McCord v. Comm'r , 120 T.C. 358 ( 2003 )

TABLE OF CONTENTS

FINDINGS OF FACT 361

rypTNnrvNr 367

367 I. Introduction .

368 II. Relevant Statutory Provisions .

369 III. Arguments of the Parties .

370 IV. Extent of the Rights Assigned.

373 V. Fair Market Value of the Gifted Interest .

373 A. Introduction .

373 1. General Principles .

374 2. Expert Opinions .

374 a. In General .

374 b. Petitioners’ Expert .

375 c. Respondent’s Expert .

375 B. Value of Underlying Assets .

376 C. Minority Interest (Lack of Control) Discount

376 1. Introduction .

376 2. Discount Factors by Asset Class .

376 a. Equity Portfolio.

377 (1) Measurement Date ..

377 (2) Sample of Funds .

(3) Representative Discount Within the Range of Sample Fund Discounts cn> CO

(4) Summary . © 00 CO

b. Municipal Bond Portfolio CO CO o

(1) Measurement Date .... CO CO I — ¹

(2) Sample of Funds . CO OO I — ¹

382 (3) Representative Discount Within the Range of Sample Fund Discounts.

383 (4) Summary .

383 c. Real Estate Partnerships .

383 (1) The Appropriate Comparables.

384 (2) Determining the Discount Factor.

386 d. Direct Real Estate Holdings .

386 e. Oil and Gas Interests .

386 3.Determination of the Minority Interest Discount.

387 D. Marketability Discount.

387 1. Introduction .

387 2. Traditional Approaches to Measuring the Discount.

387 a. In General .

388 b. Rejection of IPO Approach .

389 3. Mr. Frazier’s Restricted Stock Analysis .

390 4. Dr. Bajaj’s Private Placement Analysis.

390 a. Comparison of Registered and Unregistered Private Placements.

391 b. Refinement of Registered/Unregistered Discount Differential .

392 c. Further Adjustments .

393 d. Application to MIL .

393 5. Determination of the Marketability Discount.

393 a. Discussion.

395 b. Conclusion .

395 E. Conclusion.

395 VI. Charitable Contribution Deduction for Transfer to CFT .

395 A. Introduction.

396 B. The Assignment Agreement.

398 C. Conclusion .

399 VII. Effect of Children’s Agreement To Pay Estate Tax Liability

399 A. Introduction.

401 B. Discussion.

403 C. Conclusion .

404 VIII. Conclusion .

404 Judge Swift’s Concurring Opinion.

411 Judge Chiechi’s Concurring in Part, Dissenting in Part Opinion

416 Judge Foley’s Concurring in Part, Dissenting in Part Opinion ...

425 Judge Laro’s Dissenting Opinion .

Halpern, Judge:

By separate notices of deficiency dated April 13, 2000 (the notices), respondent determined deficiencies in Federal gift tax for calendar year 1996 with respect to petitioner Charles McCord, Jr. (Mr. McCord) and petitioner Mary McCord (Mrs. McCord) in the amounts of $2,053,525 and $2,047,903, respectively. The dispute centers around the gift tax consequence of petitioners’ assignments to several charitable and noncharitable donees of interests in a family limited partnership.

Unless otherwise noted, all section references are to the Internal Revenue Code in effect on the date of the assignments, and all Rule references are to the Tax Court Rules of Practice and Procedure. All dollar amounts have been rounded to the nearest dollar.

FINDINGS OF FACT

Some facts are stipulated and are so found. The stipulation of facts, with accompanying exhibits, is incorporated herein by this reference.

Petitioners

Petitioners are husband and wife. They have four sons, all adults (the children): Charles III, Michael, Frederick, and Stephen. In response to the notices, petitioners filed a single petition. At the time they filed the petition, petitioners resided in Shreveport, Louisiana.

Formation of McCord Interests, Ltd., L.L.P.

McCord Interests, Ltd., L.L.P. (MIL or the partnership), is a Texas limited partnership formed on June 30, 1995, among petitioners, as class A limited partners; petitioners, the children, and another partnership formed by the children (McCord Brothers Partnership), as class B limited partners; and the children as general partners (all such partners being hereafter referred to as the initial MIL partners).

On formation, as well as on the date of the assignments in question, the principal assets of mil were stocks, bonds, real estate, oil and gas investments, and other closely held business interests. On the date of the assignments, approximately 65 percent and 30 percent of the partnership’s assets consisted of marketable securities and interests in real estate limited partnerships, respectively. The remaining approximately 5 percent of the partnership’s assets consisted of direct real estate holdings, interests in oil and gas partnerships, and other oil and gas interests.

In mid-October 1995, the MIL partnership agreement was amended and restated, effective as of November 1, 1995 (such amended and restated partnership agreement being referred to hereafter as, simply, the partnership agreement). Attached to the partnership agreement is a schedule setting forth the capital contributions and ownership interests of the initial MIL partners, as follows:¹

Class and contributor Contribution Percentage interest

Class A limited partners:

Mr. McCord $10,000

Mrs. McCord 10,000

General partners:

Charles III 40,000 0.26787417

Michael 40,000 0.26787417

Frederick 40,000 0.26787417

Stephen 40,000 0.26787417

Class B limited partners:

Mr. McCord 6,147,192 41.16684918

Mrs. McCord 6,147,192 41.16684918

McCord Brothers 2,478,000 16.59480496

Total 14,952,384 100.0

Relevant Provisions of the Partnership Agreement

Among other things, the partnership agreement provides as follows:

MIL will continue in existence until December 31, 2025 (the termination date), unless sooner terminated in accordance with the terms of the partnership agreement.

Any class B limited partner may withdraw from MIL prior to the termination date and receive a payment equal to the fair market value (as determined under the partnership agreement) of such partner’s class B limited partnership interest (the put right).

Partners may freely assign their partnership interests to or for the benefit of certain family members and charitable organizations (permitted assignees).

A partner desiring to assign his partnership interest to someone other than a permitted assignee must first offer that interest to MIL and the other partners and assignees, who have the right to purchase such interest at fair market value (as determined under the partnership agreement).

The term “partnership interest” means the interest in the partnership representing any partner’s right to receive distributions from the partnership and to receive allocations of partnership profit and loss.

Regardless of the identity of the assignee, no assignee of a partnership interest can attain the legal status of a partner in MIL without the unanimous consent of all MIL partners.

MIL may purchase the interest of any “charity assignee” (i.e., a permitted assignee of a partnership interest that is a charitable organization that has not been admitted as a partner of MIL) at any time for fair market value, as determined under the partnership agreement (the call right).

For purposes of the partnership agreement, (1) a class B limited partner’s put right is disregarded for purposes of determining the fair market value of such partner’s class B limited partnership interest, and (2) any dispute with respect to the fair market value of any interest in MIL is to be resolved by arbitration as provided in Exhibit G attached to the partnership agreement.

Limited partners generally do not participate in the management of the partnership’s affairs. However, limited partners do have veto power with respect to certain “major decisions”, most notably relating to voluntary bankruptcy filings. In addition, if any two of the children are not serving as managing partners, class B limited partners have voting rights with respect to certain “large dollar” managerial decisions. Limited partners also have access to certain partnership financial information.

Southfield School Foundation

On November 20, 1995, petitioners assigned their respective class A limited partnership interests in MIL to the Hazel Kytle Endowment Fund of The Southfield School Foundation (the foundation) pursuant to an Assignment of Partnership Interest and Addendum Agreement (the Southfield agreement). The recitals to the Southfield agreement provide that “all of the partners of the Partnership desire that Assignee become a Class A Limited Partner of the Partnership upon execution of this Assignment of Partnership Interest” and “all consents required to effect the conveyance of the Assigned Partnership Interest and the admission of Assignee as a Class A Limited Partner of the Partnership have been duly obtained and are evidenced by the signatures hereto”. All of the initial MIL partners executed the Southfield agreement.

Further Assignments

On January 12, 1996 (the valuation date), petitioners, as assignors, entered into an assignment agreement (the assignment agreement) with respect to their class B limited partnership interests in MIL. The other parties to the assignment agreement (the assignees) were the children, four trusts for the benefit of the children (the trusts), and two charitable organizations — Communities Foundation of Texas, Inc. (cft) and Shreveport Symphony, Inc. (the symphony). By the assignment agreement, petitioners relinquished all dominion and control over the assigned partnership interests and assigned to the assignees all of their rights with respect to those interests. The assignment agreement does not contain language similar to that quoted above from the Southfield agreement regarding the admission of the assignees as partners of the partnership, and two of the partners of the partnership, McCord Brothers Partnership and the foundation, did not execute the assignment agreement in any capacity. The interests that petitioners assigned to the assignees by way of the assignment agreement (collectively, the gifted interest) are the subject of this dispute.

