Abuse of TNMM in Benchmarking of LATAM Electronics Distributor

Panama’s Administrative Tax Court ruled in favor of the tax authority Dirección General de Ingresos (DGI) in a January 19, 2023 decision involving remuneration for a related-party wholesale distributor of electronics determined using Transactional Net Margin Method (TMMM; CPM in the US). In earlier discussions, we noted how DGI benchmarked the return for a limited-function wholesale distributor of petroleum and the return for a high-function distributor of pharmaceuticals.

Table 1 presents the financials for the distribution affiliate under its intercompany pricing policy, the position of DGI, and two other scenarios. Operating expenses (OpEx) represented only 6.55% of sales, reflecting the limited functions performed by the distribution affiliate. Under the taxpayer’s intercompany pricing policy, the distribution affiliate’s gross profit margin (GPM) was 7.8%, which left its operating profit margin (OPM) at 1.25% or a 19.11% markup over operating expenses. Limited-function distribution affiliates would normally deserve a very modest OPM unless this affiliate had an extensive asset-to-OpEx ratio or owned valuable intangible assets.

Table 1: Income Statement of Related Party Distributor Under Alternative Applications of TNMM (in Millions)

The tax authorities successfully argued that the OPM should be 3.88% by presenting the 2011 to 2013 financials for nine wholesale distributors of electronics. Their position required that the gross margin be raised to 10.43% and would have operating profits reflect a 59.23% markup over operating expenses.

The taxpayer made two arguments against this position, with both arguments failing to convince the court. The court decision extensively discussed a debate whether the taxpayer could include other income in its calculation of operating profits. DGI successfully argued that the appropriate financials exclude other income. The other argument was that the taxpayer should have been allowed to include other companies such as SED International as possible comparables. DGI, however, noted that SED International had incurred persistent losses which eventually led to its bankruptcy. DGI successfully argued that the following nine third-party distributors be used in the analysis:

1. Anixter International, Inc. (AXE);
2. Arrow Electronics, Inc. (ARW);
3. Avnet, Inc. (AVT);
4. DXP Enterprises (DXPE);
5. Graybar Electric Co., Inc. (GRBE);
6. Houston Wire and Cable (HWCC);
7. Ingram Micro (IM);
8. Scansource, Inc. (SCSC); and
9. Tech Data, Corp. (TECD).

These companies were often used for similar TNMM analysis by representatives of multinationals for the period involved.

I reviewed the financials for these nine companies, noting that there is an extensive range of OpEx-to-sales ratios among this set with most but not all of the OpEx-to-sales ratios exceeding 8%. In fact, three of the nine third-party distributors had OpEx-to-sales ratios exceeding 16%. Companies with higher functions tend to have higher OPM, which is why many transfer pricing practitioners prefer the use of markups over operating expenses.

Table 2 presents the average annual financials for these nine companies over the three-year period ended December 31, 2013 or the three-year period ended June 30, 2014 in the case of AVT and SCSC. The high OPM for AXE and DXPE are in part due to its higher functions. IM and TECD have OPM near the 1.25% OPM for the distribution affiliate, but also have lower OpEx-to-sales ratios.

Table 2: Average Annual Financials for Nine Publicly Traded Distributors (in Millions)

While the median OPM for this set of distribution affiliates was 3.68%, the median markup was 36.89%. There were slight differences in the reported OPM noted in the court decision and the financials reported in Table 2. All financial data used in Table 2 was derived from the 10-K filings as reported by the Securities and Exchange Commission. Consider a third-party distributor that had a 13.7% gross margin with OpEX = 10% of sales. A 3.7% OPM would represent a 37% markup over operating expenses. If this company were considered comparable to a distribution affiliate whose operating expenses were only 6.5%, then the appropriate OPM would likely be only 2.4%.

Table 1 contrasts the implications of margin versus markup approaches when applied to the distribution affiliate in this litigation. While a margin approach would suggest a gross margin in excess of 10.2%, a markup approach would suggest a gross margin closer to 9%. In other words, a markup approach would suggest a much more modest decrease in the transfer pricing than what the tax authority proposed and the court suggested.

Customs Valuation Case 14.1 and the Eaton Advance Pricing Agreement

The World Customs Organization (WCO) provided a discussion of how TNMM might be used to benchmark the customs valuation for imported electronics from a related party manufacturer in case 14.1:

XCO, a manufacturer in country X sells relays to its wholly-owned subsidiary, ICO, a distributor of country  … The transfer pricing study used the Transactional Net Margin Method (“TNMM”) that, in this case, compared ICO’s operating margin with the operating margins of functionally comparable distributors of goods of the same class or kind … The transfer pricing study, using data taken from ICO’s company records, indicated that ICO’s operating profit margin on the sale of relays purchased from XCO was 2.5 percent in 2011 … In the transfer pricing study, eight distributors, unrelated to their suppliers, were selected based on the substantial similarity of their functions, assets and risks, compared to ICO ... The range of operating profit margins earned by these unrelated distributors was 0.64 to 2.79 percent, with a median of 1.93 percent.

