Recommendations
3 to acquire · 2 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 | 168.98 | SELL 50% | 168.98 | 174.01⚠ | 169.61 (+0.4%) | 175.97 (+4.1%) | 162.49 (-3.8%) | KLA has now printed a strong-sell signal for six straight closes while trading below its long-term trend line and below the stop we set, and the 12-week outlook has turned negative - so we cut half the position and keep a core. The risk line on this holding has fallen 36% from its peak without the position ever being reduced, which is the clearest case for taking size off. Wafer-inspection demand is still strong, but management has flagged memory-cost and tariff pressure on margins into 2027. OPERATIONAL: L2 satisfied (decisive -2.89% break vs 174.01, 12wk 162.49 = -3.84%, close below EMA 175.36). ADR-0020/US-36 latched at 36.03% erosion, 1st fire, VETO-IMMUNE - the US-17/18/22//34 stack does not apply. US-87: the 6th consecutive settled STRONG SELL - REDUCE (Sell_Rank 2) + negative 12wk is an independent origination and escalates the ADR-0020 floor 30% -> 50%; engine Portion 88.52% declined per US-22/US-72 (keep core). Valuation bump withheld by G2 (12wk 162.49 = 2.6x the DCF). Cell FailedBreakout_Down_20D|sell +2 (n=38, hit_4w.655, hit_12w.400). Line 174.01 HELD (US-58); printed 154.00 declined. Core; the 09--250 pre-commitment's TIME limb is now satisfied (no reclaim of 174.01 on us_signals_20260918_060012.csv) and arms the moment the cut is recorded. · news |
| NBIS | Nebius Group, N.V. | 217.99 | SELL 30% | 217.99 | 234.32⚠ | 219.88 (+0.9%) | 244.93 (+12.4%) | 291.60 (+33.8%) | Nebius still sits 7% below the stop we set for it and has never been trimmed since that stop peaked, so risk discipline requires taking 30% off regardless of how the story reads. And the story reads well: the company is raising on-demand GPU prices ~20% from 1 October, which is not something you do into weakening demand, and the 12-week view is the strongest in the book. This card knowingly gives up a live forward view to close a risk failure. OPERATIONAL: ADR-0020 LATCHED since 09-01 (US-73) - erosion prints 0.58% today only because the line was raised, not because size was cut; the latch discharges only on an observed decrease. Sole surviving origination -> PINNED at the 30% floor (US-39); valuation amplify inapplicable by construction (US-78). Signal leg would NOT originate: NO TRADE cv0 x2 (US-17 bars a bare NO TRADE), 12wk 291.60 = +33.77% and rising, close +4.12%. Signal-case trigger (settled close <=203.87) not hit; time limb (below 234.32 at the EOD on us_signals_20260923_060011.csv) not yet due. The Oct-1 price-hike catalyst is recorded as arguing against this card (US-80 dated). Line 234.32 held (US-58). · news |
| AMD | Advanced Micro Devices, Inc | 545.09 | BUY — QUEUED @527.75 | 527.75 | 488.62 | 548.95 (+4.0%) | 558.93 (+5.9%) | 645.54 (+22.3%) | AMD is the strongest buy signal in the book - a 52-week breakout at top rank with the highest conviction score of any name - and it passes the quality screen outright (F-Score 8). Data-centre revenue more than doubled year over year and now makes up 58% of the business. But the stock closed +7% on the third straight day of a chip-sector bounce and is up 156% this year, so we bid at the model's entry 3.2% below the close rather than chase the move. OPERATIONAL: Buy_Rank 1, trail NO TRADE -> BUY 68 -> BUY 76 -> STRONG BUY 90. NOT a US-16 chase family (BreakoutUp_52W is outside the three flagged momentum-chase cells) and raw 90 clears the 85 bar anyway. val_adj SUPPRESSED by US-41 (EPV vs DCF = 4.0x disagreement); size modulator still x0.5 on MoS -10.57. Cell +2 but read the whole row (US-74/L1): hit_4w.111, hit_12w.000, stop-first.357 - Target_Price 690.00 is the RISK objective, not an expectation. Entry_Distance_ATR -0.66 trips the no-chase bar -> QUEUED; US-54 clear (0.66 ATR < 1.5). US-64 keeps it a LIMIT (entry below close). Stop 488.62 = manual 1.5 ATR (US-47; printed 446.62 is 3.1 ATR), risk Sizing 0.8549 x x 20% = x 0.5 = -> = CONDITIONAL ON US-56: release only once the owed KLAC/NBIS dispositions are recorded. · news |
