Transfer Pricing and COVID-19: Loss Allocation for Retailers

By the end of May, less than six months after it was first identified, the COVID-19 pandemic had infected more than 6 million individuals in more than 180 countries and territories and led to more than 370,000 deaths. Governments around the world have reacted to the virus by restricting the movement of people, shutting down economic activities, and offering economic support to companies and individuals. All the major economies, in particular China, the United States, and the countries of Western Europe, have been severely hit by the economic fallout.

Likewise, most multinational enterprises are suffering amid an uncertain economic environment that may call for a “new normal” in the medium term. Most of the retail sector is facing lost profits and lower sales volumes along with the expectation that consumers will demand price reductions in the troubled post-lockdown period, not to mention the profound modifications to their supply channels and ways of doing business more broadly.

Large deviations in MNE revenues, costs, and profits raise questions regarding the allocation of these deviations among the various entities involved in complex value and supply chains. While transfer pricing rules and the arm’s-length principle govern the allocation of profit within MNEs, neither existing rules nor general transfer pricing general practice anticipated the unique situation of COVID-19.

Practitioners have begun to opine about the allocation of virus-related losses among group entities, and the OECD has announced plans to publish specific guidance on the subject.¹ Some tax authorities have encouraged taxpayers to contact them to identify possible solutions.

This article will add to the discussion by focusing on the specific effects of the COVID-19 crises on retailers, offering one approach for allocating the COVID-19 loss among group entities, and demonstrating the proposed approach using a sector-specific case study.

While we do not purport to provide a definitive solution to a complex matter, we will attempt to offer an approach that combines analytical rigor and adherence to traditional transfer pricing rules with pragmatism. In doing so, we rely on the fair assumption that given the magnitude of the crisis and its unique characteristics, one cannot derive significant insights from past crises, such as the 2008 financial crisis in 2008-2009 or the SARS epidemic.

Focusing on Retailers

Our Hypothetical Retailer

There are myriad retailers in the global economy, and retail includes several subsectors. Therefore, we will focus on a hypothetical retailer that has specific features.

This retailer is part of an integrated multinational group that oversees the supply chain from sourcing to manufacturing, marketing and distribution. The retailer primarily sells high-end luxury products (clothes, garments, accessories, cosmetics, and so forth) in its own shops. Our retailer has the following transfer pricing characteristics:

• The functional and risk profile of local distributors in the market has led the group to select the transactional net margin method (TNMM) as the most appropriate transfer pricing method for remunerating local distributors. This decision takes into account the centralized model of the group and the lack of a reliable basis for applying the comparable uncontrolled price method.

• Before COVID-19, the target margin for distributors was generally set in light of the remuneration of comparable companies. The quality of the comparables and the lag in financial data available — comparable data are generally from two years prior — were not necessarily a problem because overall net margins seemed reasonable in light of overall profitability of the group. Some of the group’s retailers had concluded advance pricing agreements (unilateral or bilateral) for securing their tax and transfer pricing positions.

• Originally, the group expected 2020 to be consistent with 2019, and prices had been set accordingly.

The Lockdown Period

From March to May,² sales levels were close to zero as a result of lockdowns in major markets. Many shops closed either as a direct result of local laws and administrative measures or as an indirect result of local laws forcing mall closure. Some shops closed because of a voluntary decision by the retailer itself, for instance when masks or other protective gear were not available or there was clear perception that very few people would have come to the stores given the non-necessary nature of the products sold and the general restrictions on people’s movements.

In some markets, retailers benefited from government subsidies to support labor costs. Overall, the MNEs’ cash shortage has been partially offset by postponed tax and social contribution payments, access to state-guaranteed loans, and government indemnification to support rental costs during the lockdown and prevent permanent store closures. Some businesses elected not to avail themselves of the full array of government subsidies or abstained from dismissing employees (even when legally permitted) for ethical and reputational reasons.

MNEs did experience an increase in their e-commerce sales. This did not necessarily benefit the retail entities since often these sales are directly made and invoiced by the headquarter entity, with the local entity having no or limited involvement.

As a result of the above, the profit-and-loss statement for the lockdown period shows a loss, the magnitude of which depends on the extent to which one dollar of lost sales converts into lost profits. Thus, total losses depend on the specific cost structure of each individual retailer, its ability to reduce costs, and the size of public subsidies.

