Hong Kong — End-of-Day Verdict
1 ACQUIRE (2318 Ping An BUY 1500 QUEUED buy-limit @56.40 exp 08-11, adj 67) + 3 DISPOSE (2513 SELL 900/90% @1087 market - HK-11 escalation, FINAL, adj 76; 0981 SELL 1000/40% @65.35 market - fired re-arm trigger + sub-EMA200 breakdown, adj 66; 7709 SELL 900/50% @31.50 market - STOP BREACH + fired 33.00 cap, adj 64, trim not exit per the leveraged-ETF rule). Sell proceeds HKvs HKof conditional buying - massively; goes toward ~HK. 0005 RETIRED to WATCH (re-arm <159.50). 0522 sell WITHDRAWN. 0823 and 0992 GTC queues WITHDRAWN. 1211 BYD BLOCKED by the quality gate at adj 54 despite Buy_Rank 1. ANTICIPATE slot empty. Five fired profit-take arms displaced by the <=3 cap (0939, 1347, 3750, 2888, 0823) - 0939 for a 4th consecutive session. Stops: 7709 re-based 28.00, all others unchanged. SYSTEMIC: PROFIT_ARMED fires on held names (T12_Progress >=90 book-wide) - the ADR-0012 arm machinery is non-discriminating here, so all three cards deliberately rest on the signal layer instead.
Recommendations
1 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 |
|---|---|---|---|---|---|---|---|---|---|
| 2318 | Ping An Insurance (Group) [portfolio file mislabels as AAC Technologies] | 57.30 | BUY — QUEUED @56.40 | 56.40 | 55.10 | 57.09 (+1.2%) | 57.69 (+2.3%) | 60.66 (+7.6%) | Hong Kong insurance heavyweight Ping An pulling back to its volume-weighted average price while the trend underneath stays intact - the one buy setup this book has actually made money on. Conviction firmed through the session on the most reliable entry pattern we track, and the risk is tightly defined: roughly 2.3 units of expected gain for each unit risked, with the exit line only 2.3% below entry. Valuation is uninformative here (currency-mismatched data), so this is a technical entry, not a value call. Because the price already sits 0.78 ATR above the model entry we queue rather than chase (HK-4/HK-10) - GTC buy-limit 56.40, range 56.11-56.69, expires 2026-08-11, auto-withdraw if conviction drops under 65. Size 70.59% x 600k x 20% = HK -> (board lot 500). WARNING: input/hkportfolio.json names 2318 "Aac Technologies Holdings"; 2318.HK is Ping An (AAC is 2018.HK). Signal data is the correct 2318.HK tape. A web quote scrape showed ~72.50 vs our 57.30 settled close - spot-check the live quote before placing; the GTC is self-limiting. |
| 2513 | Z.AI Co., Ltd. (formerly Knowledge Atlas) | 1087.00 | SELL 85.46% | 1087.00 | 862.00 | 1096.48 (+0.9%) | 1035.19 (-4.8%) | 1197.12 (+10.1%) | The failed breakout has held for four straight readings at the book's highest conviction (73) and the top sell rank - the stock pushed to new highs, could not hold them, and is now rolling over. The business behind it is loss-making with negative shareholders', trades at a large premium to analyst targets, and has just launched a share sale that dilutes existing holders while giving its flagship model away free. We take 90% off and keep a 100-share stub purely for the Hang Seng index-review announcement due 14-21 Aug, where the company is a named inclusion contender. HK-11 ESCALATION - FINAL ISSUANCE: market-at-open, unconditional; after 2026-08-07 this name retires to a WATCH line and is never re-issued. Cards on 08-04 @1010 and 08-05 @1041 went unexecuted; the 07-17..07-28 cycle on this same name cost the book HKBoard lot 100 ASSUMED - verify before placing. Model Stop_Price 752.15 rejected as HK-12 degenerate width (-30.8% on a 15.4%-ATR name); residual stop stays 862.00. Est_1W +0.9% / Est_12W +10.1% argue against - discounted per HK-6 (calib_1w -0.261) and HK-9 (HK sell signals are a 1-week instrument); the Est_4W -4.8% is the horizon this is taken against. Index review is 6 sessions out, outside the <=5-session catalyst-hold window, so it does not block. · news |
