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 |
|---|---|---|---|---|---|---|---|---|---|
| 0100 | MINIMAX-W | 216.00 | NO ACTION | 216.00 | 81.34 | 255.22 (+18.2%) | 346.73 (+60.5%) | 123.12 (-43.0%) | The book's top-conviction print (BUY cv70, Buy_Rank #1) and REJECTED for the third session running on the identical defect -- MiniMax is a falling knife the model cannot risk-define. (1) Un-sizable stop: Stop_Price 81.34 sits -62.3% below the 216.00 close (ATR 47.69 is ~22% of price, so the ATR-multiple stop lands two-thirds under the entry) -- the position cannot be risk-defined per CLAUDE.md section 9. (2) Negative 12-week: T12 123.12 = -43% vs close. (3) The tape confirms the knife: -15.62% today alone and -17.1% since the 07-15 reject at cv90 (260..00); RSI 30.1; close sits ~3% off the 52-week low 209.20 and below EVERY EMA (5/10/20/50/100 = 246/280/347/496/577). A deep-downtrend oversold bounce on a warrant, not buyable long-only. Est band 151.2-258.6. Ledgered direction=none/optional so the counterfactual grader prices what skipping cost. f=0.30 (L5); val_adj 0 (NA_CCY, book note B). |
| 0700 | TENCENT | 461.60 | NO ACTION | 454.16 | 427.78 | 461.41 (+1.6%) | 478.34 (+5.3%) | 443.27 (-2.4%) | The closest call of the run and still a withhold: Tencent prints the L1-favored PullbackBuy_VWAP BUY (cv65, Buy_Rank #2) with a genuinely clean entry -- 454.16 is only -1.61% below the close (ED -0.35), RSI 52.8 is neutral, price is just +2.83% above VWAP (not stretched), the stop 427.78 (-7.3%) is definable, and T4 478.34 is the only floor-clearing 4wk target above its own close (+5.32% vs entry). Rejected anyway because the risk math and the structure both fail: (a) Target_12Week 443.27 sits below both the close (-4.0%) and the entry (-2.4%) -- the model projects Tencent lower over the very horizon the trade is held; (b) R:R 0.92 -- risking 26.38 to make 24.18, negative-EV before any view; (c) structure is counter-trend, not pullback-in-uptrend: close 461.60 is below BOTH EMA100 483.34 and EMA200 510.25, Momentum=Bearish (24), Consensus=Mixed, -32% from the 683 52W high; (d) it is a fresh single-print flip (NO TRADE cv0 x2 -> BUY cv65) that appeared only because a -4.63% down day dragged price back to VWAP, landing exactly at the 65 floor with zero margin. The overlay's '0700 add when signal turns up' trigger did fire, but that override is 12 days stale and was authored for a HELD position -- this book holds zero 0700, and an add-override cannot originate a new entry. Taking 0700 at cv65 would also directly contradict yesterday's rejection of 9988 at cv73 on the identical structure; consistency governs. f=0.30 (L5); val_adj 0 (NA_CCY). |
| 9988 | BABA-SW | 112.60 | NO ACTION | 104.62 | 96.99 | 106.72 (+2.0%) | 111.61 (+6.7%) | 90.90 (-13.1%) | Rejected for the second consecutive session, now with decaying conviction (cv73 07-16 -> cv65 today) -- the 07-16 withhold is validated: BABA fell -3.7% (116..60) and the 104.11 limit never filled, so skipping cost nothing. The same structure stands: (1) the entry 104.62 is a deep sub-close pullback limit (-7.09%, ED -1.31) that fills only on a further ~7% drop; (2) Target_4w 111.61 sits BELOW the 112.60 close (-0.88%) and Target_12Week 90.90 is -19.3% vs close / -13.1% vs the entry -- both forward views point down; (3) R:R 0.92 (reward 6.99 vs risk 7.63); (4) counter-trend structure -- below EMA100 122.37 and EMA200 128.51, Momentum=Bearish, Consensus=Mixed. Note the model's internal contradiction: T12 90.90 sits far beneath its own Est_Low 103.13, the inverted-projection tell of book note B. A bear-rally pullback dressed as the L1 cell; withheld, not queued (a GTC whose 12wk target is 13% below its own entry is a negative-EV order). f=0.30 (L5); val_adj 0 (NA_CCY). |
