Hong Kong — End-of-Day Verdict
3 ACQUIRE (9618 buy-stop, 3690 queued limit, 0148 market) + 3 DISPOSE (7709 full, 2513 full, 3750 trim) - pending next open 2026-07-29
Recommendations
3 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 |
|---|---|---|---|---|---|---|---|---|---|
| 9618 | JD-SW | 124.10 | BUY — QUEUED @124.26 | 124.26 | 115.06 | 123.54 (-0.6%) | 125.50 (+1.0%) | 132.64 (+6.7%) | JD is the one name in the book showing genuine strength on a day the HK semis/AI complex broke - it closed +4.11% against a falling tape, above its 200-day average, and is the only buy candidate that passes the quality screen while still screening as undervalued on intrinsic value. The 20-day breakout is confirmed by three consecutive strong-buy prints with momentum not yet overbought, so the risk is a defined 7.4% to the stop against a 12-week view roughly 7% higher. The order is a buy-stop placed just above today's close, so it only executes if the breakout actually follows through. [HK-1: BreakoutUp_20D is a chase family, -5 scorecard haircut applied and the 1.25x undervalued size modulator DECLINED - never up-size a chase. HK-4 satisfied: entry above close = fills on strength only. L8 pending next open.] |
| 3690 | MEITUAN-W | 90.30 | BUY — QUEUED @81.64 | 81.64 | 77.08 | 89.67 (+9.8%) | 91.22 (+11.7%) | 84.18 (+3.1%) | Meituan prints the book's most reliable buy setup - a pullback-to-VWAP entry, the only buy signal family with a positive track record here - but the shares have already run 9.6% above the model's entry level, so chasing them now would be paying up for a setup whose whole edge is buying the dip. The order rests as a limit at 81.64 and simply does not execute unless the stock comes back to the level the signal actually calls for. Valuation is uninformative on this name (no usable intrinsic anchor), so the case is purely technical. [Repriced from the 07-27 queue at 80.33 to the live model entry per the provenance rule; conviction 76 >= 65 so the queue survives. HK-2 anchor cell +3. eD -2.15 -> no-chase. expires_on 2026-07-31, auto-withdraw if conviction drops below 65.] |
| 0148 | KINGBOARD HLDG | 42.22 | BUY | 42.22 | 34.01 | 45.67 (+8.2%) | 46.95 (+11.2%) | 48.53 (+14.9%) | Kingboard fell 12% in today's electronics-wide selloff into deeply oversold territory, and the model reads that as an exhaustion bounce with a 4-week view about 11% higher and a 12-week view about 15% higher; it also passes the quality screen. This is explicitly the weakest of today's three buys and the first to drop if capital is rationed: it is a single bearish-to-bullish flip on a crash day, the shares sit well below their 200-day average, and the signal family has no graded history in this book to lean on. Risk is capped by a stop 19% lower. [OversoldBounce_Up|buy has NO graded HK sample -> bias 0, reliability unknown, stated not adjusted. val_adj -1 from MoS -0.19. Book-note-B falling-knife exclusion does NOT apply - the 12wk view is +14.9%, not negative. L8 pending next open.] |
| 7709 | XL2CSOPHYNIX | 38.02 | SELL ALL | 38.02 | — | 40.59 (+6.8%) | 40.36 (+6.2%) | 41.98 (+10.4%) | This is a 2x leveraged tracker of SK Hynix that has now lost two-thirds of its value, fell 30% today alone, and closed decisively through its stop - and SK Hynix reports quarterly results tomorrow, which on doubled exposure means a plausible 28% move in either direction overnight. Holding a broken, deeply loss-making leveraged position into a binary event is the worst risk-reward on the book; the signal, the stop and the calendar all say the same thing. Exiting in full frees capital and removes the single largest source of overnight risk. [Leveraged-ETF full-exit bar met on BOTH branches: cv81 >= 70 + 3 REDUCE prints + close 38.02 < stop 44.00, AND a breakdown-family signal. 5th re-issue - EXECUTE (L3/HK-5). Printed Sell Limit @41.94 sits 10.3% ABOVE the close and can never fill -> overridden to MARKET (systemic finding #5). Est_* point mildly up but calib_1w is -0.283 on this book, i.e. anti-predictive - the estimates are reported and discounted. L8 pending next open.] |
| 2513 | KNOWLEDGE ATLAS | 1042.00 | SELL ALL | 1042.00 | — | 1082.89 (+3.9%) | 1035.92 (-0.6%) | 1159.46 (+11.3%) | The cornerstone lock-up release we flagged in advance landed today and the shares fell 19%, exactly the supply shock anticipated - and this was only one tranche of a record pipeline of HK lock-ups still to unwind. The position is 36% underwater, the company scores poorly on financial quality with negative earning power, and the signal has now escalated from avoid to an outright sell on the most reliable exit family in this book. There is no valuation floor to fall back on. Exit in full. [7th consecutive unexecuted session - the skips have cost ~HK in this session alone. RallySell_BearTrend|sell is the book's best sell cell (+5, n=12, hit_1w.75). Est_12W +11.3% is discounted: calib_1w -0.283, no valuation anchor, ongoing supply overhang. L3/HK-5 - EXECUTE. L8 pending next open.] |