Under the terms of a “formula clause” contained in the assignment agreement (the formula clause), the children and the trusts were to receive portions of the gifted interest having an aggregate fair market value of $6,910,933; if the fair market value of the gifted interest exceeded $6,910,933, then the symphony was to receive a portion of the gifted interest having a fair market value equal to such excess, up to $134,000; and, if any portion of the gifted interest remained after the allocations to the children, trusts, and symphony, then CFT was to receive that portion (i.e., the portion representing any residual value in excess of $7,044,933). The children (individually and as trustees of the trusts) agreed to be liable for all transfer taxes (i.e., Federal gift, estate, and generation-skipping transfer taxes, and any resulting State taxes) imposed on petitioners as a result of the conveyance of the gifted interest.

The assignment agreement leaves to the assignees the task of allocating the gifted interest among themselves; in other words, in accordance with the formula clause, the assignees were to allocate among themselves the approximately 82-per-cent partnership interest assigned to them by petitioners. In that regard, the assignment agreement contains the following instruction concerning valuation (the valuation instruction):

For purposes of this paragraph, the fair market value of the Assigned Partnership Interest as of the date of this Assignment Agreement shall be the price at which the Assigned Partnership Interest would change hands as of the date of this Assignment Agreement between a hypothetical willing buyer and a hypothetical willing seller, neither being under any compulsion to buy or sell and both having reasonable knowledge of relevant facts. Any dispute with respect to the allocation of the Assigned Partnership Interests among Assignees shall be resolved by arbitration as provided in the Partnership Agreement.

The Confirmation Agreement

In March 1996, the assignees executed a Confirmation Agreement (the confirmation agreement) allocating the gifted interest among themselves as follows:

Assigned partnership Assignee interest

Charles T. McCord III GST Trust. 8.24977954%

Michael S. McCord GST Trust . 8.24977954

Frederick R. McCord GST Trust. 8.24977954

Stephen L. McCord GST Trust . 8.24977954

Charles III . 11.05342285

Michael . 11.05342285

Frederick . 11.05342285

Stephen . 11.05342285

CFT. 3.62376573

Symphony . 1.49712307

Total. 82.33369836

The assignees based that determination on an appraisal report, dated February 28, 1996, prepared at the behest of the children’s counsel² by Howard Frazier Barker Elliott, Inc. (HFBE). That report (the 1996 hfbe appraisal report) concludes that, taking into account discounts for lack of control and lack of marketability, the fair market value of a 1-per-cent “assignee’s interest in the Class B Limited Partnership Interests” on the valuation date was $89,505.

Representatives of CFT and the symphony, respectively (including their respective outside counsel), reviewed the 1996 HFBE appraisal report and determined that it was not necessary to obtain their own appraisals. Furthermore, under the terms of the confirmation agreement, CFT and the symphony (as well as the other assignees) agreed not to seek any judicial alteration of the allocation in the confirmation agreement and waived their arbitration rights granted under the assignment agreement.

MIL’s Exercise of the Call Right

On June 26, 1996, MIL exercised the call right with respect to the interests held by the symphony and CFT. It did so pursuant to a document styled “Agreement — Exercise of Call Option By McCord Interests, Ltd., L.L.P.” (the exercise agreement). The purchase price for the redeemed interests was based on a two-page letter from HFBE (the HFBE letter) previewing an updated appraisal report to be prepared by HFBE. The HFBE letter concludes that the fair market value of a 1-percent “assignee’s interest in the Class B Limited Partnership Interests” as of June 25, 1996 was $93,540. CFT and the symphony raised no objections to the value found in the HFBE letter and accepted $338,967 and $140,041, respectively, in redemption of their interests.

The Gift Tax Returns

Petitioners timely filed Forms 709, United States Gift (and Generation Skipping Transfer) Tax Return, for 1996 (the Forms 709). In schedules attached to the Forms 709, petitioners reported a gross value of $3,684,639 for their respective shares of the gifted interest. Each petitioner reduced that amount by the amount of Federal and State (Louisiana) gift tax generated by the transfer that the children agreed to pay as a condition of the gift. Each petitioner further reduced that amount by a computation of the actuarial value, as of the valuation date, of the contingent obligation of the children to pay (again, as a condition of the gift) the additional estate tax that would result from the transaction if that petitioner were to die within 3 years of the valuation date. Based on those adjustments, Mr. and Mrs. McCord reported total gifts of $2,475,896 and $2,482,605, respectively.³ Mr. and Mrs. McCord each claimed an annual exclusion amount of $60,000 and a charitable contribution deduction of $209,173, yielding taxable gifts of $2,206,724 and $2,213,432, respectively.

The Notices

By the notices, respondent determined deficiencies in gift tax with respect to Mr. and Mrs. McCord in the amounts of $2,053,525 and $2,047,903, respectively, based on increases in 1996 taxable gifts in the amounts of $3,740,904 and $3,730,439, respectively. Respondent determined that each petitioner (1) understated the gross value of his or her share of the gifted interest, and (2) improperly reduced such gross value by the actuarial value of the children’s obligation to pay additional estate taxes potentially attributable to the transaction.

OPINION

I. Introduction

MIL is a Texas limited partnership formed on June 30, 1995. In exchange for their class B limited partnership interests in MIL, petitioners contributed to MIL closely held business interests, oil and gas interests, real estate, stocks, bonds, and other securities. The parties have stipulated that the value of petitioners’ contributions in exchange for their class B limited partnership interests was $12,294,384 ($6,147,192 apiece). On January 12, 1996, petitioners assigned (as gifts) their partnership interests in MIL (the gifted interest). On that date, approximately 65 percent of the partnership’s assets consisted of marketable securities and approximately 30 percent consisted of interests in real estate limited partnerships. The assignees were petitioners’ children, trusts for the benefit of the children, and two charitable organizations (Communities Foundation of Texas, Inc. (CFT) and Shreveport Symphony, Inc. (the symphony)). In calculations submitted with their Federal gift tax returns, petitioners reported the gross value of the gifted interest as $7,369,278 ($3,684,639 apiece). Respondent’s adjustments reflect his determination that the gross fair market value of the gifted interest was $12,426,086 ($6,213,043 apiece). Principally, we must determine the fair market value of the gifted interest and whether each petitioner may reduce his or her one-half share thereof to account for the children’s contingent obligation (as a condition of the gift) to pay the additional estate tax that would result from the transaction if that petitioner were to die within 3 years of the date of the gift. Preliminarily, we must determine whether petitioners transferred all of their rights as class B limited partners or only their economic rights with respect to MIL. We must also determine the amount of the gift to CFT.

II. Relevant Statutory Provisions

Section 2501(a) imposes a tax on the transfer of property by gift. Section 2512(a) provides that, if a gift is made in property, the value of the property on the date of the gift is considered the amount of the gift. In determining the amount of “taxable gifts” for any particular year, a donor is entitled to deduct the amount of gifts made during the year that qualifies for the charitable contribution deduction provided for in section 2522. Sec. 2503(a).

Section 2502(c) provides that the donor is the person liable for the payment of the gift tax. If a donor makes a gift subject to the condition that the donee pay the resulting gift tax, the amount of the gift is reduced by the amount of such gift tax. Rev. Rul. 75-72, 1975-1 C.B. 310. Such a gift is commonly referred to as a “net gift” (net gift).

Under section 2035(c) (current section 2035(b)), a decedent’s gross estate is increased by the amount of any gift tax paid by the decedent or his estate on any gift made by the decedent during the 3-year period preceding the decedent’s death. For purposes of this “gross-up” provision, gift tax “paid by the decedent or his estate” on gifts made during the relevant 3-year period is deemed to include gift tax attributable to net gifts made by the decedent during such period (i.e., even though the donees are responsible for paying the gift tax in such situation). Estate of Sachs v. Commissioner, 88 T.C. 769, 777-778 (1987), affd. in part and revd. in part on another ground 856 F.2d 1158, 1164 (8th Cir. 1988).

III. Arguments of the Parties

Petitioners contend that they correctly valued the gifted interest in determining their respective taxable gifts. Petitioners contend in the alternative that, if we determine an increased value for the gifted interest, then, by operation of the formula clause in the assignment agreement, they are entitled to an identical increase in the amount of their aggregate charitable contribution deduction under section 2522, resulting in no additional gift tax due.⁴ Petitioners also contend that, under net gift principles enunciated in Rev. Rul. 75-72, supra, and Estate of Sachs v. Commissioner, supra, they properly reduced their respective taxable gifts by the actuarial value of the children’s contingent obligation, under the terms of the assignment agreement, to pay additional estate tax under section 2035(c).