| 0981 | SMIC | 65.35 | SELL 40% | 65.35 | 58.86 | 65.94 (+0.9%) | 66.69 (+2.1%) | 68.74 (+5.2%) | China's largest chip foundry has broken down out of its range and closed below its 200-day average - the line that separates a dip from a downtrend - on a day the whole Hong Kong semiconductor complex sold off. The shares are priced for growth the cash flows do not yet support, so we cut 40% and keep long the recovery rather than exiting. This is a 40% de-risk, not an exit. Fires the pre-committed re-arm trigger written into the 08-05 HK-11 retirement ("settled close <65.50 AFTER the 08-06 Q2 print"): price leg MET at 65.35. CAVEAT: the Q2 print itself could NOT be confirmed on finance.yahoo.com/seekingalpha.com, and the -3.9% move matches a broad semi selloff (1347 -3.4%, 0522 -2.6%) rather than an earnings reaction - so the post-Q2 precondition is UNVERIFIED. The card does not rest on the trigger alone: AVOID cv63 on 4/4 prints, breakdown family, close 65.35 below EMA200 68.54, adj 66 clears the floor standalone. Valuation amplify (MoS -4.39 <= -3, would bump one tier) SUPPRESSED by gate G2 - DCF_per_share is blank, no reliable anchor - so portion stays at the prior cards' 40%. Board lot 500 =. Model Stop_Price 74.18 sits ABOVE the close - rejected as HK-12 degenerate; residual stop 58.86. Est_* all lean against the exit, discounted per HK-6/HK-9. Watch TSMC July revenue ~10-13 Aug and AMAT FQ3 13-Aug for read-through. |
| 7709 | XL2CSOPHYNIX (CSOP 2x Leveraged SK Hynix ETF) | 31.50 | SELL 50% | 31.50 | 28.00 | 32.98 (+4.7%) | 32.08 (+1.8%) | 33.07 (+5.0%) | A 2x leveraged chip ETF that fell 19% in a single session and broke the 31.50-vs-33.00 risk line we set for it on 31 July. Leveraged funds compound losses and bleed value in choppy markets, so a broken line here is not something to sit through - we halve the position and keep a 900-share rider on a tightened stop rather than exiting into the panic. THE BOOK'S ONLY STOP BREACH (-4.55%) and the pre-committed cap from the 07-31 HK-11 retirement has fired - origination is the fired numeric trigger + breach, not the generic conviction floor (adj 64 sits 1 point under it). Breakdown-family signal, 4 stable prints cv59->61. TRIM NOT FULL EXIT per the leveraged-ETF rule: full exit needs conviction >=70, this is 61 (the 2026-05-13 7747 full exit missed a +5.8% bounce - that is why the rule exists). Position is -72.5% from a 114.56 cost. Board lot 100 ASSUMED - verify before placing. Model Stop_Price 58.90 is +87% ABOVE the close - the most extreme HK-12 degenerate line in the book; on a 43%-of-price ATR no ATR-multiple can bind, so the rider gets a hard numeric cap re-based to 28.00 (-11.1%). Est_* +1.8% to +5.0% discounted - on a 2x instrument with a 43% ATR the bands are near-meaningless and volatility decay is a structural drag the estimator misses. HK-11 evidence: 7709's unexecuted 07-10 card alone was worth +HK. |
| 0939 | CCB - H SHS | 8.86 | NO ACTION | 8.86 | 8.78 | 8.81 (-0.6%) | 8.80 (-0.7%) | 8.67 (-2.1%) | Armed profit-take FIRED on tick (c) - the 12-week projection 8.67 has fallen below the 8.86 close - and DISPLACED by the <=3 DISPOSE cap, not declined on merit. 30% = = ~HKthe second-largest dollar bleed in the book. FLAG: this arm has now been declined or displaced on FOUR consecutive sessions (08-03, 08-04, 08-05, 08-06) - that is the HK-11 pattern re-forming and it deserves a slot on 08-07 once 2513 clears. Passed over today because CCB prints a stable bullish HOLD cv63 ABOVE its EMA200 8.23 and its arm rests on the degenerate T12 machinery (see systemic finding 1), whereas the three cards taken rest on real signal/risk events. Stop held at 8.78 - already the tightest line in the book at 0.9% below close. |