| CRM | SalesForce Inc | 242.85 | BUY — QUEUED @246.20 | 246.20 | 235.49 | 243.26 (-1.2%) | 250.64 (+1.8%) | 267.91 (+8.8%) | Salesforce used Dreamforce to guide to $63bn of revenue by FY2030, ahead of the 4bn the street expected, and its AI agent business is growing over 240%. It is the only buy candidate in the book not rated as priced beyond fair value, and it passes the quality screen. The order sits 1.4% above the close, so it only fills if the stock confirms the move rather than fading. OPERATIONAL: BUY cv71, Buy_Rank 8, four consecutive bullish prints (69/64/64/71). Not a US-16 chase family. val_adj SUPPRESSED by US-41 (EPV vs DCF = 4.6x); size modulator x0.75 (MoS in (-3,-1]). Cell -1 and weak on the L1 metrics (hit_4w.208, hit_12w.059, stop-first.433) - it is admissible, not favoured. Printed 'Buy Limit @246.20' is above the 242.85 close -> re-classified a buy-STOP per US-64. Stop 235.49 (printed, 1.16 ATR under entry), risk Sizing 0.4924 x x 20% = x 0.75 = ->. Ex-dividend on 09-17. CONDITIONAL ON US-56. · news |
| DELL | Dell Technologies Inc | 588.40 | BUY-STOP — QUEUED @580.31 | 580.31 | 526.95 | 591.09 (+1.9%) | 607.04 (+4.6%) | 618.15 (+6.5%) | Dell joins the S&P on 21 September, which forces index funds to buy the shares, and it comes on top of a record $95bn AI-server backlog and a fresh RBC Outperform rating with a target. The signal is the single best-performing entry pattern this book has, and it has now printed twice in three sessions - which is what we pre-committed to act on. Half-size, because anticipation entries are taken small. OPERATIONAL: the 09-16 pre-commitment (US-35) FIRES - BUY - ANTICIPATE cv75 >= 55 on us_signals_20260918_060012.csv, 2-of-3 prints, slot floor cleared. Cell AnticipationUp_VCP|buy +6 is the book's best on the L1 metrics (hit_4w 1.000, stop-first.200, n=5). val_adj SUPPRESSED by US-41 (EPV vs DCF = 2.9x); size x0.75 (MoS in (-3,-1]) then x0.5 anticipate = x0.375. Printed 'Buy Stop @580.31' sits BELOW the 588.40 close -> re-classified a buy-LIMIT per US-64; the slot's fills-only-on-strength property is lost, so note it fills on a 1.4% pullback instead - honouring the pre-committed price per US-88 rather than re-litigating it. Stop 526.95 = manual 1.5 ATR (US-47; printed 473.60 is 3.0 ATR), risk Sizing 0.6076 x x 20% = x 0.375 = -> = CONDITIONAL ON US-56. · news |
| NVDA | NVIDIA Corporation | 219.34 | NO ACTION | 219.34 | 220.48⚠ | 220.73 (+0.6%) | 221.01 (+0.8%) | 228.99 (+4.4%) | Nvidia closed +2.5% and is now only 0.5% under its stop - not a real break - and the 12-week view has recovered to +4.4%, so nothing is done. This is logged as a deferred decision purely so the scorer can price what not acting cost. OPERATIONAL: breach -0.52% is NOT decisive (US-44 bar is 1.5%). Ticks all fail: (c) 12wk 228.99 > the 223.73 band; (d) the NO TRADE streak broke to AVOID; (b) cv0 -> cv57 is a rise and a US-49 label crossing. Arm STILL unverifiable - Avg_Cost blank for a 5th week (US-55/US-66), T12-limb arm declined. Ledgered direction:none, optional:true per US-55. Printed stop 230.96 sits ABOVE the close = the US-29 bearish-family artifact, declined; 220.48 held (US-58). The 09-14 pre-commitment (price limb 201.72, time limb us_signals_20260919_060011.csv) is carried - close 219.34 is still under 220.48. ACTION NEEDED: record NVDA's Avg_Cost in input/usportfolio.json - one line of JSON. |
| OKTA | Okta Inc. | 190.02 | NO ACTION | 190.02 | 172.70 | 190.66 (+0.3%) | 197.78 (+4.1%) | 185.83 (-2.2%) | Okta's profit-take trigger technically fires again, but the position is held through Oktane (22-24 September) and the signal actually strengthened back to a strong-buy today on a new high - you do not trim into strength ahead of the company's own catalyst. Stop raised instead. OPERATIONAL: armed on the Gain_ATR limb (5.06, +35.92%, US-26/US-33 clear); ONE tick - (c) 12wk 185.83 < the 193.82 band. Tick (b) fails (cv68 -> cv80 is a RISE); no other Exit_Warning code; no NO TRADE prints. Catalyst-aware hold governs (US-85, runs to the 09-24 Investor Summit). Today's 192.55 high is a NEW high, so the confirmed-lower-high limb is not met, and 190.02 is far above the 169.38 price limb. RE-COMMITMENT CARRIED unchanged: floor 30%, 50% only on >=2 non-downgraded ticks at the firing EOD; fires at the first EOD on/after us_signals_20260925_060011.csv or earlier on a confirmed lower-high / settled close <=169.38. Trail RAISED 171.96 -> 172.70 (2.0x ATR off the 192.55 high, 1.75 ATR under the close, US-65). NOT an ADD despite STRONG BUY - ADD cv80: Entry_Distance_ATR -2.34 makes the 166.82 limit a WATCH re-arm trigger, not an order (US-54). |
