When valuing a non-controlling, minority interest in an entity for tax purposes, both the government and taxpayers generally agree that significant discounts must apply. But, what if the non-controlling interest being valued isn’t a minority interest? In limited partnerships and limited liability companies, the majority interest (for example, 75% of the entity) can easily be divorced from control by making the minority the general partner or managing member.
But, what if the interest being valued is an extremely large majority, to the point where it constitutes almost the entirety of the economic interest in the entity? Is that a case for a radically reduced discount? This was the main question recently decided in Grieve v. Commissioner.1
Pierson Grieve made two transfers in October and November 2013: (1) a transfer of a 99.8% member interest in Rabbit 1, LLC, (Rabbit) to a grantor-retained annuity trust and, (2) a 99.8% member interest in Angus MacDonald, LLC, (Angus) to an irrevocable trust. He filed a 2013 gift tax return disclosing both transfers, and the Internal Revenue Service audited the return, issuing a notice of deficiency in 2018 and increasing the value of the two transfers by almost
$7 million.2
In the appraisals submitted with the original gift tax return, net asset values (NAVs) for Rabbit and Angus were $9.1 million and $32 million, respectively; the lack of control discounts for the two entities were determined to be 13.4% and 12.7%, respectively; and the same 25% lack of marketability discount was applied to both entities.

The Expert Reports
Each expert started with the NAV of each entity, and the IRS expert accepted as his starting point the NAV claimed by the taxpayer on the original appraisals submitted with the tax return. The parties stipulated on the NAV of Rabbit at the original appraisal value. The taxpayer’s expert, however, opined on a slightly lower NAV for Rabbit and argued for a $1 million lower NAV for Angus than the original appraised amount. He arrived at this “inside” discount for some of Angus’ assets by analyzing restrictions on certain private equity fund interests. The Tax Court found “no justification to lower the NAV of Angus to a value that is significantly lower than the value to which petitioner previously admitted.”3
Beyond the NAV, the court essentially rejected the work of both experts and found entirely in favor of the taxpayer’s original appraisal submitted seven years earlier. Overall, not a great day in court for the IRS, but it was particularly not so because the approach taken by its expert in this case appears to be one that the IRS has had high hopes for. It will be interesting to see—after Grieve—whether it still does.
The taxpayer’s expert, beyond the aforementioned “inside” discounts (often referred to as “second-level discounts”), determined the value of interests transferred using two methods. In the market approach, he applied data from closed-end funds and restricted stock studies to conclude a discount just slightly higher than those taken in the original taxpayer appraisals.4 In the income approach, the expert relied on estimates of incremental required rates of return for interests, such as the subject interests in Rabbit and Angus, and determined values very close to his overall conclusions from the other approach. The court, in its analysis, “conclude[d] that the market approach is a reasonable method” and, further, that the higher discounts and lower values in the taxpayer’s expert’s report shouldn’t be substituted for the values and discounts in the taxpayer’s original report for the gift tax return.5
The Perfected Method
Even for these very large non-controlling interests, most analysts—and, in this case, also the government—would still agree that discounts should apply. However, should they be lower than in more typical situations? Do the standard discounting methods not quite apply anymore in these circumstances, as the IRS argued here, and, if so, what’s a better method?
First, it should be clear that the interests transferred were truly non-controlling ones. In the court’s summary, the subject interests for both Rabbit and Angus were 9,980 class B units, while the sole controlling owner held 20 class A units. The class A units had all the voting power. In addition, the operating agreements gave the ultimate owner of the class A units a lifetime appointment as chief manager and even gave her the right to designate a successor chief manager.6 As manager, she was entitled to compensation but chose not to take any. The agreement also required consent from all class A members before any class B units were transferred, among other transfer
restrictions.
The IRS nevertheless presented an argument that many appraisers (including me) have faced from the IRS before and that its own valuation experts sometimes refer to as the “perfected method.”7 Even though this is by now a well-known analysis, Grieve appears to be the first time it’s been presented in court.8 Let’s apply this theory to the facts in Grieve:
• The controlling owner was entitled to no more than 0.2% of either entity.
• Regardless of the current facts, future developments or anything she did to the portfolio of Rabbit and Angus, she would never be entitled to more than 0.2% of the income of the entities, no more than 0.2% of distributions made and only 0.2% of any liquidation proceeds.9 It would appear, given her minuscule share of each company, that she would have little incentive to hold on to her interest. Doing so simply meant contributing her labor and skill to improve the portfolios for the benefit of someone else entirely.

Assuming nothing about the identity of the controlling owner other than that this individual was interested in maximizing her wealth, is there not some price at which she might be persuaded to sell?
• What if she was offered five times the NAV of her ownership stake? How about 10 times? Maybe
50 times her value? What would any reasonable investor do in such a situation?