Respondent contends that petitioners undervalued the gifted interest by mischaracterizing the assignment and applying excessive discounts. Respondent also contends that the formula clause in the assignment agreement, designed to neutralize the tax effect of any upward adjustment to the valuation of the gifted interest, is ineffectual. Finally, respondent contends that petitioners improperly reduced the amount of their taxable gifts to account for the possibility that the children would be obligated to pay additional estate tax under section 2035(c) by reason of the transaction.

The parties have stipulated that respondent bears the burden of proof, and we accept that stipulation.⁵

IV. Extent of the Rights Assigned

The divergence of the parties’ valuations of the gifted interest is attributable in part to their disagreement regarding the extent of the rights assigned by petitioners. Petitioners contend that they assigned to the assignees certain rights with respect to their class B limited partnership interests in MIL but did not (and could not) admit the assignees as class B limited partners. The assignment, they argue, did not entitle the assignees to exercise certain rights that petitioners possessed (as partners) under the partnership agreement. Thus, they argue, the value of the gifted interest is something less than the value of all of their rights as class B limited partners. Respondent, on the other hand, argues that the gifted interest consists of the sum and total of petitioners’ rights as class B limited partners (i.e., that, as a result of the assignment, the assignees became class B limited partners). The parties’ experts agree that, if the gifted interest does not include all of petitioners’ rights as class B limited partners, the fair market value of the gifted interest is lower than it would be if the gifted interest did include all such rights.

Whenever the concept of “property” is relevant for Federal tax purposes, it is State law that defines the property interest to which Federal tax consequences attach. E.g., United States v. Craft, 535 U.S. 274, 278-279 (2002) (Federal tax lien attaches to property held, under State law, as tenants by the entireties). Thus, in order to determine the Federal gift tax consequences that attach to petitioners’ assignment of the gifted interest, we look to applicable State law to determine the extent of the rights transferred. Because petitioners transferred interests in a Texas limited partnership, Texas law governs our determination in that regard. Specifically, we look to the Texas Revised Limited Partnership Act (the Act), Tex. Rev. Civ. Stat. Ann. art. 6132a-l, as in effect on the date of the gift.

Under the Act, a partnership interest is personal property. Tex. Rev. Civ. Stat. Ann. art. 6132a-l, sec. 7.01 (Vernon 2001). A partnership interest is assignable in whole or in part unless the partnership agreement provides otherwise. Id. sec. 7.02(a)(1). However, an assignee of a partnership interest attains the legal status of a limited partner only if the partnership agreement so provides or all of the partners consent. Id. sec. 7.04(a). Section 8.01 of the partnership agreement governs the admission of new partners to mil. That section provides that, notwithstanding the occurrence of a valid assignment of a partnership interest in MIL in compliance with the terms of the partnership agreement, no person shall become a partner without the unanimous consent of the existing partners.

There is no evidence indicating that all of the MIL partners explicitly consented to the admission of the assignees as partners in MIL. Our inquiry does not end there, however. In Kerr v. Commissioner, 113 T.C. 449 (1999), affd. on another issue 292 F.3d 490 (5th Cir. 2002), we demonstrated our willingness to look beyond the formalities of intrafamily partnership transfers to determine what, in substance, was transferred. In that case, also involving Texas partnership law, the taxpayers argued that the interests they transferred to two grantor retained annuity trusts (GRATs) were “assignee interests”⁶ because the other general partners of the partnership (the taxpayers’ adult children, whose trusts were the remainder beneficiaries of the GRATs) did not consent to the admission of the GRATs as additional partners. Id. at 464. We found that such lack of formal consent did not preclude a finding that the taxpayers effected a transfer of all their rights as partners. Id. In essence, we inferred the necessary consent of the other general partners to admit the GRATs as partners based on all of the facts and circumstances. Id. at 464-468.

Our decision in Kerr was influenced by a number of factors that are not present in this case. For instance, the taxpayers in Kerr asked us to construe strictly the consent provision in their partnership agreement in the context of their transfers to the GRATs, notwithstanding the fact that they had disregarded that provision in other situations. Id. at 464-465. In addition, we found it difficult to reconcile the taxpayers’ characterization of the transfers with the language of their assignment documents, each of which contained the following statement: “The Assigned Partnership Interest constituted a Class B Limited Partnership interest in * * * [the partnership at issue] when owned by Assignor and, when owned by Assignee, shall constitute a Class B Limited Partnership Interest in said partnership.” Id. at 466. Finally, from an economic reality standpoint, we found it significant that the taxpayers and their children, being all of the general partners of the partnership, could have formally admitted the assignee GRATs as partners at any time without having to obtain the consent of any other person. Id. at 468.

In the instant case, there is no evidence that petitioners and the other partners of MIL ever disregarded section 8.01 of the partnership agreement, the provision on which they now rely. Indeed, when petitioners assigned their class A limited partnership interests to the foundation in 1995, all of the initial MIL partners consented in writing to the admission of the foundation as a class A limited partner, as required by said section 8.01. Furthermore, the assignment agreement with respect to the gifted interest does not contain language of the type quoted above from the Kerr assignment documents, nor does it contain any of the language in the South-field agreement relating to the admission of the assignee as a partner in mil. Finally, petitioners and their children could not unilaterally admit the assignees as partners in MIL; any such admission required the consent of the foundation, an unrelated third party.

Respondent makes note of the fact that the assignment agreement provides that “Assignors hereby relinquish all dominion and control over the Assigned Partnership Interest and assign to Assignees all of Assignors’ rights with respect to the Assigned Partnership Interest”. However, the issue in this case is not whether petitioners transferred partnership interests; under the terms of the Act, the partnership agreement, and the assignment agreement, they undoubtedly could and did. That having been said, both the Act and the partnership agreement define the term “partnership interest” in terms of economic rights (and do not equate the term with membership in the partnership).⁷ Thus, it is entirely consistent to say that petitioners assigned all of their rights with respect to their partnership interests, yet did not assign all of their rights as class B limited partners (i.e., did not cause the assignees to be admitted as substitute class B limited partners).

In sum, we conclude that the facts in this case do not permit us to infer, as we did in Kerr, that petitioners intended to transfer all of their rights as partners and that all of the other partners effectively consented to the admission of the assignees as partners. Rather, petitioners assigned only economic rights with respect to mil, and we shall proceed to value the gifted interest on that basis.⁸

V. Fair Market Value of the Gifted Interest

A. Introduction

1. General Principles

Section 25.2512-1, Gift Tax Regs., provides that the value of property for gift tax purposes is “the price at which such property would change hands between a willing buyer and a willing seller, neither being under any compulsion to buy or sell, and both having reasonable knowledge of relevant facts.”⁹ The willing buyer and willing seller are hypothetical persons, rather than specific individuals or entities, and their characteristics are not necessarily the same as those of the donor and the donee. Estate of Newhouse v. Commissioner, 94 T.C. 193, 218 (1990) (citing Estate of Bright v. United States, 658 F.2d 999, 1006 (5th Cir. 1981)).¹⁰ The hypothetical willing buyer and willing seller are presumed to be dedicated to achieving the maximum economic advantage. Id.

2. Expert Opinions

a. In General

In deciding valuation cases, courts often look to the opinions of expert witnesses. Nonetheless, we are not bound by the opinion of any expert witness, and we may accept or reject expert testimony in the exercise of our sound judgment. Helvering v. Natl. Grocery Co., 304 U.S. 282, 295 (1938); Estate of Newhouse v. Commissioner, supra at 217. Although we may largely accept the opinion of one party’s expert over that of the other party’s expert, see Buffalo Tool & Die Manufacturing Co. v. Commissioner, 74 T.C. 441, 452 (1980), we may be selective in determining what portions of each expert’s opinion, if any, to accept, Parker v. Commissioner, 86 T.C. 547, 562 (1986). Finally, because valuation necessarily involves an approximation, the figure at which we arrive need not be directly traceable to specific testimony if it is within the range of values that may be properly derived from consideration of all the evidence. Estate of True v. Commissioner, T.C. Memo. 2001-167 (citing Silverman v. Commissioner, 538 F.2d 927, 933 (2d Cir. 1976), affg. T.C. Memo. 1974-285).

b. Petitioners’ Expert

Petitioners offered William H. Frazier (Mr. Frazier) as an expert witness to testify concerning the valuation of the gifted interest. Mr. Frazier is a principal of Howard Frazier Barker Elliott, Inc. (hfbe), a Houston-based valuation and consulting firm. He is a senior member of the American Society of Appraisers and has been involved in valuation and general investment banking activities since 1975. The Court accepted Mr. Frazier as an expert in the valuation of closely held entities and received written reports of HFBE into evidence as Mr. Frazier’s direct and rebuttal testimony, respectively.

