States Should Learn from Transfer Pricing History, but Focus on The Right Lessons

Barbara Rollinson (Horst Frisch), Howard Berger, Robert Culbertson and Kandyce Korotky (Covington and Burlington) assert that state tax authorities should use the Comparable Profits Method (CPM or Transactional Net Margin Method in OECD nomenclature) with care in the evaluation of transfer pricing for tangible goods. This is a reasonable, if obvious, position. However, some of the examples cited, including the recent Coca Cola case, in their piece are misplaced for reasons we will address.

The Tangible Goods Lesson: Limited Risk Does Not Mean Zero Risk

Transfer pricing practitioners often use the notion of limited risk entities to justify low margins, but mistakenly conflate this with a scenario that resembles something closer to no-risk entities. 

For instance, the authors describe an intercompany pricing issue for tangible goods with this example:

Company A and Company B are related. Company A buys products from Company B for 300x. Company A sells the products to third parties for 400x, incurring 92x of costs in doing so. Company A’s operating income is 8x and its operating margin is 2% (8x/ 400x). The state tax authority identifies a group of companies that provide marketing and selling services to third parties. Those comparables have 8% operating margins.

Let company B be a Texas based manufacturer of inexpensive tortillas, which its domestic sales affiliate (company A) sells to groceries stories throughout the U.S. for $1 per package. Annual sales are 100 million packages so sales = $100 million per year. Company A incurs $23 million in selling costs.

The authors further assume that company B’s production costs represent 71% of sales so consolidated operating profits are only 6% of sales or $6 million. This assumption is consistent with the cost of ingredients being $0.50 per package and labor costs being $0.21 per package. The following table considers three transfer pricing policies.

Table 1: Income Statements for Three Transfer Pricing Policies (in Millions)

The taxpayer’s intercompany pricing policy set the transfer price at $0.75 per package so that the gross margin for the distributor is 25%. The authors note that under this policy, its operating profits are only 2% of sales, which represents a 8.7% markup over selling costs. The markup over total production costs (including component costs and labor costs) is 5.33%.

If a tax authority insists on an 8% operating margin, then the distributor’s gross margin would be 31%. The policy would grant the distribution affiliate with a 34.78% markup over selling costs and leave the manufacturing affiliate with an operating loss.

Our table considers an intermediate policy that allows the distribution affiliate with a 26% gross margin. This policy is consistent with a 3% operating margin, which represents a 13.04% markup over selling costs. The policy would also afford the manufacturing affiliate with a 4.05% markup over total production costs.

For reasons I noted in an earlier discussion of state transfer pricing, I would argue that this intermediate policy is a more reasonable representation under arm’s length pricing even under the authors’s assertion that the distribution affiliate has a limited asset to sales ratio. The authors’ only defense for the 2% operating margin came in the form of this footnote:

Service companies like the comparables, with no COGS and operating margins of 8%, would have a markup on costs of 8.7%

The return to total cost for service companies was the approach taken by Lawrence Dildine in his failed attempt to defend the low operating margin for Tropicana Product Sales. He presented evidence from advertising agencies and logistic companies to defend a low return to value-added expenses for a high function distribution affiliates with a 1% operating margin. This approach was misleading because the total costs for such entities include a significant amount of pass-through costs.

The ratio of operating profits to total costs (z) overstates the ratio of operating profits to value-added costs (m), depending on the proportion of pass-through costs relative to total costs (p), while the ratio of operating profits to operating expenses (w) understates the return to value-added costs if cost of goods sold includes certain value-added expenses (s > 0):

m = z/(1 – p) and m = w[(1 – s - p)/(1 – p)]

Transfer pricing practitioners who assert that the appropriate profit level indicator for service providers is the return to total costs are implicitly assuming that their comparable companies have no pass-through costs (p = 0). This assumption, however, is often not valid. Consider a simple example where a service provider incurs $25 in value-added expenses and $75 in pass-through costs. If its revenues total $107.5, then the markup over total costs is only 7.5% but the markup over value-added expenses = 30%.

Transfer pricing practitioners note that the functions and assets of the distribution affiliate as determinants of the arm’s length gross margin, which can be captured by the following model:

GP/S = a + b(E/S) + r(A/S)

where GP is gross profits, S is sales, E is value-added expenses for the distribution affiliate and A is tangible assets for the affiliate, with the key parameters being the marginal b (marginal berry ratio), r (return to tangible assets) and a (profits attributable to any intangible assets owned by the distribution affiliate). This formula can be recast in terms of operating profits (P) relative to S:

P/S = a + m(E/S) + r(A/S)

where m = b - 1. The traditional CPM approach assumes that the distribution affiliate does not own valuable intangible assets (a = 0).

The selection of publicly traded alleged comparable distributors is meant to provide evidence on the key parameters for the above formula. For simplicity, let’s assume m = 10% and r = 5%. If the asset to sales ratio for the distribution affiliate is 14%, then our model suggests a 3% operating margin and hence a 26% gross margin. An 8% operating margin is a reflective of a distributor that performs more extensive functions, holds high tangible assets relative to sales, or owns valuable marketing intangibles.

