Bloomberg Tax
July 6, 2020, 7:00 AM

INSIGHT: Transfer Pricing Adjustments to the Covid-19 Economic Downturn

Prita Subramanian
Prita Subramanian
Valentin L. Krustev
Valentin L. Krustev
HIlary Eisenberg
HIlary Eisenberg
Mariah Hughes
Mariah Hughes

Economies around the world are in the midst of a severe downturn brought about by measures taken to contain Covid-19. A question that is arising in light of the economic downturn is whether MNEs’ transfer prices need to be revised. This question is most notably arising in the context of routine entities that earn profit margins or markups within benchmark ranges based on comparable company profitability. In general, transfer prices that are based on the commonly used comparable profits method/transactional net margin method (CPM/TNMM) may require additional analysis to support their arm’s-length nature for the 2020 year.

In normal times, performing a transfer pricing analysis showing that a controlled entity earns profits within a reference CPM/TNMM range derived from comparable companies is often a simple exercise. However, during severe economic disruptions, comparability differences that normally would not have been significant may become more prominent and make it harder to establish reliable benchmarks for routine entities. This article discusses some potential approaches for adjusting profit benchmarks for controlled entities in the CPM/TNMM context, consistent with the arm’s-length standard.


One aspect of a severe downturn that makes determining appropriate transfer pricing approaches challenging is that its impact may be uneven across industries or across companies within an industry. Further, recessions are, in general, spaced further apart than expansionary periods, thus limiting the experience of practitioners in dealing with severe downturns. According to the National Bureau of Economic Research (NBER), the last recession started in December 2007 and lasted eighteen months. The result is that transfer pricing approaches that were developed in non-recessionary periods and based on market comparisons that were good benchmarks during such periods may not work well in a severe downturn such as the current one. Companies may, therefore, find it appropriate to adjust transfer pricing approaches developed during normal times to reflect current economic conditions.

Making adjustments for economic conditions is consistent with the guidance in the tax code Section 482 regulations (see generally Treasury Regulation Section 1.482-1, Treas. Reg. Section 1.482-5).In particular, Treas. Reg. Section 1.482-1(d)(1) states that:

“Whether a controlled transaction produces an arm’s length result is generally evaluated by comparing the results of that transaction to results realized by uncontrolled taxpayers engaged in comparable transactions under comparable circumstances. For this purpose, the comparability of transactions and circumstances must be evaluated considering all factors that could affect prices or profits in arm’s length dealings (comparability factors).”

The provision lists “economic conditions” as a comparability factor (Treas. Reg. Section 1.482-1(d)(1)(iv)).

In addition, Treas. Reg. Section 1.482-1(d)(3)(iv) states:

“Determining the degree of comparability between controlled and uncontrolled transactions requires a comparison of the significant economic conditions that could affect the prices that would be charged or paid, or the profit that would be earned in each of the transactions.”

Among the factors listed under economic conditions to be considered is “[t]he economic condition of the particular industry, including whether the market is in contraction or expansion.” (Treas. Reg. Section 1.482-1(d)(3)(iv)(G).)

The Organisation for Economic Co-operation and Development (OECD) Transfer Pricing Guidelines for Multinational Enterprises and Tax Administrations 2017 (OECD Guidelines) adopt a similar stance and specifically call out economic conditions as a comparability factor. The OECD Guidelines emphasize the importance of accounting for economic conditions:

“Economic circumstances that may be relevant to determining market comparability include the geographic location; the size of the markets; the extent of competition in the markets and the relative competitive positions of the buyers and sellers… The existence of a cycle (e.g. economic, business, or product cycle) is one of the economic circumstances that should be identified… The geographic market is another economic circumstance that should be identified.” (OECD Guidelines, Chapter I, Para. 1.110 – 1.112.)

