Recommendations
0 to acquire · 3 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 |
|---|---|---|---|---|---|---|---|---|---|
| AXP | American Express Co | 338.52 | SELL 50% | 338.52 | 334.75 | 340.02 (+0.4%) | n/a | 331.48 (-2.1%) | Amex's push through its 20-day high has failed and rolled over into distribution: the model now prints AVOID on a FailedBreakout_Down_20D with conviction rising to 51, and its 12-week view of sits 2.1% BELOW the close. This is the book's single largest position (11.7% of NAV) and its biggest dollar loser at -3.8% from cost, so half the line comes off while the other half stays to keep the franchise exposure. The fundamentals genuinely dissent - Q2 EPS vs, FY26 revenue growth guidance raised to 10%, a 16% dividend increase, and one published screen calls the shares 18% undervalued - but the stock is down 12.5% year-to-date and fell 4.3% on those very results. When the tape and the forward model both point down against a good story, we cut half and keep half. OPS: 3rd issuance - the 08-14 and 08-17 cards were never eligible (bad `execution` value 08-14, then blocked by the executor's CASH drift refusal on 08-17/08-18). Repriced from to the live anchor Entry_Price (+0.69%). Valuation MoS -0.46 is nowhere near the -3 tier-bump line, so 50% stands unmodified - CLAUDE.md sec.9: valuation originated nothing here. Retained keep the stop (1.11% of cushion); the scan's is a short-side stop on an AVOID row, rejected. 1wk from the model band midpoint (no Target_Price on an AVOID row), f=0.30 unused. · news |
| CB | Chubb Ltd | 345.29 | SELL 30% | 345.29 | 341.96 | 345.49 (+0.1%) | n/a | 338.28 (-2.0%) | Chubb has flipped from an unreadable NO TRADE to an outright bearish print - AVOID on FailedBreakout_Down_20D at conviction 57 - with a 12-week view of, 2.0% below the close. A failed breakout plus a negative forward view is a signal-based exit, so we take 30% off and keep 70% of a franchise we still rate fairly valued. Nothing in the business argues for selling: Q2 net income was 85bn on a beat, net premiums written grew 3.6%, a fresh $5bn buyback is authorised, the dividend goes ex on 11-Sep, and analysts peg fair value near against the last close. That is exactly why this is a 30% risk trim and not an exit. OPS: 3rd issuance (08-14 bad `execution` value, 08-17 blocked by the executor drift refusal). Repriced -> (+1.30%). BASIS CHANGED HONESTLY: the 08-17 card fired on L2 (close through the stop + negative 12wk); that stop break has since RESOLVED - the close is now 0.97% above the stop - so the card now stands on signal degradation alone (breakdown-family AVOID cv57 + negative 12wk), which is an independent sec.9 origination. Valuation UNDERVALUED (MoS +0.0003) argues the tier DOWN one under sec.9(b); 30% is already the floor tier so it is held there and explicitly NOT escalated. FailedBreakout_Down_20D|sell is ACTIONABLE at n=41, hit_4w 0.60, bias -1. Retained keep the stop; a decisive close back through it takes another 30%. No DCF_per_share (residual-income method for a financial) so no exceeds-DCF flag is possible. 1wk = model band midpoint. · news |
| AAPL | Apple Inc | 310.03 | SELL 30% | 310.03 | 300.00 | 310.03 (+0.0%) | n/a | 299.02 (-3.6%) | Apple has been unreadable to the model for five straight sessions - NO TRADE every time, reason logged verbatim as 'Consensus conflict + low conviction' - while its 12-week view of sits 3.6% below the close and our intrinsic work marks it richly priced (GROWTH_PRICED_IN, DCF vs a tape; an independent screen puts it ~10% above fair value). We take 30% off a 5.4% NAV position that the engine cannot form a view on, and deliberately leave 70% in for the bull case, which is real: fiscal Q3 revenue