...I did quite a bit of reading before selecting the Pluris DLOM Database.

Keith Sellers
University of North Alabama

New 2704 (b) Regulations - Return of Family Attribution?


By: Espen Robak | September 03, 2015

New 2704(b) Regulations – Return of Family Attribution?

In an otherwise slow news season this summer, there’s been one theme that has caused excitement among estate planners: the prospect of severe limitations on, or the outright elimination of, valuation discounts.

The Proposal

As reported in the New York Times recently,[i]  a representative of the Office of Tax Policy at the US Treasury Department, told conference attendees recently that Treasury was aiming to release new regulations by mid-September. Others have reported that these regulations are supposed to become effective immediately upon issuance.

The new regulations, if ever actually introduced, will be issued under the specific authority available under IRC Section 2704 (b),[ii] specifically Sec. 2704 (b) (4), which states:

The Secretary may by regulations provide that other restrictions shall be disregarded in determining the value of the transfer of any interest in a corporation or partnership to a member of the transferor’s family if such restriction has the effect of reducing the value of the transferred interest for purposes of this subtitle but does not ultimately reduce the value of such interest to the transferee.”[iii]

These changes – if they work as advertised – will raise an additional $18 billion over the standard budgeting window, or $1.8 billion per year, which works out to about 10 percent of the government’s total take in any given year.[iv]

The $18 billion appears based on an analysis from the government’s “Greenbook” which, for years, promised similar amounts of increased revenues from eliminating valuation discounts. However, that analysis was based on a legislative fix.

One of the first to bring the possible new regulations to the attention of tax planners was Richard Dees, who notes “although it would be highly unusual, rumors say the regulations might be immediately effective” and suggests that moving ahead with such far-reaching changes without new legislation is potentially invalid.[v] As to exactly how far-reaching the regulations could be: 

“In other words, they may revive the concept of family attribution under the pretext of limiting liquidation and other restrictions.”

Which raises an interesting question:

If new regulations are introduced, with perhaps insufficient backing in legislation, and the IRS adopts the position that they do indeed produce a result equivalent to family attribution, where does that end? With a reversal in court? Or with a long string of IRS losses until they finally abandon the cause – as they did with Revenue Ruling 93-12? That would take a long time, if history is a guide.

Effective Date

The question of whether the regulations are given immediate effect (i.e., without a notice and comment period)[vi] may also influence their validity – or at least whether courts afford them “deference” or merely “respect” – and ultimately whether the regulations can be effectively challenged.[vii]

Ron Aucutt, in a recent Capital Letter, notes that there is also a crucial difference between new regulations being applicable immediately to all new transfers versus being applicable to all new restrictions created.[viii] He adds, with a touch of understatement:

“Somehow effectiveness of the new rules immediately upon their publication in proposed form, especially if they are applicable to all transfers even of interests in entities already formed subject to restrictions already imposed, seems ambitious for regulations under a 25-year-old statute and a 12-year-old Priority Guidance Plan project.”

Reach – How Far Is Too Far?

Mr. Aucutt also has the following predictions for the content and effect of the regulations:

  1. Look for features of entities that do not look natural and compelling, and valuation discounts that do not look justified, to predict which restrictions will be disregarded.
  2. A revenue gain of $1.8 billion implies disallowing discounts on approximately $4.5 billion of assets per year. (Presumably, this is less than the total amount of assets subject to discounting on Forms 706 or 709 per year).
  3. In particular, discounts justified solely “by features of the entity holding the assets, not features of the assets themselves” may be on the chopping block before other discounts.
  4. This implies a focus on marketable securities partnerships, rather than entities that have complex operations – perhaps even including real estate entities.
  5. The regulations may provide some kind of explicit standard, rather than state law, against which restrictions are “greater than” and will be disregarded.
  6. The regulations may also distinguish between restrictions, based on whether or not the transferor created them and based on whether or not the transferee may lift them.
  7. It’s also possible that the regulations will be dialed down a bit – resulting in less revenue than estimated in the Greenbook – due to the lack of legislative cover.

