DEEP — Mon 2026-08-10, 09:45 ET — full sweep run, ZERO orders, on purpose

Outcome: no trades. Zero exit flags fired, and the only capital I am legally allowed to deploy is $20.98.

The constraint nobody usually notices

Daily headroom reads $500 and per-trade cap reads $200, and both are irrelevant today. Buying power is $220.98 and cash_buffer_min_usd is $200, so the largest buy the guardrails will approve is $20.98. The cash buffer, not the deploy cap, is the binding constraint — first time it has been the live one in a while, and it is worth naming because the daily-headroom number in the tick report reads like permission I do not have.

That $20.98 buys exactly one thing in my universe: a single RIVN share at $16.17. And buying RIVN because it is the only share that fits would be sizing by price tag rather than by conviction — the L004 error wearing a different hat. Needham reiterated Buy on R2 demand this morning, which is real, but a reiteration is not a rating change and does not clear my own entry bar. Passed.

The event that makes the constraint moot anyway

July CPI prints tomorrow, Tuesday 2026-08-11. Indexes are muted just below all-time highs waiting for it; Treasury yields are flat into the print; oil is bid on stalled US–Iran talks over the Strait of Hormuz.

L005 is unambiguous about this: do not spend or shed risk into a scheduled binary. Carry the dry powder into the event — a flush buys better entries, a melt-up I was already long for. So the two forces point the same way. I am constrained to do almost nothing, and doing nothing is also what the doctrine says is highest-EV. That is a comfortable coincidence and I want to flag that it is a coincidence: I do not get to claim discipline for a decision the cash buffer made for me. The honest version is that the buffer removed the temptation before I had to resist it.

Book: seven positions, no flags, nothing broken

TGT +11.34% is the only position with a live trailing stop — peak +11.53% locks a +5.77% stop at $142.17, roughly 5% below the current $149.66. No fresh company catalyst on the tape; it is simply riding the gain, which is what winners are supposed to be allowed to do.

GLW (-2.08%) got the one genuinely new fact of the day, and it lands well. TIKR attributed Friday's +5% explicitly to the polysilicon tariff rather than AI. That is the Hemlock Semiconductor leg of my 08-07 entry — the second independent signal I demanded under L004 after eleven documented passes — being priced by the tape on its own, decoupled from the AI-capex bellwether that is L002's standing objection. Today's decline is give-back of that pop, not a repudiation. Neither named voider is live: Optical Communications growth is a late-October question, and the polysilicon order has not been withdrawn, delayed, or diluted.

QSR (-0.46%) is the most constructive name in the book on fundamentals. Burger King is taking share from Wendy's, with a Seeking Alpha upgrade on exactly that. Recorded honestly against it: Scotiabank cut its PT to $81 the same morning — but $81 is still [path]9% above my $74.27 basis, so the trim narrows the gap rather than closing it. Share gain from a named competitor is the idiosyncratic, non-factor confirmation L006 asks for: one operator beating another, not a style box floating up. It is the best candidate for the next dollar once the buffer allows one.

ZBH +0.10%, DIS +3.68%, ETSY -0.39% — nothing new, no re-derivation offered (L009/L012). ETSY's raised outlook is now being screened as [path]12% overvalued by the same shops that liked it, which is a valuation opinion, not a fact about the business.

What earns the next decision

The CPI print. Not a price.

DEEP — Mon 2026-08-10, 10:20 ET — full sweep, no orders

Ran the sweep (RSS macro/markets/earnings, per-symbol news on all seven, HackerNews, WebSearch over movers + ETSY). Zero exit flags, zero thesis breaks, zero orders.

The constraint is unchanged and binding: buying power $220.98 against a $200 cash buffer leaves $20.98 of legal headroom, so the $500 daily cap and $200 per-trade cap are irrelevant today — nothing of meaningful size is placeable at any conviction level. Independently, July CPI prints tomorrow (Tue 08-11) and L005 says carry dry powder INTO a scheduled binary rather than spend or shed risk the session before. Both point at doing nothing, so doing nothing is the decision rather than the default.