rose 16% to 4bn with gross margin at 50.1% and EPS up 29%, Rothschild & Co Redburn carries a target on an Nvidia-models AI thesis, the Houston advanced-manufacturing centre just opened, and a foldable iPhone is expected in September. This is a partial de-risk into strength, not a call against the company. OPS: 3rd issuance (08-14 bad `execution` value, 08-17 blocked by the drift refusal). Repriced -> (+1.45%, inside the 1.5% gap threshold but the card takes the live reference on re-issue). Held at 30% and NOT re-escalated: the 08-17 card already demoted 50%->30% for the Redburn upgrade and demoting twice on the same catalyst thread would double-count. TENSION STATED: `NoTrade|sell` is ACTIONABLE at n=34 with hit_4w 0.20 - four of five NoTrade-driven sells in this book have been wrong at four weeks - and the position has just gone positive vs cost (+0.45%). Against that: the `optional_trim` counterfactual (n=90, median +2.81% cost at 4wk) says this book's structural error is deferring trims, and L3 says execute the de-risk rather than re-argue it a fourth time. 30% splits the difference. Retained keep the stop. 12wk > DCF - projection assumes growth beyond priced-in. 1wk = model band midpoint (no Target_Price on a NoTrade row). · news |
| NUE | Nucor Corp | 264.19 | HOLD | 263.55 | 259.99 | 267.78 (+1.6%) | 280.98 (+6.6%) | 322.20 (+22.3%) | Nucor is the only bullish actionable name in the book tonight and it misses the bar by three points. It prints BUY - ADD at conviction 73 on PullbackBuy_VWAP - the entry style our playbook favours - at Buy_Rank 1, sitting essentially ON its model entry (0.09 ATR past, so no chase problem at all), with a 12-week view of, a full 22% above the close. The news corroborates it: hot-rolled coil has surged toward a short ton on mill price increases, extended lead times and tariff-suppressed imports; Q2 delivered 4bn of sales, 2bn of net income and record shipments; Morgan Stanley has raised its target; the shares are up 55% this year. We are not buying because our own intrinsic model marks the stock at against a close, and that gap alone disqualifies it. OPS: adj = 73 - 10 (val_adj, MoS -2.61 x 5 = -13.05 clamped to the -10 floor) - 1 (`PullbackBuy_VWAP|buy` cell bias, n=7 ACTIONABLE) = 62, three points under the 65 floor. Quality_Pass is False but the gate blocks NEW entries only and this is an ADD to a held, so the gate did not bind - val_adj alone killed it. NOT queued: a sub-floor name does not get a GTC. This is the 20th consecutive session with no acquire and the 2nd where an EPV/DCF mark on a quality/cyclical name pinned val_adj to the -10 clamp; see systemic finding 1. Re-arms on any MoS improvement or on conviction >= 76 at the same entry. Held stop (trail-raise only - see the NUE hold card). · news |
| MCO | Moody's Corp | 485.79 | NO ACTION | 485.79 | 461.27 | 485.79 (+0.0%) | 506.45 (+4.3%) | 488.29 (+0.5%) | Moody's is repaired enough to halve the pending trim but not enough to cancel it. After four straight NO TRADE prints with a dead forward view, it now prints HOLD at conviction 68 and its 12-week target has crossed back above the close, if only by 0.5%. The news is genuinely two-sided: Q2 EPS rose 31% on 15% revenue growth with the buyback target lifted to $3bn, but 2026 operating-margin guidance was trimmed and investors are actively re-pricing the firm's exposure to private credit after weak software-loan recoveries were flagged. A ratings franchise with a margin cut and a credit-cycle question is not a name to press a 50% cut on. OPS: DE-ESCALATED 50% -> 30% on one confirming print. Full withdrawal requires a SECOND confirming print - the same two-print standard applied to BAC on 08-17, so the treatment is consistent and auditable. Bleed at the reduced tier is 4th in the book, one slot short of tonight's three. Ledgered optional so the counterfactual grader prices the omission - which matters given `optional_trim` runs n=90 at a median +2.81% cost at 4 weeks. Stop held. · news |
