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 |
|---|---|---|---|---|---|---|---|---|---|
| 3328 | Bank of Communications | 8.09 | ADD — QUEUED @8.07 | 8.07 | 7.75 | 8.06 (-0.1%) | 8.18 (+1.4%) | 8.14 (+0.9%) | Bank of Communications closed at a genuine 52-week high on the strongest trend reading in the book, and we are stepping back in with a disciplined limit rather than chasing the breakout. The bank trades on a single-digit earnings multiple with Beijing's recapitalisation of the big state lenders behind it, and the model's four-week and twelve-week views both sit above the breakout level, so the medium-term case supports adding - but the one-week view does not, which is why this is a pullback order and not a market buy. OPERATIONAL: signal BreakoutUp_52W [STRONG BUY] cv90, Buy_Rank 1, 77.96, Trend pillar 100, Regime Trending, close 8.09 = High_52W 8.09 exactly, above EMA20/50/100/200, VWAP +4.27%. HF-4 FIRES (Est_1W_Point 8.06 < the printed 8.12 buy-stop = LOW_QUALITY_ENTRY) and is applied AS WRITTEN - demote to a pullback QUEUE, not veto - so the buy-stop becomes a GTC limit at 8.07, the floor of the Entry_Price +-0.25xATR provenance band (8.074-8.166) and the level at which the one-week estimate no longer contradicts the fill. HF-3 non-inversion gate PASSES (Est_12W +0.62% vs close). Signal family upgraded off MomentumContinuation_Up (this book's 0-for-3, stop-first 1.0 cell) onto BreakoutUp_52W (1-for-1, report-only so bias 0). Volume: 77.96% x 1,500 = 1,169 -> HK board lot 1,000 ->; cost HK = 0.47% of the HKdry. Valuation NA_CCY -> val_adj 0, size x1.0, quality gate off (and it would never block an ADD). Stop RAISED 7.72 -> 7.75 (tonight's model stop; chandelier 7.723 and carried 7.72 both lower); risk 0.32/sh = 1.74 ATR. COUNTER-CASE DISCLOSED: RSI 71.70 rising, a printed RSI_DIV, and a fired profit-take tick (c) on the identical row, into a tighten-trail horizon - answered by the order's structure, a limit BELOW the market expiring 2026-09-16, the last HK session before the FOMC, so a fill is never carried into the Fed on a chased price. Counterfactual disclosed: no_chase_queue n=9, +4.97% median wh_1w [cost] - the price of the 09-09 and 09-10 flat rejections. · news |
| 2888 | STANCHART | 241.80 | SELL 26.7% | 241.80 | 233.48 | 239.46 (-1.0%) | 239.25 (-1.1%) | 241.07 (-0.3%) | Standard Chartered has run more than 5% above our cost and is now the only holding whose one-week, four-week and twelve-week model views all sit below the current price - so we are banking a partial profit rather than waiting for the forward view to prove itself right. The Hong Kong Monetary Authority follows the Fed mechanically on 17 September, and a rate cut compresses Hong Kong bank margins into exactly the window this trim steps out of. We keep running behind a raised risk line. OPERATIONAL: signal FailedBreakdown_Up_20D cv61 (trail ->, Exit_Warning PROFIT_ARMED, Regime Ranging, close above EMA20 235.83 / EMA50 229.31 / EMA200 200.61, RSI 61.65. ARMED GENUINELY - Gain_ATR 3.02 back above the 3.0 threshold (2.96 yesterday), so HF-2 does not apply and the arm does not rest on the T12_Progress 100.30 inversion artifact. FIRED: tick (c) only, Est_12W 241.07 below the 246.64 threshold; tick (b) does NOT re-fire (cv rose - the collapse was a one-shot vs the prior snapshot), no second Exit_Warning code, no NO TRADE prints. 1 tick -> profit_take_1 30%, a DE-ESCALATION from the unfilled 09-10 2-tick 50% card, which this supersedes. Volume 750 x 30% = 225 -> board lot., an exact tie -> (26.7%) rounded DOWN in deference to the n=89 optional_trim counterfactual (+1.39% median wh_1w [saved]). Proceeds HKrealised +HK (+5.76%) vs cost 228.63. Stop HELD 233.48 by ratchet - model 233.08 and chandelier 233.08 both lower; a stop is never walked down on a winner; cushion 1.91 ATR. Valuation NA_CCY, MoS -7.72 unusable -> val_adj 0, size x1.0, quality gate off. Scorecard bias -2: FailedBreakdown_Up_20D|sell turned actionable-NEGATIVE tonight (n=15, hit_1w 0.40) and argues mildly against this card - disclosed, not netted. IF THE UNFILLED 09-10 350-SHARE CARD BOOKS LATE: no further action, a 350 fill overshoots tonight's 550 destination to 400 - do not buy back. · news |
