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 | 142.80 | SELL 85.46% | 142.80 | 124.88 | 141.66 (-0.8%) | 148.20 (+3.8%) | 152.44 (+6.8%) | Hua Hong printed the strongest sell signal in the book today: a STRONG SELL - REDUCE held across five consecutive intraday prints with conviction rising into the close, on the failed-breakout-down pattern that is this book's best-sampled and most reliable exit signal. The shares closed below both their 20-day and 50-day trend lines and the stock is priced for growth it is not delivering, so we bank most of the position into today's 6.8% rally and keep a small residual holding in case the bounce extends. OPERATIONAL: adj 88 = 86 + 0 val + 2 bias (FailedBreakout_Down_20D|sell n=56). G2 SUPPRESSES the MoS -17.57 amplify - DCF_per_share blank and Quality_Pass False, anchor unreliable. HK-9: 1-week de-risk, the positive 4wk/12wk points are the acknowledged cost. Supersedes the unexecuted 08-11 card @133.30 (would have filled at the 133.70 open; inaction saved HK) - PROPOSAL #2 on this thread, retire at #3 per HK-11. Board lot 500 ASSUMED - verify. Remaining 500 trail 124.88 (2.0xATR off the 146.30 high). News low-confidence: Yahoo returned 106.80 vs the 142.80 scan close (systemic finding #4) - not allowed to override the scan per L4. · news |
| 0981 | SMIC | 67.40 | SELL 85.46% | 67.40 | 61.72 | 68.08 (+1.0%) | 70.88 (+5.2%) | 70.37 (+4.4%) | SMIC has broken down through every trend line that matters - the close sits below its 20-day, 50-day and 200-day averages - and the sell signal has now held steady across five consecutive prints. Renewed US-China trade tension and rare-earth export controls are pressuring Chinese chipmakers directly, with margin compression the standing risk to the domestic-utilisation thesis, and the shares remain priced for growth. We cut the bulk of the position and retain a small stake. OPERATIONAL: adj 75 = 73 + 0 val + 2 bias. G2 SUPPRESSES the MoS -4.66 amplify (blank DCF, Quality_Pass False). HK-11 AMBIGUITY FLAGGED: 0981 was retired to a WATCH line with a PRICE re-arm 'decisive close < 58.86' which has NOT fired at 67.40; re-armed instead on a FRESH signal (AVOID 60 -> STRONG SELL - REDUCE 73 over 5 prints), consistent with how 0939/3750 re-arms were written ('fresh SELL-REDUCE or close < X'). HK-9: all three estimates sit above the close - the acknowledged cost of a 1-week de-risk. Board lot 500 ASSUMED - verify. Remaining 500 trail 61.72 (2.0xATR off the 68.50 high). · news |
| 7709 | XL2CSOPHYNIX | 31.52 | SELL 30% | 31.52 | 28.00 | 32.84 (+4.2%) | 32.46 (+3.0%) | 33.22 (+5.4%) | This 2x leveraged SK Hynix tracker is showing a strong technical sell signal, but the underlying is in a confirmed bullish run: the AI-driven memory upcycle is seen extending into 2027 on DRAM and NAND supply tightness, six brokerages have initiated with bullish ratings since July's Nasdaq listing, and the fund rallied 12.2% today. So rather than the full cut the signal calls for, we take a smaller 30% slice - real risk off a position down heavily from cost, without fighting a catalyst the model does not see. OPERATIONAL: adj 80 = 78 + 0 (NA_ETF) + 2 bias. TWO DEMOTIONS: (1) leveraged-2x-ETF rule caps the engine's 85.46% at 50% - never a full exit on a reversal-family signal (memory feedback_leveraged_etf_exits, seeded by the 7747 2026-05-13 exit that missed a +5.8% bounce); (2) NEWS OVERRIDE 50% -> 30% - all three model estimates sit ABOVE the close (1wk +4.2%, 4wk +3.0%, 12wk +5.4%). HK-12: model Stop_Price 13.74 vs a 31.52 close (-56%) is DEGENERATE and rejected; the remaining 1,260 carry the 28.00 WATCH line as their stop. HK-11: price re-arm (close < 28.00) NOT fired at 31.52; re-armed on the fresh STRONG SELL - REDUCE. DOLLAR-RANK INVERSION: HKranks below 2888's HKfired arm - took the slot on evidence quality (5-print STRONG SELL vs a mechanical tick on a rising-conviction HOLD). Board lot 100. · news |
