When should I use a profit split instead of a margin method?
A one-sided method measures one party and treats the other as holding everything unusual about the arrangement. That only works when one side genuinely does the ordinary work. Where both sides contribute something the other could not buy in on the open market, there is no party whose margin can stand for the whole transaction, and testing either one in isolation produces an answer the other country will not accept. A profit split fits those facts, because it divides the combined result by reference to what each side actually put in. The test is not the relative size of the two entities but whether either can honestly be described as making only routine contributions.
What counts as a unique and valuable contribution?
The phrase does narrow work. It means a contribution not available from independent parties on the open market, so there is no outside price for it. Proprietary process know-how developed in-house, a workforce whose experience cannot simply be recruited, and risk that the entity both controls and funds all tend to qualify. Routine assembly, order processing, and a licence used under another party's instruction do not, however commercially important they feel to the people doing them. The distinction decides the method rather than merely describing the business: if only one side clears that bar, the other is measurable against outside comparables and a split is the wrong answer. Record the assessment before choosing the method.
How do I choose the factors that split the profit?
A splitting factor has to be something that actually drives the combined result, and it has to be measurable on both sides on the same basis. Development spending, headcount within a defined function, or the accumulated contribution of the relevant intangible are the usual candidates. Most of the work is definitional: the same cost categories in and out on each side, the same period, the same currency treatment, and a written note of what was excluded and why. A factor chosen because it produces a comfortable answer, rather than because it explains how the profit arises, is the first thing an examiner tests. Keep that reasoning contemporaneous with the year it governs.
Is a profit split allowed for a routine distributor?
Usually not, and the reason is structural rather than a matter of taste. A distributor that markets and delivers under the principal's direction, bearing little inventory or market risk, is exactly the party whose return can be measured against independent distributors. Where a measurable comparable set exists, a method built on it is preferred to one that divides the entire profit. Applying a split there hands the distributor a share of returns from intangibles it neither developed nor funded, which the counterpart country will resist. If the distributor has genuinely built and paid for local market value and carries the downside, that conclusion can change, but it has to be evidenced from conduct rather than asserted.
Will both countries accept the same profit split?
They can, but nothing makes it automatic. Each authority looks at the division from the position of its own taxpayer, and one identical calculation will read as generous on one side of the border and thin on the other. Disagreement usually appears in the inputs rather than the concept: whether a cost is operating or not, whether a loss-making period belongs in the pool, how each currency is translated. The practical protection is one agreed computation held by both entities, built on the same definitions and reconciled to each set of statutory accounts. Where the amounts justify it, the method can be put to the authorities in advance rather than defended afterwards.
Do I split the actual profit or forecast profit?
Both approaches exist and they answer different questions. Dividing the actual combined result allocates what really happened, including a loss, and is tested with hindsight. Dividing anticipated profit sets the shares at the outset from projections and then leaves them alone even if the outcome differs, because the parties fixed their expectations before the risk played out. Mixing the two is the common error: a group that sets shares from projections but recalculates whenever the outcome disappoints has insulated one party from a risk it is said to bear, which is neither variant. Decide which you are applying, say so in the documentation, and hold to it across years.
How do I actually stop being taxed twice?
In this order. Fix your residence under each country's own rules, and if both claim you, apply the treaty tie-breaker. Identify where each type of income is sourced. Read the article that covers that income type, because it decides who taxes and at what maximum rate. Then claim the relief on the residence-country return, with proof of the foreign tax. Most of the tax people lose to double taxation is lost at the last step, not the first. See how double taxation is relieved.
What is cross-border tax?
Cross-border tax is what applies when income, assets or people touch more than one tax system at once — someone living in one country and earning in another, a company selling or hiring abroad, a family holding property in a second country. The work is rarely one country's rules applied harder; it is reconciling two sets of rules and claiming the relief that stops the same income being taxed twice at full rates. See what we do.