| 0939 | CCB - H SHS | 9.75 | SELL 50% | 9.75 | 9.37 | 9.72 (-0.3%) | 9.88 (+1.3%) | 9.85 (+1.0%) | China Construction Bank sits near a 52-week high and roughly 6% above our cost, but the model's confidence in the setup collapsed by 37 points in a single session - the sharpest deterioration anywhere in the book. We take half off into that strength and let the other half run behind an unchanged stop. Our forward views here are still mildly positive, so this is profit discipline on a deteriorating signal, not a call that the bank is going down. OPERATIONAL: signal FailedBreakdown_Up_20D cv50 (trail ->, Exit_Warning PROFIT_ARMED - yesterday's RSI_DIV has DROPPED OFF. Regime Trending, close 9.75 above EMA20 9.412 / EMA50 9.086 / EMA200 8.477, RSI 66.68, VWAP +3.85%, High_52W 9.88. ARMED GENUINELY - T12_Progress 98.98 on a NON-inverted Est_12W (+1.03% vs close), so the arm is real and HF-2 does not apply at all; Gain_ATR 2.35 (lowest in the book). FIRED: tick (b) conviction collapse = -37 pts (the largest single-session drop tonight) PLUS tick (c), Est_12W 9.85 below the 9.945 threshold. 2 ticks -> profit_take_2 50%. Tonight's card rests on a DIFFERENT and STRONGER leg than the unfilled 09-10 card, which fired on tick (a) RSI_DIV + (c); it supersedes it at identical size and price reference. G1 does not shield it - the softening is scoped to fresh BreakoutUp_*/MomentumContinuation_* at cv>=80 and this row prints FailedBreakdown_Up_20D at cv50. Volume 2,000 x 50% = = EXACTLY one board lot, no rounding; tier and lot arithmetic agree. Proceeds HKrealised +HK (+5.98%) vs cost 9.20. Stop HELD 9.37 by ratchet (model 9.28, chandelier 9.283 both lower); cushion 1.63 ATR. Valuation NA_CCY -> val_adj 0, size x1.0, gate off. Scorecard bias -2 (FailedBreakdown_Up_20D|sell n=15, hit_1w 0.40, newly actionable-negative). DISCLOSED COUNTER-ARGUMENT: both the 4wk (+1.33%) and 12wk (+1.03%) estimates are POSITIVE and the name trends above every EMA - on any read but the rulebook's this is a hold; it ships because tick (b) is exactly the deterioration the rule exists to catch and ADR-0012 forbids rendering a fired profit-take as optional prose. The n=89 optional_trim counterfactual argues directly against it and is recorded, not netted away. IF THE UNFILLED 09-10 CARD BOOKS LATE: the fill stands and no further action is taken. · news |
| 3988 | Bank of China | 6.05 | SELL 20% | 6.04 | 5.79 | 6.07 (+0.4%) | 6.05 (+0.1%) | 6.03 (-0.2%) | Bank of China is our largest winner at more than 11% above cost and the most stretched name in the book - trading 5% above its volume-weighted average price with momentum near overbought. We take a minimum 20% off the top to bank some of that stretch ahead of the Fed and the Hong Kong rate follow-through, and keep running. This is a profit-discipline trim on an extended position, not a change of view on the bank. OPERATIONAL: signal FailedBreakdown_Up_20D cv61 (trail ->, Exit_Warning PROFIT_ARMED, Regime Trending, close 6.045 above EMA20 5.762 / EMA50 5.524 / EMA200 5.082, RSI 69.63 (highest in the book), VWAP +5.07% (most extended in the book), High_52W 6.14. ARMED - THE MOST GENUINELY ARMED LINE IN THE BOOK: Gain_ATR 4.30, comfortably clear of 3.0 on its own, so HF-2 suppression LIFTS; T12_Progress 100.25; +11.12% unrealised, the largest gain in the sleeve. FIRED: tick (c) only, Est_12W 6.03 below the 6.166 threshold (-0.25%); tick (b) does not fire (cv rose, no second warning code, no NO TRADE. 1 tick -> profit_take_1 30%. Volume 5,000 x 30% = 1,500 -> board lot., an exact tie -> (20.0%) rounded DOWN - a 10pp deviation BELOW the intended tier, disclosed rather than silently re-tiered. Proceeds HKrealised +HK (+11.12%) vs cost 5.44. Stop HELD 5.79 by ratchet (model 5.77, chandelier 5.764 both lower); cushion 1.81 ATR. Valuation NA_CCY -> val_adj 0, size x1.0, gate off. Scorecard bias -2. DISCLOSED WEAKNESS - THIS IS THE THINNEST TICK IN THE BOOK: a -0.25% twelve-week inversion while BOTH the one-week (+0.41%) and four-week (+0.08%) views are positive. The card ships on the strength of the ARM, not the tick, and takes the minimum expressible tier. Its 09-09 card was withdrawn on 09-10 and not re-issued; it takes tonight's third slot because 3968 dropped out on HF-2 suppression. · news |
