Menu
Tax Notes logo

Transfer Pricing Benchmark: 2022 North American Electronic Manufacturing Services

Posted on Apr. 11, 2022
Andrew Hughes
Andrew Hughes

Andrew Hughes is an economist specializing in transfer pricing and valuation. He is based in Brussels.

In this article, part of a series providing transfer pricing benchmarks that can be replicated by practitioners using readily available public data, Hughes considers a North American electronic manufacturing services benchmark for transfer pricing.

Hughes has also published a full-length benchmark that includes all the details of a traditional transfer pricing benchmark.

Transfer pricing has shifted toward a compliance-driven world. Multinational groups must issue transfer pricing documentation reports in more jurisdictions than ever before, leading to increased compliance costs. As compliance requirements have risen, consultancies have turned to automation and standardization of nearly all documentation work, decreasing the costs of performing that work.

One key aspect of nearly all transfer pricing documentation is a benchmarking analysis. For transactions driven by the comparable profits method, a search generally is required for identifying companies comparable to the tested party. However, most consulting firms have already standardized the process with dozens of off-the-shelf benchmarks that can be provided to clients on request. On the one hand, that is beneficial for multinationals because the turnaround time for benchmarking studies is significantly decreased. On the other hand, taxpayers continue to pay for a service that has nearly become commoditized and still lacks publicly available secondary resources for practitioners.

Given the proliferation of publicly available data, free screening tools, and data mining techniques, now is the time to provide clients with corroborative benchmarks and data. Enter this series of articles, meant to provide tax practitioners worldwide with transfer pricing benchmarking suggestions and information for the most widely used CPM intercompany transactions. This series also attempts to take the complexity out of benchmarking analyses that use expensive tools, demonstrating that all practitioners can perform benchmarking analyses using tools and information that are likely already at their disposal.

This series could be used as a starting point for evaluating comparable companies. Certainly, the facts and circumstances of individual intercompany transactions should be scrutinized against the series content to make an informed decision on the transaction at hand. Practitioners should always examine each company before deciding whether it would be comparable to the tested party. This article uses mining techniques to pull data from companies’ Forms 10-K. Practitioners should always double-check data and validate company financials before using them in transfer pricing analyses and not rely solely on the data in this article. Although all figures have been checked, I lack the manpower and expensive tools that consultancies use in guaranteeing complete accuracy for their clients.1

I. Electronic Manufacturing Services Benchmark

I started with identifying the North American electronic manufacturing service (EMS) companies that I wanted to benchmark. The goal was to have a set of companies that might be comparable to a tested party that manufactures various electronic products or devices, such as household goods, medical devices and display systems, audio equipment, and telecommunication products. I wanted to identify companies performing functions that would closely mirror those of a contract manufacturer, rather than vertically integrated electronic manufacturers that sell products under their own brand names and develop their own products in-house.

Selecting the initial companies to include in the search is important. Many practitioners make those selections with standard industrial classification or North American Industry Classification System codes; however, that can lead to an excessive number of companies to screen that are not even closely comparable to the tested party. Given that, I rely on the industry categorizations that investment analysts give to companies. After all, there are perhaps no better people to provide an initial industry categorization than those who are paid to study public companies in close detail.

This study uses the 154 companies in TD Ameritrade’s “Information Technology — Electronic Equipment, Instruments & Components — Electronic Components,” “Information Technology — Electronic Equipment, Instruments & Components — Electronic Equipment & Instruments,” and “Information Technology — Electronic Equipment, Instruments & Components — Electronic Manufacturing” industries.

II. Quantitative Screen

Next, I performed a series of quantitative screens on the 154 companies. I excluded companies that lack any publicly available information. I also omitted those with less than two years of revenue or operating profit metrics reported during the past three years to ensure that companies are still active and have not had a recent stoppage of activities. Also automatically excluded were companies with operating losses in the past three fiscal years. While that biases the results to some extent, I want to exclude companies that may be on the brink of financial collapse — a common practice in benchmarking studies. Next, I excluded companies with a three-year revenue average of less than $10 million to avoid wasting time on start-up or pink-sheet companies that may have much different expenses and challenges from those of the tested party. Last, I omitted companies with high annual research and development spending and a high percentage of intangible assets to total assets because a routine multinational’s contract manufacturing tested party is not likely to have those things. Table 1 summarizes my initial quantitative screens.

Table 1. Quantitative Screens and the Number of Companies Affected

Quantitative Screen

Companies Affected

Exclude companies lacking publicly available information

31

Exclude companies without at least two years of revenue or operating profit data

8

Exclude companies with operating losses in all of the last three years

49

Exclude companies with less than $10 million in average revenue in the last three years

42

Exclude companies with R&D spending as a percentage of revenue over 3 percent

80

Exclude companies with intangible assets as a percent of total assets over 15 percent

16

After applying the quantitative screens, 30 companies were left.2 Markup on total costs (MUTC) and return on assets (ROA)3 profit-level indicators (PLIs) for those 30 companies are listed in Table 2.4

Table 2. PLIs for EMS Companies After Applying Quantitative Screens

 

MUTC

ROA

Maximum

40.2%

14.6%

Upper Quartile

9.1%

10%

Median

4.9%

6.1%

Average

7.3%

5.8%

Lower Quartile

2.9%

2.7%

Minimum

-7.7%

-10.6%

At this stage, I do not report the Berry ratio PLI because I have not performed an in-depth review of the Forms 10-K. Reporting that ratio without an analysis would lead to meaningless results because companies in this industry may report cost of goods sold and operating expenses differently, which would greatly affect the PLI.

