Recommendations
2 to acquire · 0 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 |
|---|---|---|---|---|---|---|---|---|---|
| 3328 | Bank of Communications | 7.56 | ADD — QUEUED @7.53 | 7.53 | 7.42 | 7.53 (+0.0%) | 7.60 (+0.9%) | 7.71 (+2.4%) | Bank of Communications printed the strongest signal this book has recorded - a FailedBreakdown_Up_20D buy at conviction 82, the fourth consecutive rise from 54, with the shares above every one of their six moving averages and trading 2.1% above the volume-weighted average price without being overbought (RSI 63.5). The setup is not a chase: the close sits exactly at the model's entry level. Valuation is unavailable because the bank reports in renminbi against a Hong Kong dollar quote, so this is a purely technical case, and it is a strong one. We are adding to the bought earlier this week, but at half the normal size and only on a small dip, because Chinese bank interim results land in about a week and we would rather build the position through that event than into it. OPERATIONAL: half size (0.5x) on the confirmed 28-29 Aug mainland-bank 1H results window; limit set at the Est_1W point 7.53 per HF-4 (Est_1W 7.53 < Entry 7.56 = low-quality-entry signature -> demote to pullback-queue, never buy on conviction alone); 7.53 is inside the +/-0.25 ATR provenance band 7.526-7.594 off the printed Entry_Price 7.56. HF-3 inversion leg clears (Est_12W 7.71 > close 7.56). Sizing: 26.5% x 2,011,163.50 x 20% = 106,591.67, x0.5 = 53,295.84, /7.53 = 7,078 -> lot 1,000 -> = HK. Stop 7.42 is a raise from the recorded 7.33, 1.02 ATR cushion, risk HKExpiry 2026-08-26 is deliberately set BEFORE the 28 Aug BOC board meeting so no limit sits live into the print. ON EXPIRY: re-adjudicate at the then-live Entry_Price; do NOT re-post this limit (new queue_expiry counterfactual n=5 = -11.32% median 4wk cost from zombie limits). · news |
| 3988 | Bank of China | 5.51 | ADD — QUEUED @5.50 | 5.50 | 5.44 | 5.50 (+0.0%) | 5.57 (+1.3%) | 5.62 (+2.2%) | Bank of China holds a buy signal at conviction 72 with the shares above all six moving averages, 2.3% above their volume-weighted average price, and within 2% of the 52-week high - a steady uptrend rather than a spike. We already own bought earlier this week and are topping up modestly. Two things keep this deliberately small: the conviction reading has stopped rising, and the board meets on 28 August to approve first-half results, a binary event five trading days away that we would rather not increase exposure into. The limit sits just below the close, so we only add if the market offers a slightly better price. OPERATIONAL: half size (0.5x) on the CONFIRMED 28 Aug 1H board meeting. Conviction FLAT, so this ranks second behind 3328. HF-4 fires (Est_1W 5.50 < Entry 5.51) -> limit at the Est_1W point 5.50, inside the +/-0.25 ATR band 5.487-5.533. Sizing: 106,591.67 x0.5 = 53,295.84, /5.50 = 9,690 -> lot 1,000 -> = HK; position becomes at blended 5.4586 = 7.94% of book. STOP 5.44 is the engine print and a raise from 5.32, but leaves only 0.71 ATR cushion (-1.18%) - a break is a REVIEW TRIGGER, not an automatic full exit, since L2 needs stop-break AND a negative 12wk and the 12wk is +2.1% positive. Expiry 2026-08-26, before the 28 Aug print. ON EXPIRY: re-adjudicate at market, do NOT re-post. · news |
| 2888 | STANCHART | 234.00 | HOLD | 234.00 | 228.47 | 233.81 (-0.1%) | 235.66 (+0.7%) | 225.89 (-3.5%) | Standard Chartered is held at a 2.4% gain and stays held. The position was already trimmed 30% earlier this week and that sale settled today, so cutting again on the same unchanged information would be double-counting one piece of bad news. Nothing has actually deteriorated: the signal strengthened over the session, the price closed higher, and the shares remain above their volume-weighted average price. The only negative is a soft 12-week model projection, which is the same fact that triggered the earlier trim. We hold the line and keep the protective stop where it is. OPERATIONAL: PROFIT_ARMED + ADR-0012 tick (c) (Target_12Week 225.89 < 1.02 x Close 234.00) would fire a 30% profit-take. SUPPRESSED on the HF-2 mechanism: T12_Progress_Pct is Close/Target_12Week (234.00/225.89 = 103.59%), so the ARM and tick (c) are the same inverted-12wk fact read twice; Gain_ATR 1.16 is far below the 3.0 arm threshold so there is no independent winner-near-target basis; conviction ROSE; and the 30% cut on this identical state filled today at 231.40 (L3 - do not re-recommend the same trim for days). Counterfactual optional_trim n=50 median +1.72%/1wk saved by restraint. Stop held at the 228.47 PEAK; engine's live chandelier 224.74 is LOWER and is not adopted. Trail erosion 1.63%, far inside ADR-0020's 15%. RE-ARM: a second Exit_Warning code, a >=20pt conviction drop, or a decisive close below 228.47 fires the trim without further debate. |
