RPM vs. TNMM: Benchmarking a LATAM Pharmaceutical Distributor

In an earlier discussion, I criticized the abuse of the Resale Price Method (RPM) by the Panama tax authorities - Dirección General de Ingresos (DGI) - in the benchmarking of a very limited function wholesale distributor of petroleum (“Misapplication of the Resale Price Method: LATAM Petroleum Distributor”, February 23, 2021). A July 2021 litigation (Administrative Tax Court, Case No TAT-RF-066) revisits this controversy with the taxpayer abusing RPM for a high function distributor of pharmaceuticals.

The petroleum distributor case was plagued by a set of alleged comparables that performed more functions than the wholesale distribution affiliate. The recent litigation was similarly plagued by a questionable set of alleged comparables, but in this case the pharmaceutical distribution affiliate performed more functions than the third-party distributors of generic pharmaceuticals.

Table 1 summarizes the financials for the distribution affiliates. Sales were approximately $150 million each year. The gross profit margin (GPM) for 2013 was less than 10.6%, but it rose to just over 40.8% in 2014. Operating expenses (OpEx) exceeded gross profits for both years and averaged 45.3% over the two-year period, resulting in a negative operating profit margin (OPM). Table 1 presents the transfer pricing as the average of the two periods where gross profits represented only 25.7% of sales.

Table 1: 2013-2014 Average Financials for the Distribution Affiliate

The court record noted that the DGI objected to both the low GPM and the inclusion of certain intercompany expenses borne by the Panama affiliate. The taxpayer’s only defense of the intercompany pricing was a claim that third-party distributors of pharmaceuticals had lower GPMs than the GPM received by the distribution affiliate.

DGI argued that TNMM was a more reliable approach than RPM in this particular case. As in the court decision for the wholesale distributor of petroleum, this court decision redacted the names of the alleged comparable companies even though they were publicly traded companies. Three of the alleged comparables reported OPMs of less than 2%, while two of the alleged comparables reported much higher OPMs. DGI was willing to accept an OPM just over 4% based on its application of TNMM, which would imply a GPM near 50%.

Transfer pricing practitioners often rely on a selection of alleged comparables from two very different Standard Industrial Classification (SIC) codes:

• SIC 5047: wholesale distributors of medical, dental, and hospital equipment and supplies; and
• SIC 5122: wholesale distributors of drugs, drug proprietaries, and druggist sundries.

Table 2 presents key financial ratios derived from the average 2013-2015 income statements from a set of North American distributors in the broad life science sector. The companies were the six “healthcare distributors” put forward in a recent article by Andrew Hughes. Hughes notes:

The benchmark seeks to provide a range of companies distributing an array of healthcare products. It should focus on companies specializing in distribution, supply chain, and logistics functions, rather than vertically integrated healthcare providers that also develop and manufacture products. However, practitioners should keep in mind that this benchmark does not distinguish between different healthcare products, such as pharmaceuticals, devices, and consumables.

Table 2: 2013-2015 Key Financial Ratios for Select Life Science Distributors

Two companies had relatively high OPMs, which reflect the extensive functions performed by these two companies as well as certain intangible assets:

1. Henry Schein is a wholesale distributor of medical devices. Its OpEx to sales ratio was 21.0% while its GPM was 28.07%. As such, its OPM was just over 7%.
2. Patterson Companies is a wholesale distributor of dental supplies and animal health supplies. Its OpEx to sales ratio was 19.06% while its GPM was 26.41%. As such, its OPM exceeded 7.3%.

The other four companies report much lower GPMs, as their functions are more limited. Their OPMs are also relatively modest, which reflects the fact that they perform fewer functions.

Any purported use of RPM to determine the appropriate GPM must consider the functions performed by the distribution affiliate versus those of the alleged comparable companies. The three wholesale distributors of generic pharmaceuticals perform considerably less functions than the distribution affiliate.

Henry Schein and Patterson Companies performed more extensive functions, but still had an OpEx-to-sales ratio equal to less than half the distribution affiliate's. One reason for the very high operating OpEx-to-sales ratio for the distribution affiliate was the large amount of intercompany expenses that it incurred. DGI challenged these intercompany services, which would tend to reverse part of the reported operating losses.

One could still argue that the 25.7% GPM was too low, especially if the actual value-added expenses-to-sales ratio for the distribution affiliate was still 30% or more after removal of these intercompany expenses. DGI was on solid ground to propose an application of TNMM to determine the appropriate OPM.

We have argued, however, that the three wholesale distributors of generic pharmaceuticals with low OPMs are not appropriate comparables for benchmarking the tested distribution affiliate. A Berry ratio or "return to value-added expenses" would alternatively suggest a very high OPM for the distribution affiliate if based on the high Berry ratios of these three wholesale distributors of generic pharmaceuticals. This approach, however, would likely overstate the arm’s length OPM, as these three companies own valuable intangible assets.

Could a capital adjusted Berry approach using the value-added expenses-to-sales ratio of the distribution affiliate provide a reasonable estimate for the GPM under the arm’s length standard? We suggested earlier that the Berry ratios for Henry Schein and Patterson Companies were above 30% because they also owned valuable intangible assets. The Berry ratio for Owens & Minor, however, appears to be a means for estimating the routine return to distribution. If an analysis could determine the value-added expenses-to-sales ratio for this distribution affiliate, a capital adjusted "return to value-added expenses" approach using Owens & Minor could provide a plausible estimate of the routine return for this distribution affiliate.

The lesson learned from our review of this litigation is that any analysis must account for the underlying facts surrounding the functions and expenses incurred by the distribution affiliate. This would allow the analyst to better understand whether the usual financial ratios derived from third-party distributors provide insights into what would represent an arm’s length GPM. Given the wide disparity in terms of functions and assets for third-party healthcare distributors, any naïve application of either RPM or TNMM is likely to yield misleading results.

References

Harold McClure, "Misapplication of the Resale Price Method: LATAM Petroleum Distributor”, EdgarStat Blog, February 23, 2021.

"Resolución de Fondo n.° TAT-RF-066 de 9 de julio de 2021," tribunaltributario.gob.pa.

Andrew Hughes, “Transfer Pricing Benchmark: North American Healthcare Distribution”, Tax Notes International, May 3, 2021.