In his direct testimony, Mr. Frazier concludes that, based on a 22-percent minority interest discount and a 35-percent marketability discount, the fair market value of a 1-percent assignee interest in MIL on the valuation date was $89,505. Based on that figure, Mr. Frazier further concludes that the fair market value of each petitioner’s one-half share of the gifted interest on the valuation date was $3,684,634 (a figure that, due to rounding, is slightly lower than the value reported on the Forms 709).¹¹

c. Respondent’s Expert

Respondent offered Mukesh Bajaj, Ph.D. (Dr. Bajaj), as an expert witness to testify concerning the valuation of closely held entities. Dr. Bajaj is the managing director of the finance and damages practice of LEC, LLC. He also has experience as a university professor of finance and business economics. Dr. Bajaj has lectured extensively on valuation issues, and he has testified as an expert in several valuation cases. The Court accepted Dr. Bajaj as an expert in the valuation of closely held entities and received his written reports into evidence as his direct and rebuttal testimony, respectively.

In his direct testimony, Dr. Bajaj concludes that, based on an 8.34-percent minority interest discount and a 7-percent marketability discount, the fair market value of a 1-percent assignee interest in mil on the valuation date was $150,665.64. Based on that figure, Dr. Bajaj further concludes that the fair market value of each petitioner’s one-half share of the gifted interest on the valuation date was $6,202,429.67.

B. Value of Underlying Assets

The parties agree that, on the valuation date, mil’s net asset value (NAV) was $17,673,760, broken down by asset class as follows:

Asset type Value

Equities portfolio . $3,641,956

Bond portfolio . 8,040,220

Real estate partnerships . 5,194,933

Real estate . 581,553

Oil and gas interests . 215,098

Total. 17,673,760

In determining the value of the gifted interest, Mr. Frazier first (i.e., before applying any discounts) subtracts $20,000 from mil’s NAV to reflect the class A limited partner’s $20,000 priority claim against mil’s assets under the terms of the partnership agreement.¹² Dr. Bajaj makes no such preliminary adjustment. We concur with Mr. Frazier’s approach in that regard, and, therefore, we conclude that the appropriate base amount for determining the value of the gifted interest is $17,653,760.¹³

C. Minority Interest (Lack of Control) Discount

1. Introduction

A hypothetical buyer of the gifted interest would have virtually no control over his investment. For instance, such holder (1) would have no say in mil’s investment strategy, and (2) could not unilaterally recoup his investment by forcing MIL either to redeem his interest or to undergo a complete liquidation. Mr. Frazier and Dr. Bajaj agree that the hypothetical “willing buyer” of the gifted interest would account for such lack of control by demanding a reduced sales price; i.e., a price that is less than the gifted interest’s pro rata share of mil’s NAV. They further agree that, in the case of an investment company such as MIL, the minority interest discount should equal the weighted average of minority interest discount factors determined for each type of investment held by MIL: Equities, municipal bonds, real estate interests, and oil and gas interests.

2. Discount Factors by Asset Class

a. Equity Portfolio

Mr. Frazier and Dr. Bajaj both determine the minority interest discount factor for mil’s equity portfolio by reference to publicly traded, closed end equity investment funds. Specifically, they both derive a range of discounts by determining for a sample of closed end equity funds the discount at which a share of each sample fund trades relative to its pro rata share of the NAV of the fund.¹⁴ They differ in their selection of measurement dates, sample funds, and representative discounts within the range of the sample fund discounts.

(1) Measurement Date

Mr. Frazier calculates discounts for his sample of closed end equity funds on the basis of January 11, 1996, trading prices and December 22, 1995, NAV information. Dr. Bajaj, on the other hand, utilizes trading prices and NAV data as of the valuation date; i.e., January 12, 1996. We agree with Dr. Bajaj that, to the extent possible, data from January 12, 1996, should be utilized to determine discounts with respect to the sample funds.

(2) Sample of Funds

Mr. Frazier derives his sample of closed end equity funds from the list of “domestic equity funds” set forth in Momingstar’s Mutual Funds Guide. From that list, he purports to exclude from consideration “special purpose” funds (i.e., those primarily invested in a specific industry), funds with a stated maturity, and funds “that had provisions regarding votes to open-end the fund.¹⁵ That screening process produced a sample of 14 funds.

Dr. Bajaj derives his sample of closed end equity funds from the list of “general equity” funds set forth in the January 12, 1996, edition of the Wall Street Journal. For reasons not entirely clear, Dr. Bajaj excludes two of those funds from consideration, leaving a sample of 20 funds.¹⁶

Dr. Bajaj’s sample contains nine funds that Mr. Frazier excludes from his sample. With regard to the first two of Mr. Frazier’s three screening criteria, Dr. Bajaj states in his rebuttal testimony that none of those nine funds was a special purpose fund and that none had a stated maturity date. With regard to Mr. Frazier’s third screening criterion, Dr. Bajaj states that the fact that a fund’s shareholders can vote to open-end the fund does not mean that such a conversion is imminent. Dr. Bajaj also states that the summary descriptions (contained in Mr. Frazier’s direct testimony) of five of the funds included by Mr. Frazier in his sample of funds mention open-ending votes or procedures, which, according to Mr. Frazier’s criteria, should have required their exclusion.

In his rebuttal testimony, Mr. Frazier does not directly challenge Dr. Bajaj’s inclusion of any specific fund in his sample; rather, he simply asserts that “some of these funds could have announced their intent to convert to an open-end fund” and that “other funds may be non-diversified”. In the absence of more specific objections to Dr. Bajaj’s additional sample funds, we are persuaded to include such funds in our own analysis.

Mr. Frazier’s sample contains three funds that Dr. Bajaj excludes from his sample: Gemini II, Quest for Value, and Liberty All Star Growth Fund. Gemini II and Quest for Value were “dual purpose” funds, which were scheduled for either liquidation or open-ending in January 1997.¹⁷ Given the effect that the impending liquidation or conversion may have had on share prices of those funds, we exclude them from our analysis. Since Dr. Bajaj’s rebuttal testimony raises no specific objection to the inclusion of Liberty All Star Growth Fund in the sample, we include that fund in our analysis.¹⁸

(3) Representative Discount Within the Range of Sample Fund Discounts

Mr. Frazier concludes that, because an interest in mil’s equity portfolio would not compare favorably to an interest in an institutional fund, the minority interest discount factor for mil’s equity portfolio should derive from the higher end of the sample’s range of discounts. Dr. Bajaj, on the other hand, concludes that such discount factor should derive from the lower end of the range of discounts. For the reasons discussed below, we find neither expert’s arguments convincing on that point.

Mr. Frazier concludes that a higher than average minority interest discount factor for mil’s equity portfolio is warranted in part because of the relative anonymity of mil’s investment managers, the relatively small size of mil’s equity portfolio, and mil’s policy of not making distributions (other than distributions to satisfy tax obligations). However, Mr. Frazier elsewhere testifies that, based on his regression analysis, there is no clear correlation between the discounts observed in his sample of closed end funds and any of the variables he analyzed, including Morningstar rating (largely indicative of management reputation), the size of the fund, and distributions as a percentage of NAV. We are similarly unpersuaded by Mr. Frazier’s assertion (unsupported by empirical evidence) that fewer administrative and regulatory controls on mil’s investment activity (as compared to that of institutional funds) should result in a higher discount factor as a matter of course.¹⁹

Dr. Bajaj’s argument that the minority interest discount factor for mil’s equity portfolio should derive from the lower end of the range of observed discounts is based primarily on the premise that, on the valuation date, MIL was akin to a new investment fund. Dr. Bajaj’s research, along with that of others cited in his direct testimony, indicates that new investment funds tend to trade at lower discounts than seasoned funds. However, Dr. Bajaj’s analysis fails to recognize that, while MIL was a relatively new entity on the valuation date, its equity portfolio had been in place (in the hands of the contributing partners) for years. Furthermore, of the four factors that Dr. Bajaj specifically identifies as possible determinants of lower initial fund discounts, only one — lack of unrealized capital gains — perhaps would have informed the pricing decision of a hypothetical buyer of an interest in MIL.²⁰ The other factors cited by Dr. Bajaj (the initial diminution of fund NAV relative to issue proceeds due to flotation and other startup costs, the prevalence of new funds in “hot” investment sectors, and the initial lack of management inefficiencies) simply do not readily translate from the public capital markets to the hypothetical private sale of an interest in MIL.

