Comparability adjustments: In the absence of suitable local comparables in emerging and developing economies – Case Studies

Vladimir Starkov, Sébastien Gonnet, Amanda Pletz and Madhura Maitra, NERA, Chicago, Paris and London  

In the absence of suitable local comparables in emerging and developing economies - Case Studies

This series of two articles explores the options for improving comparability in cases where domestic comparables are not available. Following a request by the G8 summit held in June 2013, the OECD has just published a paper on Transfer Pricing Comparability Data and Developing Countries, which discusses four possible approaches to addressing the concerns over the lack of data on comparables expressed by developing countries, among which there is a discussion on the use on non-domestic comparables and comparability adjustments.

In the first article,1 we presented theoretical approaches, while in this second article we provide practical application and examples. As mentioned in the first article, the lack of local comparables is often addressed by practitioners by using suitable comparables from other jurisdictions. Yet, when such comparables come from countries with significantly different economic conditions than the country of the tested party, some adjustments to account for these differences are called for. This article discusses practical application of comparability adjustments for cases when non-domestic comparables have to be used by presenting two case studies.

• The first case study discusses calculation of an arm's length distribution margin for a related party in the Sub-Saharan region of Africa, focusing more specifically on how to account for the differences associated with accounts receivable and the resulting differences in credit risk supported between the tested party and the comparables that were identified in Europe.

• The second case study describes the adjustment based on the cost of capital and provides an illustration of how this adjustment could be used to compute both the median and the end points of the arm's length range of return on capital employed for two emerging markets (China and India) when the comparables come from a developed market (the US).2

Case Study A
Adjustments for differences in working capital

A Multinational Corporation (MNC) operates its main product design and manufacturing facilities in a country A, a developed country. In order to enter a new market, such as Sub-Saharan Africa, the MNC group establishes a subsidiary in a country B within the region. The main role of the entity located in country B is marketing and sales of the goods developed and manufactured by the entity in country A.

The group employs a Resale Price Method, where the transfer price is set by employing a target discount factor applied to the distributor's sales to unrelated customers. The discount factor is inferred from the results of independent comparables.

The group has a consistent transfer pricing policy worldwide, and would like to employ the same transfer pricing policy in the country B located in the Sub-Saharan region.

As is often the case when dealing with operations in emerging countries, independent comparables with suitable financials are not available in the country B, consequently other approaches need to be considered. One such approach is to consider another region where comparables with available financial data may be found (e.g., in Europe), select functionally suitable comparables, and perform appropriate adjustments.

Below, we provide the description of the adjustment related to differences in accounts receivable.3 Although this adjustment is common in transfer pricing, the differing levels of credit market risk between the region of the tested party (i.e., Sub-Saharan Africa) and the region where the comparables are derived from (i.e., Europe) add an additional layer of complexity to this case.

Rationale for the accounts receivable adjustments

In the case at hand, given the lack of local comparables, European comparable companies were identified.

These comparable companies operate in a safer economic environment. Most of them have long-standing relationships with their customers, have been operating for a long time, and have credit assessment tools to confirm the financial viability of their existing and potential customers. As a result, the comparable companies have a track record of limited bad debts and relatively fast receivables collection, as reflected in their balance sheets.

On the other hand, the tested party is a relatively new entity with newly acquired customers and limited tools to assess the financial capacity of its existing and prospective customers. The tested party also faces a higher level of political, market, and credit risks within the region. This is reflected in a significant amount of accounts receivable, higher cash requirements, more lenient repayment requirements to establish its client base and higher potential for bad debt.

Therefore it is economically justifiable to make an adjustment for the difference in risks between the comparables and the tested party.

Unadjusted results  

The unadjusted profit results of the sample of European distributors are provided in Table 1:

Table 1: 3-year weighted average RoS*

  Return on sales
(European comparables)
First Quartile 6.4%
Median   8.9%  
Third Quartile 15.4%
* European comparables

As discussed above, these profit margins take into account payment terms of companies operating in Europe. Therefore, the comparables operate in a safer (less risky) and more stable business environment. The need for an adjustment arises because operating profit observed in uncontrolled transactions (at the level of the comparables) does not incorporate any risk premium that would typically be observed in emerging countries to reflect for the above described uncertainties and market characteristics.

Adjustments performed

Working capital adjustments performed for comparables operating in the same countries or regions as the tested party involve assessing the differential in accounts receivable and remunerating this difference with an appropriate interest rate that takes into account credit risk.

