Branching Out: The Subpart F Contract Manufacturing Regulations v2.0

By James J. Tobin, Esq. Ernst & Young LLP, New York, NY

I suppose I should be embarrassed to admit it, but almost every time over the years that I've had to deal with the Subpart F sales or manufacturing branch rules, I've had to open the regulations and read through the rules -- yet again -- to get myself back up to speed on the details. And in the case of these branch rules, as much as anything, the devil really is in the details. So when the IRS issued the proposed Subpart F contract manufacturing rules in February of 2008, I was less than thrilled to see yet another set of branch rules and yet another layer of dazzling complication.

Oh, and some of the proposed rules didn't exactly make sense, even after I carefully parsed my way through them.

I made my thoughts about the early-2008 proposed regulations pretty clear on these pages in the wake of the issuance of those rules.1 In short, while I gave the government a modest attaboyfor taking on such a daunting task and making a credible first effort, I concluded that the proposed regulations were just that -- a first effort -- indicating that a good deal more thoughtful consideration was needed. I also spent some time carping about the lack of guidance provided for purposes of calculating the effective tax rate of a branch around which you're made to “draw a box” under the branch rules, as well as for purposes of figuring out how much income is attributable to such a “branch.”

So, now that the IRS has had time to think about it -- and sift through what were probably a record number of comments -- what have they wrought with version two of the contract manufacturing regulations?

Overall, the Service has done a pretty good job of addressing many of the concerns raised by taxpayers and practitioners about the proposed regulations. So, for example, they've now refined the list of “substantial contribution” activities and clarified certain aspects of the list, as well as the manner in which it is to be applied in practice. They've also provided a definition for the term “employee,” which was a big sticking point in the proposed regulations.

All in all, a decent day's work. And all very constructive in the recognition of the economic reality of the world we live in today, where physical manufacturing is often outsourced but direction and control of the manufacturing process is often the true value-added economic activity that should not, in principle, be viewed as “foreign base company”-type activity.

As to the branch rules, though, in my view what we have remains a work very much still in progress. While the progress does seem to be going in the right direction in my view, the progress is only incremental. In fact, the changes to the proposed branch rules are so extensive that this portion of the rules, unlike most of the remainder, has been issued in proposed and temporary form. Which makes one pause and wonder whether the manufacturing branch rule really is worth all this blood, sweat, and tears.

The Branch Rules: Some Answers but Many More Questions

By way of quick background, recall that if a controlled foreign corporation (CFC) conducts its manufacturing or sales activities outside its country of incorporation through a branch, a special set of rules applies for purposes of determining whether the CFC has foreign base company sales income (FBCSI).

Also recall that the branch rule is specifically authorized by the statute (§954(d)(2)) as enacted in 1962. The statutory language is centered on a sales branch paradigm, rather than on a manufacturing branch paradigm (although the legislative history provides some support for the latter). I'll also point out that in 1962, which was a good bit before my time as a tax guy, it appears there was little thought of virtual outsourced manufacturing, few if any regional or global business models as the operating philosophy of multinationals, and no formal transfer pricing documentation or penalty rules to ensure that profit wasn't shifted inappropriately from one foreign branch to another. Oh yeah, and it was also pre-check-the-box by about 35 years and operating in true branch form was pretty rare.

So now, almost 50 years after the legislative authority was provided for the branch rules, at a time when they will actually be relevant to a majority of U.S. multinationals, we are finally given somewhat detailed and very complex guidance for the first time since 1962. What does the guidance cover and, more importantly, what does it not cover?

The primary driver of the regulations has always been the so-called rate disparity test (RDT), which determines whether you have a branch that you have to worry about. In the good news category, the regulations recognize that there can be more than one substantial contributor to the manufacturing process eligible for the manufacturing exception to FBCSI; in the bad news category, this means that multiple locations within a CFC could be considered to have substantially contributed, which could give rise to multiple manufacturing branches. Likewise, a CFC can have multiple sales branches because it sells different products from different offices, it purchases a particular component or raw material in one location and sells the finished product in another (note that purchasing activities can also create a “sales” branch), or perhaps even where multiple offices contribute to the marketing or sale of a single product. To their credit, having created greater potential for multiple manufacturing branches through the operation of the new substantial contribution concept, the drafters did include in these regulations rules for prioritizing among multiple manufacturing branches in order to focus the analysis on one such branch. The drafters also helpfully included rules for coordinating the manufacturing branch rule and the sales branch rule. These are both areas where guidance had been lacking for years.

