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 |
|---|---|---|---|---|---|---|---|---|---|
| AAOI | Applied Optoelectronics, Inc. | 113.76 | SELL ALL | 113.76 | 111.93 | n/a | n/a | 80.24 (-29.5%) | Closing the book's worst holding outright at the open. Applied Optoelectronics broke its risk line decisively on Monday after announcing a share sale, and although the price has since crept back above that line by a hair, everything that matters has kept deteriorating: our twelve-week view now sits 29% BELOW today's price and has fallen for seven sessions straight, the shares remain under every meaningful trend line, and the bounce is running on shrinking volume. The tell was today itself - the optical components sector had its best day of the week, with the sector leader up more than 6%, and this stock managed +0.5%. The business is genuinely growing (record revenue, +86% year on year, strong AI-driven demand for its next-generation products), but it still loses money, burns cash, and is issuing new shares into every rally, which caps the upside the growth story would otherwise deliver. We take the 24% loss and redeploy. OPERATIONAL: fifth re-fire of the 08-25 exit (executor 7 sessions stale, last_run_date 2026-08-18); re-fired at a BETTER price, 113.76 vs the 113.15 ref = +; realized -vs cost 150.03. STOP LEG STATED PLAINLY: close 113.76 vs stop 111.93 = +1.63% = 0.14 ATR on an 11.4%-of-price ATR - reclaimed by noise width for a SECOND session; a third print starts to look like repair, not drift. OTHER LEGS DEEPENED: (a) T12 80.24 = -29.47%, monotonic 135.97->116.18->120.41->112.17->84.20->83.05->80.24, seventh consecutive dead session; (b) below EMA20 123.09 / EMA50 127.70, VWAP_Distance -11.0% (WIDER than -10.14%), Consolidation_Bearish, Trend 50 / Mom 40 / Vol 28, AVOID cv59 on a bearish family with Exit_Strategy 'immediate'; (c) volume decay 14.96m -> 6.69m -> 4.51m. UT-9 THIRTEENTH SESSION: PROFIT_ARMED is INERT (Gain_ATR -2.80, position -24.18%, arm-on-loss); T12_Progress_Pct 141.77 is the systemic-finding-3 artifact. ADR-0020 did not fire (peak_stop 111.93 = live stop, 0% erosion). L8/GAP: market order - a gap up at the 08-27 open on NVDA-beat optics sympathy only improves the fill; the basis is a dead forward view, not a price level. FALSIFIER: a decisive close above EMA20 123.09 with T12 positive repairs the basis and it re-enters as a fresh buy candidate, subject to the quality gate it currently fails badly (F 4, negative EPV, FCF_Yield -5.81%, Quality_Rank 28/31). · news |
| NVDA | NVIDIA Corporation | 209.66 | HOLD | 209.66 | 206.05 | 210.73 (+0.5%) | 220.78 (+5.3%) | 212.27 (+1.2%) | Holding after a strong earnings report that the market shrugged at. Nvidia's quarterly results, out last night, beat on every line - 2bn of revenue against 3bn expected, earnings of versus, data centre revenue up 117%, and next-quarter guidance well above what analysts were carrying. The shares nonetheless slipped around 2% on the news, a familiar pattern for a stock that had already drifted lower into its own report. Nothing in the result changes the case for owning it, and nothing in the price action gives us a reason to sell: the stock sits comfortably above our risk line and there is no sell signal. We are, however, watching one thing closely - our own twelve-week projection has flattened to barely above today's price, which means the shares need fresh news to make progress from here. If the risk line breaks, we exit in full rather than trim. OPERATIONAL: catalyst CLEARED (Q2 FY2027 printed after the 08-26 close; horizon band -8%/+9% = -, observed move well inside). NO EXIT BASIS on any origination class: HOLD label, close 209.66 = +1.75% above the 206.05 stop, no SELL print, no confirmed lower high, conviction = -7 (short of the -20 collapse tick). UT-9: PROFIT_ARMED prints at Gain_ATR -3.08 on a -7.56% position = arm-on-loss, INERT - the fired-profit-take path cannot open. L6 forbids adding into the post-earnings move. DETERIORATION STATED: T12 212.27 = +1.24% only, BELOW the 1.02x threshold 213.85 = dead forward view by the framework's own definition, decayed 246.90->240.68->236.63->230.42->216.73->219.56->212.27; now below EMA20 214.15 and marginally under EMA50 210. only EMA200 198.38. Two of four fire ticks live; the arming gate correctly holds the trim. NO TRAIL-RAISE: chandelier 2.0 x ATR (5.562) below the 213.60 high = 202.48, LOWER than the standing 206.05 - never lower a trail. ADR-0020: peak_stop 206.05 = live stop, 0% erosion. FORWARD TRIGGER: a decisive close below 206.05 arms the L2 exit (stop-break + non-positive 12wk) at the next EOD and will be taken as a FULL EXIT, not a trim. · news |