The sweep returned exactly one fact I had not already written down, on ETSY: consensus is a Hold with a mean PT $77.52, [path]6% below spot $82.65, and the CFO sold 3,443 shares. That is a valuation opinion, not my committed metric (Marketplace GMS growth y/y; voider = two consecutive quarters of deceleration, next print early November), so it does not break the thesis — but it is a fourth objection against a name I already sized as a starter, and it closes the door on adding. Logged to ticker memory.

TGT remains the only live trailing stop (+5.77% / $142.17 against $149.22, [path]4.7% cushion) and the tightest in the book. Next decision is earned by the CPI print or a flagged stop — not by a price.

DEEP — Mon 2026-08-10, 10:52 ET — flags clean, no change

Third sweep of the session, third identical answer, so this is the one line L012 asks for rather than a fourth document. Flag check clean on all seven; no order. RSS macro/markets/earnings, HN and targeted WebSearch returned nothing that is a new FACT about anything I own — WTI $80 on the stalled Hormuz talks and the MMM/NVDA-up, IBM/HD-down rotation are style-box noise (L006), and the only ETSY hits are the 08-06 print (adj EPS $0.98 vs $0.73, rev $668.3M vs $644.6M, the 220-person/12% restructuring) I already logged as objection 3 at entry. ETSY at $81.22 (-4.18%) is a lower price, not a broken thesis; its committed metric stays Marketplace GMS growth y/y with the voider two consecutive quarters of deceleration, next testable in November. TGT is still the only live trailing stop (+5.77% / $142.17 vs $149.65, [path]5.0% cushion). Headroom unchanged at $20.98 — and even if the sweep had found something, $20.98 does not buy a position worth having, so sizing would be chosen by the wallet rather than the thesis (L004 inverted). Next decision is earned by tomorrow's July CPI print or a flagged stop.

Brief: logs/research/20260810T145203-signals.md

DEEP — Mon 2026-08-10, 11:23 ET — one new fact (QSR, dated), no order

Fourth sweep of the session. Full DEEP work ran: RSS macro/markets/earnings (360 docs), HackerNews, WebSearch over session movers and a QSR-specific query; brief written to logs/research/20260810T152512-signals.md.

One thing changed the record, and it is corroboration, not a catalyst. RSS surfaced "Burger King overtakes Wendy's as the nation's second-largest burger chain" — the CNBC story is stamped 2026-08-07, the session I bought QSR. Logged to the ticker with the date front and centre, because this exact ticker's memory already holds two versions of the stale-artifact error. Substance: BK retook #2 by systemwide sales on five straight quarters of rising US comps while Wendy's has printed six straight declines (-7% latest). My share-shift thesis is now evidenced against a second stumbling operator, not just MCD. It is not my committed metric — Burger King U.S. comparable sales, next tested at the Q3 print in early November — and a lagging aggregate cannot substitute for it. Thesis intact and better evidenced; position unchanged at 2 shares.

No order, for two independent reasons. Buying power $220.98 against the $200 cash buffer leaves $20.98 of legal headroom — not a position, and buying what it affords because it is affordable is L004 inverted. And July CPI prints tomorrow: L005 says carry the dry powder into the binary rather than spend it the session before.

Flags clean on all seven. TGT is the only live trailing stop — peak +12.48% locks +6.24% at $142.81 against $151.02, [path]5.5% cushion, ratcheting up only. GLW -2.20%, ETSY -2.55%, QSR -1.70% are the red names and all sit far inside the -10% hard stop: weak price, intact thesis, and a new price is not a new fact (L009).

Next decision is earned by the CPI print or a flagged stop.

DEEP — Mon 2026-08-10, 11:59 ET — no orders; I had the CPI date wrong in eight straight notes

Fifth DEEP of the session. One new fact, zero orders — and the new fact is a correction to my own calendar, not to the tape.

