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.

US End-of-Day Verdict 31 Jul 2026, 16:00:00 GMT-4

US Semiconductors — End-of-Day Verdict

EOD 2026-08-03

Recommendations

0 to acquire · 2 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
TSMTaiwan Semiconductor (ADR)406.11SELL ALL406.11375.56n/an/a353.23 (-13.0%)Selling the whole Taiwan Semiconductor position. This is a call on the share price trend, not on the company - the business just posted record quarterly results, guided the current quarter up another 12%, and reclaimed a trillion market value. But the stock has given our model no tradeable signal in seven of the last eight sessions, its twelve-week outlook has been negative for ten sessions running, and it still trades below its 20-day, 50-day and volume-weighted average prices. We bought it as a trend and the trend is gone; we would rather hold the cash and buy it back when the chart agrees with the business again. Realising roughly a 7.3% loss. FULL BASIS: CLAUDE.md sec9 sustained-NoTrade clause - NoTrade cv0 with 'Consensus conflict + low conviction' in 7 of the last 8 sessions (07-23/24/27/28/30/31/08-03), Target_12Week 353.23 = -13.02% vs the 406.11 close and negative for TEN consecutive sessions, close under EMA20 411.11 / EMA50 414.52 / VWAP 410.84. The 375.56 stop is RECLAIMED by +8.13% so L2 is NOT claimed. Sixth re-issue of the exit first decided 07-27. HELD OVER AN EXPLICIT STRIKE: NoTrade|sell became ACTIONABLE this run at bias -3 (n=6, hit_1w 0.333) - three of five gradeable TSM exits graded good_sell:false (07-17 missed_up 3.63%, 07-27 1.29%, 07-28 3.52%) - and three structural counts improved (three consecutive higher closes 403.31->404.25->406.11, above EMA5 401.48 and EMA10 403.80, EMA100 398.91 reclaimed). Overridden because -3 is a haircut not a veto, 2-of-6 at n=6 is thin against 7-of-8 cv0 with a 12wk negative for ten, the intermediate structure is still broken, and the written re-entry trigger is only HALF met (EMA100 reclaimed but NO bullish-family print at all). IF THIS GRADES POORLY ONCE MORE THE SEC9 SUSTAINED-NOTRADE BASIS ITSELF GOES TO THE LEARNINGS REVIEW (systemic finding #1). G2 suppresses the valuation-amplify (DCF vs a price is a broken artifact) -> val_adj 0, no tier bump, sec9 amplify-only respected. Gives up a known upside catalyst: TSMC July monthly revenue lands ~08-10, five sessions out - stated, not buried. USC-5 is the lesson about NOT holding a broken technical for a catalyst. RE-ENTRY TRIGGER: a bullish-family print (PullbackBuy_VWAP preferred per L1) with conviction >=65 AND a non-negative Target_12Week. L8: executes at the 2026-08-04 open, not tonight; reprice if the open gaps >~1.5%. · news
AMDAdvanced Micro Devices484.64SELL 30%484.64475.00n/an/a489.72 (+1.0%)Cutting our AMD holding by 30% and keeping the rest. AMD reports quarterly results tonight and the options market is pricing an 8.5% move in either direction - a genuine coin flip on a position that is already about 6% underwater and has stopped generating any signal from our model for three sessions running. We are not calling the result; we are sizing for the fact that we cannot. Taking roughly a third off means a bad number costs us meaningfully less while a good one still pays on the two-thirds we keep, behind a stop at The longer-term picture is intact - the shares remain far above their 100- and 200-day averages - which is why this is a trim and not an exit. FULL BASIS: ADR-0012 deterioration ticks (d) 3rd consecutive NO TRADE cv0 and (c) dead forward view, Target_12Week 489.72 < 1.02 x 484.64 = 494.33 (+1.05% twelve-week view on a -6.09% position). Structure broken: under EMA10 489.79 / EMA20 502.12 / EMA50 486.21, VWAP -4.79%; intact long-term above EMA100 426.66 / EMA200 347.18 - hence trim, not exit. G1 does NOT protect (no fresh BreakoutUp/MomentumContinuation at cv>=80 - it is cv0); G3 SATISFIED (forward view flat +1.05%, signal not bullish). The catalyst-aware carve-out - a CONFIRMED LOWER HIGH, met 07-31 and not undone (08-03 high 490.95 < 07-31 high 515.62) - remains the licence to trim inside the <=5-session earnings window. ONE TIER ONLY, unchanged from 07-31: the 08-03 tape IMPROVED (open 461.79, low 455.30, close 484.64 near the high, +1.78%), so there is no case to escalate, and G2 