Transfer Pricing: Revolt of the Emerging Markets

Let’s face it. One important aspect of transfer pricing is that it is a competition among the government tax offices of various nations to grab as much share as they can of the income derived from an organization’s cross-border transactions.
Therefore, any difference in principles and application of transfer pricing regulations between two countries can result in serious issues for the company. Such differences were recently manifested among developing countries such as China and India (as summarized in Chapter 10 of the UN Practical Transfer Pricing Manual for Developing Countries published in October 2012).
While these positions and approaches may look provocative in the sense that they openly question traditional OECD or US-based transfer pricing regulatory frameworks, they do raise very important questions (if not perfect solutions) around the ‘practical’ issues companies face in transfer pricing, including even those involving developed countries.
In my view, these practical issues can be summarized as three important and inter-related aspects of transfer pricing:
  • lack of reliable comparables
  • location specific advantage (including market premium)
  • intangibles
In many cases, these are the three most important factors that must be dealt with in not-so-straightforward transfer pricing cases. For example, if a related-party’s actual performance falls outside of the prescribed arm’s length range, it is important to analyze the transaction from these three angles and determine if actual results can still be explained as arm’s length.
Challenging the status quo
The recent discussions appear to have common agenda – not very obvious, but indicative. Namely, they appear to challenge fundamental economic theory or principles applied in existing transfer pricing framework.
Modern finance theory has been heavily criticized and questioned particularly after the financial crisis in 2008 and the significant growth of emerging countries, where trade and financial markets are not necessarily considered as free or fully efficient.
For example, the China paper strongly argues that too much weight should not be given to risk as a driving factor in determining return as arm’s length profit. The India paper further goes on to strongly criticize the Capital Asset Pricing Model (CAPM), which forms the theoretical backbone of modern finance theory as well as some of the comparability adjustments in transfer pricing analysis.  
As we know, the application of TNMM (transactional net margin method) or CPM (comparable profits method) is common in transfer pricing, even though profit split has been used in complex cases. Arguably, finance theories (including risk-return relationship and CAPM) support the use of TNMM and CPM based on available market data.
The presumption is that the market-observed return correctly reflects the risks assumed by companies – inexact comparability can be fixed by applying CAPM-based adjustments.
Usually in TNMM or CPM, external comparables are selected and applied as benchmark to derive arm’s length profit, or the profit earned by a third party with sufficiently similar function and risk profile.
The selected and calculated profit level indicator, most often return on sales (ROS), can be used as benchmark because of the underlying assumption that companies should at least earn a return observed in the market that is consistent with function and risk. Otherwise they cannot continue as a going concern in the long run.
If there are no special circumstances such as the use of intangibles, location-specific advantage and market premium, competition would erase any excess profit. Companies are not thus entitled to a return that is higher than the observed return.
Although direct evidence of profit level indicators is found in the actual financial results of selected comparables, these results are theoretically supported by the hypothesis that companies with more functions and more risk should earn a higher return.
If directly applicable and appropriate comparables cannot be found, the financial data of available comparables can be used after adjusting their observed return, taking into account the difference between the comparables and a tested party. CAPM again provides a framework to estimate the theoretical magnitude of such an adjustment, by estimating return for difference in function and risk, which is believed to be manifested in market data (for example, equity beta).
Lack of comparables
There are increasingly larger numbers of publicly traded companies in each major country whose financial data are available and reliable as benchmark. However, serious questions have been raised with regard to whether these companies can be good comparables.
It has been pointed out that many companies become multinationals in order to enjoy economies of integration, both horizontally and vertically, or to reduce cross-border transaction costs. In the past, these complex factors were assumed to be absorbed by the parent company. It was also assumed that all subsidiaries outside the home country could be evaluated as isolated, single-function entities that were relatively easy to benchmark.
These rather simplistic views are now being questioned by tax authorities in emerging countries. India, in particular, emphasizes that research and development service providers should earn profits commensurate with total value-chain profits. They cannot be easily compared with local external independent service providers in isolation.
We at Alix Partners have come across this kind of discussion in developed countries as well. But in general we have concluded that it is not necessary to analyze the total profit of the value chain at both ends of the transaction, as long as the function and risk profile is clearly analyzed. Local, simple-function independent comparables can be applied to assess arm’s length profit.
