Watchlist scan — Financebot

Mode: full. All 129 live names repriced against a 1-year daily series. All five sleeve vacancy screens run. Catalyst calendar pulled to 2026-10-01. News sweep on the movers.

Prior: watchlist-scan-2026-08-18.md (light — no sleeve screens, no Watchlist.md edit) · watchlist-scan-2026-08-11.md

Market context

	Level	1wk	1mo	3mo
SPY	772.23	+1.3%	+5.9%	+2.1%
QQQ	720.42	+1.3%	+8.9%	−2.4%
VIX	14.46	−9.7%	−30.0%	−5.6%
10yr (^TNX)	4.67%	−0.6%	+1.0%	+4.8%

The read is correct, and the shape of the rally matters more than its size. The index did +5.9% in a month. That number understates what happened, because the rally was not broad — it was a single-sector re-rating of software and cybersecurity, and it is the third act of the pattern the regime note already describes.

This is Act III — the SaaSpocalypse unwound

Knowledge/Market/regime.md (as-of 2026-08-04, decay fast, recheck 2026-09-13) frames 2026 as two opposite routs: software crashed in March on AI-displacement fear; AI-capex crashed in July. Neither act covers what happened in August. IGV has now climbed ~42% off its April low, and the last three weeks completed the round trip on the exact thesis that caused Act I. The note is not wrong — it is incomplete, and its recheck is now the binding item, not a formality.

The August leg was driven by earnings, one name at a time, all pointing the same way:

Date	Name	Event	Move
Aug 6	TEAM	FQ4 adj EPS $1.87 vs $1.50; cloud +31% YoY	+35% in one session, +76% in 30d
Aug 26	CRM	FQ2 rev $11.35B (+10.8%); adj EPS $5.90; cRPO +14% cc; Agentforce ARR >$1.5B, +240% YoY; FY guide raised to $46.25B	+14% after hours, +21.5% on the week
Aug 26	CRWD	FQ2 adj EPS $0.31 (beat $0.02); rev $1.47B +25%; FY raised to ~$6.0B	+15%, +18.6% on the week
Aug 26	OKTA	FQ2 reframed identity as the security layer for AI agents	+20–22%, +28.2% on the week

CrowdStrike's CEO gave the sector its new sentence: "AI is driving more cyberattacks. AI is driving more cyber spending." AI stopped being the thing that kills software and became the thing software sells. That is a narrative inversion, not a multiple drift, and it is why the moves are 15–35% in a session rather than 3%.

What this scan is therefore mostly about is selling, not buying. A rally that re-rates the sleeve already held does not create entries. It creates trim decisions and stale entry zones — and this file has both, in quantity.

⚡ ACT NOW

A. The names still inside their entry band

⚠️ CORRECTION (applied 2026-08-27, after this section was first written): FOUR names were in zone, not three. This sweep missed KLAC at $183.54. Its re-entry rule was written as prose — "re-enters below $190" — not as a numeric Entry $X-Y field, so the zone parser never saw it. A rule written as a sentence is invisible to the scan that exists to surface it. KLAC [7.5] is the second-highest-conviction in-zone name. Check ADI (Add <$290), NXPI (Add <$200), GOOGL (strong buy <$290) and TEAM (Add <$68) for the same defect.

INTU has since become a fifth, on a re-derived $305-350 band from analyze-2026-08-27 — the first /analyze-from-before run, which upgraded it [7.0] → [7.5] and retracted the prior file's central claim that FCF was falling (it rose 41.7%; the "$5.23B TTM" figure was a single quarter). See [[pitfall-single-quarter-fcf-read-as-ttm]].

Out of ~90 live names with a stated zone, three were found by this sweep — plus KLAC, missed for the reason above. Two of the three found are in zone because they fell while the market rallied.

