Recommendations
0 to acquire · 1 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 |
|---|---|---|---|---|---|---|---|---|---|
| TSM | Taiwan Semiconductor (ADR) | 403.31 | SELL ALL | 403.31 | 375.56 | n/a | n/a | 341.79 (-15.3%) | Selling the whole TSMC position and taking the loss. The trading signal has said 'no trade, conviction zero' in five of the last six sessions, and the model's twelve-week outlook has been negative for eight sessions running - it currently points 15% below today's price. That is not a forward view a position earns its place with. Today's 7.6% jump was a sector-wide bounce on Microsoft's strong cloud results, not a change in TSMC's own setup, and the coverage of that bounce warns it can reverse just as fast. TSMC remains the best business in this book - record quarterly profit up 77%, capital spending raised to -64bn, revenue guided above +40% - and if the rally extends this will look like a poor exit. We are selling the signal, not the company, and the rally at least lets us leave at a materially better price than the level we first committed to. The cash stays earmarked for the next semiconductor setup that clears the bar, TSMC included. OPERATIONAL: completes the exit decided 07-27, re-decided 07-28 and 07-29, unfilled each time (executor stalled since 07-18); fills at ~403.31 vs the 07-29 reference 374.67, recovering of the delay. BASIS IS STATED PRECISELY: the 375.56 hard stop is RECLAIMED (+7.39%), so L2 does NOT fire today and is not claimed - the basis is the CLAUDE.md 9 sustained-NoTrade clause (cv0 on 07-23/24/27/28/30; the single interruption, ReversalUp_Oversold HOLD cv55 on 07-29, has already decayed back to cv0 exactly as the last card argued) plus the 8-session negative 12wk at -15.25%. Written re-entry trigger checked and only HALF met: EMA100 reclaimed (403.31 vs 398.64) but NO bullish-family print at all. Structure improved but still below EMA20 412.41 / EMA50 415.30 / VWAP 414.79. PROFIT_ARMED = 16th CONSECUTIVE FALSE ARM (Gain_ATR -1.74, T12_Progress 118.00 on a -7.92% loss): loss-control, not a profit-take, so the ADR-0012 never-100% cap does not apply. G2 suppresses valuation-amplify (broken DCF) -> val_adj 0, no tier bump, 9 amplify-only respected. G1/G3 do not block (no bullish signal, no positive forward view). Estimates: 1wk/4wk n/a (a NoTrade row prints no Target_Price so the f=0.30 interpolation has no anchor, L5); 12wk 341.79 (-15.25%), exceeds the DCF anchor. Action range 398.32-408.30 (close +- 0.25 ATR, ATR 19.95). Realized -; -book-to-date on TSM incl. the 07-18 leg. L3/USC-5 dispositive against re-litigating a thrice-committed exit on one sector-beta session (optional_trim counterfactual: median withheld 4wk -13.7%). Re-entry: bullish-family print at conviction >=65 (EMA100 condition already met). · news |
| AMD | Advanced Micro Devices | 485.39 | HOLD | 485.39 | 475.00 | n/a | n/a | 498.03 (+2.6%) | Holding AMD and cancelling yesterday's decision to sell it. That sale rested on one thing - the share price closing decisively below the 475 level we had committed to protect - and today's 13% rally has taken the price back above that line. The loss on the position has narrowed from 17% to 6%, the twelve-week outlook has turned positive again, and momentum has recovered from deeply oversold. With second-quarter results due on 4 August, our normal practice is to hold a position that is not broken through the announcement rather than sell into it, and we are deliberately leaving the stop where it is rather than tightening it into a binary event. If the shares close back below 475, the full sale goes ahead at the next opening with no further debate. OPERATIONAL: 07-29 full exit ( @ ~429.56, execute_on 07-30) WITHDRAWN - basis invalidated, not merely unfilled. Close 485.39 is +2.19% ABOVE the committed 475.00 trail; monitor reports ZERO breaches. Unreal -5.95% (was -16.76%); 12wk 498.03 = +2.60% (troughed 440.28); RSI 35..51; reclaimed EMA5 478.29, sits -0.27% under EMA50 486.69, well above EMA100 424.48 / EMA200 344.61. Label is NoTrade cv0 but it is the FIRST such print (07-27/28/29 were AVOID), so the 9 sustained-degradation clause that carries TSM does not carry AMD. Catalyst doctrine now aligns rather than being overruled: AMD Q2 08-04 AMC is 3 trading days out, inside the <=5-session hold window, and the 'decisive stop break' exception that overrode it on 07-28/29 has been reclaimed; no confirmed lower high. Stop deliberately NOT ratcheted (tightening into a binary print is a coin-flip stop-out). ESCALATION WRITTEN DOWN: a close back below 475.00 re-fires the full exit at the next open, no re-litigation. PROFIT_ARMED = 2nd FALSE ARM (Gain_ATR -0.67, T12_Progress 97.46 on a -5.95% loss). G2 suppresses valuation (broken DCF). Unrealized - (MV vs cost). 1wk/4wk n/a (NoTrade row prints no Target_Price); model band Est_Low 395.78 / Est_High 544.33. Honest counter-case recorded: a stand-down after a +13% day sits close to 'hoping', and USC-5 warns against carrying a loser through a print - the distinction applied is USC-5's own (broken technical -> de-risk; reclaimed technical with a positive forward view -> trail). · news |
| AMKR | Amkor Technology | 48.26 | WATCH | 48.26 | 33.60 | n/a | 62.49 (+29.5%) | 27.51 (-43.0%) | Passing on the only buy signal on the board. Amkor beat on its second-quarter results - earnings of 70 cents against the 48 cents expected, revenue up 26% - but guided the current quarter below what analysts wanted, and the shares fell 6.5% on the news and then 34% in four sessions before today's bounce. Buying a one-day rebound inside a fresh guidance disappointment is the trade this book has already learned not to take. The company also fails our quality screen outright - it has the weakest financial-strength score of the twenty names we follow - and the model's own twelve-week projection for the shares sits 43% BELOW today's price, which contradicts the buy label on the same row. Two brokers upgraded into the fall with targets and the packaging business has real AI exposure, so if this is the low we will have passed on a good entry. It is not one this framework can take. OPERATIONAL: rejected on five independent grounds - (1) adj conviction 60 is 5 pts under the 65 floor (cv70, val_adj -10 from MoS_FV -2.24 clamped from -11, bias 0); (2) HARD QUALITY GATE, Quality_Pass=False on a -0 name = new entry, F_Score 4 (lowest of 20), Composite_Rank 15.7, Quality_Rank 16, ROIC 7.7%, FCF yield 1.02% - USC-3: a gate override is never a licence to chase; (3) Target_12Week 27.51 vs a 48.26 close = -43.0% on a bullish BUY row, the same internal contradiction that got GLW rejected 07-28 and vindicated in one session (systemic finding #5); (4) L6 in the down direction - fresh post-earnings guidance gap-down (Q3 guided vs consensus, Communications declining, Android -20% on memory supply); (5) structure contradicts the label - below ALL six EMAs (5/10/20/50/100/200), VWAP distance -19.83%, Pillar_Consensus 'Consolidation_Bearish' ON A BUY ROW, pillars 36/37/35, and a risk-defined stop at 33.60 = -30.4% (2.08 ATR). Logged so the counterfactual grader prices the pass. Re-arm trigger: a PullbackBuy_VWAP print (L1) with the quality gate passing - not another oversold-reversal bounce. · news |
Outcome & track record
Accuracy at the time of this record.
12 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 →
-0.548 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.
24 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.