Because we are unpersuaded by the respective arguments of Mr. Frazier and Dr. Bajaj for a higher than average or lower than average minority interest discount factor for mil’s equity portfolio, we utilize the average discount of the sample funds under consideration.²¹

(4) Summary

In determining the appropriate minority interest discount factor for mil’s equity portfolio, we utilize (1) Dr. Bajaj’s price and NAV data as of January 12, 1996 (with the exception of Liberty All Star Growth Fund, for which we utilize NAV data from January 5, 1996, contained in the record); (2) Dr. Bajaj’s sample of funds, with the addition of Liberty All Star Growth Fund; and (3) the average discount of the sample funds. The resulting discount factor is 9.96 percent, which we round up to 10 percent.

b. Municipal Bond Portfolio

Both Mr. Frazier and Dr. Bajaj determine the minority interest discount factor for mil’s municipal bond portfolio by reference to publicly traded, closed end municipal bond investment funds. Once again, they disagree on measurement dates, sample funds, and representative discounts within the range of the sample fund discounts.

(1) Measurement Date

Mr. Frazier calculates discounts for his sample closed end municipal bond funds on the basis of January 11, 1996, trading prices and December 25, 1995, NAV data. Dr. Bajaj utilizes trading prices and NAV information as of the valuation date; i.e., January 12, 1996. We agree with Dr. Bajaj that, to the extent possible, data from January 12, 1996, should be utilized in determining discounts with respect to the sample funds.

(2) Sample of Funds

Mr. Frazier derives his sample of closed end municipal bond funds from the list of municipal bond funds set forth in Morningstar’s Mutual Funds Guide. In his direct testimony, Mr. Frazier indicates that he excluded from consideration funds that were “heavily weighted toward a specific sector” and funds with scheduled liquidation dates. With regard to the first screening factor, it appears that Mr. Frazier was referring to single-State funds. Mr. Frazier’s screening process produced a sample of eight funds.

Dr. Bajaj derives his sample from the list of 140 closed end municipal bond funds set forth in the January 15, 1996 edition of the Wall Street Journal. For reasons not entirely clear, Dr. Bajaj excludes six of the funds from consideration, leaving a sample of 134 funds.²² That sample includes numerous single-State funds and funds with scheduled liquidation dates.

We agree with Mr. Frazier that funds with scheduled liquidation dates should not be included in the sample. However, given the fact that Louisiana-based obligations accounted for approximately 75 percent of the value of mil’s bond portfolio, we are somewhat puzzled by Mr. Frazier’s exclusion of single-State funds from his sample. Indeed, we believe that the sample should consist entirely of single-State funds. We therefore utilize a sample consisting of the 62 single-State funds in Dr. Bajaj’s sample that do not have scheduled liquidation dates.

(3) Representative Discount Within the Range of Sample Fund Discounts

As is the case with mil’s equity portfolio, Mr. Frazier concludes that the minority interest discount factor for mil’s bond portfolio should derive from the higher end of the sample’s range of discounts, while Dr. Bajaj concludes that such discount factor should derive from the lower end of the range of discounts. Once again, we find neither expert’s arguments convincing on this point.

Mr. Frazier states that, according to his regression analysis, the three factors that are the most determinative of discounts with respect to the closed end funds in his sample are (1) distributions as a percentage of NAV, (2) built-in gain as a percentage of NAV, and (3) 3-year average annual return. We see no error in Mr. Frazier’s calculation of his first factor, although he seems to take the same factor into account as an aspect of the discount for lack of marketability. With regard to the second two factors, Mr. Frazier provides no data with respect to mil’s bond portfolio that can be compared to the data from his sample funds. Mr. Frazier also repeats factors that he deemed relevant in the context of mil’s equity portfolio, notwithstanding the fact that his own regression analysis indicates little, if any, correlation between those factors (management quality and the size of the fund) and the level of discounts in his bond fund sample.²³

Dr. Bajaj applies his “new fund” analysis, discussed above in the context of mil’s equity portfolio, to mil’s bond portfolio as well. Again, Dr. Bajaj’s analysis fails to recognize that, while MIL was a relatively new entity on the valuation date, its bond portfolio had been in place (in the hands of the contributing partners) for years. For that reason and the other reasons discussed supra in section V.C.2.a.(3), we reject this portion of Dr. Bajaj’s analysis.

Because we are unpersuaded by the respective arguments of Mr. Frazier and Dr. Bajaj for a higher than average or lower than average minority interest discount factor for mil’s bond portfolio, we utilize the average discount of the sample funds under consideration.²⁴

(4) Summary

In determining the appropriate minority interest discount factor for mil’s bond portfolio, we utilize (1) Dr. Bajaj’s price and NAV data as of January 12, 1996, (2) a sample of funds consisting of the 62 single-State funds in Dr. Bajaj’s sample that do not have scheduled liquidation dates, and (3) the average discount of the sample funds. The resulting discount factor is 9.76 percent, which we round up to 10 percent.

c. Real Estate Partnerships²⁵

(1) The Appropriate Comparables

In contrast to their opinions regarding mil’s equity and bond portfolios, Dr. Bajaj and Mr. Frazier sharply disagree on the general type of publicly traded entity from which to extrapolate the minority interest discount factor for mil’s real estate partnership interests. Dr. Bajaj argues that the discount factor should be based on data pertaining to real estate investment trusts (reits).²⁶ Mr. Frazier, on the other hand, excludes REITs from consideration “since they are primarily priced on a current yield basis because REITs are required by law to annually pay out a large portion of earnings to shareholders.” That justification overlooks the fact that the investment funds Mr. Frazier analyzes in determining the minority interest discount factors for mil’s equity and bond portfolios are also required to distribute substantially all of their income each year in order to maintain their tax-favored status as regulated investment companies (RICs). Compare sec. 852(a)(1) (income distribution requirement for RlCs) with sec. 857(a)(1) (income distribution requirement for REITs). In the absence of any explanation as to why the current distribution requirement should disqualify REITs (but not RlCs) from consideration in our analysis, we are persuaded to evaluate the REIT data.

We are further persuaded to utilize the REIT data in light of the alternative offered by Mr. Frazier. Mr. Frazier’s search for “comparable” publicly traded real estate companies yielded a sample of five companies, and he derives his range of discounts from only three of those companies. While we have utilized small samples in other valuation contexts, we have also recognized the basic premise that “[a]s similarity to the company to be valued decreases, the number of required comparables increases”. Estate of Heck v. Commissioner, T.C. Memo. 2002-34. One of Mr. Frazier’s three sample companies developed planned communities, conducted farming operations, and owned royalty interests in more than 200 oil and gas wells. Another owned and managed shopping centers and malls and developed the master-planned community of Columbia, Maryland. The assets and activities of those companies are not sufficiently similar to those of mil’s real estate partnerships to justify the use of such a small sample.²⁷

In contrast, Dr. Bajaj’s REIT sample consists of 62 companies. In recognition of the fact that two of the real estate limited partnerships in which MIL was a partner owned unimproved land that could be used for a variety of purposes, Dr. Bajaj’s sample includes REITs specializing in a broad array of real estate investments, including office, residential, and retail properties. Given the size of the sample, we believe that any dissimilarities in the assets and activities of particular REITs in the sample as compared to those of mil’s real estate partnerships are tolerable.²⁸

(2) Determining the Discount Factor

Because REITs offer investors the opportunity to invest in an illiquid asset (i.e., the underlying real estate) in liquid form (i.e., the REIT shares), investors in REITs are willing to pay a liquidity premium (relative to NAV) to invest in REIT shares. According to Dr. Bajaj, that does not imply that a minority discount is nonexistent; it only means that the difference between price and NAV for a REIT may have two components, one positive (the liquidity premium) and one negative (the minority discount). From his sample data, Dr. Bajaj calculated a median price-to-NAV premium of 3.7 percent. To be conservative and to reflect mil’s distribution policy, Dr. Bajaj looked below the median, to the 25th percentile, and began with a price-to-NAV discount of 1.3 percent (an adjustment to NAV of minus (-) 1.3 percent). Since Dr. Bajaj believes that that adjustment reflects both a minority discount and a (smaller) liquidity premium, he then proceeded to identify (and eliminate the effect of) the liquidity premium in order to determine the minority discount. Based on his opinion that, as of the valuation date, the prevailing “illiquidity” discount for privately placed restricted stock was approximately 7 percent, he calculated a 7.53-percent liquidity premium.²⁹ Based on that liquidity premium of 7.53 percent and his selected price-to-NAV discount of 1.3 percent from his REIT sample, Dr. Bajaj added the two percentages to calculate a minority discount of 8.83 percent (i.e., he increased the price-to-NAV discount to reflect the elimination of the effect of the liquidity premium), which he rounded to 9 percent.