The profit margins for controlled entities that are the subject of this article are based on comparable company benchmarks and it can be expected that there will be some self-correction to the benchmark profitability ranges as the comparable company profits change due to the economic downturn. However, the anticipation of changes in comparable company profit margins may not completely address the question of whether transfer pricing policies need adjustment as a result of the economic downturn.

Firstly, there is the practical issue that comparable company data are only available with a significant lag, so companies still need to determine if benchmarks based on historical data need to be adjusted for 2020.

Secondly, there is the more fundamental issue that differences between the comparable companies and the controlled entity being benchmarked that were not material during normal times might be quite pronounced in unusual times like these. To provide an example, the tested distributor and the comparable distributors may be quite different in terms of the diversification of their product portfolios.

A distributor with a diversified portfolio is likely to have a steadier trend in its profitability than a distributor that sells products within segments that are greatly impacted by the downturn. The latter distributor can be expected to have higher levels of extraordinary expenses as a result of the downturn. Thus, even if real-time data for 2020 were available, comparability adjustments might still be warranted.

This article presents some possible approaches to adjusting CPM/TNMM analyses in accordance with the arm’s-length standard to account for comparability challenges arising during the economic downturn. It is not meant to provide an exhaustive listing or discussion of possible adjustments. Further, there may be various approaches besides adjusting profitability benchmarks that involve different transfer pricing methods altogether, such as profit split methods, which are not the focus of this article.

Finally, any transfer pricing analysis for a routine entity should generally analyze contractual terms, functions and risks of the parties involved, and third party behavior to determine the arm’s-length return. While it may be appropriate to make adjustments to the historical pricing under certain circumstances, it may be most appropriate under other circumstances to keep the historical pricing intact. This article provides practical examples of how certain adjustments could be implemented without going into the broader question of whether adjustments should be made or which ones should be made, which will typically be a fact-specific evaluation.


We organize our discussion of transfer pricing adjustments into three general categories:

1. Adjustments to the tested party financials

2. Adjustments to the benchmarking period

3. Adjustments to the comparable companies’ profitability measures

While not every possible transfer pricing adjustment will fit under this taxonomy, in our experience most adjustments considered by practitioners can be included in one of these three categories. There are multiple possible approaches within each category, but it is the common themes which motivate the categorization.


In the first class of adjustments we review, the controlled entity’s financial data is modified to more reliably compare its financial results to unadjusted comparable company benchmarks in light of Covid-19. Such adjustments are allowed under the Section 482 regulations and the OECD Guidelines. For example, the Section 482 regulations note in the context of the CPM, “In certain cases it may also be appropriate to adjust the operating profit of the tested party and comparable parties.” (See Treas. Reg. Section 1.482-5(c)(2)(iv).) Further, the OECD Guidelines state in the context of determining the net profit indicator in the application of the TNMM, “Exceptional and extraordinary items of a non-recurring nature should generally also be excluded.” (Chapter II, para. 2.86.).

As a result of Covid-19, many companies are experiencing abnormal levels of certain costs or extraordinary costs that they would not have incurred during normal times. If comparable company results are based on recent historical financial data or if the comparable companies are impacted differently by Covid-19, it may be reasonable to adjust the tested party financial data to remove extraordinary costs related to Covid-19 since those are not reflected in the comparable company data. The profit margins from the remaining “normalized” tested party financials may then be reliably compared with the comparable company ranges.

An additional question with this approach is how to treat the abnormal or extraordinary costs that are excluded from the normalized tested party financials. One possibility may be to charge those out without a markup. Alternatively, there may be circumstances under which the tested party could assume some or all of those costs.

We discuss below some examples of extraordinary costs that companies affected financially by Covid-19 may see in 2020, which may be reasonable to treat separately from the regular CPM/TNMM benchmarking analysis.

Receivables write offs: Economic disruptions along the supply chain may result in customers of the controlled entity facing liquidity issues or even becoming insolvent. Consequently, the controlled entity may experience abnormally high bad debt and receivables write off.

