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 28 Jul 2026, 16:00:00 GMT+8

HK Tech — End-of-Day Verdict

EOD 2026-07-28

Recommendations

1 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
9618JD-SW124.10BUY — QUEUED @123.73123.73112.52130.46 (+5.4%)146.15 (+18.1%)136.88 (+10.6%)JD has broken out to a fresh 20-day high on nearly three times its recent volume, closing at 124.10 and reclaiming its 200-day average for the first time in a month - the strongest signal this book has produced since inception, at conviction 84 against a 65 floor. Crucially the model now projects the shares higher on both horizons (+18% over four weeks, +11% over twelve) while the shares trade near our estimate of intrinsic value on a passing quality screen, and the suggested stop sits a wide 9% below entry - resolving the one objection that kept us out for four straight sessions, when the stop was too tight to survive normal daily noise. Risking 11.21 a share to make 22.42 is a 2-to-1 proposition, and the price has not run away from the entry. We buy on a stop order so that a weak open simply never fills. Learnings applied: the 07-27 re-arm trigger (fresh BUY cv>=65 + forward view intact + stop >= ~0.5 ATR) fired on every clause - stop is now 3.18 ATR vs 0.31 yesterday. HT-4 forward-view screen PASSES for the first time since inception (T4 +17.8%, T12 +10.3% vs close). HT-3 does not apply - this is the BreakoutUp exit family with genuine 2.00 R:R, not the pullback family's 0.92 config artifact. L1 breakout discount and the BreakoutUp_20D|buy cell being 0-for-1 (that one sample is THIS name's 07-21 failure: cv69 -> NO TRADE cv0 in one session, expired unfilled) are expressed as a 0.5x discretionary size haircut, NOT a reject - today's print is stronger on every axis that failed then (cv84 vs 69, STRONG BUY vs BUY, portion 74.69 vs 28.39, close 124.10 clears the 07-21 high 123.40, volume 14.2M vs 5.0M). L6 does not bite - no earnings inside the 10-day horizon. L8: executes at the 07-29 open; reprice if it gaps >~1.5%. Sizing: 74.69% x 2,000,000 x 20% = 298,760 full slot; size_mod 1.0x (UNDERVALUED+Quality_Pass would permit 1.25x but MoS is knife-edge +0.0067 on a 4-day-stale price, ~-0.04 marked to today's close, so upsize declined); 0.5x haircut -> 149,380 -> /123.73 = 1,207 -> ( x 50) = HK7.4% of book. Risk to stop HK = 0.67% of book. max_position_pct and both non-binding. No DCF annotation: DCF_per_share 185.65 sits well above the 136.88 12wk target. Event Horizon names no event for 9618; FOMC 07-29 / Politburo 07-30 / PMI 07-31 are indirect and argue for the half-size, not a skip. Data caution: a cached Yahoo quote snippet showed 115.10 for 07-27 vs the scan's settled 119.20 - L4 makes the settled scan close authoritative, and the buy-stop is self-protecting if the tape is materially lower. WITHDRAW if conviction drops below 65 before fill, or on expiry 2026-07-31. · news
0148KINGBOARD HLDG42.22WATCH42.2233.1648.88 (+15.8%)64.43 (+52.6%)24.07 (-43.0%)Kingboard triggered a buy signal today, but it is a bounce in a collapse rather than a turn: the shares have fallen 26% in six sessions to 42.22 and trade far below their 20-, 50- and 100-day averages, and the company fails our quality screen on profitability and cash flow while still pricing in growth. The model's own twelve-week projection of 24.07 is 43% BELOW today's price - it is recommending a purchase it simultaneously forecasts will nearly halve. We do not buy falling knives on an oversold reading alone. Learnings applied: quality gate hard-blocks new entries on Quality_Pass=False (dispositive here - the book holds nothing, so no ADD exemption applies); HT-4 forward-view screen fails badly on the 12wk leg; note B (the HK OversoldBounce/ReversalUp buy pool with negative 12wk is the falling-knife shape this book is written to refuse). Momentum 29 / Volatility 21 pillars are weak. Fourth axis: the printed 4wk target 64.43 (+52.6%) sits far outside the model's own Est_High of 47.73 - an artifact, not an objective (logged as systemic finding 2). adj conviction 66 - 1 = 65, exactly at the floor, and rejected regardless. RE-ARM: Quality_Pass=True AND Target_12Week above close AND price back above EMA50. · news
2318AAC TECH57.10WATCHn/an/a51.94The buy signal we declined for four consecutive sessions has now disappeared entirely - AAC prints no trade at zero conviction. Our reason for passing was that the model forecast the shares lower on both the four- and twelve-week horizons even as it flagged a buy; that contradiction has resolved by the signal dissolving rather than by the forecasts improving. The book never paid 56.65 for a setup that never became sound. Learnings applied: HT-4 (forward view screened before the conviction floor) called this correctly four times running; L1 caution on extended pullback entries. Conviction trail ->. Dropped as an active candidate. RE-ARM: cv>=65 AND Target_4w/Target_12Week both above close AND entry not extended. · news
PORTFOLIOAI - HK TechWATCHn/an/an/aThe book takes its first position since inception: a half-size stake in JD.com, which today became the first name to satisfy every one of our entry conditions at once - a strong breakout signal, projected gains over both the four- and twelve-week horizons, a sensible valuation on a passing quality screen, and a stop wide enough to survive ordinary volatility. We commit HKabout 7% of the portfolio, and keep the remaining HKin reserve. That restraint is deliberate: eight other names in the book are flagged as active breakdowns with rising conviction, and a Federal Reserve decision, three chip-sector earnings reports and a Chinese policy meeting all land within four sessions. One breakout does not make a bull market. Ends the eleven-session 100%-cash streak. Sell pool empty by construction (zero holdings). Second floor-clearer 0148 hard-blocked by the quality gate. Nothing pending to reconcile coming in; monitor clean (CASH no_stop only).

Outcome & track record

Accuracy at the time of this record.

5 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 →
-0.450 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.
22 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.