MU

Two specialists signal weakness: thesis percentile has collapsed to 19 and Win-Probability is a coin flip at 44. Catalyst is supportive but a positive catalyst doesn't rescue a broken thesis — that's exactly the CRWD lesson. Memory-cycle exposure is being outranked by cleaner AI-infrastructure plays on the watchlist. Forward setup from here is weak regardless of where we entered.

Two specialists signaling breakdown: thesis has fallen to 19th percentile and win-probability is sub-50. Catalyst is only tangentially supportive (AI compute drives memory demand indirectly). The memory cycle thesis needs earnings confirmation that isn't in hand. This is exactly the deterioration-on-evidence case — trim regardless of held status.

Two specialists signal deterioration: thesis is bottom-quintile and win-probability collapsed to 38 on cyclical weakness. Momentum at 76 offers some support but is diverging from fundamentals. Forward setup is weak regardless of prior performance — trim on the evidence.

Two specialists actively signal deterioration: thesis at the 19th percentile and win_probability at 38 with the analyst explicitly flagging data anomalies and cyclical weakness. Catalyst is only generic ecosystem tailwind with no MU-specific news. Momentum at 57 is unconvincing. This is the evidence-based cut principle — held status does not rescue a name where the forward signals have broken down. Sizer can decide magnitude.

Two specialists flag deterioration: thesis has slipped to bottom-decile and win-probability at 38 signals the setup is broken from here. Even with the 87% data center gross margin catalyst, the forward view has weakened materially. Cycle turning against memory names — trim on evidence, don't anchor to prior thesis.

Two specialists flag deterioration: thesis has collapsed to the 19th percentile and win_probability at 38 signals the setup is no longer working. Catalyst is only indirect (broader AI compute) with no direct MU driver. Momentum still holding at 74 keeps it from a full exit, but the forward view has weakened materially. Score the stock, not the trade — this doesn't look like a durable winner from here.

Two specialists are signaling deterioration: thesis has collapsed to the 19th percentile and win_probability sits at 38 with active price weakness. Catalyst is only indirectly supportive via broad AI memory narrative. The forward setup is materially weaker than the position we're carrying. Trim on evidence; don't anchor to prior conviction.

Two specialists signaling breakdown: thesis has collapsed to 19th percentile and win-probability sits at 38 with data quality flags. Catalyst support exists via HBM demand but cannot rescue a weakening forward setup. Momentum is neutral. Forward view has deteriorated meaningfully from the memory bull run — trim on evidence.

Two specialists are signaling breakdown — thesis has collapsed to the 19th percentile and win-probability is at 38 with visible near-term weakness. Catalyst score is only sector-adjacent, not MU-specific. This is the deterioration pattern that demands cutting on evidence rather than anchoring to prior conviction. Forward view from here is weak.

MU has been a top winner but the forward view is deteriorating: thesis percentile has dropped sharply to 19, price is rolling over -6.7% in 5d, and momentum sits at 38. Catalyst remains supportive from AI memory demand and win_prob is balanced. Not an exit, but conviction has to come down from the 90+ zone this name held for weeks — the setup is no longer clean.

Thesis has fallen to the 19th percentile and momentum has rolled over to 38 with a -6.69% 5-day. This is exactly the deterioration signal the framework calls out — two specialists (thesis and momentum) show breakdown. The prior breakout setup is unwinding. Win-prob at 55 is only moderate and no longer justifies the prior high conviction. Score the stock, not the trade — this is not the same MU that ran.

Thesis percentile has collapsed to 19 and momentum has faded to 38 — two specialists signaling active breakdown. This was a prior winner but the forward setup is deteriorating meaningfully. Not zero (still HBM/AI exposure) but conviction has to come down hard on the evidence.

Catalyst backdrop remains supportive with named BlackRock/Asia memory commentary, but thesis percentile has slipped to 48 and the stock just took a 6.7% hit. Win-probability is only middling. Forward setup is no longer the conviction winner it was — stock-level momentum and thesis are softening. Hold-grade, not add-grade.

Momentum still strong and HBM demand thesis intact, but specialists flag near-term weakness and thesis percentile has eased as the story matures. Forward setup is constructive but no longer top-tier. Hold-grade conviction.