| COF | Capital One Financial Corp | 221.47 | NO ACTION | 221.47 | 218.60 | 221.47 (+0.0%) | n/a | 260.02 (+17.4%) | A profit-take on COF has formally fired and lost its slot to bigger dollar bleeds, not to a weaker case. The position is up 3.58 ATR above cost - well past the 3.0 ATR arming line - and the deterioration tick that fires it is (d) 2 consecutive NO TRADE prints. Under ADR-0012 that combination is a signal-based reason for a 30% trim, so this is a mandatory candidate held back only by the three-card cap, never optional advice. OPS: single tick -> profit_take_1 tier (30%,). Arming taken from Gain_ATR alone, NOT from T12_Progress_Pct, which is miscomputed book-wide (systemic finding 2). MoS -0.35 is outside the -3 amplify line so no tier bump. G1 softening does not apply - the print is not a fresh BreakoutUp_*/MomentumContinuation_* at conviction >= 80. Ledgered optional purely so the counterfactual grader prices the omission. Stop held. |
| CVX | Chevron Corp | 205.74 | NO ACTION | 205.74 | 193.56 | 207.96 (+1.1%) | 228.74 (+11.2%) | 234.18 (+13.8%) | A profit-take on CVX has formally fired and lost its slot to bigger dollar bleeds, not to a weaker case. The position is up 8.89 ATR above cost - well past the 3.0 ATR arming line - and the deterioration tick that fires it is (a) RSI_DIV on the same row. Under ADR-0012 that combination is a signal-based reason for a 30% trim, so this is a mandatory candidate held back only by the three-card cap, never optional advice. OPS: single tick -> profit_take_1 tier (30%,). Arming taken from Gain_ATR alone, NOT from T12_Progress_Pct, which is miscomputed book-wide (systemic finding 2). MoS -0.77 is outside the -3 amplify line so no tier bump. G1 softening does not apply - the print is not a fresh BreakoutUp_*/MomentumContinuation_* at conviction >= 80. Ledgered optional purely so the counterfactual grader prices the omission. Stop held. |
| OXY | Occidental Petroleum Corp | 59.80 | NO ACTION | 59.80 | 55.23 | 60.48 (+1.1%) | n/a | 65.76 (+10.0%) | A profit-take on OXY has formally fired and lost its slot to bigger dollar bleeds, not to a weaker case. The position is up 7.79 ATR above cost - well past the 3.0 ATR arming line - and the deterioration tick that fires it is (a) RSI_DIV on the same row. Under ADR-0012 that combination is a signal-based reason for a 30% trim, so this is a mandatory candidate held back only by the three-card cap, never optional advice. OPS: single tick -> profit_take_1 tier (30%,). Arming taken from Gain_ATR alone, NOT from T12_Progress_Pct, which is miscomputed book-wide (systemic finding 2). MoS -0.50 is outside the -3 amplify line so no tier bump. G1 softening does not apply - the print is not a fresh BreakoutUp_*/MomentumContinuation_* at conviction >= 80. Ledgered optional purely so the counterfactual grader prices the omission. Stop held. |
| GOOGL | Alphabet Inc Cl A | 344.20 | NO ACTION | 344.20 | 319.70 | n/a | n/a | 308.31 (-10.4%) | Four straight NO TRADE prints and a 12-week view 10.4% below the close keep a 50% cut live on Alphabet; the position is 4.4% under cost and was already cut 30% at on 08-13 without going anywhere since. Bleed ranks 5th, one slot short. The stale 07-05 overlay's 'invalidated below 345' line on this name is tripped again at but carries no delta, so nothing acts on it. Stop held. |