| 3750 | CATL | 603.50 | SELL 66.7% | 603.50 | 539.12 | 599.97 (-0.6%) | 581.73 (-3.6%) | 585.06 (-3.1%) | CATL has deteriorated steadily across three consecutive readings - from no-signal to avoid to an outright sell - which makes this a stable trend breakdown rather than a one-day wobble. The position is 20% underwater and, unusually, the model's own forward view now sits below the current price at both 4 and 12 weeks, so there is no projected recovery to wait for. Selling two-thirds banks the de-risk while leaving a residual stake in case the China policy meeting later this week reverses sentiment. [RallySell_BearTrend +5, the book's best sell cell. Negative 12wk (585.06 < 603.50) satisfies the trend-exit contrast rule - the trim-discipline gates G1-G3 do not apply to a bearish signal with a negative forward view. 85.46% of 300 = rounded DOWN to the 100-lot so it stays a trim, not an accidental full exit; residual on the 539.12 stop. MoS -1.24 is not <= -3 -> no valuation tier bump. L8 pending next open.] |
| 0005 | HSBC HDGS(USD0.50) HK O'SEAS | 162.00 | HOLD | — | 158.50 | n/a | n/a | 167.19 | RSI 72.87 and three straight no-trade prints, but the 12-week view is still modestly higher and the shares held up on a heavy day. Hold into the H1 results on 4 August with a raised trailing stop. [trail 157..50; horizon stance hold-through, +-2.16 ATR the window's largest single-name band.] |
| 0522 | ASMPT | 139.50 | HOLD | — | 133.00 | n/a | n/a | 160.89 | ASMPT fell 13.7% and gapped straight through its stop, so the level was never executable. The exit test still is not met: the model's 12-week view sits 15% above the current price and the shares remain above their 200-day average, and interim results land tomorrow - selling into a binary print after a 14% drop is the worst of both. Hold through the event on a re-based stop; a decisive close below the 200-day average after the results turns this into a sell. [L2 does NOT fire - stop-break requires a NEGATIVE 12wk and this is +15.3%. L4 - the cv25/27 intraday REDUCE is withdrawn at the settled close (AVOID cv50). Stop re-based 156..00, below EMA200 135.12 and today's low 136.00. Catalyst-aware hold.] |
| 0700 | TENCENT | 447.20 | HOLD | — | 430.00 | n/a | n/a | 442.83 | Tencent prints a sell on a failed breakout, trades below its 200-day average and has a flat-to-negative forward view - a genuine fourth disposal candidate that misses out only because three larger, cleaner exits take the slots. The label whipsawed twice within today alone and the standing 430 downside trigger is still intact at a 447 close, so the position is kept on a tightened stop rather than sold on an unstable print. [4th DISPOSE candidate, adj 75, only 3 slots. Stop tightened 415..00 (the standing overlay floor, tighter than the printed level). Promote to DISPOSE on a close below 430 or a second stable REDUCE print.] |
| 0823 | LINK REIT | 39.26 | HOLD | — | 37.89 | n/a | n/a | 38.98 | Link REIT was one of the few holdings to finish higher and its signal remains constructive, but the Fed decision tomorrow feeds straight through the HK dollar peg into REIT discount rates, so the trailing stop is tightened ahead of it. [trail 37..89 (printed stop). Horizon: FOMC +-1.87 ATR, stance tighten-trail.] |
| 0981 | SMIC | 69.10 | HOLD | — | 66.00 | n/a | n/a | 74.08 | SMIC turned mildly bearish in today's semis selloff but held up far better than its peers and remains clear of its stop, with a 12-week view about 7% higher. Hold into Q2 results on 6 August. [Printed Stop_Price 80.00 is ABOVE the 69.10 close and unusable (systemic finding #5) - carried stop 66.00 retained. Q2 06-Aug, +-9%, hold-through.] |
| 1347 | Hua Hong Semiconductor | 138.30 | HOLD | — | 133.00 | n/a | n/a | 149.20 | Hua Hong fell 6.9% and closed below its stop, but the break is shallow, the shares are still far above their 200-day average, and the 12-week view remains about 8% higher - the exit test is not met. Hold into Q2 results on 6 August on a re-based stop. [L2 does NOT fire - 12wk is +7.9%, not negative. Stop re-based 145..00, below today's low 135.60. NO TRADE cv0 x4: 'Consensus conflict + low conviction'.] |
| 2888 | STANCHART | 222.80 | HOLD | — | 215.07 | n/a | n/a | 224.85 | StanChart's conviction has faded from over four readings but the signal is still constructive and H1 results plus the buyback decision land tomorrow. Hold through the print on the existing trailing stop. [trail held at 215.07 (tighter than the printed 212.38). Horizon: H1 29-Jul, +-5%, hold-through.] |
Outcome & track record
Accuracy at the time of this record.
124 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 →
-0.283 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.
21 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.