Question: Per TASKS-RESEARCH.md:79 (filed 2026-05-25 from the quantum-bottlenecks blind-test verdict [[2026-05-25-quantum-computing-physical-bottlenecks-blind-test-verdict]]), where does KEYS (Keysight Technologies) belong? The blind trace flagged it as a coverage gap — unanimous in the quantum trace as the Layer-3 control-electronics player (~45% quantum-grade T&M share), and a broad artificial-intelligence / automated-test-equipment (ATE) end-market beneficiary, but it is not in any thesis-aligned watchlist. Decide placement: ai-power-delivery (T&M for artificial-intelligence capex broadly), supply-chain-traces (quantum/semis test layer), or a new test-and-measurement cohort.

Verdict: Add KEYS to supply-chain-traces.json as the test-and-measurement / quantum-control-electronics layer — it sits naturally beside the test/packaging names already there (ATEYY = Advantest ATE, KLIC = Kulicke & Soffa, AMKR/ASX = OSAT, BESIY = hybrid bonding). Keep KEYS in monster-discoveries.json (it qualifies — $59B mcap, −6% from 52wk high, SMA50 well above prior base). Do NOT add to the ai-power-delivery perspective key_tickers — that frame is the 30%-parasitic-loss power-conversion chain (MPWR/VICR/IFNNY/VRT + capital equipment), and T&M instrumentation is a different layer that would dilute it. Do NOT spin a new T&M watchlist — single-name over-tagging, same anti-pattern flagged in the [[2026-05-26-bot-robostrategy-cef-verification-coverage-decision]] and [[2026-05-27-brun-boost-run-ai-infra-coverage-decision]] decisions. Do NOT build a quantum perspective — already deferred in the source verdict (cohort too cross-thesis with HBM/optical/ai-power). The watchlist edit is filed as a mechanical follow-up; this artifact is the decision, not the execution.

What we're asking

The 2026-05-25 quantum blind-test verdict confirmed the six-layer quantum-hardware physics chain (cryogenics → He-3 → control electronics → cryo test → ultra-pure materials → precision lasers) but found the investable surface narrow: five of six layers sit on names already in our coverage for non-quantum reasons, and the one quantum-pure name (OXINF) is LSE-primary with thin US ADR. KEYS was the unanimous name across all three trace agents — Layer-3 quantum control electronics (microwave AWGs; ~45% quantum-grade T&M share) — and was explicitly flagged as the lone control-electronics coverage gap.

The verdict's own follow-up #2 framed this as a ~15-min watchlist-placement decision with three candidate homes:

• ai-power-delivery (T&M for artificial-intelligence capex broadly)
• supply-chain-traces (the quantum/semis test layer)
• a new T&M cohort

What we found

1 — KEYS profile + current tape

deno task fetch KEYS (data_as_of 2026-05-27):

• Issuer: Keysight Technologies, Inc.
• Sector / Industry: Technology / Scientific & Technical Instruments
• Exchange: NYQ (NYSE)
• Market cap: $59.4B
• Segments: Communications Solutions Group + Electronic Industrial Solutions Group — electronic design + test software, instrumentation, systems and services across design / simulation / validation / manufacturing / optimization.
• Price: $347.65 (close 2026-05-27); 52wk range $152.85 — $370.18; −6.1% from 52wk high
• Trend: SMA20 $352.05, SMA50 $326.88 (price ~flat against SMA20, SMA50 rising and well above the prior base); RSI_14 38.9 (pulled back from the high, not overbought)
• 1y trajectory: roughly doubled off the $152.85 52wk low — a genuine multi-quarter uptrend, which is why it surfaced in monster-discoveries.

KEYS is a large-cap, near-52wk-high, uptrending name — not a speculative micro-cap. The coverage question is purely which thesis watchlist, not is it real / liquid.

