IRS preemptive strike?

This coming Monday, Tax Notes will be publishing my short article on the EU state aid case, which is entitled "Friends Without Benefits?: The Treasury and EU State Aid."  But here is a kind of teaser as it's on the same subject, addressing a development that came out too recently (i.e., today) for me to discuss it in the article.

The IRS has just issued Notice 2016-52, entitled “Foreign Tax Credit Guidance Under Section 909 Related to Foreign-Initiated Adjustments.”

While I don’t have time right now to parse through it technically (I’m teaching a class in about an hour), it’s pretty clear what they are doing and why.  I’m personally sympathetic to their aim here, although I’ll leave it to others to assess whether, how, and how well the approach that they announce here works technically.

Section 909 addresses foreign tax credit "splitter" transactions like that in the Guardian Industries case, in which taxpayers created what were called super-charged or turbo-charged foreign tax credits, by finding workarounds to avoid the basic requirement or assumption of US law that you can only claim credits by repatriating the associated income.  The basic trick, using transparent entities and the like, was to place the profits in Entity A and the tax liability in related Entity B, so one could bring home just Entity B’s foreign tax credits, without Entity A’s earnings and profits also being included in the repatriation, and use the credits to offset the U.S. tax on other foreign source income.

It’s easy to see why Apple, in the EU state aid setting, would no doubt love to do a version of that trick this time around, if it can.  Say it owes $14.5B of Irish taxes under the state aid adjustment that the European Commission has proposed.  Since Ireland has a 12.5% tax rate, that would imply that the EC required a $116B increase in Irish taxable income.  Suppose Apple responded by repatriating $116B from Ireland (i.e., via a $101.5B dividend, grossed up for the tax liability) in order to be able to claim the credits.  That wouldn’t be so great for Apple, since the US pre-credit tax liability on this amount, at 35% (and assuming no other foreign taxes, for simplicity) would be $40.6B, reduced by the credits to $26.1B.

Now suppose instead that Apple could gin up a way to claim the $14.5B in FTCs without actually repatriating any Irish profits.  Then it would have built-in shelter for other repatriations in the amount of $41.4B that had no associated foreign taxes (and thus that would yield $14.5B in pre-credit US tax liability).  Voila, although Apple is still unhappy about the EU state ruling, at least it got to bring home lots of money for US tax purposes without paying any US repatriation tax.

Would this actually cost the Treasury $14.5B?  In a way no, since Apple wouldn’t have done the repatriation but for having the credits hypothetically made fully available.  But it might cost Treasury money in the future – e.g., if we went to a territorial system but had a deemed repatriation on pre-enactment foreign profits. Apple would have gotten to reduce by $40B the foreign earnings to which such a hypothetical tax on deemed repatriations would have applied.

Plus, as a general matter, the Treasury has good reason to dislike “splitter” transactions, which not only encourage repatriations that might not otherwise have occurred, but also can shelter repatriations that would have occurred anyway (and hence been taxable).  The EU state aid cases present a new wrinkle, by reason of their involving the sudden emergence of lots of taxes from past years that weren’t previously being accrued by the taxpayer, a scenario which certainly would motivate tax planning activity to figure out how to get around pre-existing section 909 and regulations.