Recommendations
1 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 | 26.28 | BUY | 26.39 | 21.31 | 29.44 (+11.6%) | 36.54 (+38.5%) | 37.34 (+41.5%) | Lenovo is the one name in this portfolio that is simultaneously rising hard, high quality, and still not asking us to pay more than we think the business is worth. It has just been named FIFA's Official Technology Partner for the 2026 World Cup - pairing near-real-time AI video infrastructure with a refurbished-server tie-up in Hong Kong - on top of the Saudi sovereign-AI build-out with Napster and DETASAD under Vision 2030, and AI-related revenue is now more than a third of sales. The shares broke to a new 20-day high on nearly double the usual volume, absorbing a dividend mark-down on the same day and still closing up 4.5%, with every moving average reclaimed and the strongest trend reading in the book. Crucially, our twelve-week projection still sits BELOW our own estimate of intrinsic worth, which is true of almost nothing else here - seventeen of twenty-one names we follow are priced beyond it, several dramatically so. We are buying half of the model's suggested size because the shares have already doubled in a month and this is a breakout, the pattern family that has been least reliable for us. | OPERATIONAL: STRONG BUY cv79 -> adj 77 (val_adj -2 on FAIR_VALUE MoS -0.3925; scorecard_bias 0 - BreakoutUp_20D|buy is n=2, hit_1w 0.50, below the n>=5 floor, report-only per HT-2). Buy_Rank 1, the only floor-clearing bullish row in a 21-name snapshot with 11 AVOIDs. Four-session rising base WATCH63 -> WATCH61 -> BUY76 -> STRONG BUY79 that STRENGTHENED through the anchor session ( 28.39% -> 74.69%) - unlike this book's prior one-print acquires. HT-4 screened before the floor and passes every leg: T4 +38.5%, T12 +42.1%, stop 21.31 = 3.08 ATR (definable, well clear of session noise), measured R:R 2.00. T12 37.34 sits INSIDE DCF_per_share 43.47 so no 'exceeds DCF anchor' annotation - Quality_Pass True, Quality_Rank 2/21, hard gate passes. SIZING: model slot 74.69% x 2,000,000 x buy_pct 20% = 298,760; size_mod 1.0 (MoS band -; L1 breakout-family haircut 0.5x -> 149,380; / 26.39 = 5,661 -> rounded DOWN to the of 2,000 -> = HK = 5.28% of book, risk-to-stop HK = 1.02%. The L1 haircut is applied by hand today (unlike 08-04) BECAUSE the engine printed a full 74.69% slot rather than the 28.39% partial that already embedded the family discount - 07-28 precedent: L1 governs size, measured geometry governs admission. REPRICED NOT CARRIED: the 08-04 card at 25.16 is closed withdrawn - it was never executable (execution 'market_open' is not in the executor's _EXEC_KINDS) - and this card uses the canonical 'market' value at the anchor's own live Entry_Price per the provenance rule. HT-6 re-checked and CLEAR: the US$2bn Alat/PIF convertible was priced in June 2026 and CAUSED the +114.8% 30-day re-rating rather than overhanging it; no new placement or registration found. RISKS: parabolic 30-day run; T4 36.54 overshoots Est_High 28.99 by 26% and the 52W high 27.42 - it is the chandelier exit's risk objective, not a forecast; Q1 FY27 results land 2026-08-13 (6 sessions out) and are ABSENT from the hk-parent horizon file - the hold-through-earnings call belongs to the 08-12 run, which the half-size is partly. f=0.30 on the 1wk per L5 (calib_1w -0.45, 7 graded buys < 10). · news |
| 0700 | TENCENT | 492.20 | WATCH | 496.23 | 447.63 | n/a | 593.44 (+19.6%) | 516.95 (+4.2%) | Tencent has turned back up and traded through the trigger level we cancelled yesterday, but it is still not a buy for us. The signal strength has only partly recovered and remains meaningfully short of the bar we require, the shares are priced above our estimate of what the business is worth, and the company reports second-quarter results in five trading days. We would rather miss a bounce than take a position on a half-recovered signal days before an event that can move the stock 5-10% either way. | OPERATIONAL: BUY cv59 -> adj 56 (val_adj -3 on GROWTH_PRICED_IN MoS -0.5502; scorecard_bias 0, BreakoutUp_20D|buy n=2 report-only). Nine points under the 65 floor. Note the geometry is sound - T4 593.44, T12 516.95 (+5.0% vs close 492.20), stop 447.63 = 2.83 ATR, measured R:R 2.00 - so this is a SCORE reject, NOT an HT-4 structural reject; that distinction matters if conviction recovers. Today it traded a 497.80 high vs the 496.28 buy-stop withdrawn yesterday at cv53, i.e. the withdrawn order would have triggered intraday - logged, but the withdrawal terms were conviction-based and are explicitly NOT loosened to capture a bounce. Q2 results 2026-08-12 (5 sessions), carried in decisions/hk_horizon.json. RE-ARM unchanged: fresh BUY/STRONG BUY at cv >= 65 with a definable stop, T12 above the action price, and the 08-12 earnings either past or explicitly sized for. |
| PORTFOLIO | hktech | — | WATCH | — | — | n/a | n/a | n/a | One buy, no sells. The book still owns nothing but cash, so there is nothing to sell. We are re-entering Lenovo - the position we should already have owned, and would have, but for an execution fault on our side that quietly dropped yesterday's order. The case has strengthened rather than faded in the meantime, so we are taking it again at today's price, at half the size the model suggests, because the shares have doubled in a month and report results in six trading days. Roughly 95% of the book stays in cash: eleven of the twenty-one names we follow are outright avoids with twelve-week projections 30-43% below today's prices. | OPERATIONAL: buy pool = 2 bullish-actionable rows (0992 STRONG BUY cv79 -> adj 77; 0700 BUY cv59 -> adj 56, 9 under floor); 11 AVOID, 4 WATCH, 4 NO TRADE. Sell pool empty by construction - zero holdings, no stop breaches, no Exit_Warning codes, no PROFIT_ARMED positions. ANTICIPATE slot empty (0 BUY - ANTICIPATE prints across the last 3 snapshots). Commits HKof HKdry; HK (94.72%) retained;. Anchor 2026-08-05 live settled HK close (L4). Fairvalue FetchedAt 2026-08-03 = 2d, FRESH. Scorecard: 63 entries, graded 7, calib_1w -0.45 -> f=0.30 mandatory (L5); all cells below n>=5 so scorecard_bias 0 everywhere (HT-2); counterfactual no_chase_queue FLIPPED to cost (n=3, median wh_1w +5.07%) - advisory only, but it is why today's card is a market order rather than another buy-stop trigger. Overlay 31d stale and inert. Horizon = hk-parent fallback, does not carry 0992's 08-13 earnings (systemic #3). PRIOR-SESSION EXECUTION FAILURE reconciled this run - see the 0992 reconcile line. Step 8a NOT run (batch mode). |
Outcome & track record
Accuracy at the time of this record.
7 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.
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.