Inventory write downs: Mandated shutdowns may translate into inventory buildups. If there is no market for the excess inventory or if disruptions in the logistics supply chain prevent market access, some of the company’s inventory may be written off.

Asset impairments: If the disruption has resulted in a permanent adjustment to a company’s business expectations, some of the company’s fixed assets may be subject to impairment for accounting purposes. The impairment would generally be recognized in the current accounting period.

Underutilized production capacity: Supply chain disruptions and reduced demand due to Covid-19 may lead to abnormally low manufacturing capacity utilization. A company may recognize unallocated overheads resulting from abnormally low capacity utilization as an expense in the period in which they are incurred rather than as a portion of the inventory cost.

Idle sales force: Covid-19 disruptions may result in unusual revenue losses. While companies may be able to scale back certain sales costs in proportion to the revenue losses, they may be unable to scale back other sales costs. Such costs will then be disproportionately higher as a percentage of revenues as compared to normal times. It may be possible to identify idle sales force costs during Covid-19.

We illustrate the adjustment to a tested party’s financials for some these items in Table 1 for a hypothetical manufacturer. (See article by Felgran, et al., (2009) for another example involving abnormal selling, general, and administrative (SG&A) expenses).

TABLE 1: Identifying Potential Covid-19 Related Costs

The tested party manufacturer in Table 1 is facing inventory write-downs, receivables write-offs, and impairment losses in 2020 due to Covid-19. If the extraordinary expenses related to these items are excluded from the financials, the operating margin moves up from 1.8% loss to 4.0% profit. The CPM/TNMM approach may be applied to the adjusted tested party data showing the normalized operating margin of 4.0%. This margin may be compared to the unadjusted pre-Covid range. Alternatively, it may be compared to comparable company benchmarks that include data from the Covid-19 period. In that case, the reference comparable range should also be reviewed. If some of the comparable companies incurred similar Covid-19 related expenses, these may also be adjusted for so as to ensure a more reliable comparison.


The second category of adjustments we discuss relates to adjusting the period over which comparable company profitability is measured to account for the impact of the downturn. Typically, CPM/TNMM analyses rely on the most recent publicly available comparable data. In practice, comparable company financial data are not audited and released until after the end of the applicable fiscal year end reporting date. Thus, the comparable company information relied upon in setting prices will usually lag behind the tested party data, and even more so when considering European or Asian databases. As such, the most recently available data for analyzing 2020 tested party results contemporaneously will likely be information from 2018 or 2019 (for documenting 2020 results for US tested parties, generally 2020 data would be available given the delay in filing tax returns after the fiscal year end).

This may cause practical challenges with regard to comparability, as the economic conditions in 2020 due to the Covid-19 outbreak are very different from those of the immediately preceding years, when most industries operated under fairly normal market conditions. Accordingly, comparable company data from the years immediately preceding the downturn years may not provide appropriate benchmarks in economic downturns.

In view of these considerations, we outline some possible approaches involving adjustments to the benchmarking period below.

Limit Benchmarking Data to Current Downturn: While benchmark data for the complete 2020 year won’t be available contemporaneously, it may be possible to estimate 2020 results for comparable companies based on quarterly financial filings. Given how fast economic conditions are evolving, quarterly data as such may not provide a full picture of the year. Calendar year companies may have at most three quarters of information to rely upon contemporaneously for the year 2020. In that regard, quarterly financial data supplemented with historical data from the last downturn and a company’s disclosures on its projections and forward-looking statements to its investors may provide reasonable benchmarks. In a recent public statement released on April 8, 2020, the SEC urged companies to “provide as much information as is practicable regarding their current operating status and their future operating plans under various Covid-19 mitigation conditions.” (SEC Chairman Jay Clayton, William Hinton, 2020). In this statement, the SEC stated that company disclosures should cover how the company’s “operations and financial condition may change as all our efforts to fight Covid-19 progress.” We note that such information may not be available for private companies, which could be a shortcoming for European or Asian comparable company sets since they often include private companies.