Ticker	Conv	Price	1wk	1mo	Zone	Catalyst	Recommendation
AVGO	[8.5]	$369.87	+1.6%	−0.1%	✅ IN $360-420 — now LOW-band (was top edge on 8/11)	Q3 print Sep 2	🟡 The highest-conviction in-zone name in the file, and it sat out the entire rally. It moved from the top edge to the bottom third of its own band without the thesis changing. But the print is six days away — a buy now is a buy into an event. Either act before Sep 2 deliberately, or wait for it deliberately. Do not drift.
APP	[6.0]	$311.17	+0.8%	−22.1%	✅ IN $290-330, mid-band	none in window	🟡 Still in zone, and now for a worse reason. −22.1% in a month and −57.6% off its 1-year high while software rallied 20%+. The 8/05 file's issue (first-ever miss vs its own guidance midpoint, sequential growth halved) is compounded by the market disagreeing with the rebound thesis for four straight weeks. 🚩 The SEC data-collection investigation is unresolved. The 8/05 analysis is now 22 days and −11% old. Re-analyze or explicitly pass — a third scan of "still in zone" is not a decision.
FIS	[5.0]	$40.90	+0.8%	−12.6%	✅ IN $36-42 (entered 8/18)	none in window	⚪ In zone, but conviction 5.0 is the reason it is in zone. −41.4% off its 1y high and falling while its own sector rallied. Cheap is not the finding; the finding is that nothing bid for it in the best month software has had all year. That is information. Treat as a pass unless the [5.0] gets re-earned.

The signal in this table: the only names still cheap are the ones that did not participate. AVGO, APP and FIS averaged −11.6% over the past month against QQQ's +8.9%. Whatever gets bought here is a name the August bid skipped — so the question is always "why was it skipped," never "why is it cheap."

B. The sell side — nine positions have breached or reached their trim level

This is the actionable half of the scan and it is bigger than the buy side.

A portfolio-specific passage was removed from the public build.

Approaching trim (within ~5%): WDAY $197.99 vs $200+ (reports today) · ANET $202.01 vs $210+ · MSFT $504.57 vs $520+ · ARCC $19.92 vs $20 · TSLX $18.79 vs $19+ · BMY $67.11 vs $70+.

⚠️ Read the trim levels before acting on them. Six of the nine above are fixed-dollar trims ($360+, $250+, $200+), not multiples. CLAUDE.md's valuation rule says a fixed-dollar trim goes stale the moment earnings grow — and CRM, CRWD and OKTA just grew earnings and raised guidance on the same day. A $250+ trim set before an EPS beat of 80% is not a valuation judgment any more, it is a leftover. Do not sell CRM, GWRE or V on a dollar number that predates the print that moved them. Convert these to forward multiples first — that is the one number needing a human's judgment, and it is the actual work item this scan produced.

👀 WATCH CLOSELY — six names the rally pushed just out of reach

Each was in its entry band recently. All are now out by less than 5%, which is within one bad session of returning.

Ticker	Conv	Price	1wk	1mo	Entry zone	Above by	Note
INTU	[7.0]	$348.75	−3.6%	+4.7%	$300-345	+1.1%	Only name here that fell this week. −50.3% off its 1y high; ACCUMULATE 7.0 on the 8/04 file. Closest to re-entry, and moving the right way.
BR	[8.5]	$183.80	+3.0%	+15.5%	$150-180	+2.1%	Highest conviction of the six. +15.5% in a month took it out.
DOX	[7.0]	$63.56	+4.4%	+11.9%	$50-62	+2.5%	Exited this week.
GDDY	[6.5]	$97.64	−0.6%	−7.1%	≤$95 (strong <$78)	+2.8%	Second name falling into the rally. −7.1% on the month.
LDOS	[8.0]	$140.49	−0.6%	+22.8%	$115-135	+4.1%	+22.8% in a month on the defense bid. Was ❌ EXITED on 8/18 and has kept going.
G	[6.5]	$37.69	+2.1%	+2.7%	$30-36	+4.7%	Dropped off the ⭐ Shortlist this week — it was the in-zone name on the prior file.

INTU and GDDY are the two to actually watch, because they are the only two whose price is moving toward the zone rather than away from it. The other four are being carried out by the tide, and chasing a tide is how a $150-180 band becomes a $180-210 band with no new analysis behind it.

🚨 The structural finding: the entry zones are now systematically stale

This is the most important thing in the scan and it is not about any one name.

