Transfer Pricing Essentials for Multinational Groups

Transfer pricing essentials for multinationals: the arm's length principle, documentation, methods, and how to manage audit risk across borders.

Transfer pricing is the most consequential tax discipline most multinational groups underestimate. The moment a business operates through more than one entity in more than one country, the prices charged on intercompany transactions, goods, services, financing, licences, become a live question for every tax authority involved. Get it right and the group's profits sit where the value is genuinely created. Get it wrong and you face double taxation, penalties, and years of disputes.

The stakes have risen sharply since the OECD's base erosion and profit shifting programme reshaped global norms. Tax administrations now share information, coordinate audits, and apply substance-over-form thinking with confidence. For groups of any real size, transfer pricing is no longer a compliance afterthought; it is a board-level risk that touches effective tax rate, cash flow, and reputation.

This guide sets out the essentials: the governing principle, the accepted methods, what documentation you must hold, and where audits most often go wrong.

The arm's length principle

At the heart of transfer pricing sits a single idea: transactions between related parties should be priced as if they were between independent parties dealing at arm's length. This is the standard adopted, in some form, by almost every developed tax system and codified in the OECD Transfer Pricing Guidelines, which most jurisdictions either follow or echo.

The principle sounds simple and is anything but. It requires you to identify what genuinely comparable independent parties would have agreed, then test your intercompany pricing against that benchmark. Where no perfect comparable exists, and they rarely do, you must apply judgement supported by economic analysis.

The practical consequence is that intercompany pricing cannot be set to suit a desired tax outcome. It must reflect the functions performed, the assets used, and the risks assumed by each entity, the so-called functional analysis. An entity that bears genuine risk and contributes valuable intangibles should earn a return commensurate with that contribution. A routine distributor or contract manufacturer should earn a routine, more modest, return.

The recognised methods

The OECD framework recognises five principal methods, and choosing the right one is half the battle. The comparable uncontrolled price method compares the price charged in a controlled transaction directly with the price in a comparable uncontrolled one; it is the most direct but demands close comparables. The resale price method and the cost plus method work from gross margins and suit distribution and manufacturing arrangements respectively.

The two transactional profit methods are now the workhorses in practice. The transactional net margin method tests a net profit indicator, such as operating margin or return on costs, against comparable independent companies, and is widely used because reliable net-margin comparables are easier to find. The profit split method allocates combined profit between related parties according to their relative contributions, and is the natural choice where both sides contribute unique and valuable intangibles, integrated trading operations being the classic example.

There is no universal "right" method. The guidance asks you to select the most appropriate method to the circumstances, supported by reasoning. Choosing a method and then reverse-engineering the facts to fit it is a frequent and dangerous error.

Documentation and the three-tiered approach

Most jurisdictions now expect a structured, three-tiered documentation package, again following the OECD model. The master file gives a high-level picture of the global group: its structure, its business, its intangibles, its intercompany financing, and its overall transfer pricing policies. The local file drills into the specific transactions of each local entity, with the detailed functional and economic analysis. Country-by-country reporting, generally required of the largest groups above a revenue threshold that is commonly set around EUR 750 million, gives tax authorities a jurisdiction-by-jurisdiction summary of revenue, profit, tax, and employees.

Thresholds, formats, filing deadlines, and language requirements vary considerably between countries, and they change. Some jurisdictions require contemporaneous documentation prepared by the time the return is filed; others allow it on request within a short window. As a general rule, treat documentation as something you prepare before the transaction is priced, not after a query lands.

Good documentation does more than discharge a filing duty. In many systems it is the difference between a defensible position and an exposed one, often shifting the burden of proof and providing protection against penalties even where an adjustment is ultimately made.

Where audits go wrong

Several patterns recur in disputes. The first is a mismatch between paperwork and reality, intercompany agreements describing one allocation of functions and risks while the actual conduct of the parties tells a different story. Tax authorities increasingly look at what people actually do, not only at what contracts say.

The second is intangibles. The OECD's framework on development, enhancement, maintenance, protection, and exploitation of intangibles means that legal ownership of a brand or patent is not enough to justify the return; the entity must actually perform and control the relevant functions. Routing royalties to a low-substance holding company is now among the most heavily scrutinised structures in international tax.

Intercompany financing is a third flashpoint. Interest rates on related-party loans, guarantee fees, and the very characterisation of an instrument as debt rather than equity all attract attention, and many jurisdictions layer thin-capitalisation and interest-limitation rules on top of the arm's length test.

The fourth is the loss-making local entity. A subsidiary that consistently reports losses while the group prospers invites the question of whether it is being correctly remunerated for the functions it performs. There may be a perfectly good commercial explanation, but you should expect to provide it.

Managing the risk proactively

Reactive transfer pricing is expensive. The groups that fare best build the discipline into their operating model. That means setting policies that reflect the real value chain, documenting them contemporaneously, and reviewing them as the business changes, after an acquisition, a new product line, or a shift in where key people sit.

For higher-value or higher-uncertainty positions, advance pricing agreements offer a route to certainty. By agreeing a methodology with one or more tax authorities in advance, a group can remove a transaction from audit risk for the agreement's term. They take time and resource to obtain, but for material recurring flows they are often worth it. Where double taxation does arise, the mutual agreement procedure under the relevant treaty provides a mechanism for the affected authorities to resolve it between themselves.

Above all, align substance with structure. Post-BEPS, the surest protection is a structure where profit follows people, functions, and genuine decision-making, and where the documentation simply records a commercial truth rather than constructing a convenient one.

How HPT helps

We help multinational groups design and defend transfer pricing positions that hold up under scrutiny: functional analyses, method selection, master and local file documentation, intercompany agreements that match operational reality, and support through audits, advance pricing agreements, and mutual agreement procedures. We coordinate across the jurisdictions where you operate so the policy is coherent rather than a patchwork.

If your group's intercompany pricing has outgrown its documentation, we would be glad to review your position and tell you, candidly, where the exposure lies.