Textbook winning pattern: AI-infrastructure name pressing 52w high with +8.7% 5d move, direct ticker-relevant catalyst (memory-chip boom article), and win-probability confirming. Thesis percentile is muted at 48 but that's the transient low print our learnings warn against trimming on. Let the tape and catalyst speak — this is exactly the MU/MRVL setup we want to ride.

Win-Probability is what we weight most and it reads 78 with price breaking to new highs on HBM demand. This is exactly the pattern that worked for us before — strong tape, hyperscaler-linked catalyst, override the noisy thesis percentile. Thesis_pct at 48 reflects ranking compression vs newer AI names, not deterioration. Forward setup remains strong; let it run.

This is the textbook winner pattern from our learnings: AI-infrastructure name breaking to 52wk highs with strong momentum and supportive catalyst. Win-probability and momentum are the strongest in the batch. Thesis percentile slip is noise — the tape and HBM cycle are clearly working. Strong forward setup.

Textbook winner pattern: AI-infrastructure name breaking out to 52w high on direct Nvidia/hyperscaler partnership news. Win-prob 78 with momentum 92 and concrete catalyst confirmation is exactly the setup we've learned to lean into. Thesis percentile slipped to 48 on relative ranking but forward setup is among the strongest in batch.

Win-Probability is the strongest in this batch at 78, with price essentially at 52wk high and HBM demand explicitly cited across multiple hyperscaler capex headlines. Thesis percentile compressed to 48 as it's already worked, but the forward setup remains strong on the AI infrastructure leg. This is the pattern our biggest winners followed — trust the price/news combination.

Win-probability is the strongest signal here — price essentially at 52w high with 8.7% 5d gain and explicit HBM/AI compute tailwinds from today's NVDA news flow. This matches our winning pattern playbook: trust price/news combination over middling thesis percentile. Forward setup remains strong.

Textbook winner pattern: breaking to 52w highs with explosive momentum on confirmed AI-memory demand. Per our learnings, this is exactly the pattern (MRVL, MU, AMAT) we should ride through score noise. Thesis percentile relative-rank slipped because newer AI names rose, but the forward setup is among the strongest in the batch. Sizing is the Sizer's call; the stock itself is high conviction.

Momentum is exceptional and the catalyst stack (HBM for GB200, storage bottleneck narrative, HPE-NVIDIA expansion) is exactly the name-specific AI-infra setup that worked for us before. Thesis percentile is modest but Win-Probability and the breakout-character momentum carry the forward view. Strong hold from here.

Classic winning-pattern setup: AI memory name with momentum_pct 88, concrete HBM3E/NVIDIA GB200 catalysts, and supportive sector news. Thesis percentile is mediocre at 48 but win-prob and momentum more than compensate — this is exactly the MRVL-style profile our learnings tell us to bias toward. Forward setup remains strong; sizing decisions belong to the Sizer.

Best-in-batch momentum with concrete HBM3E/NVIDIA catalyst — this is the winning pattern we identified. Thesis percentile is mediocre but Win-Probability and catalyst are strong, and the stock is making new highs. Forward setup remains constructive; sizing is the Sizer's call.

Momentum at 100 and a concrete HBM/Nvidia catalyst — textbook winner pattern from our learnings (MRVL/MU profile). Thesis percentile is mediocre but win-probability and catalyst are the more important forward signals here. Sector confirmation today from multiple GPU deployment headlines reinforces.

Momentum at 90 and HBM3E demand for GB200 is exactly the AI-infrastructure hardware pattern that worked. Win-prob solid at 66, catalyst supportive with LPDDR6 article. Thesis percentile relatively low but the forward setup — breakout proximity plus named hyperscaler demand — is what matters. Constructive forward view.

Momentum is exceptional and the HBM cycle remains the dominant AI memory story. Thesis percentile slipped to 48 as the watchlist crowded with newer AI names, but Win-Prob and the breakout profile are exactly what we've gotten right recently. Forward setup stays constructive.

MU is the cleanest expression of the winning pattern from our learnings: AI-infrastructure hardware (memory/HBM) at top of its 52w range with momentum at 100 and a concrete NVIDIA GB200 HBM3E catalyst. Thesis percentile is only 48 but that's exactly what our learnings flagged — initial mediocre thesis scores on breaking-out semis names (MRVL, MU) have been our best winners. Win-probability 66 and strong sector flows justify continued holding with high forward conviction.

