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 |
|---|---|---|---|---|---|---|---|---|---|
| KLAC | KLA Corporation | 180.64 | SELL 50% | 180.64 | 174.01 | 181.16 (+0.3%) | 188.08 (+4.1%) | 177.44 (-1.8%) | KLA has now printed a sell signal at the top of the board for two consecutive closes, and our own 12-week view on it has turned negative - the stock has already reached the price target we set for it. Meanwhile our protective stop has drifted 36% below its high water mark without the position ever being reduced, which is a risk-management failure on its own. We are banking half and keeping half, because the wider chip sector actually rose today - this is taking profit on a stock that has run its course, not fleeing a breakdown. OPERATIONAL: STRONG SELL - REDUCE cv76, Sell_Rank 1, Portion 88.52%; 2nd consecutive settled print (09-10 cv61 RallySell_BearTrend -> 09-11 cv76 FailedBreakout_Down_20D) SATISFIES the US-18/US-34 lone-flip bar for the first time on this position. L2 leg 2 satisfied (12wk 177.44 vs 180.64 close = -1.77%). ADR-0020/US-36 underwrites the floor: 36.03% erosion vs the 272.00 Peak Stop (07-01), no size cut since, FIRST FIRE, VETO-IMMUNE. Sized at the 50% tier not the engine's 88.52% under US-23 (target-overrun winner, T12_Progress 101.8%, sector up on the session); exact precedent 08-01 which fired 50% on identical cv76/Sell_Rank 1/2-print conditions. MoS -4.28 tier bump WITHHELD by G2 (12wk 177.44 vs a DCF anchor). adj 76 +0 (val, G2-suppressed) +1 (FailedBreakout_Down_20D|sell, n=37) = 77. Core : trail ratchet HELD at 174.01 - the printed Stop_Price 165.37 is DECLINED as the loosening US-29/S1 artifact; Peak Stop re-bases to 174.01 post-cut. 12wk exceeds DCF anchor. · news |
| GOOGL | Alphabet, Inc.- Class A | 338.50 | SELL 30% | 338.50 | 322.86 | 338.55 (+0.0%) | 343.67 (+1.5%) | 383.14 (+13.2%) | Alphabet is up 175% from our cost and has covered nearly 90% of the ground to our 12-week target, while the model has gone quiet on it for two straight sessions. We are taking 30% off the table and letting the rest run - the business news is good (the new Gemini desktop assistant launched today and the stock rose on it), so this is locking in a large gain, not an exit. OPERATIONAL: ADR-0012 armed + fired. Arm is US-26-GENUINE via the real-gain limb (Avg_Cost 123.00, Unreal +175.2%, Gain_ATR 27.56), so US-33 is NOT implicated. Tonight's sole tick is (d), two consecutive NO TRADE prints: tick (b) is GONE (cv0 -> cv0, no drop, so the S14 polarity defect is moot) and tick (c) does NOT fire (12wk 383.14 vs a 345.27 bar, forward +13.19% alive). Soft-tick only, so US-19 governs: FIRST fire at 30%, NO escalation - held at 30% for a third consecutive night. MoS -0.61 is above -3, no tier bump. Core : trail RAISED 318.24 -> 322.86 (2.0x ATR chandelier). PRE-COMMITTED escalation line re-bases to the live EMA200 328.34 and carries forward UNFIRED (close 338.50 = +3.09% above). 12wk exceeds the DCF anchor. GOOGL ad-tech redaction deadline 09-16 is on the horizon. · news |