| 3968 | China Merchants Bank | 51.95 | HOLD 1,500 | 51.95 | 50.88 | 51.77 (-0.3%) | 52.14 (+0.4%) | 51.84 (-0.2%) | China Merchants Bank is held unchanged at and yesterday's proposed trim is cancelled. The only reason to sell was a twelve-week forecast fractionally below today's price - two-tenths of one percent - and that same fractional gap is what generated the profit-take trigger in the first place, so the case was circular rather than real. We hold the position behind an unchanged stop. OPERATIONAL: signal FailedBreakdown_Up_20D cv50 (trail ->, Exit_Warning PROFIT_ARMED, Regime Trending, close 51.95 above EMA20 50.84 / EMA50 49.54 / EMA200 48.77, RSI 62.50, +6.78% unrealised. HF-2 SUPPRESSION RE-APPLIES - Gain_ATR fell back 3.22 -> 2.92, BELOW 3.0, so the arm now rests solely on T12_Progress 100.21, which is generated by the -0.21% Est_12W inversion itself: the exact artifact HF-2 exists to catch. The only tick is (c), that same inversion read a second way; conviction fell (-7, short of the 20-pt tick (b) bar); NO independent bearish basis. The exact mirror of the 09-10 lift, in reverse. Stop HELD 50.88 by ratchet (model 49.69, chandelier 49.69 both lower) - cushion 0.95 ATR, THE TIGHTEST IN THE BOOK AND THE ONE TO WATCH. Valuation NA_CCY -> val_adj 0, gate off. · news |
| 3328 | Bank of Communications | 8.09 | HOLD 1,500 | 8.09 | 7.75 | 8.06 (-0.4%) | 8.18 (+1.1%) | 8.14 (+0.6%) | The existing of Bank of Communications are held and the stop is raised to HK. A profit-take trigger did fire, but taking it would have meant selling two-thirds of a position that just broke to a 52-week high - the Hong Kong minimum trading lot makes any smaller trim impossible - so we tighten the risk line instead and let the breakout work. OPERATIONAL: signal BreakoutUp_52W [STRONG BUY] cv90, Trend pillar 100, close 8.09 = High_52W 8.09, RSI 71.70, +7.87% unrealised, Exit_Warning RSI_DIV,PROFIT_ARMED, Gain_ATR 3.21 (genuine arm), T12_Progress 99.39 on a NON-inverted Est_12W (+0.62%). Tick (a) RSI_DIV is G1-DOWNGRADED to a trail-raise - fresh BreakoutUp_52W at cv >= 80 with a Tier-1 code, the NBIS protection. Tick (c) fires by only 0.62% (8.14 vs the 8.2518 threshold) -> 1 tick -> profit_take_1 30% = against a 1,000-share board lot -> the minimum expressible cut is 66.7%, TWO TIERS WORSE, on a line already cut 40% effective on 09-09. STOP RAISED 7.72 -> 7.75 (tonight's model stop; chandelier 7.723 and carried 7.72 both lower) - the mandatory ADR-0012 trail-raise carrying the downgraded RSI_DIV. Cushion 1.85 ATR. Paired with tonight's separate ADD QUEUED @8.07 card on the same name. · news |
| 2388 | BOC Hong Kong Holdings | 52.30 | WATCH | 52.30 | 51.00 | 52.13 (-0.3%) | 52.61 (+0.6%) | 53.01 (+1.4%) | BOC Hong Kong produced the highest-scoring buy signal in tonight's screen but we are not buying it: the stock fails our balance-sheet and cash-flow quality screen, and we do not open brand-new positions in names that fail it, however good the chart looks. We will revisit if the quality reading improves. OPERATIONAL: BUY cv72, Buy_Rank 2, FailedBreakdown_Up_20D, entry 52.30, stop 51.00, Target_Price 54.62, Entry_Distance_ATR 0.00 (not extended), Trend 77, above every EMA. Adjusted conviction 75 (val_adj -1 on MoS -0.290; scorecard bias +4 from FailedBreakdown_Up_20D|buy, n=7 hit_1w 0.714 - this book's BEST actionable buy cell) - clears the >=65 floor by ten points. BLOCKED BY THE HARD QUALITY GATE: Quality_Pass = False, F_Score 4, Quality_Rank 6.5 - the F-Score/FCF screen failed and the gate blocks NEW ENTRIES only (; it would never block an ADD to a held name). Independently HF-4 also fires: Est_1W_Point 52.13 < the 52.30 entry (LOW_QUALITY_ENTRY). Passes HF-3 (Est_12W 53.01, +1.36% non-inverted). RE-ARM TRIGGER: Quality_Pass flips True on a refreshed fairvalue AND Est_1W_Point >= Entry_Price. |
Outcome & track record
Accuracy at the time of this record.
43 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 →
5.000 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.