How Many Comparables Does a Transfer Pricing Study Need?

There is no fixed number of comparables required for a transfer pricing benchmarking study. What matters is that the final set is genuinely comparable, independent and documented: quality over quantity. In practice a defensible study often ends with a final set in the single or low double digits, after screening thousands of companies.

Why Is the Number Not the Point?

A large set of loosely comparable companies is weaker than a small set of closely comparable ones. The interquartile range only needs enough reliable data points to be meaningful; adding poor comparables widens the range and undermines the conclusion. Reviewers look at how the set was built, not just how big it is.

How Does the Screening Funnel Work?

Stage	What happens	Typical effect
Starting universe	Search by region and industry code	Thousands of companies
Quantitative screens	Independence, size, data availability, persistent losses	Hundreds remain
Qualitative review	Manual reading of business descriptions and accounts	Dozens remain
Final comparable set	Closely matched independent companies	A small, defensible set

In the UAE the search follows the FTA’s order: UAE comparables first, then the GCC and wider Middle East, then further afield where needed. Because published financials for UAE private companies are limited, wider regions are often needed to reach a reliable set.

What Does a Strong Comparable Set Look Like?

• companies that match the tested party on functions, risks, industry and market;
• only independent companies, without significant related-party shareholdings;
• current, multi-year financial data, usually three years;
• a recorded reason for every acceptance and rejection.

When the set is thin, a multi-year average helps stabilise the range.

What Do the FTA and OECD Expect?

Following the OECD guidance the FTA relies on, the emphasis is on a transparent, replicable search and a documented rejection matrix, not on reaching a target count. A study that shows its working is far more defensible than one that simply lists comparables.

Expert tip: if your final set has fewer than five or six companies, explain why in the study: the search steps, the regions covered and why the remaining companies are reliable. A short, well-argued set is accepted more readily than a padded one.

Related guides: Transfer pricing benchmarking · Transfer pricing services · TNMM benchmark studies · Benchmarking tools and databases · How many comparables · Internal vs external comparables