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 |
|---|---|---|---|---|---|---|---|---|---|
| — | AI - HK Leveraged ETFs | — | NO ACTION | — | — | n/a | n/a | n/a | Stay 100% in cash (HK) for a 35th session. Today delivered one of the most violent two-way sessions this book has recorded and it still produced nothing worth buying. Both leveraged trackers gapped sharply lower at the open, sold off hard into the morning, and then rallied all session to close near the high of the day on record volume: the 2x SK Hynix ETF (7709) opened 11.9% down, fell as much as 19.6%, and finished at 36.46 - only 1.2% lower on the day and 22.9% above its own low, on 352 million shares, the heaviest volume of the window and a record that was one session old. The 2x Samsung ETF (7747) opened 10.6% down, fell 15.4%, and closed at 75.34, still 6.5% lower but 10.6% off its low, also on record volume. A close that finishes near the top of a 19%-wide range looks like capitulation and a bottom, and that is precisely the pattern this book has learned to decline: the model printed no bullish signal on either name - both are still tagged a failed breakdown - and its confidence fell on both rather than rising. Its twelve-week view still sits about 43% below today's price on both listings, and neither offers a usable stop-loss level, so there is no way to take the trade with defined risk even if one wanted to. The two names also split decisively for the first time: 7709 rebuilt its short-term moving averages while 7747 lost its 5-, 10- and 20-day lines in a single session and now sits 12.5% below its 200-day line, having given back 60% of the entire four-day rally. Volatility expanded again on both to the widest readings of the window - a direct and compounding cost on daily-reset 2x vehicles. Cash stays fully intact ahead of the Korea chip-export print on 21 Aug and the NVIDIA results reaction on 27 Aug, with no pre-positioning into either. OPERATIONAL: buy pool empty for a 12th session; HL-2 vetoes the reversal bar (no bullish label, conviction down on both, T12 ~43% below close, zero 2nd print, 0-for-3 on oversold flips holding a 2nd session); HL-3 binds (Stop==Entry==Close, Target_Price blank, Order_Type Market - 15th consecutive zero-width band on 7747, 6th on 7709); HL-4 - the acquire floor screens bullish actionable labels only, so neither name is eligible before conviction is read. Finding 13's 5th boundary crossing (7709, bracket [SELL]->[HOLD], 2nd downward). NEW finding 16: 7747's EMA5 and EMA20 both ROSE on a -6.5% close that finished below them - arithmetically impossible, likely a restated price history between the 08-18 and 08-19 fetches; invalidates yesterday's 'reclaimed EMA200' narrative, does not change the verdict. Finding 15 superseded - the momentum pillar finally diverged (41 vs 24) after six identical sessions. News NOT-FOUND (all retrieved Seoul articles dated 2026-07-29 / 2026-06-05; the Hang Seng quote returned the identical figure logged for 08-18, i.e. stale cache, so no same-day divergence cross-check today). Counterfactual optional_trim n=22, median -21.4% 1wk / -41.5% 4wk - MIS-SIGNED for a cash-only book, read as avoided-buy discipline that paid; advisory only. f=0.30 (calib_1w null, L5/HL-6). |
| 7747 | XL2CSOPSMSN | 75.34 | WATCH | 75.34 | 75.34 | n/a | n/a | 42.94 (-43.0%) | No position, and none taken - this is now the broken half of the pair. It opened at 72.00, down 10.6%, fell to 68.14, and closed at 75.34, recovering to finish 10.6% above its low but still 6.5% lower on the session, on 17.1 million shares - a second consecutive record volume day. The recovery off the low is not the story; the damage to the structure is. Every moving average this tracker spent the four-day rally reclaiming is gone in one session: it closed below its 5-day line at 79.08, below its 10-day at 76.27 and below its 20-day at 79.63, having held all three yesterday, and it now sits 12.5% below the 200-day line at 86.14 - two sessions after briefly closing above it for the first time in this book's history. It has given back 60% of the entire advance off the 11 Aug low. The model's confidence dropped again to 42, the lowest reading either name has printed in this window, with its trend score at a window low of 30 and its momentum score collapsing to 24. Volatility widened from 11.2% to 14.4% of price, the widest of the window - on a fund that resets its 2x exposure daily, that is a compounding drag that erodes returns even in a flat market. And despite falling 6.5%, it is still trading 9.7% above its volume-weighted average price, so it is not yet even neutral, let alone cheap. The twelve-week view remains around 42.94, roughly 43% below today's close. Korea's 1-20 August chip export figures land on 21 Aug and NVIDIA reports on 26 Aug with the Hong Kong reaction on the 27th - both would move this listing at double the