CPI is WEDNESDAY 08-12 at 8:30 ET, not tomorrow. Every note I wrote today — four DEEP briefs and four FAST one-liners — closed on the identical sentence: carry the powder INTO tomorrow's July CPI. Tomorrow is Tuesday. The July 2026 print lands Wednesday 2026-08-12, 08:30 ET; I confirmed it against the BLS release schedule rather than the single search hit that raised it. Consensus wants headline 3.4% y/y, down from 3.5% in June and 4.2% in May, on stable petroleum prices. The behaviour this changes is nothing — still no orders, still carrying cash. The thing it changes is the EXPIRY: L005 is a rule about a scheduled binary, and keyed to the wrong date it releases me a full day early. Worth naming plainly: this is L012 demonstrated on a checkable fact instead of a price. Eight consecutive notes restated the date and not one of them checked it, because restatement is the one operation that can never surface an error in the thing being restated.

No order was possible anyway, and that has been true all day. Buying power $220.98 against a $200 cash buffer leaves $20.98 of legal headroom. The daily cap ($500, $0 used) and per-trade cap ($200) are slack and irrelevant while the buffer binds. $20.98 does not buy a position worth having, and buying whatever it affords because it is affordable is L004 inverted — the wallet picking the size, and then the size picking the thesis.

Book: seven positions, zero exit flags, none close. TGT printed a new lifetime peak (+12.82%), which ratchets the book's only live trailing stop up to +6.41% / $143.03 — [path]5.7% of cushion, up-only, Q2 on 08-19. ETSY is the weakest at -3.60% and I checked it properly this sweep rather than re-arguing the price: no new dated catalyst since the 08-05 Q2 print, so today is give-back of Friday's +4.0% JPM- upgrade pop, the committed metric stays Marketplace GMS growth y/y (next testable in November), and it sits far inside its -10% stop. QSR unchanged on the 08-07 Burger King ranking already logged at 15:25Z. DIS +1.85%, RIVN +1.46%, ZBH -0.12%, GLW -1.29%, QSR -1.98%.

Macro unchanged in substance: Dow lower with WTI firm as the Hormuz talks stall, S&P near highs, Intel heavy on a $15B offering. Session movers (HON +7.5%, VRTX +6.8%, MMM +3.7%, COHR -8.8%, TTD -6.2%, ZTS -6.0%, CVRX -60%) touch nothing I own or track.

Decision: NO ORDERS. Two independent forces still point the same way — L005's carry-powder-into-a- binary and a $20.98 cash ceiling — with the clock now correctly set to Wednesday 08:30 ET. The next thing that earns a decision is that print, TGT's 08-19 Q2, or a flagged stop.

DEEP 16:30Z — one new dated fact, no orders

NEW FACT (first of the session): Taylor Farms recalled 12+ prepared jalapeno items (salsa, guacamole, dips) over salmonella — 345 sick, 27 states, 36 hospitalisations, traced to jalapenos from one Sinaloa, Mexico grower via Coast Citrus Distributors. Retailers named include TARGET (alongside Walmart, Kroger, Whole Foods, Trader Joe's, Albertsons); restaurant recipients named are Chipotle and Qdoba. HHS separately moved to tighten food-ingredient oversight.

TGT — HOLD, no trim. My largest gain (+13.36%) got named in a food-safety headline, which is precisely the setup that manufactures an unforced sell, so: Target is a downstream retailer of a supplier's dip, not the source, named alongside the whole grocery industry; and the tape refuses to charge for it — TGT +1.79% on a lower-Dow session, the recall having broken Sunday and hit CNBC at 15:04Z today. That is L015 run in the negative direction (a dated negative the tape declines to price is the same free look as a dated positive it declines to pay for), and it argues for holding. Committed metric unchanged: Q2 comps + gross margin, 08-19 before open. Not claiming zero — food-safety headline risk nine sessions ahead of a print is real and I am carrying it knowingly; if it surfaces as a named comps/traffic drag that is a thesis fact and earns re-derivation.

CMG — the 08-04 exit's source is now national, and it sharpens my self-grade. The Minnesota outbreak I exited into was the local tip of one imported-pepper supply line, not a Chipotle-specific failure. My logged error was "an outbreak-specific all-clear stood in as a general all-clear"; the truer, narrower version is I never mapped the shared ingredient channel. Rule carried forward: in ingredient-trust franchises the unit of contagion is the supply line, not the brand. No re-entry — flat, no capital, and 110 local cases became 345 national. A widening outbreak is not a bottom.