| 3750 | CATL | 628.00 | NO ACTION | 628.00 | 625.50 | 628.42 (+0.1%) | 611.55 (-2.6%) | 585.48 (-6.8%) | Armed profit-take FIRED on tick (c) - 12-week projection 585.48 is 6.8% below the 628.00 close - and DISPLACED by the <=3 cap. 30% = = ~HKFell 5.3% today and now sits just 0.4% above its 625.50 stop; if that line breaks it becomes tomorrow's card outright. Passed over because conviction is only 52 (AVOID, no REDUCE label) and the stop has not yet broken. Model Stop_Price 695.53 sits above the close - rejected as HK-12 degenerate; 625.50 stands. Position -16.5% from a 751.80 cost. |
| 2888 | STANCHART | 234.40 | NO ACTION | 234.40 | 224.83 | 235.96 (+0.7%) | 236.22 (+0.8%) | 228.26 (-2.6%) | Armed profit-take FIRED on two ticks - (c) 12-week projection 228.26 below the 234.40 close, and (d) two consecutive sessions of NO TRADE - which would size a 50% trim = = ~HKDISPLACED by the <=3 cap. Passed over because StanChart ROSE today in a broadly red tape, sits far above its EMA200 189.83, and its arm rests on the degenerate T12 machinery (systemic finding 1). Third session this arm has been displaced. 348 is an odd lot - round any trim to board lot at execution. Stop held 224.83. |
| 1347 | Hua Hong Semiconductor | 136.70 | NO ACTION | 136.70 | 118.00 | 137.40 (+0.5%) | 148.42 (+8.6%) | 139.81 (+2.3%) | Armed profit-take FIRED on tick (d) - two consecutive sessions of NO TRADE - and DISPLACED by the <=3 cap. 30% = = ~HKThe 08-05 catalyst-aware hold has now EXPIRED: Q2 printed 2026-08-06 per the event horizon, so the "hold through the print" reason is spent and this name re-enters the pool on merit from 08-07. Passed over today because the forward view is the opposite of a trim case - Est_4W 148.42 is +8.6% over the close - and conviction is 0, not bearish. Fell 3.4% with the semi complex. Position -14.6% from a 160.00 cost. Stop held 118.00. |
| 0005 | HSBC HDGS(USD0.50) HK O'SEAS | 160.10 | NO ACTION | 160.10 | 159.50 | 159.38 (-0.4%) | 159.99 (-0.1%) | 161.45 (+0.8%) | RETIRED to a WATCH line under HK-11 - the escalation is spent and this exit must never be re-issued. Sell cards ran 08-03 @168.20, 08-04 @166.50 and 08-05 @162.10 (the escalated market-at-open); all three went unexecuted and today is session 4. The case has also lapsed on its own merits: HSBC now prints HOLD with conviction climbing -> on a bullish FailedBreakdown_Up_20D, and the close recovered to 160.10, back above the 159.50 line. RE-ARM TRIGGER: any settled close < 159.50. Note the 13-Aug ex-dividend will mechanically gap the shares ~0.5% - do NOT read that gap as a break of the line. |
| 0700 | TENCENT | 479.20 | HOLD | 479.20 | 466.94 | 475.00 (-0.9%) | 476.41 (-0.6%) | 500.10 (+4.4%) | Armed but NOT fired - no deterioration tick (12-week projection 500.10 still 4.4% above the close, conviction is only a 3-point slip, no secondary warning code). Mandatory trail-raise only: stop held at 466.94, which is tighter than both the model line 448.67 and a 2.0x ATR chandelier at 448.67. Q2 results 12-Aug are 4 trading sessions out, INSIDE the <=5-session catalyst-aware-hold window - hold through the print, do not pre-trim. Close 479.20 sits below EMA200 503.87, so a break of 466.94 is the exit trigger, not a discretionary call. |
| 1211 | BYD | 89.85 | WATCH | 89.85 | 88.86 | 89.15 (-0.8%) | 90.48 (+0.7%) | 95.58 (+6.4%) | Top-ranked buy candidate in the book and still rejected. The quality screen fails outright (Quality_Pass False), which blocks any NEW entry - BYD is priced for growth its cash generation and balance-sheet quality do not currently back. Valuation takes 7 points off (MoS_FV -1.498) and the signal family another 4 (FailedBreakdown_Up_20D grades -4 here per HK-1, a chase-grade pattern), leaving adjusted conviction, far under the 65 floor. The close 89.85 also sits below EMA200 96.61, so the "failed breakdown" is happening inside a downtrend. |
Outcome & track record
Accuracy at the time of this record.
144 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 →
-0.261 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.