2 — Where the quantum-trace peers actually sit today

Correction to the source verdict: the quantum verdict's adjacency table said "ENTG (ai-power-delivery)". That is stale — ENTG is in monster-discoveries only, not in any ai-power-delivery-linked watchlist. There is no ai-power-delivery.json watchlist file; the perspective carries a key_tickers array instead (see §3). This does not change the KEYS decision but is worth noting so the placement reasoning isn't anchored on a wrong premise.

3 — Candidate home evaluation

Candidate A — ai-power-delivery perspective key_tickers. The perspective's key_tickers are: MPWR, AOSL, VICR, IFNNY, NVTS, POWI, BESIY, $ON, STM, WOLF, VRT, ETN, AIXA.DE, VECO, AMAT, LRCX, ASML, MRAAY, INTC, TSM. This is the power-conversion / 30%-parasitic-loss chain (power-management semis + the capital equipment that builds them). T&M instrumentation (oscilloscopes, network/spectrum analyzers, microwave AWGs) is a different layer — it tests the devices, it is not in the power-delivery path. Adding KEYS here would blur the perspective's clean "watts-per-rack delivery loss" frame. Reject.

Candidate B — supply-chain-traces.json. Current 11 symbols: ALMU, KMT, CRS (tungsten), MP, TDY, WOLF (rare-earth/materials), BESIY, ATEYY, KLIC, AMKR, ASX (HBM packaging + test). This watchlist is explicitly "tickers found by our autonomous research agents across all traces" — KEYS is a trace-discovered name (the 2026-04-12 quantum blind trace), and the test/ATE layer is already represented here by ATEYY. KEYS as the T&M / quantum-control-electronics test-layer member fits the existing taxonomy cleanly. Accept.

Candidate C — new T&M cohort. A dedicated test-and-measurement.json would need a population to justify it. The only other obvious member would be ATEYY (already in supply-chain-traces) and arguably TER (Teradyne, not currently tracked). Spinning a watchlist for one-to-two names is the single-name over-tagging anti-pattern already rejected in the BOT and BRUN coverage decisions. If a T&M cohort accumulates 4+ names with a distinct editorial thesis later, revisit. Reject for now.

Secondary — optical-supply-chain.json? KEYS does provide T&M for optical / CPO validation (and FORM/COHR live there). But KEYS's trace-discovery role was quantum control electronics, and its end-markets are broad (comms, aerospace/defense, automotive, general electronics), not optical-specific. Tagging it into optical-supply-chain would over-claim an optical-pure-play role it doesn't have. Leave optical-supply-chain untouched. Reject.

4 — Monster-discoveries: keep

KEYS sits in monster-discoveries legitimately — it is a large-cap multi-quarter uptrend (doubled off the 52wk low, SMA50 rising above the prior base). Unlike BRUN/BOT (de-SPAC / fresh-IPO, no SMA200, fails the golden cross gate), KEYS is exactly the established-uptrend shape monster-discoveries is meant to hold. No reason to remove it. The supply-chain-traces add is additive — it gives KEYS a thesis lane alongside its discovery-bucket membership.

Verdict + reasoning

1. Add KEYS to supply-chain-traces.json as the test-and-measurement / quantum-control-electronics layer. It joins the existing test/packaging cohort (ATEYY ATE, KLIC bonders, AMKR/ASX OSAT, BESIY hybrid bonding) and is a genuine trace-discovered name. This is the smallest, most taxonomy-consistent home.
2. Keep KEYS in monster-discoveries.json — it qualifies on the established-uptrend mechanism (unlike de-SPAC/IPO names); the trace add is additive, not a move.
3. Do NOT add KEYS to the ai-power-delivery perspective key_tickers — T&M is a test layer, not part of the power-conversion-loss chain that perspective frames.
4. Do NOT spin a new T&M watchlist — single/double-name over-tagging; revisit only if a 4+-name T&M cohort with a distinct thesis accumulates.
5. Do NOT build a quantum perspective — already deferred in the source verdict; the cohort is too cross-thesis (HBM / optical / ai-power) to justify a separate editorial layer, and KEYS's quantum exposure is one segment among many.
6. Producer-only output. No perspective log entry, no candidate, no promotion without reviewer sign-off. Reviewer follow-through executed the mechanical watchlist add after Codex flagged that a closed placement decision with only an inline follow-up could drop the actual coverage change.