The first column of Table 3 presents the financials noted in this WCO case study.  The cost of goods represented 82% of sales and the expenses borne by the distribution affiliate represents 15.5% of sales. As such, operating profits represent 2.5% of sales.

Table 3: Customs Policies (in Millions)

The case study does not specify the manufacturer of the relays or what nation was importing them. Let’s assume the transaction was between a U.S. affiliate of Tyco Electronics (now TE Connectivity) and Tyco Electronics Canada with the latter being a Canadian distribution affiliate. The very modest OPM noted in the case study raises several questions. One question is why the IRS would accept a Canadian OPM above the median margin. Of course, the Canadian Revenue Agency (CRA) might have insisted on a capital-adjusted Berry ratio or capital-adjusted markup approach if the OpEx-to-sales ratio for the third-party distributors was less than that of the Canadian distribution affiliate.

If we relied on the financials noted in the column labeled Customs-a, the markup over reported OpEX would be only 16.13%, which is well below the markups reported in Table 2. The financials for Tyco Electronics suggests its selling expenses are only 12.5% of sales rather than the reported 15.5% OpEx-to-sales ratio in the case study. If these expenses included third-party logistic expenses borne by the Canadian affiliate, the financials should be restated to include logistic costs in the cost of goods, which is shown in the column labeled Customs-b. The restated gross margin would be 15% with the ratio of value-added expenses-to-sales being 12.5%. Under this reporting, the markup over value-added expenses would be 20%.

Eaton manufactures electrical and industrial products. The IRS entered into a series of Advance Pricing Agreements covering a variety of transfer pricing issues, including goods produced in Puerto Rico and sold in the U.S. These agreements benchmarked the appropriate gross margin as a Berry ratio between 1.2 and 1.27 times the 13% operating expense to sales ratio for the U.S. distribution affiliate. Table 3 shows the implications of this agreement as the distribution affiliate having a 16% GPM and a 13% OPM. Case 14.1 and the Eaton APA agreement involve the distribution of electronics, with both reaching results similar to the taxpayer’s original transfer pricing policy in the Panamanian litigation.

A Capital-Adjusted Markup Approach and the Role of Assets

The taxpayer’s representatives raised a potentially important issue by arguing that adjustments for differences in the relative amounts of inventories, receivables, and payables should have been incorporated into the TNMM analysis. The DGI and the tax court, however, choose not to address any comparability adjustments with the tax court noting:

... upon verifying the transfer pricing studies for the 2013 period, submitted by the taxpayer, we agree with the Tax Authority in the sense that upon appreciating Appendix D of the aforementioned study visible at pages 70 to 71 of the background file, it only describes the definition of capital adjustments and we do not see that it contains a comparability analysis of the taxpayer and the companies that were selected as comparable companies in the items of accounts payable, accounts receivable and inventories, so it is insufficient to demonstrate that the application of such items improved the reliability of the results and increased the degree of comparability.

A standard model for the GPM of a distributor under arm’s length pricing is:

GPM = gross profits/S = b(E/S) + r(A/S),

where S = sales, E = value-added or OpEX, A = the distributor’s assets including working capital (inventories + receivables – payables), r = the return to asset, b = the impact of higher functions on the gross margin, and a represents the profits attributable to intangibles (if any). Since the OPM = the GPM - OpEx, the model can be restated as:

OPM = P/S = m(E/S) + r(A/S),

where P = operating profits and m = b -1. This model can be expressed in terms of a capital-adjusted markup approach where:

(P/E)t = (P/E)c + r[(A/E)t – (A/E)c],

where t denotes the tested party (distribution affiliate) and c denotes the comparables (third-party distributors). While the taxpayer suggested that the distribution affiliate had a lower asset-to-sales ratio than the third-party distributor, whether comparability adjustments lower or raise the markup depends on whether the asset-to-expense ratio for the distribution affiliate is less than or greater than that of the third-party distributors.

The court record suggested that the representatives of the taxpayer failed to provide sufficient information on the assets of the distribution affiliate to properly conduct the comparability adjustments suggested by our model. Since the representatives of the taxpayer were trying to defend a markup of less than 20%, providing such information would have been in their interest.

References

Republica de Panama Tribunal Administrativo Tirbutario, Resolucion de Fondo no.TAT-RF-006, Expediente: 115-19.

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

Harold McClure, “RPM v. TNMM: Benchmarking a LATAM Pharmaceutical Distributor,” October 6, 2021.

"Use of Transfer Pricing Documentation When Examining Related Party Transactions Under Article 1.2(a) of the Agreement," www.wcoomd.org.