Transfer Pricing Issues in Italian Imports of Russian Natural Gas

Italy’s Supreme Court remanded on May 17, 2022 a case (No. 15668/2022) to the Regional Tax Commission of Lombardy in which the Italian tax authorities had challenged the transfer pricing between Promgas S.P.A. and Gazprom Export, finding alleged deficiency in the tax authority’s application of the Transactional Net Margin Method (TNMM). Prior to being considered by the Supreme Court, Promgas' had won its appeal to the Tax Commission of Milan, but the Regional Tax Commission of Lombardy

Promgas performed a mere intermediary function between Russia’s Gazprom Export as an exporter of natural gas and the Italian company Edison s.p.a. as the final purchaser of natural gas. Sales in 2005 were approximately €300 million and the intermediary affiliate’s operating margin was just under 0.25%. The Lombardy Regional Tax Commission asserted that the operating margin should be approximately 1.4%. Russia’s tax rate at the time was only 24% and was lowered in 2009 to 20%. Italy’s tax rate was 37.3% but was lowered to 31.4% in 2008 and further lowered to 24% in 2017. Since Italy’s tax rate was higher than Russia’s, one can understand why Gazprom would wish to limit profits for Promgas.

The litigants agreed that TNMM using the financials for Promgas was an appropriate transfer pricing policy, but disagreed with the application of TNMM performed by the tax authorities. The key concern of the Supreme Court was the selection of alleged comparables by the tax authorities. The Supreme Court agreed that the companies selected as alleged comparables did not meet the requirements in relation to the business activity carried out and independence. The Supreme Court also noted that while an application of the Resale Price Method (RPM) could be used to determine the appropriate gross margin, its preference would be an application of TNMM as long as there was appropriate consideration of the expenses of the intermediary affiliate and its asset intensity in comparison to the key financial ratios of the alleged comparable companies.

While these important considerations were mentioned, the Supreme Court decision did not state the operating expenses to sales ratio for Promgas or its asset intensity. In an earlier blog post, we noted an inappropriate application of the RPM for a similar case involving a Panamanian distribution affiliate selling imported fuel oil. In this litigation, the tax authority tried to raise the gross margin of the affiliate from 3.75% to over 8% even though the operating expenses represented only 2.75% of sales.

In our discussion of this Italian litigation, let’s assume that the intermediary affiliate’s operating expenses = 1.5% of sales. The following table considers gross margins as low as 1.75% representing the taxpayer’s position to as high as 3%. The tax authority’s position is equivalent to a 2.9% gross margin. The following table presents the implied operating margin as well as the markup over operating expenses.

Alternative Transfer Pricing Policies for Promgas

The taxpayer’s position that the gross margin should be only 1.75% is consistent with a 0.25% operating margin, which represents a 16.67% markup over operating expenses. A 3% gross margin would be consistent with a 1.5% operating margin, which represents a 100% markup. The tax authority’s position that the operating margin should be 1.4% implies a 2.9% gross margin and a 93.33% markup over operating expenses. Consider also an intermediate position that the gross margin should be 2.5% implying a 1% operating margin and a 66.67% markup over operating expenses.

This litigation represents an incomplete application of TNMM, but is not quite the misapplication of the RPM we discussed in an earlier blog post:

Administrative Tribunal, Case No TAT-RF-062 is a September 2020 decision involving a Panamanian affiliate that acts as a wholesale distributor of petroleum products. The Panama tax authority Dirección General de Ingresos (DGI) issued an assessment that resulted in the affiliate owing $14 mil. in additional tax payments. The taxpayer appealed, but the Tribunal ultimately ruled in favor of the tax authority … the taxpayer and the tax authority both focused on the gross margin for the distribution affiliate as compared to the gross margins of third-party distributors, and the court accepted the premise that the RPM was an appropriate approach.

The gross margin for the Panamanian affiliate was 3.75% while its operating expenses were 2.75% of sales. The DGI argued for an 8.5% gross margin based on the gross margins for third parties that had higher functions than the Panamanian affiliate. I noted the following model for the gross margin of a limited function sales affiliate under arm's-length pricing:

GP/S = a + b(E/S) + r(A/S),

where S = sales, GP = gross profits, E = operating expenses (or more correctly value-added expenses) and A = tangible assets held by the distributor. This equation has three key parameters:

• The marginal Berry ratio (b), which will generally be less than the average Berry ratio (GP/E);
• The return to tangible assets held by the distributor (r); and
• Profits attributable to intangible assets (a).

To illustrate, let b = 1.1 and r = 5%. Table 1 assumed E/S = 2.75%. While the court record did not specify the tangible asset to sales ratio for the distribution affiliate, let’s assume A/S = 14.5%. Under these assumptions, the arm’s length gross margin would be 3.75% if the distribution affiliate does not own valuable intangible assets. Our illustration justifies the 1% operating margin for the distribution affiliate on the grounds that its functions are very limited, and its asset intensity is quite modest. The tax authority’s position that the gross margin should be 8.5%, which implies a 5.75% operating margin, leads to an absurd conclusion that the markup over value-added expenses in excess of 200%.

This model can be applied to the Promgas issue if we knew the asset to sales ratio for the Italian affiliate. If this ratio were a mere 2%, then the taxpayer’s position would be supported by this model as the Italian affiliate would be both limited in functions but also very limited in asset intensity. Note, however, the tax authority’s position would be supported by this model if the asset to sales ratio were 25%. An intermediate assumption with respect to this affiliate’s asset intensity might be to assert that the asset to sales ratio were 17%. In this case, the model would suggest a 2.5% gross margin and a 1% operating margin under the arm’s length standard.

