Discovery › watchlist
Watchlist scan
The July AI-capex rout reversed in four sessions. This scan is not a normal pass. Between 7/29 and today the entire premise of the last two scans inverted, and the practical consequence is that the shortlist emptied itself by running away, for the third consecutive time.
Market context
| Marker | 2026-07-29 | 2026-08-04 | Δ |
|---|---|---|---|
| S&P 500 (SPY) | — | $770.24 · all-time high (−0.0% off high) | +1.7% today |
| Dow | — | 53,178.41 record close (8/3) | +1.32% (8/3) |
| Nasdaq Comp | — | 25,913.9 (8/3) | +2.1% (8/3) |
| SOX | 10,447.49 | 12,166.22 | +16.5% |
| SOX off June high | −28.6% | −17.0% | 11.6pts recovered |
| VIX | 20.66 | 16.19 | −21.6% |
| QQQ | $682.12 | $720.53 | +5.6% |
| SCHD | $33.43 | $33.77 | +1.0% |
| 10yr Treasury | 4.68% | ~4.70% (off an 18-month high) | +2bp |
What drove it — three things, none of them rates:
- Earnings. All 11 S&P sectors are showing YoY earnings growth, three of them triple-digit. AMZN printed $200.6B of net sales and $1.88 adj EPS and rose 15.3% in a session.
- Geopolitical relief. Trump called off a strike on Iran to negotiate opening the Strait of Hormuz. Oil fell; the risk premium came out of everything at once.
- The AI-capex fear unwind. The July rout was a multiple de-rating with no earnings revision behind it — the July 29 AMAT analysis said so explicitly ("~100% multiple de-rating, 0% earnings revision"). Multiples that de-rate on sentiment can re-rate on sentiment, and did.
⚠️ This is NOT a rate-cut rally, and that matters for one whole sleeve. The 7/29 FOMC held 9–3 with all three dissents hawkish, Chair Warsh gave no forward guidance, and the market now prices a ~63% probability of a 25bp HIKE in September. The rate-duration cluster (VICI · O · ARCC · MAIN · TSLX · OBDC · NLCP) is the one group that did not participate — VICI +1.2%, O −4.2%, ARCC +1.4%, NLCP −0.2% against a +5.6% QQQ. That divergence is the factor working exactly as
Knowledge/Themes/reits.mdwarned, and it is unresolved.
🟢 The headline: the shortlist ran — and this time the top pick was taken
A portfolio-specific passage was removed from the public build.
| Scan date | Shortlist published | What it did next | Taken? |
|---|---|---|---|
| 7/14 | buy list | ran +8 to +22% | ❌ No |
| 7/29 | 6 names, "TOP PICK KLAC" | ran +12 to +27% in 4 sessions | ✅ Top pick taken, both accounts |
| 8/04 | 3 names left | — | — |
The 7/29 shortlist, four sessions later:
| Ticker | 7/29 | Now | Δ | Zone | Status |
|---|---|---|---|---|---|
| KLAC 🥇 top pick | $170.19 | $195.06 | +14.6% | $165-190 | ❌ Above zone. fwd P/E 25.9x → 29.8x — back at the sector median it had fallen below. The entire edge the 7/29 scan identified was consumed in four days. |
| AVGO | $370.32 | $416.42 | +12.4% | $360-420 | ⭐ Still in zone — top edge only. 21.4x fwd, tgt $527.88 (+26.8%), strong_buy. Days left, not weeks. |
| MSFT | $390.54 | $494.62 | +26.7% | $380-400 | ❌ Gone. The Azure >$100B print was correctly read on 7/29 as an unearned discount. It closed in a week. |
| ADBE | $263.43 | $253.64 | −3.7% | $230-265 | ⭐ The only name still cleanly in zone — and it got there by falling while the market hit a record. See below; this is not a free lunch. |
| BR | $159.20 | $168.15 | +5.6% | $150-162 | ❌ Above zone. Q4 print 8/4 was good (below). |
| UEC | $9.04 | $10.73 | +18.7% | $8.50-10.50 | ❌ Just above the top edge. |
Six in-zone names became two in four sessions, and one of the two only qualifies because of bad news. The 7/29 instruction was "keep cash ready rather than forcing entries." That was the right posture for a VIX-20 tape. It was the wrong outcome for the fourth-largest four-day SOX rally on record. This is now a repeated, documented pattern and it belongs in the Playbook, not just in a scan note.