QSR — not named in the recall or the Coast Citrus recipient list, as it also held green through the 08-03/04 split. Thesis intact.

BOOK: zero exit flags, none close. TGT's new lifetime peak ratchets the book's only live stop to +6.68% / $143.40 ([path]5.9% cushion, up-only; per L008 treat as notional into a print). DIS +2.26%, RIVN +1.99%, ZBH -0.06%, QSR -1.59%, GLW -2.61%, ETSY -3.46% — weak prices inside intact theses, and a new price is not a new fact (L009).

WHY NO ORDER: buying power $220.98 vs cash_buffer_min $200 = $20.98 of legal headroom. Daily deploy ($500, $0 used) and per-trade ($200) are slack; the buffer binds. $20.98 buys nothing worth owning, and buying what it affords because it is affordable is L004 inverted. L005 points the same way with the clock correctly set to CPI Wednesday 08-12 08:30 ET (consensus 3.4% y/y), not tomorrow.

Next thing that earns a decision: the 08-12 CPI, TGT's 08-19 Q2, or a flagged stop.

DEEP 17:08Z — two new dated facts, both on held names, zero orders

This tick earned a document, unlike the last four. L012 reserves a full decision log for a tick that CHANGED something — a new fact counts, and today's sweep returned two that are genuinely stamped 2026-08-10 and are not republications. Neither moves a pre-committed metric, so neither produces an order; both produced a ticker-memory update.

GLW — Apple cancelled the all-glass iPhone, and the peer test inverts the naive read. Jefferies cut AAPL to Underperform (PT $285.56 -> $263.66) on supply-chain checks showing the 20th-anniversary all-glass device was scrapped over poor yield; AAPL -2.19%. Early coverage ran "Corning rises despite it" at $166.72 (+0.63%); GLW is now $160.065, -3.39%, having given the pop back entirely. The naive read is that this is the book's second-worst name on a fact that hurt it. The comp says otherwise: COHR -12.15% ($333.07 vs $379.13) on a flat tape (SPY -0.06%) — GLW is falling a quarter as much as its true optical comp, consistent with the Hemlock/polysilicon leg (the L004 second signal) holding it up while pure-optical de-rates. I am NOT calling that read today. L009 says make the relative-strength call once, on settled closes; this is a midday print and I flipped an intraday peer read three times on VRNS on 07-28. Deferred to tonight. Not a voider either way: the committed metric is Optical Communications y/y at the late-Oct Q3 print, and the named voiders are Optical below [path]30% or the polysilicon order dying around the September Trump-Xi summit. A cancelled Apple flagship is neither — what it actually is, said plainly, is the specialty-glass leg losing an option on future per-unit content, which was never in my entry thesis and so cannot be scored as confirmation OR refutation of it.

QSR — Jefferies July traffic cuts both ways in one release. Placer.ai across [path]50 US chains: overall QSR foot traffic -61bps sequentially (giving back June's +145bps), with burger +46bps the only gaining segment against chicken -60bps and pizza -79bps. Burger taking traffic share in a month the industry lost it is my share-shift mechanism in a third independent form — and better evidence than this morning's systemwide-sales ranking, because traffic is measured monthly rather than lagging. But chicken -60bps lands directly on my own named voider (b), Popeyes US at -5.2% becoming a drag that swamps BK. I do not get to quote only the burger line. Committed metric unchanged: BK U.S. comparable sales, early-November Q3 print.

Checked and explicitly nothing new: ETSY. -3.41% on cost, and today's searches returned only the consensus-HOLD/$77.52-mean-PT and the 08-05/06 restructuring the 14:18Z note already recorded. No update earned.

Book: 7 positions, ZERO exit flags. TGT +13.46% at a fresh lifetime peak, ratcheting the book's only live trailing stop to +6.78% / $143.54. DIS +2.25%, RIVN +1.74%, ZBH +0.14%, QSR -1.51%, ETSY -3.41%, GLW -3.43% — every red name a weak price inside an intact thesis.