In the case at hand, the companies operate in a region where not only the economic circumstances, but also the underlying credit risk, are significantly different from the tested party, requiring an accounting for the differences in the level of accounts receivable between the tested party and the comparables, as well as differences in the interest rates between the regions of the tested party and the comparables4.

For simplicity purposes we divide the adjustment into two distinct steps:

1. The first step consists of adjusting the accounts receivable of the comparables down to zero and applying a Eurozone interest rate to the differential, which results in lower revenues for the comparables and a lower margin (in line with the lower risks associated with carrying no accounts receivable)

2. The second step consists of adjusting the accounts receivable of the comparables up from zero to the level of the tested party and applying to them an interest rate of the Sub-Saharan country or region, which increases the revenues for the comparables and results in a higher margin (in line with the higher risks associated with carrying higher accounts receivable in the Sub-Saharan region)

 

The two-step approach is necessitated by differences in local credit conditions. For example when evaluating 3 month short term sovereign bond yields across various countries, we note significant variations in sovereign rates. Similarly, when evaluating short term corporate rates one would see significant differences. Chart 1 shows a comparison of three month sovereign bond rates across a sample of randomly selected countries:BNAI Chart One

The first step is described below:

• Equation 1: Balance sheet adjustment

BNAI Eq1
 

Where:  

ΔART = change in accounts receivable when setting the target (T) days accounts receivable to zero

Comparables Days in AR is the days in accounts receivable of the European comparables

The "0" is the target days receivable to remove the impact of days receivable

 

•  Equation 2: Income statement adjustment  

BNAI Eq2
 Where:

ΔSalesT is the adjustment to sales after removing the impact of accounts receivable

ΔART is the impact on accounts receivable estimated as part of the balance sheet adjustment

i(non-domestic) is the short term interest rate reflecting the underlying credit risk (in Europe)

After having removed the working capital related effects of operating in the European market from the comparables accounts, the next step is to introduce the working capital related impact of operating in the more risky developing market. It corresponds to adjusting the comparables' accounts receivable (that have been adjusted to zero) up to the level of the tested party's accounts receivable, and applying an interest rate corresponding to the country/region of operations of the tested party. In particular, an interest rate should be chosen that represents a typical short term financing of receivables in the local market (i.e., country B). The approach to obtain such a rate would typically depend on the available credit market data for a particular country and industry. Then, formulas similar to the above can be employed again.

• Equation 3: Balance sheet adjustment

BNAI Eq3
 

Where:  

ΔARLT = change in accounts receivable when setting days of accounts receivable to the tested party days of receivables

Target Comparables Days AR is the days in accounts receivable of the tested party

The "0" is accounts receivable of the comparables following the above described first step of the adjustment

 

• Equation 4: Income statement adjustment

 

ΔSalesLT = ΔARLT * ilocal

Where:  

ΔSalesLT is the adjustment to sales by adjusting the accounts receivable in line with the local target company

ΔARLT is the impact on accounts receivable estimated as part of the balance sheet adjustment

ilocal is the short term interest rate of the local market in which the target company operates (Sub-Saharan region)

Adjusted Results

Table 2 provides the adjusted comparables data after adjustments.

Table 2: Three-year weighted average return on sales (unadjusted, adjusted)

  Pre-adjusted weighted average return on sales Adjusted to "Zero" accounts receivable thereby removing local/European market credit conditions: Weighted average return on rales Adjusted to tested party level of accounts receivable and local/Sub-Saharan market credit conditions: Weighted average return on sales
Comparable 1 17.2% 16.8% 22.7%
Comparable 2 8.0% 7.4% 13.3%
Comparable 3 9.9% 9.0% 13.8%
Comparable 4 5.8% 5.2% 11.1%
Comparable 5 69.0% 68.3% 69.9%
Comparable 6 5.6% 4.4% 9.7%
1st Quartile   6.4%   5.7%   11.6%  
Median   8.9%   8.2%   13.5%  
3rd Quartile   15.4%   14.8%   20.5%  

This economic analysis shows that comparable distributors operating in Europe earn a return on sales between 6.4 percent and 15.4 percent, with a median return of 8.9%. This range, however, does not reflect arm's length results appropriate for the tested party operating in Sub-Saharan Africa. The results adjusted to the tested party's level of accounts receivable and the local market credit conditions do represent such arm's length range of returns as the adjustments account for the higher accounts receivable and greater credit risks in the tested party's country. Based on these adjustments, the target return on sales should be between 11.6 percent and 20.5 percent, with a median return of 13.5 percent.