The Supreme Court decision itemized what it argued are the necessary steps for a reliable application of TNMM including:

... determination of the financial results relating to the controlled transactions … identification of comparable companies … accounting adjustments to the financial statements of the tested party and differences in accounting practices, provided that such adjustments are appropriate and possible; assessment of whether adjustments are appropriate or necessary to take account of differences between the tested party and the identified comparable companies in terms of risks assumed or functions performed; selection of a reliable profitability profit level indicator (so-called Profit Leverage/ Indicator, or PLI).

Our model for the evaluation of the arm’s length profit margin noted how the determination of the intermediary affiliates operating expenses to sales ratio and assets to sales ratio are key elements, while we made reasonable assumptions with respect to the return to value-added expenses and the return to tangible assets for an intermediary affiliate.

Standard practice, however, asks the transfer pricing analysts to select alleged comparable third-party sales affiliates. If these third-party entities had the same expense to sales ratio, the use of an operating margin versus a gross margin standard is a matter of convenience. Third-party distributors, however, often have higher functions than such a limited function affiliate, which not only suggests the analyst use TNMM over RPM, but also explores a capital adjusted markup over value-added expense approach. This litigation like the Panamanian suffered from the abuse of alleged comparables that may not have been reliable applications of either RPM or TNMM using a naïve operating margin approach.

Should the Distribution Affiliate be the Tested Party?

Natural gas prices in Europe have been very volatile. The following table shows one measure of these prices from July 2002 to June 2022. Over this period, the price of a million British Thermal Units (BTU) averaged $8.52 and has ranged from $2.52 during the summer of 2002 to $42.39 as of March 2022.

Natural Gas Prices in Europe from July 2002 to June 2020: Dollars per Million BTU

My 2014 discussion in the Journal of International Taxation of transfer pricing approaches for commodity transfer pricing explored three basic approaches including the use of the producer affiliate as the tested party. I critiqued this approach noting:

Trying to evaluate whether the intercompany price of mined goods is at arm's length by examining the profitability of a mining affiliate is akin to using a cost-plus approach to evaluate a pharmaceutical manufacturer responsible for R&D costs of new drug applications—swings from enormously profitable in one year to losses in another would be expected even under arm's-length pricing.

During periods of high commodity prices, this approach inappropriately grants the economic rents to distribution affiliate instead of the producing affiliate.

Tax authorities for the producer affiliate often advocate the use of the Comparable Uncontrolled Price (CUP) approach, which is also known as the Sixth Method. The usual application of the CUP approach envisions the use of spot markets for homogenous goods. Spot-market comparisons should be based on the market price for third-party transactions with the intercompany price on the same day if market prices exhibit day-to-day volatility. The CUP approach has the advantage of granting the economic rents from a commodity price boom to the producer affiliate. This approach, however, tends to ignore the value of the functions performed by the distribution affiliate. An adjusted CUP approach would deduct an appropriate gross margin from the third- party price received by the multinational for the commodity as determined either by an adjusted RPM analysis or an appropriate application of TNMM with the distribution affiliate as the tested party.

On July 21, 2011, Edison and Promgas signed a long-term agreement for natural gas, which was described by the Edison press release:

Today Edison and Promgas have successfully completed the renegotiations on the price review for the long-term contract for the supply of natural gas from Russia in order to comply with changed market conditions. The agreement has an overall impact on 2011 accounts of Edison, which is estimated to approximately 200 million euro. “Today’s agreement with Promgas – stated Bruno Lescoeur, CEO of Edison – confirms the commitments of our Russian partner to maintain a balanced long-term relationship and the competitivity of its product in the market. The availability of Promgas to review the contract price was due to the dramatic decrease of the Italian market gas prices over the last years”.

The press release noted that Edison began renegotiations with Promgas as early as late 2008 when natural gas prices reached $16 per million BTU. From that period to June 2011, this price declined from $16 to $10. While the press release did not specify the contractual long-term price, let’s assume that it was $10, which was equivalent to €8 given the fact that the Euro was worth approximately $1.25 at the time. If Edison purchased 25 trillion BTUs per year at this price, Promgas revenue would be $250 million (€200 million) per year.

During the first 3.5 years of this long-term agreement, natural gas prices rose slightly about $10. From January 2015 to May 2021, natural gas prices fell below $10. Since June 2021, natural gas prices in Europe have been very high. The long-term nature of the July 2011 agreement insulated Promgas and Gazprom from this volatility as long as this contract remained in force. As such, who gets the economic rents was not a material question during the period when this long-term contract was in force.

Recent events, however, suggest that natural gas multinationals would be earning significant economic rents if they are receiving revenues under spot contracts or if their long-term contracts were renegotiated. If the transfer pricing approach were to grant the distribution affiliate with an arm’s length gross margin, its profitability would not include the substantial economic rents from the commodity boom.

References

Harold McClure, “Misapplication of the Resale Price Method: LATAM Petroleum Distributor”, Edgarstat Blog, February 23, 2021.

“Evaluating Whether a Distribution Affiliate Pays Arm's-Length Prices for Mining Products”, Journal of International Taxation, July 2014.