← All Scores / ARM
BULL ai-chips Arm Holdings

ARM

54 +7 all-time
Last scored 2026-07-14

Arm architecture dominates edge and mobile AI inference, enabling agentic AI to run on billions of devices beyond the data center.

Reading — Higher scores = stronger evidence of agentic AI tailwinds benefiting this company.
§ 01 — Why this score

Conviction breakdown

Confidence slipped this period because, despite a strong underlying business, the stock's price is considered extremely stretched relative to earnings, making the odds of a good outcome unfavorable right now.

Case for the company
How strongly the company fits the agentic-AI disruption story
71 0
Odds it pays off
How likely the call actually works out
58 0
Near-term catalysts
Upcoming news or events that could move the stock
60 0
Price momentum
Whether the stock's own recent price trend is rising
26 -32
Financial strength
Balance-sheet and business soundness — higher means safer
15 0
How the score works. These five signals are weighed together into one forward-looking conviction from 0–100 — it's a judgment, not a sum of the bars. That number is then ranked against every other stock we track, so it can also move when other stocks move. It reflects the outlook from here, not how the price has done for us.
Initial read this run: 42 → ranked to 54.
§ 02 — Score History
100
Ceiling
54
Latest
0
Floor
§ 03 — Analyst notes

Run-by-run detail

The fund's full reasoning for each scoring run, in its own words. More technical than the plain-English summary above.

2026-07-14 54 -2

Fails convergence bar (win_prob 58 < 65). Thesis intact but extreme valuation and lowest risk_quality in batch make it uninvestable at current levels without a clear catalyst. Skip.

2026-07-13 56 +10

Interesting royalty ramp story but Win-Probability under the 65 convergence bar and intrinsic risk_quality is dangerously low with an extreme multiple. Doesn't clear the 'great company AND great price' hurdle for a new buy. Watch, don't act.

2026-07-12 46 +17

Priced for perfection at 370x PE with deteriorating momentum and lowest risk quality in the batch. Win-probability at 58 doesn't come close to clearing the new-buy bar. Classic expensive AI-narrative name where the composite should fade — pattern-matches to ORCL/ARM prior correct skips.

2026-07-11 29 -10

Fails convergence (win_prob 58 < 65) and the intrinsic quality score is the worst in the batch. Priced for perfection with a 370x PE while momentum weakens. This is exactly the pattern (CRWD-like) recent learnings warn against — narrative name, no valuation discipline, no confirming momentum. Skip.

2026-07-10 39 +10

Classic expensive AI-narrative name with diverging momentum and win-probability — the exact pattern the system correctly faded on ORCL/ARM historically. 370x PE and risk_quality of 15 leave zero cushion for any guidance stumble. Skip.

2026-07-09 29 -5

Extreme valuation, deteriorating price action, and win_probability well below the new_buy bar. Prior ARM miss (-8%) validates fading expensive AI-narrative names when momentum and win-prob diverge. Hard skip.

2026-07-08 34 +8

Fails new_buy bar (win_prob 58). Extreme valuation with weakening momentum is exactly the ORCL/ARM fade pattern we correctly skipped historically. Positive catalyst can't override a 370x PE with -6.58% weekly action. Hard skip.

2026-07-07 26 -34

ARM has thesis appeal as foundational AI IP but at 370x PE with weakening momentum and a 15 risk-quality score, the forward setup is poor. Fails convergence (win_prob 58 < 65). Any guidance disappointment would be punishing at this multiple. Hard skip.

2026-07-06 60 +18

Fails the new-buy convergence bar hard: win_probability at 58 is well below the 65 threshold, and risk_quality of 15 is a red flag on its own. 370x PE means any guidance miss is catastrophic. Structural ARM v9 story is real but the entry point offers no margin of safety. Skip.

2026-07-05 42 -15

Fails convergence bar and is the lowest-quality name in the batch. Extreme valuation, weak momentum, and high volatility make this the wrong AI-chip vehicle when NVDA/AVGO/TSM are cleaner expressions of the same thesis. Skip.

2026-07-04 57 +25

Extreme valuation and worst risk quality in the batch. Win-probability fails the new-buy convergence bar and there's no breakout momentum to justify overriding. Hard skip.

2026-07-03 32 -27

PE 398, risk quality 15, momentum 42, and 35% below highs with no direct catalyst. Fails both convergence tests. Exactly the profile of stocks we should be avoiding — expensive, weak-quality, not breaking out. Hard skip.

2026-06-30 59 -19

Win-prob 58 is below the 65 new-buy bar and thesis at 73 doesn't carry it alone. PE 398 with -3.87% recent weakness and 35% gap to highs is the worst risk/reward in the batch. Not the breakout-with-anchor pattern that worked for us. Hard skip.