⭐ ACT NOW
A portfolio-specific passage was removed from the public build.
Sizing note, unchanged and now more urgent: AVGO/KLAC/AMAT/AMD/ANET/INTC are one bucket and that bucket just returned 12-21% in four days. Do not treat a post-rally entry as the same trade as a mid-panic entry.
🔴 INTC — the trim thesis has inverted, and so has the tax math
The 7/29 scan called INTC "the only real trim" and noted it had just lost $1,923 of gain in a single session at $81.88. Trim instruction: "30-50% on strength."
This is the strength. $81.88 → $100.27, +22.5% in four sessions, +10.2% today, a fresh 26-year high, up ~225% YTD.
A portfolio-specific passage was removed from the public build.
The catalyst is thinner than the move. The driver is a preliminary, unconfirmed Apple–Intel Foundry agreement plus Xeon AI-server demand. Verified state of that deal:
- ❌ Neither Apple nor Intel has confirmed it. Trump confirmed it (June 18); the counterparties have not.
- ❌ 18A-P is in risk production only. Formal customer commitments are not expected until H2 2026.
- ❌ Lower-end chips only — TSMC retains >90% of Apple's supply.
- ✅ Real: 18A-P entered risk production (VLSI Symposium, June 16), and Xeon datacenter demand is genuine.
So: 49.2x forward, no trailing earnings, consensus "hold", consensus target only +15%, on a deal none of the parties will put their name to, in a position that is now ~19% of an account that was already flagged for having its top two names at 38%.
A portfolio-specific passage was removed from the public build.
A portfolio-specific passage was removed from the public build.
The free trim is gone. A meaningful de-risking of a $42K position now costs real tax in a
heavily-taxed taxable account. That converts a mechanical decision into a judgment call — and
the file backing that judgment is analyze dated 2026-03-17, when the stock was roughly a
third of today's price. That is the most dangerously stale document in this repo.
Recommendation: /analyze INTC before touching it. Not to delay the trim — to price it.
👀 WATCH CLOSELY
| Ticker | Conv | Price | Δ vs 7/29 | fwd P/E | Tgt (upside) | What changed |
|---|---|---|---|---|---|---|
| IT (Gartner) | [6.0] | $180.29 | +15.7% (+19.0% today) | 11.8x | $160.38 (−11.0%) | 🟢 GATE SATISFIED. See below — the single cleanest thesis resolution of the week. |
| INTU | — | $319.75 | +2.2% | 11.7x | $454.65 (+42.2%) | 🔬 Still never analyzed. Third scan carrying it. −59.7% off high, cheapest wide-moat upside on the whole list, and it barely moved in a record-high week. |
| LDOS | [7.5] | $128.73 | +8.8% (+8.4% today) | 9.8x | $150.67 (+17.0%) | ✅ Beat + raise. Adj EPS $3.26 vs $2.91, rev $4.56B +7%, book-to-bill 1.1x, backlog $48.7B, FY26 revenue/EPS/OCF guidance all raised. Every break trigger cleared. Above the $115 add but still below the $140-170 FV. |
| BR | [8.5] | $168.15 | +5.6% (+6.9% today) | 16.1x | $206.50 (+22.8%) | ✅ Strong Q4. Adj EPS $3.82 (cons $3.75), rev $2.22B +7.5%, FY26 recurring +8%, adj EPS +12% to $9.60. 20th consecutive dividend raise, +12% to $4.36. New $1.5B buyback. FY27 guide 8-12% EPS. Thesis strengthened, zone blown. |
| FDS | [7.5] | $271.30 | −4.4% | 13.8x | $255.06 (−6.0%) | 📉 Drifting toward the ≤$265 add while the market ran. One of only three names on the list that fell. |
| CEG | — | $269.09 | +3.6% (−1.7% today) | 20.2x | $351.91 (+30.8%) | 🔬 /analyze priority. Fell on a record day — the only AI-power name that did not participate. −34.8% off high. |
| FISV | [5.5] | $54.51 | +0.6% | 6.1x | $66.63 (+22.2%) | ⚠️ Earnings 8/6 — still binary, still gated. Did not participate in the rally at all. Do not add until Q2 organic ≥0%. |
| RBRK | — | $79.81 | +11.7% | 122.7x | $95.23 (+19.3%) | Top edge of $70-80 and about to exit. Cyber sleeve 4-deep. Swap only. |
| BDC (Belden) | — | $128.35 | +22.3% | 12.8x | $152.75 (+19.0%) | 🔬 The 7/29 note said dismissing this on 7/14 "looks like the wrong call." It was — +22.3% in four sessions. Still strong_buy at 12.8x. |
🟢 IT (Gartner) — the gate written into the watchlist was explicitly met
The watchlist condition, verbatim: "Gate: positive net CV additions 2 consecutive quarters before any upgrade."
