Understanding and responding to corporate inversions

[This post has been corrected since it was initially posted.]

The last few days have seen a flurry of discussion about corporate inversions.  Examples include this Washington Post op-ed by Treasury Secretary Jacob Lew, and today’s NYT op-ed by Paul Krugman.

Corporate inversions are transactions in which a U.S. multinational transmutes itself into a foreign multinational by arranging to convert the U.S. entity into a mere subsidiary, rather than the company at the top of the chain.  Once this has been done, there is still a U.S. company conducting U.S. operations.  But there are two big changes on the ground.

First, unrepatriated foreign earnings of the U.S. company's remaining foreign subsidiaries are easier than previously to redeploy without a taxable repatriation.  Once there is a parent that isn't under the U.S. affiliate, they can simply loan their trapped cash to that company, or use it to buy the company's stock.  Even if this generates interest or dividend income of the foreign sub that is taxable under subpart F (and low-dividend stock can be attractive in minimizing this problem), the inversion has usually yielded plenty of interest deductions from loan financing that was used in making the deal.  Apparently the term of art for this, as Edward Kleinbard notes in a Wall Street Journal op-ed, is "hopscotch transactions," since one skips over the U.S. parent in directing one's funds where one likes.

By the way, the importance that companies attach to keeping deferral going shows how unrealistic it is to analyze the value of deferral on the assumption that repatriation is inevitable at some point at the currently applicable repatriation tax rate.  It's well-known in the literature that deferral does nothing to reduce the present value of the expected repatriation tax if it is certain to be incurred eventually and if the U.S. repatriation tax rate will always be the same.  Are companies too stupid to know this?  I don't think so.  I will say more below on why they evidently value continuing deferral despite its not reducing the present value of the deferred liability under the above assumptions.  

Second, inversion makes it vastly easier to reduce taxes on one’s U.S operations.  For example, the use of both intra-company and third-party debt to strip earnings out of the U.S. and make sure they will show up elsewhere instead can be a highly effective tax planning strategy post-inversion.  The key is simply that they use affiliates that are not subsidiaries of the U.S. company.  That prevents both the creation of subpart F income when the U.S. company pays interest to foreign affiliates (since only its subsidiaries can give rise to subpart F income) and the application of the interest allocation rules to treat interest deductions as reducing foreign source income (which would potentially lead to foreign tax credit disallowance).  For fuller accounts, see Kleinbard's op-ed and Stephen Shay’s article in Monday’s Tax Notes.

An earlier flurry of inversions was shut down by the 2004 enactment of an anti-inversion statute in 2004 that basically prevents pure paper-shuffling transactions from working, as in the case where a U.S. company creates a Caymans affiliate that emerges at the end of the day as the parent on top even though nothing substantive has actually happened.  But the rules are now being beaten by transactions that have just enough economic substance to work - e.g., where there is an actual acquisition or merger of some kind, with an actual company located in a country where the U.S. group had some prior activity, but the whole thing is still in fact highly tax-influenced.  So the Obama Administration has been calling for legislation to address these transactions and prevent them from being tax-effective.

On Sunday I was on Arise TV, an Africa-based international news and entertainment station, to discuss inversions for a few minutes on a news show.  The host wanted to address whether inversions are immoral.  My response was, that’s not a very useful way to think about the problem.  Companies are going to do what they do (although it’s true that I myself would feel bad about spending my time doing this sort of deal), and the important thing is that we have rules in place that lead to the results we want.  More on what I mean by that in a moment.

But inevitably, moralistic vocabulary is going to be a part of these discussions.  In 2004 there was talk of “corporate Benedict Arnolds,” and President Obama has been speaking of “corporate deserters.”  This is pretty much how discussion needs to proceed in general public debate if one accepts – as I do – the bottom line conclusion that allowing the transactions to be tax-effective is undesirable.

As summarized by Treasury Secretary Lew, the proposed legislation would cause inversions to be ineffective for tax purposes if the multinational company “is still managed and controlled in the United States, does a significant amount of its business here and does not do a significant amount of its business in the country it claims as its new home ….  [In addition, t]o make sure the merged company is not merely masquerading as a non-U.S. company, shareholders of the foreign company [that ostensibly acquired the U.S. company] would have to own at least 50 percent of the newly merged company – the current legal standard requires only 20 percent.”

I support this legislation because I see no reason why we should let it so be easy for U.S. companies to avoid the repatriation tax that ostensibly has merely been deferred, and because I also see no reason to make it easier for them to strip away U.S. earnings.  But the fact that these are the two problems has four broader implications worth noting here.

First, as Kleinbard has noted, anti-hopscotch rules, making the loan to the new foreign parent subpart F income, could be deployed to address the sidestepping of repatriation taxes that the transactions faciliate.

Second, as a tougher alternative to merely addressing "hopscotch," Congress could consider enacting an “exit tax” when U.S. companies with unrepatriated foreign earnings cease to be U.S. companies – even in deals that would pass muster under the new proposed legislation.  The exit tax could be based on the amount of U.S. tax that the company would have paid had it repatriated all of its earnings just before the change in legal status occurred.  Under the simplest version, the tax due would be the amount of unrepatriated foreign earnings times the U.S. corporate tax, minus the amount of tax reduction that would have resulted from claiming foreign tax credits.  But one could also consider other approaches, such as providing “rough justice” in the form of a lower tax rate in lieu of counting up all the credits.  Obviously, if this were done, subsequent repatriations to the U.S. company by foreign affiliates that had kept their status as such would not be duplicatively taxable all over again.

Third, so far as stripping income out of the U.S. tax base is concerned, it would be desirable to move towards having “residence-neutral” rules that applied similarly to U.S. and foreign multinationals.  In particular, there ought to be ongoing multilateral discussion (such as through the OECD) of causing information from the entire global corporate group to be available to countries that merely host subsidiaries, rather than the parents.  This, for example, would permit the U.S. to apply interest allocation rules neutrally as between U.S. and foreign multinationals, for purposes of measuring U.S. source income.

Fourth, in the tax policy literature, we ought to be operating from realistic, rather than unrealistic, assumptions.  I noted above that deferral does nothing to reduce the present value of the expected repatriation tax if it is certain to be incurred eventually and if the U.S. repatriation tax rate will always be the same.  The reason is that, with further standard assumptions, the amount of tax ultimately due grows at the discount rate, so its present value stays the same.

It’s important for people who are working in the field to understand this point, which often is called the “new view” of dividend repatriations but is in fact a provable theorem (aka a tautology) under the requisite assumptions.  But what they should NOT do is assume that the assumptions are true.  Inversions raise the issue of the deferred tax simply disappearing.  But consider as well the fact that the U.S. could change the repatriation tax rate – for example, by lowering it to zero through the adoption of a territorial system without enactment of a transition tax, or through the enactment of further tax holidays.  (I say “further” because a dividend tax holiday was enacted in 2004, and there has been lots of discussion of doing it again.)

A better way of framing one’s understanding of deferral is to say that it has option value.  By deferring, one gets the option to benefit from future disappearance of the repatriation tax and/or from a reduction in the repatriation tax rate.  This would be economically valuable even if that rate were as likely to go up as down – which almost surely is not the case, especially given that the company will generally control the timing of repatriations.

A recent article by Al Warren did a nice job of explaining the logic behind the new view if one assumes (as he expressly does) not only that repatriation at some point is inevitable, but that the current tax rate on it is fixed.  But from a broader perspective, if one wants to inform current policy debate, what Warren did was assume away the most interesting and important aspects of the real world problem that he addresses (i.e., how to think about deferral when different possible models of international tax reform are on the table).