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 27th session - but this was the first genuinely constructive day since the crash. Both 2x Korean-memory trackers absorbed an early flush and closed at the session high: 7709 fell as much as 11% intraday to 28.04 before finishing at 31.06 (-1.4%) on the heaviest volume of the week, holding a higher low than the 30-July bottom, while 7747 added 2.6% to 63.72. 7709 also printed its first bullish-family signal since 20-July (ReversalUp_Oversold). The genuinely new fundamental item is regulatory: Korea's curbs on single-stock leveraged ETFs took effect 31-July - a tripled minimum cash deposit and a ban on new listings - and turnover in the flagship SK Hynix tracker has collapsed to a two-month low, meaning the forced retail-liquidation mechanism that drove the whole 30% drawdown is being administratively throttled. That improves the backdrop; it does not make a 2x tracker buyable. Nothing cleared a gate: the signal is a WATCH, not an actionable buy, at conviction 62 against a 65 floor, on a single print (the identical 20-July print reverted bearish the next day); the 12-week view still sits 43% below the close; and the model's own stop is 76.6% below the entry - no usable risk control at any conviction. 7747 moved the wrong way entirely, its conviction rising on a breakdown label, which is higher confidence in a down move, not a softening. Valuation is blind on both (NA_ETF). Holding cash into TSMC July revenue (~10-Aug), US CPI (12-Aug) and Applied Materials (13-Aug), all doubled through a 2x vehicle. Learnings applied: HL-2 (no one-bar oversold flip off a crash), HL-3 (no entry without a usable stop - the binding gate), HL-4 (rising conviction on a bearish label is a stronger avoid), L6/HL-6 (no pre-positioning into events). Scorecard 0 graded of 67, calib_1w null, f=0.30. Counterfactual optional_trim n=22 median -21.4% is MIS-SIGNED for a cash-only book - it measures declined buys, i.e. avoided-buy discipline that paid; advisory only. |
| 7709 | XL2CSOPHYNIX | 31.06 | WATCH | 31.06 | 7.26 | n/a | 51.83 (+66.9%) | 17.70 (-43.0%) | No position, none taken - but this 2x SK Hynix tracker had its best-looking session since the crash and now sits closest to a real entry. It gapped down 7% to 29.22, sold off to 28.04 (-11% intraday), then reversed to close at 31.06, the exact high of the day and 10.8% off the low, on 322 million shares - the heaviest volume of the week and a third more than yesterday's capitulation. Crucially the low held above the 30-July bottom of 25.36, so the base is intact, and daily volatility is compressing (ATR down from 43.5% to 37.5% of the price). The signal flipped from a breakdown label to ReversalUp_Oversold, the first bullish-family read since 20-July. Context helps: yesterday's 19% drop is now attributable to pre-payrolls de-risking across Korean chips (SK Hynix -9.7%, Kospi -4.5%), and that event has passed. Still not buyable, on four independent counts. The engine itself says 'monitor for entry trigger', not buy - WATCH is not an actionable label; conviction 62 is below the 65 floor; this is print one of a pattern that needs two, and the identical 20-July print reverted bearish the very next session; and the model's stop of 7.26 against a 31.06 entry is a 76.6% risk band, i.e. no risk control at all on an instrument that moves 19% in a day. The 51.83 'target' is simply the 20-day moving average restated, not a projection. The 12-week estimate of 17.70 remains 43% below the close, though that ratio has been mechanically pinned for 18 sessions and is treated as non-binding. Re-arm ladder: reclaim the 5-day average at 34.31 (now only 9.5% away, narrowed from 12.3%), then the 10-day at 38.89, on a second confirming bullish print, at conviction 65+, with a usable stop. HL-2/HL-3/HL-4 all still bind; HL-3 is the one that will block the next otherwise-qualifying entry. Events ahead: TSMC July revenue ~10-Aug, US CPI 12-Aug, Applied Materials 13-Aug - no pre-positioning (L6/HL-6). · news |
| 7747 | XL2CSOPSMSN | 63.72 | WATCH | 63.72 | 63.72 | n/a | n/a | 36.32 (-43.0%) | No position, none taken. This 2x Samsung Electronics tracker recovered 2.6% to 63.72 and, like its sibling, closed at the session high after dipping to 60.52 - a steadier session than yesterday's 11% fall, and it is holding 22.8% above the 30-July low. Structure improved at the margin: the price is now only 2.8% under its 5-day average, the closest since the crash, and the 200-day average overhang narrowed from 40.3% to 36.3%. But the signal went the wrong way and that decides it. The model still reads this as a failed-breakout-down, and its conviction in that bearish call jumped from - the highest of the week. On a leveraged breakdown name, rising conviction on a bearish label is a stronger avoid, not a softening one (HL-4): it is high confidence in a further down move, and the buy screen only ever considers bullish labels. The 12-week estimate of 36.32 sits 43% below the close, and there is still no usable stop - the model prints stop, entry and close all at 63.72 with a market order and a blank target, a zero-width risk band for the seventh consecutive session (HL-3). Participation was thin: volume up just 3.1% on a 2.6% up day, a drift rather than accumulation. Both direct catalysts (SK Hynix Q2 on 29-July, Samsung Q2 on 30-July) have already printed and the next ten sessions carry no 7747-specific event, so there is nothing scheduled to force the issue. Backdrop is slowly improving - Korea's leveraged-ETF curbs effective 31-July are throttling the forced-selling that drove the drawdown - but Morgan Stanley still calls that unwind only past its midpoint. Re-arm: reclaim the 5-day 65.58, then the 10-day 69.35, then the real test, a reclaim-and-hold of the 200-day at 86.86, on a bullish actionable label with a second confirming print and a usable stop. · 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.