A benchmarking study is the evidence behind your transfer price. It searches financial databases for independent companies doing comparable things and produces an arm’s length range. Done properly it is the strongest defence you have. Done badly it is worse than nothing, because it documents an indefensible position in writing.
What a transfer pricing benchmarking study does
You cannot show a price is arm’s length by assertion. You show it by finding what independent companies in similar circumstances actually earn, and demonstrating your result sits within that range.
The output is usually an interquartile range — the middle 50% of comparable results, with the top and bottom quartiles discarded as outliers. If the tested party’s margin falls inside it, the position is supported.
How the search works
- Identify the tested party. Normally the less complex entity — the one performing routine functions without unique intangibles. You test the simple side because comparables exist for it.
- Choose the profit level indicator. Operating margin for distributors, full cost mark-up for service providers, return on assets where the activity is capital-intensive.
- Run the database search. Screen by activity code, geography and independence — excluding companies that are themselves part of groups, since their results are not independent.
- Apply quantitative screens. Remove dormant companies, those with missing data, and persistent loss-makers.
- Review manually. The step people skip. Read each remaining company and reject the ones that are not actually comparable, recording why.
- Compute the range and compare your tested party’s result to it.
Step five is where studies are won or lost. An automated search with no documented manual rejection is the clearest sign of a study nobody actually thought about, and it is the first thing an experienced reviewer looks for.
The UAE comparables problem
There is no meaningful public database of UAE private company financials. Most UAE companies do not file publicly available accounts, so a search restricted to the UAE returns almost nothing usable.
The practical answer is a regional or pan-European search, using markets where financial data is publicly filed, with adjustments and a documented explanation of why those comparables are appropriate. This is accepted practice, but it must be reasoned rather than assumed — and the reasoning belongs in the study.
Adjustments
Where comparables differ in ways that materially affect margins — working capital intensity being the common one — adjustments can improve comparability. Applied sparingly and documented, they strengthen a study. Applied to force a result, they draw attention to exactly the number you were trying to defend.
How often to refresh
Common practice is a full search every three years, with the financial data of the accepted comparables updated annually in between. That balances cost against reliability.
Refresh sooner if your business changes materially — a new function, a restructured supply chain, a shift in who bears risk. The study describes a set of facts, and when the facts move the study stops describing you.
Why studies fail on review
- The functional analysis does not match reality. Everything rests on it — see the arm’s length principle and methods. A study built on a mischaracterised entity fails regardless of search quality.
- No documented rejection reasons. A list of accepted comparables without the reasoning behind exclusions is not a defensible search.
- Wrong tested party. Testing the entrepreneur rather than the routine entity produces a meaningless range.
- Loss-makers excluded automatically. Independent companies do make losses. Blanket exclusion can bias the range upward, and reviewers know it.
- Stale data. A search run four years ago, never refreshed, against a business that has changed.
- Result at the edge. Consistently sitting at the extreme of your own range invites the question of whether the range was built to accommodate the answer.
Who actually needs one
Formally, businesses meeting the documentation thresholds — AED 3.15 billion group revenue or AED 200 million standalone — because the Local File requires the analysis.
Practically, any business with material related party transactions benefits from something. A full study is expensive and disproportionate for a small company. A short internal analysis — what the transactions are, how the price was set, and any market evidence you can point to — costs little and is far better than an unsupported number when the FTA asks. Our guide to what triggers an FTA audit covers where related party transactions sit in that risk picture.
Frequently asked questions
What is a transfer pricing benchmarking study?
A database search for independent companies performing comparable functions, producing an arm’s length range of financial results against which your related party pricing is tested.
What is the interquartile range?
The middle 50% of comparable results, discarding the top and bottom quartiles as outliers. A result within it is generally accepted without adjustment.
Can I use UAE-only comparables?
Rarely. UAE private company financials are not publicly available in usable volume, so regional or pan-European searches with documented justification are standard practice.
How often should a study be refreshed?
A full search every three years, with comparables’ financial data updated annually. Refresh sooner if your functions, risks or supply chain change materially.
Which entity should be tested?
The less complex one — the routine entity without unique intangibles — because reliable comparables exist for it.
Should loss-making comparables be excluded?
Not automatically. Independent companies do incur losses, and blanket exclusion biases the range. Exclusions need a documented reason.
Do small businesses need a full study?
Usually not. A short documented rationale for how related party prices were set is proportionate and far better than nothing.