Using the same procedure as Dr. Bajaj, but substituting an illiquidity discount of 18 percent for his 7-percent figure, we arrive at a liquidity premium of 22 percent and therefore conclude that the minority discount embedded in the 1.3-per-cent price-to-NAV discount selected from the REIT sample is 23.3 percent, which we shall apply to mil’s real estate partnership interests. We have substituted 18 percent for 7 percent because, as discussed infra in section V.D.5.a., Dr. Bajaj has not been clear in distinguishing between the apparently different (but overlapping) concepts of “marketability” and “liquidity”. Our substitute percentage derives from a published study referenced in his direct testimony (the Wruck study)³⁰ which reported that, on average, discounts observed in private placements of unregistered shares exceeded those observed in private placements of registered shares (freely tradable in the public market) by 17.6 percentage points, which we round up to 18 percent. The theory, discussed in more detail infra in section V.D.4., is that such additional discount represents, to some degree, pure illiquidity concerns, since a ready, public market is available to owners of registered shares and unavailable to owners of restricted shares.

d. Direct Real Estate Holdings

Respondent has instructed Dr. Bajaj to base his value conclusion regarding mil’s direct real estate holdings on the 40-percent minority interest discount factor for those assets appearing in the 1996 HFBE appraisal report. On that basis, we find that the minority interest discount factor for mil’s direct real estate holdings is 40 percent.

e. Oil and Gas Interests

Mr. Frazier assigns a 33.5-percent minority interest discount factor to mil’s oil and gas interests. Respondent has instructed Dr. Bajaj to base his value conclusion regarding the oil and gas component of mil’s portfolio on the 33.5-per-cent minority interest discount factor for those assets appearing in the 1996 HFBE appraisal report. On that basis, we find that the minority interest discount factor for mil’s oil and gas investments is 33.5 percent.

3. Determination of the Minority Interest Discount

The minority interest discount factors determined above yield a weighted average discount of 15.18 percent, determined as follows:³¹

Percent Percent Percent of discount weighted Asset class NAV factor average

10.0 Equities CD O O <N

10.0 Bonds lO to to ^

Asset class Percent Percent Percent of discount weighted NAV factor average

R.E. partnerships 29.4 23.3 6.85

Real estate 3.3 40.0 1.32

Oil and gas 1.2 33.5 0.40

Discount 15.18

Rounding to the nearest percentage point, we conclude that the appropriate minority interest discount with respect to the gifted interest is 15 percent.

D. Marketability Discount

1. Introduction

The parties agree that, to reflect the lack of a ready market for assignee interests in mil, an additional discount (after applying the minority interest discount) should be applied to the net asset value of mil’s assets to determine the fair market value of the gifted interest. Such a discount is commonly referred to as a “marketability discount”. The marketability discount analyses of Mr. Frazier and Dr. Bajaj differ from their minority interest discount analyses in that they seek to identify a single, “entity-wide” discount rather than a weighted average of discount factors specific to each category of assets held by MIL. The parties disagree as to the amount of that discount.

2. Traditional Approaches to Measuring the Discount

a. In General

Mr. Frazier and Dr. Bajaj agree that empirical studies of the marketability discount fall into two major categories. The first major category, the IPO approach, consists of studies that compare the private-market price of shares sold before a company goes public with the public-market price obtained in the initial public offering (IPO) of the shares or shortly thereafter. The second major category, the restricted stock approach, consists of studies that compare the private-market price of restricted shares of public companies (i.e., shares that, because they have not been registered with the Securities and Exchange Commission, generally cannot be sold in the public market for a 2-year period)³² with their coeval public-market price. Mr. Frazier relies primarily on data from the restricted stock approach to support a marketability discount of 35 percent, although he also contends that data from the IPO approach strongly support that level of discount. Dr. Bajaj relies on a variant of the restricted stock approach, which we shall refer to as the private placement approach, to support a marketability discount of 7 percent.

b. Rejection of IPO Approach

Dr. Bajaj argues that the IPO approach is flawed both in concept and in application. His principal criticism is that the IPO premium (over the pre-IPO private market price) may reflect more than just the availability of a ready market. He believes that buyers of shares prior to the IPO are likely to be insiders who provide services to the firm and who are compensated, at least in part, by a bargain price. More importantly, he believes that a pre-IPO purchaser demands compensation (in the form of a lower price) for bearing the risk that the IPO will not occur or will occur at a lower price than expected. His opinion is: “[T]he IPO approach probably generates inflated estimates of the marketability discount. Consequently, it is of limited use in estimating the value of closely held firms.”

In his rebuttal testimony, Mr. Frazier fails to offer any rebuttal of Dr. Bajaj’s criticism of the IPO approach. Mr. Frazier’s support for the IPO approach consists only of his reference to a series of studies undertaken by Shannon Pratt, chairman of Willamette Management Associates, Inc., a national business valuation firm (the Willamette studies). Without explaining the basis of his testimony, Mr. Frazier’s opinion is: “[T]he evidence from the Willamette study was quite compelling and offered strong support for the hypothesis that the fair market values of minority interests in privately held companies were and should be greatly discounted from their publicly-traded counterparts.”

By contrast, in his rebuttal testimony, Dr. Bajaj offers a compelling criticism of both the Willamette studies and another series of studies undertaken by John D. Emory, vice president of appraisal services at Robert W. Baird & Co., a regional investment banking firm.³³ He concludes that the latest study conducted by Mr. Emory is biased because it does not adequately take into account the highest sale prices in pre-IPO transactions, and he criticizes the Willamette studies for not disclosing enough data to reveal whether they suffer from a similar bias. Dr. Bajaj has convinced us to reject as unreliable Mr. Frazier’s opinion to the extent it is based on the IPO approach. We shall proceed to consider Mr. Frazier’s restricted stock analysis and Dr. Bajaj’s private placement analysis.

3. Mr. Frazier’s Restricted Stock Analysis

Mr. Frazier reviews four studies under the restricted stock approach³⁴ and then attempts to infer an appropriate marketability discount by “placing” MIL within the range of observed discounts in those studies and the Willamette studies, on the basis of certain characteristics of MIL (revenue, income, and NAV) and the gifted interest (size of the interest, expressed both as a percentage of MIL and as a dollar amount). The results of that attempt are set forth in Table 31 of the report constituting Mr. Frazier’s direct testimony (Table 31). Based on data from the five studies, Mr. Frazier identifies 10 hypothetical discount levels for the gifted interest (some expressed as a specific percentage, e.g., “33.6%”, and some expressed as being greater or less than a specific percentage, e.g., “>35%”). Six of the hypothetical discounts were greater than 35 percent and four were less than 35 percent. He states: “Based on these studies alone, we concluded that the discount applicable to the Partnership’s block should approximate 35 percent.”

We find several flaws in Mr. Frazier’s analysis. For example, Table 31 indicates that mil’s projected revenue of $681,000 is consistent with a discount of 51.9 percent based on data from the Willamette studies. The Willamette studies are IPO studies rather than restricted stock studies, and they do not, so far as we can tell from Mr. Frazier’s testimony, analyze firm revenues.³⁵ Table 31 also indicates that one can infer a discount from the “Hertzel and Smith” study based on the proportional size of the offering, although Mr. Frazier gives no indication that that study drew any correlations between that variable and the level of discount. Furthermore, under the heading “Size of Block as % Total Outstanding” in Table 31, the entry corresponding to MIL is “1.0%”, when in fact each half of the gifted interest represents a greater than 40-percent interest with respect to MIL. Similarly, although there is no entry in Table 31 for the dollar size of the gifted interest, it is evident that the “>35%” discount inferred from that variable in Table 31 is based on the same mischaracterization of each half of the gifted interest as a 1-percent interest in MIL.³⁶ In light of those numerous defects, we give little weight to Mr. Frazier’s restricted stock analysis.

