Recommendations
1 to acquire · 1 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 |
|---|---|---|---|---|---|---|---|---|---|
| 3968 | China Merchants Bank | 48.90 | BUY — QUEUED @48.65 | 48.65 | 47.47 | 48.78 (+0.3%) | 49.09 (+0.9%) | 49.16 (+1.0%) | China Merchants Bank has pulled back into its rising trend and is now the strongest setup this book has produced since inception: a PullbackBuy_VWAP signal at conviction 71 and rank 1, with the shares reclaiming their 20-, 50- and 200-day averages and momentum turning up while RSI sits at a comfortable 53.9 -- strong enough to buy, not so extended that we are chasing. We are bidding HK, a quarter below the close, so the order only fills if the market comes to us; the HKstop keeps risk at HK (0.12% of the book) against a HKobjective, a 2.5:1 structure. Valuation is unreadable for this name (CNY financials against an HKD quote), so the case is purely technical and no valuation credit is claimed. Operational: (board lot 500) = HKnotional, from HKsized as 26.5% x HK book x 20% buy_pct, size modulator 1.0x (NA_CCY, no tier adjustment). Clears every gate by name -- floor 71 >= 65; HF-4 Est_1W 48.78 >= entry 48.65, which is the exact re-arm trigger written into the 08-05 card and the specific condition that blocked it yesterday; HF-3 Est_12W 49.16 = +0.53% vs close, non-inverted; L1 favoured PullbackBuy family; Entry_Distance_ATR -0.22 inside the -0.25 no-chase bar; quality gate OFF on NA_CCY though Quality_Pass is TRUE independently, the only such name in the book. Honest caveat: PullbackBuy_VWAP|buy has NEVER been graded in this book (scorecard bias 0, no actionable cells), and the model's own central path is flat (+0.27%/+0.90%/+1.05% at 1/4/12wk) -- the trade is carried by the 2.5:1 structure, not by a projected move. GTC expires 2026-08-11 (anchor + 3 sessions); withdraw earlier if conviction falls below 65. US Non-Farm Payrolls lands on the 08-07 open -- a below-market limit is the correct instrument into a two-sided macro print, since a gap up simply leaves it unfilled. WebFetch unavailable this run, so the news pass was search-only. · news |
| 2888 | STANCHART | 234.40 | SELL 30% | 234.40 | 222.07 | 235.96 (+0.7%) | 236.22 (+0.8%) | 228.26 (-2.6%) | Standard Chartered has now gone three straight sessions with the signal engine unable to form a view -- NO TRADE at zero conviction, pillars in open conflict (trend bullish, momentum weak) -- and the 12-week outlook has turned negative, widening from -0.8% to -2.6% below the current price. The long-term uptrend is genuinely intact and that is why we are cutting 30% rather than exiting: the shares still sit 23.5% above their 200-day average, H1 results in late July were a record (US8bn pre-tax, +9%, wealth income +38%, a fresh $1bn buyback), and momentum actually ticked up today. The decisive point is that our HKstop sits BELOW our HKcost, so a stop-out would realise a LOSS of ~HK -- the +HKopen gain is not protected by anything. Selling at ~HKbanks HKthe stop cannot defend and cuts stop-risk on the line from HKto HKwhile stay long the trend. RE-ISSUE of the 08-05 card, which NEVER EXECUTED and never could have: it was written with execution 'market_next_open', a token paper_executor.py does not recognise (_EXEC_KINDS = market|gtc_limit|gtc_stop), so _eligibility silently skipped it -- this card uses 'market'. The section 9 count did NOT reset because the reset clause in _state.json is conditioned on an EXECUTED cut; N is therefore 3, one past the 08-04 pre-commitment. Origination is CLAUDE.md section 9 N-consecutive-NoTrade ONLY -- NOT a profit-take (PROFIT_ARMED suppressed a 10th session under HF-2: Gain_ATR 0.78 vs the 3.0 threshold, T12_Progress 102.69 merely 234.40/228.26 on an inverted 12wk), NOT valuation (NA_CCY blind; the MoS -7.556 is a USD-financials-vs-HKD-price artifact and is deliberately NOT used to bump the tier to 50%), NOT a stop break. Volume: 30% x 1,550 = 465 -> board lot 50 -> (29.03%), proceeds HKrealised gain +HK, remainder on UNCHANGED stop 222.07 (2.0-ATR chandelier prints 219.62, below the ratchet floor). HF-5: the HKdividend went ex TODAY -- the 230.20 open reflected it and the 234.40 close more than recovered it, so this is mechanical, not deterioration, and no L8 reprice applies. Counterfactual optional_trim (n=40, +2.44% median saved @1wk, 4wk effect weakening 9.05% -> 6.71%) argues against and is consciously overridden: advisory-only by construction, its samples are trims against INTACT signals, and L3 forbids re-deferring a written pre-commitment. Horizon: 08-07 NFP is macro with a tighten-trail stance, compatible with a 30% cut; no 2888-specific event inside 5 sessions. Post-execution: re-base the section 9 count from 08-07 -- a further trim needs a NEW basis (decisive close below 222.07, a non-PROFIT_ARMED Exit_Warning, a fresh high-conviction SELL, or a fresh 2-consecutive-NoTrade run); raise the stop only when the 2.0-ATR chandelier exceeds 222.07. · news |
| 0939 | CCB - H SHS | 8.86 | NO ACTION | 8.86 | 8.71 | 8.81 (-0.6%) | 8.80 (-0.7%) | 8.67 (-2.1%) | CCB is the book's best-improving name that we are still not buying. Conviction has climbed -> and the shares rose 0.68% today with momentum turning up, but the label is WATCH rather than an actionable BUY, and the model's own 12-week projection sits 2.1% BELOW today's price -- we do not buy a name the forecast says will be lower in a quarter. Withheld, not rejected. Two independent blocks: WATCH is not an actionable label so it never enters the buy pool, and HF-3 fails (Est_12W 8.67 vs 8.86 close, -2.1% inverted). Valuation NA_CCY -> val_adj 0, quality gate off, Quality_Pass FALSE anyway. Ledgered as a counterfactual so the grader prices the non-purchase. Re-arm: an actionable BUY label at cv >= 65 with a non-inverted Est_12W vs close. Calendar: China CPI/PPI 08-10, aggregate financing ~08-14. · news |
| 0388 | HKEX | 410.00 | NO ACTION | — | — | 410.21 | 410.51 | 417.81 | HKEX was yesterday's top-ranked buy candidate and has already fallen apart -- from BUY at conviction 64 to NO TRADE at zero in a single session, with the shares down 1.11% and the signal pillars back in conflict. It also remains expensive on our valuation screen and fails the balance-sheet quality test, so we are not buying it. This is the second time in three days the quality gate has been vindicated by the signal decaying on its own. Blocked three ways: signal collapsed to NoTrade cv0 (consensus Conflicted, Momentum back to 15); OVERVALUED_VS_RI MoS -1.118 -> val_adj -6; hard quality gate, Quality_Pass FALSE on an F-Score of 4, which bars a new entry regardless of signal. The one thing that improved is Est_12W 417.81, now non-inverted at +1.9%. Also L1-discounted (breakout family) and the only graded buy cell in this book, BreakoutUp_20D|buy, is 1/2 at n=2 (report-only, no bias applied). Ledgered as a counterfactual. Re-arm: Quality_Pass TRUE plus an actionable BUY at cv >= 71 raw (to clear the 65 floor after the -6) with a non-inverted 12wk. · news |
Outcome & track record
Accuracy at the time of this record.
3 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 →
0.184 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.