When evaluating a company for transfer pricing purposes, it is common to find that there are no functionally comparable companies located in the same country. As a result, foreign companies with similar functions are often used as comparables. Given this, the idea of making additional adjustments to account for structural differences between comparable companies and the analyzed company has been a persistent recommendation—even by the OECD (2018). A broad set of factors, such as the legal and tax framework, human capital, natural resources, and geographical location, can influence the profit margin a company can achieve and cause it to vary depending on the country where it operates.

However, it is of interest to determine whether there are significant differences in profit margins depending on the country in which companies are located. If not, it would be sufficient to consider companies with the same functions as comparables without requiring any additional jurisdiction-based adjustments.

For example, Meenan (2004) conducted a statistical analysis to assess whether the interquartile ranges of net margin distributions for companies in developed European countries differ significantly from those of individual countries within the same region. The study concluded that, in most cases, the differences were statistically insignificant.

The following sections will discuss some key points related to this topic, and a model will be proposed to evaluate whether a company’s country of operation is a sufficiently relevant variable to justify a country risk adjustment in transfer pricing analysis.

RISK-RETURN TRADEOFF

The risk-return tradeoff hypothesis suggests a positive relationship between expected returns and asset risk. This is because investors will only be willing to invest in a riskier asset if they expect higher returns compared to a less risky asset. As a result, investment options that compensate for higher risk with a premium in expected returns will persist in the market.

Merton (1980) highlights that, in practical terms, there is a positive relationship between the mean and variance of the market portfolio, referring to expected returns and risk, respectively. However, various econometric analyses have produced mixed results regarding this proposed relationship between risk and return.

For instance, Campbell (1985) finds a negative relationship between risk, measured using conditional variance, and returns. Similarly, Baillie (1990) employs GARCH-in-mean models with data at different frequencies to study the relationship between portfolio returns and their conditional variance or standard deviation, consistently finding a weak relationship between the two variables.

Conversely, Guo (2001) finds that market return variance has a strong predictive power for excess market returns—i.e., the spread between portfolio returns and the risk-free rate. Engle (1987), using a series of ARCH models, confirms the theoretical relationship between the conditional variance of a long-term bond and its return.

It is important to note that risk, in quantitative terms, is an unobservable variable. This has led to a wide variety of models designed to measure asset risk. Consequently, empirical analysis of the risk-return relationship is challenging, as risk is not a condition that can be perceived entirely subjectively by economic agents, unlike the observable returns obtained from an asset or economic unit.

DEVELOPED VS. EMERGING COUNTRIES

Another topic of analysis in recent years has been the contrast between developed and emerging markets. The question arises as to whether there are significant differences in investment performance between these two groups. Emerging markets are generally characterized by social, economic, and institutional conditions that make them perceived as riskier investment destinations compared to developed countries.

Given that emerging countries may be structurally riskier than developed ones, the risk-return tradeoff hypothesis suggests that emerging markets should offer higher returns to compensate for the additional risk.

In this regard, Salomons and Grootveld (2003) find a significant difference in the risk premium (the return over the risk-free rate) between emerging and developed markets in financial assets. They also observe that this difference follows a cyclical pattern rather than being associated with structural changes.

However, it is crucial to recognize that these analyses are often based on financial market asset prices, which reflect market expectations about companies rather than solely their actual profitability. Nevertheless, real company performance also plays a crucial role in how their equity securities are valued by economic agents.

COUNTRY ADJUSTMENTS IN COMPARABILITY ANALYSIS

In the context of transfer pricing, the implications of the discussed concepts relate to fundamental differences in profit margins between similar companies due to the different risks they face in different countries. This could mean that an adjustment is necessary when incorporating a foreign company as a comparable for a company being analyzed.

For example, Starkov, Gonnet, and Pletz (2014) conduct two case studies. In the first, they adjust the margins of comparable companies located in Europe to analyze a related entity in Sub-Saharan Africa using an adjustment based on working capital and accounts receivable levels. In the second case, they adjust margins based on the cost of capital for two emerging markets (China and India) while using U.S. companies as comparables.

In both cases, due to the higher risks faced by the analyzed companies compared to the comparables, the adjusted profitability range is higher than the initial range before applying the adjustment.

DATABASE DESCRIPTION

The database was constructed using TP Catalyst data and is described as follows (updated as of March 2019):

• Active companies
• Companies with financial data from 2018
• Publicly traded companies
• Companies with a website
• Companies with available net cost-plus (CAN) data
• Companies with available operating margin (OM) data
• Companies located in the following countries:

Developed countries: Australia, Austria, Belgium, Canada, Denmark, Finland, France, Germany, Ireland, Italy, Japan, Netherlands, Norway, New Zealand, Portugal, Singapore, South Korea, Spain, Sweden, Switzerland, United Kingdom, United States.

Emerging countries: Brazil, Chile, China, Colombia, Egypt, Greece, Hungary, India, Indonesia, Malaysia, Mexico, Peru, Philippines, Poland, Russia, South Africa, Taiwan, Thailand, Turkey.

From this selection, a sample of 7,609 companies was obtained.

KOLMOGOROV-SMIRNOV TEST

To summarize, the hypothesis under evaluation is whether profit margins of companies in emerging markets differ significantly from those in developed markets. The Kolmogorov-Smirnov test is used to determine whether the probability distribution of margins in both groups differs across sub-industries.