Knowledge/Playbook/pitfall-stale-entry-zone-suppresses-a-name.md already names this exact failure mode. The August rally triggered it across the whole file at once. Counting live names with a numeric entry band:

How far above entry	Count	Names
In zone	3	APP, AVGO, FIS
0–5% above	6	INTU, BR, DOX, GDDY, LDOS, G
5–20% above	~9	HON +4.5%, LHX +5.4%, LULU +5.9%, TSM +7.3%, HONA +8.5%, DPZ +9.1%, ADBE +11.3%, O +12.6%, HLNE +16.4%
>20% above	~19	IBM +17%, QBTS +20%, KT +22%, DBX +23%, DT +24%, IT +25%, NOW +25%, DIOD +27%, UEC +29%, ETN +31%, GPN +32%, EMR +34%, AMAT +36%, DOCN +36%, GLW +38%, MTCH +38%, TGT +41%, SHOP +44%, VEEV +70%

Nineteen names carry an entry zone more than 20% below spot. Those zones are not "on deck" — they are unreachable without a crash, and every one of them silently suppresses the name from every future scan. A watchlist where 19 entries can never trigger is not an opportunity list; it is a museum.

The three worst offenders — VEEV (+70% above entry, conviction 8.0), SHOP (+44%, 7.5), TGT (+41%, 5.5) — are high-conviction names that have been priced out and never re-derived.

This is a zone-reset job, not a scan finding, and it needs a decision before anything gets rewritten. See What this scan did not do below.

🔒 HOLD ON LIST — no action

At or near a 1-year high, nothing to do but not chase: CRWD ($225.70, at high), DT ($53.38, at high), FTNT ($169.71, at high — and still parked in the Graveyard with a stale $200-215 entry while up +98.7% over 52 weeks), MANH ($226.35, at high, below trim), OKTA ($171.95, at high), TEAM ($183.65, at high), RBRK (at high — but see trim table), PANW ($385.16, −2.7% off), QLYS ($189.27, −3.4% off, +40.1% in a month), SNOW ($332.70, −1.4% off), V, GPN, MTCH, TGT.

Falling, no zone reached, no action: BABA −11.4% on the week (structural-risk sleeve, [7.0], no zone stated) · MU −6.4% · SNDK −7.7% · RKLB −8.3% · U −6.6% · TTD (−75.6% off 1y high, conviction 2.5, correctly parked).

Graveyard, confirmed dead: LGHL $0.53, −17.7% wk, −50.6% mo, −99.8% off 1y high · EVGO $1.42 (−71.7% off high) · VRRM $4.54 (below the sub-$6 entry, but all three gates still unmet — correctly parked, do not buy on price alone) · SMR $9.77 (−81.7% off high, conviction 2.0, held deliberately as a lottery ticket).

🗓 Catalyst calendar — next 35 days

Date	Ticker	Event	Why it matters here
Aug 27 (today)	RBRK	Q2 print	Above trim, at 1y high, `[—]` conviction.
Aug 27 (today)	WDAY	Q2 print	1% below a `$200+` trim.
Sep 1	PANW	Q4 print	−2.7% off 1y high, `[—]` conviction
Sep 2	AVGO	Q3 print	The only high-conviction in-zone name. Buy before or after — decide.
Sep 2	SNOW, NTAP	prints	SNOW at 1y high
Sep 3	DOCU, GWRE, LULU, PATH, ZS	prints	GWRE just breached trim; LULU's entry is 5.9% below spot and the Sept print was already the gate
Sep 10	ADBE, ORCL	prints	ADBE 11% above entry; ORCL −54.5% off 1y high
Sep 17	FDS	print
Sep 24	UEC	print	+50% in a month, +29% above the top of its entry band
Sep 30	MU	print
Oct 1	ACN	print
Sept	FOMC	rate decision	The 8/04 regime note recorded the market pricing ~63% odds of a 25bp HIKE, not a cut. VIX at 14.46 is not priced for that.

🔍 Sleeve Vacancy Scout

All five screens run. Every one returned exactly the cap (25/25/25/25, smallcap 21) — see Knowledge/Playbook/pitfall-fin-py-screen-silently-caps-at-25-rows.md. These lists are truncated, not exhaustive. Absence from them is not evidence.

Reported below: only names absent from both Watchlist.md and Portfolio*.md. Every one is a nomination on a TTM snapshot, not a find — none has been checked against the §1 health bar (multi-year FCF, share-count direction, dividend coverage). Route survivors to /analyze.

