By Dr. J. Harold McClure, New York City
The UN Committee of Experts on International Cooperation in Tax Matters will consider updates to the United Nations Practical Manual on Transfer Pricing for Developing Countries at its 21st session, which commences October 20.
UN documents released in advance of the meetings show that China has updated its country practice chapter. This revised China chapter, written by Chinese government officials, provides an interesting analysis of the appropriate remuneration for contract manufacturing affiliates and distribution affiliates.
China distribution affiliates
Paragraphs 4.3.8 and 4.3.9 of the revised China country practice chapter states:
Sales, marketing and distribution are another set of functions where it has been seen that MNEs often underestimate the contribution of developing countries. Chinese experience shows that many MNEs treat its Chinese distribution entities as a limited risk distributor, and use a set of simple distributors performing limited functions in a mature market such as Japan as the comparables … there often is a mismatch in terms of functional profile, as the Chinese entity may perform significantly more functions than these so-called comparables, which is evident as it incurs significantly more operating expenses relative to sales … the Chinese tax administration has attempted to correct such deficiencies by using a more appropriate transfer pricing method, such as profit split in the cases where significant local marketing intangibles or LSAs is identified, or performing comparability adjustments when TNMM is used.
China’s view of transfer pricing for distribution affiliates can be understood by reference to information reported by an Australian government-funded entity called Wine Australia.
Wine Australia reported that total revenues for Australian wine producers reached almost 6.4 billion Australian dollars (A$) with exports = A$2.9 billion during the 2018–2019 season. China is Australia’s largest market for wine exports, followed by the UK and the US.
The companies with the largest market share include Treasury Wine Estates Limited, Pernod Ricard Pacific Holding Pty Ltd, Casella Wines Pty Limited, and Accolade Wines Holdings Australia Pty Limited.
Consider an Australian wine multinational with a Chinese distribution affiliate that sells 5 million bottles of wine to Chinese customers for A$20 per bottle with the following cost information:
- The parent incurs production costs = A$7 per bottle and shipping costs = A$1 per bottle;
- The Chinese affiliate incurs selling costs = A$6 per bottle; and
- Advertising expenses paid to a third party marketing entity = A$2 per bottle.
Table 1: Wine multinational income statement (Australian dollars)
|Cost of goods||$40||$0||$40||$0||$40|
Consolidated profits = A$4 per bottle or A$20 million, which would be allocated between the Australian parent and the Chinese distribution affiliate according to the transfer pricing policy.
Table 1 presents two such policies under the assumption that the Chinese affiliate pays the advertising expenses.
The taxpayer asserted that all valuable intangible assets are owned by the Australian parent. Its transfer pricing policy afforded the Chinese distribution affiliate with a 45 percent gross margin, which allowed it to retain only A$5 million in profits.
The usual justification for limiting the operating margin for the distribution affiliate is an analysis based on the transactional net margin method (TNMM).
The China State Taxation Administration would like to argue for a 50 percent gross margin on the grounds that the Chinese distribution affiliate owns the marketing intangibles.
This alternative transfer pricing policy would raise Chinese taxable income to A$10 million, lowering Australian tax income from A$15 million to A$10 million.
If the Australian Tax Office agreed with the taxpayer’s transfer pricing policy, the divergence of views regarding the appropriate gross margin would lead to a double tax issue.
Contract and toll manufacturing affiliates for smartphones
Your new deluxe smartphone likely cost you more than USD 1000, but the Chinese contract manufacturer was paid something closer to USD 260, of which USD 225 represented the cost of components with only USD 25 in Chinese assembly wages and USD 10 in profits for the Chinese assembly company.
When Apple pays Foxconn for the contract assembly of its iPhones, it pays the market price. But let’s imagine that Apple or Samsung has established a Chinese affiliate to perform the assembly function.
Let’s further imagine that after the Chinese affiliate was established, it was converted to a toll manufacturer with a Hong Kong affiliate responsible for sourcing the components and ultimately selling the finished product to the parent corporation.