| OKTA | Okta Inc. | 166.50 | SELL 30% | 166.50 | 163.59 | 169.09 (+1.6%) | 177.30 (+6.5%) | 181.03 (+8.7%) | Okta is up 19% from our cost and has covered 92% of the ground to our 12-week target, but it fell nearly 3% today and the model has gone quiet on it two sessions running. Our stop now sits less than 2% below the price - the tightest line in the book - so we are banking 30% here rather than risk giving the gain back. We committed to this trim in writing last night and are honouring it. OPERATIONAL: ADR-0012 armed + fired; HONOURS the 09-10 written pre-commitment to fire FIRST at the next EOD while arm and tick hold. Arm is US-26-GENUINE AND rests on the Gain_ATR 3.11 >= 3.0 limb (not the T12 limb), so US-33 is clear; Unreal +19.1%, T12_Progress 91.97. Tick (d) fires on two consecutive NO TRADE prints (09-10, 09-11). Ranked 4th on banked dollars last night and excluded only by the <=3 cap; the LLY withdrawal frees the slot and the pre-commitment is honoured BEFORE re-ranking (the point of the US-35 device). MoS -5.25 tier bump SUPPRESSED BY G2 (12wk 181.03 vs a DCF anchor; the -5.25 derives from a EPV on a stock) - same adjudication as 09-10. Core : trail ratchet HELD at 163.59 (chandelier 149.34 is lower); close is only +1.78% above it. 12wk exceeds DCF anchor. · news |
| DELL | Dell Technologies Inc | 567.29 | BUY — QUEUED @563.71 | 563.71 | 446.25 | 570.36 (+1.2%) | 580.28 (+2.9%) | 685.15 (+21.5%) | Dell jumped 12% to an all-time high after RBC started coverage with a buy rating and a target, and after Oracle confirmed -95bn of capital spending that flows straight to AI server makers. It is the strongest signal on our board and the only candidate that passes both our conviction floor and our quality screen. Because it closed right at its high we are placing a limit order slightly below, at the model's own entry price, rather than chasing it. OPERATIONAL: STRONG BUY cv88, BreakoutUp_52W, Buy_Rank 1. adj 88 -10 (MoS -2.48 degree-scaled, clamped) +2 (BreakoutUp_52W|buy, n=12, hit_1w.60) = 80, the highest in the pool by 9 pts. Quality gate PASSES (Quality_Pass True, F=7) - the only pool name that clears the floor AND the gate (BAC adj 71 was gate-killed, US-20 again). L1 does NOT bar it: BreakoutUp_52W is not one of the three flagged momentum-chase families. L6 does NOT bar it: this is not a post-earnings gap - Q2 ($47bn rev, $95bn AI backlog) was already public; today is an RBC initiation plus the Oracle capex read-through (HPE +11%, SMCI +7%). But the close is 0.08% off the high at an all-time high, so LIMIT not market, per US-9. Entry_Distance_ATR -0.09 does NOT trip the -0.25 no-chase bar; entry range 554.27-573.15. Size: 0.8549 x x 20% = x0.75 modulator = -> =; well under the max-position cap and the. Stop 446.25 = 3.1 ATR, at risk. WARNING: Target_Price 798.62 is the RISK OBJECTIVE, not the 4wk expectation - this cell's hit_4w is only.125 (the US-24 caveat on a second cell). 12wk 685.15 exceeds the DCF anchor - projection assumes growth beyond priced-in. Context: the 09-10 QUEUED-ANTICIPATE @541.74 was withdrawn and would have been +4.72% today (systemic S16). · news |
| SPCX | Space Exploration Tech Corp | 151.21 | NO ACTION | 151.21 | 140.06 | 150.15 (-0.7%) | 151.12 (-0.1%) | 167.21 (+10.6%) | SpaceX is up 132% from our cost and the model has gone quiet on it, which would normally trigger a 30% trim. We are holding instead: the Nasdaq-100 rebalance that could force index funds to buy this stock does not take effect until 21 September, and trimming into that flow would be selling right before a mechanical buyer arrives. The price is also still 8% above our stop with no confirmed downturn. OPERATIONAL: ADR-0012 armed (US-26-GENUINE: Avg_Cost 65.28, Unreal +131.63%, Gain_ATR 11.91) + tick (d) FIRES (two consecutive NO TRADE). Outranks OKTA on banked dollars (vs) but STOOD DOWN for a 2nd session on the CATALYST-AWARE HOLD: the 09-11 N100 quarterly ANNOUNCEMENT