| 0005 | HSBC Holdings | 161.40 | HOLD | 161.40 | 159.50 | 160.45 (-0.6%) | 160.77 (-0.4%) | 162.57 (+0.7%) | HSBC is flat-to-slightly-positive on every forward horizon and still trading constructively, so we hold the full position and let the trailing stop do the work rather than trimming into strength. OPERATIONAL: ARMED (T12_Progress 99.28) with tick (c) LIVE (12wk 162.57 < 1.02 x 161.40 = 164.63) - a 30% trim worth ~HKthe book's LARGEST candidate, deliberately NOT taken. HK-11: retired to a WATCH line on 08-06, re-arm 'close below 159.50', close is 161.40 - NOT fired, may not be re-issued. Stop held at the 159.50 watch line (tighter than the 157.30 model stop and the 157.02 chandelier). 14-AUG EX-DIVIDEND (second interim): the mechanical gap is NOT a stop break. Cost of this discipline recorded for measurement. |
| 0700 | Tencent | 461.60 | HOLD | 461.60 | 466.94⚠ | 459.41 (-0.5%) | 457.43 (-0.9%) | 465.76 (+0.9%) | Tencent closed just below its stop, but the results landed after the market shut - so today's close tells us nothing about how the market reads the quarter. We wait one session for the verdict rather than selling blind into an event that has already happened but not yet been priced. OPERATIONAL: STOP BREACH close 461.60 vs stop 466.94 (-1.14%), breaching close PRE-DATES the print. Require a CONFIRMED close below 466.94 on 2026-08-13 (the session after results) before exiting. ADR-0012 arm (PROFIT_ARMED + tick (d), 6 consecutive NO TRADE) OVERRIDDEN by the catalyst hold for one more session. DATA ARTIFACT: Avg_Cost 1.00 in the portfolio JSON yields a +46,060% and 34.99 Gain_ATR on this row - meaningless, ignored (systemic finding #6). · news |
| 2888 | Standard Chartered | 233.80 | NO ACTION | 233.80 | 224.39 | 235.37 (+0.7%) | 232.24 (-0.7%) | 225.69 (-3.5%) | StanChart is trading above its own 12-week target, which would ordinarily argue for banking some gain, but its signal is still strengthening and three genuinely deteriorating semiconductor names had a better claim on today's limited sell capacity. We raise the trailing stop instead and keep the position intact. OPERATIONAL: ARMED (T12_Progress 103.59) + tick (c) FIRES (225.69 < 1.02 x 233.80 = 238.48). A 30% trim = @233.80 = HKDISPLACED from the <=3 for a SECOND consecutive session - PROPOSAL #2 on this thread (HK-11: escalate once at #3, then retire). Conviction RISING on a bullish FailedBreakdown_Up_20D label independently supports the displacement. Trail raised 223.39 -> 224.39. |
| 0823 | Link REIT | 38.44 | HOLD | 38.44 | 37.86 | 38.53 (+0.2%) | 38.63 (+0.5%) | 39.52 (+2.8%) | Link REIT's forward view is positive across all three horizons, so trimming it would be cutting a position the model still expects to gain. We hold and keep the stop where it is. OPERATIONAL: ARMED (T12_Progress 97.27) + tick (d) (6 consecutive NO TRADE) technically fires a 30% trim worth HK - the 2nd-largest dollar bleed in the book. EXCLUDED ON G3 (forward-view gate): 1wk 38.53, 4wk 38.63, 12wk 39.52 = +2.8%, all positive; NoTrade_Reason is the low-information 'consensus conflict + low conviction'. Stop HELD at 37.86 (tighter than the 37.49 chandelier). Second session excluded on essentially this basis. |
| 2513 | Knowledge Atlas | 1208.00 | HOLD | 1208.00 | 989.29 | 1210.15 (+0.2%) | 1167.93 (-3.3%) | 1267.75 (+4.9%) | Knowledge Atlas is holding up with a rising conviction score and a potential index-inclusion catalyst on 21 August, so we stay in and simply lift the trailing stop to protect the position ahead of it. OPERATIONAL: ARMED (T12_Progress 95.29) but NO TICK FIRES - (c) does not fire (12wk 1,267.75 > 1.02 x 1,208 = 1,232.16), conviction RISING, label AVOID not NO TRADE, no second warning code. MANDATORY TRAIL-RAISE 975.86 -> 989.29 (2.0xATR chandelier off today's 1,285 high). CATALYST HOLD: Hang Seng quarterly review 21-Aug, inclusion contender. HK-12: model Stop_Price 1,503.71 is 24% ABOVE the close - degenerate, rejected (systemic finding #3). |
| 0522 | ASMPT | 169.50 | HOLD | 169.50 | 151.79 | 171.91 (+1.4%) | 184.75 (+9.0%) | 197.59 (+16.6%) | ASMPT carries the best forward outlook of anything in the book - up 9% over four weeks and 17% over twelve on the model's estimates - so despite a quiet signal we hold the full position and raise the trailing stop. OPERATIONAL: the ONE held name with NO arm at all (T12_Progress 85.78 < 90, Gain_ATR -2.66) despite 6 consecutive NO TRADE prints. Trail raised 145.50 -> 151.79 (2.0xATR off the 172.00 high). |
