Transfer pricing benchmarking is used to evaluate whether pricing between associated enterprises is consistent with the arm’s-length principle. It generally involves identifying comparable uncontrolled transactions or independent companies and comparing their prices, margins, or financial results with those of the tested transaction or entity.
Benchmarking may appear straightforward, but small errors in functional analysis, comparable selection, filters, financial data, or adjustment methodology can significantly affect the final arm’s-length result.
The OECD Transfer Pricing Guidelines describe comparability analysis as being at the heart of applying the arm’s-length principle. The analysis requires both an accurate understanding of the controlled transaction and a comparison with sufficiently comparable independent transactions or enterprises.
What Is Benchmarking in Transfer Pricing?
Transfer pricing benchmarking is the process of identifying independent market data that can be used to evaluate related-party pricing.
For example, if a company provides routine support services to an associated enterprise, benchmarking may be used to identify independent companies carrying out similar functions and determine the margins they earn.
The objective is not simply to find companies operating in the same industry. The selected comparables should have sufficiently similar economic characteristics to make the comparison meaningful.
- Functions performed
- Assets used
- Risks assumed
- Products or services
- Contractual terms
- Geographic markets
- Business strategies
- Market conditions
- Economic circumstances
Why Benchmarking Errors Occur
- Understanding the related-party transaction
- Conducting functional analysis
- Selecting the tested party
- Selecting the transfer pricing method
- Defining search criteria
- Identifying potential comparables
- Applying quantitative and qualitative filters
- Reviewing financial data
- Making appropriate adjustments
- Calculating an arm’s-length result
An error at any stage can affect the outcome.
1. Incomplete Functional Analysis
One of the most significant benchmarking errors occurs when the functions, assets and risks of the tested party are not properly analysed.
For example, two companies may both provide IT-related services, but one may own intellectual property, perform product development, bear market risk, employ highly specialised technical staff or undertake significant research activities. The other may only provide routine support services.
Treating these companies as comparable purely because both operate in the IT sector can distort the benchmarking result.
What Should Be Reviewed?
The functional analysis should clearly identify functions performed, assets employed, risks assumed, decision-making responsibilities, intellectual property involvement and contractual arrangements. The benchmarking search should then reflect these characteristics.
2. Selecting the Wrong Tested Party
The tested party is generally the entity or transaction to which the selected transfer pricing method is applied.
Problems can occur where the more complex entity is selected without examining whether reliable comparable data is available.
A tested party should normally be selected based on the facts of the transaction and the reliability of available information.
3. Using an Inappropriate Transfer Pricing Method
Benchmarking depends heavily on the transfer pricing method being applied.
- Comparable Uncontrolled Price Method
- Resale Price Method
- Cost Plus Method
- Transactional Net Margin Method
- Profit Split Method
The appropriate method depends on the nature of the transaction and available data.
4. Using Broad Industry Filters
A benchmarking search that starts with an overly broad industry classification may return companies that are commercially different from the tested party.
For example, a search for software companies could include software developers, software product owners, cloud-service providers, IT consulting firms, routine support-service providers and cybersecurity businesses.
Industry codes should therefore be treated as a starting point rather than the only comparability criterion.
5. Selecting Functionally Different Companies
A company may appear similar based on its business description but may not be functionally comparable after detailed review.
- Ownership of intangibles
- Research and development activities
- Brand ownership
- Manufacturing functions
- Distribution responsibilities
- Entrepreneurial risk
- Market-development functions
Qualitative review is therefore essential after the database search.
6. Ignoring Ownership of Intangibles
Companies owning valuable intellectual property may earn returns that differ significantly from routine service providers.
- Proprietary software
- Patents
- Trademarks
- Significant technology
- Recognised brands
- Unique business processes
Intangible ownership should therefore be reviewed carefully during comparable selection.
7. Not Reviewing Related-Party Transactions of Comparables
Independent comparable companies are generally expected to derive a significant portion of their activity from uncontrolled transactions.
If a potential comparable has substantial related-party transactions, its financial results may themselves be influenced by controlled pricing.
8. Applying Turnover Filters Without Analysis
Turnover filters can be useful, but applying them mechanically can also create errors.
A company with much larger operations may benefit from economies of scale, stronger bargaining power, broader customer base, greater market reach or specialised assets.
However, turnover alone does not automatically make a company incomparable.
9. Using Incorrect Financial Data
Benchmarking results depend on reliable financial information.
- Outdated financial statements are used
- Consolidated data is used instead of standalone data
- Exceptional items are included incorrectly
- Operating and non-operating income are mixed
- Segmental information is ignored
- Financial numbers are transcribed incorrectly
Even a small error can materially change the operating margin of a comparable.
