Tax policy colloquium, week 13: "Helen of Troy" anti-inversion regulations

Yesterday at the colloquium, Deborah Paul presented "Has Helen's Ship Sailed? A Re-Examination of the 'Helen of Troy' Regulations." This paper, which is closer to the ground-level institutional details of federal income tax practice than most of our fare this semester, addresses a kind of coelacanth of the federal regulatory process, although the time frame for this "living fossil" is 25 years rather than 400 million.

The "Helen of Troy" regulations are so known because they were issued in response to an inversion transaction involving a company of that name. They came out in 1994, or a decade before an ensuing wave of inversion transactions gave rise to the enactment of IRC code section 7874, responding to the phenomenon both legislatively, and far more broadly and systematically.

A corporate inversion, as presumably is known to most readers who were interested enough to read this far, involves a U.S. multinational company with foreign subsidiaries seeking, through tax-free reorganization transactions, to substitute a foreign corporate parent (often located in a tax haven) on top of the prior U.S. parent, and also to change the corporate structure so that the foreign subsidiaries are under the new parent, rather than the U.S. company, which remains on hand just to engage in the broader group's U.S. operations.

Pre-2017, the main tax planning aims served by inversions were (1) to allow dividends to be paid up from the foreign subsidiaries to the company on top of the chain without triggering the U.S. repatriation tax, and (2) to facilitate earnings-stripping out of the U.S. tax base. If this is done by having a U.S. parent pay interest to foreign subsidiaries (which might have made the "loan" by simply round-tripping equity previously inserted by the parent), it ends up being foiled because the U.S. interest deduction is offset by subpart F income taxable to the U.S. parent by reason of the subs' interest income from the loan. But this doesn't happen if the interest is paid to foreign group members that have a sibling or parent, rather than subsidiary, relationship to the U.S. company in the corporate ownership chain.

The 2017 act eliminated the repatriation tax that used to motivate inversions, and created some additional barriers around interest-stripping. But its enactment of GILTI (a quasi-minimum tax on U.S. companies on their foreign subsidiaries' profits) it created a new reason for wanting to invert.

Anyway, back in the day (1994) the Treasury wanted to clamp down on inversions, but didn't have all the tools it has now. I don't know why there was no legislative push - this was before the November 1994 elections swept Gingrich et al into power - but conceivably the politics had something to do with it. What they decided to do was issue regulations under section 367(a).

Let's pull back the camera now for some broader background. In general under the U.S. federal income tax (and most others), gain from asset appreciation (or loss from its declining in market value) is not taken into account for tax purposes until there is a realization event, such as sale. This rule leads to numerous distortions and tax planning opportunities - the late William Andrews called it the "Achilles heel of the income tax" - but it has generally been though necessary in response to problems of asset value measurement and taxpayer liquidity. (There are now proposals around to apply mark-to-market taxation, or retrospective systems that aim for equivalence thereto, but that's another topic.)

 But once realization events were made taxable, it was thought desirable to create exceptions, by allowing nonrecognition for certain transactions, such as incorporating one's business, turning one corporation into two or two into one, etcetera. The rationale was that these transactions not only might be doing little to address measurement and liquidity issues, but also were not convenient occasions for levying the tax on appreciation - for example, because they were merely reshuffling how one's assets were held, and would tend not to happen (rather than yielding taxable gain) if they were taxed.

But then the next step was the tax authorities' learning the hard way that taxpayers could exploit nonrecognition transactions to achieve tax planning aims beyond business-motivated reshuffling. A classic example is the Gregory case from the late 1930s, which established modern economic substance & business purpose doctrine. A simplified version of that case might go as follows. My company has two types of assets: boring stuff and cash. I want to get the cash out into my own pocket, but dividends were subject to high tax rates at the time. So step 1, I do a tax-free spin-off so there are now 2 companies, one holding the boring stuff and the other holding the cash (both wholly owned by me). Step 2, I liquidate the company holding the cash. Now I'm taxed at the capital gains rate rather than the dividend rate (today they're the same, but at the time CG rates were much lower), plus I get some basis recovery with respect to the cash company's stock. If this had been allowed to work, there would never have been a taxable dividend transaction again - everyone would have done these two-steps instead. So tax-free reorg treatment was denied.

Section 367(a), the provision under which the Helen of Troy regs were issued, responded to another type of taxpayer planning trick. Say I own an appreciated asset of any kind - be it a painting, Facebook shares that I got back in the day, etc. - and want to move towards converting it to cash. As per the legislative history of the provision's 1932 enactment, I might contribute it to a new foreign corporation (FC) in exchange for all its stock, have the FC sell the asset outside of the U.S. (generating no U.S. tax), and I now have 100% control of an entity that's sitting on the cash (although it remains in corporate solution, and paying myself a dividend would be taxable. Congress viewed this as undue avoidance, so it passed a provision stating that otherwise tax-free reorganizations in which one ended up with foreign stock would be taxable, subject to the Treasury's creating exceptions.

The statutory language was quite broad. In current form, section 367(a) says that FC stock won't count as stock received for purposes of determining gain recognition, subject to the Treasury saying otherwise. So it went well beyond the specific situation that Congress had most directly in mind.

Section 367(a) imposes a shareholder-level sanction - gain recognition - and is widely thought of as responding to shareholder-level, not entity-level, tax planning fun and games. But in 1994, when the Treasury announced and then adopted the Helen of Troy regs, they aimed it at inversions, which are an instance of entity-level tax planning. This led some to argue that the regs were beyond the provision's statutory purpose (since Congress in 1932 presumably had no idea that inversions would become a problem 60+ years later), and also that it was in tension with principles of sound system design. E.g., it might be good drafting to have the provisions aimed at entity-level planning issues over here, and those aimed at the shareholder level over there. As an example of the mismatch, the Helen of Troy regs leave inversion transactions unscathed if the shareholders are tax-exempt, because in that case they aren't going to face taxable gain recognition anyway.

The regs apparently are a bit of a mess - reflecting, for example, that the state of the art so far as drafting provisions applying to the issues presented has improved since then - as is reflected in section 7874 and its regs. So the Deborah Paul paper that we discussed yesterday goes through a lot of the problems, and urges that the Helen of Troy regs be addressed. For example, they might be eliminated, or alternatively they might be updated, improved, conformed more to section 7874.

I don't know enough about conditions on the ground to evaluate the cost-benefit analysis that would be involved in deciding whether this distinctly tertiary means of discouraging inversions should be streamlined or eliminated. The first tool at hand is section 7874, while the second, which I gather has been quite effective, is the 2016 regulations, issued during the Obama Administration, under the guise of section 385 (addressing debt vs. equity). The fate of the latter remains uncertain, although so far the current administration has merely tinkered around the edges, rather than more substantially scaling them back. Perhaps they're worried about the headlines if they throw out the 2016 regs and more inversions ensue.

Another piece of this whole story is the increasing difficulty, given the current state of U.S. politics, of using legislation to respond to new developments in tax practice that seem to undermine the existing system (as a wave of inversions can do). Regulators increasingly will and (given the totality of circumstances) should address urgent problems that might better have been left to Congress, as a matter of design flexibility and also inter-branch comity, if things weren't the way they are. One wild card left behind by doing more through regulations, and less through legislation, is that there may be a rise of back-and-forth seesaws when the presidency changes hands. A second is that the courts may increasingly be following their own ideological (and even partisan) preferences in deciding when to rein in regulation, and when to approach it deferentially. These of course are bigger problems than just Helen of Troy, even if it was the transaction that launched a thousand regs.