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 7 Sept 2026, 16:00:00 GMT+8

HK Leveraged ETFs — End-of-Day Verdict

EOD 2026-09-07 (hklev)

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
7709XL2CSOPHYNIX41.06WATCH41.0641.06n/an/a23.40 (-43.0%)The 2x SK Hynix ETF exploded 16.1% higher to 41.06 - its biggest single day ever - reclaiming its 5-, 10- and 20-day averages for the first time in weeks, closing in the top fifth of its range on the heaviest volume of the window. The move is genuine and explained: Korean chipmakers rallied hard on 07-Sep with SK Hynix itself up 8.1% and the Kospi up 4.6% on renewed AI-memory demand, and the ETF tracked that exactly. We still do not buy it, for a reason that has nothing to do with direction: the engine rates it AVOID on a breakdown signal, and - more decisively - the model prints its stop-loss at exactly the entry price with no profit target, so there is no way to build a risk-defined position in an instrument that swings 10% a day. We would be buying with no floor under us. The name also remains 21% below its 200-day trend line, so this is a sharp rally inside a downtrend, not a repaired uptrend. We stay in cash and wait for a usable stop and a bullish label. Operational: NINTH consecutive AVOID; AVOID is a disqualification label, never enters the acquire pool at any conviction (Step 5); so no sell can originate. Adj conviction 63 (flat d/d) is 2 pts below the 65 floor - but the floor is moot while the label is AVOID. HL-3 is the BINDING veto: Stop == Entry == Close 41.06, blank Target_Price, ATR/price 9.9%. HL-4: conviction on a breakdown-family label reads as confidence in a DOWN move. HL-2 re-arm: 2nd-strong-close leg NOW MET, bullish-label leg unmet, T12 leg inert (T12/Close clamp pinned 0.56990 for a 5th session - Systemic finding 1). Pillars trend 48 flat / momentum / volatility netted to zero conviction change on a +16% day (Systemic finding 3). Re-arm trigger: bullish label + adj conv >=65 + a REAL ATR stop below entry + a 2nd confirming bullish print; watch EMA20 38.50 as the give-back level and EMA200 52.25 (+27%) as the trend objective. Horizon: 7709 is tagged in SIX events in the next 10 sessions - 09-10 TSMC revenue and 09-11 Oracle + Korea 1-10 Sep semi exports (both add-after-print, Korea band anchored on 7709), then 09-14 CPI, a coin-flip FOMC 09-16/17 and a live BoJ hike 09-18 where the horizon names 7709 the book's top carry-unwind casualty. Stage any future re-arm AFTER the 09-11 prints. val_adj 0 (NA_ETF, valuation blind by construction); scorecard_bias 0 (0 graded, ever); f=0.30 per L5/HL-6. · news
7747XL2CSOPSMSN77.88WATCH77.8877.88n/an/a44.39 (-43.0%)The 2x Samsung ETF rose another 10.5% to 77.88, a two-day gain of 16.6%, and this is the strongest buy case this book has ever declined: its conviction score jumped to 76, clearing the 65 threshold by a wide margin, and its trend reading actually improved for the first time in the window - reversing the exact deterioration we cited against it on Friday. It reclaimed all three short-term averages, closed near the day's high on double the prior volume, and the rally has a verified cause: Samsung Electronics rose 5.7% in Seoul on 07-Sep as Korean AI-memory names surged. We are still not buying, and the reason is mechanical rather than a disagreement with the tape: the engine labels it AVOID on a breakdown signal, and it prints a stop-loss identical to the entry price with no target - there is no risk-defined trade to construct. On a leveraged fund that moves 8% a day, entering with no defined loss floor is the one thing this mandate does not do. It also sits 8% below its 200-day line, which is now within reach and is the level that would confirm a genuine trend repair. We stay in cash. Operational: NINTH consecutive AVOID. Adj conviction 76 CLEARS the floor by 11 pts - the first clean clear with a repairing trend pillar rather than the 07-17 cv88 STRONG SELL configuration - but HL-2 says the floor is necessary-not-sufficient and Step 5 keeps AVOID out of the buy pool at any score; HL-4 reads high conviction on a breakdown family as stronger confidence in DOWN. HL-3 is the binding veto: Stop == Entry == Close 77.88, blank Target_Price, ATR/price 8.3% - the veto would fire identically on a bullish label, which is the point (Systemic finding 2). so no sell origination. T12 44.39 is 43% below close with the T12/Close clamp pinned at 0.56998 - inert, not informative. Re-arm trigger: bullish label + adj conv >=65 (already met) + a REAL ATR stop below entry + a 2nd confirming bullish print; EMA20 75.12 is the give-back level, EMA200 84.65 (+8.7%) the trend objective. Not tagged in the horizon (parent hk does not hold it) but carries the same Korean-memory and 09-18 BoJ yen-carry read-through as 7709. val_adj 0 (NA_ETF); scorecard_bias 0; f=0.30 per L5/HL-6. · news
AI - HK Leveraged ETFsNO ACTIONn/an/an/aNo trades today, but this was the closest call in the book's history. Both instruments - the 2x SK Hynix and 2x Samsung ETFs - surged again on a broad Korean AI-memory rally, up 16.1% and 10.5%, and every objection we raised on Friday has now inverted: both closed near their highs instead of their lows, on expanding rather than falling volume, reclaiming their short-term averages, with a real dated catalyst behind the move. One of them cleared our conviction threshold outright. We still bought nothing, because the model is publishing a stop-loss identical to the entry price on both names, leaving no way to define the downside on instruments that move 8-10% a day. That is a plumbing failure in our own tooling, not a view on the sector, and we are recording it as the top fix for this book. The cost was real: staying in cash forwent roughly HKtoday and about HKacross the two-day bounce, and it has erased the HKof avoided drawdown we booked on Friday. The book remains 100% in cash at HKOperational: 0 ACQUIRE + 0 DISPOSE, ninth consecutive zero-action session. No open GTC queues to reprice or withdraw; no stop breaches; no trades detected; no unledgered drift; the two monitor warnings are the expected all-cash CASH row artefacts. Valuation permanently BLIND (both rows NA_ETF; regeneration SKIPPED per Step 3, CSV 64 days old) -> val_adj 0, quality gate off, size modifier 1.0x. Scorecard 121 entries / 0 graded / calib_1w None -> f=0.30 mandatory (L5/HL-6); no cell can reach the actionability floor because the book has never taken a trade - and per Systemic finding 4 the identified reason is that no row has ever printed a usable stop, so ADR-0003 tuning should be sequenced AFTER the exits fix. counterfactuals.optional_trim n=22 unchanged and still MIS-SIGNED for a cash-only book (HL-advisory), and note its sample is entirely July crash-leg vintage - today is the first evidence the same discipline can forgo a large gain. Overlay STALE (2026-07-05, 64 days; its 7709 'below 101 -> full exit' line is 59% out of the money and unheld). Event Horizon FRESH (hk parent fallback, regenerated today): 7709 tagged in SIX of the window's events - 09-10 TSMC, 09-11 Oracle + Korea semi exports (add-after-print), 09-14 US CPI, 09-16/17 coin-flip FOMC (~49.4% odds of a 25bp HIKE), 09-18 live BoJ hike naming 7709 the book's top carry-unwind casualty. Step 8a learnings review NOT run (skipped by instruction).

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.

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.