The null hypothesis states that both samples exhibit the same distribution and is rejected if the following condition is met:

Results from the Catalyst database show that, for both ratios analyzed, the Kolmogorov-Smirnov test rejects the null hypothesis in only 9 out of 45 sub-industries, meaning that only 20% of the cases exhibit a statistically significant difference in the distribution of ratios between developed and emerging countries.

ANOVA MODEL

In addition to the previous analysis, we use an ANOVA model to test whether profit margin differences follow the expected theoretical pattern for country risk adjustments.

The results indicate that, in only 11 sub-industries, there is a statistically significant difference in CAN, with higher margins observed in emerging countries in 9 of these cases. Similarly, for OM, significant differences are found in 11 cases, with 10 indicating higher margins in emerging markets.

The model structure is as follows[1]:

The model takes companies located in developed countries as the benchmark. Therefore, the beta parameter captures whether there is a difference in the profit margins recorded by companies in emerging markets compared to those in developed countries in the j-th industry.

Table 2 presents the model results for each sub-industry, where it can be observed that in the case of CAN, only in 11 instances does the model indicate a statistically significant difference between the ratio value of a company in an emerging market and a developed country. In 9 of these cases, the results suggest that the ratio is higher in emerging markets compared to developed ones.

Similarly, in the case of MO, we find very similar results. As in CAN, there are only 11 instances where a statistically significant difference is observed in the ratio value depending on the company’s location (in an emerging or developed country). In 10 of these cases, the model indicates that the MO ratio is higher if the company is located in an emerging market compared to a developed country.

Overall, these results are quite similar to those obtained with the Kolmogorov-Smirnov test, as only about one-fifth of the cases analyzed show an impact of the company’s location on the ratios used.

It is also important to highlight that in cases where localization impact was found, the results align with expectations based on literature (tradeoff). That is, profit margins tend to be higher in emerging markets compared to developed countries.

CONCLUSIONS

Based on the above, we can conclude that there is no strong statistical evidence to support the hypothesis of a structural difference between developed and emerging countries that would consistently result in higher profit margins for a company solely due to its location in an emerging market rather than in a developed country. In a significant majority of cases, no substantial effect of this variable was observed.

For transfer pricing purposes, we might reconsider the suitability of establishing the country risk adjustment as a standard practice. Given the results obtained, it may be that when a prior functional selection of comparables is conducted, applying an additional country risk adjustment may not be entirely appropriate in a general sense, as this effect seems to be neutralized.

  BIBLIOGRAFÍA.

Baillie, R. y Deggenaro, R. (1990). “Stock Returns and Volatility”, Journal of Financial and Quantitative Analysis, vol. 25, issue 02, 203-214.

Campbell, J. (1987). “Stock Returns and the Term Structure,” Journal of Financial Economics, Vol. 18, No. 2, pp. 373-399. 

Guo, H. (2001). “Understanding the Risk-Return Tradeoff in the Stock Market”, Federal Reserve Bank of St. Louis, Working Paper, 2002-001A.

Engle, R., Lilien, D. y Robins, R. (1987). “Estimating Time Varying Risk Premia in the Term Structure: The Arch-M Model”, Econometrica, vol. 55, issue 2, 391-407.

Merton, R. (1980). “On estimating the expected return on the market: An exploratory investigation”, Journal of Financial Economics, vol. 8, issue 4, 323-361.

Meenan, P., Dawid, R. y Hülshorst J. (2004). “Is Europe One Market? A Transfer Pricing Economic Analysis of PanEuropean Comparables Sets”, Deloitte White papers, European Commission.

Organización para la Cooperación y el Desarrollo Económicos (2018). “A Toolkit for Addressing Difficulties in Accessing Comparables Data for Transfer Pricing Analyses”, The Platform for Collaboration on Tax, Author.

Salomons, R. y Grootveld, H. (2003).  “The equity risk premium: emerging vs. developed markets”, Emerging Markets Review, vol. 4, issue 2, 121-144.

Starkov, V., Gonnet, S., Pletz, A. y Maitra, M. (2014). “Comparability adjustments”, Transfer Pricing international journal, Bloomberg BNA, ISSN 2042-8154.

Publicado en el número especial de precios de transferencia de la revista IDC, Asesor Fiscal, Jurídico y Laboral.

Las opiniones expresadas en este artículo son emitidas a partir de los considerandos señalados en el mismo, y no deben aplicarse a casos específicos sin el debido cuidado y revisión del contexto particular del contribuyente en cuestión. Este artículo no representa una opinión en particular sobre caso concreto alguno, y sugerimos consultarnos para la revisión de casos específicos.

[1] El modelo es una regresión lineal simple en la que se incluye una variable binaria para capturar el efecto que tiene la ubicación de la empresa en sus márgenes de utilidad. Se calcula una regresión por cada industria a evaluar tomando como datos las empresas de países desarrollados y emergentes que se encuentran en esa industria. Así, se obtiene el resultado de una regresión por cada industria para analizar si la localización de las empresas tiene realmente impacto en sus utilidades.

[2]  Se considera el margen de utilidad registrado por la compañía en la base de datos de TP Catalyst. De esta forma se prueba con tres distintos múltiplos utilizados regularmente en el análisis de precios de transferencia, que son: Costo adicionado neto (CAN) y Margen operativo (MO).

Autor: José Chamorro