Extension of the Benchmarking Period to Include Prior Downturn: For a CPM/TNMM analysis, a multiple-year tested period is generally considered appropriate in order to reduce short-term variances that may be unrelated to transfer pricing.(Treas. Reg. Section 1.482-1(f)(2)(iii)(D).) The multiple year tested period commonly uses three years of comparable data, which usually includes the taxable year under review and the two preceding taxable years. (Treas. Reg. Section 1.482-5(b)(4).) As the use of financial data from the previous three years alone, when the U.S. economy generally was in better circumstances, may not be appropriate for benchmarking, taxpayers might consider adding data from the previous recession years.

Limit Benchmarking Period to Prior Downturn Years: Transfer pricing practitioners may consider including only the profitability levels for years which reflected prior economic downturns. For example, taxpayers may consider using comparable company data from the 2008-09 recession. This approach, however, must include careful consideration of industry and market level factors in determining the appropriate years of data to establish a comparable range. For example, while airlines did experience a decline in profitability during the 2008-09 recession, an analysis must consider whether the level of decline is comparable to that experienced during Covid-19.

We illustrate the first approach described above, i.e., limiting the benchmarking period to the current downturn and using earnings guidance, in Table 2 below.

TABLE 2: Illustration of Estimated 2020 Results Based on Company Projections

The above table illustrates a hypothetical estimate of the 2020 results for a company based on its publicly available quarterly financials and disclosures to investors projecting future prospects for the remainder of 2020. In this example, Company A has available quarterly 2020 financials through Q2, which shows a decline in operating income of approximately 79% compared to Q1. Company A’s public disclosures project declines in sales over the next two quarters, noting an expected decline of 15% for Q3 and a slightly lower decline in Q4.

Based on Company A’s disclosures, the table above presents one possible approach to estimating the 2020 profit margin of Company A. It assumes a decline of 10 percent in revenues between Q3 and Q4 given the company’s disclosure that its Q4 decline is expected to be lower than the Q3 decline. Importantly, this example assumes that costs and profits will decline proportionately to revenues in Q3 and Q4, resulting in a constant operating margin over Q2 through Q4. Thus, while the observed operating margin over the first half of the year is 2.6%, the estimated margin based on the company’s public disclosures is lower at 1.9%.

Alternatively, if there is reason to believe that costs will not decline proportionately with revenues (for example, based on an analysis of trends in costs and revenues in previous periods of revenue declines for the company), it may be reasonable to decline costs less than revenues to estimate Q3 and Q4 profits. For instance, if we assume half the operating expenses in the previous quarter will carry over unchanged to the following quarter while the other half will scale proportionately with revenues, the estimate of full year 2020 operating margin will be -0.3% instead of the 1.9% shown in the table.

The 2020 estimate may provide a reasonable benchmark that can be used to guide transfer pricing policy in real time, before the 2020 year is closed. However, it is important to monitor additional company results as they become available, since guidance may change as external conditions related to Covid-19 develop. Further, as the example above illustrates, it is important to factor in as much relevant information as possible in estimating the full year profit margins.


The third category of adjustments we consider are adjustments to the comparable companies’ profitability measures. Within this category we discuss the following three types of adjustments:

1. Adjustments to the range used for benchmarking

2. Adjustments to comparable company results based on company metrics

3. Adjustments to comparable company results based on macroeconomic indicators

Adjustments to the range used for benchmarking

In the absence of contemporaneous comparable company information during the downturn, a simple option to account for the downturn could be to vary the range used for the benchmarking. For example, a lower 50 percentage point interval within the full range may be considered to reflect the impact of the downturn, such as the 10th to 60th percentile levels rather than the customary interquartile range (the 25th to 75th percentile levels). Another possibility may be to use the full range to account for the increased volatility and uncertainty that typically characterize periods of severe economic downturn. A third possibility could be to adjust the interquartile range by a percentage based on the percentage change in the interquartile range of similar companies in the last recession.

