Recommendations
0 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 |
|---|---|---|---|---|---|---|---|---|---|
| 3328 | Bank of Communications | 7.93 | SELL 50% | 7.93 | 7.54 | 7.91 (-0.3%) | 7.97 (+0.5%) | 8.03 (+1.3%) | Bank of Communications has run into a clear overbought extension and the model now flags it as a sell for the second session running - the strongest and most stable sell signal in the book. The shares are up 5.7% on our cost, but RSI has pushed to 74 while the twelve-week outlook offers only about 1.3% more upside, less than a single day's typical price swing. We are banking half the position into that strength and letting the rest ride behind a raised stop. Formal STRONG SELL - REDUCE, Sell_Rank #1 on two consecutive sessions (cv90 -> cv72). Book note B: HK sell cells are reliable. Engine 84.59% (target full exit) held back to the 50% tier because conviction is FALLING and the originating cell ReversalDown_Overbought|sell is n=3 report-only. SUPERSEDES the unfilled 09-01 card (3,000 @7.88) - does not stack. The 'single print' discount that cut it to 3,000 is retired by the second print. Board lot 1,000. Proceeds HKrealised +HKResidual 3,500; stop 7.48 -> 7.54. Valuation NA_CCY (val_adj 0, quality gate off). L8: executes at the 09-03 open; reprice if it gaps >1.5%. · news |
| 3988 | Bank of China | 6.00 | SELL 30% | 6.00 | 5.73 | 6.06 (+1.0%) | 6.00 (+0.0%) | 5.98 (-0.3%) | Bank of China is our best winner, up 10.3% on cost, and it has now run to an RSI of 78.5 - the most stretched reading in the book - while the model's own twelve-week forecast has slipped fractionally below today's price. That combination is the textbook case for taking money off the table: we bank 30% of the position, keep 70% invested, and raise the protective stop on the remainder. We are also declining the engine's simultaneous invitation to buy more. ADR-0012 profit-take: ARMED on the trustworthy leg (Gain_ATR 4.06 >= 3.0 on a REAL Avg_Cost 5.44, +10.29%) - the only holding armed on Gain_ATR rather than the T12_Progress artifact. FIRED by tick (c), dead forward view: Target_12Week 5.98 < 1.02 x 6.00 = 6.12. Tick (b) checked, does NOT fire (cv = -16, short of -20); (a) no second warning code; (d) no NO TRADE prints. 1 tick -> profit_take_1 = 30%. G1 does not engage (cv74 < 80) and tick (c) fires regardless of G1. HF-2 does NOT suppress: it conditions on oversold-family signals and this is BreakoutUp_20D, and the arm is the Gain_ATR leg not the inversion artifact. RE-ISSUE of the unfilled 09-01 card, repriced 5.975 -> 6.00 per the provenance rule (live anchor close, not a carried-forward limit). Concurrent BUY - ADD cv74 DECLINED on three grounds: RSI 78.54 clear of the 75 line; buy-stop 6.01 ABOVE the close (EDA +0.07) = chase; HF-3 (every graded hkfin buy has stopped first; inverted-12wk names are not buyable, 5.98 < 6.00). Board lot 1,000. Proceeds HKrealised +HKResidual 14,000; stop 5.63 -> 5.73. L8: executes at the 09-03 open. · news |
| 3968 | China Merchants Bank | 51.75 | SELL 25% | 51.75 | 49.39 | 51.80 (+0.1%) | 52.09 (+0.7%) | 51.89 (+0.3%) | China Merchants Bank was a buy candidate yesterday and the model has now withdrawn that bid entirely, flipping to a bearish reversal read as the shares pushed to an RSI of 70. We are 6.4% ahead on cost and the twelve-week outlook has gone flat, so we take a quarter of the position off and tighten the stop on the rest. Label flipped BUY - ADD cv74 -> AVOID cv62 on ReversalDown_Overbought - the same signal family originating 3328's REDUCE. ADR-0012: ARMED via T12_Progress 99.73 on a real cost 48.65 (+6.37%, Gain_ATR 2.63); NOT an HF-2 artifact because the 12wk (51.89) sits ABOVE the close (51.75), so the arm reflects a position near its objective rather