The hierarchy principle: Vietnamese data first, and why
Benchmarking in a transfer pricing file seeks independent companies performing comparable functions to build the arm’s length price or margin range for the enterprise’s related-party transactions. Vietnam’s related-party framework (currently Decree 255/2026/ND-CP from 1 July 2026, carrying over Decree 132/2020/ND-CP’s principles) sets a clear hierarchy: prefer comparables data from independent parties in Vietnam, because market conditions, costs and business cycles within one country are far more alike than foreign data. Foreign data — including regional Asian data — is used only where no domestic comparable can be found, and then with an explanation of why domestic data is unavailable plus market-difference adjustments where quantifiable.
This hierarchy has a major practical consequence: a benchmark built entirely on foreign company data, without proving a thorough domestic search, is discounted by Vietnam’s tax authority at the file review stage however good the statistics are. Conversely, a Vietnamese-data benchmark with fewer companies but rigorous screening usually persuades more. The search must therefore be documented as a process: screening criteria set in advance, company counts through each filter round, reasons for each rejection — this search log is part of the analysis file, not an optional appendix.
Usable sources of Vietnamese company data
Source one: commercial databases specialised for transfer pricing and company research — platforms aggregating Vietnamese companies’ financial statements from published and registration sources, filterable by industry code, revenue and asset scale, independence status, exporting multi-year financials for margin computation. This is the working source for most Vietnam benchmarks because of its structure and reproducibility. Source two: published financial statements of companies listed and registered for trading on HOSE, HNX and UPCoM via the State Securities Commission and exchange portals — high-quality audited data with notes on industry and related transactions, well suited to high-reliability analyses though narrower in industry coverage (mostly larger companies).
Source three: the National Business Registration Portal and statistical agency data — used to verify a candidate’s legal status, registered industries, capital and independence (for example removing a company that is in substance another group’s subsidiary), but not providing financial detail sufficient for margin computation. The three sources complement each other: commercial databases for the wide screening round, published statements for the verification round, and the registration portal for the independence check. Standard screens include: independence (no controlling related-party relationship), same or adjacent industry code, comparable functions per the activity description, sufficient years of financial data (multi-year averages are commonly used to smooth cycles), and removal of persistent abnormal loss-makers or companies with material related-party transactions where identifiable.
When no comparable Vietnamese company can be found
This is real in narrow Vietnamese industries: too few independent companies in the same function, or a market dominated by foreign groups themselves. The correctly sequenced response is: progressively relax screening criteria with documentation ( widen to adjacent industry codes, accept larger scale differences, use multi-year data) and record each step’s result; if still insufficient, move to regional or international data with a market comparability argument and difference adjustments where quantifiable; in parallel, consider switching to a pricing method less dependent on external comparables — for example the profit split where both transaction sides contribute unique intangibles, or an internal comparable where the enterprise has similar transactions with third parties, which is always the strongest evidence.
What to avoid is jumping straight to foreign data, or accepting a benchmark of two or three weak companies just to have a number: on audit, the authority runs its own analysis on Vietnamese data, and a weak benchmark loses on home ground immediately. Where the domestic search is genuinely thin, the best protection is full transparency of the search process in the file, presenting the options rejected and why, and choosing a prudent price point within the range (not the most favourable edge) so the result survives when the range shifts under a different dataset. In Vietnamese transfer pricing audit practice, process transparency often weighs no less than the final number itself.
Frequently asked questions
Where can independent Vietnamese comparables be found?
Three main sources: commercial databases aggregating Vietnamese companies’ financial statements for screening by industry, scale and independence; published financial statements of companies listed on HOSE, HNX and UPCoM; and the National Business Registration Portal to verify a candidate’s legal status, industries and independence.
Can foreign data be used?
Yes, but as the last resort: Vietnam’s transfer pricing framework prefers domestic independent comparables; foreign or regional data is used only after a documented thorough domestic search shows no comparable exists, with explanation and market-difference adjustments where quantifiable. Jumping straight to foreign data with no domestic search log is discounted on audit.
Which databases are reputable?
The commonly used groups are commercial databases specialised in Vietnamese and international company financials for transfer pricing, the official disclosure data of the securities market (HOSE, HNX, UPCoM) and national business registration data. Reputability means clear data provenance, reproducible search results and multi-year figures exportable for margin computation.
What if no comparable company can be found?
Relax screening criteria step by step and record each step; if still insufficient, move to regional or international data with a market comparability argument and difference adjustments; consider methods less dependent on external comparables, such as profit split, or an internal comparable if one exists. Document the whole search transparently in the file and choose a prudent point within the range.
This article has been quality-reviewed by the FLAW AI Legal Council