| GOOG | Alphabet Inc Cl C | 341.28 | NO ACTION | 341.28 | 320.20 | n/a | n/a | 304.47 (-10.8%) | The C-line moves with GOOGL and carries the same case - four NO TRADE prints, a 12-week view 10.8% below the close, 4.2% under cost. Immaterial in isolation at of bleed; it executes with GOOGL when GOOGL gets a slot. Stop held. |
| DVA | DaVita Inc | 177.91 | NO ACTION | 177.91 | 174.64 | n/a | n/a | 111.26 (-37.5%) | DaVita is the worst holding in the book on every measure that is not dollar size - down 24.3% from cost, a 12-week view 37.5% below the close, and MoS -3.35, the ONLY name inside the -3 amplify line, which bumps the tier 50% -> 100% the moment it is slotted. It has now ranked 5th-6th on dollar bleed for five consecutive sessions and been passed over every time, which is precisely the ADR-0020 'never got cut at all' failure arriving by a different route (systemic finding 3). The one counterweight is a washed-out RSI of 26.2. Stop held. |
| SIRI | Sirius XM Holdings Inc | 28.62 | NO ACTION | 28.62 | 28.60 | n/a | n/a | 26.98 (-5.7%) | Sirius keeps a 30% trim live on a 12-week view 5.7% below the close and a 6.2% loss from cost, but the close is clinging 0.07% above its stop so the escalation basis stays withdrawn (it was de-escalated 50% -> 30% on 08-17 when the stop break reversed). Bleed ranks 6th. A decisive close below re-escalates to 50%. Stop held. |
| NYT | New York Times Co | 65.07 | NO ACTION | 65.07 | 60.57 | n/a | n/a | 43.67 (-32.9%) | The Times carries the worst forward view in the book - a 12-week target 32.9% below the close - on an AVOID print at conviction 56 on a failed breakout, with the position 11.9% under cost. The case is easily strong enough for a near-exit; only its size defeats it, at of bleed on a position. It will not reach a top-3 slot on dollar ranking, which is the argument for the materiality-adjusted tiebreak in systemic finding 3. Stop held. |
| KHC | Kraft Heinz Co | 24.82 | NO ACTION | 24.82 | 24.15 | n/a | n/a | 24.62 (-0.8%) | Kraft Heinz stays the weakest of the carried set: NO TRADE, but a 12-week view only 0.8% below the close - barely negative rather than decisively so - and the stock has drifted up two sessions running. It was flagged a withdrawal candidate on 08-17 and remains one; a second print with a flat-or-positive 12wk should retire the card outright. Bleed Stop held. |
| ALLY | Ally Financial Inc | 43.63 | NO ACTION | 43.63 | 43.26 | n/a | n/a | 41.03 (-6.0%) | Ally has softened from an AVOID to NO TRADE with a 12-week view 6.0% below the close and the position 3.6% under cost, so the 30% trim stays live but unpressed. Its UNDERVALUED reading argues the tier DOWN one under the sec.9(b) amplify rule and 30% is already the floor, so it is held there. Bleed Stop held - the close sits just 0.85% above it, so this is the next name likely to breach. |
| KR | Kroger Co | 56.34 | NO ACTION | 56.34 | 54.87 | n/a | n/a | 44.73 (-20.6%) | Kroger prints the book's strongest sell signal tonight - STRONG SELL - REDUCE at conviction 68, Sell_Rank 1, portion 88.5%, on a failed breakout with a 12-week view 20.6% below the close - and it is excluded on materiality for a fifth session: the entire position is worth Recorded so the signal is not silently lost. Its UNDERVALUED mark argues the tier down in any case. Stop held. |
| BAC | Bank of America Corp | 64.23 | HOLD | 64.23 | 62.57 | n/a | n/a | 75.80 (+18.0%) | Bank of America holds the best forward view in the book - a 12-week target 18.0% above the close - on a steady HOLD at conviction 61 for a third session, with the position 6.8% and 4.12 ATR above cost. ADR-0012: armed on gain but ZERO fire ticks (conviction flat, no warning code, forward view strongly positive), so the rule mandates a trail-raise and forbids a trim. Stop HELD at - the scan's would loosen the line and is rejected. The 08-17 withdrawal of its pending sell stands. |