Thus the question ultimately comes down to the classic problem of what are the appropriate comparables.
But finding comparables has never been easy. It is probably over-simplistic to argue that comparables can be selected almost automatically through electronic databases. Such an approach seems to be close to just applying seemingly appropriate numbers without understanding underlying transactions and the difference from available comparables.
In US intellectual property litigation, the use of industry standard (for example, the standard royalty rate for the auto-parts industry or any other similarly fixed percentages) has been now heavily criticized by the courts. The same probably applies to transfer pricing analysis.
At the end of the day, depending on importance and complexity, we should pay appropriate attention and make reasonable efforts to select and apply appropriate comparables.
It is true that the globalization of business and the emergence of a more flexible relationship between parent and subsidiaries in sharing function and risk make the selection of comparables more difficult. But the efforts to find appropriate external and internal comparables should not be easily discarded.
As I explain later, the alternative of resorting to profit split is an even more difficult task.
Location Specific Advantage
Location Specific Advantages (LSAs) may be summarized as the specific price or cost advantages due to the specific location. This encompasses very difficult elements such as the meaning of ‘quality of labor force.’
It is not realistic to assume, for example, that workers in developed countries are ten times as productive as those in emerging countries because they earn ten times more. Productivity-adjusted labour cost differential can be considered as part of LSA so long as the incremental profit due to the cost differential is not “passed over” to the next level of the value chain.
If there is no complete pass-through, there may be LSAs that can be considered. In attributing LSAs, the question is whether they should remain in the local country or should be recaptured by an affiliate that is providing special technology so that other companies cannot compete (which is why an LSA is observed, to start with).
The incremental profit may have to be primarily attributed to the owner of the technology. Thus, discussion of LSAs must address such issues as where exactly the relevant LSAs come from, and how related parties contribute to the creation of profits associated with LSAs.
It is not very clear how the distinction between LSAs and intangibles should be made. But in deciding the attribution of LSAs, intangibles can be an important factor.
In discussions within the OECD, it appears to be clear that this group of developed countries is inclined to define intangibles rather broadly, not confined only to legally protected IP such as patents and trademarks.
How to distinguish intangibles from other source of excess profit such as LSA and market premium is also challenging. This distinction is related to the question regarding intangibles in transfer pricing, i.e. attribution of intangibles between related parties.
When the attribution can be judged to be just to one side of the parties, it is often easy to evaluate intangibles. In fact, there is no need to explicitly evaluate the intangibles owned by the other party. The party without intangibles is evaluated simply by using external comparables.
However, when intangibles are deemed to be attributable to two or more related parties, how to attribute the given intangibles is often very difficult. To begin with, it is not easy to identify what “profit” needs to be analyzed.
The relationship between analyzed profit and intangibles need to be established so that appropriate relative attribution can be quantitatively determined. Contribution analysis is increasingly important for this purpose.
Contribution analysis consists of an analysis of the relevant expenditures or capitalized amount of such spending, man-hours spent, strategic importance, third-party availability and relative bargaining power.
It is often hard to find detailed accounting or engineering evidence directly supporting such analysis, but best effort, including the use of expert opinion, should be fully explored.
Recent developments in transfer pricing at the UN are potentially a good source to estimate where and global transfer pricing regulations and practices are heading. The ideas and concepts advanced by emerging economies such as China and India may become part of the new thinking around transfer pricing.
The OECD or developed countries may have given too much emphasis on analysis based on modern finance theory, resulting in one-sided analysis based on CPM/TNMM. While there are still lot of cases where the CPM/TNMM analysis is appropriate, the questions are often and increasing made whether such analysis is always applicable, and if not, how ‘exceptions’ should be treated.
But when the number of exceptions increases so much, they are no longer exceptions. It is important for companies, advisors and tax authorities to consider and develop an alternative framework to incorporate the reality of more complex and integrated multinational operations of global companies.
We need to go back to the basics and strive to really identify the sources of profit and develop the best transfer pricing models that are rooted in the realities of today.
About the Author

Nobuo Mori is Managing Director of AlixPartners, a global business-advisory firm offering comprehensive services in four major areas: enterprise improvement, turnaround and restructuring, financial-advisory services and information-management services. Visit for more information.