| 3968 | China Merchants Bank | 49.44 | HOLD | 49.44 | 48.10 | 49.63 (+0.4%) | 49.88 (+0.9%) | 50.70 (+2.5%) | China Merchants Bank has turned constructive. The signal flipped from a bearish reading a few sessions ago to a buy-grade one at conviction 72, and the shares closed above all six of their moving averages for the first time since we bought them, now 1.6% ahead of cost. The 12-week projection is still pointing higher. Nothing here argues for taking profit; we simply hold and let the protective stop do its work. OPERATIONAL: PROFIT_ARMED for a 4th session but ZERO fire ticks - no second Exit_Warning (a), conviction ROSE (b), forward view alive at Target_12Week 50.70 > 1.02 x 49.44 = 50.43 (c), no NO-TRADE run (d). Armed-but-not-fired = mandatory trail-raise, which yields nothing: the 2.0x ATR chandelier is 49.44 - 1.902 = 47.54, BELOW the recorded 48.10, so the stop stays 48.10 (never lowered). Cushion 1.41 ATR. NOTE the engine prints a signal-level BUY at cv72 but Action HOLD with 0, so 3968 is NOT in today's add pool; a genuine 'BUY - ADD' print would make it one. |
| 0939 | CCB - H SHS | 9.24 | HOLD | 9.24 | 8.89 | 9.23 (-0.2%) | 9.36 (+1.2%) | 9.20 (-0.5%) | China Construction Bank was bought two days ago and is a shade ahead. The signal eased just below the level at which we would add more, so we hold the existing without topping up. The shares are still above all their moving averages and well above their volume-weighted average price, so the uptrend is intact - it simply is not strong enough right now to justify more money. OPERATIONAL: WATCH cv64, ONE POINT under the 65 add floor after ->64; held, so this was an ADD candidate not a new entry. STOP KEPT AT 8.89: the engine prints 9.15, a large raise, but that is only 0.53 ATR of cushion on a two-day-old position, below this run's 0.70 ATR adopt-a-raise floor - 9.15 is the next trail step once the line has room. Target_12Week 9.20 is marginally below the 9.245 close (-0.49%), worth watching but not an exit leg alone. RE-ARM: a BUY print at cv>=65 makes it an add candidate again. |
| 0388 | HKEX | 421.00 | NO ACTION | 421.39 | 403.82 | 421.90 (+0.1%) | 424.33 (+0.7%) | 425.83 (+1.1%) | Hong Kong Exchanges gave the cleanest buy signal in the book today - a 20-day breakout at conviction 79, above every moving average, with the model's own one-week and twelve-week projections both pointing higher. We did not buy it. The exchange operator trades well above what its earnings and book value justify, and it fails our financial-strength screen, and that screen is a hard block on opening any new position no matter how good the chart looks. This is the second session running that the signal has been blocked on those grounds. OPERATIONAL: STRONG BUY cv79, val_adj -6 (MoS_FV -1.186) -> adjusted 73, clears the 65 floor; BLOCKED by the hard Quality_Pass=False gate (F-Score 4) on new entries. HF-4 CLEARS here (Est_1W 421.90 > entry 421.39) - the only actionable name today where the estimation layer agrees with conviction - and HF-3's inversion leg clears too (Est_12W 425.83 > close 421.00). Ledgered optional/none so the counterfactual grader prices what the gate costs. SYSTEMIC: every mainland bank in this book reports in CNY against an HKD quote, returns NA_CCY, and therefore PASSES the gate with it switched off; 0388 and 2388 are the only HKD/HKD names the residual-income model can price and so the only two it can reject. The gate is screening exactly the subset it has data for - a code-level defect in swingtrade/fairvalue/, not licence to override today. |
| 2388 | BOC Hong Kong Holdings | 51.00 | NO ACTION | 51.00 | 50.00 | 50.80 (-0.4%) | 51.36 (+0.7%) | 52.48 (+2.9%) | BOC Hong Kong turned to a buy signal for the first time in weeks at conviction 72, but we passed. The shares screen as expensive against what the bank's book value and returns support, and the company fails our financial-strength test, which blocks new positions outright. Underneath the headline signal the internals are also thinner than they look - the trend is strong but momentum is only middling and the volatility reading is outright negative. OPERATIONAL: BUY cv72 after ->; val_adj -1 (MoS_FV -0.271) -> adjusted 71, over the floor; BLOCKED by hard Quality_Pass=False (F=4) on new entries. HF-4 also fires (Est_1W 50.80 < entry 51.00). Pillars: Trend 77 / Momentum 41 / Volatility BEARISH 25, Consolidation_Bullish only. Ledgered optional/none for counterfactual grading; see the 0388 card for the shared NA_CCY gate-asymmetry defect. |
Outcome & track record
Accuracy at the time of this record.
16 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 →
1.330 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.
50 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.