Adjustments to comparable company results based on company metrics

A company’s profit margins depend on its performance on various operational (and consequently, financial) metrics, e.g., its ability to trim down its sales force or its ability to utilize existing production capacity during the economic downturn. If the comparable companies are different from the tested party in terms of their performance on key metrics that are significantly impacted by the downturn, an adjustment to comparable company results based on such metrics may be reasonable. The differences between the comparable companies and the tested party could be a result of using pre-Covid-19 comparable company financial data or because the comparable companies are impacted differently by Covid-19.

We illustrate one such adjustment based on differences in companies’ cost structures below.

Fundamentally, companies’ cost structures have variable costs, which scale with sold volumes and revenue (e.g., cost of materials), and fixed costs (e.g., cost of fixed assets or certain general and administrative overheads), which cannot be adjusted as quickly when the economic conditions change. When volumes decrease, fixed costs do not decrease at the same rate, which results in a disproportionate reduction in operating profit. The higher the level of fixed costs in a business, the more sensitive will the business’ profitability be to revenue fluctuations.

If full information about a company’s fixed and variable costs is available, the concept can be illustrated using the calculation in Table 3 below. For another example, see McClure (2019).

TABLE 3. Variable vs. Fixed Costs Illustration

If we denote revenue by “R”, variable costs by “VC”, fixed costs by “FC” and the degree of operating leverage as “DOL”, then DOL is defined according to the following formula:

Fixed costs decrease the denominator and therefore increase the DOL. For Company A in Table 3 the DOL is 1.66x while for Company B, which has lower fixed costs, the DOL is 1.13x. Based on the revenue levels shown in the table, both companies earn the same profit amount (i.e., R – VC – FC).

However, Company A’s and Company B’s profit levels will respond very differently to a revenue reduction of 50%. For Company A, the R – VC component will fall from 25 to 12.5, as both revenue and variable costs will drop by 50%. However, since fixed costs will remain the same, the profit will fall to 2.5 which is an 83% reduction from 15 before the revenue decrease. On the other hand, Company B’s profit will only decline to 6.5 given its lower level of fixed costs and DOL (calculated based on a 50% reduction in R – VC to 8.5 and a fixed cost of 2, leading to profit of 8.5 – 2 = 6.5.). When revenue declines are significant, if the tested entity has significantly different DOL than the comparable companies, adjustments may be performed to the reference range to account for such differences.

In practice, it is not always straightforward to estimate a benchmark company’s fixed and variable costs solely based on information disclosed in public filings. In lieu, the DOL may be estimated empirically by observing how a company’s operating profit tracked revenue during a historical period. Specifically, the DOL can also be defined as follows (See Damodaran, 2002):

By examining past years, we can estimate the percentage change in operating profit, on average, for a single percent change in revenue for a given company. We can then compare the DOL of the tested entity to the DOL of the benchmarking group; if there are significant differences, we can adjust the benchmarking group’s operating profit to be consistent with the cost structure of the tested entity. We illustrate the adjustment process through a case study. Another possible approach is to estimate fixed costs more directly by considering SG&A to revenue ratio. The drawback of this method is that accounting classifications into COGS vs. SGA may not always follow a fixed vs. variable cost distinction. For example, manufacturing assets depreciation which is a fixed cost would typically be a COGS item.

Operating Leverage Adjustment Case Study

Our case study is based on the information for hypothetical companies in Table 4 below.

TABLE 4. Illustration of Operating Leverage Adjustment

The example assumes that at the time of the analysis, information for 2020 is available. The approach described here is applicable, with possibly some adjustments, even if comparable company data are only available through 2019. All benchmarking companies in this example have experienced decrease in revenue and profits; however, the tested party’s profit has fallen much more, resulting in its operating margin being below the 2020 interquartile range.