The Post-Lockdown Period

In June retailers slowly began reopening their shops.

The level of sales may not be at pre-pandemic levels for several reasons: people’s attitude and cautiousness, including feelings about the pandemic and the general economic uncertainty; limited entry to the shops and potential wait times may discourage shoppers; and most of the spring-summer collection is likely to be discounted, with the retailers expecting almost no full-price sales. At the same time, the retailers have incurred several costs to comply with post-lockdown regulations. Resizing of the operations is limited because of concerns about the brand’s reputation.

Therefore, the MNE expects a loss in the post-lockdown period through the end of 2020. The outlook for 2021 remains uncertain.

Full-Year 2020

Because of all the foregoing issues, 2020 budget goals may not be hit, and the retailer’s accounts may show an overall loss before any transfer pricing adjustment.

The Key Transfer Pricing Questions

From a transfer pricing perspective, we have identified two key questions:

• Should the limited-risk retailers bear, under arm’s-length conditions, a portion of the negative financial consequences of the COVID-19 disruptions?

• If the first question is answered in the affirmative (which we and others assume will be the case³), how should the portion of this COVID-19 loss that pertains to the limited-risk retailers, as identified above, be calculated?

The Arm’s-Length Rule in a Pandemic

Given its very nature, COVID-19 can be seen as an “extraordinary” hazard risk. In paragraph 1.72 of the 2017 Transfer Pricing Guidelines for Multinational Enterprises and Tax Administrations, the OECD defines hazard risk as:

COVID-19 presents a hazard risk beyond the “normal” hazard risks that can be anticipated — notably, the OECD does not mention the hazard risk of a pandemic in its definition — and it differs by size and the simultaneity of its worldwide impact from the 2008 financial crisis and the SARS outbreak. COVID-19 is also not within the range of normal business-cycle crises.

In a transfer pricing context, transacting parties within the same group obviously did not anticipate (for example, in their contractual arrangements) the question of which parties would bear a risk like COVID-19. Even third parties did not anticipate this risk, which will likely lead to a wave of disputes among them.

If the allocation of such a risk could not be anticipated, the typical transfer pricing framework should apply to establish appropriate transfer pricing adjustments. An analysis of the contribution of parties within the value chain to determine how third parties would have negotiated under similar circumstances can thus be relevant.

We note the external nature of the risk makes it more difficult — that is, compared with other strategic, financial, and operational risks — to isolate which entity within the value chain has control of the risks, as paragraph 1.60 of OECD’s guidelines requires.

If it is clear that transacting parties could not anticipate the risk, and could not truly control it, the effects the situation had on each individual retailer’s business must be investigated to determine if and to what extent the loss suffered by the retailer should be shifted with the principal and, if yes, how. Importantly, the OECD does not entirely exclude the possibility that a routine entity may end up in a loss position. Paragraph 1.129 of the OECD guidelines states:

There are different ways to evaluate a one-off 2020 adjustment to the usual transfer pricing policy. One could fairly assume that the determination should be grounded in the specific financials of the MNE itself, as well as relevant benchmarks, if at all possible.

Many published articles discuss alternative techniques for identifying more reliable comparables to be used for testing the year 2020 results based on the assumption that the 2017 to 2019 data will not be suitable because they will not reflect the impact of the crisis. The following approaches, inter alia, have been proposed: adjustment of historical comparables through regression analysis, use of forecast comparables (for example, leveraging public data of representative Chinese- or Hong Kong-listed retailers), reexamination of the existing set, and moving from interquartile to a full range.⁴

However, we believe that testing the retailer’s full-year results, even using a suitable benchmark, may not necessarily offer the most reasonable results in compliance with the arm’s-length principle. Instead, the specific features of each of the three distinct segments of the 2020 financials — the pre-lockdown, lockdown, and post-lockdown periods — should be analyzed in isolation; the underlying facts are different, and they may call for different solutions to reach the heart of this exceptional situation.