The new regulations include a complicated scheme for determining which manufacturing branch (or aggregation of branches) must be examined under the RDT. Once you determine the manufacturing branch that will be the focus of the analysis, the RDT is then applied by comparing the actual effective tax rate incurred in a sales branch with the hypothetical tax rate that would have been incurred had that income been earned in the appropriate manufacturing branch; if the actual tax rate is too disparate (less than 90% of and more than 5% points lower than the hypothetical tax rate), then a box is to be drawn around the sales branch and Subpart F consequences are analyzed as if there were two separate CFCs. If there are multiple sales branches, this analysis is repeated for each sales branch.

In my view, the basic conceptual approach used in these new branch rules is fairly well thought-out and recognizes the realities of complex global or regional business models, as far as it goes. Of course, the rules will also be incredibly complex to apply in practice for all but the most simple fact patterns.

My biggest concerns, however, are still with respect to the issues the regulations do not address, as well as a potential trap for the unwary created by the regulations in cases where a virtual manufacturer is purchasing and selling from and to unrelated parties.

Areas that are not addressed in the regulations include the definition of what is a branch, details as to how to compute the effective tax rate of a sales branch or the hypothetical tax rate of the manufacturing branch, and details on how to compute the amount of Subpart F income if the RDT is failed and the deemed existence of a second CFC could result in the creation of Subpart F income. Not exactly minor details!

The Preamble to the regulations does indeed acknowledge the lack of guidance on these issues in stating, “The IRS and the Treasury Department concluded that other questions and requests in this area, including further clarification of the methodology for calculation of hypothetical tax rates, and for changes to the assumptions used in applying the tax rate disparity tests and determining the hypothetical effective tax rate, are beyond the scope of this regulatory project.” The Preamble similarly states that defining the term “branch” is outside the regulations' scope.

My reaction to all this was going to be to say that, in my view, this is unacceptable from a tax policy standpoint. I was then going to go on to note that if these detailed calculation issues are too tough for the IRS national office to figure out for inclusion in the regulations, how can it be acceptable to toss these unanswered questions over the fence for taxpayers and IRS agents to figure out? Admittedly, we have lived without any real guidance on these issues since 1962. However, under existing law, based on the Ashland and Vetco cases (and notwithstanding the questionable position espoused in Rev. Rul. 97-48, which revoked Rev. Rul. 75-7), a virtual manufacturer should not be considered to have a manufacturing branch, so the issue of the RDT has only infrequently had application. However, the way the new substantial contribution concept is structured in these regulations now will result in many companies needing to grapple with the manufacturing branch rule, only to find themselves on their own in attempting to interpret how to do the most fundamental of calculations.

Then, on March 20, the IRS issued “technical corrections” to the regulations, portions of which seem to go way beyond the usual meaning of that term. For example, the Preamble was revised to provide that uniformly available tax incentives should be considered in determining the hypothetical effective tax rate used in applying the RED; if a ruling is available and the manufacturing branch does not obtain one, then the hypothetical effective tax rate that would be paid by that branch (or remainder of the CFC), were it to derive the sales income, should be the jurisdiction's effective rate not taking into account the “forgone” relief. It is thus clear that the manufacturing branch (not the sales branch) must affirmatively obtain a ruling that includes sales income. In addition, the Preamble removed the portion of the regulations that said the principles of the §954 regulations addressing the high-tax exception were to be used in determining the effective rate of tax when applying the RDT. The technical corrections also revise the rules to be applied when no location independently satisfies the manufacturing test in such a way that one can now have multiple tested sales locations for the same item of property.