We assume there is reason to believe that the tested party experienced worse results due to its higher fixed costs which magnified its losses in the face of falling volumes. To test this hypothesis we use the historical sales and profit data to estimate the DOL for the tested party and the benchmarking companies.

For example, for Company A, the DOL for 2019 would be calculated as follows:

That is, Company A saw a 15% decline in revenue from 2018 to 2019, but a 56% operating profit decline, implying DOL of 3.85x. The degree of leverage may also appear more visibly during downturn years; this may drive the analyst’s considerations which historical years the calculation should be based on.

We next calculate the DOL for all companies for each year 2017-2019 (the calculation requires 2016 data as well). We average the 2017-2019 DOL measures for each company to approximate the longer-term relationship between revenue changes and operating profit changes. The results are summarized in Table 5; the full calculations are omitted for brevity.

TABLE 5. Degree of Operating Leverage Summary

It is clear from the table that the tested party has higher DOL than the benchmarking group. One way to calculate an adjustment for the differences in cost structures of the benchmark companies and the tested party is to estimate how the benchmark companies’ operating profit might have changed from 2019 to 2020 if they also had DOL of 4.06.

For example, Company A’s revenue in 2020 declined by 52%, from 6,024 to 2,880. If Company A had DOL of 4.06, we would expect an operating profit decline of 52% x 4.06 = 212%. This would put its 2020 operating loss at 303 instead of 181, leading to an adjusted operating margin of -10.5% instead of -6.3%. (If complete 2020 data for Company A are not available, it may be possible to obtain just revenue estimates based on interim financial statements and other Company A disclosures). After all benchmarking companies are adjusted, we arrive at a lower interquartile range as summarized in Table 6 below.

TABLE 6 Summary of Adjusted Results

After the adjustments, the tested party result would be within the interquartile range.

Other variations of the operating leverage adjustment described here are possible too. For example,

— Another possibility is to consider differences in capacity utilization between the tested entity and the benchmarking companies. For some industries, regression-based approaches can be used to estimate the effect of capacity utilization on operating margins. If the tested entity incurred losses due to significantly lower capacity utilization, the relationships estimated from the regression approach can be used to adjust the benchmarking range downward. This approach is discussed in more detail by Felgran et. al. (2009). The cited authors describe how a semiconductor capacity utilization index can be used to extrapolate cost plus markups via a regression approach.

— A simpler approach could be to change the profit level indicator used for the benchmarking from an operating profit measure to a gross profit measure, since cost of goods sold are more likely than operating expenses to be variable costs. For example, it may be reasonable to use gross margin as a benchmark rather than operating margin during the downturn.

— As a less quantitative alternative, transfer pricing practitioners could try to estimate qualitatively which benchmarking companies have similar operating leverage as the tested entity and eliminate companies with significantly different cost structures.

The adjustments described here are a few among many possibilities; however, the underlying theme is the same, i.e., to more closely align the operational metrics of significant import in an economic downturn of the benchmarking companies and the tested entity.

Adjustments to comparable company results based on macroeconomic indicators

Transfer pricing adjustments may also be merited to account for systematic differences between economic conditions of the tested party and the comparable companies, as evidenced by macroeconomic indicators. Such differences could be a result of using comparable company information from a historical period subject to different economic conditions than the current downturn or due to the comparable companies and tested party operating in industries or countries experiencing the economic fallout of Covid-19 differently.

For example, it is often the case, particularly for tested parties outside the U.S., that comparable companies are based in different countries. While under normal economic conditions, companies operating in the same region but not the same country as the tested party may provide reliable profitability benchmarks, what we have seen so far with regard to the Covid-19 economic fallout is that its effects have not been uniform across countries. Some countries are suffering longer shutdowns than others and different economies are dominated by different industries which in turn may be affected differently.

Ideally, CPM/TNMM benchmarks will be derived from the same market in which the tested entity operates. However, in practice it is difficult to identify sufficient comparable company data from many jurisdictions in commercially available databases. That is why CPM/TNMM benchmarks frequently rely on regional datasets. However, even within the same region (e.g., Europe or Asia-Pacific), countries may have been differently affected by Covid-19 related disruptions.