Q2 CY2026, reported August 4:
| Metric | Result |
|---|---|
| Adj EPS | $4.37 vs $3.76-3.77 consensus (+$0.61 beat), up from $3.53 |
| Revenue | ~$1.7B, above consensus |
| Adj EBITDA | $466M |
| FCF | $378M |
| Contract Value | CEO Gene Hall: "Contract Value growth accelerated again" |
| Buyback | $547M in the quarter — share count down >5% sequentially; authorization raised to ~$1.2B |
| FY26 guidance | EBITDA raised to ≥$1.57B · adj EPS to ≥$14 · FCF to ≥$1.185B |
That is the gate, plus a guidance raise on all four metrics, plus a 5%-in-one-quarter share count reduction. The June 11 analysis is built on the opposite fact pattern — CV growth "collapsed to +1% YoY with −$56M net quarterly addition."
⚠️ But the price took the news too. +19.0% today, +15.7% since 7/29, and consensus target
$160.38 now sits 11% BELOW spot. The $130-145 entry is stale in both directions: the business
earned an upgrade and the price ran past it. Both the [6.0] and the entry zone are now
unsupported by the current file. → /analyze IT
⚠️ ADBE — in zone, but check why it is in zone
ADBE is the only shortlist name that fell. It did not fall on the tape:
- Morgan Stanley downgraded to Underweight — free-tier shift, leadership uncertainty, elevated AI spend, AI-native competition.
- Bank of America downgraded — generative-AI competitors and Figma structurally capping long-term growth.
- CFO Dan Durn departed — the stock fell after a beat-and-raise quarter.
None of that is in the 2026-06-04 analysis. The recorded break triggers (Digital Media ARR <$400M 2Q, or Firefly growth <20% YoY) have not fired, so the thesis is not broken on its own terms — but the entry is now supported by a two-month-old file that predates two downgrades and a CFO exit. 9.2x forward with 89% gross margins is either the best value on this list or a value trap, and the current file cannot tell you which.
Consensus is not helping: target $269.61 = +6.3%, rating hold.
✂️ TRIM ZONE — the whole cohort is now above sell-side targets
| Ticker | Held | 7/29 | Now | Δ | Trim at | fwd P/E | Off high | Tgt vs spot | Verdict |
|---|---|---|---|---|---|---|---|---|---|
| INTC | $81.88 | $100.27 | +22.5% | 30-50% on strength | 49.2x | −29.6% | +15.0% | 🔴 Act — see above. | |
| SNOW | $270.36 | $318.47 | +17.8% | $220-230 | 118.0x | −0.1% | −5.1% | Hold per the tax rule, but note it is at its 52-week high and past consensus. | |
| CRWD | $181.80 | $209.04 | +15.0% | $140+ | 133.8x | −3.9% | −7.6% | Hold. Category leader; needs a huge fall to erase the gain. | |
| PANW | $319.00 | $360.24 | +12.9% | $230+ | 87.3x | −2.3% | −6.5% | Hold. | |
| OKTA | $136.21 | $146.65 | +7.7% | $110+ | 34.2x | −6.6% | −12.0% | Hold — still below the threshold of mattering. | |
| AMD | $454.62 | $523.89 | +15.2% | $360+ | 37.6x | −10.4% | +10.5% | ⚠️ Beta lots were harvest candidates at a loss on 7/29 — the rally likely erased that. Re-check before assuming a harvest is available. | |
| TWLO | $194.33 | $198.35 | +2.1% | $175+ | 29.8x | −16.8% | +6.2% | Hold. |
🔎 Cluster signal worth naming: SNOW, CRWD, PANW, OKTA, QLYS and FTNT now all trade ABOVE their consensus targets simultaneously, three of them within 4% of a 52-week high. The market has run the entire cyber/high-multiple software complex past the sell-side in four sessions. That is a distribution-of-outcomes statement, not a sell signal — but the tax hold-rule is what is keeping these positions, not the valuation case.
