Recommendations
0 to acquire · 0 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 |
|---|---|---|---|---|---|---|---|---|---|
| 0992 | Lenovo Group Limited | 30.36 | HOLD | 30.36 | 30.09 | 31.56 (+4.0%) | 34.37 (+13.2%) | 43.26 (+42.5%) | Holding all of Lenovo and letting the stop do the work. The stock gave back its entire AI-product pop today, closing 6.2% lower on nearly twice the usual volume, but the underlying trend is intact - it still trades above every major moving average, the model's conviction actually improved, and the twelve-week view remains 42% above today's price. Valuation is fair with the best quality score in the book, and no news explains the fall, so there is nothing that justifies selling into it. Our protective stop sits just below the current price and above our average cost, which means the worst case here is being taken out at a small profit rather than a loss. OPERATIONAL: no sell origination class fires - label is HOLD (not SELL-REDUCE), Exit_Warning blank, Sell_Rank blank, conviction, Trend_Score 77, close above EMA20/50/100/200. L2 fails on both legs (close 30.36 > stop 30.09; T12 43.26 = +42.5%). ADR-0012 NOT ARMED: Gain_ATR 1.42 on the snapshot's stale 27.498 basis and 0.63 on the true 29.0925 cost, both < 3.0; T12_Progress 70.18 stale / 8.9% true, both < 90; PROFIT_ARMED cleared. ADR-0020 n/a (off outside us; live stop 29.55 is 1.8% below the 30.09 peak). Valuation cannot originate and has nothing to amplify (MoS -0.65 not <= -3, FAIR_VALUE not UNDERVALUED). The executor's unauthorized 6,000-share add is NOT a sell reason - undoing it would be a portfolio-shape origination, never permitted (CLAUDE.md S9). STOP HELD at the 30.09 ratchet watermark (live model stop 29.55 is lower; stops never walk down); Peak Stop 30.09. WARNING: 30.09 is only 0.9% / 0.13 ATR below the close on a 2.01-ATR name - a break exits all, but at +HK (+3.4%) vs the 29.0925 cost, so this is shake-out risk, not capital risk. No trail-raise (raise is mandatory only when armed-but-not-fired). f=0.30 (L5 mandatory, calib_1w -0.183, negative, 4th value in 5 sessions). DCF caveat CLEARED: T12 43.26 is now 5.2% BELOW the 45.64 DCF anchor, where 46.11 sat above it on 09-07. Re-arm for an ADD: two consecutive prints of conviction >= 65 with a live buy trigger at or below the close - today's Buy Limit @32.06 is 5.6% ABOVE the market and unusable (systemic finding #7). · news |
| 0100 | MINIMAX-W | 331.20 | WATCH | 331.20 | 296.61 | n/a | 396.47 (+19.7%) | 243.89 (-26.4%) | Passing on MiniMax despite it screening as the book's top-ranked buy. The model's own twelve-week projection sits 26% BELOW today's price - it is forecasting a decline while labelling the stock a buy - and the shares are 75% off their high, below their fifty- and hundred-day averages, with the weakest momentum reading of any candidate this book has seen. The buy signal appeared for the first time today, on a session the stock fell 5.6%. We have no valuation anchor for it either: the fair-value model cannot price the company. A single fresh signal on a falling stock with a negative forward view is not an entry, it is a knife. OPERATIONAL: adj conviction 65 clears the floor by exactly zero. HT-4 forward-view screen FAILS - T12 243.89 vs Close 331.20 = -26.4% (7th dead-12wk-on-a-bullish-row instance in 12 sessions, systemic finding #3); only the 4wk leg passes (396.47, +19.7%). Below EMA50 350.95 and -23.3% below EMA100 431.65; EMA200 blank (insufficient listing history); -75.1% from the 1,330.00 52W high; Momentum_Score 24. HT-3 R:R artifact reproduced a 4th time: (348.71-296.61) risk vs (396.47-348.71) reward = 0.92. Entry UNUSABLE as printed - Buy Limit @ HKis 5.3% ABOVE the close (Entry_Distance_ATR +0.42), and the close falls outside the Step 7 provenance band 338.29-359.13, so there is no compliant action price (systemic finding #7). Single print (WATCH/45 -> BUY/65). Quality_Pass False, F-Score 5, Quality_Rank 8/21; Val_Signal NA so val_adj 0 and the quality gate is off by rule. Volume fell 23.9M -> 17.1M on the decline. The horizon file's 2026-09-08 lock-up tag on this name was CHECKED AND WITHDRAWN as inaccurate (major cornerstone/employee lock-up expired 2026-07-08; executive tranche runs to Jan-2027) and forms no part of this rejection. Re-arm: two consecutive BUY prints with Target_12Week >= Close and a reclaim of EMA50. · news |
| — | AI HK Tech | — | NO ACTION | — | — | n/a | n/a | n/a | No trades today. The book holds one stock and keeps 76% of its capital in cash. The only name that screened as a buy is forecast by our own model to fall over the next quarter, so we passed on it, and the one position we hold gave no sell signal - its trend and forward view are intact even after a sharp down day. Cash is unchanged and ready. OPERATIONAL: 0 ACQUIRE + 0 DISPOSE + 0 ANTICIPATE. Buy pool had exactly one member (0100), withheld on the HT-4 forward-view screen; rows print AVOID; zero BUY - ANTICIPATE prints across 09-03..09-08. BLOCKING INCIDENT: the Auto-Executor filled the WITHDRAWN 09-04 ADD (6,000 0992 @31.75, txn hktech-20260908-0001) and skipped the live 09-07 SELL 3,000 @32.36 - the book did the opposite of its decision on both legs. One-session cost ~HK (-HKmark-to-market on the unwanted add, HKforgone on the un-executed trim). Position 10,000 ->, cost 27.4980 -> 29.0925, cash. See systemic findings #1 (executor must honour withdrawals + staleness guard) and #2 (scan/executor race corrupts the Held columns that drive ADR-0012 arming). Book HK = +1.01% vs the 2026-07-06 HK seed. |
Outcome & track record
Accuracy at the time of this record.
13 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 →
-0.183 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.
50 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.