| NUE | Nucor Corp | 264.19 | HOLD | 264.19 | 259.99 | n/a | n/a | 322.20 (+22.0%) | Nucor is the book's best-performing holding at +19.7% and 5.78 ATR above cost, and its conviction ROSE this session from on a fresh BUY - ADD. Armed on gain with zero deterioration ticks, so the ADR-0012 outcome is a trail-raise, never a trim. Stop HELD at - the scan's would loosen it and is rejected - which leaves only 1.59% of cushion below the close; that tightness is deliberate on a name this extended. See the NUE near-miss card for why the add was not taken. |
| KO | Coca-Cola Co | 88.82 | HOLD | 88.82 | 86.05 | n/a | n/a | 97.11 (+9.3%) | Coca-Cola gave back its BUY - ADD label from 08-17 and prints HOLD at 53, conviction down 13 points - short of the 20-point collapse that would fire a profit-take - with the position 5.6% and 3.38 ATR above cost and a 12-week view 9.3% above the close. Armed, not fired: trail-raise only. Stop RAISED -> (the scan's live level, a genuine tightening), leaving 3.12% of cushion. |
| VRSN | VeriSign Inc | 276.74 | HOLD | 276.74 | 276.41 | n/a | n/a | 276.61 (-0.0%) | VeriSign's stop breach has RESOLVED - the close is back 0.12% above the line after two sessions below it. Gain_ATR 2.37 is under the 3.0 arming threshold so no profit-take applies, and the 12-week view is flat at -0.05%. Stop HELD; the scan's would loosen it materially and is rejected. Cushion is thin enough that another down session puts this back on the breach list. |
| M | Macy's Inc | 23.23 | HOLD | 23.23 | 23.35⚠ | n/a | n/a | 23.31 (+0.3%) | Macy's closes 0.51% below its stop for a second session - 0.17 ATR, sub-noise on a position sitting flat to cost with a 12-week view 0.3% above the close. Not actioned on materiality. Stop held. |
| NVR | NVR Inc | 6248.37 | HOLD | 6248.37 | 6069.66 | n/a | n/a | 5683.06 (-9.0%) | NVR prints AVOID at conviction 63 on a failed breakout with a 12-week view 9.0% below the close and the position 7.4% under cost - a real sell case on a 3-share, line. Excluded on materiality; recorded so the signal is not lost. Stop held. |
| LEN | Lennar Corp Cl A | 84.94 | HOLD | 84.94 | 82.64 | n/a | n/a | 77.01 (-9.3%) | Lennar A flipped from HOLD to AVOID at conviction 53 with a 12-week view 9.3% below the close. The whole position is worth - immaterial. Stop held. |
| LEN-B | Lennar Corp Cl B | 83.50 | HOLD | 83.50 | 81.78 | n/a | n/a | 76.59 (-8.3%) | Lennar B carries the identical AVOID print and a 12-week view 8.3% below the close, on a 9-share, line. Immaterial. Stop held. |
| STZ | Constellation Brands Inc | 131.00 | HOLD | 131.00 | 128.06 | n/a | n/a | 112.81 (-13.9%) | Constellation reverted from HOLD to NO TRADE with a 12-week view 13.9% below the close, on a 23-share, line. Immaterial. Stop held. |
| LLYVA | Liberty Live Group Cl A | 100.00 | HOLD | 100.00 | 96.79 | n/a | n/a | 104.69 (+4.7%) | Both Liberty Live lines have collapsed from HOLD cv83 to NO TRADE cv0 in the three sessions since the GTC buys filled - the queue expired straight into a signal fade. The 12-week view is still 4.7% above the close and the position is only 3.8% under cost, so this is a watch, not a sell. Stop held. |
| LLYVK | Liberty Live Group Cl C | 103.73 | HOLD | 103.73 | 100.60 | n/a | n/a | 109.16 (+5.2%) | The C-line mirrors LLYVA: HOLD cv83/80 down to NO TRADE cv0 since the fills, 12-week view 5.2% above the close, 3.2% under cost. Watch, not a sell. Stop held. |
Outcome & track record
Accuracy at the time of this record.
117 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 →
-0.010 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.
46 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.