One possibility to adjust for such differences could be to use models that link macroeconomic indicators such as GDP (or more specific indices relevant to particular industries) to company performance metrics such as revenue or profits. These models use a “regression” or an equation specifying a statistical relationship between the chosen macroeconomic indicator and firm revenue or profits. For example, we could consider how GDP growth is correlated with company operating margin growth for a given industry or country. Once the statistical relationship is estimated, it can be used to predict the operating margin growth of companies within the industry or country if only GDP growth is known.

Table 7 shows hypothetical GDP growth and average operating margin growth in 2020 for companies comparable to the tested party for five countries. The assumption is that in each of these five countries operating margin growth can be observed directly. It could be determined as an average or median of comparable company operating margin growth. Alternatively, average or median data for a broader pool of companies in each country, e.g., within industry or SIC code, could be collected.

TABLE 7. GDP and Operating Margin

Among the five countries listed above, the lowest growth country shows a decline of 5% in GDP. However, suppose that Country F, where the tested party is located, saw a 7.0% decline in GDP because it was affected by Covid-19 worse than Countries A-E. If we believe economic conditions are an important comparability factor, we could adjust the results of the benchmarking companies downward to account for the different economic conditions in Country F and the countries in which the comparable companies operate.

It can been seen in Table 7 that there is a correlation between GDP decline and operating margin decline. In fact, if we estimate a linear regression, we would obtain the following relationship between the two indicators:

The regression in this above hypothetical example has a high “R-squared” measure of fit at 89% (higher R-squared indicates greater explanatory power for the regression).

We could now use the estimated relationship to adjust the profit margins of the comparable companies in Countries A-E to the levels they would have been had they operated in Country F.

For example, suppose a comparable company in Country A had an operating margin of 10.0% in 2019, which declined to 9.1% in 2020.We know that GDP growth in Country A in 2020 was -3.0%, i.e., 4 percentage points higher than the GDP growth in Country F of -7.0%. Based on the regression equation above, the additional 4 percentage point decline in GDP in Country F compared to Country A would be associated with an additional 16.1% decline in operating margin (i.e., change in operating margin growth = 4.02 * change in GDP growth = 4.02 * -4% = -16.1%, for an adjusted operating margin of 7.5% in 2020 because 9.1% + 10.0%*-16.1% = 7.5%).

The example above presented a simple illustration of an adjustment to comparable company results for the varying impact of Covid-19 on different countries using macroeconomic data. It should be noted that macroeconomic data by definition are aggregate data, which can mask differential trends at more disaggregated levels, so should be used with caution. Other adjustments using macroeconomic data may be possible too. For example,

— The adjustment described above used cross-sectional GDP and profit data for countries, i.e., the regression data included one data point for each country. A variant of this approach could be to perform separate time series regressions (i.e. regressions that derive a correlation between operating margin growth and GDP growth over time within a country) for each country to derive a relationship between operating margin growth and GDP growth within each country. Then a similar adjustment as outlined above could be performed for each comparable company based on the regression equation for that country.

— Importantly, a time series regression could be used to account for comparable company data lags—a challenge for 2020, as discussed above. For example, a regression analysis could be used to establish a relationship between GDP growth and operating margin growth in the tested party country, and possibly each country with a comparable company in the benchmark set, based on historical data. If 2020 comparable data reflecting Covid-19 disruptions are not available, the estimated relationship can be used to extrapolate a 2020 range. In practice, GDP and other macroeconomic measures are published in real time, with small delays and will be available much before public companies file their financial reports.

— Multiple other variations of these adjustments are possible and the focus may be on industry specific drivers rather than broader measures such as GDP. Examples include various manufacturing, durable goods, and other indices routinely published for different countries.

— Further, the adjustment can also be performed qualitatively, for example, by excluding benchmarking data from countries with significantly different economic distress compared to the jurisdiction of the tested entity.


