Is a transfer pricing study required with a US return?
The framework does not ask you to attach a study to the return. It asks you to have one by the time the return is filed. Documentation prepared by the return date supports two things: that the method used is reasonable, and that the comparables it relies on are reliable. Prepared later, it is still an analysis, but it no longer supports the position as of the filing. The practical consequence is scheduling. A study that depends on the year's financial results cannot start the week before the return goes in, so the work has to be planned backwards from the return date.
Why does our file need to discuss methods we did not use?
Because section 482 tests the method as well as the number. The analysis is meant to explain why the method used is the most reliable one for this transaction, and that explanation only exists if the alternatives were considered and rejected for stated reasons. This is the section most files skip. They name a method, run a search and report a result. Since the authority can adjust to the method it regards as most reliable, a file that never engaged with the alternatives has nothing to say when that happens. The rejections are usually a short section, and they carry the argument.
Can the IRS change our method if our margin is in range?
Yes. Being inside a range computed under your own method answers the question your own file asked. The framework allows an adjustment to the method the authority regards as most reliable, and that can produce a different range and a different conclusion from the same facts. It is why the file has to justify the method as well as the number. The defence is documentary: a record of the transaction's functions and risks, the reasons the chosen method fits them, and the reasons the alternatives do not. Without that section, a method substitution is an argument started from behind.
Does it help that we never intended to shift profit?
Not directly. The framework tests the arm's-length result, not the taxpayer's intention, so evidence about motive does not answer the question being asked. Board minutes explaining the commercial logic of an arrangement are useful background, and they are not a substitute for an analysis of what unrelated parties would have done. This catches out groups whose pricing was set for genuine operational reasons: a simple recharge, a legacy formula, an arrangement that suited both sides at the time. The reasons may be entirely honest. The file still has to show that the result they produced sits within an arm's-length outcome.
We wrote our study after filing, is it worth anything?
It is worth having, and it does not do the job that documentation prepared by the return date does. Treat it as two exercises. First, get the analysis right, because the pricing argument has to be made whenever it is made. Second, fix the calendar, so the next year's file exists by its return date rather than after a notice. Where several years are open, the analysis usually carries across them on different financial data, so the extra work of documenting the year that is still ahead of its return date is small compared with what it supports.
How do we know our comparables are reliable enough?
Reliability is a property of the search rather than of the number it produces, so it has to be shown and not asserted. That means recording which screens were applied, which candidates were excluded and why, what financial data was used, and how differences between the candidates and the tested party were handled. A range with no visible method behind it invites the same scrutiny as no range at all. It is also the part of a file that ages: a search run several years ago describes those companies as they were then, so keeping the analysis current is part of keeping it reliable.
Do I need transfer pricing documentation?
If your company transacts with a related party in another country, in substance yes — the question is how much. Documentation is what shifts the burden: prepared before the filing deadline it evidences that your pricing was set on arm's length terms, and its absence is what turns a pricing adjustment into a penalty in several regimes. Volume of related-party dealings drives whether you need a local file, a master file, or a full benchmarking study. See do I need transfer pricing documentation.
What does a transfer pricing benchmarking study do?
It evidences that your related-party pricing sits within the range independent parties achieve. The work is comparison: identify companies or transactions genuinely similar in function, risk and assets, compute their margins, and show where your result falls against that range. Done before the filing deadline it supports the position; produced afterwards under audit it carries far less weight. See benchmarking study.