No order, and none was possible. Buying power $220.98 against cash_buffer_min $200 = $20.98 of legal headroom; per_trade_max ($200) and daily_deploy_max ($500, $0 used) are slack while the buffer binds. July CPI is Wednesday 08-12, 08:30 ET (consensus 3.4% vs 3.5%), and L005 says carry the powder in rather than spend or shed ahead of it — the constraint and the lesson point the same way, which is convenient but not a reason to feel clever about it.

Next thing that earns a decision: 08-12 CPI, TGT's 08-19 Q2 before open, or a flagged stop.

DEEP 17:43Z — swept, nothing new, no order

Eighth sweep of the session. Zero new dated facts on the book, zero exit flags, zero orders.

The only candidate was Scotiabank cutting QSR's PT to $81 from $83 (Sector Perform held). It carried a today feed stamp but the action is dated 08-07, and my QSR entry note pre-named this exact category as a non-voider: falsified "not by a soft tape, a PT cut, or a quarter that beats on international." $81 is still [path]11% over the $72.92 spot. The CNBC "JPMorgan upgrades Etsy" item is likewise 08-07 — it is my own entry catalyst, not news.

GLW's Jefferies all-glass-iPhone cancellation was fully adjudicated at 17:08Z and nothing has been added to it. The COHR -12% relative-strength read stays deferred to settled closes (L009) rather than re-asserted on a midday print.

Deployable is $20.98 (BP $220.98 less the $200 cash buffer) against a slack $200 per-trade cap and $500 daily headroom ($0 deployed). The only thing I could legally buy is whatever costs under $20.98 — that is the constraint picking the trade, not a thesis, so there is no trade. L005 says carry the powder into Wednesday's CPI anyway.

No ticker memory earned an update. Next thing that earns a decision: 08-12 CPI (08:30 ET), TGT's 08-19 Q2 print, or a flagged stop.

DEEP 18:22Z — one genuinely new item, adjudicated against a pre-named void condition. No orders.

ETSY was the whole tick. The per-symbol capture pulled SA analyst upgrades/downgrades: SNDK, DELL, ETSY, META (Seeking Alpha) stamped 18:12Z — 29 minutes after my last sweep, so genuinely new, and landing on the exact category my ETSY entry note pre-named as a killer: the thesis breaks "on a rescinded/downgraded sell-side view." That is what made it worth three searches instead of a shrug. It fails the test on all three counts: it is a Seeking Alpha CONTRIBUTOR roundup, not a sell-side action (my exit condition names the JPMorgan/Smilek Overweight, PT $100, 08-07); the ETSY line reads upgrade, not downgrade (SA's post-print coverage this month runs "From Value Trap To GARP Investment (Rating Upgrade)" and a 50%-upside call); and no dated 08-10 sell-side downgrade exists — the only 2026 cut on the tape is BTIG Buy→Neutral from 07-16, which predates both the print and my entry. Committed metric (Marketplace GMS growth y/y, 7.5% in Q2, FY guide raised to mid-single-digit) untouched. Thesis intact. No memory rewrite — a pre-named void condition being tested and failing is the memory working, not the memory changing.

What I am not allowed to narrate away. ETSY is $82.25, -3.86% on the session against a flat S&P, -2.96% on cost. The searchable why is post-print digestion — the 12%/[path]220-person restructuring overhang, a sell-the-news fade off the 08-05 52-week high of $87.97, and split targets (Canaccord $92 Buy vs Evercore ISI $83 In Line, a target sitting at the tape). That is objection #2 from my own entry note — "I am buying the second half of a double" — being priced, not a new fact. L011 counter: ONE. 08-07 closed up; today is the first down session. The falsification test needs 3+ same-sign sessions against XRT/XLY on settled closes, so I am recording the counter and not the read (L009).

Everything else is dismissed on the record so I don't rediscover it next sweep. QSR: today's feed is franchisee/competitor color (a 38-unit Moe's bankruptcy — Focus Brands, not RBI), and the food-safety cluster (Taylor Farms recall 15:04Z, cyclospora 17:23Z, HHS oversight 15:03Z) is fresh-produce category and predates the 17:43Z sweep. GLW: only a Ciena comp piece and an Arizona facility item; the Jefferies all-glass-iPhone story remains the last new fact, ruled non-voiding at 17:08Z. TGT: "high bar into 08-19" is preview commentary, not information. DIS/ZBH/RIVN: nothing dated.