| DELL | Dell Technologies Inc | 463.82 | HOLD | 463.82 | 420.84 | 468.48 (+1.0%) | 512.36 (+10.5%) | 587.69 (+26.7%) | Holding the full position and raising the protective stop; the position is in profit for the first time. Yesterday's decision to stand down the planned half-sale has been vindicated - Dell rose another 2.7% and has now recovered every level that the sell case depended on, closing above the line whose break originally triggered it. Conviction in the signal has improved for a fourth straight session and our twelve-week projection points 27% higher. The business case is intact and improving: Dell booked 1bn of AI server revenue in a single quarter, 4bn of AI orders, and ended with a record 3bn backlog, and the shares are up 262% this year yet still trade below where analysts value them. Earnings land in about a week, so we hold through the event with a raised stop rather than taking anything off beforehand - the known risk there is margin pressure as lower-margin AI servers dominate the sales mix. OPERATIONAL: repair complete - close 463.82 is +3.46% above EMA20 448.32, +2.28% above VWAP, and +1.87% ABOVE the 455.32 line the exit was built on (from -4.86% two sessions ago); today's high 469.38 extends the higher-high sequence. Conviction ->. Exit_Warning blank, Consolidation_Bullish, Trend_Score 70, RSI 55.1. Unrealized + (+0.55%), positive for the first time. TRAIL RAISED 419.19 -> 420.84 (chandelier 2.0 x ATR 24.269 below the 469.38 high). ADR-0020: peak_stop 455.32 vs live 420.84 = 7.57% erosion, under the 15% threshold - DID NOT FIRE. CATALYST-AWARE HOLD BINDS: FQ2 earnings 09-03 per horizon (external coverage says Sept 1), inside the <=5-session window - no pre-trim; a trim requires a CONFIRMED lower high or a decisive stop break, neither exists. NOTE the signal's own Stop_Price prints 451.53 (today's low, 0.51 ATR below close) - far tighter than the managed chandelier and whipsaw-prone into a print; the CHANDELIER GOVERNS (systemic finding 4). · news |
| LITE | Lumentum Holdings | 939.03 | NO ACTION | 939.03 | 876.09 | 953.33 (+1.5%) | 1087.97 (+15.9%) | 1076.02 (+14.6%) | Passing for a second day on the strongest buy signal this book has produced, because the company fails our fundamental quality screen - and recording what that has cost. Lumentum rose another 6% today and is up 13% over the two sessions since we first stood aside; the shares have more than doubled this year on surging demand for the optical components that connect AI data centres, the company beat its latest quarter and guided the next one strongly, and Washington is preparing restrictions on Chinese optical imports that would favour US suppliers. Against all that, the business currently converts almost none of its revenue into free cash flow, which is precisely what our screen tests, and our own valuation work puts fair value at a small fraction of the market price. We would also be buying at the top of a 6% day with no pullback to enter on, which is the kind of chase this book has been burned by before. Recorded, not bought. OPERATIONAL: adj 66 CLEARS the 65 floor (cv 76, val_adj -10) but Quality_Pass False blocks the new entry - F-Score 6, Quality_Rank 22.5/31, FCF_Yield 0.149%, Earn_Yield 0.388%, MoS_FV -32.505 (FV/sh). Gate cost now measurable: 830.17 -> 885.57 -> 939.03 = +13.11% over two blocked sessions. Two independent secondary vetoes: signal is FailedBreakdown_Up_20D (the family L1/UT-12 discount) and Entry_Distance_ATR 0.0 means the entry IS the close. Ledgered direction:none per ACTIVE advisory (d)(iii) so the F-Score/FCF gate's cost keeps being priced - this is systemic finding 1 and should headline the next learnings review. · news |
| MU | Micron Technology, Inc. | 938.40 | NO ACTION | 938.40 | 914.71 | 947.50 (+1.0%) | 1033.17 (+10.1%) | 997.77 (+6.3%) | Not yet - but this is the one we are closest to buying. Micron is the highest-quality company in this portfolio on our own screen, and the only name in today's buy list that passes the fundamental quality test at all. The backdrop is genuinely strong: its entire 2026 production of the high-bandwidth memory that AI accelerators depend on is already sold out, 2027 is pre-sold too, and the newest generation carries roughly 80% higher pricing than the last, with management expecting tight supply to persist beyond 2027. The signal has strengthened for two sessions and the valuation gap has narrowed. What stops us is that the signal's conviction is still five points short of the level this book requires to commit capital, and buying today would mean paying the market price with no pullback to enter on. The counter-argument worth holding in mind is that at least one large broker has begun trimming targets on the view that memory pricing peaks in 2027. We wait for either a stronger signal or a better entry. OPERATIONAL: adj 60 = cv 66 + val_adj -6, five short of the 65 floor; Quality_Pass TRUE (the only buy-pool name that passes), Quality_Rank 1.0/31, F-Score 8, Rev_Growth +259.5%, FCF_Yield 2.44%, MoS_FV -1.107 (shallowest in the pool, tightened from -1.349). The 08-21 re-arm (raw conviction >= 72, OR MoS_FV above -1.0) is still unmet but BOTH legs moved toward it: conviction ->66. Entry_Distance_ATR 0.0 = chase. Ledgered direction:none as a floor-miss counterfactual so the cost of the 65 floor gets priced alongside the quality gate. · news |
Outcome & track record
Accuracy at the time of this record.
23 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 →
0.096 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.