2026-06-29 78 +4

Named Oracle partnership is exactly the catalyst pattern we like, but win_probability of 58 fails the convergence bar (need 65+), risk_quality is the lowest in the batch at 15, and PE 398 is binary around any earnings event. Catalyst creates urgency, not direction — and direction here isn't yet confirmed. Skip.

2026-06-28 74 +9

Per recent learnings, ARM is a name where we want to cap the risk_quality penalty when catalyst + price action align. But here momentum_pct is only 33 and the catalyst is indirect (Qualcomm-driven, not ARM-specific). Win-probability of 60 is below the standard 65 bar and the ARM-specific loosened bar requires range_position breakout, which is borderline. Pass for now; reassess on a direct ARM catalyst.

2026-06-27 65 -13

Our ARM learning says cap the risk penalty when catalyst and breakout align — and they partially do here. But win-probability of 60 still falls short of even the relaxed 50 bar applied only when range_position is at extremes, and momentum_pct=33 says the tape isn't actually breaking out yet. Close but not converged.

2026-06-26 78 +20

Per ARM-specific learning, when catalyst is strong and price is breaking 52w range, we treat risk_quality more leniently and lower the win-prob convergence bar. Catalyst at 70 and analyst upgrade present, but win_prob 60 and thesis 73 still fall short of a high-conviction new buy. Constructive lean but not pulling the trigger; valuation extreme keeps this from being a layup.

2026-06-25 58 -5

Our recent learnings warn against reflexive risk-quality vetoes when catalysts and price align — but here the catalyst is generic AI-architecture commentary rather than an ARM-specific hyperscaler/partnership headline, and win_probability=60 + thesis_pct=73 fails the convergence bar. 529x PE leaves no margin. Skip.

2026-06-24 63 -4

ARM learning pattern says cap risk penalty when catalyst+performance+range_position align with a name-specific headline — but today's catalyst is sector-derived, not ARM-specific, so the override doesn't trigger cleanly. Win_prob 60 below the 65 bar. Extreme valuation at 529x PE limits forward asymmetry. Skip without convergence.

2026-06-23 67 +6

Learning notes flag ARM as a past missed opportunity, but the specific override (catalyst >=65 AND range_pos >=95 AND ticker-specific headline) isn't fully met today — catalyst is sector-validating, not ARM-specific. Win-probability of 60 falls short of the 65 new-buy bar even with the relaxed framework, and momentum=38 doesn't confirm a breakout. Skip but keep on watchlist for a cleaner setup.

2026-06-22 61 -18

Despite the ARM-pattern learning about not over-penalizing risk_quality on breakouts, this lacks a ticker-specific catalyst in the news window and win-probability sits at 60. Valuation at 529x PE prices in years of perfection and near-peak positioning caps upside. Convergence bar not met.

2026-06-21 79 -10

Per the explicit ARM learning, I'm not over-penalizing the PE — risk floor respected. But win_probability at 58 is below the 65 convergence bar for new buys, and the 11% 5-day rip raises mean-reversion risk near highs. Catalyst is real but not name-specific enough. Skip until win-prob confirms or it pulls back.

2026-06-20 89 +5

Per learnings, we should not over-penalize ARM on valuation when momentum and a named hyperscaler catalyst (Graviton5 GA, +25% perf for agentic workloads) are both lighting up. But win_probability of 58 fails the >=65 new_buy convergence bar — Win-Prob agent's mean-reversion concern after an 11% 5d run is reasonable. Conviction respects the strong forward setup; action respects the hard rule. Watch for pullback re-entry.

2026-06-19 84 -1

Per learnings I won't let the high PE alone kill it, but here win-probability is genuinely sub-60 and catalysts are second-derivative (others' chips using ARM IP), not ARM-specific. Misses the 65 win-prob convergence bar. Re-engage on a pullback or a direct royalty/earnings catalyst.

2026-06-18 85 +3

Per recent learnings I should cap risk-quality penalty on high-PE breakout AI-infra names, so I'm not zeroing this out. But ARM lacks a name-specific catalyst in today's tape and win_prob at 58 sits below the 65 convergence bar. Sharp 5d run raises mean-reversion risk. Skip for now, watch for confirmed pullback or named catalyst.

2026-06-17 82 -3

Per our learnings, don't over-penalize ARM valuation when breakout + named catalyst is in place — but here win-probability is only 58 and the move already happened. Doesn't clear convergence bar (win_prob<65). Watch for pullback or fresh catalyst.

2026-06-16 85 +2

Per the ARM learning, I'm not over-penalizing valuation — but the convergence bar requires both thesis>=60 AND win-prob>=65, and win-prob is only 58. Catalyst is indirect (AMD/Meta read-through), not name-specific. After an 11% 5-day pop without a direct headline, this is exactly the 'riding thematic tide near highs without specific catalyst' pattern that should de-rate, not chase.