It should, however, be noted that the above adjustment does not account for the increased volatility that may be observed in profit margins of companies that operate in emerging markets. An analysis adjusting for volatility between markets can be relied upon as an alternative approach.

Case Study B
Cost of capital adjustment5

As the name implies, the cost of capital adjustment corrects for the differences in costs of capital across different geographic regions in which the comparables and the tested parties operate. This adjustment is necessary because the profitability of a company is directly related to its cost of capital, and companies operating in markets with different costs of capital will have different profitability requirements. In case there are significant differences in cost of capital, margins observed for the comparables in a non-domestic location will not appropriately reflect those that can be expected for the tested party. Thus, an inter-country cost of capital adjustment is required to account for the fact that the comparables are based in a developed country, for example, in the US, whereas the tested party is based in an emerging country, such as China or India. In general, the cost of capital varies not only across countries but also over time and the cost of capital adjustment described below attempts to capture both.

We demonstrate the cost of capital adjustment for two different emerging countries' capital environments (i.e., China and India) using an illustrative set of 16 independent US publicly traded durable goods distribution companies. When carrying out cost of capital adjustment, we chose the US as the benchmark country and China and India as “target” countries to which the adjustments are made. We also chose December 31 as the fiscal year end (i.e., the fiscal year end for the hypothetical tested party in either country). The numerical examples below are prepared using a five-year (2008-2012) weighted average Return on Capital Employed (ROCE) as the PLI.

The operating profit of each comparable in each year was adjusted to account simultaneously for inter-period and inter-regional differences in the constituent elements of the cost of capital, such as the cost of debt, the risk-free interest rate, the equity risk premium, the corporate tax rate, the level of financial leverage, and the betas of the stocks of the comparable companies,6 as follows:

 

• Equation 5

 

Adjustment to operating profit = (WACCT - WACCC) × Capital Employed

Where:  

WACC T = pre-tax weighted average cost of capital (WACC) of a comparable company C adjusted to the market conditions of the target country T (e.g., China or India) and adjusted to the target fiscal year end, i.e., December 31 of 2008-2012, computed as follows:  

BNAI Eq4  

i = the month of the fiscal year end for the tested party (i.e., December  

j = the month of the fiscal year end for the comparable company C

LGB RateTi = the 10-year government bond rate in the target country in month i

Beta = 1

ERPTi = equity risk premium in the target country in month i (source: Morning Star Cost of Capital Report)

DCj/CCj = debt to capital ratio for comparable C in month j

CB RateTi = corporate bond yield in the target country in month i

TTi = corporate tax rate in the target country in month i

WACCC = weighted average cost of capital of comparable company C in its own country (i.e., the U.S. in this case) and its own fiscal year end, found a follows:

BNAI Eq5  

LGB Rate C j = the 10-year government bond rate in the comparable's country in month j  

Beta = 1

ERPCj = equity risk premium in the comparable's country in month j

CB RateCj = corporate bond yield in the comparable's country in month j

TCj = corporate tax rate in the comparable's country in month j

Table 3 shows the five-year (2008-2012) weighted average ROCE for the 16 US companies with and without Cost of Capital (WACC) adjustment.

Table 3: Five-year weighted average ROCE with and without cost of capital adjustment*

Company Name ROCE without Cost of Capital Adjustment ROCE with Cost of Capital Adjustment (China) ROCE with Cost of Capital Adjustment (India)
Company 1 13.7% 13.9% 23.7%
Company 2 26.9% 27.0% 36.9%
Company 3 40.8% 41.6% 53.4%
Company 4 -2.2% -2.2% 4.5%
Company 5 1.4% 1.4% 9.4%
Company 6 -5.8% -5.5% 7.6%
Company 7 53.5% 53.9% 66.8%
Company 8 -15.1% -14.5% -3.1%
Company 9 11.4% 11.8% 25.6%
Company 10 1.5% 2.0% 15.7%
Company 11 24.5% 25.1% 41.0%
Company 12 1.8% 2.4% 12.4%
Company 13 16.1% 16.3% 29.7%
Company 14 14.5% 15.5% 29.2%
Company 15 11.1% 11.4% 22.9%
Company 16 58.3% 58.9% 74.3%
First Quartile   1.4%   1.7%   10.9%  
Second Quartile (Median)   12.5%   12.9%   24.7%  
Third Quartile   25.7%   26.1%   39.0%  
Mean   15.8%   16.2%   28.1%  
Standard deviation   20.7%   20.8%   22.0%  
Width of Interquartile Range   24.3%   24.4%   28.1%  
* Note: The data are for the years 2008-2012

