A transfer pricing benchmarking study compares the price, margin or interest rate in a transaction between group companies with what independent parties earn in comparable circumstances. It is the core evidence that a price is arm’s length under section 31 of the Income Tax Act, and it sits at the heart of every local file.
Key points
- A benchmarking study produces an arm’s length range, usually the interquartile range of comparable results.
- The quality of a study depends on the functional analysis and on transparent, repeatable search and screening steps.
- SARS uses the OECD Transfer Pricing Guidelines as its reference point for how comparables should be selected.
- Studies need regular updating, and a full refresh when the business changes.
When do you need a benchmarking study?
Any material cross-border transaction with a connected person should be supported by evidence that its pricing is arm’s length. Benchmarking is especially important for:
- Management fees and other intra-group services
- Distribution of products bought from group companies
- Contract or toll manufacturing for a related party
- Royalties for trademarks, technology or know-how
- Intra-group loans and guarantees
For groups that must prepare a local file, the benchmarking is a core component. For smaller groups below the documentation threshold, a proportionate benchmark for the largest transactions is still the best defence if SARS asks questions. See transfer pricing documentation requirements in South Africa for the thresholds.
How a benchmarking study works, step by step
The OECD Transfer Pricing Guidelines describe a typical comparability process in paragraph 3.4. In practice, a well-run study follows these steps.
1. Understand the transaction
The starting point is the functional analysis: what each party does, which risks it controls and bears, and which assets it uses. A South African distributor that only resells products has a very different profile from one that owns the brand locally and runs its own marketing.
2. Choose the method and the tested party
The OECD recognises five methods: the comparable uncontrolled price (CUP) method, the resale price method, the cost plus method, the transactional net margin method (TNMM) and the profit split method. Most benchmarking studies use the TNMM, testing the less complex party, often the South African entity, on a profit level indicator such as operating margin or return on total costs.
3. Search the databases
Comparable companies are identified in commercial databases using industry codes, keywords and financial filters. Because there are few listed independent companies in South Africa for many activities, searches often extend to other regions with similar markets, with the reasons documented.
4. Screen the results
Each potential comparable is reviewed and accepted or rejected for a recorded reason: for example, it is not independent, performs different functions, has persistent losses or lacks reliable financial data. A transparent screening matrix is what makes a study defensible when SARS reviews it.
5. Calculate the arm’s length range
The results of the accepted comparables, usually over several years, are used to calculate the range. If the tested party’s result falls within the interquartile range, the pricing is generally regarded as arm’s length. If it falls outside, an adjustment towards the median is normally required.
6. Document the conclusion
The final report records every step, so that someone else could repeat the search and reach the same conclusion. That is exactly what SARS will try to do in an audit.
What makes a study defensible?
- A functional analysis that matches the company’s actual conduct and agreements
- Clear reasons for the method and tested party chosen
- Search criteria and rejection reasons that are recorded, not just applied
- Financial data that is current and consistent with the period being tested
- A conclusion that ties back to the company’s actual results
Benchmarking intra-group loans
Loans need a different approach. The question is both whether the amount of debt is arm’s length and whether the interest rate is. SARS sets out its views in Interpretation Note 127, including credit rating approaches and the use of market data. Our guide to intra-group loans under Interpretation Note 127 explains the detail.
How often should a study be updated?
Comparables age quickly. The OECD accepts a full new search every three years if nothing material has changed, with the comparables’ financial data updated every year. We cover the detail in how often you should update a transfer pricing benchmarking study.
If your group needs a study prepared or reviewed, Impact prepares transfer pricing benchmarking studies for services, distribution, manufacturing, royalties and financing, written to stand up to SARS scrutiny.
Frequently asked questions
What is the interquartile range?
It is the range between the 25th and 75th percentiles of the comparables’ results. Using it narrows the range to the more reliable middle of the sample.
Can foreign comparables be used for a South African company?
They are commonly used where reliable local comparables are not available. The choice should be explained and, where needed, adjusted for differences in market conditions.
Is a benchmarking study the same as a transfer pricing policy?
No. A policy tells your team how to price transactions; a benchmarking study provides the evidence that the resulting prices are arm’s length. Most groups need both.