Data to verify this is hard to get by, but $4.5 billion is presumably less than the total amount of assets subject to discounting on Forms 706 or 709 per year. In addition, since the revenue gain is only supposed to result in savings around 10 percent of the total estate and gift tax take, the new regulations may not be quite applicable to all discount situations. That would also eliminate the possibility that the regulations will bring family attribution back in through the back door.

Given that, Mr. Aucutt’s way of prioritizing between entities based on their asset holdings, rather than distinguishing between transferred interests based on their features may make sense. That may in fact be how the government gets to its revenue estimates.

But there’s another possibility: that the government is simply wrong about how much the proposed regulations will actually bring in.

One thing is clear about the new proposed Treasury regulations: thus far, we know nothing about them. It’s been reported that they will be issued specifically under the authority granted by Sec 2704(b)(4), and that they will yield additional revenues of $1.8 billion per year, but we really don’t even know that.

In fact, the two things we do, supposedly, know about the new regulations may be mutually exclusive.

As I will argue here, eliminating discounts altogether – and bringing back a 21st century version of family attribution – would require much more than simply disregarding even a very expansive list of restrictions: it would require establishing a new legal fiction only applicable for transfer tax purposes imputing economic value to interests in privately-held entities which they do not possess; or a brand-new definition of value; perhaps both.[ix]

Also, let’s consider the ability and inclination of high-net-worth families to change strategies when the rules change. Assuming advisors modify their plans, for the Government to reap a meaningful increase to its tax take would require much more than simply disregarding restrictions.

Imagine… No Restrictions?

Here, I will for the sake of space simply ignore the possibility that Treasury will establish a new definition of value. Fair market value – of the property transferred – has been the standard so long, and imbedded so far in legislation, regulations, and the legal, accounting, and valuation professions that it would seem outlandish for the Government to change it without legislation, without even hearings, and at the drop of a hat.

Let’s instead focus on the text and intent of Section 2704: the treatment of restrictions. Sec. 2704(b) specifically pertains to restrictions on liquidation, but let’s take an expansive view and assume that the term “other restrictions” in Sec. 2704(b)(4) really means all other restrictions, meaning: restrictions on transfer, assignability, liquidation, dissolution, distributions, voting rights, and all others.

Let’s do away with them all.

It’s Easy If You Try…

As an example, let’s consider a family-owned corporation with 1,000 shares outstanding. All shares have the same voting rights, and a majority elects the board. There is no limitation on transfers, or right of first refusal, or any other kind of transfer restriction. The corporation may sell its assets at any time, distribute all proceeds to shareholders, and dissolve itself, with no restrictions whatsoever. The by-laws and shareholder agreements merely formalize the rights and responsibilities of management and the board and provide for meetings, notices, etc. The company holds marketable securities.

When valuing 100 transferred shares, wouldn’t it be logical to discount those shares more or less fully? The interest is clearly non-controlling, since it represents a minority. And it is non-marketable, in the sense that it lacks access to the public market (more on that below). Further, disregarding restrictions cannot possibly change the value, since the interest isn’t subject to any.

Analogies and Evidence

Fundamentally, valuation discounts aren’t a function of features or restrictions. They are a function of the economics of the entities and interests valued, and the operations and assets owned by the entities. Restrictions may sometimes influence the magnitude of valuation discounts, but the exact weight of restrictions in the determination of discounts is often exaggerated. The fundamental questions to answer when determining discounts are:

  1. Does the holder of the interest have any control, or constructive control, over the entity?
  2. Can the interest be traded in a public, liquid, marketplace?
  3. What are the economic characteristics of the operations and assets of the entity, the entity itself, and the interest, in terms of risk, cash flow yield, and other characteristics?

For illustration, let’s consider two analogies, unregistered corporate shares and interests in closed-end funds.