Follow-ups filed

• Execute the watchlist add: cd stonks && deno task watchlist:add supply-chain-traces KEYS, then deno task precompute supply-chain-traces. Done 2026-05-27 by reviewer follow-through; KEYS now appears in stonks/watchlists/supply-chain-traces.json, stonks/data/summaries/supply-chain-traces.json, and research/api-data/watchlist-feeds/supply-chain-traces.json.
• Stale-reference fix (optional, low priority): the 2026-05-25 quantum verdict's adjacency table lists "ENTG (ai-power-delivery)" — ENTG is actually in monster-discoveries only. Correct the table or note the drift if that artifact is ever revised. ~5 min.
• Watch trigger for a future T&M cohort: if TER (Teradyne) or another T&M/ATE name surfaces from a trace or capture, re-evaluate whether supply-chain-traces is still the right home or whether a dedicated T&M cohort (KEYS + ATEYY + TER) has reached critical mass. No action until a 3rd-4th name appears.

Sources

• TASKS-RESEARCH.md:79 — source backlog row scoping the coverage decision.
• research/investigations/2026-05-25-quantum-computing-physical-bottlenecks-blind-test-verdict.md — the blind-test verdict that surfaced KEYS as the unanimous Layer-3 control-electronics coverage gap and framed the placement question.
• research/projects/autonomous-research/traces/2026-04-12-quantum-computing-physical-bottlenecks-w.json — original 3-agent blind trace (KEYS unanimous).
• stonks/data/stocks/KEYS/info.json + stonks/data/stocks/KEYS/ohlc/3m.json — yfinance confirmation (NYSE, Scientific & Technical Instruments, $59.4B mcap, $347.65 close, −6.1% from 52wk high, SMA50 $326.88 rising, RSI 38.9).
• stonks/watchlists/supply-chain-traces.json — destination watchlist (11 trace-discovered names incl ATEYY/KLIC/AMKR/ASX/BESIY test+packaging layer).
• stonks/watchlists/monster-discoveries.json — current KEYS home (discovery bucket, 160 names).
• stonks/watchlists/optical-supply-chain.json — evaluated and rejected as KEYS home (FORM/COHR live there; KEYS is not optical-specific).
• research/perspectives/ACTIVE.json — ai-power-delivery key_tickers list, evaluated and rejected (power-conversion-chain frame, not test layer).
• research/investigations/2026-05-26-bot-robostrategy-cef-verification-coverage-decision.md + research/investigations/2026-05-27-brun-boost-run-ai-infra-coverage-decision.md — precedent for the "don't spin a single-name watchlist" anti-pattern.

Carrier notes

• This is a placement decision, deliberately emitted as investigation (not deep-dive) because the scope is "which thesis lane," not "full company financials." KEYS is an operating company, so a future deep-dive (if the user wants the full revenue-segment / quantum-T&M-share / ATE-cycle treatment) would need the Massive/filings financials gate per CLAUDE.md — that is out of scope here and not claimed.
• The ~45% quantum-grade T&M share figure originates in the blind trace and the source verdict; it is a trace-level / trade-press claim, not a KEYS-disclosed figure. Per [[capture-paraphrases-are-not-primary-sources]], if that number is ever carried into a thesis-shaping artifact it should be anchored to a primary KEYS disclosure or a named third-party estimate first. The placement decision does not depend on the exact share number — KEYS is a top-2 quantum-control-electronics T&M vendor regardless of whether the figure is 40% or 50%.
• KEYS's quantum exposure is a small slice of a $59B broad-instrumentation business. The trace surfaced it because the quantum control-electronics layer maps onto KEYS, but the investment case (if any) is the broad artificial-intelligence / ATE / comms instrumentation uptrend, not a quantum-pure bet. The supply-chain-traces placement reflects "trace-discovered test-layer name," not "quantum-pure play."