Keeping in mind the profile of our hypothetical retailer, the following key principles and suppositions should serve as a framework for this analysis:

• COVID-19 Is Not a Reason to Discard the TNMM: The extraordinary crisis alone does not justify a change in transfer pricing methodology; any departure must be independently justified by a change in the functional and risk profile of the parties. Also, a change in method is not needed to handle loss sharing between principals and retailers. The TNMM is not disqualified: Instead, it must be adapted to the exceptional circumstances, and a one-off adjustment to the transfer pricing policy may be justified.

• Identification of Decision Takers (Centralized Versus Decentralized Groups): MNEs have had to make many decisions as they manage the COVID-19 crisis, and they will have to make more. A clear understanding of who within the group made these decisions may tell a lot about which parties are in control of the decisions and thus should bear the related consequences. For example, some MNEs deliberately decided not to avail themselves of particular public subsidies for reputational reasons.

• Retailer Loss – Optical Effects: When it comes to the loss shown in the retailer accounts, one should carefully consider the extent to which: (a) the loss is because of COVID-19 (if a drop in profits started before the pandemic, including because of the use of a transfer pricing methodology that wasn’t fully consistent with the TNMM, one should isolate those losses); and (b) the loss has already been shifted to the principal, at least in part. This may happen when a full price return policy is in place. While such policy is in principle consistent with the application of TNMM, there may be situations where — especially for the spring/summer season — the returns exceed the historical average and serve, per se, as an extraordinary support from the principal for local retailers.

• Relationships With Unrelated Parties (Such As Wholesalers): These relationships could be relevant (but not decisive) when testing the reasonability of the approach taken. Examples of relevant matters include the sustainability of losses left with the retailer, special contributions by the principal, and price changes during the year. That noted, one should bear in mind that the original choice of the TNMM already implicitly disqualified the arrangements with these unrelated parties from serving as reliable internal comparables (otherwise the CUP method would have been selected).

• Consistency: The final position of the retailer must be consistent with the way it was treated in the past and will be treated in the future. Any reduced profitability (or loss) should not be grounds to call for a higher profit than was reported in past years when group sales were strong, nor should it justify a change in the functional profile in future years (for example, treating a retailer as if it became more entrepreneurial and should be entitled to higher returns in excess of applicable interquartile ranges).

While we present one method for apportioning COVID-19 losses among parties — a method that we believe has several merits — we also acknowledge that other solutions could achieve similar results.

A Proposal for Allocating COVID-19 Losses

A Case Study: Key Facts

Tables 1 and 2 show the difference between the pre-pandemic budgeted sales, costs, and profits of a retailer and the expected sales, costs, and profits of the same distribution affiliate as reforecast in early June, assuming the transfer pricing policy between the principal company (which supplies the goods) and the distributor is unchanged. The gross margin is kept at 50 percent in both.

Table 1. Pre-Pandemic Budget

Budget 2020	Retailer		Jan	Feb	Mar	Apr	May	June	July	Aug	Sep	Oct	Nov	Dec
Turnover		1,200	100	100	100	100	100	100	100	100	100	100	100	100
Cost of Goods Sold	50%	600	50	50	50	50	50	50	50	50	50	50	50	50
Gross Margin	50%	600	50.0	50.0	50.0	50.0	50.0	50.0	50.0	50.0	50.0	50.0	50.0	50.0
Selling and Labor Costs	28%	336	28.0	28.0	28.0	28.0	28.0	28.0	28.0	28.0	28.0	28.0	28.0	28.0
Rents	11%	132	11.0	11.0	11.0	11.0	11.0	11.0	11.0	11.0	11.0	11.0	11.0	11.0
Advertising	2%	24	2.0	2.0	2.0	2.0	2.0	2.0	2.0	2.0	2.0	2.0	2.0	2.0
G&A	5%	60	5.0	5.0	5.0	5.0	5.0	5.0	5.0	5.0	5.0	5.0	5.0	5.0
Operating Expenditures	46%	552	46.0	46.0	46.0	46.0	46.0	46.0	46.0	46.0	46.0	46.0	46.0	46.0
Operating Margin	4%	48	4.0	4.0	4.0	4.0	4.0	4.0	4.0	4.0	4.0	4.0	4.0	4.0
Operating Margin %			4%	4%	4%	4%	4%	4%	4%	4%	4%	4%	4%	4%
Assumptions: Sales of 100 per month. No inventory, no seasonal effect. 100 percent of goods is purchased from the group (Principal). No currency risk. Target margin 4 percent - Median (TNMM). Benchmark Q1 2 percent - Q3 6 percent.