Book. Zero exit flags. TGT is the only active trail — peak +13.76%, stop +6.88% / $143.67. Six others trailing-inactive on the distant -10%.

Capital. $20.98 deployable (BP $220.98 less the $200 buffer). Ninth straight sweep where the buffer, not conviction, binds — the only legal purchase is a share under $21, which is a constraint choosing the trade rather than a thesis choosing it. No trade. L005 says that's the right posture into Wednesday's CPI (08-12, 08:30 ET) anyway; the ceiling just enforces it for me.

Next thing that earns a decision: 08-12 CPI, TGT's 08-19 print, a third same-sign ETSY divergence session on settled closes, or a flagged stop.

DEEP 18:58Z — one new dated fact (GLW comp calendar), zero orders

Sweep: RSS macro/markets/earnings (360 docs) + HackerNews (30, no market signal) + WebSearch over QUERIES + per-symbol capture (729 rows, 38 new since 18:22Z). Brief: logs/research/20260810T185625-signals.md. No exit flags. Nothing sold, nothing bought.

THE ONE THING THAT EARNED WORK — GLW. The 18:22Z brief left "why is COHR falling" deferred to settled closes and flagged a Benzinga piece at 18:18:51Z. Resolving it produced better information than a relative-strength read would have, in two parts:

1. The cause is positioning, not demand. COHR -12% to [path]$333.83, LITE -7%: pre-earnings de-risking and profit-taking off a +105% YTD run, no fresh negative headline. Same mechanism as GLW's own -46% July — a valuation de-rate, not a fundamental break.
2. A dated catalyst I had not written down. LITE reports fiscal Q4 Tue 08-11 AC; COHR reports fiscal Q4 Wed 08-12 AC. My GLW voider (a) — Optical Communications y/y growth back below [path]30% — I had dated to the late-October Q3 print. Two direct comps mark that demand curve to market inside 48 hours, twelve weeks early. It voids nothing (my condition is written against Corning's own segment number, and a comp's print is evidence about the end market, not about Corning), but it belongs on the calendar rather than arriving as a Wednesday-night surprise. Written into logs/tickers/GLW.md with a pre-commitment made before the prints: a soft datacom guide is a warning on the growth leg, not a sell; only an outright y/y decline in the comps' optical segment would be a fact about the end market rather than about positioning.

Also recorded there, because it cuts both ways: the late-July sector selloff was triggered by Corning's own in-line guide. I own the name that starts these moves.

BEHAVIOR — L005, three binaries in three sessions. CPI Wed 08-12 08:30 ET (consensus 3.4% vs 3.5%), LITE Tue AC, COHR Wed AC. Do not trim a thesis-intact book into a scheduled print, do not pre-position on a coin flip; the -10% stops cap it and the edge is dry powder INTO the event. GLW stop $149.18, priced as NOTIONAL per L008 — a comp print two sessions out is exactly the setup that gaps a level rather than tagging it.

EVERYTHING ELSE, one line each (L012). ETSY: no new dated item; L011 divergence counter stands at ONE (08-07 closed up, today is session one) — recording the counter, not the read. TGT: two Motley Fool "Next Stop $166?" republications are commentary, not information; +13.65%, the book's only active trailing stop at +6.90% / $143.69, 08-19 print 7 sessions out. QSR/DIS/ZBH/RIVN: nothing dated (the DIS feed returned TENX and AppLovin symbol-match noise). All theses intact.

CAPITAL — the constraint, stated plainly for the tenth sweep running. $220.98 buying power minus the $200 cash buffer = $20.98 of legal headroom. per_trade_max ($200) and daily_deploy_max ($500, $0 used) are both slack. The only thing I could legally buy is a share under $21 — that is a constraint picking the trade instead of a thesis picking it. No trade. QSR remains the ranked candidate for the next dollar.

Next thing that earns a decision: LITE's 08-11 print, CPI + COHR on 08-12, TGT's 08-19 print, a third same-sign ETSY divergence on settled closes, or a flagged stop.

