Why is the interquartile range used instead of the full range?
Because the extremes of a comparable set usually say more about the individual company than about the transaction being tested. A search that survives honest screening still contains businesses with an unusually good year, a one-off cost buried in operating expenses, or a cost structure nobody noticed. Cutting to the middle half removes those tails without anyone having to argue company by company about which ones deserve to go, which is why authorities on both sides of a cross-border file tend to accept it as the working range. It is a discipline applied to the comparables, not a licence to discard survivors whose results are inconvenient.
Is being inside the interquartile range enough to be safe?
It is the result you want, but it is not the whole answer, because the range is only as good as the set it was computed from. An examiner who disagrees with the tested activity, the screens or the rejections recomputes the range and your position moves with it. There is also the question of pattern: a result that sits inside for one year and swings across the range from year to year invites the question of what is actually driving it. Treat falling inside as the conclusion of the analysis rather than a substitute for it, and keep the search that produced the range as carefully as the comparison itself.
Where in the range should my result sit?
Any point within the range is a defensible result on its own terms, and there is no rule that pushes a compliant taxpayer toward the middle. What matters more is that the point is explained by the facts rather than chosen for comfort. A tested activity bearing slightly more risk than most of the surviving companies has a reason to sit in the upper part; one performing a stripped-down version of their function has a reason to sit low. Where the two countries involved take different views on the point used for an adjustment, being comfortably inside rather than sitting on the lower edge is worth something in practice.
Do I test one year or an average of years against the range?
Whichever you choose, both halves of the comparison have to be built the same way. A multi-year average for the tested activity measured against a range computed from single-year comparable data compares two different things and the mismatch is easy for an examiner to spot. Averaging across a cycle is often the fairer picture where the activity is seasonal or the period included an unusual year, but then the comparable data has to be averaged over the same span. State the basis in the documentation, apply it to both sides, and keep it stable from year to year so the comparison remains meaningful over time.
What happens if my result falls below the range?
The result is compared, an adjustment is quantified, and the working is kept. The first step is to be sure the range itself is right, because a wrongly defined tested activity or an unsupported screen produces a range the result was never going to meet. If the range stands, the adjustment brings the tested result to a defined point within it, and which point is used has to be stated and then applied consistently rather than varied by year. The second country's treatment matters here too: an adjustment on one side without a corresponding change on the other leaves the same profit taxed twice, so raise both at once.
Can two countries calculate different ranges from the same comparables?
Yes, and it happens more often than the arithmetic suggests, because the quartiles can be computed by more than one statistical convention and different conventions land on different figures from an identical set. Add the ordinary sources of divergence, such as which period is included, how each company's accounts are adjusted to a common basis, and whether a loss year is kept, and two competent analyses of the same data produce two ranges. The practical response is to name the convention used in the documentation and be able to show the calculation step by step, so a disagreement is about method rather than about who typed what.
Do I get credit for all of the foreign tax I paid?
Only up to your own country's tax on that same income, and only for tax you were legally obliged to pay. Two consequences follow. Living somewhere that taxes you more heavily than your residence country does leaves an excess that becomes a carryover rather than a refund. And withholding suffered above the treaty rate is not creditable — the route back to that money is a refund claim in the country that took it. See claiming the credit.
Can I avoid capital gains tax on a foreign property?
Not by virtue of it being foreign — there is no exemption for that, and the "keep it offshore" advice you may have read is how people acquire penalties rather than savings. What genuinely reduces the gain is ordinary and legitimate: principal residence relief where the property qualifies and the designation is made correctly, a properly built cost base including acquisition costs and capital improvements, the timing of the disposition, the treaty rules for real property, and credit for the foreign tax paid. See principal residence and foreign property.