was only the information leg; the REBALANCE IS EFFECTIVE 09-21, 6 sessions out and entering the <=5 window on Monday, and the mechanical buying the hold rests on has not occurred. Neither exit condition exists: highs 154.86 -> 153.00 -> 154.70 -> 151.85 is ONE lower high, not a confirmed sequence; close 151.21 is +7.96% above the 140.06 trail. Trail-raise unavailable (chandelier 136.78 < 140.06); ratchet holds. Ledgered optional:true WITH price and volume - the US-13 form, not prose. PRE-COMMITTED: fire the 30% at the first EOD AFTER 09-21, or IMMEDIATELY on a confirmed lower-high or a settled close below 140.06, whichever comes first. |
| NBIS | Nebius Group, N.V. | 224.55 | HOLD | 224.55 | 234.32⚠ | 225.26 (+0.3%) | 252.70 (+12.5%) | 300.93 (+34.0%) | Nebius has slipped below our stop level, but we are holding. Palantir has just named it their preferred AI infrastructure partner, revenue grew 454% last quarter, and our own 12-week view still points 34% higher - this looks like a pullback inside a strong uptrend rather than a breakdown. OPERATIONAL: monitor stop breach -4.17% (close 224.55 vs a 234.32 trail, widened from -2.65%), NOT carded. L2's second leg fails decisively (12wk 300.93 = +34.0%). Not armed (Gain_ATR 0.89, T12_Progress 74.62, Exit_Warning blank) so no ADR-0012 path exists. ADR-0020 silent (0.58% erosion). News leg outright CONFIRMING: Palantir preferred sovereign-AI partner, Q2 revenue +454% y/y to, run-rate guide -9bn exiting 2026. The 07-05 overlay override on NBIS is 69 days stale - US-31 strips its news-override force; it may amplify, never originate. MoS -364.66 is an artifact of a EPV on a hyper-growth name (Rev_Growth 13.8x) and is G2-unreliable by construction. PRE-COMMITTED escalation: a settled close below 210.00 with the 12wk under +15% opens an exit case. · news |
| NVDA | NVIDIA Corporation | 218.29 | HOLD | 218.29 | 220.48⚠ | 219.40 (+0.5%) | 221.01 (+1.2%) | 231.16 (+5.9%) | Nvidia is fractionally below our stop after a flat session, and we are holding. Our 12-week view is still positive and the shortfall is under 1% - not a break. OPERATIONAL: monitor stop breach -0.99% (close 218.29 vs a 220.48 trail), NOT carded; L2's second leg fails (12wk 231.16 = +5.9%). Exit_Warning prints PROFIT_ARMED but Avg_Cost is NULL, so US-26's mandatory verification (Gain_ATR > 0 AND Unreal_PnL_Pct > 0) CANNOT BE SATISFIED - the profit-take path is DECLINED BY RULE, not by judgement. Independently moot: conviction is a 12-pt drop, under the 20-pt tick-(b) bar, and the label is HOLD not NO TRADE so tick (d) cannot fire. ADR-0020 silent (at peak). FOMC 09-16 and triple witching 09-18 are the near events. |
| MU | Micron Technology, Inc. | 975.26 | HOLD | 975.26 | 953.77 | 981.58 (+0.6%) | 1092.52 (+12.0%) | 1279.55 (+31.2%) | Micron is flat on the day, above its stop, and our 12-week view still points 31% higher. Nothing to do. OPERATIONAL: NO TRADE cv0 for a 2nd session but NOT ARMED (Gain_ATR 2.67 < 3.0, T12_Progress 76.22 < 90, Exit_Warning blank), so tick (d) has nothing to fire - the ADR-0012 path does not open. Close 975.26 is +2.25% above the 953.77 trail. ADR-0020 silent (at peak). 12wk 1279.55 = +31.2%. The 07-05 overlay override on MU is 69 days stale (US-31). Unreal +13.47% vs a 859.49 cost. |