than an inversion. FIRED by tick (c): Est_12W 51.89 < 1.02 x 51.75 = 52.79. Tick (b) does not fire = -12). Portion 30% = is untradeable at the 500-share board lot; adjacent lots 500 (25%) / 1,000 (50%); FLOORED to 500 per the book's standing convention (3968; 2888; 0005. The 08-12 round-UP precedent needed a simultaneous stop break as tie-breaker; there is none here (close is 2.75 ATR clear of the 49.01 stop). Proceeds HKrealised +HKResidual 1,500. STOP HAND-SET 49.01 -> 49.39 (chandelier 2.0xATR): the engine's printed Stop_Price 54.11 is a SHORT-side stop above the close and is unusable for this long. L8: executes at the 09-03 open. · news |
| 2888 | STANCHART | 231.00 | NO ACTION | 231.00 | 228.47 | 230.15 (-0.4%) | 231.13 (+0.1%) | 221.50 (-4.1%) | Standard Chartered is held unchanged. The position is barely above cost (+1.0%) so there is no profit to bank, the shares remain above their protective stop, and no exit condition has been met. The system's 'profit armed' flag on this name is a known false positive and is disregarded. HF-2 FALSE ARM, second observation: T12_Progress 104.29% is produced by an INVERTED 12-week target (221.50 BELOW the 231.00 close), not by a winner nearing objective - the position is +1.04% with Gain_ATR 0.50. No profit exists to take. HF-6 independently holds it: no SELL label (AVOID cv58), no second Exit_Warning code, stop 228.47 intact (close 231.00). Stale overlay override (2026-07-05, 59d) 'hold 0005/2888 unless below the recorded stop' also holds. Stop UNCHANGED at 228.47 - chandelier 2.0xATR would be 221.58, BELOW the live stop, and trails raise only. |
| 2388 | BOC Hong Kong Holdings | 52.65 | NO ACTION | 52.97 | 49.97 | 52.68 (-0.5%) | 53.55 (+1.1%) | 53.97 (+1.9%) | BOC Hong Kong is the only buy signal the book produced today and we are passing on it for the third session running. It fails our quality screen on fundamentals, it screens as expensive against its residual-income value, and the model's own one-week price forecast sits below the price we would have to pay to get in - a combination that has historically preceded losses in this book. Quality gate HARD BLOCK: Quality_Pass = False, F-Score 4, OVERVALUED_VS_RI - and this is a NEW ENTRY, exactly what the gate blocks. Note the gate, not the floor, is what stops it: MoS_FV -0.3135 -> val_adj -2, adj conviction 72, which WOULD have cleared the 65 floor. Three further independent objections: (i) HF-4 LOW_QUALITY_ENTRY fires - Est_1W 52.68 < Entry_Price 52.97, the estimation layer contradicting the conviction layer (the same signature that preceded 2888's -2.2% day-one fill on 07-17); (ii) L1 - BreakoutUp_52W 4wk targets overshoot +55-75% and mean-revert; (iii) HF-3 - every graded hkfin buy has stopped first. Re-arm: quality gate passing AND Est_1W >= Entry_Price. · news |
| 0005 | HSBC HDGS(USD0.50) HK O'SEAS | 162.60 | NO ACTION | 162.60 | 160.20 | 161.75 (-0.5%) | 161.67 (-0.6%) | 159.82 (-1.7%) | HSBC is building strength but is not yet a buy. Conviction has climbed steadily over four sessions and it sits in the one signal family that has actually worked in this book, but it remains below our action threshold and the twelve-week forecast is still below today's price - we do not buy names the model expects to be lower in three months. WATCH cv62, below the 65 floor; conviction trail ->. FailedBreakdown_Up_20D|buy is the book's one +6 ACTIONABLE cell (n=7, hit_1w 0.80), which is why this is worth tracking. Blocked on HF-3's inversion test: Est_12W 159.82 < 162.60 close. Re-arm: conviction >= 65 AND a non-inverted Est_12W reprint. |
Outcome & track record
Accuracy at the time of this record.
23 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 →
3.000 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.