2026-06-15 83 +26

Per ARM learning, I cap the valuation penalty given momentum+breakout+catalyst alignment, but win_probability of 58 fails the 65 convergence floor for new_buy. The 5-day 11% rip is overextended and catalyst is sector-derivative, not ARM-specific (unlike AMD's named Citi upgrade). Skip; revisit if win_prob improves on pullback.

2026-06-14 57 -20

Catalyst and momentum are loud but win-prob is below the 65 convergence bar and risk_quality is bottom-of-batch at 15. 398x PE leaves no margin. Pass on new buy.

2026-06-14 77 +18

Real catalyst with Amazon Graviton5 and strong momentum, but win probability sits below 50 and PE near 400x leaves zero margin. Does not clear the 60/65 convergence bar for new_buy. Skip.

2026-06-13 59 +43

Strong catalyst and momentum but win_probability 48 fails the 65 convergence threshold for new buy. Extreme valuation and risk_quality=15 make this a chase, not a setup. Skip per rule 3.

2026-06-12 16 -2

Fails convergence bar (win_prob 48 < 65). Extreme valuation, severe recent breakdown, and exacerbates chip concentration. No reason to initiate now.

2026-06-11 18 -6

Fails new_buy convergence bar on win_prob (48 vs 65 required). PE 398x leaves zero margin for error, risk_quality is bottom-tier at 15, and sector would become dangerously concentrated. Catalyst is real (Supermicro/Arm agentic AI) but cannot override broken risk/win-prob profile. Skip.

2026-06-10 24 +6

Fails convergence bar — win_probability=48 well below 65 floor. PE of 398x with risk_quality=15 is uninvestable at this level. Hard skip.

2026-06-09 18 -2

Watchlist — fails new_buy bar (win_prob 48 < 65). PE 398x with -12.8% 5d move and extreme volatility makes this a poor risk/reward entry despite the AI narrative. Sector already saturated with chip exposure. Hard skip.

2026-06-08 20 -68

Fails new-buy convergence. Extreme valuation with no margin for error, worst risk quality in batch, and severe recent selling. Catalyst mention doesn't override broken setup.

2026-06-07 88 -1

ARM at breakout point with 5.4% 5d surge, top performance percentile, and direct Barclays CPU/agentic-AI catalyst. Risk_quality=15 is brutal but per the ARM-specific learning, cap the risk penalty when thematic breakout conditions are met. Convergence bar cleared (thesis 73, win_prob 65). New buy, smaller starter size for the sizer to determine.

2026-06-06 89 +8

ARM hits the thematic breakout override pattern: direct named catalyst tying agentic AI to ARM growth, price at 52wk high, strong recent surge. Thesis 73 and win_prob 65 just meet the convergence bar. Cap risk_quality penalty per learnings — this is exactly the ARM setup we missed before. Small new buy warranted; sizing handled downstream.

2026-06-05 81 -6

Constructive setup near breakout but the thematic-breakout override requires 3+ named-catalyst articles and momentum>=70 — catalyst is only indirectly supportive (55) and momentum is 62. Win-prob just barely at the 65 bar. Without clean convergence, skip and revisit if catalyst score firms.

2026-06-04 87 +1

Strong thematic breakout candidate matching the ARM-miss pattern (named catalysts, momentum 94, range at high). Thesis 73 and win_prob 65 just meet the convergence bar. However, PE 415x and risk_quality 15 are extreme, and concentration is real. With several already-held chip winners (NVDA/AVGO/MRVL) capturing the same thesis, marginal add value is low. Skip; watch for pullback entry.

2026-06-03 86 -2

ARM is at 99% of 52wk range with momentum 95 and a 5.37% 5d surge — the thematic breakout pattern. Thesis 73 + win_prob 65 clears the convergence bar (60/65). Per learnings, cap the risk_quality penalty rather than letting valuation veto entry — that's the AMD/ARM lesson. New buy at constructive but not core conviction; size will be modest given risk_quality.

2026-06-02 88 0

ARM hits the thematic-breakout override criteria from learnings: named-catalyst article explicitly highlighting ARM leading the AI software surge, range_position at 99%, momentum=98. Thesis=73 and win_prob=65 sit right at convergence threshold. Per ARM-specific learning, cap risk_quality penalty rather than letting it veto. Small starter position warranted.

2026-06-01 88 +47

Thematic breakout setup with named catalyst (NVDA RTX Spark Arm-based) fits the ARM-pattern learning. Win-prob 65 just below 65 bar for new_buy. Applying thematic-breakout override would justify entry, but extreme valuation and sector concentration argue for the Sizer to handle this. Score it high enough to be a candidate but not auto-buy here.

2026-05-31 41 -10

Fails new-buy convergence on both axes — thesis 51 and win_prob 60 are below the 60/65 floor. PE 356 with risk_quality 15 near all-time highs is exactly the late-cycle setup we want to avoid. Skip.