Perhaps my planned carping was not quite strong enough, because, among other things, this last change will likely result in a greater amount of foreign base company sales income and add even more draconian compliance burdens, and very little has been clarified (or “corrected”).

In any case, notwithstanding these so-called technical corrections -- or, to some degree, because of them -- companies will really have their hands full in attempting to comply with the new rules.

Consider a simple fact pattern. A CFC European holding company, CFC Holdco, owns a Swiss checked subsidiary that substantially contributes to the manufacturing of a product. Swiss “branch” sells to Belgian sales company, which is also a checked subsidiary of CFC Holdco. Belgian “branch” sells to non-Belgian customers. Assume Belgian “branch” buys for 100 and resells for 175 and deducts 20 for local marketing costs, 25 for interest to CFC Holdco, 10 for royalties to Swiss branch, and 15 in NOL carryovers, which results in net taxable income in Belgium of 5 and Belgian tax of 2. Questions that one might ask -- in determining how to apply the RDT -- include:

• Do I use local Belgian taxable income rules or U.S. rules to compute the effective tax rate?

• Do I regard disregarded payments in the calculation?

• Do I take into account NOL deductions?

- Does that answer depend on whether the NOL relates to the sales activity?

• Do I respect the terms of sale for the purchase from Swiss branch and thus the gross margin that the sales branch records?

• How should U.S. or Belgian or Swiss transfer pricing concepts or adjustments enter into the calculation?

• How do I deal with forex issues, in this case between the Euro and the Swiss Franc?

• If I do conclude that the RDT is flunked, how do I compute Subpart F income -- which raises all the same questions as above?

• After I have done all that, do I follow the same principles I use above for purposes of §987 in considering what constitutes a branch remittance?

Again, not exactly minor details.

My last whine relates to the trap for the unwary. Assume CFC Holdco in the above example is purchasing all raw materials from unrelated parties and selling all finished products to unrelated parties, albeit through branches. And assume it doesn't fail the sales branch rule RDT, so it only could have Subpart F income if the manufacturing branch rules apply. Under Ashland/Vetco and other existing authority, there would not seem to be a concern about Subpart F in such a case; after all, the CFC is not doing either related party sales or related party purchases. And indeed, if the Swiss “branch” is not a substantial contributor, the manufacturing branch rule would not apply and there would still be no Subpart F risk. However, if the Swiss branch were found to be substantially contributing, all of the calculation headaches above would apply and some Subpart F income could well exist. The irony, of course, is that now the IRS will have an incentive in some cases to assert a CFC is a manufacturer and the taxpayer will need to argue why its alleged contribution is not substantial.

Going back to one of my opening comments, the world in 1962 was very different from the world today. The substantial contribution rules provide appropriate guidance on when activities of a CFC are substantial enough to not be considered tainted base company activities. But we still can't figure out how to apply the branch rule income test. At this point, it is appropriate to remember that the authority for the manufacturing branch rule is vague and exists only in the legislative history and not in the statute. I suggest some bolder policy thinking is in order and ask why not just drop the manufacturing branch rule altogether? The policy underlying the branch rule is rooted in concern about the potential for inappropriate shifting of a multinational's foreign tax base from a high-tax location to a low-tax location. The vigilance of foreign tax authorities, the proliferation of transfer pricing documentation and penalty regimes, and the intense focus on the tax implications of “business restructurings” as evidenced by the recent OECD report on this subject should all provide the IRS with comfort that the complexity and burden required to police this concern are no longer needed. We all have too many other things to worry about these days!

This commentary also will appear in the May 2009 issue of the Tax Management International Journal. For more information, in the Tax Management Portfolios, see Yoder, 928 T.M., CFCs--Foreign Base Company Income (Other than FPHCI), and in Tax Practice Series, see ¶7130, U.S. Persons' Foreign Activities.

1 Tobin, “The Proposed Contract Manufacturing Regulations: A Journey of a Thousand Miles Begins With a Single Step,” 37 Tax Mgmt. Intl J. 407 (7/11/08).