An auditor’s query on intercompany pricing usually arrives in a recognisable form: a request for the basis on which a management fee, a royalty or an intragroup interest margin was set, followed by an observation that the supporting evidence appears thin. It is not an assessment. It is not a correction. It does not, by itself, create a liability. What it does create is a written record that the group’s own reviewers considered a position uncertain, and that record tends to outlive the audit cycle in which it was made. The period between the flag and the sign-off is the last stretch in which the group still controls the sequence of events, and it should be used accordingly.
What the flag is, and what it is not
The auditor is not applying tax law. The auditor is testing whether the financial statements are fairly stated, which includes whether an uncertain tax position requires recognition, measurement or disclosure. The question is therefore probabilistic and accounting-led: how likely is it that a tax authority, with full knowledge of the facts, would disturb the position, and what would that cost. A tax inspection is a different instrument entirely, with statutory information powers, a defined burden of proof, penalty exposure and an appeal route. The auditor has none of those. What the auditor does have is the ability to withhold or qualify an opinion, and a working paper file that may later be read by someone else.
Two further points are frequently missed. First, the flag often originates with a component auditor in a subsidiary jurisdiction rather than with the group auditor, which means the concern may be framed by local practice and may not match how the group sees the same transaction. Second, an audit flag typically surfaces well before an inspection would. That is an advantage, provided it is treated as one.
The first task is an inventory, not an answer
The instinct to reply immediately with a justification should be resisted. Before any position is defended, the group needs a complete list of related-party transactions across the open years: goods and services flows, headquarters and management services, licences and other intellectual property charges, financing in all its forms including loans, guarantees, cash pooling and deferred trade balances, cost contribution arrangements, secondments and shared personnel, and the use of tangible assets. It also means the transactions nobody ever papered, which are usually the ones that generated the query.
For each item, record the counterparties and their jurisdictions, the quantum per year, the mechanism by which the price was set, whether a written agreement exists, and, critically, whether conduct matches that agreement. The most common failure is not an absent contract; it is a contract describing a division of functions and risks that the business stopped following some years ago.
Whether a documentary basis exists at all
Dutch law imposes the arm’s length principle and an associated documentation obligation without any turnover threshold, so the absence of a group-level filing requirement is not the absence of an obligation. The Master File and Local File requirement attaches from fifty million euro in consolidated revenue, and country-by-country reporting from seven hundred and fifty million, but a group below both thresholds must still be able to explain how its intercompany prices were arrived at. The scope and mechanics are set out in our note on the article 8b documentation requirement.
The assessment should distinguish between four states, because they carry different consequences: no analysis at all; an analysis that exists but has not been refreshed since the business changed; an analysis describing a policy the operating entities never implemented; and an analysis prepared to another jurisdiction’s specification that does not address the Dutch position. Only the first is a pure gap. The others are, in varying degrees, evidence against the group’s own case.
Exposure is measured jurisdiction by jurisdiction
There is no such thing as a group-level transfer pricing exposure. A primary adjustment in one country produces a deduction that must be defended, or surrendered, in another. The Dutch corporate income tax rate in the upper bracket is 25.8 per cent, with a reduced rate in the first bracket, but the headline rate is rarely the relevant figure. What matters is the net of the primary adjustment, any corresponding adjustment the counterpart jurisdiction is willing to grant, and the treatment of cash that has already moved.
Several secondary consequences deserve modelling before any position is taken. A recharacterised charge may be treated as a deemed distribution, engaging dividend withholding tax at the general rate of 15 per cent subject to treaty reductions and EU exemptions, all of which remain conditioned on anti-abuse tests. Interest and royalty flows towards low-taxed or listed jurisdictions have attracted a conditional withholding tax since 2021, and a pricing correction can move a payment into or out of that perimeter; we set out the interaction in our note on Dutch withholding taxes. Restating an intragroup interest margin also changes the base for the earnings stripping limitation, whose parameters have been amended more than once. For groups within the seven hundred and fifty million euro perimeter, an adjustment shifts covered taxes and income between jurisdictions for Pillar Two purposes, so a correction that is neutral in group terms may not be neutral once the fifteen per cent minimum is computed. Penalty regimes and burden-of-proof rules differ as well; in several jurisdictions the absence of documentation affects how the burden is allocated.
Correcting forward, correcting back, or both
Correcting prospectively is the lighter option: the policy is reset, the agreements are rewritten to match how the business actually operates, and the new pricing applies from a defined date. It stops the exposure accruing. It does not close the open years, and it carries a presentational risk, because a visible change of policy invites the question of what was wrong with the previous one.
Correcting retrospectively is a board decision rather than a technical one. The inputs are which years remain open to assessment in each jurisdiction, whether a voluntary disclosure or equivalent procedure is available and what relief it offers, whether the counterpart jurisdiction will grant a corresponding adjustment and through what route, whether an advance pricing agreement or mutual agreement procedure is realistically available, and how any true-up will be characterised. That last point matters: a balancing payment between group companies is an invoice, a capital contribution or a distribution, and each carries its own consequences. Where a cross-border arrangement is restructured, the DAC6 hallmarks should be tested, since the reporting obligation falls on the intermediary or, failing that, on the taxpayer.
What must not be done
The single action that converts a manageable problem into an unmanageable one is the creation of documentation bearing a date it does not deserve. That includes benchmarking studies commissioned now and presented as contemporaneous, agreements signed today and dated to a prior year, and board minutes recording deliberations that did not occur. It includes instructing an adviser to prepare a file as at a historic date. The exposure then ceases to be a tax adjustment and becomes a question of the integrity of the records, with consequences for directors, for the audit relationship and, in some jurisdictions, criminal ones.
A transfer pricing file written after the question has been asked is worth less than no file at all: it converts a documentation gap into a credibility problem, and credibility is the one asset the group cannot rebuild inside the audit cycle.
Three lesser errors are common. Adjusting unilaterally in one jurisdiction, leaving the group asymmetric. Allowing the operating business to change its invoicing mid-review without recording why. And giving the auditor a firm conclusion the group cannot support if it is quoted back later. Material prepared for the auditor is generally not protected from disclosure, and should be written on that assumption.
The question underneath is usually substance
Pricing follows functions, assets and risks. A challenge to a management fee is almost always a challenge to whether the entity charging it has the people, the authority and the decision-making to perform the service. A challenge to a financing margin is a challenge to whether the lender controls the risk it is paid to bear. Remediating the paperwork without remediating the underlying facts produces a better-drafted version of the same weakness, which is why the review should run alongside an honest assessment of the group’s Dutch substance position. It also determines whether the ruling route is open: Dutch practice since July 2019 requires genuine economic nexus, and no advance certainty is given where the decisive motive is tax saving or where listed jurisdictions are involved.
Closing the cycle properly
The deliverables are modest in volume and specific in nature: a complete transaction inventory, a gap analysis against the documentation obligation, a short position paper for each material flow setting out the functional analysis and the pricing rationale, revised agreements that match conduct, and a remediation plan with named owners and dates. Above all, a recorded decision on whether prior years will be corrected, together with the reasoning. That record is what an inspector, a successor auditor or an acquirer’s diligence team will eventually read, and it is worth more than any single technical argument inside it.
Montclare structures and operates Dutch and cross-border platforms for international groups. Our services are set out on our services page.
This article is informational and does not constitute tax, legal or investment advice. Each engagement is subject to scope and applicable regulation.