| 0939 | CCB - H Shares | 8.75 | HOLD | 8.75 | 8.44 | 8.75 (+0.0%) | 8.76 (+0.1%) | 8.64 (-1.3%) | CCB is going nowhere in either direction and its exit case was already retired to a watch line, so we hold and nudge the trailing stop up. OPERATIONAL: ARMED (T12_Progress 101.27) + tick (c) fires (8.64 < 8.925), but RETIRED 08-11 with re-arm 'fresh SELL-REDUCE or decisive close < 8.46' - label is HOLD(37) and close 8.75, so NOT re-armed (HK-11 forbids re-issue). Trail 8.38 -> 8.44. |
| 3750 | CATL | 654.00 | HOLD | 654.00 | 601.81 | 650.82 (-0.5%) | 638.77 (-2.3%) | 627.51 (-4.1%) | CATL's signal has improved again and its exit case is retired to a watch line that has not been triggered, so we hold and raise the trailing stop. OPERATIONAL: ARMED (T12_Progress 104.22) + tick (c) fires (627.51 < 667.08), but RETIRED 08-11 with re-arm 'close < 625.50 or fresh SELL-REDUCE' - HOLD(60) rising, close 654.00, NOT re-armed (HK-11). Trail 600.81 -> 601.81. |
| 9618 | JD.com | 123.00 | NO ACTION | 124.35 | 120.20 | 123.19 (-0.9%) | 124.77 (+0.3%) | 126.55 (+1.8%) | JD is the single best-quality buy candidate in the book right now - strongly cash-generative, fairly valued, with a positive outlook on every horizon and roughly four units of upside for each unit of risk. We are still standing aside, because the company reports Q2 results tomorrow evening and the stop that defines our risk sits only 2.3% away - an earnings gap would jump straight over it, leaving the position effectively unprotected. We would rather buy this after the numbers than gamble on them. OPERATIONAL: adj 71 = 65 + 0 val + 6 bias (PullbackBuy_VWAP|buy, n=12, hit_1w.778). THE ONLY buy-pool name that PASSES the quality gate (F-Score 6, FCF +HK6bn, Quality_Rank 6.5). Entry_Distance_ATR +0.37 - no-chase clean. Consistent with the 08-11 decision not to re-queue into this print. RE-ARM: fresh BUY at the 14-Aug SETTLED close at the then-printed Entry_Price - NO gap-chase (L6). |
| 0316 | OOIL | 148.60 | NO ACTION | 148.20 | 143.63 | 149.69 (+1.0%) | 152.18 (+2.7%) | 164.96 (+11.3%) | OOIL is the cleanest chart in the book - a genuine pullback entry, not a chase, trading above all its major trend lines with the best twelve-week outlook of any candidate at +11%, and it screens as undervalued. We are still not buying it, because the company is burning free cash flow and fails our financial-quality screen, and that screen is a hard rule for new positions. A cheap price on a cash-negative business is not the bargain it looks like. OPERATIONAL: adj 72 = 63 + 3 (UNDERVALUED, MoS +0.2985 -> +6 HALVED on Quality_Pass False) + 6 bias. The 08-11 re-arm FIRED cleanly (first settled-close BUY, Entry_Distance_ATR -0.10 inside the -0.25 bar, RSI 58.3). BLOCKED SOLELY by the hard quality gate - and legitimately: F-Score 5, FCF -HK, FCF_Yield -0.02%. 5TH CONSECUTIVE SESSION declined while being the best technical setup in the book - logged as a learnings-review candidate (systemic finding #1), NOT overridden. |
| 9988 | Alibaba | 122.60 | NO ACTION | 119.20 | 113.90 | 122.19 (+2.5%) | 120.42 (+1.0%) | 117.56 (-1.4%) | Alibaba scores highest of any buy candidate on signal strength, but it fails on everything that matters underneath: it is burning free cash flow at scale, the price has already run nearly 3% past the level the model wanted to pay, and the projected return is negative over one, four and twelve weeks with the shares still below their long-term trend line. Paying up for a weakening business with a negative forecast is how this book has lost money before. OPERATIONAL: adj 76 = 73 - 3 (GROWTH_PRICED_IN, MoS -0.687) + 6 bias. BLOCKED ON THREE INDEPENDENT GROUNDS: quality gate (F-Score 3, FCF -HK7bn, Quality_Rank 23.0 - worst in pool); no-chase (Entry_Distance_ATR -0.73); negative forward view on all three horizons with close below EMA200 127.24. Textbook HK-B falling-knife profile. |
| 3690 | Meituan | 91.65 | NO ACTION | 89.75 | 86.13 | 91.73 (+2.2%) | 90.73 (+1.1%) | 85.16 (-5.1%) | Meituan has already run well past the price the model wanted to pay, and the forecast turns negative beyond a week - down 7% over twelve. Chasing it here means buying high into a declining outlook. OPERATIONAL: adj 69 = 63 + 0 (bare NA) + 6 bias. 5th consecutive session declined: no-chase (Entry_Distance_ATR -0.59) plus negative 12wk (85.16 = -7.1%) and negative 4wk (90.73 = -1.0%). |
Outcome & track record
Accuracy at the time of this record.
161 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.