Financebotresearch desk研究台

Playbook › Themes

Hardware, networking, storage, components & electrical infrastructure

fast 2026-08-05

Claim

Five screens covering enterprise hardware, IT distribution, connectors/passives, power & electrical infrastructure, and data infrastructure/observability. The verdict was consistent and explicit: "the moats are narrower and the FCF profiles less compelling" than the software names screened the same week. See [[pattern-value-depth-varies-by-software-layer]].

Names of interest

Name Case at the time
NTAP 12.1x fwd, storage infrastructure, 2.1% yield
CDW 10.6x fwd, IT-distribution picks-and-shovels — but FCF declining
ZBRA 10.2x fwd, 43% off high, turnaround — later re-surfaced as a July survivor
TDC Deep-value data infra with FCF yield + buyback support
DDOG / ESTC Quality observability; entry targets $80–90 / $45–50

Sub-areas explicitly covered so a re-screen doesn't repeat them: enterprise storage/data management · IT distribution · warehouse/industrial automation hardware · enterprise networking beyond data-center · connectors & passives · grid/transformers/ switchgear/UPS · observability & data integration.

Rejected

PD — value trap (decelerating growth, extreme SBC, unsustainable buyback). MDB — best developer moat available, but SBC exceeds FCF.

⚠️ Staleness warning

Every figure here predates both the April recovery and the entire July AI-capex rout ([[regime]] Act II). ZBRA in particular has moved materially and re-surfaced in July at conviction 7.0. Treat the coverage as current and every number as void.

What would falsify this

A hardware-led selloff, or the March-cohort retrospective showing these names beat the software picks — which would invert the layer ordering this note rests on.

History

  • 2026-03-19 → 03-21 — five screens.
  • 2026-07-30 — recorded; flagged as the oldest un-refreshed field in the repo.

Partial refresh, 2026-08-05 — the IT distribution channel is NOT a correlated block

CDW fell −9.0% on 8/05 (as much as −22.5% pre-open) on a Q2 that beat both lines (adj EPS $2.91 vs $2.79, revenue $6.57B vs $6.2B, +10% YoY) but showed a profitability squeeze alongside an executive departure.

The obvious hypothesis was a channel-wide margin squeeze — component/memory cost inflation passing through every value-added reseller at once. It was tested and it failed:

Ticker Price 8/05 52wk high Off high
NSIT (Insight) $140.28 $141.91 −1.1%
SNX (TD SYNNEX) $263.07 19.0x ttm / 12.3x fwd, not repriced

Insight closed 1.1% off its 52-week high on the same session CDW dropped 22% pre-open. Whatever hit CDW did not hit the channel. Treat CDW's margin problem as company-specific until a second distributor confirms it.

Method note: this is a clean ring-1 falsification — "sells to/through the epicentre" is a hypothesis about shared economics, and distributors of the same goods can have genuinely uncorrelated outcomes when the cause is mix or management rather than input cost. Do not assume the channel moves together; check the second name before spending effort on the ring.

Rest of the field remains stale from March — this refresh covers IT distribution only.