Recommendations
2 to acquire · 3 to dispose
| Ticker | Name | Close | Action | Action Price Where to act — a buy-limit on a pullback or a buy-stop above resistance, not the current price. Sanity-check vs Close: on a held name, an action price stranded above the close means the queued level was set when the stock was higher — stale, awaiting reprice. | Stop The risk-defined exit, set below the entry for a long (~1.2–1.8×ATR). A held long whose Stop sits at/above the Close is already below its stop — a stale box awaiting reprice. | 1wk Model estimate of the likely price at this horizon — a forward projection, not a guarantee, and separate from the trade's Action Price or Stop. Where the ML estimator is enabled these are per-ticker calibrated forecasts (~80% coverage band) that are refreshed periodically and can drift between refreshes; otherwise they're interpolated from the signal's swing target. Confirm against the settled close. | 4wk Model estimate of the likely price at this horizon — a forward projection, not a guarantee, and separate from the trade's Action Price or Stop. Where the ML estimator is enabled these are per-ticker calibrated forecasts (~80% coverage band) that are refreshed periodically and can drift between refreshes; otherwise they're interpolated from the signal's swing target. Confirm against the settled close. | 12wk Model estimate of the likely price at this horizon — a forward projection, not a guarantee, and separate from the trade's Action Price or Stop. Where the ML estimator is enabled these are per-ticker calibrated forecasts (~80% coverage band) that are refreshed periodically and can drift between refreshes; otherwise they're interpolated from the signal's swing target. Confirm against the settled close. | Rationale |
|---|---|---|---|---|---|---|---|---|---|
| KLAC | KLA Corporation | 194.79 | SELL 30% | 194.79 | 193.38 | 196.57 (+0.9%) | 195.24 (+0.2%) | 189.15 (-2.9%) | Cut a third of KLA after the risk line walked away from the position: the trailing stop has fallen 28.9% from its peak while the holding was never once reduced, and the shares now print their first outright bearish signal of the run with a 12-week model view that sits BELOW the current price. The stock is a semi-capital-equipment name caught in a rate-driven chip selloff (30-year Treasury above 5.2%, highest since 2007) and China export restrictions, and it closed 5.3% lower on the session just 0.7% above its stop. We keep because the business itself is intact - five straight earnings beats, record 66bn quarterly revenue, and a average analyst target well above the market - so this is a risk-discipline trim, not a verdict on KLA. || ADR-0020 Trail Erosion FIRST FIRE, session 15, veto-immune on this Peak Stop; ALSO an ADR-0012 fired profit-take (PROFIT_ARMED via T12_Progress 102.98 + tick (c) dead forward view, T12 189.15 < 1.02x Close 198.69). Valuation amplify (MoS -4.88 would bump SUPPRESSED by G2: T12 189.15 vs DCF 61.81 = 3.06x, anchor unreliable. Tier held at 30% - news caps escalation. A decisive settled close below 193.38 escalates to a fuller exit next session. Realised ~+vs 175.75 cost. · news |
| GOOGL | Alphabet, Inc.- Class A | 344.20 | SELL 30% | 344.20 | 332.73 | 346.70 (+0.7%) | 338.20 (-1.7%) | 377.14 (+9.6%) | Bank a third of a position that has nearly tripled (cost now) and has stopped generating a tradeable signal - five consecutive sessions of no actionable read, with the price stalled below its 20- and 50-day averages and 91% of the way to the 12-week objective. The business case is still strong (revenue +24%, cloud +82%, eleven straight earnings beats, and 19 of the 20 largest funds on Wall Street own it), so we keep; but the AI-division leadership shake-up including Jeff Dean is a genuine overhang and there is no reason to carry full size in a name the model has gone quiet on. || ADR-0012 fired profit-take: PROFIT_ARMED (Gain_ATR 29.86, T12_Progress 91.27) + tick (d) 5 consecutive NO TRADE prints. 1 tick -> profit_take_1 30%. MoS -0.684 > -3, no tier bump. Carried unexecuted since 08-14, session 6 owed (US-37/ADR-0019). 12wk 377.14 EXCEEDS DCF anchor 145.29 - projection assumes growth beyond priced-in. Realised ~+. · news |