This is the gist of the perfected method: In such a lopsided ownership structure, in which the controlling owner has almost no “skin in the game,” the controlling owner should have very little incentive to cling onto her interest, if given some incentive to sell. In essence, it views the holder of the non-controlling interest as holding an interest in its pro rata share of the NAV that’s only temporarily clouded by the control exercised by the other owner. The only logical thing to do, then, is to “perfect” this interest by buying out the other party.
Framed like this, the IRS’ argument here appears logical. The IRS’ appraiser also claimed that “there was no empirical data on the sale of a 99.8% non-controlling interest.”10 The court simply lets this statement stand without comment. But, it’s obviously true: There’s no discount data for such large non-controlling interests.11 What appraisers do instead is use data from sales of smaller interests and hope it’s accepted for larger blocks too.
So, what premium did the IRS appraiser apply? First, he had estimated reasonable discounts for a minority interest in each entity.12 Then, he determined that 5% of the theoretical dollar amount of each discount would be a reasonable premium to pay the controlling owner to give up her control and sell her interest.13 In the case of one entity, the 5% “purchase premium” applied was $130,000. This compares with a pro rata NAV of $18,134 for the controlling owner’s interest. For the other entity, the purchase premium was $450,000, which compared with a pro rata NAV of $63,941, In both cases, the total amount offered to the controlling owner to sell was approximately seven times her NAV.14
What the IRS’ approach in Grieve doesn’t have is supporting evidence. As the Tax Court points out in Grieve:
• The IRS’ appraiser submitted no “evidence to show support for his valuations.”15
• “His reports did not include empirical data which back up his calculation of the […] premium to purchase the […] units of either entity.16
• “He provided no evidence showing that his metodology was subject to peer review”; and
• “Respondent cited no case law in support of [its appraiser’s] methodology.”17

It didn’t help the IRS’ case that the sole owner of the controlling interests in this case testified in court and stated that she had no intention of selling. And, even if she did sell, she claimed that she would have demanded a higher premium than the IRS’ appraiser determined. But, does that not implicitly admit that there’s some premium at which she would sell? According to the controlling interest owner on the stand, the seven times repurchase premium argued here would have been insufficient. The court believed her. Would another court, in another set of circumstances? We don’t know—but we also don’t know if there’s any analysis that would win this argument for the IRS.
Analysis
As noted, the Tax Court rejected the opinions of both appraisers, found no reason to object to the discounts in the taxpayer’s original appraisals and thus adopted them. The rejection of the IRS’ argument here is the interesting part of the case. And, it raises the questions: Is this it for the IRS on this theory?
The court notes: “[e]lements affecting the value that depend upon events within the realm of possibility should not be considered if the events are not shown to be reasonably probable.”18 Further, the “facts do not show that it is reasonably probable that a willing seller or a willing buyer of the class B units would also buy the class A units and that the class A units would be available to purchase.”19
That is, even if the purchase premium determined by the government’s appraiser here had been better supported, it sounds like at least in this case, the court would have been inclined to reject the analysis anyway. It remains to be seen if Grieve is the final nail in the coffin of the IRS method here. But, for planners and appraisers who are challenged by this methodology in audits going forward, Grieve provides strong encouragement.

Endnotes
1. Grieve v. Commissioner, T.C. Memo. 2020-28, (March 2, 2020).
2. Some concessions and stipulations were made before trial, most notably, the Internal Revenue Service stipulated that the petitioner would owe no gift tax for the transfer to the grantor-retained annuity trust (GRAT) if, assuming the government prevailed on the valuation, the GRAT would “true up” its annuity payments due back to petitioner within a reasonable time.
3. Citing Estate of Hall, 92 T.C. 312.
4. The assets of both entities were predominantly marketable securities and cash, plus some fund interests and notes. The approach of discounting the net asset value (NAV) of such interests (here referred to as the “market approach,” often also considered a cost approach to the valuation) is the standard valuation method.
5. It’s not clear to what degree the court’s analysis implies a rejection of the income approach used. It’s certainly possible that the entire analysis of the taxpayer’s expert here could have found favor, had it not also arrived at values lower than those “admitted” by the taxpayer on Form 709. Reading entirely between the lines: The taxpayer’s expert’s chief value-add in Grieve may have been shoring up the original valuations and, maybe even more so, rebutting the IRS’ expert’s novel analysis.
6. She could be removed as manager, but only for cause.
7. The taxpayer also argued that, while the IRS’ appraiser was personally qualified as an expert, his report and testimony shouldn’t have been admitted into evidence (forfeiting the government’s case altogether) as it was “irrelevant, speculative, and unreliable.” The court rejected this.
8. If so, it would appear to be a well-chosen case for the government. Rarely do we see ownership structures as lopsided as we see here.
9. There’s no evidence that the controlling owner was receiving any non-pro rata benefits from either company. In fact, she wasn’t even taking compensation, even though she might be spending considerable time managing the portfolios.
10. Supra note 1, at p. x.
11. Furthermore, the obvious reason why there’s no discount data on 99.8% blocks selling at more than 30% discounts in the literature or in the many collections of studies or databases that appraisers rely on is that such interests simply don’t sell at significant discounts (or at all).
12. The discounts were 10% for lack of control and 20% for lack of marketability, which he viewed as “not inconsistent with relevant market data and quantitative methodologies.” The court notes that not enough evidence was provided to support these estimates. They were, however, only the first step in the government’s appraiser’s method.
13. This seems like an obvious weak point: Why determine the purchase premium with reference to the lack of control suffered by the non-controlling owner? Rather, the focus should be on the selling owner in such an analysis and what, in this hypothetical scenario, might induce her to sell. This analysis seems to get the motivations of the principals backwards.
14. It’s unclear from the court’s summary of the case whether the analysis was presented exactly in this way or not, but the actual multiple assumed “offered” to the controlling owner as purchase premium is never specified.
15. Supra note 1, at p. x.
16. Ibid., at p. x.
17. Ibid., at p. x.
18. Citing Olson v. United States, 292 US. 246,257 (1934).19. The best interpretation of this sentence may be that it may not be sufficient to show that such a large interest holder suffering from a lack of control would want to take control (obviously, he would prefer to, that’s what gives rise to the discount). You also need to show that’s it’s likely to happen. (The term “reasonably probable” is doing quite a bit of work here). A future court, with somewhat different facts, might see things differently, but this particular sentence should be very troubling to the IRS if it intends to pursue this analysis in the future.