Recommendations
1 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 |
|---|---|---|---|---|---|---|---|---|---|
| DELL | Dell Technologies Inc | 588.40 | SELL 30% | 588.40 | 517.26 | 612.94 (+4.2%) | 670.21 (+13.9%) | 838.47 (+42.5%) | Dell has run +23% above our cost on the record $95bn AI-server backlog and now closes at another high, but the rally is showing its first momentum crack - an RSI divergence alongside a gain worth 3.1 ATR - so we bank roughly a third of the position at the high and let the other 70% keep running behind a raised stop. OPERATIONAL: ADR-0012 profit-take FIRED - armed on the GAIN branch (Gain_ATR 3.11 >= 3.0, the first materially-armed print this book has seen; X2/Z6 floor cleared with room, not a T12_Progress_Pct artifact) + tick (a) RSI_DIV on the same row. G1 carve-out does NOT reach it (AnticipationUp_VCP at cv 75, not a BreakoutUp_*/MomentumContinuation_* >= 80) per Y6/Z6. G3 does not block a FIRED profit-take (ADR-0012). Portion 30% (1 tick); no valuation tier bump (MoS -2.66 > -3). Residual, stop ratcheted 491.76 -> 517.26 (Y5: 2.0x ATR chandelier tightest; scan 509.17 eligible at 2.23 ATR but looser). Forward view is still strongly positive (12wk +42.5%) - this is risk-banking on a 50%-of-equity position, not a call against Dell. Executes into 09-18 triple witching (horizon: tighten-trail, high conf). U6: Target_Price 793.73 above Est band top 670.21 -> band top is the 4wk objective. Re-arm: a second tick (another Exit_Warning code, cv drop >=20 from 75, or 12wk < 1.02x close) -> trim a further 30%; close < 517.26 is a Y2 ALERT (news-check first), not an automatic exit. · news |
| TSM | Taiwan Semiconductor Manufacturing C | 430.26 | SELL 43.2% | 430.26 | 409.22 | 433.35 (+0.7%) | 440.56 (+2.4%) | 422.78 (-1.7%) | Our own 12-week projection has now sat below the share price for four straight sessions - the model has said 'no upside from here' all week - which is the exact pre-committed trigger we wrote down three days ago, so we complete the planned cut into this week's chip rebound rather than re-argue it on a green day. The remaining stay on a tighter stop. OPERATIONAL: Z3 armed trigger FIRED - 4th consecutive Target_12Week < Close (09-15 416.97<418.01, 09-16 403.87<413.75, 09-17 404.30<417.72, 09-18 422.78<430.26). Fired WITHOUT re-litigation by design: Z3 exists so a contested hold cannot quietly become a permanent one, and waiving a pre-committed trigger because the tape was good is how a book never cuts. Noted against it: conviction recovered ->76 and the close rallied +3.0%; W3(i) says this series lags by construction. Not a fundamentals call - Aug revenue +53% y/y, Q2 op margin 60.3%; the 09-16 ex-dividend explains part of the recent drift. UT-9: PROFIT_ARMED is inert here (Gain_ATR -0.74, underwater). G2/Y7: valuation amplify SUPPRESSED (DCF 45.44 is the broken-unit ADR row). U6: Target_Price 451.30 above Est band top 440.56 -> band top. Residual, stop 397.82 -> 409.22; 12wk counter re-bases to 0/4 (Y6(b)). · news |
| AAPL | Apple | 337.00 | SELL 30% | 337.00 | 322.88 | 340.67 (+1.1%) | 349.23 (+3.6%) | 369.09 (+9.5%) | Apple is at 91% of its 12-week target with the signal's conviction collapsing from in a single session - the move is done before the target is - so we take a third off at the high and hold the rest through the iPhone 18 Pro retail weekend on a tightened stop. OPERATIONAL: ADR-0012 profit-take FIRED - armed (T12_Progress_Pct 91.31 >= 90) with Gain_ATR 1.42, clearing BOTH UT-9 (>0) and the X2 materiality floor (>=1.0) - unlike the 09-12 arm at 0.60. Fire tick (b): conviction = -26 pts, >= the 20-pt bar; ticks (a)/(c)/(d) all absent. G1 irrelevant (cv 43, far below 80). Portion 30% (1 tick). Valuation tier bump SUPPRESSED by G2 - MoS -3.06 would otherwise bump %, but Target_12Week 369.09 exceeds the DCF anchor 237.39 by +55%, so the DCF reading is unreliable and must not amplify. News is mildly supportive, not a cancel (BofA reiterates Buy; +1.38% on 09-17) - the basis is signal decay, not a call against the iPhone cycle. U6: Target_Price 372.38 above Est band top 349.23 -> band top. Residual, stop 318.84 -> 322.88 (chandelier). Y6(b): this trim re-bases Avg_Cost and may legitimately un-arm the residual - manage by trail, not a second trim. Re-arm: close < 322.88 -> trim 23. · news |
