“Tax Affecting” Allowed Regarding the Vanderbilt’s Biltmore Estate
Cecil v. Commissioner provides insights on how to weigh evidence on valuation discounts
By Espen Robak, CFA
Ever since Gross,[1] the Tax Court has struggled with the valuation of pass-through entities, such as S corporations (S corps). The question is: should the earnings of an S corp be tax affected by imposing an assumed corporate tax rate (based on the taxes of similar C corps, for example) to pretax earnings, and then capitalizing those earnings in some way?
Recently, in Jones,[2] after a 20-year history of denying tax-affecting, one court agreed that tax affecting was allowable. However, just two years later, in Jackson,[3] another court didn’t find tax affecting appropriate.
Now, here comes Cecil,[4] decided yesterday, where the court is again allowing tax affecting in a valuation case. Cecil also provides another look at how the court weighs evidence on valuation discounts. Finally, the case is also instructive in its treatment of the asset approach when valuing an operating business.
Gifts of Shares in Biltmore Company
The case revolves around gifts made by William A.V. Cecil, Sr., and Mary Ryan Cecil, of class A (voting) and class B (generally non-voting) shares in the Biltmore Company (TBC) in November 2010. William’s mother was Cornelia Cecil nee Vanderbilt, the only child of George W. Vanderbilt, who built the Biltmore House in the Blue Ridge Mountains in Asheville, N.C., between 1889 and 1895.
Today, TBC owns the Biltmore House, a national tourist attraction, and its surrounding acreage, and operates a significant (1,300 employees) travel, tourism and historic hospitality business. Visitors to the historic estate can stay at its inn, shop at five retail outlets, eat at eight restaurants and engage in a variety of activities including tours, river rafting, fly fishing and equestrian training. The business generates significant revenue (about $70 million in 2010) and has been profitable every year since 1995, except for the recession year of 2008.
In total, the Cecils transferred to their children and certain trusts one share of class A common stock and 9,337 shares of class B common stock in total. The shares weren’t publicly traded and, beyond the normal illiquidity of privately held shares, were also restricted by the by-laws of TBC and its shareholders’ agreement.[5] The Cecil family owned all of the shares of TBC and had adopted formal written policies aimed at perpetuating family ownership and management of the company and its historic assets.
After the gift, the Cecils filed Forms 709 for 2010, valuing the gifted shares at a total appraised value of $20.9 million, based on a going-concern valuation.[6] The Internal Revenue Service, on the other hand, in its notice of deficiency, disregarded the existence of TBC entirely and attributed no weight to its going-concern value, valuing the gifts instead based on the liquidation value of the company’s assets. The court’s opinion doesn’t specify the exact NOD but based on the net asset value (NAV) method presented by the IRS expert at trial, it may have been more than $140 million. However, as we shall see, the IRS made a major climb-down before trial.
Valuation Battle
The taxpayers produced two appraisals from expert witnesses in court, both of whom appear to have valued the TBC shares at significantly lower values than taken on the tax return. The valuation methods used by the taxpayer’s experts included the discounted cash flow (DCF) method under the income approach and the guideline public companies (GPC) method and the similar transactions method under the market approach (both rejected the asset approach).[7]
Since the vast majority of the shares of TBC were gifted, for comparison, the total value of TBC (discounted) claimed on the gift tax return was approximately $22 million, while the total value implied by the taxpayer’s experts at trial was approximately $11 million.[8]
The IRS took a radically different view – its main expert first, using the NAV method under the asset approach, relied on appraisals of some assets along with his own estimate of certain other assets (such as TBC’s many trademarks) to arrive at a total NAV of $146.6 million. He then applied discounts based on discount data to this number to arrive at a total value of $92 million.[9]
However, he also applied what he called the discounted future benefits method, which appears to be a standard DCF, under the income approach. Using the DCF, he arrived at a value of approximately $15.2 million, but added non-operating assets to reach a total value of $36 million. Finally, he put almost all the weight on his DCF conclusion and took further discounts, to arrive at a value of $4,000 per class A share and from $3,066 (smaller blocks) to $3,276 (larger blocks) per class B share. Running these numbers, this implies a total value for TBC of approximately $31 million.
Thus, from a $22 million versus $140 million valuation gap at the NOD stage, to a $11 million to $31 million valuation gap at trial. The court still had quite a valuation assignment!
Court Ruling
The court had three major valuation judgments to make in arriving at its opinion:
- To what extent should the asset approach be considered in this valuation exercise?
- Should TBC earnings be tax affected in the income and market approaches?
- What valuation discounts are appropriate?
The asset approach. “In that TBC is an operating company whose existence does not appear to be in jeopardy, and not a holding company, we believe that TBC’s earnings rather than its assets are the best measure of the subject stock’s fair market value.” The court also found the IRS’ expert’s use of the asset approach to be inconsistent with appraisal standards as the holder of any of the gifted interests lack control and, thus, the ability to cause liquidation.[10]
The court held that the liquidation of TBC is unlikely, as any one holder of the subject interests would have to:
- “acquire additional shares in order to cause TBC’s liquidation”;
- “convince other shareholders to vote for liquidation”; or
- “wait until the shareholders of their heirs decide to liquidate”
The court also gave weight to the voting trust and shareholders’ agreement and testimony from family members.
Tax affecting. The court, after summarizing its 20-year long string of anti-tax affecting decisions from Gross to Giustina,[11] provided a deeper dive on the reasoning in Jackson and Jones. Noting that in Jones, “the parties agreed that a hypothetical buyer and seller would take into account the entity’s business form when determining the value of a limited partner interest” and that “the Commissioner disagreed with his experts,” the court noted that in Jones, the experts mostly agreed that tax affecting was appropriate, it was just the lawyers who disagreed.
