Methodology
LeadTimePulse is an independent monthly synthesis of published electronics component lead-time data. One analyst, assisted by scripts and an LLM for collection and drafting, reviews every number and writes every judgment. This page is the full method — versioned, with a public corrections policy.
The status scale
Each category gets an editorial status score every month:
- 1 · Stable normal lead times, predictable supply.
- 2 · Watch early signals — rising indicators, selective stretch. Monitor.
- 3 · Alert stretching lead times or selective allocation. Act on exposed BOMs.
- 4 · Critical broad allocation or severe extension. Assume constraint in planning.
The score is the analyst's call, informed by — but not mechanically derived from — the data. It is published alongside the computed Pulse Index; when they disagree, that disagreement is information.
The Pulse Index (v1.0)
A computed 0–100 reading per category, in the spirit of the NY Fed's GSCPI: it measures how unusual current conditions are relative to that category's own history, so an 8-week jump in passives and an 8-week jump in memory are scored on their own terms.
- Level — z-score of the current lead-time range midpoint against the category's trailing history (minimum 12 months, winsorized at ±3σ).
- Momentum — z-score of the 3-month change in that midpoint.
- Trend consensus — mean of source trend readings (up = +1, flat = 0, down = −1).
composite = 0.5·level + 0.3·momentum + 0.2·consensus, mapped logistically to 0–100.
50 ≈ the category at its own historical norm. 65+ is elevated, 80+ severe pressure, below 30 unusually relaxed. Categories with under 12 months of history show no index. The market-wide Pulse is the weighted mean of category indices, weighted by fixed annual revenue shares (disclosed in the repo, revisited each January). Sub-components are published with every reading so the number is auditable.
Published values are frozen: the index is computed once at publication and never silently revised. Formula changes bump the version and are listed in the changelog below.
The Consensus Meter
Because we read multiple sources, we can publish something no single distributor can: how much they agree. High consensus — all sources point the same way (spread ≤ 1.5×). Mixed — partial agreement. Contested — sources point in opposite directions or ranges spread more than 2×. Contested categories deserve extra caution with any single quote you receive.
Sources
Most distributors stopped publishing lead-time guides after 2023; we synthesize the published survivors plus association, spot-market, and trade statistics — five methodologically distinct source types. We publish our synthesis and cite sources; we never republish a source's tables.
| Source | Cadence | Collection |
|---|---|---|
| Avnet Abacus Lead-Time Guide | monthly | Automated fetch + parse |
| ECIA Industry Pulse (public summary) | monthly | Automated fetch, analyst-reviewed extraction |
| Fusion Worldwide Greensheet | monthly | Automated fetch, analyst-reviewed extraction |
| Future Electronics Market Conditions Report | monthly | Automated fetch, analyst-reviewed extraction |
| Sourceability Lead Time Report | quarterly | Manual analyst input |
| TrendForce DRAM spot prices | daily | Automated fetch + parse |
| TTI Lead Time Trends | monthly | Automated fetch + parse |
| WSTS global semiconductor billings | monthly | Automated fetch + parse |
Policy and market events (the "why it's moving" layer) are curated monthly from primary sources — Federal Register documents, official notices (via CSET translations for Chinese ministry documents), and named analyst institutions. Every factual claim carries a dated link; interpretation is clearly the analyst's own.
Cadence, backfill & corrections
Published monthly, first week of the month. Months reconstructed after the fact from archived briefs and snapshots are flagged as backfilled (currently: December 2025, January 2026, February 2026, March 2026, April 2026, May 2026, June 2026, July 2026).
Corrections log: no corrections to date. Any correction to a published value will be listed here with the original value, the corrected value, and the reason.
Methodology changelog: v1.0 — initial publication.
Disclaimers
LeadTimePulse is informational only — not procurement, legal, trade-compliance, or investment advice. Lead times are multicausal; drivers we tag are context, not complete explanations. Data reflects published sources at the stated date and can be revised by those sources at any time. Verify anything load-bearing with your suppliers and counsel. LLM assistance is used for collection and drafting; a human analyst reviews every published number and sentence.