SENTINEL

proposes; never places a live order

prices: Yahoo Finance (yfinance), cross-checked vs Finnhub · scanned 04 Sep 08:46 Asia/Jerusalem (10 hours ago)

SCAN BLOCKED 2026-09-04T13:02:39+00:00 - PStarStateIncoherent: P* state is not exhaustive or disagrees with the rendered line - AVGO: state NOT_ASKED but a line was rendered: no price clears this while the news cap is active; the cap is re-e.
Showing data from the last successful scan (10 hours ago). These rankings are stale.
news screen self-check: 2 past false positive(s) resolved by the current rules; no category needs escalation. details
Nothing on this page changes until it finishes.

Top picks · 2 buys of 150 scanned · 1 strong

NVDAcomputeSTRONG BUY
228.45 → 326 (+43%) · est +4.1%/90d
NXPIsemiconductorsBUY
224.99 → 311 (+38%) · est +2.0%/90d
opportunity score 0.82: profitable toll collector, cheap valuation; held back by leveraged balance sheet: net debt/EBITDA 1.8x

Full scan

#TickerVerdictScorePriceUpsideValuationLayer
1NVDASTRONG BUY0.97228.45+43%CHEAPcompute
2NXPIBUY0.82224.99+38%CHEAPsemiconductors
3CRDOWAIT0.78164.17+69%CHEAPnetworking
4AVGOWAIT0.77357.16+47%CHEAPcompute
5STXWAIT0.70798.61+41%CHEAPmemory
6MRVLWAIT0.69208.83+36%FAIRcompute
7RMBSWAIT0.6784.34+75%FAIRsemiconductors
8AMDWAIT0.67456.16+35%FAIRcompute
9TSMWAIT0.64417.01+32%EXPENSIVEfoundry
10GOOGLWAIT0.59342.48+25%EXPENSIVEcloud
11VRTWAIT0.58268.83+26%EXPENSIVEpower
12NXTWAIT0.58 83.50+69%CHEAPsemiconductors
13AEISWAIT0.58274.30+56%CHEAPequipment
14ONTOWAIT0.57 252.46+54%FAIRequipment
15MPWRWAIT0.571,213.65+52%EXPENSIVEsemiconductors
16GEVWAIT0.57 941.84+31%EXPENSIVEpower
17QWAIT0.56118.47+49%FAIRsemiconductors
18CDNSWAIT0.56304.88+32%EXPENSIVEeda
19ORCLWAIT0.56154.04+57%EXPENSIVE (1.5x own history; FAIR vs peers)cloud
20MTSIWAIT0.56259.95+52%EXPENSIVEsemiconductors
21ALABWAIT0.55282.82+38%EXPENSIVEnetworking
22STMWAIT0.5551.28+46%FAIRsemiconductors
23ANETWAIT0.54191.44+26%EXPENSIVEnetworking
24ASMLWAIT0.521,646.19+31%EXPENSIVE (2.1x own history; FAIR vs peers)equipment
25KEYSWAIT0.51322.62+29%EXPENSIVEcomponents
26ADIWAIT0.51356.50+32%EXPENSIVE (1.8x own history; FAIR vs peers)semiconductors
27LINWAIT0.51482.19+13%FAIRmaterials
28AMATWAIT0.50435.91+47%EXPENSIVE (2.2x own history; FAIR vs peers)equipment
