This website is for educational purposes only and does not constitute investment advice. Always do your own research and assess your own risk tolerance before making investment decisions.

HK End-of-Day Verdict 7 Sept 2026, 16:00:00 GMT+8

HK Semiconductors — End-of-Day Verdict

EOD 2026-09-07 (hksemicon)

Recommendations

0 to acquire · 0 to dispose

TickerNameCloseActionAction 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
0992Lenovo Group Limited32.36HOLD32.3629.96n/a35.52 (+9.8%)46.11 (+42.5%)We hold all 30,000 Lenovo shares and raise the trailing stop rather than take any profit. Friday's queued limit filled at the open at 31.82, roughly 1.8% below where a market order would have gone, and the position is now up 3.9% on a blended cost of 31.15. The signal cooled from a conviction-84 pullback buy back to a conviction-63 hold, but the price itself was flat on the day and the trend picture is intact - the shares sit above their 20-, 50- and 200-day averages and 2.7% above the volume-weighted average price, with a 12-week projection of 46.11 implying about +43% from here. Valuation is neutral at fair value, so nothing argues for selling and nothing argues for adding at this level. Operational nuance: conviction fell 21 points, which is an ADR-0012 deterioration tick (b), but the position is NOT armed (Gain_ATR 0.76 vs the 3.0 arm; T12_Progress 70.18 vs 90), so there is no profit-take to fire - note it, do not trim. Gain_ATR recomputed by hand per HS-1: the scan ran 4 seconds before the Auto-Executor fill so its row still shows @ 30.39 and a stale Gain_ATR 1.24 (Systemic #2). Stop RAISED 29.71 -> 29.96 = 2.0x ATR chandelier off today's 33.12 high, monotonic up, 1.52 ATR below close, HKat risk. Model Stop_Price 31.76 REJECTED - 0.38 ATR below close and exactly today's low, i.e. already touched on the day it printed; the exit strategy flipped fixed_then_trail -> fixed_with_time_stop this session, which is the controlled comparison behind Systemic #1. ADR-0020 not fired (Peak Stop re-based up to 29.96, 0% erosion). L2 fails both legs. No news in window; the 09-03 Qira/AirJet items are two sessions out and carry no credit. Horizon: 09-08 economic band [-2,+2] med touches this name. · news
2382Sunny Optical Technology69.55HOLD69.5567.02n/a74.98 (+7.8%)84.48 (+21.5%)We hold all 3,800 Sunny Optical shares and raise the trailing stop again rather than trim into a catalyst. The position is 4.0% underwater, but a losing position is not by itself a sell case: the signal is still a constructive hold at conviction 60, the 69.55 close sits above our stop, and the 12-week projection of 84.48 is about +21% higher, so none of the three sell-origination classes fires. Valuation nudged from undervalued to fractionally rich, but at a margin of safety of -0.03 that is a hairline crossing of zero, nowhere near the level that could amplify a trim - and there is no trim to amplify. Apple's 09-10 product event is a direct optics read-through three sessions away, and the catalyst-aware rule says trail through the event, never pre-trim. Operational nuance: conviction (-7, short of the 20-pt tick); ADR-0012 NOT armed (Gain_ATR -1.07; T12_Progress 82.33 is the HS-2 target-exhaustion artefact on a loss, not a profit-lock). Stop RAISED 66.64 -> 67.02 = 2.0x ATR chandelier on the 72.45 entry with today's ATR 2.7126, monotonic up, HKat risk. CAVEAT: the buffer compressed to 0.93 ATR below close (from 1.33) because price fell into a rising stop, and the keynote's two-sided [-5,+5] band would print 66.07 on a -5% day - through the stop. That tension stands deliberately: hold through the event, but if the stop breaks decisively HS-3 fires unconditionally. Model Stop_Price 68.65 REJECTED for the FIFTH consecutive session at 0.33 ATR below close (Systemic #1). ADR-0020 not fired (Peak Stop re-based up to 67.02, 0% erosion). Scorecard cell FailedBreakdown_Up_20D|buy +4 is actionable but held report-only per learnings section C. No news in window. · news
AI - HK SemiconductorsNO ACTIONn/an/an/aNo trades today. Friday's queued Lenovo add filled at the open at 31.82 -, taking the position to 30,000 - so the book is now 54% invested with HK (46%) still in cash, and tonight there was simply nothing new to do. Not one of the twelve names printed a buy trigger: two holds, two watch-list names, five bearish avoids and three stuck on consensus conflict, so there was nothing to buy at any price. Nothing was sold either - neither holding broke its stop, carried a sell label, or qualified for a profit-take. Both trailing stops were raised instead. Lenovo is up 3.9% on its new blended cost and Sunny Optical is 4.0% underwater, leaving the book about HKahead on open positions. Operational nuance: buy pool empty, anticipate slot empty for a 4th session, no open orders remain after the fill so there is no zombie limit to reprice or withdraw. Loudest near-misses: 3750 CATL printed the book's highest bullish conviction (69) but only as WATCH, with a 12-week view 28% BELOW spot - the negative-12wk falling knife the B-note forbids; 6869 gapped +9.9% but sits 18 points under the 65 floor with MoS -12.38 and Quality_Pass False (hard gate); 0068 ripped +29.9% on a NO TRADE cv0 with a 09-08 lock-up expiry pending. 1347's valuation was revised UNDERVALUED +0.048 -> GROWTH_PRICED_IN -18.63 tonight, a third independent disqualifier that retroactively confirms both index-inclusion non-chases. Valuation was 10 days stale and was REGENERATED (FetchedAt 2026-09-07). Horizon refreshed same-day; the 09-10 Apple event [-5,+5] on 2382 is the book's biggest scheduled risk. Learnings ACTIVE block 51 days old with 7 PROPOSED blocks pending; Step 8a skipped per batch directive.

Outcome & track record

Accuracy at the time of this record.

18 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.

This website is for educational purposes only and does not constitute investment advice. Always do your own research and assess your own risk tolerance before making investment decisions.