Recommendations
0 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 |
|---|---|---|---|---|---|---|---|---|---|
| 0992 | Lenovo Group Limited | 30.36 | HOLD | 32.14 | 29.96 | n/a | 37.18 (+15.7%) | 43.26 (+34.6%) | Holding Lenovo and declining the model's add. The shares fell 6.2% on nearly twice the usual volume and briefly traded below our protective sell level before recovering to close above it, so the position stays - the underlying uptrend is intact, the stock is still well above its long-term averages, and the 12-week view of HKis 42.5% above the close. But we are not adding: the model wants to buy at HK, which is 5.9% ABOVE where the shares actually settled, and our protective level now sits just 1.3% below the price. Paying up into a heavy selling day, on a position that is one bad session from being sold entirely, is not a risk we will take. Fair value is neutral (broadly fairly priced), and no company news explains the fall - it looks like a broad China-tech risk-off day. || OPERATIONAL: BUY-ADD cv77 PullbackBuy_VWAP Buy_Rank 1; adj conviction 77 + val_adj -4 + bias 0 = 73, clears the 65 floor; size mod 1.0x; Entry_Distance_ATR +0.88 (below entry, not a chase in model terms). REJECTED on (a) risk geometry - live monotonic trail 29.96 is 0.20 ATR/1.32% below close and was pierced intraday (low 28.62); model Stop_Price 29.55 was ALSO breached on the session it printed; today's range 3.38 = 1.68 ATR, so a 39.22% size-up sits at maximum stop-out probability; and (b) provenance - the only publishable action band is Entry 32.14 +- 0.25 ATR = [31.64, 32.64], i.e. +4.2% to +7.5% over the 30.36 settled close, a dollar chase forbidden by L1/HS-5 (no_chase_queue counterfactual n=9, 4wk median +23.19% saved). Cell PullbackBuy_VWAP|buy is n=4, hit_4w 0.000 - below the n>=5 floor so bias 0 applied, warning printed, no haircut. RE-ARM (WATCH, not queued): settled close back above VWAP 31.56 WITH >=1.0 ATR of chandelier buffer restored (settled close ~>=31.98 vs the 29.96 trail) and PullbackBuy_VWAP still printing - then reconsider at the THEN-CURRENT Entry_Price, never at 32.14. STOP: HOLD 29.96 (chandelier from the 32.00 high with ATR 2.0132 = 27.97, lower, so monotonic rule keeps 29.96; model 29.55 rejected per HS-3(a) as a downward drift). Not ADR-0012 armed (Gain_ATR -0.39, T12_Progress 70.18); conviction ROSE so tick (b) points the other way; Exit_Warning blank; L2 fails both legs; ADR-0020 not fired (peak stop = live stop, 0% erosion). A decisive close below 29.96 fires HS-3 unconditionally. |
| 2382 | Sunny Optical Technology | 67.30 | HOLD | 67.30 | 67.02 | n/a | 72.90 (+8.3%) | 79.88 (+18.7%) | Holding Sunny Optical into Thursday's Apple keynote rather than trimming ahead of it. The shares fell 3.2% and dipped below our protective sell level intraday before closing just above it, but the trend read is still constructive - price remains above its 20-, 50- and 200-day averages and the 12-week view of HKis 18.7% above the close, so there is no sell case on the signals. The position is 7.1% underwater, which by our rules is precisely when we must NOT dress a defensive cut up as profit-taking. The Apple iPhone 18 launch on 10 September sets camera-module volumes for two quarters and cuts both ways (roughly +/-5%); our discipline is to hold through a scheduled catalyst and let the pre-set sell level do the work rather than guess the outcome. Valuation is essentially neutral. No company news in the window. || OPERATIONAL: HOLD cv67 FailedBreakdown_Up_20D, conviction IMPROVED (+7), no ADR-0012 tick. NOT armed - Gain_ATR -1.84, T12_Progress 84.25 (<90) and the position is -7.11%: the HS-2 target-exhaustion