29TXNWAIT0.50253.83+28%EXPENSIVE (2.3x own history; FAIR vs peers)semiconductors
30SMTCWAIT0.50135.40+55%EXPENSIVE (1.7x own history; FAIR vs peers)semiconductors
31AMZNWAIT0.48 258.90+27%EXPENSIVEcloud
32CRUSWAIT0.47113.13+46%CHEAPsemiconductors
33NVMIWAIT0.47352.70+51%EXPENSIVE (1.6x own history; FAIR vs peers)equipment
34KLACWAIT0.47172.94+35%EXPENSIVE (2.9x own history; FAIR vs peers)equipment
35MSFTWAIT0.47510.12+12%EXPENSIVEcloud
36NVTWAIT0.44152.09+34%EXPENSIVE (2.0x own history; FAIR vs peers)power
37TERWAIT0.44338.59+32%EXPENSIVEequipment
38CSCOWAIT0.43108.61+27%EXPENSIVE (2.4x own history; FAIR vs peers)networking
39CIENWAIT0.42317.46+74%EXPENSIVEnetworking
40SIMOWAIT0.42235.70+57%EXPENSIVE (1.9x own history; FAIR vs peers)components
41LRCXWAIT0.40292.66+27%EXPENSIVE (3.0x own history; FAIR vs peers)equipment
42GFSWAIT0.38 44.53+71%CHEAPfoundry
43LSCCWAIT0.38113.63+45%EXPENSIVEsemiconductors
44SITMWAIT0.37574.47+50%EXPENSIVEsemiconductors
45ONWAIT0.3673.65+46%EXPENSIVE (2.0x own history; FAIR vs peers)semiconductors
46ARMWAIT0.36242.59+18%EXPENSIVEcompute
47FORMWAIT0.3496.76+43%EXPENSIVEsemiconductors
48DIODWAIT0.3489.94+81%EXPENSIVE (1.7x own history; FAIR vs peers)semiconductors
49KLICWAIT0.3378.82+35%EXPENSIVE (2.5x own history; FAIR vs peers)semiconductors
50SNPSWAIT0.33416.31+31%EXPENSIVEeda
51ETNWAIT0.32397.12+20%EXPENSIVE (2.6x own history; FAIR vs peers)power
52TSEMWAIT0.29206.16+53%EXPENSIVEsemiconductors
53IPGPWAIT0.2676.46+65%EXPENSIVEsemiconductors
54ACLSWAIT0.26110.37+46%EXPENSIVE (1.8x own history; FAIR vs peers)equipment
55ALGMWAIT0.23 35.56+54%FAIRsemiconductors
56POWIWAIT0.2350.54+53%EXPENSIVEsemiconductors
57PENGWAIT0.2248.63+53%CHEAPsemiconductors
58SNDKAVOID0.511,554.99+37%CHEAPcomponents
59CLSAVOID0.44 309.84+54%FAIRhardware
60METAAVOID0.34610.68+24%CHEAPapplication
61EMEAVOID0.30741.32+39%FAIRpower
62APHAVOID0.2982.07+17%EXPENSIVEhardware
63OLEDAVOID0.2982.64+40%FAIRhardware
64PLABAVOID0.2728.18+47%CHEAPequipment
65NOWAVOID0.26145.59-3%FAIRapplication
66AGXAVOID0.26411.77+62%EXPENSIVE (2.3x own history; FAIR vs peers)power
67SPXCAVOID0.25193.92+40%EXPENSIVE (1.6x own history; FAIR vs peers)thermal
68MYRGAVOID0.25284.10+45%EXPENSIVE (1.3x own history; FAIR vs peers)power
