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 |
|---|---|---|---|---|---|---|---|---|---|
| 1347 | Hua Hong Grace Semiconductor | 130.20 | SELL 50% | 130.20 | 122.39 | 133.55 (+2.6%) | 138.85 (+6.6%) | 134.77 (+3.5%) | Hua Hong broke its stop on a brutal session - it opened at yesterday's close of 147.20, printed its high there and collapsed to 130.20, an 11.6% fall that took it below the 131.49 risk line we set yesterday and below its 20-, 50- and 100-day averages. We are selling half and keeping half: the stop break is a real risk-control signal and the position is already 19% underwater against cost, but the model's own forward view is still positive at every horizon, so this is a tactical de-risk rather than a verdict that the story is over. No company announcement explains the drop, and SMIC actually rose the same day, so this looks like a give-back of a sharp three-session run rather than sector news. OPERATIONAL: 08-13 WATCH line re-armed on BOTH legs (fresh STRONG SELL - REDUCE conv 65 at 10:30/11:30, AND settled close 130.20 < 131.49 trail) - a re-armed WATCH line is a FRESH thread #1, not an HK-11 re-issue. L2 combo INCOMPLETE (12wk +3.5% positive) so 50% not 100%, per HK-9 (sell = 1-week instrument, RallySell_BearTrend|sell hit_1w.667 -> hit_4w.250). Adj conviction 55 is BELOW the 65 floor - fires anyway as a CLAUDE.md section 9 stop break, same treatment as 0700's conviction-0 exit on 08-13. G2 SUPPRESSES the MoS -17.569 tier bump (DCF blank, Quality_Pass False) - no escalation. HK-12: model stop 155.01 is 19% ABOVE the close, degenerate, rejected; remaining 1,500 carry a 122.39 trail (2.0xATR off the 147.20 high). Board lot 100. · news |
| 0700 | Tencent Holdings | 440.00 | SELL ALL | 440.00 | 466.94⚠ | 443.06 (+0.7%) | 433.85 (-1.4%) | 430.62 (-2.1%) | We are closing Tencent entirely. The shares have now closed below our 466.94 risk line for a third straight session, finishing at 440.00, and the model's twelve-week view has turned negative - the combination we treat as a genuine exit signal regardless of what label the engine prints. The second-quarter results are the cause and they got worse on a closer read: revenue grew 11% but profit was essentially flat at RMB56bn against a RMB61.8bn consensus, while capital spending jumped 65% in a single quarter to RMB52.8bn on AI infrastructure. The stock is down 26% this year. A 15.6x earnings multiple is not expensive, but a broken trend with spending running ahead of returns is not a swing-trade setup. OPERATIONAL: HK-11 thread position #2 - re-issue permitted (08-13 was #1). If unexecuted again, 08-17 is #3 = escalate-then-retire. Close below EMA20 463.74 / EMA50 462.76 / EMA100 477.36 / EMA200 501.95, RSI 40.35. Full exit not a trim: stop-loss break, which the trim-discipline gates explicitly leave untouched. PROFIT_ARMED on this row is MEANINGLESS - Avg_Cost 1.00 is a data artifact (+43,900%, 31.94 Gain_ATR) and was ignored. 12wk 430.62 < DCF 482.54 so no DCF-anchor annotation. Stop retired on fill. Board lot 100. · news |
| 0823 | Link REIT | 38.78 | SELL 30% | 38.78 | 37.86 | 38.87 (+0.2%) | 38.78 (+0.0%) | 39.87 (+2.8%) | Link REIT has stopped going anywhere and we are banking 30% of it. The units sit at 97% of the model's own twelve-week target, the four-week forecast is dead flat, and the signal engine has printed no actionable view for eight straight readings - the definition of money that has finished working. Nothing is broken here: the price is above every major moving average, volatility is minimal, and there is no adverse news. That is exactly why this is a partial trim and not an exit - we are recycling capital that has stopped compounding, not fleeing a problem. OPERATIONAL: ADR-0012 fired profit-take - PROFIT_ARMED (T12_Progress 97.27) + tick (d) 2+ consecutive NO TRADE (eight prints). One tick -> profit_take_1 = 30%. Not gated by the conviction floor (fired profit-take is its own origination class). HK-11 position #3 after three consecutive displacements (08-11/08-12/08-13) - acted rather than retired because two slots freed up when 0981 and 7709 retired, and HK-5 argues for it (optional_trim n=51 verdict 'saved'). NoTrade_Reason verbatim: 'Consensus conflict + low conviction' - low-information. 12wk 39.87 EXCEEDS DCF anchor 37.84 on a GROWTH_PRICED_IN name - projection assumes growth beyond priced-in. Sizing 30% x 2,800 = (board lot 100). Remaining 2,000 hold the 37.86 stop (tighter than the 37.61 chandelier). · news |
| 0522 | ASMPT | 176.00 | HOLD - trail raised | 176.00 | 163.69 | 179.23 (+1.8%) | 192.84 (+9.6%) | 196.68 (+11.8%) | ASMPT keeps the best forward view in the Hong Kong book - the model sees roughly 10% upside over four weeks and 12% over twelve - and it is the only holding not flagged for any kind of profit-taking. The signal improved from avoid to hold. We are holding the full position and raising the protective stop to 163.69 to lock in more of the recovery. OPERATIONAL: The ONE held name with NO arm at all (T12_Progress 89.49 < 90, Gain_ATR -1.99). Printed a 184.10 high and CLOSED ON ITS LOW at 176.00 - the AMAT FQ3 read-through was sold into. Trail raised 161.13 -> 163.69 (2.0xATR off the 184.10 high). HK-12: model stop 172.94 is just 0.3xATR below the close - degenerate-narrow, rejected. |