| 1211 | BYD | 88.70 | NO ACTION | 82.03 | 76.67 | 83.50 (+1.8%) | 86.94 (+6.0%) | 76.22 (-7.1%) | Floor-clearing BUY cv65 (Buy_Rank #3) on the L1-favored cell, withheld on the same three counts as the rest of today's pullback pool -- BYD's own numbers argue against the trade. (1) The entry 82.03 is -7.52% below the 88.70 close (ED -1.56), filling only on a further -7.5% slide; (2) Target_4w 86.94 sits BELOW the close (-1.98%) and Target_12Week 76.22 is -14.1% vs close / -7.1% vs entry -- the model projects it lower; (3) R:R 0.92 (reward 4.91 vs risk 5.36). Structure is counter-trend, not a pullback-in-uptrend: close is below EMA100 92.30 and EMA200 97.29 with Momentum=Bearish (24) and Consensus=Mixed. Sharpest tell: the engine's entry limit 82.03 sits BELOW its own Est_Low 84.77 -- it is asking to buy beneath the bottom of its own estimate band. The cv65 print appeared only because a -2.47% down day pulled price back toward VWAP. f=0.30 (L5); val_adj 0 (NA_CCY). |
| 1810 | XIAOMI CORP-W | 26.88 | NO ACTION | 24.67 | 22.77 | 25.19 (+2.1%) | 26.40 (+7.0%) | 21.29 (-13.7%) | Floor-clearing BUY cv65 (Buy_Rank #4) and the worst forward view of today's pullback pool -- withheld. (1) The entry 24.67 is -8.22% below the 26.88 close (ED -1.46), filling only on a further -8% drop; (2) Target_4w 26.40 sits BELOW the close (-1.79%) and Target_12Week 21.29 is -20.8% vs close / -13.7% vs entry -- the deepest negative 12wk of the four pullback names; (3) R:R 0.91, the worst of the group. Structure: close is below EMA100 29.78 and EMA200 34.49, Momentum=Bearish (24), Consensus=Mixed, and Xiaomi sits -55% from its 59.90 52W high -- a bear-market rally, not a pullback in an uptrend. Same model contradiction as BYD: the entry 24.67 is below its own Est_Low 24.73. f=0.30 (L5); val_adj 0 (NA_CCY). |
| — | hktech book | — | NO ACTION | — | — | n/a | n/a | n/a | NO ACTION (0 ACQUIRE / 0 DISPOSE) -- FIFTH consecutive session in 100% cash, through a broad HK-tech risk-off day (0100 -15.6%, 6869 -16.6%, 1347 -11.9%, 0700 -4.6%, 9988 -3.7%). The buy pool WIDENED to five floor-clearing names (0100 cv70; 0700/1211/1810/9988 all cv65) but did not IMPROVE: the engine flipped four names from NO TRADE cv0 -> BUY cv65 purely because the selloff dragged price back toward VWAP, and all four share one defect set -- (a) counter-trend structure: every one closes below BOTH EMA100 and EMA200 with Momentum=Bearish and Consensus=Mixed (bear-rally pullbacks inside long-term downtrends, book note B, not the pullback-in-uptrend the L1 cell should capture); (b) a negative model 12-week view on every name (0700 -4.0%, 1211 -14.1%, 9988 -19.3%, 1810 -20.8% vs close); (c) structurally sub-1 R:R on every row (0.91-0.92), because the tuned HK exits set the stop at 1.25 ATR below entry against a 4wk target only 1.145 ATR above -- a config artifact that makes the L1-favored PullbackBuy_VWAP cell mathematically unbuyable in this book (systemic note 1, the run's key finding). 0100 is a third-strike falling knife (stop -62.3% of close = un-sizable, T12 -43%, RSI 30.1, ~3% off its 52W low). Three of the five (1211/1810/9988) print an entry limit BELOW their own Est_Low. The one structurally sound long -- 0992 Lenovo, the only bullish name above EMA50/100/200 with Consolidation_Bullish consensus and the only positive forward view in the book (T4 +17.6%, T12 +41.4%) -- is blocked at cv55, 10 points sub-floor; re-arm on a fresh PullbackBuy_VWAP BUY reclaiming cv>=65. Sell pool empty (zero holdings). ANTICIPATE slot empty (no BUY - ANTICIPATE prints). Every prior withhold is now validated by tape: 7709 -31.3%, 0100 -17.1%, 9988 -3.7% since their rejects. HKdry fully retained. Valuation technical-only (NA_CCY, quality gate off, val_adj 0). Overlay stale (12d); the 0700 add-override fired its trigger but cannot originate an entry on an unheld name. Horizon: SK Hynix Q2 07-22 reads through to 0981/1347/7709, all sub-floor -- wait for the print. f=0.30 (calib_1w -0.45). |
Outcome & track record
Accuracy at the time of this record.
3 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.