10. Using Consolidated Data Instead of Relevant Segmental Data
A potential comparable may operate in several business segments.
Where reliable segmental data exists, it may provide a more meaningful comparison than the company’s overall margin.
11. Incorrect Treatment of Operating and Non-Operating Items
Profit-level indicators depend on correct classification of income and expenses.
- Interest income
- Investment income
- Exceptional gains
- Foreign-exchange items
- Asset-sale gains
- Financing costs
The operating/non-operating classification should therefore be applied consistently to both the tested party and comparables.
12. Ignoring Persistent Loss-Making Comparables
A company reporting losses should not automatically be excluded.
Persistent losses may indicate different business circumstances, commercial failure, start-up conditions, extraordinary events, functional differences or higher risk exposure.
13. Excluding High-Profit Companies Without Proper Reason
High profitability alone does not necessarily make a company incomparable.
- Valuable intangibles
- Unique services
- Extraordinary events
- Market leadership
- Business restructuring
- Functional differences
Exclusion should be based on objective comparability factors rather than the desired result.
14. Ignoring Extraordinary Events
A potential comparable may have experienced unusual events during the relevant period.
- Merger
- Demerger
- Acquisition
- Restructuring
- Major asset sale
- Business closure
- Exceptional litigation
- Significant change in business model
Such events can affect profitability and may reduce comparability.
15. Incorrect Working Capital Adjustments
Working capital differences can affect the profitability of businesses.
- Trade receivables
- Trade payables
- Inventory
- Credit terms
Where appropriate, working capital adjustments may be used to improve comparability, but the methodology should be documented and applied consistently.
16. Using Only Current-Year Data Without Reviewing Market Conditions
Single-year data may sometimes be sufficient, but reviewing multiple years can provide useful context regarding business cycles, unusual profitability, start-up periods, market disruptions and recurring losses.
Prior-year data should not automatically replace relevant current-period information.
17. Copying the Previous Year’s Benchmarking Study
- Business model
- Related-party transactions
- Functions
- Risks
- Assets
- Market conditions
- Comparable companies
- Financial data
- Search databases
A previous study can be a useful reference, but the benchmarking analysis should be reviewed for the current transaction and period.
18. Search Strategy Does Not Match the Transaction
A database search should be designed around the actual controlled transaction.
Search keywords, industry codes and qualitative criteria should reflect the actual services performed.
19. Accepting Database Results Without Manual Review
Databases provide potential comparables, not final comparables.
- Annual reports
- Company websites
- Business descriptions
- Notes to accounts
- Segment information
- Related-party disclosures
Automated search results may otherwise include companies with materially different activities.
20. Benchmarking Does Not Match Transfer Pricing Documentation
- The functional analysis describes one activity
- Form 3CEB reports another transaction description
- Financial statements classify the transaction differently
- The benchmarking search uses a different functional profile
These inconsistencies can raise questions about the reliability of the analysis.
How to Review a Transfer Pricing Benchmarking Study
- Step 1: Confirm the Controlled Transaction
- Step 2: Review the Functional Analysis
- Step 3: Confirm the Tested Party
- Step 4: Review the Selected Method
- Step 5: Review Database Search Criteria
- Step 6: Perform Qualitative Comparable Review
- Step 7: Verify Financial Calculations
- Step 8: Review Adjustments
- Step 9: Reconcile With Financial Statements
- Step 10: Reconcile With Transfer Pricing Reporting
Conclusion
Benchmarking errors in transfer pricing commonly arise from weak functional analysis, incorrect comparable selection, unsuitable filters, unreliable financial data, inconsistent adjustments and differences between benchmarking studies and underlying documentation.
A reliable review should start with the controlled transaction rather than the database search. The functions, assets and risks of the parties should first be understood, followed by selection of the tested party, method, comparables and appropriate adjustments.
The final benchmarking result should also reconcile with transfer pricing documentation, financial statements, intercompany agreements and applicable reporting. A well-documented review process can make the benchmarking analysis more consistent, transparent and supportable.
FAQs
Q: What is benchmarking in transfer pricing?
It compares related-party pricing with independent market data to assess whether the result follows the arm’s-length principle.
Q: What causes transfer pricing benchmarking errors?
Errors may arise from weak functional analysis, unsuitable comparables, incorrect filters, unreliable data or poor adjustments.
Q: What are working-capital adjustments in benchmarking?
They adjust for differences in receivables, payables, inventory and credit terms that may affect business profitability.
