Lab vertical integration (acquisitions and acquihires, trailing four quarters)
1What this measures
Distinct acquisitions or acquihires by AI labs and hyperscalers of application, tooling, data or infrastructure companies announced in the trailing four quarters, derived metric lab_vertical_integration_events_4q over the acquisition ledger (acq.<acquirer>.<target>.pt rows, each with a fetched announcement).
Why it matters. The application layer's exit bell. When the labs buy the layer above them, the surplus that layer was capturing moves down the stack; a rising count is the capture thesis's clearest leading signal.
- Proxy types
- capital_flow, market_structure
- Unit
- count
- Cadence
- quarterly
2How we track this
- series
acq.*.pt - derived metric
lab_vertical_integration_events_4q(formula in the semantic layer)
Direction over 4 periods, dead-band 1; higher = concentrating.
Four quarters; one event is not a trend. More deals per year concentrate the application layer into the labs.
Applied to metric:lab_vertical_integration_events_4q.
3Tracker interpretation
Fifteen deals in the ledger, from Meta's $14.3B Scale stake and Google's $2.4B Windsurf licence to Nvidia's $12.9B Hugging Face purchase and Stripe's $7B-plus OpenRouter deal; the pace roughly doubled in 2026.
4Evidence
Direction over the last 4 periods; the band is ±1 around the window's first point. Higher reads concentrating.
16 observations. Hollow points are disputed (see counterevidence). Every point links to its observation.
Derived rows (6)
| as of | dims | value | inputs |
|---|---|---|---|
| 2026-09-30 | 10 | obs:07e3555cobs:129c29f7obs:323f4b23+8 more | |
| 2026-06-30 | 10 | obs:07e3555cobs:129c29f7obs:323f4b23+7 more | |
| 2026-03-31 | 8 | obs:10c4c0b2obs:129c29f7obs:4217fe77+6 more | |
| 2025-12-31 | 5 | obs:10c4c0b2obs:129c29f7obs:4217fe77+3 more | |
| 2025-09-30 | 4 | obs:10c4c0b2obs:4217fe77obs:8646dd26+2 more | |
| 2025-06-30 | 1 | obs:10c4c0b2obs:87066654 |
5Status and reasoning
Evaluator: 10 (CNBC, 2026-09-30) reads concentrating over 4 periods with a dead band of 1. Auto-reason; rule rationale: Four quarters; one event is not a trend. More deals per year concentrate the application layer into the labs.
6Timeline notes
7Counterevidence
What cuts against this reading
The ledger holds what the press and filings announce; undisclosed prices count as events; Stripe is a payments company, in the ledger because OpenRouter is a routing layer.
8Update history
- 2026-09-10unmeasured to concentratingconf — → 55 · evaluate
Evaluator: 10 (CNBC, 2026-09-30) reads concentrating over 4 periods with a dead band of 1. Auto-reason; rule rationale: Four quarters; one event is not a trend. More deals per year concentrate the application layer into the labs.
9Confidence
55 / 95 — mixed or hard to operationalise
Confidence is independent of status: 90–95 multiple strong independent sources; 70–89 good evidence, some ambiguity; 50–69 mixed or hard to operationalise; below 50 limited or vague.