This intermediate transfer pricing policy leaves the manufacturing affiliate with a markup over total production costs just over 4%. Note that most of these production costs represent the cost of ingredients with labor costs being approximately 30% of total costs. If production of tortillas requires modest assets relative total costs, then a 4% markup over total costs may be reasonable. Cambridge Brands v. Commissioner involved the markup over total costs for a contract manufacturer of inexpensive chocolates. Its markup was also modest, as the overall asset to total costs was modest.

The Evaluation of Intercompany Royalty Rates: CPM Abuse Cuts Both Ways

The issue in the Coca Cola litigation was not a tangible goods issue, but an issue of what the arm’s length royalty rate should be when consolidated operating profits were 45% of sales. While there were significant differences in how the various expert witnesses estimated the routine returns to production and distribution, these differences were secondary to the issue of how residual profits should be allocated between the US parent as the licensor and the foreign affiliate licensees. My discussion of this issue noted:

The IRS position rested on two propositions. The first proposition was that the overall routine return was only 5% of sales, which would suggest residual profits would be 40% of sales. While the taxpayer representatives severely criticized the specific approach used by the IRS expert witness, the taxpayer’s expert witnesses agreed that the routine returns for distribution and production were very modest. The other proposition is that the US parent is entitled to all of the residual profits in the form of intercompany royalties. This position more than doubles the intercompany royalty rate leaving income equal to only 5% of sales for both the production and distribution functions.The IRS position was an application of the comparable profits method (CPM).

I also noted two potential objections to the use of CPM to evaluate arm’s length royalty rates:

The primary taxpayer argument for lower royalty rates rested on the claim that the foreign licensees owned valuable marketing intangibles because the foreign affiliates paid the ongoing marketing expenses … Even if we accept the IRS premise that the US parent holds all of the valuable intangible assets, a lower royalty rate is suggested by a model grounded in sound financial economics that properly considers licensee risk and the implications for expected returns. My critique of CPM approaches notes that licensees deserve a portion of residual profits for two reasons: compensation for profits attributable to the value of intangible assets owned by the licensee; and compensation for the leverage risk of using valuable intangible assets owned by another entity.

Taxpayers often use CPM to justify high royalty rates. For example, Utah State Tax Commission v. See’s Candies involved a 16.5% royalty rate that left the related-party licensee with only a 5.5% operating margin.

The royalty rates in Sherwin Williams and Lowes’ were 3% of sales, which the expert witnesses for the New York Division of Taxation demonstrated were excessive. The expert witnesses for the taxpayer in both litigations tried to justify the 3% royalty rates by CPM reports that showed that the licensee affiliates receive routine returns for the functions they performed. The problem with the taxpayer position was not only did a mere routine return to tangible assets ignore the commercial risk borne by the related party licensee but also ignored the fact that the affiliate licensees owned valuable intangible assets. In the Sherwin-Williams case, the trademark holding company owned only the marketing intangibles whose value was less than the value of the product and process intangible assets retained by the licensee affiliates.

State transfer pricing structures often involve the transfer of trademarks to the intangible holding company, which charges an intercompany royalty rate that leaves the parent corporation with only a routine return for distribution and production functions. While CPM reports are often drafted to defend these high royalty rates, these analyses are flawed if the parent corporations own other valuable intangible assets or if the parent corporation bears licensee risk.

References

Howard Berger, Robert Culbertson, Kandyce Korotky and Barbara Rollinson, “States Can Learn Much From Transfer Pricing History Or Be Condemned to Repeat,” Tax Notes State, February 22, 2021.

Coca-Cola Co. v. Comm'r, 155 T.C. No. 10, November 18, 2020.

Harold Mcclure, “State Transfer Pricing: The Economics of a Few Key Issues,” Journal of Multistate Taxation and Incentives, May 2015.

Matter of Tropicana Products Sales, Inc., NYS Tax Appeals Tribunal, ALJ Determination, DTA Nos. 815253 and 815564, November 25, 1998.

Cambridge Brands, Inc. v. Commissioner of Revenue, Commonwealth of Massachussets Appellate Tax Board Findings of Fact and Report, 2003-358.

Harold McClure, “US IRS Prevails in Coca-Cola Transfer Pricing Dispute over Intercompany Royalties,” MNEtax.com, November 23, 2020.

Sees Candies v. Tax Commn, 2018 UT 57.

Sherwin-Williams Co. v. New York Tax Appeals Tribunal, 12 App Div 3d 112 784 NYS2d 178 2004 NY Slip Op 7737n2004 WL 2403994 (N.Y. Sup. Ct. App. Div. 2004). In re Lowe's Home Centers, Inc., N.Y. Div. Tax App., ALJ determination, No. 818411, September 30, 2004.