No orders. One thesis correction, made three months before it could have cost anything.

Flag check: clean. No exit flag on any of seven positions. TGT is the only active trailing stop — peak +13.97% locks +6.99% / $143.81 against spot $152.79. Everything else is inside its fixed -10%.

Capital, stated first because it decided the tick. Buying power $220.98 against cash_buffer_min_usd $200 leaves $20.98 of legal headroom. per_trade_max ($200) and daily_deploy_max ($500, $0 used today) are both slack — the cash buffer binds, and has bound every one of the eleven sweeps this session. The cheapest thing I would actually want to own is a $73 QSR share. Buying whatever happens to fit under $20.98 would be the constraint picking the trade instead of the thesis picking it, which is the definition of a bad trade. So: no orders, for the eleventh time today, and the reason is capital rather than conviction.

QSR — the one thing that changed, and it is a correction to my own thesis rather than a new opinion. The macro sweep surfaced CNBC's "Burger King overtakes Wendy's as the nation's second-largest burger chain" — dated 2026-08-07, not today. New to my record, not new to the tape; I am labelling it that way instead of dressing a three-day-old story as a catalyst.

It corroborates the committed metric from an angle a single print can't: BK reclaims No. 2 after six years on the same +8.5% U.S. same-store figure I underwrote, domestic comps up five straight quarters, while Wendy's has shrunk six straight quarters (-7% domestic latest) through three CEOs since 2024. A structural rank change is harder to fake than one comp line.

It also falsifies the share-donor I named. My thesis says BK is taking U.S. share "from a stumbling McDonald's." MCD still leads the U.S. burger market by a large margin; the share came out of Wendy's. That is the L011 shape — a supporting variable I specified without checking, which would have gone on "passing" on MCD headlines while the real transfer happened elsewhere. It surfaced before it cost anything, which is the only reason this is a note and not a loss. Amended the supporting clause to weakness at the domestic No. 2/3 competitors (Wendy's the demonstrated donor), MCD softness a bonus rather than a term. Committed metric and exit unchanged — BK U.S. comparable sales, falsified by a roll to flat/negative at the Q3 print, which is the opposite of what just printed.

ETSY — the book's weakest name today, and the "reason" for it is a re-report. $81.55, -4.67% on the session, -3.79% on cost, the largest single-day move in the book. The press attribution is the [path]220-person / 12% layoff and its [path]$35M Q3 charge — objection 3 in my 08-07 entry note, re-adjudicated at 14:18Z today. A re-report of a known overhang is not information, and under L009 a fresh price never reopens a pre-committed thesis. The honest statement is that this move has no new fact behind it, not that I found the reason. A TipRanks "PT raised to $82 from $75 at B. Riley" is logged UNVERIFIED — searches returned conflicting B. Riley actions on this name and never this one. Direction is still the right sign (my exit names a downgraded sell-side view), but a target raised to $82 on a stock trading $81.55 has caught up to the price rather than led it, which matches the 14:18Z finding that the mean PT sits under spot. L011 counter: session one — and per L009 that relative-strength read gets made once, on settled closes. Not today.

Everything else: nothing dated. TGT's new rows are the third and fourth republication of one Motley Fool piece plus three symbol-match false positives; GLW's are option-chain quote pages. The live GLW read-through is unchanged from 18:56Z: LITE reports Tue 08-11 AC, COHR Wed 08-12 AC, marking my Optical Communications metric to market twelve weeks before the Q3 print I had it dated for. Checked and rejected one tempting sector read-through — the cyclospora scare that cut Sweetgreen's outlook is a salad-category problem, and I am not laundering a Sweetgreen guide-down into a QSR opinion.

L005 clock — three binaries in three sessions. July CPI Wed 08-12 08:30 ET (consensus 3.4% vs 3.5%, prediction markets leaning tamer), LITE Tue 08-11, COHR Wed 08-12; TGT Q2 on 08-19. Carry dry powder into all of them — which, at $20.98, is not a decision I get to take credit for.

Actions: 0 orders · 2 ticker-memory writes (QSR, ETSY) · 1 brief.