Recommendations
1 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 |
|---|---|---|---|---|---|---|---|---|---|
| KLAC | KLA Corporation | 175.45 | SELL 88.52% | 175.54 | 160.55 | 176.25 (+0.4%) | 181.46 (+3.4%) | 177.07 (+0.9%) | Exit the position. KLA closed at 175.54, decisively through its 184.00 trailing stop (-4.6%), on a RallySell_BearTrend signal with price below both its 20-day (191.46) and 50-day (200.40) averages and only 0.3% above its 200-day line. The 12-week projection of 177.07 is essentially flat to the current price, so the forward case has gone dead while the downside stays open. News confirms the derating is policy-driven and unresolved -- tightening US export rules on China sales plus tariff overhang -- rather than a fundamentals stumble, which is why five straight earnings beats have not stopped the slide. The valuation layer also reads rich (MoS -4.24). Take the loss at roughly breakeven versus the 175.75 cost and redeploy the capital. OPS: signal-originated exit under L2 (stop break + dead 12wk); Portion 88.52% is the model's own 'target full exit'. ADR-0020 Trail Erosion INDEPENDENTLY fires here and is veto-immune on first fire: peak stop 272.00 (2026-07-01) -> live 184.00 = 32.4% erosion with ZERO size reduction in 7 sessions. This card has now been written 4 EODs running (08-25/26/27/28) and never executed -- US-37. Valuation amplify SUPPRESSED by G2 (Est_12W 177.07 / DCF 61.80 = 2.87x, anchor unreliable), so no tier bump. scorecard_bias -5 from RallySell_BearTrend|sell (n=11, hit_1w.273). Residual sits on the model stop 160.55. · news |
| GOOGL | Alphabet, Inc.- Class A | 339.35 | SELL 30% | 346.59 | 333.68 | 349.28 (+0.8%) | 341.01 (-1.6%) | 366.78 (+5.8%) | Bank a third of a very large gain. The position is up 181.8% on a 123.00 and has reached 94.5% of its 12-week objective, but the signal engine has printed NO TRADE with zero conviction for five consecutive snapshots -- the trend that drove the gain is no longer readable. News confirms a real catalyst behind the stall: a leadership shake-up in Alphabet's AI division, including Jeff Dean's departure, knocked the shares despite strong Q2 results (revenue +24%, Google Cloud +82% with a $514bn backlog). The 4-week projection of 341.01 sits below today's price. Taking 30% locks in profit while leaving to run on an intact long-term story. OPS: ADR-0012 armed + fired -- PROFIT_ARMED (T12_Progress 94.5%, Gain_ATR 33.8) plus fire tick (d), 5 consecutive NO TRADE prints. One tick -> profit_take_1 = 30%. No valuation tier bump (MoS -0.69, not <= -3). G1 does not apply (no fresh breakout to protect). Third consecutive EOD this card has been written (08-26/27/28) unexecuted. 12wk 366.78 exceeds DCF anchor 145.29 by 2.52x. Residual on trail 333.68. · news |
| NVDA | NVIDIA Corporation | 220.78 | SELL 30% | 217.55 | 217.07 | 218.53 (+0.5%) | 219.85 (+1.1%) | 227.92 (+4.8%) | Take 30% off after the post-earnings pop faded. NVIDIA reported blowout Q2 results on 26-Aug (2bn revenue, EPS, FY28 growth guided to 70%) and the stock added $442bn of value on 27-Aug, the second-largest one-day gain in market history. Friday gave back 4.6% of that to close at 217.55 -- just 48 cents above our 217.07 trailing stop -- and the engine's conviction collapsed from in one session. The position has reached 95.5% of its 12-week objective and the forward view now offers only +4.8% over twelve weeks. The business news stays excellent; this is a risk decision, not a change of thesis, and stay on to carry it. OPS: ADR-0012 armed + fired -- PROFIT_ARMED (T12_Progress 95.45%) plus fire tick (b), conviction -27 pts vs the prior EOD, which fires REGARDLESS of G1. One tick -> profit_take_1 = 30%. No valuation bump (MoS -2.28 > -3). This is the exact 07-01 NVDA round-trip pattern (200 -> 194.5) that ADR-0012 was written to close -- it is a mandatory DISPOSE, not an optional trim. 12wk 227.92 exceeds DCF anchor 163.98 by 1.39x. Residual on trail 217.07; a settled close below 217.07 takes the remaining 70. · news |