Table 2. Post-Pandemic Reforecasts — No Change of Transfer Pricing Policy (No Change of Markdown)

P&L Retailer	2020 Budget			Pre- COVID-19			Lockdown Period				Post-Lockdown Period								2020 Reforecasts
				Jan	Feb		Mar	Apr	May		June	July	Aug	Sep	Oct	Nov	Dec
Turnover		1,200		100.0	100.0		0	0	0		50	60	70	70	70	80	80		100%	680
Cost of Goods Sold	50%	600		50.0	50.0		0	0	0		25	30	35	35	35	40	40		50%	340
Gross Margin	50%	600		50.0	50.0		0	0	0		25	30	35	35	35	40	40		50%	340
Selling and Labor Costs	28%	336		28.0	28.0		8.4	8.4	8.4		25.2	25.2	25.2	25.2	25.2	25.2	25.2		38%	258
Rents	11%	132		11.0	11.0		11.0	11.0	11.0		11.0	11.0	11.0	11.0	11.0	11.0	11.0		19%	132
Advertising	2%	24		2.0	2.0		1.0	1.0	1.0		2.0	2.0	2.0	2.0	2.0	2.0	2.0		3%	21
G&A	5%	60		5.0	5.0		4.8	4.8	4.8		5.0	5.0	5.0	5.0	5.0	5.0	5.0		9%	59
Operating Expenditures	46%	552		46.0	46.0		25.2	25.2	25.2		43.2	43.2	43.2	43.2	43.2	43.2	43.2		69%	470
Operating Margin	4%	48		4.0	4.0		-25.2	-25.2	-25.2		-18.2	-13.2	-8.2	-8.2	-8.2	-3.2	-3.2		-19%	-130
Operating Margin %				4%	4%						-36%	-22%	-12%	-12%	-12%	-4%	-4%
			Profit = 8 Profit = 4%				Loss = (75)				Loss = (62) Loss = (13%)								Loss = (130) Loss = (19%)
Note: Summary figures in the table are rounded. Assumptions: Lockdown period: March through May 2020. Shops are closed by law during that period. Key figures (as a percentage of budget): Sales — 0 percent. Selling expenses (labor costs subsidized by the government): 30 percent. Rents: 100 percent. Local advertising: 50 percent. General and administrative expenses: 95 percent. Post-lockdown period: June throught December 2020. Key figures (as a percentage of budget): Sales (gross margin unchanged): from 50 percent to 80 percent. Selling expenses: 90 percent. Rents: 100 percent. Local advertising: 100 percent. General and administrative expenses: 100 percent.

We acknowledge that the case study is a simplification and that financial data shown will not represent the situation of all retailers. The calculations are for illustration purposes only. We also note that local distributors will have different abilities to reduce costs in crisis times. For instance, rents in the United States tend to be more variable than rents in Europe. Another example is labor market flexibility: Retailers would have more flexibility to make staff redundant in Asia than in France, Germany, Italy, or the European Nordic countries.

Table 3. Decisions/Facts Matrix

COVID-19-Decisions	Principal	Retailer
Closure	NA (decided by law)	NA (decided by law)
Not avail of subsidies	NA in this casea
Labor costs (hiring, freeze, dismiss)	NAa	X
Shop layout and operations (reduced hours, protection gear)/investments: temporary measures	Limited	X
Shop layout/operations/investments: ongoing	X	Limited
Rent (unilateral/negotiated reduction, suspension of payments)	X	Limited
Local advertising (unchanged/reduced)	X	Limited
G&A reduction	X	Limited
Sell-out price and discounts	X
Orders for autumn/winter	X	Limited
Full price returns versus restrictions in the sale to outlets of unsold products	X
E-commerce sales	Flat/no sizable increase	Not involved
Other information
2019 results	Profit	Profit
2020 group results	Profit
aThe principal may govern this decision given wider reputational concerns. bSimplified assumption.

The Proposal

We propose the following logical steps and an approach that segments the financial result in three sections (pre-lockdown, lockdown, and post-lockdown).

Year 2020

Pre-Lockdown

For this period, TNMM is applied as usual with any seasonality factored in by reference to past years. In the simplified facts of our example, no adjustment is needed because budget was hit.