III. Qualitative Screen

For the remaining 30 companies, I reviewed all business descriptions, company websites, and Form 10-K filings to determine whether a company could be comparable for inclusion in the broad EMS benchmark. I rejected companies for some common reasons:

  • the company is a subsidiary whose results are not ideal for inclusion because its actual financial results could be affected or determined by transfer pricing;

  • the company owns brands and sells products under its own brand names;

  • the company manufactures dissimilar products (metal-based products such as aluminum compartments or connectors); and

  • the company bears dissimilar risks (for example, a substantial amount of profit is derived from holding marketable securities or stocks).

That list includes the most critical rejection reasons in my benchmark. I rejected companies based on their general comparability for several other reasons (for example, manufacturing of nonelectronic products). Further, most companies reviewed had some operations outside North America, and few reported the actual breakdown of revenue generated in North America versus outside North America (one company even specified in its Form 10-K that although half its revenue was directly generated in Asia, nearly 90 percent of the products were for North American customers). That made it difficult to distinguish whether a company should be considered for the benchmark.

Therefore, I decided to accept a company if it was an EMS company performing functions, bearing risks, and owning assets similar to those of a contract manufacturer and generated sales and had a manufacturing presence in North America, even if the exact ratio of North American versus non-North American presence was not ascertainable.

The goal of that review was to provide a range of potentially comparable companies that perform contract manufacturing functions in the EMS industry. I wanted to include companies that manufacture various electronic products, and most of the accepted companies manufacture a range of products for clients in the medical, automotive, aviation, and consumer goods industries.

After the qualitative screen, I accepted 10 companies for the general North American EMS benchmark. (See Table 3.)

Table 3. Companies Accepted as Potential Contract EMS Comparables

Benchmark Electronics Inc.

Kimball Electronics Inc.

Fabrinet

Nortech Systems Inc.

Flex Ltd.

Plexus Corp.

Jabil Inc.

Sanmina Corp.

Keytronic Corp.

SigmaTron International Inc.

Table 4 shows the resulting PLIs for those 10 companies.

Table 4. PLIs for Accepted Contract EMS Companies

 

MUTC

Berry Ratio

ROA

Maximum

8.7%

3.14

10.1%

Upper Quartile

4.7%

2.03

7.3%

Median

3.6%

1.87

6.2%

Average

3.4%

1.78

4.9%

Lower Quartile

1.9%

1.3

2.6%

Minimum

-1.9%

0.8

-3.5%

After reviewing the companies’ Forms 10-K, I found there may be differences in the way companies classify employee salary expenses — that is, COGS versus operating expenses. Although that would not affect the MUTC and ROA results, it means a practitioner wanting to use a Berry ratio for benchmarking an intercompany transaction for contract EMS should scrutinize the financial accounts of potentially comparable companies to ensure there are no differences in the way the tested party and comparable companies classify salary expenses. In theory, the differences in expense classification for manufacturers should be less than those in another industry, such as consulting services.

IV. Conclusion

As transfer pricing benchmarking turns toward compliance and automation, tax practitioners should have readily available resources to assist them in reviewing and challenging their intercompany pricing and benchmarking analyses. The goal of this series is to provide multinationals with an additional benchmarking resource using publicly available data and resources.

I simplified the benchmark by including broad industry categories and allowing for a range of EMS in the accepted comparable companies. I then reported the PLIs resulting from the benchmarking analyses. Practitioners should consider that not all companies report financial data by region and that it will thus be difficult to determine whether some companies should be classified as North American. Practitioners performing that search will also want to pay special attention to whether a potential comparable company manufactures products under its own brand name because that might not be similar to the way a contract manufacturer would operate in an intercompany context.

I welcome any feedback on data points or figures that practitioners would like to see in future articles in this series.

FOOTNOTES

1 For a full-length benchmark that includes all the details of a traditional transfer pricing benchmark, see Andrew Hughes, “Transfer Pricing Benchmarking: 2022 North American Electronic Manufacturing Services,” Tax Notes Today Int’l (Apr. 6, 2022).

2 One hundred twenty-four companies were eliminated, some for multiple reasons.

3 ROA is defined as operating profit divided by the average of the last two years’ total assets to eliminate large discrepancies that might arise from taking a value at a single point in time.

4 The data come from companies’ Form 10-K filings or annual reports, where applicable. Interquartile ranges are based on the OECD’s definition and may differ from the IRC section 482 definition.

END FOOTNOTES

Copy RID