There are two schools of thought when it comes to comparables searches for transfer pricing analyses: Should you use publicly listed companies, which most likely means that the comparables will be regional in nature; or should you include privately held companies so that comparable searches are specific to the tested party’s jurisdiction?
Tax authorities like those in the US and Argentina expect publicly traded comparables—with good reason. The financial data is easily verifiable against audited financial statements, which are filed annually with regulatory authorities (such as the Securities and Exchange Commission in the US).This offers certainty when calculating the comps’ PLIs—a byproduct of protecting investors.
Another advantage? You are guaranteed that the financials are complete, and you can ensure that the PLIs do not contain non-operating items that can skew the arm’s length range.
Many other tax authorities, however, mandate or strongly prefer local comparables and they don’t take a stance on private vs public. In fact, many can’t force taxpayers to use public comps, as in many countries, there are simply not enough publicly traded companies to create comparable sets for any given function (distribution, contract manufacturing, contract R&D, management services, etc.). The tradeoff (in economics there are no free lunches) is that private company financial data is nowhere near as robust, and it’s much less reliable since even if financials are audited, they aren’t publicly submitted to regulatory bodies annually.
While I myself would prefer a world where we use publicly traded comparables as comps, the fact is that this ship sailed a long time ago. Most tax authorities fall into the school of local comps at all costs, regardless of whether this means a lack of reliable financial statements, less detail with fewer line items, or both.
Game Plan
So, when you have no choice but to use privately held companies as comps to get a local set, what do you do?
First, make use of the other tools that tax authorities have legislated in transfer pricing regulations. Use the statistical methods they require when constructing the arm’s length range—in many cases, an interquartile (IQ) range—but sometimes a variation thereof where the range points are not 25% and 75%. The whole point of the IQ range is to remove outliers from the arm’s length range. So, when we see a wacky PLI, like an 80% operating margin for a distributor that is privately held, let’s use the IQ range for its intended purpose.
Rather than bemoan the fact that the comparable data is poor and find some reason to reject the comp—let’s be realistic and call a spade a spade. Any comparable can be rejected through the art of transfer pricing, so—keep it in the set and let the IQ range calculation eliminate it. Take the opportunity to discuss (and justify) your approach in the body of the transfer pricing report, explaining that it makes more sense to accept a functionally comparable company and rely on the statistical method to remove it, rather than artificially rejecting it, which would be hypocritical.
If the tax authority doesn’t like this, perhaps they should move away from this pedantic notion that comps must be local at all costs.