4. Dr. Bajaf s Private Placement Analysis

a. Comparison of Registered and Unregistered Private Placements

Dr. Bajaj believes that the discounts observed in restricted stock studies are attributable in part to factors other than impaired marketability.³⁷ In support of his position, Dr. Bajaj analyzes data from studies (including his own unpublished study)³⁸ involving both registered private placements and unregistered private placements (the private placement studies). He observes that, if discounts found in unregistered (restricted) private placements are attributable solely to impaired marketability, then there should be no discounts associated with registered private placements (i.e., because such shares can be sold in the public market). However, the results of the private placement studies indicate that even registered private placement shares are issued at a discount, although such discounts tend to be lower than those observed in unregistered private placements. Dr. Bajaj explains that phenomenon by positing that privately placed shares, whether registered or unregistered, tend to be issued to purchasers of large blocks of stock who demand discounts to compensate them for assessment costs and anticipated monitoring costs. He states: “The discount offered to buyers is a compensation for the cost of assessing the quality of the firm and for the anticipated costs of monitoring the future decisions of its managers.”

b. Refinement of Registered /Unregistered, Discount Differential

Dr. Bajaj further contends that the additional discount typical of unregistered private placements (as compared to registered private placements) is not entirely attributable to the fact that unregistered shares, unlike registered shares, generally cannot be sold in the public market. Rather, he contends that such differential is attributable in part to higher assessment and monitoring costs incurred in unregistered private placements as compared to registered private placements. In support of his theory, Dr. Bajaj suggests four factors that might have a correlative relationship to assessment and monitoring costs and, by extension, to private placement discounts: (1) The size of the private placement relative to the issuer’s total shares outstanding, (2) the volatility of the issuer’s recent economic performance, (3) the overall financial health of the issuer, and (4) the size of the private placement in terms of total proceeds. Dr. Bajaj posits that the additional discount observed in unregistered issues could be attributable solely to impaired marketability only if those four additional factors were present in equal measure among both registered and unregistered private placements.

Dr. Bajaj analyzes the effects of the four additional factors listed above and concludes that the first three (but not the fourth) of those factors are systematically related to the level of private placement discounts. Specifically, he concludes that, relatively speaking, a high ratio of privately placed shares to total shares of the issuer, high issuer volatility, and weak financial health of the issuer tend to be indicative of higher discounts. Dr. Bajaj then demonstrates that, as compared to registered private placements, unregistered private placements tend to involve a higher percentage of the issuer’s total shares, higher issuer volatility, and financially weaker issuers. That being the case, Dr. Bajaj concludes that the registered-unregistered private placement discount differential must be attributable in part to those three factors rather than just impaired marketability. In other words, the additional discount typical of unregistered private placements as compared to registered private placements does not represent solely compensation for impaired marketability but represents in part compensation for the relatively higher assessment and anticipated monitoring costs normally associated with unregistered issues.

Having concluded that factors unrelated to impaired marketability play a variable role in the total discounts observed in private placement transactions, Dr. Bajaj then attempts to isolate the effect that impaired marketability has on such total discounts. To that end, he adds a variable for stock registration to variables representing the three additional correlative factors and uses the statistical technique of multivariate regression to determine the effect of each such variable on the discounts observed in his sample of private placements. He concludes from that analysis that, over the 1990 to 1995 period of his study, a private issue that was registered (thereby allowing purchasers to immediately resell in the public market) would have required a discount that was 7.23 percentage points less than an otherwise identical issue (in terms of the three additional correlative factors) that was unregistered.

c. Further Adjustments

Dr. Bajaj considers and rejects any additional adjustment (discount) on account of the long-term impaired marketability of an assignee interest in MIL³⁹ as compared to the limited impaired marketability of restricted shares of stock. His rejection is based primarily on his opinion, supported by the economic analysis of others,⁴⁰ that the level of discount does not continue to increase with the time period of impaired marketability, because investors with long-term horizons would provide a natural clientele for holding illiquid assets and would compete to purchase all or a portion of the gifted interest.

d. Application to MIL

Dr. Bajaj concludes:

Considering the available data, the Partnership’s holdings and history, and the marketability discount of 7.23% suggested by my regression analysis involving a broad range of economic sectors, I conclude that a marketability discount of 7% [rounded from 7.23 percent] is appropriate for all the assets held by MIL when valuing the subject interest. * * *

5. Determination of the Marketability Discount

a. Discussion

Mr. Frazier, in his testimony in rebuttal to Dr. Bajaj, criticizes Dr. Bajaj for focusing narrowly on “liquidity” at the expense of other factors that contribute to a lack of marketability. Mr. Frazier states that “[t]he impediments to value associated with inability to easily sell an interest in a closely-held entity go well beyond the narrowly defined ‘liquidity costs’ Dr. Bajaj has isolated in his analysis” and that “the [marketability] discount is caused not by just ‘liquidity’ but the other negative characteristics that attend securities issued by small closely-held entities.”

Dr. Bajaj has indeed been helpful in focusing our attention (and Mr. Frazier’s attention) on the distinction between illiquidity and other factors (e.g., assessment and monitoring costs) that contribute to private placement discounts. However, his apparent confusion regarding the nature of the discount for lack of marketability (i.e., whether such discount can be explained purely in terms of illiquidity or whether other factors may be involved) is troubling. In his direct testimony, Dr. Bajaj is fairly clear that assessment and monitoring costs associated with private placements are outside the realm of the marketability discount. In his rebuttal testimony, however, he indicates that such costs may contribute to the marketability discount for a closely held entity. That leads us to question whether other “negative characteristics” (in the words of Mr. Frazier) associated with closely held entities may contribute to the appropriate marketability discount for an assignee interest in MIL. Therefore, while we are impressed by portions of Dr. Bajaj’s analysis, he has not convinced us that the appropriate marketability discount in this case can be inferred from the illiquidity cost associated with private placements.

Although we reject Dr. Bajaj’s quantification of the appropriate marketability discount in this case, we look to the data from his private placement study for two reasons. First, we believe that, given mil’s status as an investment company,⁴¹ what Dr. Bajaj refers to in the context of private placements as assessment and monitoring costs would be relatively low in the case of a sale of an interest in MIL. That belief, coupled with Dr. Bajaj’s persuasive argument that such costs are relatively high in unregistered private placements, leads us to conclude that a sample consisting entirely of unregistered private placements would be inappropriately skewed. Second, only Dr. Bajaj’s study (and not the other private placement studies on which he relies) covers the period (1990-95) immediately preceding the valuation date.

In Table 10 of the report constituting his direct testimony, Dr. Bajaj separates the 88 private placements in his sample into three groups according to the level of discounts (i.e., the 29 lowest discounts, the 29 middle discounts, and the 30 highest discounts). Presumably, the “low” category is dominated by registered private placements which, unlike an assignee interest in MIL, did not suffer from impaired marketability. Similarly, it is likely that the “high” category is dominated by unregistered private placements which, unlike the sale of an interest in an investment company, entailed relatively high assessment and monitoring costs. Accordingly, we look to the “middle” group of private placements in Dr. Bajaj’s sample in determining the appropriate marketability discount for an assignee interest in MIL. The average discount of that group of private placements was 20.36 percent.⁴² We are not persuaded that we can refine that figure any more to incorporate characteristics specific to MIL.

b. Conclusion

We find that a discount for lack of marketability of 20 percent (rounded from 20.36 percent) is appropriate in determining the fair market value of each half of the gifted interest.

E. Conclusion

We conclude that the fair market value of each half of the gifted interest is $4,941,916, determined as follows:⁴³

Total NAV . $17,673,760

Less:

Class A preference . (20,000)

“Net” NAV. 17,653,760

1 percent of net NAV . 176,538

Less:

15-percent minority interest discount . (26,481)

Marketable value . 150,057

Less:

20-percent marketability discount . (30,011)

FMV of 1-percent interest . 120,046

FMV of 41.16684918-percent interest . 4,941,916

VI. Charitable Contribution Deduction for Transfer to CFT

A. Introduction

The gift tax is imposed on the value of what the donor transfers, not what the donee receives. Shepherd v. Commissioner, 115 T.C. 376, 385 (2000) (citing, inter alia, Robinette v. Helvering, 318 U.S. 184, 186 (1943)), affd. 283 F.3d 1258 (11th Cir. 2002). In essence, petitioners contend that because (1) they transferred to CFT a portion of the gifted interest corresponding to the excess of the fair market value of that interest over $7,044,933, and (2) we have determined the fair market value of the gifted interest to be $9,883,832, it follows from the maxim beginning this paragraph that they are entitled to a charitable contribution deduction in the amount of $2,838,899 for their gift to CFT. Because the assignment agreement does not equate the term “fair market value” with the term “fair market value as finally determined for Federal gift tax purposes,” petitioners’ argument must fail.