| AMD | Advanced Micro Devices, Inc | 545.09 | BUY — QUEUED @528.27 | 528.27 | 461.97 | 545.88 (+3.3%) | 586.97 (+11.1%) | 600.69 (+13.7%) | AMD has broken to a 52-week high on the third straight session of the semiconductor rebound and prints the book's strongest conviction of the day at 90, but it closed 7% up on sector flow rather than company news, so we bid below the market at the model's own entry rather than chase the candle. Half-size, because the shares already trade far beyond any defensible fair-value anchor. OPERATIONAL: adj 80 = cv 90, val_adj -10 (MoS -11.02, clamped), bias 0 (BreakoutUp_52W|buy n=3, report-only - flagged, not applied). Quality gate PASSES. Extended (Entry_Distance_ATR -0.64) -> no-chase -> GTC LIMIT at the live Entry_Price 528.27, -3.1% from the 545.09 close; provenance = current anchor snapshot, not a carried level. Size: 66.95% x equity x 20% buy_pct = x0.5 modifier (MoS <= -3) = -> @528.27 = (6.7% of equity; does not bind). L1 caveat recorded: this is a breakout, not the favoured PullbackBuy cell, and the BreakoutUp_52W 4wk target overshoots - U6 applied, Target_Price 660.87 sits 12.6% above the Est band top, so 586.97 is the credible 4-week objective and 660.87 is a family artifact. 12wk 600.69 exceeds the DCF anchor 171.74 - projection assumes growth beyond priced-in. Risk: fills into 09-18 triple witching (horizon: high-conf +/-3%, tighten-trail), so the limit can be tagged on dealer-gamma flow rather than information; stop 461.97 = -12.6% / 2.5 ATR. expires_on 2026-09-22 (3 sessions); auto-withdraw if conviction falls under 65. · news |
| MU | Micron Technology, Inc. | 977.50 | HOLD | 977.50 | 889.31 | 994.21 (+1.7%) | 1033.21 (+5.7%) | 1026.77 (+5.0%) | Micron rallied 5.5% to 977.50 and the signal firmed from, leaving the position all but flat against cost with the forward view still alive - nothing to do but raise the stop into FQ4 results. OPERATIONAL: no origination. Dead-view clause NOT fired (12wk 1026.77 > 1.02x close 997.05, margin 3.0% - much wider than yesterday's 1.0%). Price trigger 876.77 not touched. UT-9: PROFIT_ARMED inert (Gain_ATR -0.26). Stop 846.14 -> 889.31 (scan = chandelier, 2.00 ATR). L6: no adds into FQ4 on 09-30. U6: Target_Price 1065.69 above Est band top 1033.21 -> band top. Re-arm: close < 889.31 OR 12wk < 1.02x close -> SELL 33. |
| PLTR | Palantir Technologies Inc. | 176.24 | HOLD | 176.24 | 167.07 | 177.47 (+0.7%) | 180.33 (+2.3%) | 219.42 (+24.5%) | Palantir's signal strengthened sharply with the position sitting essentially at cost and no warning flags - hold, no add. OPERATIONAL: no origination, no Exit_Warning. Not armed (Gain_ATR 0.07). Stop HELD at 167.07 (Y5: scan 172.61 is only 0.57 ATR below the close - rejected under the >=1.0 ATR bar; chandelier 163.57 would lower it). No ADD proposed - MoS -12.96 would carry val_adj -10 and the name is not in the buy pool. U6: Target_Price 188.91 above Est band top 180.33 -> band top. 12wk 219.42 exceeds the DCF anchor 46.62. Re-arm: decisive close < 167.07 -> trim 116. |
| CRM | SalesForce Inc | 242.85 | NO ACTION | 244.75 | 235.49 | 247.55 (+1.1%) | 258.51 (+5.6%) | 346.06 (+41.4%) | Salesforce flipped to the book's most reliable buy signal - a pullback to VWAP at conviction 71 - but it adjusts to 61 against a 65 floor once fair value is applied, and we do not average down into a position already 5.8% underwater, so we hold the without adding. OPERATIONAL: refused ADD, counterfactual only. adj 61 = 71 + val_adj -10 (MoS -1.96) + bias 0 (PullbackBuy_VWAP|buy n=2, report-only). UT-10 independently refuses it - do not pyramid a PullbackBuy ADD into a held name already underwater. NEAR-MISS worth flagging: the close 242.85 cleared the standing 242.76 trim trigger by nine cents. Stop HELD at 238.25 (both candidates below prior: scan 235.49, chandelier 224.36). Target 258.51 sits inside the Est band - no U6 adjustment. Re-arm: decisive close < 242.76 -> trim 76. |
