ANALY SIS (Vol.12,No23) s11 is a common situation in the smaller European coun-I 3. Finalizing the Comparability Analysis tries and in Germany, where available data is limited Although the database contains over 1.000, 000 com- One of the last step anies for any given taxpayer. Many companies are pany's functions,risks,characteristics, etc.using, for subsidiaries of multinational or national companies, so esults could be influenced by transfer pricing issues the shares. shareholder owns more than 49 9 percetinn or d s iling mhs his analysis is based on specified crite- Companies were eliminated from further consideration ia d lost important comparability facto re the taxpayer/tested party profits. Typically ompanies that may have unique facts and circum- the transfer pricing practitioner will review the previous stances inarily, depending on the facts and criteria that were employed to confirm the facts. but the cumstances, potentially rable companies classi- greater amount of factual information will allow the fied as"small companies"based on the relevant Euro practitioner to further assess the comparability of the t be considered. All com anies acts and circumstances than otherwise could be the with a date of incorporation less than five years ago are case. For the research in the next section, this last step excluded from the set, because their results could be af- was not performed. The statistical rationale for this fected by start-up activities. Finally, any companies that proach is that the authors needed to ensure there were did not report their financial data for at least the pre ceding three years are excluded statistically reliable conclusions could be generated 2. Refining the Comparability analysis D Summary Once the high-level database comparability analysis Since there is no clear definition of a market in the ECd guidelines that would allow someone to deter- is performed, comparability assessment turns to a more mine whether a pan-European or country-specific com pecific focus on the relevant functional and risk profile for the potentially comparable companies. For example parables analysis is appropriate for applying the TNMM NACE codes--statistical classifications of the economie and related methods, the authors have developed an empirical testing approach to analyze this issue. This activities of companies in the European Union-are se- approach is based on the TNMM comparability criteria lected to broadly describe the relevant functions and product characteristics of the taxpayer. NACE codes as articulated in the OECd guidelines and a standard eral of which are often assigned to companies, help ach to performing TNMM analyses based on the ECd guidelines. The research and the results of the the transfer pricing practitioner to determine the prod- statistical tests are presented in the next section uct categories in which the whether the company is classified as a manufacturer distributor, service company, etc. Next, keyword searches, which will ordinarily specify"key"facts that V Empirical Analysis of should be included or excluded in a “ s Europe One Market? rable company's description, are specified and pro- cessed.For example, if the search is for a wholesale dis- scribes the results of their empirical anarsearch and de- The following details the aut tributor, keywords that describe a company as perform ing "retailing, " manufacturing, "or"proprietary per takes two hes to the autho tical test research"will likely cause the company to be rejected ing. The first approach, referred to as from further consideration as potentially comparable arability tests(Specific Tests), is designed to generate epending on the results of the keyword search, high testable comparability data that closely replicates the evel screening of certain diagnostic ratios may also be standard comparability analysis process described in helpful in assessing comparability. For example, a high Section rV. The effort required to generate comparabil- fixed assets-to-sales ratio could indicate that a poten ity data under the Specific Tests approach is significant, tially comparable company is involved in manufactur- so that the authors have concentrated the Specific Tests g rather than wholesale distribution Or a large per on selected functional and risk comparability areas that ent age of interest income and expense could indicate analysis we have performed over the last several earsy. For each functional and risk profile chosen, comparable companies are selected by using comparability selec of le comparability analys tion criteria that match the characteristics of the spe- cific profile. The authors then developed arms-length database ranges by country and on a pan-European basis to sta- n an" ability to control"criteria, but th specific and must be performed when factually analyzing in istically test whether the country-specific arm's-length dual companies. Also note that consolidated parent compa- ranges deviate from the ranges for the pan-Europear set of comparable companies nies were not eliminated based on this comparability criteria 4z For example, financial statements that may not be au wners could be compensated by The data used in this research is available to interested alary or as resid claimants in the profits, etc al parties upon request, subject to the approval of Bureau van TAX MANAGEMENT TRANSFER PRICING REPORT ISSN 1C63-2069 NA TAX 414 04