🏛 Evergreen Compounders

Ticker	Price	PE	52wk	Why it is a nomination	Beats which current member
MA	$591.60	32.5	+1.4%	Mastercard is flat over a full year while the market rallied — the direct peer of a name already at trim. The sleeve holds V at $381.20, which just breached its `$380+` trim. If V is trimmed on price, MA is the same toll-booth economics at a lower relative starting point.	Pairs with V rather than replacing it — a rotation candidate, not an addition
BKNG	$203.91	22.7	−7.6%	Cheapest large-cap asset-light compounder on the screen, and down over 52 weeks. The sleeve has no travel/marketplace exposure at all.	TGT [5.5] — weakest evergreen member, and TGT is now +10.7% above its own trim

Also surfaced, not nominated: NVDA (35x, known), ASML (58x), UI (37x), MTD (32x), MEDP (36x), AMP (13x asset manager — sleeve 3 fit, not sleeve 1), STX/MU (peak-cycle — see pitfall-multiple-trim-inverts-on-peak-cycle-cyclicals).

📈 Income — Yield Tomorrow

Ticker	Price	PE	Yield	52wk	Why	Beats which current member
AFL	$116.79	12.6	2.08%	+10.7%	Cheapest quality dividend grower on the screen by a wide margin. The sleeve has zero insurance exposure — it is BR/ADP/TRI/ACN/DOX (services), EMR/ETN (industrial), AZN/ZTS/BMY (pharma). A genuine vacancy, not a better version of something held.	Fills a gap — nothing to displace
AOS	$62.06	17.3	2.30%	−12.8%	Only name on the yield-tomorrow screen that is down over 52 weeks. 30-year dividend grower.	ZTS [5.5] — weakest member, −49.9% over 52wk with a broken re-rating case

Also surfaced: CNI (22.6x rail — ⚠️ pitfall-yahoo-mixes-cad-usd-on-cross-listed-issuers applies before trusting any figure), DOV/DCI/ITW/BRC (industrial, already covered by EMR+ETN), JNJ/CVX/RTX/LIN (mega-cap, none cheap).

💰 Income — Yield Today

The screen produced almost nothing usable. 19 of 25 hits are commodity shippers (HAFN, INSW, FRO, BWLP), Brazilian state-linked issuers (PBR, PBR-A, VIV, TIMB), MLPs (MPLX, WES, HESM, BSM, ARLP) or preferred shares (AXS-PE, ACGLO, RNR-PF, CTA-PB, ALL-PH). A 9% yield on a tanker at the top of a rate cycle fails the sleeve's durability test on arrival.

Ticker	Price	Yield	52wk	Why	Beats which current member
OWL	$12.15	6.69%	−36.9%	Blue Owl — permanent-capital alternative asset manager, down 37% over 52 weeks in a year when everything else re-rated. Fee-related earnings are structurally more durable than a BDC's spread income. ⚠️ The screen shows PE 101 — do not use it. `pitfall-up-c-structure-corrupts-six-vendor-fields` applies to Blue Owl's Up-C structure directly.	NLCP [5.0] — the sleeve's weakest member (cannabis REIT, single-tenant-class concentration, $15.85 against a $21 trim it has not approached)

Explicitly rejected: WU — 12.95% yield at 5.85x is the textbook value trap (structurally declining remittance volumes); CPB 6.72% at 11.5x, −26.6% (packaged food in secular decline).

🔧 Re-Rating Plays

The richest screen this month, because Act III lifted software and left everything else behind. One of the two names originally listed here (BSX) has been withdrawn — see the correction below.