| NVDA | NVIDIA Corporation | 219.34 | HOLD | 219.34 | 207.72 | n/a | n/a | 220.20 (+0.4%) | Nvidia rose 2.5% with the chip group but the engine now labels it AVOID on a failed-breakout pattern - that is a warning not to add, not a reason to sell that sit 6% above their stop, so we hold and tighten. OPERATIONAL: no origination class is satisfied. AVOID is not in the sell pool; the armed exit (close < 200.06 with 12wk < close) is untouched at 219.34, and the 12wk leg has REVERSED (220.20 > close, after printing below on 09-17). UT-9: PROFIT_ARMED inert (Gain_ATR -1.28). No Exit_Warning code besides the inert arm; MoS -1.80 is not extreme -> no valuation elevate. Worth noting against the hold: the FailedBreakout_Down_20D|sell cell is this book's BEST actionable cell (n=6, bias +7, 0.833 at 1wk) and NVDA is the held name printing it - but a scorecard bias adjusts conviction, it never originates. Stop 202.38 -> 207.72 (chandelier; the scan's 230.96 is a short-side line ABOVE the close - rejected). No Target_Price under AVOID -> 1wk/4wk n/a; Est band 212.21-232.91. Re-arm: close < 207.72 with 12wk < close -> exit; close < 207.72 alone = Y2 ALERT. · news |
| TSLA | Tesla Inc | 366.20 | NO ACTION | 366.20 | 359.94 | 371.09 (+1.3%) | 382.50 (+4.5%) | 348.66 (-4.8%) | Tesla jumped from AVOID to a conviction-74 BUY in one session, but it fails the book's quality screen and trades at roughly 10x any defensible fair value, with the model's own 12-week projection below today's price - no entry. OPERATIONAL: quality gate blocks new entries (Quality_Pass False). adj 61 = 74 - 10 (MoS -18.83, clamped) - 3 (FailedBreakdown_Up_20D|buy bias) - under floor anyway. Negative forward view (12wk 348.66 < close 366.20). L1: FailedBreakdown chase, discounted. Counterfactual only. |
| SAP | SAP AE | 213.66 | NO ACTION | 213.66 | 210.84 | 215.06 (+0.7%) | 218.34 (+2.2%) | 260.08 (+21.7%) | SAP holds a third consecutive BUY print at conviction 69 - the steadiest buy trail in the book - but it still fails the free-cash-flow quality screen, so no entry. OPERATIONAL: quality gate (FCF). adj 59 = 69 - 7 (MoS -1.31) - 3 (bias). Third print (BUY -> 69) - the most stable unheld candidate; would clear the floor at 69 raw if the gate lifted. U6: Target_Price 224.94 above Est band top 218.34. Counterfactual only. |
| SNOW | Snowflake Inc. | 338.39 | NO ACTION | 338.39 | 328.77 | 343.39 (+1.5%) | 355.06 (+4.9%) | 433.44 (+28.1%) | Snowflake flipped back to BUY at 69 but fails the quality screen and its signal has whipsawed four times in five sessions - no entry. OPERATIONAL: quality gate (Quality_Pass False). adj 57 = 69 - 9 (MoS -1.88) - 3 (bias). Unstable trail (WATCH/BUY/WATCH/WATCH/BUY). U6: Target_Price 372.32 above Est band top 355.06. Counterfactual only. |
| QCOM | Qualcomm Incorporated | 188.71 | NO ACTION | 191.61 | 169.90 | 196.65 (+2.6%) | 208.42 (+8.8%) | 204.20 (+6.6%) | Qualcomm broke out of a 20-day range on a jump from NO TRADE to BUY, but at conviction 63 it does not clear the floor and RSI is already near 68 on a single print - pass. OPERATIONAL: adj 57 = 63 - 6 (MoS -1.24), bias 0 (BreakoutUp_20D|buy n=1, report-only). Quality gate passes. Entry 191.61 sits ABOVE the 188.71 close (Entry_Distance 0.34), so it would be a buy-stop, not a discount. U6: Target_Price 235.04 above Est band top 208.42. Counterfactual only. |
| NET | Cloudflare, Inc. | 333.94 | NO ACTION | 335.28 | 291.25 | 347.13 (+3.5%) | 374.78 (+11.8%) | 440.99 (+31.5%) | Cloudflare's buy case has faded from STRONG BUY 78 to BUY 58 over four sessions and it fails the quality screen at 11x fair value - no entry. OPERATIONAL: quality gate + adj = 58 - 10 (MoS -10.00, clamped). Decaying trail. U6: Target_Price 423.34 above Est band top 374.78. Counterfactual only. |
Outcome & track record
Accuracy at the time of this record.
40 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 →
-0.280 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.