suppresses the MoS_FV -15.00 valuation-amplify that would otherwise bump the tier (DCF vs a price is a broken artifact) -> val_adj 0. PROFIT_ARMED IS A FALSE ARM for a 4th straight session (Gain_ATR -0.74 on a -6.09% position; the arm fires off the T12-progress leg at 98.96% on a target BELOW cost) - the profit-take path is NOT the basis of this card; this is a sec9 loss-control de-risk. Backlog item (a) unfixed. STOP 475.00 HELD AND DELIBERATELY NOT RATCHETED: close is only +2.03% above it and an 8.5% implied gap goes straight through a 2% stop, so tightening buys nothing and invites a whipsaw at the open. COMMITTED ESCALATION STANDS UNCHANGED for the retained : a decisive close below 475.00 re-fires the full exit, no re-litigation. Dispersion arithmetic: 8.5% implied = +/-on, +/-on 120 - the trim removes of binary risk. Scorecard bias -3 (NoTrade|sell actionable, n=6) noted and argues against the sell; overridden by the two live ADR-0012 ticks plus tonight's binary. USC-5: de-risk to the technical's own tier and let a DEFINED core ride the catalyst. L8: executes at the 2026-08-04 open - PRE-PRINT, since earnings are after that session's close; reprice if the open gaps >~1.5%. · news
QCOMQualcomm Incorporated151.57WATCH151.57133.29157.01 (+3.6%)169.69 (+12.0%)86.39 (-43.0%)Passing on Qualcomm again, and this time it was close. Our model upgraded it to an outright buy today - the first buy signal anywhere on this board in eight weeks - and the entry is not stretched. We are still not taking it. Our quality screen on the company's own financials is failing, which blocks new positions outright; the model's own twelve-week price projection sits far below today's price, contradicting the buy signal beside it; and the reason the shares are down 44% from their high is not a passing dip - management has said Apple is winding down its modem business faster than expected, next quarter's profit guidance came in about 10% below what analysts wanted, and cost inflation across chip manufacturing is squeezing margins industry-wide. Wells Fargo cut its price target from to A four-day-old guidance cut is not a base. We want to see the quality screen pass and a proper pullback entry before we pay for this bounce. REJECT BASIS (four independent counts; note the FLOOR IS CLEARED - conviction 70, val_adj -4 on MoS_FV -0.80, bias 0 because OversoldBounce_Up|buy has NO cell in this book, adj 66 vs a 65 floor, Entry_Distance_ATR -0.0 so the no-chase rule is NOT what kills it, RSI 32.8 nowhere near overbought, Buy_Rank 1): (1) HARD QUALITY GATE - Quality_Pass False (flipped True->False on the 08-01 regeneration, unchanged since); the gate is an outright block on NEW entries and QCOM is. USC-3 is explicit that a gate override is never a licence to chase. (2) Target_12Week 86.39 = -43.0% on a BUY-labelled row - taking the row's bullish half while ignoring its own forward half is not a selective read this book gets to make (systemic finding #2). (3) NEWS OVERRIDES AND IT IS STRUCTURAL: -44% from the high, Apple modem revenue declining FASTER than previously forecast, Q4 guided -2.25 vs consensus on revenue -10.5B vs expected, pressure attributed to industry-wide input inflation across wafers, packaging and memory - not a demand air-pocket that mean-reverts; Wells Fargo PT -> Equal Weight with the sell side broadly following. (4) THE TAPE UNDER THE LABEL IS BEARISH: Pillar_Consensus Consolidation_Bearish, Trend and Momentum both Neutral_Bearish_Context, Volatility Bearish, close below ALL SIX EMAs (5/10/20/50/100/200), VWAP distance -10.66%; the 08-03 +2.68% print is the first up-close in six sessions - a bounce inside a downtrend, which is what OversoldBounce_Up is built to catch and what L6 in the down direction says not to pay for on day one. RE-ARM TRIGGER: a PullbackBuy_VWAP reprint (L1) OR a BUY at conviction >=65 WITH Quality_Pass restored to True, and given count (2) a Target_12Week that is not deeply negative. Two of four counts are single-refresh flips, so this name stays ON the board rather than off it. Logged direction:none so the counterfactual grader prices the pass. · news

Outcome & track record

Accuracy at the time of this record.

15 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.
53 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.