| PLTR | Palantir Technologies Inc. | 186.38 | BUY — QUEUED @166.99 | 166.99 | 158.46 | 186.51 (+11.7%) | 197.54 (+18.3%) | 192.63 (+15.4%) | Bid for a starter position on a pullback, not at the market. Palantir is the strongest signal in the book -- maximum conviction 100, top buy rank, on a pullback-to-VWAP setup that has been the most reliable pattern we trade. The business news confirms it: Q2 revenue grew 94% to 94bn, US commercial revenue +149%, full-year guidance raised to 15bn, and the Pentagon formally designated the Maven Smart System a program of record now tracking toward $1bn of annual run-rate revenue. The catch is price: the stock has run 43% in a month and trades 12% above where the model wants to enter, and on intrinsic value it is the most expensive name in the universe. So we queue a limit 10.4% below Friday's close rather than chase, and size at half the normal amount. If it never fills, we never overpaid. OPS: QUEUED GTC limit @ 166.99 -- the anchor snapshot's own printed Entry_Price (repriced from the 08-27 queue @165.23 under the provenance rule; NOT a carried level). Entry range 165.15-168.83. Entry_Distance_ATR -2.62 -> no-chase queue mandatory. Fusion 100 -10 +1 = adj 91. Quality gate PASSES (Quality_Pass True). Size modifier x0.5 (MoS -13.08 <= -3): 0.8549 x x 20% = -> -> = and max_position_pct (60% x) neither binding. scorecard_bias +1 from PullbackBuy_VWAP|buy (n=24, the L1-favoured cell). 12wk 192.63 exceeds DCF anchor 46.62 by 4.13x -- projection assumes growth beyond priced-in. AUTO-WITHDRAW if: family leaves PullbackBuy_VWAP, conviction < 65, Quality_Pass flips False, or 2026-09-02 passes unfilled. If PLTR gaps BELOW 166.99, do NOT fill at the gap -- reprice at the next EOD. · news |
| OKTA | Okta Inc. | 173.04 | HOLD | 166.23 | 160.05 | 167.67 (+0.9%) | 175.00 (+5.3%) | 188.52 (+13.4%) | Hold the full and raise the stop. Okta is up 18.9% on a 139.80 cost after last week's beat-and-raise, but conviction has cooled to 56 and the shares gave back 3.9% on Friday. Not enough deterioration to sell, enough to tighten the risk line to 160.05. OPS: the 08-27 ADD QUEUED @156.65 is WITHDRAWN -- conviction 56 < 65 floor and the row prints HOLD with no buy trigger (provenance rule). Trail RAISED 147.61 -> 160.05 (the model's own Stop_Price, 3.7% below close). NOT profit-armed (T12_Progress 88.2 < 90, Gain_ATR 2.14 < 3.0) so no profit-take is available. Quality_Pass False -- irrelevant, gate blocks new entries only, never ADDs. |
| NBIS | Nebius Group, N.V. | 206.32 | HOLD | 209.18 | 213.00⚠ | 213.16 (+1.9%) | 229.14 (+9.5%) | 256.47 (+22.6%) | Hold, but the stop is now live. Nebius closed at 209.18, 1.8% below our 213.00 stop -- a breach, but not a decisive one, and the forward case is still the strongest in the book at +22.6% over twelve weeks with a bullish consolidation reading. The business news is intact and confirms the hold: Q2 cloud revenue grew 514% year over year, 2026 AI-compute capacity guidance was raised from 3GW to 4GW, and the $27bn five-year Meta compute deal plus Nvidia's $2bn investment are both untouched -- which is precisely the condition our standing