69FFIVAVOID0.24392.15+11%FAIRnetworking
70FSLRAVOID0.24207.42+33%EXPENSIVE (1.5x own history; FAIR vs peers)semiconductors
71HUBBAVOID0.24460.15+23%EXPENSIVE (1.5x own history; FAIR vs peers)hardware
72TTAVOID0.23445.19+18%EXPENSIVE (1.6x own history; FAIR vs peers)thermal
73JBLAVOID0.23305.92+44%EXPENSIVE (2.7x own history; FAIR vs peers)hardware
74TELAVOID0.23209.41+18%EXPENSIVE (1.5x own history; FAIR vs peers)components
75DGIIAVOID0.2270.28+22%FAIRnetworking
76CAMTAVOID0.22 138.90+34%FAIRequipment
77UMCAVOID0.2219.86-7%EXPENSIVEfoundry
78ARWAVOID0.21210.87+11%CHEAPcomponents
79FLEXAVOID0.20107.89+49%EXPENSIVE (2.7x own history; FAIR vs peers)hardware
80IBMAVOID0.19234.71+5%EXPENSIVE (2.1x own history; FAIR vs peers)cloud
81MCHPAVOID0.1973.11+49%EXPENSIVE (4.2x own history; FAIR vs peers)semiconductors
82PWRAVOID0.19620.06+24%EXPENSIVEpower
83EXTRAVOID0.1921.89+53%EXPENSIVE (1.5x own history; FAIR vs peers)networking
84OSISAVOID0.19206.62+35%CHEAPsemiconductors
85FTNTAVOID0.19156.36+4%EXPENSIVEhardware
86PLTRAVOID0.18182.53+5%EXPENSIVEapplication
87NTAPAVOID0.18185.38+2%EXPENSIVE (1.4x own history; FAIR vs peers)components
88MUAVOID0.18958.16+58%EXPENSIVEmemory
89FNAVOID0.18395.00+86%EXPENSIVE (1.6x own history; FAIR vs peers)networking
90LITEAVOID0.18847.37+36%EXPENSIVEnetworking
91PLXSAVOID0.18242.07+26%EXPENSIVEhardware
92PANWAVOID0.16331.94+16%EXPENSIVEhardware
93MXLAVOID0.1659.45+59%FAIRsemiconductors
94AMKRAVOID0.1646.94+63%EXPENSIVE (1.7x own history; FAIR vs peers)semiconductors
95CRMAVOID0.16 264.43+2%CHEAPapplication
96BEAVOID0.16235.55+17%EXPENSIVEpower
97GNRCAVOID0.16183.02+55%EXPENSIVE (1.7x own history; FAIR vs peers)power
98ICHRAVOID0.16 52.82+81%CHEAPsemiconductors
99ACMRAVOID0.1669.42+70%EXPENSIVE (1.4x own history; FAIR vs peers)equipment
100LOGIAVOID0.16101.71+5%FAIRhardware
101AXTIAVOID0.1456.20+63%EXPENSIVEsemiconductors
102MKSIAVOID0.14249.65+67%EXPENSIVE (2.3x own history; FAIR vs peers)equipment
103AAPLAVOID0.14328.21-1%EXPENSIVEhardware
104AMBAAVOID0.13 63.38+50%EXPENSIVEsemiconductors
105VIAVAVOID0.1333.63+83%CHEAPsemiconductors
106AAOIAVOID0.13 100.38+63%FAIRsemiconductors
107CBRSAVOID0.13190.44+53%EXPENSIVEsemiconductors
108QCOMAVOID0.12168.57+15%CHEAPcompute