Tax rule reminder (adopted 7/28, unchanged): hold unless gains are at genuine risk of being wiped out or significantly cut; the bigger the winner, the stronger the case to wait. INTC is the exception because the issue is concentration (~19% of Beta) on an unconfirmed catalyst, not valuation drift.
🔻 THESIS RESOLVED AGAINST — CHKP
Q2 2026, reported July 30. The 7/29 scan called this "the densest print on the list — ignore EPS (they'll beat), watch product revenue, billings, churn." It beat EPS and failed on exactly the thing that mattered:
| Metric | Result |
|---|---|
| EPS | $2.55 vs $2.50 — beat, as predicted, and irrelevant |
| Revenue | $673.6M vs $689.9M — MISS, +1% YoY |
| Product revenue | Sharp decline — "product slump", weak firewall appliance demand |
| Operating income | $260M vs $271M prior year; op margin 39%, down |
| Other | Sales-force disruption ongoing; soft near-term outlook; launched an AI Network Firewall |
The buy condition was "$105-115 and only if Q2 disproves churn." Q2 did not disprove it — product revenue slumped and margins contracted. Price fell −10.6% to $123.11, still above the buy zone, so there is nothing to buy and the condition failed.
Recommendation: downgrade [5.0] → [4.0] and move to 🪦 Graveyard (🌱 Revisit). Re-entry trigger: product revenue growth positive for two consecutive quarters. FTNT was retired on exactly this logic on 7/29 after three passes with the same verdict; CHKP has now had three too.
📋 HOLD ON LIST — condensed
Blown further out of zone by the rally (nothing to do): AMAT $546.13 +20.8% (32.1x fwd; entry $300-350 now 56% below spot) · ORCL $148.19 +23.5% (Google Gemini partnership, $7B defense contract, $638B RPO — but FCF −$50B FY27 unchanged; $110-115 stage is 24% below) · GLW $160.43 +27.3% · SANM $208.29 +20.8% · ADI $376.92 (add <$290) · ETN $441.17 +14.2% · ANET $194.10 +14.4% (at 52wk high, tgt below spot) · TDC $33.66 +15.3% · SNPS $399.85 +7.0% (zone $310-335; gated on 8/26) · TCEHY $62.87 +8.9% · BABA $129.03 +12.0% (strong_buy, tgt $190.16 = +47.4%) · LEU $190.29 +12.2% · MAIN $56.76 · PTC $139.70 · SAP $193.44 · CRM $189.31 · WDAY $168.01 · TEAM $108.59 +8.6% · ZS $161.07 · CDW $151.09
QLYS $156.85 (+15.5%) — no company news. Q2 is August 10, not August 4. The 7/23 note recorded QLYS falling −20% in nine days on no news; this is the exact mirror image. Consensus target $133.35 is now 15% below spot. Nothing changed at the company either time.
Rate-duration cluster — the group that did not rally: VICI $26.59 (+1.2%) · O $63.00 (−4.2%) · ARCC $19.43 (+1.4%) · MAIN $56.76 (+2.6%) · TSLX $17.93 (+1.4%) · OBDC $11.12 (+0.9%) · NLCP $15.77 (−0.2%). With September now pricing a hike, the VICI [5.5] "hold for the coupon" and ARCC [5.5] "hold, coverage broke below 1.0x" calls from 8/3 both stand unchanged and unpressured. O at −4.2% is the closest this sleeve has come to interesting, and it is still $8 above the $50-55 entry.
Fell while the market hit a record (only four names): O −4.2% · FDS −4.4% · ADBE −3.7% · NVO −5.2% (Q2 tomorrow, 8/5 — the DO NOT ADD stands) · MANH −8.2% · SHOP −7.6% · EFX −6.5% · CTSH −2.7% · AAPL −9.3%.