| AMD | Advanced Micro Devices, Inc | 516.13 | NEAR-MISS | 516.13 | 501.35 | 519.43 (+0.6%) | 538.70 (+4.4%) | 617.69 (+19.7%) | AMD is one point short of our buy threshold. The signal itself is solid and it passes our quality screen, but the stock is far above what our valuation model says it is worth, and this particular setup type has a losing record. Passed over. OPERATIONAL: adj 76 -10 (MoS -11.22, clamped) -2 (FailedBreakdown_Up_20D|buy, n=15, hit_1w.417) = 64 - ONE POINT under the 65 floor. Quality_Pass True (F=8), so the gate is not the blocker here; the valuation clamp is. Conviction improving -> across the four closes. Closed +2.49% at 516.13, Entry_Distance_ATR 0.0 (at entry, not extended). The nearest miss on the board - re-check at the next settled close. |
| TSM | Taiwan Semiconductor Manufacturing C | 433.24 | NEAR-MISS | 433.24 | 428.34 | 436.12 (+0.7%) | 445.14 (+2.7%) | 472.09 (+9.0%) | TSMC misses our threshold by a point, and its signal has been switching on and off all week, which we treat as unreliable. It is also the most expensive name in the pool against our valuation model. Passed over. OPERATIONAL: adj 76 -10 (MoS -14.04, the DEEPEST in the book, clamped) -2 (FailedBreakdown_Up_20D|buy) = 64, one point under the floor. Quality_Pass True (F=9). The disqualifier beyond the number is stability: the trail is ANTICIPATE cv83 -> BUY cv68 -> NO TRADE cv0 -> BUY cv76, the most unstable in the pool, and the Step 7 floor requires stable or fresh signals. Closed +1.22% at 433.24, at entry. |
| BAC | Bank of America | 62.69 | NEAR-MISS | 62.69 | 62.11 | 62.79 (+0.2%) | 63.11 (+0.7%) | 63.99 (+2.1%) | Bank of America clears our conviction threshold comfortably and is the steadiest chart on the board, but it fails the quality screen - it looks overvalued against the book-value model we use for banks - so we cannot open a new position in it. OPERATIONAL: adj 75 -2 (MoS -0.45) -2 (FailedBreakdown_Up_20D|buy) = 71, SIX POINTS CLEAR of the floor - and then killed by the HARD quality gate (Quality_Pass False, OVERVALUED_VS_RI via the residual-income method for financials, F=7). This is US-20's structural interaction firing again: a name that clears the floor gets gate-vetoed, leaving the book with one admissible buy. Conviction ->. Lowest ATR in the pool (1.16, 1.9% of price). |
| PL | Plant Labs PBC | 16.45 | NEAR-MISS | 16.45 | 14.08 | 16.66 (+1.3%) | 16.62 (+1.0%) | 17.25 (+4.9%) | Planet Labs is deeply oversold and has just flipped to a buy signal, but it is both below our conviction threshold and failing our quality screen, so we pass. OPERATIONAL: adj 70 -10 (MoS -9.00, clamped) +0 (OversoldBounce_Up is report-only - no graded sample) = 60, five under the floor. ALSO gate-blocked (Quality_Pass False, F=3 - the worst in the pool, EPV NEGATIVE at -6.58). RSI 23.5, the most oversold reading on the board, after AVOID x3 -> BUY on a -1.44% session. The stop 14.08 is 2.0 ATR wide on a stock - 14.4% of price. |
| MRVL | Marvell Technology | 236.10 | NEAR-MISS | 236.10 | 229.28 | 239.16 (+1.3%) | 245.97 (+4.2%) | 241.27 (+2.2%) | Marvell had the strongest move of the semiconductor group today but is well below our conviction threshold and fails the quality screen. Passed over. OPERATIONAL: adj 69 -10 (MoS -7.04, clamped) -2 (FailedBreakdown_Up_20D|buy) = 57, eight under the floor. Gate-blocked (Quality_Pass False, F=6). Best single-session move in the pool at +4.03% but the trail is choppy (AVOID cv64 -> BUY cv76 -> WATCH cv61 -> BUY cv69). |
Outcome & track record
Accuracy at the time of this record.
210 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 →
-0.191 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.