Plain English
After screening a database down through quantitative filters and reading business descriptions, you end up with a short list of companies — usually somewhere between five and fifteen — whose financial results you actually use to build the arm's length range. These are the comparable companies. They do not need to be identical to the tested party, just similar enough on the factors that actually drive profitability: what they do, what risks they carry, and roughly what market they operate in.
Technical definition
The subset of candidate independent entities remaining after quantitative and qualitative screening under OECD Transfer Pricing Guidelines Chapter III, whose financial data is used to compute the profit level indicator or price metric forming the arm's length range against which the controlled transaction is tested.
Why it matters
The composition of this list is the single biggest driver of the resulting range, and therefore of whether an adjustment is proposed. Disputes over transfer pricing routinely turn on whether specific comparables should have been included or excluded.
How it works in practice
- 01Start from the screened shortlist produced by the database search.
- 02Read annual report and website descriptions for each candidate.
- 03Exclude companies with materially different functions, risks, or business models despite matching industry codes.
- 04Exclude companies with data quality issues: mergers, restructurings, extreme outliers.
- 05Confirm the remaining set is large enough to support a statistically meaningful range, typically at least five to eight companies.
Worked example
Narrowing a shortlist
A quantitative screen for contract R&D service providers returns 45 companies coded under the relevant industry classification. Reading descriptions eliminates 20 that are product companies mis-coded as service providers, 8 that are loss-making in more than two of three years, and 6 that underwent a merger distorting comparability. The remaining 11 form the final comparables set, with interquartile operating margins of 6.2% to 9.8%, used to test the tested party's 7.5% result.
Common mistakes
- Retaining a comparable purely because it fits the desired range.
- Failing to re-verify comparables in each annual refresh, allowing stale companies to remain despite changed business models.
- Including comparables with related-party transactions that were not properly screened out.
Audit red flags
- A final set with unusually tight clustering suggesting selective inclusion.
- Comparables that changed business substantially during the tested period without adjustment or exclusion.
- No documented reason for excluding companies that appeared in the initial screen.
Documentation & data
Documents to hold
- Final comparables list with company names, countries, and financial summaries.
- Individual justification for each inclusion where the match is not obvious.
- Exclusion log for companies dropped at the qualitative stage.
Data you need
- Multi-year segmented or company-level financial statements for each candidate.
- Business descriptions and, where available, annual reports.
- Related-party transaction disclosures to confirm independence.
Who owns this internally: The advisor or in-house economist conducting the benchmarking study, reviewed by group tax.
Jurisdiction notes
- European Union
- Tax authorities frequently re-run searches during audit and challenge specific inclusions or exclusions.
- United States
- Section 482 practice places heavy weight on functional similarity over industry coding alone.
Notes by role
CFOs & finance leaders
A thin or fragile comparables set is a leading indicator of controversy risk worth flagging before filing, not after an audit letter arrives.
Advisors & consultants
Keep a one-line justification for every company on the final list — it is the fastest thing to lose track of and the first thing challenged.
Frequently asked
- How many comparables are enough?
- There is no fixed OECD minimum, but most practitioners aim for at least five, and many tax authorities become uneasy below that.
- Can loss-making companies be comparables?
- Occasionally, if the loss is not structural and does not reflect a fundamentally different risk profile, but persistent losses are usually grounds for exclusion.
Sources & status
- Primary source
OECD Transfer Pricing Guidelines, Chapter III, paragraphs 3.1-3.46
OECD, 2022
- Our interpretation
Practical screening thresholds
This glossary, 2026
Reference material only, not advice on a specific fact pattern. Reviewed 2026-06-30.
Careers
How this shows up in the job
Explaining why a specific comparable was rejected is a classic case-study interview question in TP graduate recruitment.
Careers in transfer pricing