Two specialists signal active deterioration: thesis sub-50, win-probability sub-50, and a sharp technical break. Momentum percentile is misleadingly high given the recent gap down. Forward setup has weakened materially; conviction down hard regardless of held status.

Two specialists flag active deterioration: thesis only 48 and win-prob 44 with a sharp 13% weekly drop. Momentum percentile is high but masks a real near-term breakdown. Cyclical memory exposure without conviction — forward setup has weakened materially.

Held position barely positive. Direct NVIDIA HBM4 validation is the saving grace, but thesis and win-probability specialists are both lukewarm and 5d action is poor. Not breakdown territory yet — hold but do not add. Watch for follow-through on HBM4 narrative.

Held position only modestly green (+1.96%) — doesn't trigger winner-protection rule (needs +20%). Specialists are mixed: catalyst and momentum strong, but thesis and win-prob both sub-50. Hold rather than add; not enough deterioration to trim either. Memory cycle volatility keeps conviction muted.

MU is the weakest held position in this batch: thesis and win-prob both sub-50, near-term technical breakdown of 13% in 5 days, and memory cyclicality risk. PnL only -5% so not deep loss, but two specialists signal deterioration. Trim to free capital for higher-conviction names.

Held memory play down 7% with deteriorating technicals and two specialists flagging concerns (win_prob 44, thesis 48). Doesn't meet the 15% loss threshold for forced exit, but the cyclical setup is weakening and the catalyst tailwinds aren't translating to price action. Trim to reduce concentration.

Held loser down 10% with thesis and win-prob both below 50. Momentum weak despite supportive HBM/DRAM catalyst narrative. Not yet at the -15% cut threshold but specialists are not constructive enough to defend. Lean trim to free capital for higher-conviction AI infra.

Held at -1.7% (not yet a loser by 15% rule) but two specialists flag deterioration: win-prob agent sees negative near-term setup, thesis agent only mid-tier. Catalyst supportive but not enough to override breakdown. Trim to reduce cyclical exposure.

MU is down 11% from entry, below the -15% cut threshold. Specialists are mixed-to-constructive on the memory cycle and HBM demand, but no specialist signals active breakdown. Hold the position without adding given semi concentration.

Held position down 11% with mediocre specialist support across the board: thesis=48, win_prob=55, catalyst=55, momentum=38. Per core principle 2, losers without conviction get cut. Not yet at the -15% threshold for forced exit, but conviction is deteriorating and there's no specialist screaming to add. Trim to free capital for higher-conviction names.

Held long at modest -6% loss, below the 15% cut threshold. Memory cycle thesis intact with HBM demand, but specialist scores are mediocre across the board. Maintain position size; not adding given concentration.

MU is a clear breakout winner — momentum 84, performance 82, near 52w highs on confirmed HBM demand. This matches our winning pattern (Micron returned +79% in similar setup). PnL only +4.77 so still early in the move. Protect the winner; do not let mediocre thesis_pct=48 (already-priced-in concern) suppress conviction.

Held memory winner up ~10% with momentum pinned at 98 and supportive AI memory catalyst flow. Thesis percentile dipped to 48 as newer names rose but recent learnings explicitly flagged Micron as a big winner from staying long. Protect the winner — hold without trim.

Held winner with strong momentum and supportive catalysts from AI memory demand. Thesis percentile is modest (48) but performance and momentum are top-tier. Protect the winner — no reason to trim a profitable position with momentum=98 and confirmed sector tailwinds.

Held winner with explosive momentum (top decile) and confirmed AI memory catalyst directly from Intel and NVIDIA announcements. PnL +6.9% is positive though not yet a protected winner (sub-20%). Thesis percentile dropped as other names rose, but Win-Probability acknowledges HBM-led earnings revisions. Hold the position — momentum + catalyst alignment is the winning pattern.

MU is the textbook winner from our recent learnings — up 89%, momentum 99, with today-dated thesis-confirming articles directly naming Micron's HBM segment. Thesis_pct drift to 57 reflects newer names rising, not MU breaking down. Protect aggressively; this is a top-of-book position.

MU is the archetype winner from our learnings — up 89%, top decile performance and momentum, with direct AI memory demand catalyst confirmed today. Thesis percentile compression to 57 is exactly the rotation trap we are told to ignore. Protect aggressively; this is a core position deserving outsized weight.