In Jackson, however, the court didn’t find tax affecting appropriate. While in Jones, the “experts agreed to take into account the form of the business entity and agreed on the entity type,” in Jackson, “we held that tax affecting would not be appropriate because the estate’s experts had not persuaded us that the buyers would be C corporations.” However, even in Jackson the court didn’t find that there’s “a total bar against the use of tax affecting when the circumstances call for it.”
Thus, the Cecil court observed that “each side’s experts (…) totally agree that tax affecting should be taken into account (…) and experts on both sides agree on the specific method that we should employ to take that principle into account, we conclude that the circumstances of these cases require our application of tax affecting.” However, the court emphasized “that while we are applying tax affecting here, given the unique setting at hand, we are not necessarily holding that tax affecting is always, or even more often than not a proper consideration for valuing an S corporation.”[12]
Valuation Discounts
The court applies a discount for lack of control of 20%. It rejected the (higher) discount of the IRS expert because it was based on RELP and CEF data. These databases are in turn based on trading in holding companies, while TBC is an operating company.[13]
For the lack of marketability discount, the taxpayer’s experts applied discounts from 25% to 30%, while the IRS’ expert applied a discount in a range between 19% and 27%, depending on voting rights and size of the block.
The Court first rejected the analysis of one of the taxpayer’s experts, because it was based on: (1) studies of restricted stock that were too old (apparently, the data was mostly from the 1970s and 1980s), (2) studies of pre-initial public offering transactions, which the expert had admitted were unreliable, and (3) a put option analysis, which the court noted produced a range of discounts from 11.6% to 22.6% p and stated “we cannot fathom how that analysis supports his final discount rate of 30%.”
Finally, because the IRS’ expert had differentiated between smaller and larger blocks and between differing voting rights, the court felt that these discounts were appropriate. Interestingly, the court rejected the application of a voting rights discount (again, based on old studies) because each expert already accounted for “valuing a nonvoting minority interest.”
Refund for Taxpayer
In the end, the court accepted one of the taxpayer’s expert’s conclusions before tax affecting and before discounts. Discounts are applied at the levels opined at by the taxpayer’s expert for lack of control and at the IRS’ expert for lack of marketability. The court didn’t land on an exact value conclusion, but it appears that final judgment will be entered at a value substantially below the value on the Form 709, so this taxpayer will be getting a very nice refund!
[1] Gross v. Commissioner, TC Memo 1999-254.
[2] Estate of Jones v. Comm’r, TC Memo 2019-101.
[3] Estate of Jackson v. Comm’r, TC Memo 2021-48
[4] Estate of William A.V. Cecil, Sr., v. Comm’riand Estate of Mary R. Cecil, v Comm’r T.C. Memo 2023-24 (Feb. 28 2023).
[5] The by-laws provided for one vote per each of 7 class A shares, with majority voting for the board of directors. The board declared dividends by majority vote. Under the shareholders’ agreement, shareholders could transfer shares to any other shareholder or to any lineal descendent of such shareholder. Transfers outside of the family required written notice to each other shareholder, all of whom may then purchase such shares according to a formula clause. Furthermore, a voting trust agreement provided that any vote to sell any land, structure, asset or stock of TBC required the vote of two-thirds of the family trustees.
[6] The gifts were one share of class A stock from Mrs. Cecil, valued at $3,308, and 9,337 shares of class B stock from Mr. Cecil, valued at $2,236 per share. The Cecil’s split the gift and reported each a taxable gift of $10,438,766 (the math doesn’t quite work out, so presumably there’s rounding in some of the numbers). The appraisal attached with the 709 was done by Dixon Hughes.
[7] David Adams, of Adams Capital, applied the DCF, GPC, and similar transactions methods, and arrived at a value of $1,614.71 per class A share and $1,019 per class B share. George Hawkins, of Bannister Financial, applied the capitalization of net cash method and GPC method, and arrived at a value of $1,131 per class A share and $1,108 per class B share.
[8] TBC had 10,000 shares outstanding, 7 class A shares and 9,993 class B shares. Based on an average per share value of $1,063 per share, this implies a value of $10.6 million. (However, the Court’s opinion is unclear on whether the value of Mr. Hawkins noted above, supra, is his final concluded value, or just the value from the GPC method.)
[9] Two databases were used: on closed-end funds (CEF) and real estate limited partnerships (RELP). Both of these data-sources are widely used in the valuation profession as indicators of discounts for lack of control and (to some extent, for RELPs) lack of marketability. Confusingly, this is described in the Tax Court opinion as “multiples” applied. However, applying P/NAV multiples to a NAV is really just the inverse of discounting. Thus, he appears to have discounted the NAV by 37 percent ($92 million divided by a NAV of $146.6 million is an average “effective” P/NAV of 0.63, or a discount of 37%).
[10] USPAP Standards Rule 9-3 is cited. The Rule cautions that the approach should be used only when the interest being valued has the ability to “cause liquidation”.
[11] Estate of Giustina v. Commissioner, TC Memo 2011-141.
[12] In the end, it’s somewhat unclear how the Court resolves the tax affecting issue. Reference is made to “tax affecting” percentages between 17% and 25% (the Court adopts one of the appraisers’ 17% figure) but these are S-corp premiums (developed by the SEAM model). It isn’t explicitly stated in the opinion what tax rate the various appraisers tax-effected the pretax income of TBC at. Presumably, all the appraisers used the same tax rate, so the Court felt it was unnecessary to address the issue.
[13] To not leave the obvious unstated here: Yes, it’s odd that the IRS had a higher discount, but it’s even more odd that the data would result in higher discounts for a holding company versus an operating company. This is highly unusual.