We have discussed the practical application of three general categories of adjustments that may be useful in economic downturns. While we have discussed these categories of adjustments separately, they are by no means mutually exclusive. For example, adjustments to the tested party financials could be performed in conjunction with an adjustment applied to the comparable companies. There may also be approaches not discussed in this article, or the option of making no adjustment at all, that may be more appropriate depending on circumstances.

The adjustments and the CPM/TNMM analysis in general should be based on an understanding of the specific factual circumstances and contractual terms surrounding the controlled transactions. While the focus of this article has been on CPM/TNMM adjustments, a broader analysis of contractual terms, functions and risks of the parties involved and third party behavior to determine the arm’s-length return should be an inherent part of the overall transfer pricing approach.

Under the current economic environment, there will likely be an increased need for adjustments in order to improve reliability of the application of the CPM/TNMM. Additionally, there is a greater likelihood of the adjustments being novel or ones with which tax authorities have less experience. Thus, the likelihood of challenge by tax authorities may be greater than under normal circumstances, which makes it especially important to carefully analyze, support and document any adjustments.

In particular, it is important to carefully consider the reason for the adjustment, i.e., whether it is to estimate 2020 results in the absence of real-time data and/or to account for the differential impact of the downturn on the tested party and comparable companies. It is important to keep in mind that when 2020 data do become available, they may look different from what was estimated. Thus, it is advisable to monitor 2020 information in conjunction with developing careful arguments for any adjustments.

Given the uncertain economic conditions, it may take companies some time to determine whether and how their CPM/TNMM analyses need to be modified. While companies are making such determinations, it will be helpful for them to collect information that will be important in these determinations. For example, in addition to monitoring the financials of comparable companies, it will be useful for companies to start organizing their own financial data with the aim of aiding future analyses. This could involve tracking extraordinary expenses and quantifying the impact of operational disruptions on their financials.

Finally, there may be some differences in approaches of tax authorities regarding the acceptance of various types of adjustments. For example, some tax authorities might view adjustments to tested party results more favorably than adjustments to comparable company results. Tax authorities may also differ in their levels of comfort with complex analyses. Thus, it will also be important to consider the viewpoints of the relevant tax authorities in devising, supporting and documenting adjustments.


This article has outlined the application of three types of transfer pricing adjustments which may be considered in CPM/TNMM analyses in view of the unprecedented global economic disruption due to Covid-19: adjustments to the tested party financials; adjustments to the benchmarking period; and adjustments to the comparable companies’ profitability measures. We believe that comparability is more difficult to establish in times of severe economic disruption and these adjustments may be more warranted for 2020 compared to ordinary years. Practitioners may use one or more of the adjustments we describe to show the arm’s-length nature of a routine entity’s transactions although the specific facts and contractual terms surrounding a transaction will ultimately determine the appropriateness of any approach.

This column doesn’t necessarily reflect the opinion of The Bureau of National Affairs Inc. or its owners.

Author Information

Prita Subramanian is a principal in the Economic Valuation Services (EVS) group of KPMG LLP’s Washington National Tax (WNT) practice, Valentin Krustev is an EVS senior manager in Houston , Hilary Eisenberg is a senior associate in WNT’s EVS group, and Mariah Hughes is an EVS senior associate in New York.

The following information is not intended to be “written advice concerning one or more Federal tax matters” subject to the requirements of section 10.37(a)(2) of Treasury Department Circular 230. The information contained herein is of a general nature and based on authorities that are subject to change. Applicability of the information to specific situations should be determined through consultation with your tax adviser. This article represents the views of the author(s) only, and does not necessarily represent the views or professional advice of KPMG LLP.


Learn more about Bloomberg Tax or Log In to keep reading:

Learn About Bloomberg Tax

From research to software to news, find what you need to stay ahead.

Already a subscriber?

Log in to keep reading or access research tools.