Ticker	Price	PE	52wk	Why	Beats which current member
~~BSX~~	$47.32	19.2	−54.5%	❌ WITHDRAWN — NOT A NOMINATION. Already analysed twice: 2026-08-12 (WATCH 5.5) and 2026-08-26 (WATCH 5.0) — the day before this scan ran. Conviction is trending down. Worse, the justification was circular: `pattern-console-lock-in-is-symmetric` was written by the 8/26 BSX analysis itself, so citing it as a reason to look at BSX cited the agency's own prior work as independent evidence. If BSX is revisited it is an `/analyze-from-before` on the 8/26 file, not a discovery.	—
NKE	$38.49	18.3	−50.5%	18x with a 4.25% yield after a 50% drawdown. The sleeve is entirely tech/fintech (CTSH, EPAM, FISV, FIS, GPN, PLAB, NXPI, TDC, CDW, PYPL) — there is no consumer re-rating exposure at all, and the watchlist's only consumer names (DPZ, TGT, LULU) are all above their zones.	PLAB [5.0] — weakest member, −46.2% off high, semi-cyclical with no re-rating trigger

Also surfaced, deliberately not nominated: HUBS (−50.2% over 52wk but 90x PE — a SaaS name that did not join the Act III rally, which is interesting, but 90x is not a re-rating entry); CSGP/Z/ZG (150–178x); PODD 26.8x (same console-lock-in frame as BSX, second choice); COIN (crypto beta — the KB records that BTC is held directly, so coin-beta equities are leverage on a position already owned); WING/TME/PLNT/LRN/TSCO/POOL/APTV.

🎲 Small-cap

Weakest screen of the five. ARLO is already held. ❌ AMSC was raised here and is WITHDRAWN — already analysed 2026-08-17 as analyze-smallcap, verdict AVOID [3.5], ten days before this scan. The instinct in the original note ("a single-digit PE on a small industrial after a 45% drawdown usually means the earnings are the thing that is wrong") was right, and a file had already established it. No nomination from this screen. Everything else on the list (LKFT, LUXE, ZVRA, OMDA, NPKI, LINC, PKE, JOUT, CINT, QNST, NWPX) is noise or outside the framework.

Sleeves with no screen

Sleeve 4 (AI & Infrastructure Capex), 6 (Structural-Risk) and 7 (Speculative & Micro) were not screened — their membership tests are thematic, legal and binary respectively. No filter was fabricated for them. They stay hand-curated.

🎯 Unscored entries — priority for `/analyze`

Thirty-plus live names carry [—]. After a month in which the software sleeve moved 20–76%, the unscored held positions are the problem, because several now sit above trim levels that nobody set with a conviction score behind them.

Rank	Ticker	Held	Price	1wk	Why it needs a score now
1	CRM	Self	$249.49	+21.5%	The largest unscored event in the file. Blowout FQ2 on 8/26 — rev $11.35B, adj EPS $5.90 (beat by 80%), cRPO +14% cc, Agentforce ARR >$1.5B (+240%), FY guide raised. The position sits exactly at a `$250+` fixed-dollar trim that predates all of it. Selling on that number without a score is selling on a stale artifact. `/analyze CRM`
2	CRWD	Self	$225.70	+18.6%	At a 1-year high, beat and raised on 8/26 (FY to ~$6.0B), and has no conviction score and no trim level at all. A held position at a relative high with no exit discipline defined. `/analyze CRWD`
3	OKTA	—	$171.95	+28.2%	Biggest single-week move on the entire watchlist. Sits in the ✂️ Trim Zone section with no score and no trim number — listed as needing a decision and never given one. The FQ2 print reframed the whole thesis (identity as the AI-agent security layer), so the old read is void either way. `/analyze OKTA`

Next tier, in order: RBRK (above trim, prints today — analyze after the print, not before) · PANW (prints Sep 1, same reason) · GWRE (breached trim, prints Sep 3) · PCTY (5.8% above trim, no score) · AMD (31.7% above trim, no score — the single largest unharvested gap in the file) · ANET (approaching trim, no score) · SANM, CEG, BWXT (sleeve-4 members, never scored).

➕ Additions / ➖ Removals

⚠️ Correction — two nominations withdrawn after a coverage check

This scan's Sleeve Vacancy Scout filtered candidates only against Watchlist.md and Portfolio*.md and never checked Output/. The watchlist holds tracked names, not considered ones, so the filter re-nominated names the agency had already analysed and passed on:

Withdrawn	Prior coverage	Verdict
BSX	`analyze-2026-08-26` (and `2026-08-12`)	WATCH 5.0, down from 5.5 — analysed the day before this scan
AMSC	`analyze-smallcap-2026-08-17`	AVOID 3.5

HUBS, mentioned but not nominated, is also already covered (analyze-2026-07-27).