AAPL $308.33 (−9.3%) — Q3 beat (rev $109.4B +16%, EPS $2.02) but Q4 guided to 9-11% growth vs consensus, on DRAM/NAND cost inflation and component bottlenecks; Greater China $18.8B vs $19.6B expected. ⚠️ Cross-read for MU: Apple's own guidance is direct corroboration that memory pricing is spiking now. That does not refute the "peak-cycle earnings, not value" call on MU — it does mean the peak has not rolled over yet, and MU rose +9.2% to $896.09 while its loss shrank from −$1,494 to −$889. The harvest window on MU is closing.
🔬 Unscored / never-analyzed — priority for /analyze
| # | Ticker | Price | fwd P/E | Off high | Tgt upside | Why now |
|---|---|---|---|---|---|---|
| 1 | INTU | $319.75 | 11.7x | −59.7% | +42.2% | Never analyzed. Flagged 7/14, 7/29, and again now. Biggest upside-to-consensus of any wide-moat name on the list, and it did not participate in the rally. |
| 2 | CEG | $269.09 | 20.2x | −34.8% | +30.8% | Never analyzed. Fell 1.7% on a record day. Fills the AI-power gap without owning an AI-capex semi. |
| 3 | BWXT | $174.00 | 33.5x | −28.0% | +36.9% | Never analyzed. Sole-source Navy propulsion + BWRX-300. |
| 4 | BDC (Belden) | $128.35 | 12.8x | −19.8% | +19.0% | +22.3% in four sessions and still strong_buy at 12.8x. Twice deferred. |
Recommendations
Top 3 actionable now
- INTC — decide, and price it first. The trim you have waited three weeks for is live at
$100.27 (+22.5%), the catalyst is an unconfirmed preliminary deal for Apple's lower-end
chips, the position is ~19% of Beta with $27,993 embedded, and the tax-neutral path shrank
from ~66 shares to ~16. Run
/analyze INTC— the current file is from March 17. - META $580.90 — in the add zone on a fresh (7/29) analysis. No new work required. It walked into $560-590 by falling on a record-high day. Either act on the existing file or explicitly pass.
- AVGO $416.42 — final day(s) inside $360-420. ⚠️ Half-size at most: Beta already holds 52sh at $380.93, and the AI-capex bucket just returned 12-21% in four sessions.
Worth a fresh /analyze — ranked
A portfolio-specific passage was removed from the public build.
Remove / downgrade
- CHKP [5.0] → [4.0], move to 🪦 Graveyard (🌱 Revisit). Q2 failed the exact test the watchlist set. Re-entry: product revenue growth positive 2 consecutive quarters.
- QLYS — reconfirm no action. Target 15% below spot; Q2 is 8/10, not 8/4.
No additions recommended
Nothing surfaced in the news sweep that belongs on a watchlist bought at a record high after a 16.5% four-day SOX rally. The correct posture is the opposite of adding names.
📌 The finding that outlives this scan
Three consecutive shortlists (7/14, 7/29, 8/04) ran 8-27% within two weeks, so the analysis layer works. The 7/29 top pick was executed in both accounts — KLAC bought at $183-186 against a $165-190 zone, inside a window that lasted four sessions. The buy discipline is there, and it required overriding the scan's own "keep cash ready rather than forcing entries" advice to get it.
The asymmetry is on the sell side. The 7/14 trim list fell 12.5% unharvested; INTC has now
been "trim on strength" for three weeks across a −24% swing and a +22.5% swing, and the
tax-neutral capacity to do it fell from ~66 shares to ~16 while nobody acted. Buys have been
executed; trims have not. That is the narrower, better-supported version of this finding →
Knowledge/Playbook/pattern-shortlist-runs-when-unexecuted.md.
Prices live 2026-08-04. Previous pass: watchlist-scan-2026-07-29.md
Sources: CNBC — market 8/3 · Yahoo Finance — Dow record 8/3 · Gartner Q2 CY2026 · Gartner EPS beat detail · Leidos Q2 + raise · Broadridge Q4 FY2026 · Check Point Q2 mixed · Check Point product slump · Intel–Apple deal status · Intel 18A-P risk production · Apple Q3 FY26 guidance · Adobe downgrades · Oracle catalysts · 10yr yield / Sept hike odds · Qualys Q2 date