underlying's rate, and neither is a reason to be early. OPERATIONAL: 15th consecutive session of FailedBreakout_Down_20D, bracket [HOLD]. HL-3 binds for a 15th session - Stop_Price == Entry_Price == Close 75.34, Target_Price blank, Order_Type Market, Exit_Strategy immediate: no risk-defined entry is constructible. HL-2 re-arm bar - (1) reclaim EMA5 79.08 LOST -4.73%; (2) reclaim EMA10 76.27 LOST -1.22%; (3) reclaim EMA20 79.63 LOST -5.38%; (3b) reclaim EMA200 86.14 LOST -12.53%; (4) bullish actionable label + 2nd confirming print NOT MET; (5) usable stop NOT MET. Target_12Week 42.94 pinned at ratio 0.5700 for a 28th session (finding 4). NEW finding 16 applies to THIS ticker: EMA5 77.82->79.08 and EMA20 79.18->79.63 both ROSE on a close of 75.34 that sits below both - impossible for a standard EMA; likely restated price history between fetches, which retroactively invalidates the 08-18 'first close above EMA200' claim. Treat this name's cross-session MA narrative as unverified until re-fetched. Momentum pillar diverged from 7709 for the first time in six sessions (24 vs 41), superseding finding 15. Declined-print marks deepened: 07-14 @95.30 now -20.9% (was -15.5%), 07-27 @83.42 now -9.7% (was -3.4%). Still no name-specific event before 08-21. No gap-chase (L6/HL-6). · news |
| 7709 | XL2CSOPHYNIX | 36.46 | WATCH | 36.46 | 36.46 | n/a | n/a | 20.78 (-43.0%) | No position, and none taken - but this is the half of the pair that held, and it is worth recording why that still is not enough. It opened at 32.50, down 11.9%, fell all the way to 29.66 - down 19.6% from yesterday's close - and then rallied for the rest of the Hong Kong session to finish at 36.46, down just 1.2% on the day. That is a 22.9% recovery off the low, closing within 4% of the day's high, on 352 million shares: the heaviest volume of the window by a fifth, taking out a record set only one session earlier. Two of yesterday's three broken items were repaired: the 5-day average crossed back above the 10-day, and the close reclaimed both. On any conventional reading this is what a capitulation low looks like. The model does not agree, and the model's objections are the ones that carry: it still tags the listing a failed breakdown rather than any bullish setup, its confidence fell from, and its twelve-week projection sits at 20.78 against a close of 36.46 - about 43% lower. There is also still no usable stop: the model's entry, stop and last price are the same number, so there is no defined risk to take. This ticker has now attempted an oversold bounce three times in five weeks and has failed to hold the gain into a second session every time. Volatility widened again to 19.6% of price, the widest of the window - on a fund that resets 2x exposure daily, that decay compounds against a holder. The listing is also the most event-exposed name in the book, with Fed minutes tonight, Korean chip export data on 21 Aug, and NVIDIA's results reaching Hong Kong on 27 Aug - all amplified twofold. Waiting for a second confirming session costs nothing here; being early has cost 30-44% on three of the four prior attempts. OPERATIONAL: 6th consecutive session of FailedBreakout_Down_20D, bracket flipped [SELL] -> [HOLD] as conviction fell back through the 65 line - finding 13's 5th boundary crossing, 2nd downward. Under HL-4 that is weaker DOWN-confidence, mildly constructive, never a buy trigger on its own; the label did not move. HL-3 binds for a 6th session - Stop == Entry == Close 36.46, Target_Price blank, Order_Type Market. HL-2 re-arm bar - (1) reclaim EMA5 36.17 MET, close +0.80%; (2) reclaim EMA10 36.09 MET, +1.02%, and the EMA5/EMA10 crossover flipped BACK bullish (36.17 vs 36.09); (3) EMA20 43.09 NOT MET, gap -15.39%, roughly a full ATR-move away; (4) bullish label + 2nd confirming print NOT MET - count is zero and this ticker is 0-for-3 on oversold flips holding a second session (07-20, 08-07, 08-17); (5) usable stop NOT MET. Target_12Week 20.78 pinned at ratio 0.5700 (finding 4) - the re-arm gate of T12 >= close fails by 43%. VWAP discount narrowed -5.34% -> -4.19%, still below. EMA200 blank all window - the key structural gate remains UNEVALUABLE here (finding 6). Momentum pillar 41 vs 7747's 24 - first divergence in six sessions, superseding finding 15. Declined-print marks: 07-14 @65.20 -44.1%, 07-20 @52.58 -30.7%, 07-27 @54.50 -33.1%, 08-07 @31.06 +17.4%. Events - densest in the book: FOMC minutes 08-19 (HK 08-20, +/-2.5%), Korea 1-20 Aug chip exports 08-21 (+/-6%), US flash PMIs 08-21 (+/-1.5%), options expiry 08-21 (+/-2%), NVIDIA Q2 08-26 (HK 08-27, +/-14%), US July PCE 08-26 (HK 08-27, +/-3%), Jackson Hole 08-28 (+/-3%) - all bands doubled, no pre-positioning (L6/HL-6). · news |
Outcome & track record
Accuracy at the time of this record.
0 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.
0.000 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.