2026-05-30 51 -20

Fails new_buy convergence test: win_probability 60 < 65 floor and thesis_pct 51 < 60 floor. Momentum is spectacular but neither the thesis nor win-prob agents are convinced, and PE 356 is uncomfortable even allowing for catalyst override. The MU/PANW winners had thesis conviction; ARM doesn't. Skip.

2026-05-29 71 +30

Direct agentic AI product headline (Metis security) is exactly the catalyst pattern we want, but thesis_pct=51 and win_prob=60 both miss the 60/65 convergence bar. PE 356 and extreme momentum suggest late entry. Skip - chase risk too high without thesis support.

2026-05-28 41 -31

Fails convergence bar decisively: thesis_pct 51 and win_prob 60 both below threshold. PE 356 is extreme, risk_quality 15 is the weakest in the batch. Catalysts are all indirect (ARM-based chips in others' deals). Momentum-only setup near ATH with stretched valuation — this is the late-stage chase that ends badly. Skip.

2026-05-27 72 +41

ARM has a direct product validation catalyst (NVIDIA Vera CPU with Olympus cores) and strong momentum, but thesis percentile is only 51 and win-prob 60 — fails the convergence bar (need both >=60/>=65) for new buy. Valuation extreme even by AI standards. Stay on watchlist.

2026-05-26 31 -27

Fails new-buy convergence bar decisively: thesis_pct=51 below 60 floor, win_prob=60 below 65 floor. Momentum is the only standout but PE 356 and lowest risk_quality in batch suggest most of the move is priced in. Indirect catalyst only. Hard skip.

2026-05-25 58 -10

ARM fails the convergence bar (thesis 51, win_prob 60 — both below threshold). Momentum is extreme but valuation is stretched and catalyst is indirect (ARM benefits tangentially from ASIC thesis). No held position to protect. Skip — would chase momentum at extreme multiples without thesis support.

2026-05-24 68 -6

ARM has a screaming catalyst (5x profit growth headline, ATH) and momentum, but win_probability at 44 and risk_quality at 15 are disqualifying for new_buy. Thesis only 51. The -8.5% 5-day pullback signals momentum deterioration despite the headline. Even with the ARM-specific learning adjustment, the convergence bar isn't met. Skip but keep on radar.

2026-05-23 74 +36

Named agentic AI partnership headline (ARM/Red Hat) is exactly the catalyst pattern our learnings flagged. However, win_probability=44 is far below the 65 convergence bar, and recent -8.46% 5d move and PE 243x indicate genuine deterioration, not just valuation worry. Catalyst alone doesn't override broken near-term setup. Skip but track.

2026-05-22 38 -15

ARM fails the new_buy convergence bar on both axes — thesis only 51, win-prob only 44. Despite near-highs and momentum scores, recent 5d down -8.46% shows the breakout is cracking. Extreme valuation, lowest risk-quality in batch. Skip cleanly.

2026-05-21 53 +21

Despite the headline-perfect agentic-AI catalyst, ARM fails the convergence bar badly: win_prob 44 and thesis 51 are both well below 60/65 thresholds. Recent -8.46% 5d move and extreme PE compound the risk. Catalyst alone doesn't override broken win-prob per core principle 5. Skip.

2026-05-20 32 +7

ARM fails convergence bar decisively: win_prob 44, thesis only 51, recent sharp selldown (-8.46% 5d), PE 243x, risk_quality 15. No direct catalyst. Three of four specialists are skeptical. Hard skip — would be quality-deteriorating add to an already concentrated chip book.

2026-05-19 25 -24

Watchlist name failing convergence test - thesis_pct only 51 and win_probability 44, well below 60/65 bar. Recent sharp selling and 243x PE with deteriorating momentum. Even with structural ARM ecosystem tailwinds, no reason to initiate here. Skip.

2026-05-18 49 +4

ARM fails the new_buy convergence bar on both thesis (51) and win_probability (44). Sharp recent selling, extreme valuation, and momentum deterioration all argue against entry despite the Huawei Armv9 catalyst. Skip.

2026-05-17 45 +3

Thesis and win_prob both ~50, no ARM-specific news, extreme valuation, and would add correlated sector risk. Fails new_buy convergence on both axes. Hard skip.

2026-05-16 42 -5

ARM fails the convergence bar: thesis 51 and win-prob 50 both mediocre, PE 250x, no ticker-specific catalyst, 11% off highs (not a breakout). Only sector tailwinds support it. Skip in favor of cleaner setups already held.

2026-05-15 47

ARM fails the new-buy convergence bar: thesis 51 and win-prob 50 both below thresholds, no breakout (11% below highs), extreme valuation with no specific catalyst. Generic sector tailwind isn't enough. Skip.

Supervisor: ATTENTION — the scoring pipeline has a non-blocking warning.