. Then the honest problem: the win-rate field reads zero resolved and awaiting results. There is no record here to lean on, because every signal fired in the measured window is still open. High-volatility strict tier on top of that. This ranks first on location and structure, not on proof, and the size instruction reflects exactly that.
2ARKKLong🔒 Members88% win · Conf 6 (2F) · Building 71% · Dev OPEN · Conv Diverged 72% · 32-bar avg
Highest confidence score on the board at 6, the tightest coil anywhere on the sheet at 71% and still loading, and a deep 88% record behind it. The entry gate is open with the next signal qualifying, and a bullish tier-2 carrying nine confirming factors fired off the low a few bars back, which is the highest factor count on any long here bar one. What holds it at second rather than first is the disagreement underneath: intraday bias reads bull while the multi-bar swing structure still reads bear, and institutional anchors read Diverged at 72%, meaning the resistance overhead is spread out rather than stacked. The 32-bar average hold is roughly five and a half sessions, so this is a carry, not a day trade, and the gate is close to shutting on the upside.
3XYLLong🔒 Members80% (5 resolved of 7) · Signal 3 · 10 factors · Conf 4 (2F) · Dev OPEN · Expanded 26% · 21-bar avg
The strongest single signal on the sheet by factor count. A tier-3 bullish signal fired carrying ten independent confirming factors, which is the deepest confluence in this brief and two tiers above anything else on the board. Eighty percent across five resolved signals of seven fired is a usable sample, the entry gate is open, and the anchor sits on a round number that price has respected twice in the past fortnight. Bias reads bull. The swing structure still reads bear and the compression is Expanded at 26%, so there is no coil contributing anything and this is a signal trade rather than a breakout trade. Mid-range tier means standard size is appropriate, which makes it the most straightforward position on today's sheet even though it ranks third.
4BTCUSDShort🔒 Members63% (8 resolved of 8) · Conf 6 (1F) · Building 66% · Dev OPEN · 26-bar avg
The only short on the board and the only instrument on the whole sheet where intraday bias and multi-bar structure agree. Both read bear. Confidence is 6, tied for highest here, compression is Building at 66% and still loading, and the record is complete: eight signals fired, eight resolved, nothing left open, which makes 63% an honest number rather than a flattered one. Price is sitting just under the anchor with the gate open on both sides. The awkward part is the regime. Bitcoin is the risk proxy in the Cantillon layer and that layer reads on, so a short here is being taken against the framework's own risk read. A stacked cluster of bearish divergences overhead through the second half of July is what the setup is leaning on instead.
5HOODLong🔒 Members67% · Conf 4 (2F) · Dev OPEN but closing · Conv Diverged 64% · Expanded 47% · 39-bar avg
Real structure and a window that has already shut. The setup is sound: a bullish tier-2 fired off the mid-July base, price has reclaimed the institutional anchor and is holding above it, and confidence reads 4 across two timeframes. But the deviation is now 2.85% against a 3.0% band on the strict high-volatility tier, and the overnight print has carried price past the level where that gate closes. There is no qualifying entry left at the current price. The 39-bar average hold is the longest on the board, roughly six and a half sessions, which is a long time to carry a high-beta broker into an extended tape. Alert, not a position, unless price comes back to the anchor.
6DGLong🔒 Members83% (6 resolved of 7) · Conf 4 (2F) · Dev OPEN · Expanded 7% · 21-bar avg · Event marker ahead
Best resolved hit rate of any long here at 83% across six settled signals of seven, with bias and the multi-bar structure both reading bull, which only one other name on this board can say. Two bullish tier-2 signals have fired on the way up from the June base. It fell 0.48% on the session, which improves the location rather than damaging it, and the gate is still open. The reasons it ranks last are ordinary rather than fatal: confidence is 4, compression is Expanded at 7% so nothing is loaded, and the capture shows an event marker sitting ahead of price against a 21-bar hold. Confirm the report date before carrying it, because three and a half sessions is long enough to run into one.

If you watch one: SoFi. On a morning when four indices out of four failed their entry filter, the one instrument that passed is worth more than the one with the best statistics. It carries the only active institutional convergence in the brief, the only live signal tier, and confirmation across three timeframes, which is more agreement than anything else here can show. It beats ARK Innovation because ARK's gate is nearly shut and its structural trend still points the other way. What it does not have is a single resolved signal to its name, so the position it deserves is small. The entry band, the trigger and the invalidation are below.