artefact, never a profit-lock. Exit_Warning blank. L2 fails both legs (close 67.30 > trail 67.02; T12 +18.7%). HS-3 does not fire on the 66.60 wick - L4 settled-close authoritative. Valuation GROWTH_PRICED_IN MoS -0.0326 is a hairline crossing of zero (Systemic #6), nowhere near the -3 elevate bar, and amplify-only means nothing to amplify. Scorecard cell FailedBreakdown_Up_20D|buy +4 (n=8, hit_1w.714, hit_4w 0.000) is ACTIONABLE but held report-only per learnings section C - nothing actionable rides on it. STOP: HOLD 67.02 (chandelier from the 69.75 high with ATR 2.8000 = 64.15, and from the 72.45 entry = 66.85 - both lower, so the monotonic rule keeps 67.02; model 66.46 rejected as a downward drift, SIXTH consecutive rejection on this name). Buffer has compressed to 0.10 ATR / 0.42% below the close, from 0.93 ATR on 09-07. CATALYST: 09-10 Apple keynote two-sided [-5.0,+5.0] pct / [-0.9,+0.9] ATR, confidence med, stance hold-through (band anchor 0522), PLUS TSMC August monthly revenue the same session. A -5% day from 67.30 prints 63.94, decisively THROUGH the 67.02 stop - that tension stands deliberately: trail through, never pre-trim, but a decisive break fires HS-3 unconditionally. ADR-0020 not fired (peak stop = live stop, 0% erosion, position never cut). |
| — | AI - HK Semiconductors | — | NO ACTION | — | — | n/a | n/a | n/a | No trades today. A broad risk-off session took the book down 3.0% - nine of twelve names closed red and both holdings briefly traded below their protective sell levels before recovering to close above them. The one buy signal the model produced (adding to Lenovo) asked us to pay 4-8% more than the shares actually settled at, on the same day they fell 6.2% on double the usual volume, so it was declined. The only other buy candidate, YOFC, fails our conviction floor, our valuation screen and our balance-sheet quality screen. Neither holding gives any reason to sell. We keep HK - 48% of the book - in cash and wait for a better entry. || OPERATIONAL: 0 ACQUIRE, 0 DISPOSE, 2 STOP HOLDS (both at monotonic peaks, both chandeliers now compute lower). Buy pool = 0992 BUY-ADD cv77 -> adj 73, REJECTED on risk geometry (live trail 0.20 ATR below close, pierced intraday; model stop 29.55 also breached on the printing session) and on action-price provenance (band [31.64, 32.64] = +4.2%/+7.5% over the 30.36 close = dollar chase, L1/HS-5); and 6869 BUY cv62 -> adj 52, triple-blocked (3 under the floor raw, val_adj -10 on MoS -12.383, Quality_Pass False hard gate, plus the 09-07 +9.9% gap under L6/HS-5). ANTICIPATE slot EMPTY, fifth consecutive session. open_orders EMPTY - nothing to reprice, withdraw or expire. Book HK (cash 1,059,710 = 47.6%; 0992 910,800 = 40.9%; 2382 255,740 = 11.5%), down HK (-2.99%); unrealized swung +25,160 -> -43,390. Valuation FRESH (FetchedAt 2026-09-07, 1 day). Scorecard 86 entries / graded 19 / pending 0, calib_1w -0.097 degenerate -> f=0.30 per L5, not consumed (no action carries an estimate). News none found in window - both Yahoo news pages returned HTTP 503 and the one macro item surfaced could not be date-verified, so it is recorded and NOT credited. Overlay 65 days stale; learnings ACTIVE block 52 days stale with 7 PROPOSED blocks pending; Step 8a explicitly SKIPPED per the batch directive. Systemic #1 ESCALATED and last night's 'defect tracks the strategy' conclusion WITHDRAWN - 0992 switched back to fixed_then_trail and still printed a breached stop, so the defect spans both exit strategies. |
Outcome & track record
Accuracy at the time of this record.
19 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 →
-0.097 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.