109AAONAVOID0.1277.98+83%FAIRthermal
110ENTGAVOID0.12130.70+33%EXPENSIVE (2.2x own history; FAIR vs peers)equipment
111VECOAVOID0.1242.55+44%EXPENSIVE (4.8x own history; FAIR vs peers)equipment
112SYNAAVOID0.1297.68+33%CHEAPsemiconductors
113XNDUAVOID0.1210.02+229%UNMEASUREDhardware
114NBISAVOID0.11 210.63+36%UNMEASUREDcloud
115BHEAVOID0.1171.21+21%EXPENSIVEhardware
116DELLAVOID0.10516.39+8%EXPENSIVEservers
117HPQAVOID0.1031.93-8%EXPENSIVE (1.6x own history; FAIR vs peers)hardware
118CEGAVOID0.10 285.05+22%FAIRpower
119LASRAVOID0.0941.47+114%EXPENSIVEsemiconductors
120SMCIAVOID0.09 37.87+12%CHEAPservers
121LFUSAVOID0.09414.24+18%EXPENSIVEcomponents
122POWLAVOID0.08175.16+60%EXPENSIVE (8.6x own history; FAIR vs peers)power
123HPEAVOID0.0854.44+24%EXPENSIVEservers
124PAVOID0.0898.15+33%EXPENSIVEcomponents
125SANMAVOID0.07195.69+33%EXPENSIVE (2.4x own history; FAIR vs peers)hardware
126NRGAVOID0.07111.84+69%EXPENSIVE (2.9x own history; FAIR vs peers)power
127AZTAAVOID0.0731.18+13%EXPENSIVEequipment
128VSHAVOID0.0730.45+28%EXPENSIVE (12.0x own history; FAIR vs peers)hardware
129TTMIAVOID0.07115.39+86%EXPENSIVE (2.1x own history; FAIR vs peers)hardware
130UCTTAVOID0.0766.59+106%CHEAPsemiconductors
131VSTAVOID0.06144.22+51%CHEAPpower
132WDCAVOID0.06441.57+51%EXPENSIVEmemory
133JCIAVOID0.06142.23+15%EXPENSIVE (1.9x own history; FAIR vs peers)thermal
134QRVOAVOID0.06100.38-10%CHEAPsemiconductors
135INTCAVOID0.0691.67+25%EXPENSIVEcompute
136ENPHAVOID0.0636.31+30%CHEAPsemiconductors
137SLABAVOID0.05220.71+1%EXPENSIVEsemiconductors
138MODAVOID0.05186.63+66%EXPENSIVE (2.6x own history; FAIR vs peers)thermal
139SEDGAVOID0.0533.80+14%EXPENSIVEsemiconductors
140COHRAVOID0.04264.41+57%EXPENSIVE (2.6x own history; FAIR vs peers)networking
141ATKRAVOID0.0493.65-0%CHEAPpower
142COHUAVOID0.0445.98+54%FAIRequipment
143SWKSAVOID0.0471.50-4%EXPENSIVE (2.2x own history; FAIR vs peers)semiconductors
144MRCYAVOID0.0384.71+37%EXPENSIVEhardware
145APDAVOID0.03304.22+13%UNVERIFIEDmaterials
146CRWVAVOID0.0384.56+71%UNMEASUREDcloud
147SNOWAVOID0.02356.47-7%EXPENSIVEapplication
148AVTAVOID0.0290.82+17%EXPENSIVE (2.7x own history; FAIR vs peers)components
149TLNAVOID0.01 305.52+51%CHEAPpower
150NVTSAVOID0.0011.10+27%UNMEASUREDsemiconductors