| SPCX | Space Exploration Tech Corp | 154.81 | NO ACTION | 154.81 | 142.72 | 152.60 (-1.4%) | 155.39 (+0.4%) | 171.91 (+11.0%) | SpaceX closed at a new high, +2.6%, into the Nasdaq- re-weight that trades at tomorrow's close. A divergence warning did appear, but trimming a position making new highs directly into the index flow it has been waiting for is the exact mistake the catalyst rule exists to prevent - the stop is raised instead. OPERATIONAL: armed genuinely (Gain_ATR 12.65, +137.14%). Tick (a) RSI_DIV fires; tick (d)'s two NO TRADE prints are DECLINED by US-71 as uncorroborated - price +2.6% to a fresh 156.87 high, 12wk rolled UP 168.63 -> 171.91, no deterioration input. Tick (b) blocked by US-49 (STRONG BUY - ADD cv83 -> NO TRADE cv0 is a label crossing); tick (c) fails (12wk 171.91 > the 157.91 band). Catalyst-aware hold: N re-weight traded at the 09-18 close, EFFECTIVE 09-21 (US-85), then the 09-24 lock-up of up to 328.4m shares. No stop break (+10.5% above the line), no lower high. Re-commitment's first eligible EOD stays us_signals_20260922_060011.csv. Trail RAISED 140.06 -> 142.72 (2.0x ATR off 156.87, 1.71 ATR under the close, US-65). |
| GOOGL | Alphabet, Inc.- Class A | 347.33 | HOLD | 347.33 | 334.31 | 352.09 (+1.4%) | 350.72 (+1.0%) | 377.45 (+8.7%) | Alphabet closed +1.3%, comfortably above both its stop and its long-term trend line, with the 12-week view still +8.7%. Nothing triggers; the stop is ratcheted up. OPERATIONAL: the 09-14 pre-commitment is NOT fired - the NO TRADE run broke to AVOID so the 3rd-consecutive limb fails, and 347.33 is far above the 328.71 price limb and the 329.13 EMA200. Armed (Gain_ATR 28.77, +182.38%) but no tick: (b) cv0 -> cv57 is a rise and a US-49 crossing; (c) 12wk 377.45 > the 354.28 band; (d) streak broken. Printed stop 362.92 sits ABOVE the close = US-29 artifact, declined. Trail RAISED 333.72 -> 334.31 (2.0x ATR off the 349.90 high, 1.67 ATR under the close, US-65). Pre-commitment carried; its time limb lapses after the EOD on us_signals_20260919_060011.csv. ADR-0020 erosion 6.62%, below the 15% trigger. |
| NET | Cloudflare, Inc. | 333.94 | HOLD | 333.94 | 302.25 | 334.79 (+0.3%) | 335.04 (+0.3%) | 372.77 (+11.6%) | Cloudflare closed +2.9% at a new high with the 12-week view at +11.6% - a winner doing what winners do. The profit-take is armed but nothing has deteriorated, so the rule says raise the stop, not trim. OPERATIONAL: armed genuinely (Gain_ATR 3.45 >= 3.0, +21.76%). No tick: (a) only PROFIT_ARMED; (b) cv61 -> cv58 = -3, far short of 20; (c) 12wk 372.77 > the 340.62 band; (d) none. G1 also protects - fresh BreakoutUp_52W. Printed stop 300.28 declined as lower than the carried line (US-58). Trail RAISED 300.66 -> 302.25 (2.0x ATR off the 336.90 high, 1.83 ATR under the close, US-65). ADR-0020 erosion 0.36%. |
| VIST | Vista Energy, S.A.B. de | 73.31 | NO ACTION | 74.58 | 71.84 | 73.56 (-1.4%) | 72.80 (-2.4%) | 76.15 (+2.1%) | Vista prints an add signal that clears the conviction bar, but the position is already 4.5% underwater and sitting just 2% above its stop, and the 12-week view is only +3.9% - adding size there is the wrong risk for the return on offer. Held, not added. OPERATIONAL: adj = cv71 + 0 (MoS -0.02 rounds to 0) - 1 (PullbackBuy_VWAP|buy). L1/US-90 ranking declines it: that cell grades hit_4w.208 and stop-first.433 - nearly one entry in two hits its stop first - which is unacceptable on a name this close to its line. Entry 74.58 is above the 73.31 close = a buy-STOP per US-64, needing +1.73%. PROFIT_ARMED on a -4.54% position is the US-26 loser artifact, declined by rule. Printed stop 67.88 declined (US-58); 71.84 held. |
Outcome & track record
Accuracy at the time of this record.
228 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.
91 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.