| 2888 | Standard Chartered | 234.20 | NO ACTION | 234.20 | 228.47 | 236.19 (+0.8%) | 233.09 (-0.5%) | 225.98 (-3.5%) | Standard Chartered is the closest thing to a missed trim today. The shares trade above the model's own twelve-week target, the forward view has gone negative, and conviction collapsed from in a single session - a 27-point drop that would normally justify taking half off. We are not acting because we retired this exit yesterday with an explicit condition attached, and that condition has not been met: the shares closed at 234.20, above our 230.19 trigger. One decisive session resolves it either way. OPERATIONAL: ARMED (T12_Progress 103.64) + THREE ticks live - (a) dead-view class, (c) 12wk 225.98 < 1.02 x 234.20 = 238.88, and (b) conviction collapse = -27, well past the 20-point bar. Mechanically a 50% trim. NOT taken: HK-11 binds (retired 08-13, re-arm 'decisive close < 230.19 OR fresh SELL-REDUCE', neither fired). Ledgered optional:true so the counterfactual grader prices what not acting cost. Stop held 228.47 (tighter than the 227.22 chandelier); the 230.19 re-arm sits just above it. |
| 0005 | HSBC Holdings | 160.80 | NO ACTION | 160.80 | 159.50 | 160.41 (-0.2%) | 160.48 (-0.2%) | 163.50 (+1.7%) | HSBC went ex-dividend today, so most of the 1.4% dip is mechanical rather than a signal - the second interim payment lands on 25 September. On the numbers a 30% trim would be justified, and at roughly HKit would be the largest single action available in this book, but we retired that exit on 6 August with a clear condition - a close below 159.50 - and at 160.80 it has not triggered. We hold and keep the line where it is. OPERATIONAL: ARMED (T12_Progress 98.35) + tick (c) fires (12wk 163.50 < 1.02 x 160.80 = 164.02). NOT taken - HK-11 forbids re-issuing a retired exit. Conviction FELL. EX-DIV 14-Aug (record 14-Aug, pay 25-Sep): the mechanical gap must NOT be read as a stop break. Stop held at the 159.50 watch line (tighter than the 158.08 chandelier). Ledgered optional:true for counterfactual grading. |
| 2513 | Knowledge Atlas | 1270.00 | HOLD - trail raised | 1270.00 | 1114.23 | 1254.02 (-1.3%) | 1217.89 (-4.1%) | 1344.36 (+5.9%) | Knowledge Atlas gave back 3.6% in a wild session that ranged from. We are holding it through next Friday's Hang Seng index review, where it is a genuine inclusion candidate - selling into that kind of binary catalyst is how you give away the upside you were positioned for. The protective stop moves up sharply to 1,114 to define the downside instead. OPERATIONAL: ARMED (T12_Progress 94.47) but NO tick fires - (c) does not fire (12wk 1,344.36 > 1.02 x 1,270 = 1,295.40), (b) conviction fell only (-9, under the 20-pt bar), label AVOID not NO TRADE, no second warning code. MANDATORY trail-raise 1,041.04 -> 1,114.23 (2.0xATR off the 1,435 high). CATALYST HOLD: Hang Seng quarterly review announcement 21-Aug, within 5 trading sessions - trim only on a confirmed lower-high or a decisive stop break, never pre-trim into the catalyst. HK-12: model stop 1,590.77 is 25% ABOVE the close - degenerate, rejected. |
| 0939 | CCB - H Shares | 8.74 | HOLD | 8.74 | 8.52 | 8.71 (-0.3%) | 8.77 (+0.3%) | 8.72 (-0.2%) | CCB is doing nothing in either direction - flat forward estimates, a quiet 1% dip, and no signal worth acting on. We hold and nudge the stop up to 8.52. China's July credit data and the loan prime rate fixing on 20 August are the next things that could move it. OPERATIONAL: ARMED (T12_Progress 100.23) + tick (c) fires (8.72 < 8.91). RETIRED 08-11; re-arm 'fresh SELL-REDUCE or decisive close < 8.46' NOT fired at 8.74, label HOLD(39). HK-11 binds. Trail raised 8.48 -> 8.52 (2.0xATR off the 8.84 high). |
| 3750 | CATL | 641.50 | HOLD | 641.50 | 614.95 | 635.14 (-1.0%) | 631.96 (-1.5%) | 624.00 (-2.7%) | CATL slipped 1.7% with the forward view mildly negative, but our exit condition here is a close below 625.50 and at 641.50 it has not been met. We hold the position and keep the stop at 614.95. OPERATIONAL: ARMED (T12_Progress 102.80) + tick (c) fires (624.00 < 1.02 x 641.50 = 654.33). RETIRED 08-11; re-arm 'close < 625.50 or fresh SELL-REDUCE' NOT fired, label HOLD(39) is not a SELL-REDUCE. HK-11 binds. Trail HELD 614.95 (tighter than the 610.76 chandelier). |
Outcome & track record
Accuracy at the time of this record.
169 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 →
-0.249 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.
49 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.