The check is one line against the generated index and has been added to /watchlist-scan and /screen: for t in <CANDIDATES>; do row=$(grep -E "^\| $t \|" Output/INDEX.md); echo "$t -> ${row:-NOT COVERED}"; done

The five surviving nominations below were re-checked and are genuinely uncovered.

Recommended additions (all require `/analyze` first — none goes straight to `Watchlist.md`)

Ticker	Sleeve	One-line case
AFL	📈 Yield Tomorrow	12.6x quality dividend grower into a sleeve with zero insurance exposure
NKE	🔧 Re-Rating	18x + 4.25% yield into a sleeve with zero consumer exposure
BKNG	🏛 Evergreen	22.7x asset-light compounder, down over 52wk, no travel exposure anywhere in the file
MA	🏛 Evergreen	Flat over 52wk while its direct peer V just breached trim — a rotation question, not an addition
OWL	💰 Yield Today	6.69% on fee-related earnings, −36.9% 52wk ⚠️ Up-C pitfall corrupts six vendor fields
AOS	📈 Yield Tomorrow	Only down name on the yield-tomorrow screen; 30-year dividend grower

Recommended removals

No thesis-broken removals this scan. What the file has instead is ~19 names with unreachable entry zones (table above), which is the same problem wearing different clothes — a name with a permanently out-of-reach zone is removed in effect, without anyone deciding to remove it. pitfall-stale-entry-zone-suppresses-a-name is firing at scale.

Worst three, needing an explicit keep-or-cut: - VEEV [8.0] — entry $155-168, spot $285.48 (+70%), and already above its $250+ trim. A conviction-8.0 name that is simultaneously un-buyable and un-owned. - SHOP [7.5] — entry $96-108, spot $155.16 (+44%). - DBX [5.0] — entry $26-29, spot $35.67 (+23%), +32.7% over 3 months, conviction 5.0. Lowest conviction of the three; the strongest actual delete candidate in the file.

Graveyard hygiene: FTNT is filed in the Graveyard with a stale $200-215 entry while trading at a 1-year high, +98.7% over 52 weeks. Either the graveyard entry is wrong or the name belongs back in a sleeve. It cannot be both.

What this scan did NOT do

Watchlist.md was not rewritten. The zone-state changes below are recorded for a decision, not applied. ⚠️ This is the second consecutive scan to defer the edit (8/18 deferred it too) — the file is now drifting against its own prices.
The Chinese sidecar (Watchlist.zh.json) was not refreshed and the site was not rebuilt — correct, because no English thesis, break trigger or regime note was edited. The moment Watchlist.md is rewritten, both must follow.
Knowledge/Market/regime.md was not updated, though this scan establishes that it should be. It describes a two-act 2026 and there are now three acts. Its recheck is 2026-09-13; the Act III evidence is in this file.
No name was health-checked. Every sleeve-screen nomination is a TTM snapshot only.
The screens are capped at 25 rows. All five hit the cap.

Zone-state changes pending a `Watchlist.md` edit

Ticker	Was (8/18)	Now (8/27)
G	⭐ IN $30-36	❌ EXITED — $37.69, +4.7% above
BR	⭐ IN $150-180	❌ EXITED — $183.80, +2.1% above
DOX	⭐ IN $50-62	❌ EXITED — $63.56, +2.5% above
AVGO	✅ IN, mid-band	✅ IN, LOW-band — $369.87
META	🔻 BELOW $455-495	❌ $570.74 — +15% above the reset band, +4.6% wk
INTU	❌ EXITED	❌ still out, but closing — +1.1% above
GDDY	⚠️ EXITED by $0.81	❌ +2.8% above, and falling toward it
FIS	✅ ENTERED $36-42	✅ still IN — $40.90
CRM	—	🔴 AT trim `$250+`
GWRE	—	🔴 BREACHED trim `$200+`
V	—	🔴 BREACHED trim `$380+`
RBRK	—	🔴 ABOVE trim `$103+` — prints today
SAP	—	🔴 ABOVE trim band $210-220
PCTY	—	🔴 ABOVE trim `$150+`
VEEV	—	🔴 ABOVE trim `$250+`
TGT	—	🔴 ABOVE trim `>$150`
AMD	—	🔴 +31.7% above trim `$360+`