Lockdown

The financial consequences of the lockdown period are in principle isolated; their causes are investigated; and they are treated accordingly:

• Any cost that stems from the principal’s decisions should be borne by the principal. Examples include a decision not to avail itself of subsidies for reputational reasons or a decision not to dismiss employees for ethical rather than business reasons.

• To the extent any meaningful increase in the principal’s e-commerce sales to customers in the retailer’s region is clearly the result of the closure of physical stores, it may necessitate shifting a portion of the operating margin earned by the principal to the retailer.

• Temporary costs that the retailer incurs to allow it to resume operations may remain with the retailer because they pertain to local day-to-day operations — that is, unless the principal requires the retailer to take extra steps, possibly as part of a global strategy, in which case the principal may pay the relevant costs. A different conclusion may apply to permanent measures, although that will be part of a wider discussion on the ongoing profitability of the retailer in the post-pandemic climate.

• Except as specifically noted below, the rest of the loss should remain with the retailer.

Post-Lockdown

Again, financial results for this period are isolated and duly analyzed and benchmarked. In principle, there may be two cases⁵:

• The retailer may quickly return to pre-pandemic levels of sales, ending up in a profitable position before any transfer pricing adjustment. In this case, the same approach that was used pre-lockdown may apply. If the level of profit is below the first interquartile of the (unchanged) 2017-2019 set of comparables, there may be room to not make an adjustment depending on the reason for these results and provided that this approach is supported by a revised set.

• The retailer makes a slow and gradual recovery, ending the year in a loss position (this is the result shown in our case study). In this case, we believe the use of tailored comparables — for example, targeting start-up companies — may support the position that the retailer’s targeted profit for the period should be set at break-even (or slightly more) and a profit adjustment made accordingly.

Sanity Checking

The final full-year 2020 result should be checked to confirm the following:

• sustainability: any loss should be recoverable over the following three to five years, and a multiyear approach should be applied to confirm an average profitability in line with the comparables;

• reasonability: corroborative analysis possible may be possible, for example, through full-year forecast comparables or regression analysis);

• consistency: both with prior and future years (retailer to remain a routine entity); and

• tax-related considerations: includes potentially aggressive positions or those based on explicit guidelines from local authorities as well as customs duty implications.

Year 2021 and Beyond

A wait-and-see approach is advisable for 2021 and beyond.

Proposed Post-Pandemic Adjusted Results

Applying the proposed solution to the hypothetical case study leads to Table 4, which shows adjusted results for the retailer. For the sake of simplicity, we assume that the loss in the lockdown period is left entirely with the retailer. However, according to the foregoing discussion, part of it could well be shifted to the principal under specified circumstances.

Table 4. Adjusted Results

P&L Retailer	2020 Budget			Pre- COVID-19			Lockdown Period				Post-Lockdown Period									2020 Adjusted
				Jan	Feb		Mar	Apr	May		June	July	Aug	Sep	Oct	Nov	Dec	H2 Adj
Turnover		1,200		100	100		0	0	0		50	60	70	70	70	80	80			100%	680
Cost of Goods Sold	50%	600		50	50		0	0	0		25	30	35	35	35	40	40			50%	340
Gross Margin	50%	600		50.0	50.0		0.0	0.0	0.0		25.0	30.0	35.0	35.0	35.0	40.0	40.0			50%	340
Selling and Labor Costs	28%	336		28.0	28.0		8.4	8.4	8.4		25.2	25.2	25.2	25.2	25.2	25.2	25.2			38%	258
Rents	11%	132		11.0	11.0		11.0	11.0	11.0		11.0	11.0	11.0	11.0	11.0	11.0	11.0			19%	132
Advertising	2%	24		2.0	2.0		1.0	1.0	1.0		2.0	2.0	2.0	2.0	2.0	2.0	2.0			3%	1
G&A	5%	60		5.0	5.0		4.8	4.8	4.8		5.0	5.0	5.0	5.0	5.0	5.0	5.0			9%	59
Operating Expenditures	46%	552		46.0	46.0		25.2	25.2	25.2		43.2	43.2	43.2	43.2	43.2	43.2	43.2			69%	470
Operating Margin	4%	48		4.0	4.0		-25.2	-25.2	-25.2		-18.2	-13.2	-8.2	-8.2	-8.2	-3.2	-3.2	-62.4		-19%	-130
Transfer Pricing Adjustment*																		62.4		9%	62
Final Operating Margin				4.0	4.0		-25.2	-25.2	-25.2		-18.2	-13.2	-8.2	-8.2	-8.2	-3.2	-3.2	0.0		-10%	-67
Operating Margin %				4%	4%						-36%	-22%	-12%	-12%	-12%	-4%	-4%	0%
	Profit = 48 Profit = 4%			Profit = 8 Profit = 4%			Loss = (75)							Profit = 0 Profit = 0%						Loss = (67) Loss = (10%)
Note: Summary figures in the table are rounded. *The transfer pricing adjustment can be made in various ways leading to the targeted profit.