Unregistered corporate shares are often referred to as “restricted stock” or “Rule 144 restricted shares” but those names are misleading at best in this context. Unregistered shares are not restricted in the sense that there is a prohibition on their transfer, as commonly seen in limited partnerships with transfer restrictions, rights of first refusal, or similar restrictions. And Rule 144 itself is not a restriction; rather it’s a safe harbor exemption, itself providing a path to liquidity for unregistered shares.

In fact, my firm, Pluris Valuation Advisors, has for years compiled a database on sales of unregistered shares.[x] The average discount in the LiquiStat database is close to 30 percent. And, as noted, these shares are subject to no transfer restrictions whatsoever, other than a lack of access to the public markets – in fact, the discounts are themselves measured based on prices paid for these shares in arm’s-length transfers.

With a lack of access to public markets being their sole detriment to value – and the sole factor giving rise to the discount – the shares in the LiquiStat database are close analogies to our example of 100 shares in a family-owned corporation.

Interests in closed-end mutual funds are publicly traded and fully liquid. Their holders, however, lack control over the portfolios of the funds. This lack of control is by virtue of the interests being minority interests. These discounts typically range from low single-digits to high teens. Again, the analogy with our 100 shares in a family-owned corporation above is very close. And again, explicit restrictions (which might be disregarded) are not the cause of the discount.

Exceptions – And More Planning Opportunities

It’s obvious from our example above that this is a simplified example. Family entities – with some exceptions – that hold marketable securities are usually not corporations. A more common entity form would be a family limited partnership with marketable securities.

And what about transfers of limited partner interests in such entities? Isn’t it at least arguable that such an entity – with all restrictions disregarded – is the rough equivalent of our family-held corporation in the example above? If so, that would impose a very heavy additional tax burden on a transfer of a 90 percent interest – all the discounts would be eliminated, and rightly so. But would the same be the case for a set of three transfers, of 30 percent each, to three separate trusts?

This is why I believe the government will ultimately find its revenue gain from the new regulations – assuming they are approximately as outlined herein – disappointing. Many transfers would not be affected by the new rules. As for the transfers that are affected, careful additional planning, and making somewhat different transfers, might mitigate the effect of the new rules altogether.

[ii] 26 U.S. Code Sec 2704 – Treatment of certain lapsing rights and restrictions. https://www.law.cornell.edu/uscode/text/26/2704

[iii] Emphasis added. The emphasized phrase is obviously nebulous in the extreme. What does “ultimately” mean? It could only truly be discovered in hindsight, and is clearly unknowable as of any relevant valuation dates, and probably not even by the lapse of statutes of limitation.

[iv] The Estate Tax Provides Less than One Percent of Federal Revenue; Tax Foundation, Tax Policy Blog; April 7, 2015. http://taxfoundation.org/blog/estate-tax-provides-less-one-percent-federal-revenue

[v] New 2704 Regs May Increase Estate & Gift Taxes for Family Businesses; LinkedIn Pulse, May 24, 2015. https://www.linkedin.com/pulse/new-2704-regs-may-increase-estate-gift-taxes-family-businesses-dees?trk=prof-post

[vi] Supreme Court’s Mayo Foundation Opinion Grants Chevron Deference to Treasury Regulations, Tax Executives Institute, June 7, 2011. https://www.tei.org/news/articles/Pages/TTE_SPRING11_supreme_courts_mayo_foundation_opinion.aspx

[vii] The ‘Death Knell’ for Family Discounts?; April 24, 2013. http://articles.jmbm.com/2013/04/24/the-death-knell-for-family-discounts/

[viii] Capital Letter No. 38, Anticipated Valuation Discount Regulations, July 20, 2015. http://www.actec.org/public/CapitalLetter38.asp

[ix] One caveat: applicable restrictions to be disregarded include restrictions which can be removed by the transferor and/or his family. That could include almost anything in the family-owned entity context. However, it should still only apply to "restrictions" - i.e., specific contractual provisions in entity documents. A reading of Sec. 2704(b)(2)(ii) to the effect that it gives the government the ability to impute control and marketability to interests who fundamentally - even absent contractual restrictions - lack them (essentially treating minority interests as majority interests), seems too far removed from reality.