Recommendations
2 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 |
|---|---|---|---|---|---|---|---|---|---|
| 0992 | Lenovo Group | 34.32 | BUY — QUEUED @34.18 | 34.18 | 28.74 | 34.63 (+1.3%) | 34.76 (+1.7%) | 35.74 (+4.6%) | Lenovo's breakout has matured into a top-conviction anticipation setup (97) printed on every snapshot today; valuation is roughly fair and quality passes, so we keep the queued purchase and move its limit up to the model's current entry rather than chasing. OPS: repriced from the 09-16 GTC @33.58 (touchable at the 09-17 low 33.52 but never placed/recorded); live row is a Buy Stop @34.18 below the 34.32 close, so queued as a LIMIT; 48% x 600k x 20% = 57.6k -> 1,685 -> one; anticipate slot 0.5x would round to, not used separately; expires_on unchanged 2026-09-21; L7 hktech/hksemicon hold 0992 - execute once from hk. · news |
| 1024 | Kuaishou Technology | 29.94 | BUY — QUEUED @29.94 | 29.94 | 28.01 | 29.88 (-0.2%) | 32.25 (+7.7%) | 39.49 (+31.9%) | Kuaishou turned to a buy at the close from a deeply oversold level (RSI 26), and it is the book's highest-quality name screening undervalued with a strongly positive three-month view, so we queue a position at the model entry with a tight stop. OPS: single closing BUY print after 4 intraday AVOIDs -> size x1.0 not x1.25; 70.59% x 600k x 20% = 84.7k / 29.94 = (lot 100); stop 28.01 = 1.8 ATR; OversoldBounce_Up|buy n=2 report-only; 08-31 BUY @33.58 unexecuted, name -10.8% since. · news |
| 0005 | HSBC Holdings | 159.80 | SELL 50% | 159.80 | 155.31 | 159.66 (-0.1%) | 159.91 (+0.1%) | 160.40 (+0.4%) | HSBC has now closed below the protective stop we committed to for a third session, the model has dropped to no-trade and its three-month view is flat, and it is the book's largest position, so we halve it. OPS: 159.80 vs pre-execution 163.18 (-2.07%, 1.24 ATR); 12wk 160.40 dead; adj 2 below floor, carried by the L2 break; HK-11 FINAL ISSUE (already market-at-open) - if unexecuted at the 09-18 open retire to WATCH 155.31 (re-arm: close < 155.31 OR REDUCE cv>=65); retire early on close >= 163.18; 15:30 printed STRONG SELL - REDUCE cv39; US-ADR 'Sept 17' quote discarded (S8). · news |
| 2513 | Z.AI (Knowledge Atlas) | 740.00 | SELL 85.46% | 740.00 | 660.80 | 797.58 (+7.8%) | 831.51 (+12.4%) | 829.29 (+12.1%) | Z.AI's rally to 787.50 failed back to 740 and the model now ranks it the book's top sell at STRONG SELL conviction 76, steady all session; with a share placement adding supply and the position 55% under cost, we cut most of it. OPS: Sell_Rank 1, 85.46% x = (lot 100, R7 downward); fresh REDUCE cv>=65 re-opens proposability (2513 precedent); estimator +7.8%/+12.4% disagrees (T5/P2 down-weight); model stop 739.99 = close refused; residual WATCH 660.80. · news |
| 0823 | Link REIT | 36.72 | SELL 50% | 36.72 | 35.41 | 36.98 (+0.7%) | 37.31 (+1.6%) | 35.85 (-2.4%) | Link REIT closed below the watch line we pre-committed to act on while the model re-issued a STRONG SELL - REDUCE and its three-month view points lower, so we halve the position. OPS: re-arm 'close < 36.74 OR REDUCE cv>=65' price leg met by 0.03 ATR (T6); adj 59 below floor -> 50% tier not engine 85.46%; 12wk 35.85 negative; residual stop 35.41 (2.0 ATR), pre-execution line 36.74; takes slot 3 over 0981 (live REDUCE label vs decayed AVOID). · news |
| 0981 | SMIC | 62.40 | NO ACTION | 62.40 | 62.78⚠ | 62.72 (+0.5%) | 63.32 (+1.5%) | 66.30 (+6.2%) | SMIC slipped back under its pre-committed 62.78 line, but only marginally and its sell signal has faded to avoid with a positive three-month view, so it waits fourth in line behind the live sells. OPS: 0.15 ATR break; label S.SELL-RED 76 -> AVOID 63 (Portion 0); promotes on settled close < 61.55 (0.5 ATR) OR REDUCE cv>=65; T4 latch - promotion vs HK-11 retirement conflict flagged; model stop 67.31 above close refused. · news |
| 3750 | CATL | 507.00 | HOLD | 501.00 | 486.40 | 511.02 (+2.0%) | 516.51 (+3.1%) | 515.36 (+2.9%) | CATL's breakdown signal has turned into an oversold-reversal hold, so yesterday's unexecuted sale is retired to a watch on the recent capitulation low. OPS: 09-16 SELL @501.00 unexecuted; HK-11 amended (signal decay); close 507.00 still 0.24 ATR under 511.94; WATCH 486.40 - re-arm: close < 486.40 OR REDUCE cv>=65; model stop 465.24; HK-12 US-30 caution noted. · news |
| 1347 | Hua Hong Semiconductor | 110.20 | HOLD | 111.90 | 105.00 | 113.48 (+1.4%) | 116.47 (+4.1%) | 121.14 (+8.3%) | Hua Hong's high-conviction sell signal faded to avoid within a day, so the unexecuted sale is retired to its 105 watch line. OPS: 09-16 SELL @111.90 unexecuted; S.SELL-RED 76 -> AVOID 63; close 110.20; re-arm: close < 105.00 OR REDUCE cv>=65; model stop 122.47 above close refused. · news |
| 0522 | ASMPT | 160.00 | HOLD | — | 152.83 | 161.53 | 167.87 | 164.56 | Holding ASMPT on its 152.83 line; the model shows no trade and no sell case. OPS: NO TRADE x2 would be tick (d) but P1 bars profit-takes from PROFIT_ARMED on hk (position -18.5%); close 160.00. · news |
| 0700 | Tencent | 426.00 | HOLD | — | 416.17 | 421.89 | 419.37 | 415.80 | Holding Tencent on the 416.17 watch line; the avoid label is not a sell. OPS: close 426.00 (+2.4% above); stale overlay 430 exit line met but carries no force at 74 days; model stop 445.83 above close refused; re-arm close < 416.17 OR REDUCE cv>=65. · news |
| 2888 | Standard Chartered | 243.40 | HOLD | — | 236.74 | 241.31 | 240.76 | 237.36 | Holding Standard Chartered above its 236.74 stop. OPS: NO TRADE (not consecutive); 12wk negative but no stop break (L2 needs both); hkfin holds 600 - no conflict. · news |
| 0939 | CCB | 9.68 | HOLD | — | 9.45 | 9.69 | 9.75 | 9.65 | Holding CCB above its 9.45 stop on a steady hold signal. OPS: close 9.675; model stop 9.25 lower; hkfin holds - no conflict. · news |
| 7709 | XL2 CSOP SK Hynix (2x) | 39.50 | HOLD | — | 34.40 | 41.37 | 44.80 | 40.72 | Holding the leveraged Hynix tracker above its 34.40 stop; the leveraged-ETF exit bar is not met. OPS: AVOID 60; 2x rule needs cv>=70 + 3 prints + stop break; model stop 47.26 above close refused. · news |
Outcome & track record
Accuracy at the time of this record.
227 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 →
-0.220 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.
88 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.