Cost Plus Method: Cost Base, Mark-Up and Common Mistakes (2026)
The cost plus method for Indian transfer pricing — building the cost base, what belongs in it, mark-up logic, the tested party profile, comparability issues and a worked example.
Cost plus prices a controlled transaction by taking the costs incurred in performing it and adding an arm’s-length mark-up. It is the natural method for the entity that performs a defined service or manufactures to specification — the routine, cost-plus end of the routine/entrepreneurial spectrum. In the OECD framework it appears as the “cost of services plus” method; in Indian law it sits within CPM (Comparable Profits Method) as one of the objective-measures-of-profitability approaches, with TNMM as its company-level benchmarking expression.
How cost plus works
- Define the cost base — the costs of performing the controlled function (the “relevant costs” definition is where most of the method lives).
- Benchmark the mark-up — the mark-up on that base, tested against independent comparables performing the same function (the PLI is mark-up-on-cost — the same arithmetic as OP/OC).
- Apply — price = cost base × (1 + arm’s-length mark-up).
The economics: the provider is compensated for its cost plus a routine return for the function it performs — no more (it keeps no entrepreneurial residual), no less (its cost is fully funded).
Building the cost base: what belongs in it
The cost base definition must be stated once and applied consistently to the tested party and every comparable. The standard construction:
| Component | In the base? | Note |
|---|---|---|
| Direct production / service delivery costs | Yes | Materials, direct labour, direct overheads |
| Employee costs (incl. ESIC/EPF, incentives) | Yes | The dominant component for service entities |
| Administrative and support overheads | Yes (allocated) | Allocated on a documented key; the allocation method itself is a comparability factor |
| Depreciation | Yes, consistently | Where the function employs owned assets |
| Finance costs | No | Finance costs are excluded from operating-cost bases under the standard operating definitions |
| Interest income / other non-operating income | No | Non-operating items distort the mark-up |
| Exceptional / one-off items | No (or documented and consistent) | See the extraordinary events playbook |
| Provision for warranties (where the risk is the provider’s) | Yes | A cost of performing the function |
Two rules that prevent most cost-base disputes:
- The base must match the PLI denominator. If the PLI is mark-up on operating costs, the base is operating costs — the same definition, the same exclusions, for every company in the pool. A mixed base (operating costs for the tested party, total costs for the comparables) is an automatic TPO finding.
- State the exclusions. “Finance costs excluded” must be written, not assumed. The exclusion list is part of the method documentation.
Mark-up logic and the tested party profile
The mark-up is not negotiated — it is measured against comparables that perform the same function under the same risks:
- The function must be the benchmarked function. A contract manufacturer’s mark-up is benchmarked against contract manufacturers; a service provider’s against service providers. The functional analysis drives this, not the invoice description.
- The mark-up reflects risk and assets. A routine service provider (captive, cost-plus economics, no inventory risk) benchmarks in a different band than a full-risk service firm carrying credit and delivery risk. The comparable pool is the operationalisation of that difference.
- The PLI is mark-up on cost — the same ratio as OP/OC in TNMM terminology. This is why cost plus and TNMM are often described as the same test at two levels of abstraction: cost plus at the transaction level (price = cost + mark-up), TNMM at the entity level (entity profit tested against a pool). See cost plus vs TNMM for the boundary.
When cost plus fits — and when it doesn’t
Fits:
- Contract manufacturing and toll manufacturing (cost of production + mark-up)
- Routine R&D and development services (cost of services + mark-up)
- Shared services and captive service provision (cost + mark-up, or the LVAS/10TD safe harbour where eligible)
- Service entities whose remuneration is contractually cost-based
Doesn’t fit:
- Distributors — their return is earned on revenue, not cost; the cost base under-measures what they actually do (use TNMM on OP/Sales or RPM).
- Entrepreneurial functions — where the value is in the residual (innovation, brand, market creation), cost plus systematically under-pays the function and the TPO (or the other side) will challenge it.
- Where cost data is unreliable — cost plus is only as clean as the cost data; shared costs allocated across many functions make the base fuzzy.
Worked example: contract software development
An Indian captive develops software for its overseas parent at cost plus. FY 2025-26:
| Item | ₹ lakh |
|---|---|
| Employee cost (incl. statutory) | 1,800 |
| Allocated support overheads | 240 |
| Infrastructure and other direct costs | 160 |
| Operating cost base | 2,200 |
Four comparable contract developers (NIC 62011, same function, same cost-based economics) screen to mark-ups of 14.2%, 19.8%, 27.4% and 33.1% on operating cost. The arm’s-length range (IQR) is 19.8%-27.4%.
If the captive’s effective mark-up is 25%: inside the range — done. If it is 12%: below the range — the adjustment question opens, and the defence is the functional analysis (does the tested party really perform what the pool performs?) plus the documented cost base.
Common mistakes
- A cost base that changes year to year without explanation — the definition is a method choice, not an annual election.
- Overhead allocation without a documented key — “allocated administration” is not a definition; the key (revenue, headcount, floor area) and its rationale are.
- Including finance costs in the base for the tested party and not for the pool — the exclusion-list failure.
- Benchmarking a cost-plus function against a pool that carries risk — full-risk service providers sit in a higher band; the mark-up then looks “too high” for a captive. The pool must mirror the risk position.
- Cost plus for a distributor — the wrong base for the function; the mark-up on cost of a distributor says nothing about its distribution return.
Documentation checklist
- The cost base definition — included items, excluded items, allocation keys.
- The function benchmarked — the FAR summary driving the pool.
- The comparables — screens, accept/reject, the same base for each.
- The mark-up range and the tested party’s result against it.
- Any adjustments — working capital, extraordinary items — with the formula and the rate.
Cost plus is the workhorse for the routine half of the Indian captive economy — and a cost base built with the discipline above is very hard for a TPO to move. The method selection framework covers when cost plus yields to TNMM, GMM or profit split.
Run the screens as a study, not a spreadsheet
Quartyl applies the method, PLI and screening steps above as a pipeline — and keeps a documented reason for every exclusion.
Related docs
Cost Plus vs TNMM: Gross Mark-Up vs Net Margin Testing
Cost Plus and TNMM both use the cost base, but they answer different questions. The mechanics of each, the risk profile of each, and the documentation that survives audit.
Read docHow to Choose a Transfer Pricing Method: The Decision Framework (2026)
A step-by-step framework for transfer pricing method selection in India — comparability first, data second, tested party logic third — with the decision tree and documentation of the choice.
Read doc