| NVDA | NVIDIA Corporation | 219.74 | SELL 30% | 219.74 | 217.07 | 221.01 (+0.6%) | 222.81 (+1.4%) | 235.94 (+7.4%) | Take off a winner that has gone four straight sessions without a tradeable signal and now sits 93% of the way to its 12-week objective, while trading barely 1.2% above its trailing stop. Tuesday's 2.2% fall was macro - a hawkish Fed hold pushed the 30-year Treasury above 5.2% and hit the whole semiconductor complex - not company news, and the setup into the 26 August Q2 report is constructive (consensus EPS on ~$92bn revenue, +96% year on year; GF Securities just raised its target to). That is exactly why we trim 30% and no more: stay on for the print. || ADR-0012 fired profit-take: PROFIT_ARMED printed on the anchor row via the T12_Progress 93.13 limb (Avg_Cost blank - US-31, arming no longer depends on it) + tick (d) 4 consecutive settled NO TRADE. 1 tick -> profit_take_1 30%. This is the SEVENTH session this tick has formed and the first time it is carded - "optional trim" language is banned (ADR-0012 / L3, the 07-01 200->194.5 round-trip). News classified OVERRIDES but applied as a CAP not a cancel: earnings 08-26 is 6 trading sessions out, just OUTSIDE the <=5 catalyst-hold window, so the trim stands at tier 1 with no escalation. MoS -2.542 > -3, no bump. 12wk 235.94 EXCEEDS DCF anchor 163.98. · news |
| VIST | Vista Energy, S.A.B. de | 69.74 | BUY | 69.74 | 68.13 | 69.93 (+0.3%) | 70.59 (+1.2%) | 74.24 (+6.5%) | Buy the only genuinely cheap, quality-passing name on the board. Vista is an Argentine Vaca Muerta shale producer trading below our estimate of intrinsic worth (discounted-cash-flow anchor against a price) with production up 32% year on year and revenue up 102%, while the company keeps buying back its own stock (another on 17 August) and Peter Thiel holds a ~1% stake. The Iran stalemate lifting crude gives the position a live catalyst that works against the rate-driven pressure hurting the rest of the book, and the 12-week view of sits comfortably inside the valuation anchor rather than beyond it. Sized at normal weight rather than the enlarged weight cheapness would allow, because the shares gapped to on the oil spike and gave it all back to close at. || Adj 67 = raw 72 + val_adj +2 (UNDERVALUED, MoS +0.087, Quality_Pass True) -7 scorecard (FailedBreakdown_Up_20D|buy n=7, hit_1w.143, ACTIONABLE). Entry_Distance_ATR 0.00, Order_Type Market -> ACT NOW, no chase violation. Size modulator permits up to 1.25x; TAKEN AT 1.0x deliberately on the outside-down reversal day. GRADUATES the 08-17 QUEUED-ANTICIPATE buy-stop @71.44 to a full acquire at a BETTER price (provenance rule: repriced to the live entry, stale 71.44 limit retired). WARNING - that buy-stop TRIGGERED at the 08-18 open (open 73.51 > 71.44) and no fill was recorded: if the were taken at ~73.51 this is an ADD and size should be cut to; book shows so it is carded as a new entry. Record TAKEN or SKIPPED (ADR-0019). No anticipate slot used this session - VIST graduated into ACQUIRE. · news |
| LMT | Lockheed Martin Corporation | 607.17 | BUY — QUEUED @583.11 | 583.11 | 565.15 | 606.42 (+4.0%) | 608.24 (+4.3%) | 618.85 (+6.1%) | Queue a defensive-growth position that is doing well for reasons the rest of the book cannot access. Lockheed just booked one of the largest contract awards in its history - a seven-year PAC-3 missile deal worth up to 6bn - ended the quarter with a record $230bn order backlog, raised full-year profit and cash-flow guidance, and posted its biggest one-day share gain since 2001. Analyst targets have moved to -against our buy limit, and our own discounted-cash-flow anchor of sits above the 12-week view of so the projection is supported rather than stretched. Rising long-term interest rates and a Middle East flashpoint are hurting almost everything else we own; defence earnings are the one place that mix helps. The order only fills on a 4% pullback - we are not chasing it. || Adj 65 = raw 72 -9 val_adj (MoS -1.764 x5) +2 scorecard (PullbackBuy_VWAP|buy n=22, hit_1w.591, the only positive buy cell). Clears the floor EXACTLY. Quality_Pass True, F-Score 8, FAIR_VALUE. Size modulator 0.75x (-3 < MoS <= -1). Entry_Distance_ATR -1.55 = extended -> GTC LIMIT, no chase (L1). Est_* are anchored to the 607.17 close, so the quoted upside is what the fill earns if 583.11 trades. Expires 2026-08-21 (anchor + 3 sessions); auto-withdraw if conviction < 65. · news |
| LLY | Eli Lilly & Co | 1225.73 | HOLD | — | 1159.93 | 1237.05 | 1257.04 | 1359.07 | Hold and raise the stop. Lilly was the one thing that worked on a brutal session, up 3.6% to a 52-week high as money rotated out of rate-sensitive technology into defensive healthcare, and the 12-week view of is 11% above the price. || ADR-0012 ARMED (PROFIT_ARMED, T12_Progress 90.19) but NO tick fired - (a) no second warning code, (b) conviction ->60 no collapse, (c) T12 1359.07 >> 1.02x Close 1250.24, (d) no NO TRADE run. Armed-without-tick => MANDATORY TRAIL RAISE, not a trim. Stop 1159.23 -> 1159.93 (model print). Peak Stop 1180.30 unchanged - still 1.7% of erosion outstanding. |
