The cost-plus method prices an intragroup transaction as the costs borne by the supplying entity plus a markup, tested against what independent parties earn for comparable activity. It is one of the five OECD transfer pricing methods, and the one most used for services, shared functions and contract manufacturing.
It applies where the supplying entity performs a routine function without owning valuable intangibles or bearing significant market risk. A shared services centre handling accounting and IT for the group is the standard case: it renders a service, it takes little risk, and its remuneration should reflect that.
Two components have to be defended, and the first is where most disputes start. The cost base must be defined explicitly: direct costs only or full costs, which overheads are included, whether pass-through costs bear a markup. Costs recharged without added value, such as a third-party invoice simply passed on, are generally excluded from the base. The markup is then benchmarked against comparable independent companies, and the resulting intercompany markup is usually a single-digit percentage for routine services.
The method fails in a recognisable way. The base is set once and never revisited, so it silently absorbs new costs. The rate stays fixed for years while the benchmark moves. Costs that should have been excluded creep in. Nobody checks the invoiced amount against the method the agreement actually describes.
That last point is a control problem rather than a doctrinal one. Phacet's invoice versus contract control agent checks intragroup invoices against the stated terms, and cost allocation covers how the base itself is spread across entities.