The adjusted results show a margin of -67 on a full-year basis, instead of -130 (the result based on the reforecast).

The term “lost profits” refers to the profits lost as a result of COVID-19 compared with the initial budget margin. Compared with the budget margin (established pre-pandemic), the total lost profits kept at the distributor level under our proposed method is -115 (that is, a -67 loss instead of 48 profit).

Using the proposed method, the principal company (supplier of the goods) absorbs/bears
-63 of the market losses. Thus, out of a total loss of -130, -67 is left with the retailer. This corresponds to 48 percent of the total loss under the reforecast and 35 percent of the total lost profits under the reforecast, which equal 178 (a reforecast loss of
-130 versus a budgeted profit of 48).

The loss left with the retailer appears reasonable. In fact, assuming the sales get back to pre-pandemic levels in 2021, and also assuming a 5 percent margin in 2021, 2022, and 2023,⁶ the average operating margin in 2020-2023 would be 2.64 percent,⁷ which would likely fall within the range of the expected set of comparables.

We believe the proposed solution has the following benefits:

• It preserves the transfer pricing policy and method while acknowledging the extraordinary features of the hazard risk being faced in 2020, which also calls for a one-off adjustment of the policy (rather than a “business as usual approach” implemented through adjusted benchmarks).

• It limits the otherwise huge shift of losses to the principal, which could severely affect the principal’s dividend distribution capacity and the covenants agreed by it with financing banks. Given its size, such a shift would also likely trigger scrutiny by the tax authorities.

• It factors in the centralized and decentralized features of the decision-making process and control over risk to the extent doing so makes sense (some occurrences are out of the parties’ control).

• It is mindful of the sustainability of the retailer’s financial position and looks at market practice as well as what independent parties would have done.

Finally, it is worth highlighting that aside from the analysis of the retailers’ economic results, there are also financial considerations that must be taken into account — funds must be provided to retailers to finance their operations. To this end, a choice must be made between equity contributions and intercompany loans considering local tax restrictions on interest deductibility and the risks of recharacterization of debts (as highlighted in the 2020 OECD guidelines on financial transactions⁸ and in the practice of several jurisdictions). While analyzing this issue is beyond the scope of this article, we want to highlight that (a) there are interconnections between this issue and the foregoing discussion on the transfer pricing of goods, and (b) simply keeping intercompany trade payables outstanding, although pragmatic, may not be the safest long-term solution.

Conclusion

This article offers a conceptual path that some MNEs may follow to deal with the extraordinary situation of COVID-19 — a path that is consistent with OECD guidelines and TNMM philosophy. The proposal has the merits of being practical, justifiable from an economic and arm’s-length perspective, and allowing room for a degree of sophistication in the analysis that factors advanced standards on decision-making and risk control, to the extent applicable. As always in transfer pricing, it may be necessary to confirm the outcome of the method by corroborating analyses.

We fully acknowledge that we are at the forefront of a new scenario — and possibly a new normal — that poses new challenges and questions in the transfer pricing area. We hope this article is a useful contribution and that the OECD will hopefully build on this proposal to provide timely guidelines to the operators, thus relieving companies from the pain of the technical uncertainty.

Celerity and pragmatism in the release of guidelines at all levels, general consensus among most advanced jurisdictions, reasonability in future audits phase, and the proper functioning of mutual agreement procedures are more essential than ever to get through this period — and even turn it into an opportunity to build an enhanced transfer pricing world.