RoyaltyStat Blog

Transfer Pricing Memento

Posted by Ednaldo Silva

We started (1988) in transfer pricing oriundo from academia, and the IRS dominant paradigm was forcing single point estimates of gross profit indicators in audit. This IRS practice included gross profit margin (expressed as a percent of net sales) for controlled inbound distributors, and gross profit markup (expressed as a percent of cost of goods sold (COGS)) for controlled outbound manufacturers. The Berry ratio (gross profits/operating expenses (XSGA)), introduced by Charles Berry in Du Pont, was an unspecified “fourth” method applied to service providers (like DISA) because they may not recognize labor costs in COGS. Functions performed are reflected in COGS and XSGA. See United States Court of Claims. E. I. Du Pont de Nemours and Company v. The United States, Nos. 256-66 & 371-66, April 18, 1978. (Cite as: 1978 WL 3449 (Ct. Cl. Trial Div.)). 

Another fourth method, introduced by Irvin Plotkin in Du Pont, used as a test of reasonableness in litigation, was based on net profit rates of return on assets, without using comparables. Plotkin's multi-industry net profit rates of return on assets was not used to establish arm's length profits. Plotkin used net profits before tax (i.e., pretax net profits after interest deductions) from large samples of multi-industry listed non-financial enterprises. This pretax net profits assay method was morphed into the BALRM (basic arm's length return method) on the well-researched White Paper (1988); however, BALRM was rejected in the 1994 US regulations despite the ITC (International Tax Counsel) efforts to make it live.

We were instrumental driving nails on BALRM's coffin because we held that assets are an unreliable base to determine arm's length operating profits. Aside from different cross-enterprise assets composition, which is difficult to adjust to comparable standards, we face different asset vintages (different initial or acquisition costs) and different depreciation, amortization and depletion schedules. Moreover, regarding assets as questionable profit bases, we face nebulous capital theories that confound assets as stock with a flow variable like CAPX or investment that (by implausible assumption) gravitates across enterprises in multple industries to a uniform rate of return. E.g. see: Robert Solow, Capital Theory and the Rate of Return, Rand McNally, 1964. Can't find DOI, but see:

Ruling when we started, the 1968 US transfer pricing regulations specified three methods:

(1) comparable uncontrolled price (CUP),

(2) resale price method (gross profit margin) applied to controlled importers, and

(3) cost plus method (gross profit markup) applied to controlled exporters.

Under the IRS industry economist assistance program to international examiners (IE), large case transfer pricing audit consisted of using gross profit industry-level indicators obtained from MANA Surveys or primarily from RMA Annual Statement Studies. As restated below, no systematic analysis of enterprise-level comparables were used, and IRS economists had no access (subscription) to Compustat. See

Until we introduced the comparable profits method (CPM) in 1989, which we later difused in the new APA Program, no enterprise-level comparables were employed by IRS audit economists. Together with the CPM, we introduced Standard & Poor's Compustat enterprise-level financials to find comparables. We got this idea (during our prior to IRS teaching post-graduate economics) from the US enterprise net profit research of Dennis Mueller, “Persistence of Profits Above the Norm,” Economica, New Series, Vol. 44, No. 176 (November 1977). See Mueller on Stable URL: 

We extended this US enterprise-level profit analysis to satisfy transfer pricing comparability standards prompted by legal advice from Christine Halphen, Kim Palmerino and John Dean (later appointed Tax Court Judge). As attorney advisors from the IRS Office of Chief Counsel, they suggested that to be defensible under audit scrutiny we should refocus our enterprise-level economic analysis from net profits after interest deduction to operating profits because controlled interest deduction was regulated under thin-capitalization rules. Later we discovered that Plotkin used Compustat enterprise-level financials in Du Pont, but not to find comparables as already mentioned above. In contrast to previous practice, while working at IRS, we employed Compustat-sourced comparables to determine arm's length operating profits of controlled importers that showed substantial profit shifting to offshore affiliates. 

In IRS audit of controlled distributors that imported foreign goods for resale in US, we discovered that profit shifting was not restricted to overcharged Purchases recognized in COGS. We discovered persistent profit shifting in operating expenses (XSGA), including excess tax deductions for controlled professional, scientific, and technical services, for management and administrative support services, for outbound royalties, and for advertising expenses. After acquiring more audit skills, we discovered hidden royalties and management fees embedded in COGS or XSGA unstructured other costs and expense schedules.

Because of verified double-dipping, we introduced a new transfer pricing method called CPI (comparable profits interval) and then renamed CPM (comparable profits method) in US, which was later copied as TNMM by OECD. Unlike unreliable gross profit methods, the novel CPM is based on operating profits because we can correct intercompany profit shifting in COGS or XSGA. After we introduced the CPM in US (called TNMM by OECD), gross profits methods, including the Berry ratio, were debunked and became legacy baggage. These traditional gross profit methods are unreliable and should not be used in transfer pricing practice. Gross profit methods should be junked because they reflect an incipient idea proved to be unreliable and now they are superseded by operating profit methods such as CPM and profit split.

We also introduced statistical intervals to test comparable operating profit indicators (instead of then prevailing point estimates) and requirements that transfer pricing analysts representing tax administration and corporate taxpayers must demonstrate reliable estimates of arm's length taxable income.

Discovery of excess advertising expenses led to specific rules regarding “assistance provided to the legal owner of intangibles,” i.e., new rules distinguishing legal versus economic beneficial owners of intangibles and challenges to intercompany outbound royalty payments when the controlled licensee incurs deductible advertising expenses.

Published on Nov 11, 2018 4:40:36 PM

Ednaldo Silva (Ph.D.) is founder and managing director of RoyaltyStat. He helped draft the US transfer pricing regulations and developed the comparable profits method called TNNM by the OECD. He can be contacted at:

RoyaltyStat provides premier online databases of royalty rates extracted from unredacted license agreements
and normalized company financials (income statement, balance sheet, cash flow). We provide high-quality data, built-in analytical tools, customer training and attentive technical support.

Schedule an Online Demo