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 4 Aug 2026, 16:00:00 GMT+8

HK Semiconductors — End-of-Day Verdict

NO ACTION (both sides)

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 Limited25.16HOLD25.1623.0426.13 (+3.9%)28.40 (+12.9%)34.55 (+37.3%)Lenovo remains a hold, and the case got stronger. The position is up 7.2% on a bullish setup whose conviction has now risen three sessions running to, the shares sit 9% clear of our trailing stop, and the 12-week view improved to +37% - the only positive forward anywhere in this book. Valuation is neutral (fair value: no discount, no excess), so nothing argues for taking profit here. Operationally: not PROFIT_ARMED (Gain_ATR 1.04 vs 3.0; T12 progress 72.8% vs 90) so ADR-0012 has no fire rule to evaluate; the RSI_DIV warning is NOT elevated because MoS_FV -0.39 is far from the -3 that CLAUDE.md Sec9c requires, and G1/G3 independently forbid trimming a strengthening winner on divergence alone. Trail HELD at 23.04 - not lowered to the model's drifted Stop_Price 21.92 (HS-3a), and not raised because the position is not armed and 23.04 is already 1.31x ATR below the close, tighter than a 2x-ATR chandelier. Ex-dividend HKon 08-05 marks the reference down ~1.3% (mechanical, no stop threat). From the 08-06 EOD the 08-13 Q1 FY27 earnings enters the <=5-session catalyst window: default hold/trail through the print, no pre-trim. · news
2382Sunny Optical Technology57.80NO ACTION57.80n/an/a38.59 (-33.2%)Sunny Optical is still the cheapest thing in this book and we are still not buying it. Our valuation work puts fair value near HKagainst a HKprice - a real ~19% discount on a business that passes our quality screen, and the market's own consensus target sits at HKBut the price still trades 12% below its 200-day average, the 12-week model view is -33%, and this session the signal actually got worse rather than better. A cheap stock in a confirmed downtrend is a value trap until the tape turns; we buy the turn, not the discount. Valuation never originates a buy (CLAUDE.md Sec9) - recorded as a declined-buy counterfactual so the scorer prices what waiting costs or saves. Re-arm trigger unchanged: a bullish-family label (Pullback/Reversal) PLUS a reclaim of EMA200 ~65.58. Note the label degraded RallySell_BearTrend -> FailedBreakout_Down_20D while bearish conviction rose. · news

Outcome & track record

Accuracy at the time of this record.

10 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 →
-0.157 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.