NBIS override requires. Selling a name into a 4% market-wide risk-off day when its own news is unchanged is how you lose the position and keep the loss. OPS: NOT a dispose -- trim-discipline G3 FAILS (Target_12Week solidly positive AND signal bullish -> 'cutting the winner'); L2 requires stop-break PLUS negative 12wk and the 12wk is +22.6%. §9 forbids the MoS -8.69 from originating. Stop is NOT lowered to the model's 201.13 -- re-basing down is the exact ADR-0020 erosion failure (peak 235.69 -> 213.00 is already 9.6%, and 201.13 would take it to 14.7%, one notch under the 15% mandatory-trim line). PRE-COMMITTED: a SECOND consecutive settled close below 213.00 converts to a MANDATORY 30% trim, no re-litigation. · news |
| LLY | Eli Lilly & Co | 1156.73 | HOLD | 1174.61 | 1168.46⚠ | 1179.18 (+0.4%) | 1188.77 (+1.2%) | 1201.03 (+2.2%) | Hold on a tight leash. Lilly is flat on the session and holding just above its 50-day average, with conviction recovering to 57 and 97.8% of the 12-week objective already banked. OPS: PROFIT_ARMED but NOT fired -- no second Exit_Warning code, conviction flat vs prior EOD, 12wk 1201.03 clears the 1.02x dead-view test (1198.10) by a hair, no NO TRADE prints. Armed-without-a-tick = mandatory trail-raise, NOT a trim (ADR-0012). Trail held at 1168.46; the model's 1106.94 chandelier is LOWER and is refused as a loosening. Close sits only 0.5% above the line -- this one resolves itself early next week. |
| MU | Micron Technology, Inc. | 958.73 | HOLD | 932.86 | 929.07 | 940.11 (+0.8%) | 1030.47 (+10.5%) | 1101.42 (+18.1%) | Hold. Micron is up 8.5% on cost and barely moved on a heavy down day, but the engine reads it as conflicted with no tradeable edge either way. The 12-week projection is the most attractive in the book at +18%, so there is no case to sell -- just no case to add either. OPS: NO TRADE, NoTrade_Reason 'Consensus conflict + low conviction'. NOT profit-armed (Gain_ATR 1.54 < 3.0, T12_Progress 84.7 < 90) so the profit-take path is unavailable. Trail 929.07 unchanged; close 932.86 is only 0.41% above it. |
| NET | Cloudflare, Inc. | 305.11 | HOLD | 299.84 | 283.72 | 300.82 (+0.3%) | 303.61 (+1.3%) | 328.07 (+9.4%) | Hold. Cloudflare gave back 2.7% on Friday but is still well clear of its 283.72 stop and holding the re-rating it earned when Salesforce's results lifted the enterprise-software group. Conviction 61 is below the 65 action floor, so no add. OPS: WATCH label with no buy trigger -- hold only. Trail 283.72 unchanged (a 2.0x ATR chandelier off the close would be 267.22, LOWER, and is refused). DATA ISSUE: input/usportfolio.json carries TWO NET rows -- 'Clouadflare Inc' and 'Cloudflare, Inc.' -- so the scan reads Held=N and the monitor reports phantom unledgered_drift. Dedupe the file; the is real. |
| SPCX | Space Exploration Tech Corp | — | HOLD | — | 129.00 | n/a | n/a | n/a | Hold. SpaceX is privately held with no public quote, so it carries no signal, no conviction and no price targets -- it is held on a manual stop and reviewed on news only. Nothing material surfaced this week on marks, secondary prints or registration. OPS: permanent no_close mute -- absent from the scan CSV by construction, never synthesized. Manual stop 129.00 unchanged. Auto-unmutes to full signal + fusion treatment the moment it appears in the scan with a non-null Close. |
Outcome & track record
Accuracy at the time of this record.
168 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 →
-0.191 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.
52 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.