Applying the PS method
Example 16 — Greenfield development split — and the CCA comparison
Annex II Ch.II · Example 16 · ¶83–85
OECD Guidelines 2022At a glance
Companies A, B and C jointly develop a new product greenfield — no pre-existing contributions. Each is responsible for one of three key components. The functional analysis concludes that the relative expenses incurred by each company in development directly proxy relative value contributed. How should the resulting profits (or losses) be allocated?
The parties
Company A
Develops & manufactures component 1joint dev
Company B
Develops & manufactures component 2joint dev
Company C
Develops & manufactures component 3What happens
- Companies A, B and C jointly agree to share the greenfield development of a new product.
- None of them brings existing intangibles or other contributions of value to the project.
- Each company is responsible for developing and manufacturing one of the three key components.
- Functional analysis concludes that relative expenses incurred by each company in component development directly correlate with relative value contributed.
Analysis
Profit split is the right method. Relative development spend across A, B and C is a reliable proxy for relative value contributed.
How should profits (or losses) be allocated between A, B and C?
OECD analysis · Split on relative development costs
Per Annex II Ch.II · Example 16 (¶84–85), relative development costs directly correlate with relative value contributed, so they are an appropriate splitter. The result is similar to an analogous CCA outcome (see ¶8.4) because parties with similar economic characteristics should receive similar expected returns regardless of whether the arrangement is formally termed a CCA. An equal split ignores the established correlation between cost and value.
Key takeaway
Splitting profits in a greenfield development based on relative development costs converges economically with an analogous CCA outcome. Form follows substance — labelling an arrangement "profit split" or "CCA" should not change the arm's length result.
Related: ¶2.165–2.171 (splitting factors) · ¶8.4 (CCA vs. profit split convergence)