The rule for today, stated once. There is no index trade. Not a long, not a fade, not at any size. All four failed the deviation filter and the misses are not marginal: 6.39% on the Dow, 5.3% on the Russell, 4.88% on the S&P and 2.83% on the Nasdaq, every one of them against a 1.5% band. The Dow reading is the widest recorded in this brief all year and its gate sits roughly 2,500 points below a record close. So the whole session is a single-name session, and the single names are being scored under DISTRIBUTION at Composite +1/9 with a −1 macro headwind rather than the +1 tailwind the indices get. Size accordingly. And note what the correlation is doing: at 0.51 against a 0.35 stress line, a long equity book has no hedge available in bonds today, which is the argument for fewer positions rather than smaller ones.
Elite Members Content

Full Board, Execution Detail & Every Annotated Chart

Enter your Cantillon Research subscriber email to unlock the complete August 5 brief: exact entries, triggers and invalidations for all six ranked setups, the SoFi anchor and the only passing entry band on the sheet, the ARK coil with the price where its gate shuts, the ten-factor tier-3 level on Xylem, the Bitcoin short band and the number that ends it, the exact prices where all four shut index gates reopen, and the annotated IVT terminal chart for every instrument in the scan.

Not a subscriber? Subscribe for Elite access →
Or get the free daily regime read
Includes the IVT Regime Playbook (PDF) + the daily regime read. No spam.
Index readings (US30, SP500, NAS100, RUSSELL) were captured Wednesday 5 August 2026 between 05:56 and 05:57 UTC+2 pre-market from the IVT v13.3 terminal, are sourced from cash CFD and futures feeds, will not match the cash index tick for tick, and must be re-verified against your own instrument before acting; the last bar had not closed on any of them. Single-name and crypto readings (SOFI, ARKK, XYL, BTCUSD, HOOD, DG, MELI, RBC) were captured the same morning between 05:54 and 05:56 UTC+2. Overnight prints shown on the captures were SOFI $18.54 against a capture of $18.68, ARKK $76.49 against $76.79, XYL $122.17 against $122.16, HOOD $93.73 against $93.48, DG $127.29 against $127.29, MELI $1,893.81 against $1,886.56 and RBC $565.60 against $566.08. Entry bands in this brief are set against both figures and must be re-checked against live prices. All timeframes are 4-hour. Institutional anchor prices are derived from the terminal's deviation percentage rather than read directly and are approximate; the deviation percentage is the governing figure. Cantillon Flow reads INFLATIONARY PUMP on the index feeds with liquidity off, equities on and risk on, Macro Bonus +1, Composite +2/9, VIX 1 Low Fear at 17, Leader Risk, Transition Stable, and bond-equity correlation at a STRESS RECOVERY 0.51, well above the 0.35 stress threshold and up from 0.36 on 4 August. Single-name feeds read DISTRIBUTION at Composite +1/9 with a −1 macro bonus, Leader Equities, VIX 1.5 Low Fear at 16 and a Normal 0.25 correlation; indices and single names in this brief are therefore scored against different regime backdrops. The IVT breadth panel on the index feeds reads CONFIRMING at +8/8 with Divergence Clean, Trend +2, 68.2% of names above the 50-day, 71% above the 200-day and McClellan +79.5; the same panel on the single-name feeds reads CONFIRMING at +7/8. The separate universe advance-decline scan shown in this brief was generated 4 August 2026 at 21:46 UTC and reads a BREADTH DIVERGENCE warning still active, A/D net +10 on 42 advancing, ratio 1.31, cumulative A/D line +1,760, McClellan −113.9, Summation +11,851, 56.8% above the 50-day and 59.5% above the 200-day; it covers the Cantillon scan universe rather than the full exchange and is not directly comparable to the index breadth panel. The SOFI win-rate field reads zero resolved signals awaiting results, so no hit rate is claimed or implied for that instrument. The HOOD resolved-signal count is partially obscured on its capture and only the 67% headline and the 39-bar average hold are reported here. MercadoLibre reports after the close on Wednesday 5 August; an earnings print overrides every technical level listed for that name. An event marker appears ahead of price on the Dollar General capture and the date must be confirmed independently before carrying a position through it. Several names show intraday Bias disagreeing with the underlying multi-bar Swing structure; verify the higher-timeframe trend on your own screen before acting on any level here. The Nasdaq short published in the 4 August brief invalidated on a 4H close above 29,050 as written and is closed. Charts are unmodified terminal captures from the session listed and are provided for reference, not as entry instructions. This brief is for informational and educational purposes only. Not financial advice. Past win rates do not guarantee future results.