MU is our flagship winner up 88% with two direct headlines today explicitly tying agentic AI to structural memory demand expansion. This is exactly the fingerprint from our learnings — breakout name with direct AI-infrastructure catalyst. Protect aggressively; thesis_pct compression to 57 is the classic 'newer names rose' artifact, not a thesis break. Do not trim.

This is the protect-winner archetype. Up 83% since entry, top decile performance and momentum, HBM thesis still intact. Thesis percentile slipped to 57 only because newer names rose in the rank, not because the thesis broke. Per learnings, MU earned an oversized weight — do not trim a +83% winner on a thesis-rank rotation. Top conviction in batch.

MU is the textbook protect-winner case: +78% PnL, performance and momentum at the ceiling, and the HBM/AI memory thematic remains the cleanest direct catalyst in the portfolio. Thesis_pct dropped to 57 only because newer names rotated up the percentile rank — that is not a signal to trim our best performer. This is the kind of name our learnings explicitly say to size more aggressively, not less.

MU is our flagship winner up 72% since entry — exactly the protect-winners scenario. Three of four specialists remain constructive; only thesis percentile drift reflects relative ranking against newer names, not deterioration. Per the recent learnings, MU earned aggressive sizing and persistent top score. Do not trim. Keep at top conviction.

Held winner up 46% — protect winner rule applies. BofA 'unprecedented chip cycle' and HBM demand directly support the thesis. Thesis_pct softened to 57 as newer names rose, but Win-Prob and catalyst remain supportive and no specialist signals breakdown. Per recent learnings, persistently top-scoring winners deserve aggressive sizing. Hold core position.

Protect this winner. MU is +46% since entry with strong performance and momentum percentiles, and a specific, dollar-attached catalyst (CEO commentary on $1000 path tied to agentic AI token consumption driving HBM demand). Thesis_pct compressed to 57 only because newer names rose — does not justify trimming. Hold size.

MU is a clear winner up 46% since entry with top-decile performance and momentum. Per protect-winners rule, do not reduce conviction on a +20%+ position without 3 of 4 specialists signaling breakdown — only catalyst is soft, not negative. Memory cycle thesis intact. Hold and let it run.

Micron up nearly 49% since entry — this is exactly the winner we must protect. Performance and momentum percentiles in the top decile, sector catalysts from NVIDIA prints support continuation. Recent 5d weakness is noise against the longer trend. Hold conviction firmly in protect-winner zone.

Micron is a clear winner at +46% with strong performance and momentum percentiles and a directly relevant HBM catalyst. Per protect-winners rule, do not reduce conviction below 75 absent specialist breakdown — only the win_prob agent is balanced, not negative. Hold and let it run.

Held winner up nearly 40% — protect-winner rule applies. Specialist scores are middling but no active breakdown signal across 3+ agents. HBM/AI memory thesis is exactly the pattern our biggest wins came from. Hold, do not trim on a single weak week.

Held winner up 38%. Per protect-winners rule, do not cut below 75 absent broad specialist breakdown — only one specialist (catalyst light) is muted, others remain constructive. Recent pullback is noise within a cyclical winner. Hold the position; let it run.

Held winner up 33% — protect-winner rule applies. Specialists are mixed but none signal active breakdown; only one (risk) flags concentration which is a portfolio-sizer issue, not intrinsic. Recent pullback noted but PnL cushion and thesis intact. Hold, do not trim.

Textbook winner-protection case. Up 41% since entry, +15.5% in 5 days, at all-time highs with HBM/AI memory thesis intact. This matches the exact pattern from learnings (MU range 99.5, perf 83, mom 75 -> +98% historically). Thesis_pct middling at 57 is irrelevant — breakout + AI-spend headline combo dominates. Run the winner.

Textbook winner per our recent learnings: breakout to 52wk high + AI memory infrastructure thesis = run the position. Up 41% since entry with explosive +15% week. Protect-winner rule applies; no specialist signals breakdown. Hold firmly.

Textbook protect-the-winner: +43% PnL, performance percentile 90, near 52wk high with +15% 5d move, direct AI memory/HBM catalyst. This is exactly the MU pattern flagged in learnings (breakout + AI-spend headline dominates middling thesis_pct). Hold with high conviction - do not trim a winning HBM play in the middle of the AI capex cycle.