From Table 3 we see that the interquartile range without cost of capital adjustment is from 1.4 percent to 25.7 percent with a median of 12.5 percent. Once a cost of capital adjustment has been made to China's capital environment, the outcome remains very similar to the original one. The interquartile range becomes 1.7 percent to 26.1 percent with a median of 12.9 percent. When a cost of capital adjustment is made to India's capital environment, the interquartile range shifts up to become between 10.9 percent and 39.0 percent with a median of 24.7 percent.7

When the cost of capital adjustment is made from a country with a lower cost of capital to a country with a higher cost of capital, profitability range given by the returns of comparables shifts up, consistent with the basic finance theory of risk-return trade-off which implies that an investor investing in a riskier investment would ex ante expect a higher return.

However, in higher-risk countries there is also a higher probability to both lose and gain more, so one would expect to find a wider interquartile range of returns when making adjustments to a higher-risk country. Thus, one shortcoming of the cost of capital adjustment discussed above is that it does not take into account such volatility of market outcomes i.e., it does not explicitly adjust the width of the interquartile range to the target country's risk environment. Thus, a more appropriate adjustment technique should widen the range as well as increase the median and standard deviation when the target country has a higher level of risk than the country where the comparables are located. One such method for risk adjustment can be found in the article by Curtis and Ruhashyankiko8which shows that under the assumption that comparables' returns are normally distributed, adjustments can be made to the standard deviation of the outcomes to calibrate the range of the comparables' profitability results. Such an adjustment would complement the cost of capital adjustment discussed above and ensure that increased risk will be reflected in increased standard deviations and larger predicted interquartile ranges. However, the empirical analysis conducted in that article calculates the risk-return adjustment in the context of the US market only. Expanding this analysis beyond the US market is a topic of further research.

Conclusion

In application of profit-based methods of transfer pricing analysis, cases when comparables are not available in the country of the tested party may present challenge to practitioners. This series of two papers is aimed at describing some options for improving comparability when comparables from other jurisdictions are used to calculate profitability range of a local tested party. This series discusses, from the theoretical and practical perspectives, several such adjustments including the working capital adjustment, adjustment for the difference in the cost of capital among countries, adjustment related to the differences in accounting standards, and adjustment for the economic cycles. The applicability of these adjustments to a particular case will depend on facts and circumstances, and, thus, such adjustments should be used judiciously.


Vladimir Starkov and Sébastien Gonnet are Vice Presidents at NERA Economic Consulting, in Chicago and Paris, respectively, and Amanda Pletz is a Senior Consultant in London. Madhura Maitra is a former NERA employee. They may be contacted at:
Sebastien.Gonnet@nera.com
Vladimir.Starkov@nera.com
Amanda.Pletz@nera.com
www.nera.com

NOTES

1 “Comparability adjustments in the absence of suitable local comparables in emerging and developing economies”, Bloomberg BNAsTransfer Pricing International Journal, Vol.14, No.7, July 2013.

2 The authors acknowledge that returns on capital in different countries may have different volatilities, which may impact the width of the interquartile ranges when an adjustment for the difference in the cost of capital is made between different countries. However, this type of adjustment is not illustrated in this paper.

3 Other adjustments, such as adjustments for accounts payable and inventories may be necessary, as well as adjustments associated with the differences in the cost of capital between the regions, as discussed in section B of this article.

4 It should be noted that to manage the risk factors that firms operating in emerging markets are exposed to, an additional functional return in some circumstances may need to be considered.

5 A detailed account of the theoretical motivation behind the cost of capital adjustment is provided in the first article of the series, “Comparability adjustments in the absence of suitable local comparables in emerging and developing economies”, Bloomberg BNAsTransfer Pricing International Journal, Vol.14, No.7, July 2013.

6 In the examples shown below, betas for all the companies have been assumed to be 1. The market beta for the companies, if available, may also be used.

7 The larger cost of capital adjustment for India than for China is due to the higher yields on the government and corporate bonds in India as well as higher expected equity risk premium in India than in China.

8 “Risk-Adjustments to the Comparables Range”, S. L. Curtis and J. F. Ruhashyankiko, BNA's Transfer Pricing International Journal Vol.4, No.8, August 2003.