The key passages from paragraphs 4.3.4 to 4.3.7 of the updated China country practice chapter state:
In evaluating a contract manufacturer’s return, the TNMM is often used as the transfer pricing method with the FCMU being the most commonly used profit level indicator. The arm’s length principle involves testing controlled transactions with uncontrolled transactions to determine how independent parties would have acted in broadly comparable situations. This principle becomes challenging to apply where a company relies on its related parties for both input purchase and output sales. If such a company is to be evaluated on a cost-plus basis, a low intercompany purchase price results in an undervalued cost base that will ultimately undercompensate the contract manufacturer…. The Chinese approach to evaluating such companies is to start with the general presumption that the related party purchase price of materials is at arm’s length, and evaluate the reasonableness of the mark-up earned by the contract manufacturer on its cost base … Toll manufacturing is a common form used by MNEs in developing countries, but its proper return is difficult to determine since there are only a few independent listed companies that perform such activities. Some taxpayers simply use the FCMU for contract manufacturers as the mark-up for toll manufacturers. This grossly underestimates the return to toll manufacturers. Others use return on assets as a profit level indicator based using contract manufacturers as comparables, and this may also underestimate the return, particularly for labour intensive toll manufacturers as often being the case in developing countries…the Chinese tax administration has sought to first estimate the total cost of the toll manufacturing operation as if it were a contract manufacturer, usually by adding back costs of raw materials which may be obtained from Customs. It then estimates the appropriate returns (say, FCMU) for contract manufacturing based on contract manufacturing comparables, and apply this to the estimated total cost to arrive at the total contract manufacturing profit, from which it then adjusts for factors such as inventory carrying costs, to arrive at the total profit for the toll manufacturer. This approach works well when reliable customs information on raw materials is available.
FCMU is short for full cost markup, that is, the ratio of operating profits to total costs, including both the cost of components and labor costs.
I noted that the Chinese approach to toll manufacturing in many ways captured the thinking of Ronald Simkover when he worked for the Canadian Revenue Agency (see “Made in China, Sold by Hong Kong: Processing Trade and Transfer Pricing,” Journal of International Taxation, October 2017).
Simkover’s approach would suggest that the concern over the price of components does not play a key role in these issues as the markup for the cost of components is very modest, whereas the markup for labor costs is high because the ratio of fixed assets to labor costs is high.
Simkover’s basic argument was that the markup over total costs (m) for a turnkey contract manufacturer should be seen as a weighted average of the return to value-added expenses and the return to pass-through costs:
m = x.v + (1 − x)z;
x = ratio of value-added expenses relative to total costs, which is 10 percent in our example;
v = ratio of operating profits attributable to employing extensive fixed assets relative to labor costs; and
z = ratio of operating profits attributable to modest working capital relative to pass through costs.
These markups can be seen as the product of asset intensities times the appropriate return to assets with:
v = Rf (fixed assets/value-added expenses)
z = Rw (working capital/pass through costs)
Rf = return to fixed assets
Rw= return to working capital.
Let’s assume that the value of fixed assets = $50 million and the value of working capital = $50 million. Total operating assets would therefore represent 40 percent of total costs. The ratio of fixed assets to labor costs, however, would be 200 percent, while the working capital to pass through cost ratio would be only 22.22 percent.
Table 2: Markups for a toll manufacturer under alternative assumptions
|Return to assets||2%||10%||14%||16%|
Table 2 evaluates the implications for the appropriate markup over labor costs for a toll manufacturer under alternative assumptions with respect to the return to working capital (Rw).
The first scenario is based on using the 4 percent markup before the contract manufacturer was converted to a toll manufacturer for the markup after the conversion to the toll manufacturing structure.
The UN transfer pricing manual’s China country practice chapter rightfully rejects this assumption. In our example, pass-through costs represent 90 percent of total costs, while working capital represents only 50 percent of total operating assets.
Conversion to a toll manufacturing structure removes the working capital from the balance sheet of the Chinese affiliate and eliminates pass-through costs from the income statement.
If the same 4 percent markup were used for the markup over labor costs for the toll manufacturer, its return to fixed assets is lowered to only 2 percent. Such a result is inconsistent with the arm’s length standard.
The second column of Table 2 presents a sensible result under the assumption that the return to working capital is 10 percent. Under this assumption, the conversion to a toll manufacturer structure removes half of the profits for the Chinese affiliate as it reduces the total cost base by 90 percent. In other words, the appropriate markup over labor costs would be 20 percent.
The China country practice chapter notes that if the original return to total costs for the contract manufacturer were arm’s length, then the analysis should simply deduct the return to working capital.
We should note two caveats. First, how would an analyst verify that this original markup was arm’s length. Our discussion notes that a proper return to operating assets analysis could prove useful.
The other caveat is that one does not necessarily assume Rw = Rf =10 percent as one could assume a lower return to working capital.
Table 2 considers scenarios where the return to working capital ranges from 4 percent to 6 percent. The implications of these alternative assumptions is that the markup over labor costs would range from 28 percent and 32 percent.
Our model justifies the China State Taxation Administration position that the conversion from a contract manufacturer structure to a toll manufacturer structure involves a reduction of the affiliate’s profits by simply the return to working capital.
This position is reasonable if the original markup over total costs was consistent with arm’s length pricing.
The most straightforward means for evaluating this assumption is a return on assets approach, which would ultimately lead to the sensible conclusion that the toll manufacturing affiliate is entitled to an appropriate return to its fixed assets.