Saumyanil Deb and I noted that the UK Diverted Profits Tax should allow taxpayers a chance to defend its intercompany pricing based on a reasonable model of the intercompany transaction in question. In an earlier paper, I posed a simple model of the appropriate gross margin for a sales affiliate, which we applied to the Google issue. I will offer a similar comment on the recent UK ruling by the High Court of Justice in a matter between Glencore Energy UK Ltd and the Commissioners of HM Revenue and Customs. While this ruling was a procedural decision, Chris Lallemand provided a summary of the issue:
The principal activity of GENUK is trading in oil and gas from the UK. This involves transporting oil from producers, storing it to benefit from mismatches in supply and demand, and the transformation of oil products through refining and blending operations. The company employs around 41 traders. It had turnover in 2015 of approximately £28bn and accounting profit before tax of around £36m. GENUK is an indirect subsidiary of Glencore international AG (GIAG), a Swiss resident subsidiary of Glencore plc …GIAG provides GENUK with working capital financing, the assumption of net losses suffered by GENUK that remain after GENUK has hedged and insured trading transactions. It also covers a range of other services including priority access to the worldwide network of fuel offices of GIAG and subsidiaries, storage and transportation facilities and regular detailed consultation and advice relating to market risk, counter-party credit risk and financing matters. In return for these services GENUK pays GIAG 80% of its net profit.
While the precise details of this dispute are not clear, the following table offers one interpretation. We have assumed that third party sales are £28 billion per year and that the overall enterprise pays third party suppliers 98.85 percent of sales thereby earning a 1.15 percent overall gross margin. The overall enterprise incurs operating expenses equal to 0.5 percent of sales with 80 percent of those expenses borne by the UK affiliate and the remaining borne by the Swiss affiliate. Under the transfer pricing policy at issue, the UK affiliate retains gross profits equal to 0.53 percent of sales and hence operating profits equal to 0.13 percent of sales, while the Swiss affiliate retains gross profits equal to 0.62 percent of sales and operating profits equal to 0.52 percent of sales.
|Millions of pounds||Consolidated||UK||Swiss|
|Cost of goods||27678.0|
The Canadian Revenue Agency (CRA) evaluates situations like this arrangement in terms of a co-distribution model. While consolidated operating profits represent a 130 percent markup over operating expenses, the markup for the UK affiliate is only 32.5 percent and the markup for the Swiss affiliate is 520 percent. The CRA approach often is to afford both affiliates with an equal markup over operating expenses, which in this case would grant the UK affiliate with 80 percent of the profits rather than only 20 percent of the profits. We shall argue, however, that this approach ignores the role of which legal entity holds the working capital. Brad Rolph summarizes and critiques this CRA approach:
The CRA applies the Berry ratio, another distribution-related PLI, to the “own costs” of limited-function distributors …Since it is generally difficult to determine with certainty what returns are associated with which distribution activity, it follows that each of the co-distributor entities would earn the same return given this data limitation. This return is usually measured by reference to the Berry ratios realized by comparable distributors…. It is submitted that service providers do not earn the same return as the distributors to which they are providing services, and vice-versa. Since each type of entity performs different functions, assumes different risks, and owns and employs different assets (e.g., distribution rights and working capital), it stands to reason that the return to each entity would be unique and consistent with the arm’s-length principle. Accordingly, each of the service providers in the example presented above should realize mark-ups on costs that are consistent with arm’s-length returns for service providers, not distributors. Similarly, limited-function distributors should earn mark-ups on costs that are consistent with arm’s-length returns for distributors, not service providers.
If the Swiss affiliate is the distributor holding working capital while the UK affiliate is effectively a commission agent, one might be able to defend the higher markup received by the Swiss affiliate. My presentation of the standard model for sales affiliates was noted in my earlier paper:
GP/S = a + b(OE/S) + r(WC/S), where S = sales, GP = gross profits, OE = operating expenses (or more correctly value-added expenses), and WC = working capital (inventory + trade receivables – trade payables). This equation has three key parameters:
the marginal Berry ratio (b), which will generally be less than the average Berry ratio (GP/OE);
the return to working capital (r); and
profits attributable to intangible assets (a).
This model can be stated alternatively in terms of the operating margin:
OP/S = a + m(OE/S) + r(WC/S), where OP = operating profits and m = b – 1.
Let’s assume the working capital to sales ratio for this Glencore operation was 10 percent and the return to working capital (r) = 5 percent. For the overall operations, the markup over operating expenses not counting the return to working capital (m) would be 30 percent. Applying our model to the facts for Glencore Energy UK would suggest its commission rate should be 0.52 percent. As such, the actual transfer pricing policy was likely appropriate if not generous.
Glencore’s application for judicial review was declined on the grounds that the UK tax authority questioned whether Glencore adequately addressed the tax authority’s concerns. We have treated this issue in terms of arm’s length pricing and have speculated on what the precise facts were. Whether the UK affiliate should be treated as a commission agent versus a distributor depends in part on the intercompany contractual arrangements with respect to its products in the UK. While we do not know whether Glencore has presented a transfer pricing analysis, we offer up this possible interpretation as an illustration of how a reasoned transfer pricing analysis may address controversies such as this one.