| MU | Micron Technology, Inc. | 940.76 | HOLD | — | 921.42 | 947.56 | 1020.41 | 1116.46 | Hold, but this is the position most likely to be stopped out this week. Micron fell 7% as rising long-term rates hit the memory-chip trade hardest, and the stop we raised on Monday now sits only 2% below the price - well inside a single day's normal range for this stock. The 12-week view is still 19% higher and the position is 9.5% ahead of cost, so we do not cut it; but if breaks we are out of a third and the trade thesis needs re-underwriting. || No longer PROFIT_ARMED (Gain_ATR 1.25, T12_Progress 84.26 - de-armed by the drawdown). Model prints no stop (NoTrade). Trail HELD at 921.42 = Peak Stop, erosion 0%; NOT loosened (US-29). Close 940.76 is 0.30 ATR above the line on a 6.9%-ATR name - see systemic #5. · news |
| NBIS | Nebius Group, N.V. | 248.43 | HOLD | — | 235.69 | 247.56 | 269.06 | 283.99 | Hold. Nebius gave back 6.2% in the same rate-driven AI-infrastructure selloff that hit CoreWeave, and the signal has gone quiet, but the position is still 17% ahead of cost with the stop 5% below the price and a 12-week view 14% above it. || Not armed (Gain_ATR 1.38, T12_Progress 87.48). ADR-0020 watermark re-based to 235.69 on the 08-16 discharge - erosion 0%, veto-immunity spent. Overlay override live: invalidates if the Meta deal is materially at risk. MoS -11.75 is extreme but valuation NEVER originates a sell and no signal-based reason is present. |
| NET | Cloudflare, Inc. | 301.36 | HOLD | — | 301.75 | 301.98 | 305.81 | 321.50 | Hold on a hair-thin stop breach. Cloudflare closed at against our line - 0.13% through, which is noise, not a break - and the 12-week view is still 6.7% above the price. We are not cutting on a 39-cent miss, but the tolerance is now zero. || MARGINAL BREACH THE MONITOR DID NOT REPORT - portfolio_monitor.py reads NET as because input/usportfolio.json lists NET TWICE and the loader takes the last row (16th session of this defect, systemic #1). L2 NOT satisfied: needs a decisive break AND a negative 12wk; T12 321.50 is positive. Stop HELD at 301.75, NOT reset down (avoids the OKTA/KLAC erosion path). TRIGGER: two consecutive settled closes below 301.75, OR one decisive close below the 08-18 low of 297.63, fires a 30% trim. |
| OKTA | Okta Inc. | 143.92 | HOLD | — | 140.55 | 144.32 | 155.95 | 169.02 | Hold and tighten. Okta stabilised for a third straight session while the market fell around it, holding just above cost, and the model has lifted its stop back up - so we take the tighter line. Earnings on 26 August keep this at hold rather than add. || Stop RAISED 136.11 -> 140.55 (model print) - the erosion clock opened on 08-17 partially REWINDS; vs Peak Stop 144.84 the erosion improves from -6.03% to -2.96%, well clear of the ADR-0020 -15% line. Not armed (Gain_ATR 0.61, T12_Progress 85.15). The 08-17 BUY-ADD remains DECLINED (adj 63 < 65) and earnings 08-26 carries a TRIM-INTO stance. |
| SPCX | Space Exploration Tech Corp | — | HOLD | — | 129.00 | n/a | n/a | n/a | Hold through the second lock-up release. SpaceX passed its first insider-selling test on 6 August far better than feared - the shares rose 35% in five sessions and added roughly $500bn of market value instead of being swamped by supply - and Morgan Stanley called the fundamentals "largely unchanged" with the level an attractive entry. The next tranche, up to 319 million shares (7% of restricted stock), unlocks on 20 August, two sessions away and inside the catalyst window, so we hold through it: no trim, no add. || MUTED - absent from the scan CSV, permanent no_close, no synthesized signal. News-assessed as required by the muted-name rule. Manual stop 129.00 stands. Mark carried at 141.29 (08-15) may UNDERSTATE the position by against the reported post-lockup rally. 65.282 is CORRUPTED (19,584.60/300 - the original basis was overwritten and is unrecoverable). · news |
Outcome & track record
Accuracy at the time of this record.
140 Graded decisions Past recommendations old enough to score against what actually happened (target hit / stopped out / closed). This is the evidence base — the bigger the number, the more history behind the calls. Click to inspect every one. See the decisions →
-0.132 1-week calibration How much of the eventual 4-week move typically shows up in the first week (0 = almost none, 1 = most of it). Use it to read the 1-week estimates: a low number means prices drift toward the target slowly, so don't over-read a soft week-1.
52 Published records How many dated end-of-day verdicts exist for this book — the length of the public track record you can browse in the Archive below.