Score (0-1) is a single OPPORTUNITY measure - quality relative to price - with BOTH thesis-fit AND actionability folded in: valuation, quality, growth, upside, reduced for disqualifiers (not a toll-collector, overbuild, value trap, falling estimates, governance/regulatory) and for actionability (expensive, premium to own history, chasing, no buy confirmation). The verdict is DERIVED from the score, so sorting by score always tracks the tiers - no BUY sits below a WAIT. A great company at the wrong price scores like an opportunity that isn't there. ⚠ marks a provisional score (own-history unverified). ⏳ a stale score; the red glyph a governance risk.

RANK FRAGILE: 3 name(s) rank outside their own 5-95% interval (FSLR, IPGP, QCOM) - the published rank is not robust to input noise in a dense band. NOT a path disagreement: verify with rank_intervals(magnitude=0), which must collapse every interval onto its point estimate

Falsification monitors · ESCALATED posture

capex-growth (demand)GREEN →
Hyperscaler capex GROWTH decelerating (TTM YoY rolling over = buildout losing conviction)
health 0.94 · measured, as-of 2026-06-30
SEC XBRL companyfacts (5/5 cohort; as-of 2026-06-30) [MEASURED]
cloud-growth (demand)GREEN →
Cloud-segment revenue growth decelerating below trend for 2+ quarters
health 1.00 · proxied, as-of 2026-03-31
SEC Financial Statement Data Sets, segment revenue (3 members); META excluded (not a cloud vendor - structural, not missing data); MSFT (Microsoft Cloud - PROXY: bundles M365 commercial, Azure, Dynamics and LinkedIn commercial. Azure dollars are NEVER disclosed separately, so no exact Azure series exists at any price.); as-of 2026-03-31 [PARTIAL]
memory-supply (demand)GREEN →
Memory OVERSUPPLY building (cohort inventory-days rising + gross margin falling) - PROXY for contract-vs-spot pricing. Samsung/SK Hynix (~70% of DRAM) are non-SEC filers, so oversupply there reaches this only once it compresses the US cohort's margins: CONFIRMATION, not early warning (2-3q lag with the 2-consecutive rule)
health 1.00 · proxied, as-of 2026-07-03
proxy for contract pricing: memory-cohort inventory days + gross margin; excludes Samsung and SK Hynix (largest DRAM makers, non-SEC filers); as-of 2026-07-03 [PROXIED]
this GREEN is driven by the cohort gross-margin peak alone - the same fact that collapsed the P/E of the memory names flagged CYCLE-PEAK below, so the two are one observation, not two (MU, WDC)
backlog (demand)GREEN →
Remaining-performance-obligation (RPO) coverage of revenue falling. NON-GATING: its deviation-from-trailing-median construction is one-sided for a monotonically rising series (never went negative in 34 replayed quarters), so it has not demonstrated it can fire - informational until it does
health 0.91 · partial, as-of 2026-06-30 untested
SEC XBRL companyfacts (4/5 cohort; as-of 2026-06-30) [PARTIAL]
supply-response (supply)GREEN →
Capacity being ADDED into the peak: capacity-owner cohort TTM capex/revenue vs its own history. The cohort is MU/INTC/STX/WDC - the intended foundries (TSM, UMC, GFS) file IFRS and carry no us-gaap capex concept, so there is NO foundry in it. This is the mechanism that ends a margin cycle - capex committed at the top arrives two years later into weaker demand - and it is observable YEARS before it reaches inventory or gross margin, which is what memory-supply reads. NON-GATING, and NOT demonstrated: the 2018-2026 replay is RED in 32 of 32 periods, so it has shown the opposite of discrimination (defect 32)
health 1.00 · partial, as-of 2026-07-03 untested
SEC XBRL companyfacts, TTM capex and revenue summed across 4 capacity owners; Samsung and SK Hynix (largest DRAM makers) and the private/non-SEC capacity additions do not file with the SEC; as-of 2026-07-03 [PARTIAL]
capex-financing (supply)RED →
Capex FINANCING conditions: buildout self-funded by OCF, or leaning on external debt/equity that could tighten - supply-side
health 0.49 · measured, as-of 2026-06-30 untested
SEC XBRL companyfacts (5/5 cohort; as-of 2026-06-30) [MEASURED]

falsification gate closed (ESCALATED posture): no demand monitor has confirmed RED; sentiment wobbles are not falsification; UNTESTED (never fired in the replay, GREEN not yet validated against an event): backlog, supply-response, capex-financing - and capex-financing GATES the escalator

Are the picks working?

Are the BUY / STRONG BUY picks beating VOO?
+6.1% vs VOO
STRONG BUY +6.1% (n=17) · BUY -4.3% (n=87) over closed 3mo horizons
NOT YET SIGNIFICANT - fewer than 30 observations per tier.
Do the rankings match which way prices moved?
+0.11 rank correlation - weak link
Higher score vs higher return across 1668 names. Tier order (STRONG BUY > BUY > WAIT > AVOID) does NOT hold right now.
Not yet significant.
TIER SIZES FLOAT WITH THE MARKET - STRONG BUY has held 2, 3 and 4 names inside one week - so these means are NOT comparable across anchors and are DISPLAY ONLY. They answer "is the bar in the right place", never "is there skill", and can never trigger a recalibration.
Every tier, excess return over VOO (a tier that ranks higher should return more):
tiernamesvs VOObeat VOO
STRONG BUY17 ANECDOTE - not a statistic+6.08%-
BUY87-4.25%-
WAIT678-2.66%-
AVOID886-4.21%-
Evidence so far: 90 daily record(s) since 2026-07-26, 150 names each (every name scored, not just the buys - otherwise the sample is survivorship-biased). Closed 3mo horizons: 0. A conclusion needs 4 of them and n≥30 per tier; until then these numbers are shown but not trusted, and no recalibration is proposed.

Not financial advice - you decide and place the trade. Re-verify live price + FX at execution.