B. The Assignment Agreement

By way of the assignment agreement, petitioners transferred to CFT the right to a portion of the gifted interest. That portion was not expressed as a specific fraction of the gifted interest (e.g., one-twentieth), nor did petitioners transfer to CFT a specific assignee interest in MIL (e.g., a 3-percent assignee interest). Rather, CFT was to receive a fraction of the gifted interest to be determined pursuant to the formula clause contained in the assignment agreement. The formula clause provides that CFT is to receive that portion of the gifted interest having a fair market value equal to the excess of (1) the total fair market value of the gifted interest, over (2) $7,044,933. The formula clause is not self-effectuating, and the assignment agreement leaves to the assignees the task of (1) determining the fair market value of the gifted interest and (2) plugging that value into the formula clause to determine the fraction of the gifted interest passing to CFT.

Petitioners argue that, because the assignment agreement defines fair market value in a manner that closely tracks the definition of fair market value for Federal gift tax purposes, see sec. 25.2512-1, Gift Tax Regs., the assignment agreement effects a transfer to CFT of a portion of the gifted interest determinable only by reference to the fair market value of that interest as finally determined for Federal gift tax purposes. We do not believe that the language of the assignment agreement supports petitioners’ argument. The assignment agreement provides a formula to determine not only CFT’s fraction of the gifted interest but also the symphony’s and the children’s (including their trusts’) fractions.⁴⁴ Each of the assignees had the right to a fraction of the gifted interest based on the value of that interest as determined under Federal gift tax valuation principles. If the assignees did not agree on that value, then such value would be determined (again based on Federal gift tax valuation principles) by an arbitrator pursuant to the binding arbitration procedure set forth in the partnership agreement. There is simply no provision in the assignment agreement that contemplates the allocation of the gifted interest among the assignees based on some fixed value that might not be determined for several years. Rather, the assignment agreement contemplates the allocation of the gifted interest based on the assignees’ best estimation of that value. Moreover, each of the assignees’ percentage interests was determined exactly as contemplated in the assignment agreement (without recourse to arbitration), and none can complain that they got any less or more than petitioners intended them to get.⁴⁵ Had petitioners provided that each donee had an enforceable right to a fraction of the gifted interest determined with reference to the fair market value of the gifted interest as finally determined for Federal gift tax purposes,⁴⁶ we might have reached a different result. However, that is not what the assignment agreement provides.

Of course, the assignees’ determination of the fair market value of the gifted interest, while binding among themselves for purposes of determining their respective assignee interests, has no bearing on our determination of the Federal gift tax value of the assignee interests so allocated. Since we find that the fair market value of a 1-percent assignee interest in MIL on the valuation date was $120,046, the following table expresses the fair market values of the percentage assignee interests passing to the various assignees:

Assignee Percentage assignee interest Fair market value

Children and trusts 77.21280956 $9,269,089

Symphony 1.49712307 179,724

CFT 3.62376573 435,019

9,883,832

C. Conclusion

We find that the fair market value of the property right transferred by petitioners to CFT was $435,019.⁴⁷ Taking into account annual exclusions, see secs. 2503(b) and 2524, each petitioner is entitled to a charitable contribution deduction under section 2522 of $207,510 resulting from the transfer to CFT.⁴⁸

VII. Effect of Children’s Agreement To Pay Estate Tax Liability

A. Introduction

Recently, in Ripley v. Commissioner, 105 T.C. 358, 369 (1995), revd. on another issue 103 F.3d 332 (4th Cir. 1996), we described the nature of a net gift as follows:

Where a “net gift” is made, the donor and donee agree that the donee will bear the burden of the gift tax. The value of the property transferred is reduced by the amount of the gift tax paid by the donee, resulting in the net amount transferred by gift, or the “net gift”. The IRS has provided an algebraic formula for determining the amount of gift tax owed on a “net gift” in Rev. Rul. 75-72, 1975-1 C.B. 310. It is important to keep in mind that once the “net gift” is calculated, the full amount of the gift tax is paid on the “net gift”.

When a “net gift” is made, a portion of the property is transferred by gift and the remaining portion is transferred by sale. * * *

The net gift rationale flows from the basic premise that the gift tax applies to transfers of property only to the extent that the value of the property transferred exceeds the value in money or money’s worth of any consideration received in exchange therefor. See sec. 2512(b); sec. 25.2512-8, Gift Tax Regs.

Petitioners each reported his or her transfer of one-half of the gifted interest as a net gift. Each treated as sales proceeds (consideration received) (1) the amount of Federal and State gift taxes that he or she calculated was to be paid by the children (the gift tax amount) and (2) an amount described on brief as the “mortality-adjusted present value” (mortality-adjusted present value) of the children’s contingent obligation to pay the additional estate tax that would have been incurred on account of section 2035(c) (the 2035 tax) if that petitioner had died within 3 years of the date of the gift. Petitioners describe their computation of the mortality-adjusted present value as follows:

Petitioners * * * estimated the amount of estate tax that would be owed under I.R.C. § 2035(b) based on an expected 55% marginal estate tax rate. Then Petitioners adjusted that amount to present value at the applicable discount rate under I.R.C. § 7520 for January 1996, with further adjustment for the possibility that they would survive each year of the three-year period with no estate tax actually being owed. The probability of death in each of the ensuing three years was calculated, and then the probability-weighted tax amounts were discounted to present value at the required interest rate. All calculations were made, as required under I.R.C. § 7520, by reference to Petitioners’ ages as of their nearest birthdays, the applicable interest rate under I.R.C. § 7520 for January, 1996, and mortality factors provided by Table 80 CNSMT (as found in Respondent’s Pub. 1457, “Actuarial Values, Alpha Volume”).

Petitioners computed the mortality-adjusted present values of the above-described obligations as being $149,813 and $139,348 with respect to Mr. and Mrs. McCord, respectively.

Respondent does not take issue with petitioners’ treatment of the gift tax amounts as sale proceeds. However, he disputes petitioners’ treatment of the mortality-adjusted present values as sale proceeds, primarily on the grounds that those amounts are too speculative to be taken into account. Respondent cites Armstrong Trust v. United States, 132 F. Supp. 2d 421 (W.D. Va. 2001), affd. sub nom. Estate of Armstrong v. United States, 277 F.3d 490 (4th Cir. 2002). In Armstrong Trust, a gift was made and, because of (current) section 2035(b), the donees were subject to a potential liability, as transferees, for estate taxes. See sec. 6324(a)(2). Plaintiffs argued that liens or encumbrances were created on the gift by reason of the potential estate tax liability assumed by the donees, thereby reducing the value of the gift. Id. at 430. The District Court found that the possibility of future estate tax liability was too speculative to reduce the value of the gift.

Relying on an opinion of the U.S. Court of Claims, Murray v. United States, 231 Ct. Cl. 481, 687 F.2d 386 (1982), the Court of Appeals for the Fourth Circuit affirmed the District Court’s decision in Armstrong Trust, also on the basis that the donees’ potential liability for the donor’s estate tax was too speculative to reduce the value of the gifts. Estate of Armstrong v. United States, 277 F.3d at 498. It was of no moment to the Court of Appeals that the donor had in fact died within 3 years of the gift, thus causing a 2035 tax to be due, or that the plaintiffs apparently had produced expert calculations of an amount similar to the mortality-adjusted present value at issue in this case. The Court of Appeals said:

The litigation attempts of the Estate and the Trust to quantify through expert calculations the value of potential estate taxes at the time of the transfers is irrelevant. What is relevant is that the children’s obligation to pay any estate taxes was then “highly conjectural,” Murray, 687 F.2d at 394, and so provides no ground for applying net gift principles. [Id,.]

In Murray v. United States, supra, the donor had made gifts in trust pursuant to an instrument that obligated the trustees to pay, among other debts, the donor’s estate and death taxes liabilities. The plaintiffs (executors of the donor’s estate) argued that the obligation to pay the donor’s estate and death taxes rendered the gifts without value when made. The Court of Claims disagreed, finding that the obligation to pay estate and death taxes “was not * * * susceptible to valuation at the date of the gifts because the economic burden of paying these taxes was then unknown.” Id. at 394. The court questioned whether it was even possible to approximate the value of the trustee’s obligation to pay the donor’s estate and death tax liabilities:

the amount of estate and death taxes payable from the * * * [trusts] was highly conjectural. If Oliver lived until May 1971, the value of the 1968 Family Trust would no longer have been included in his estate as a gift in contemplation of death under section 2035, significantly reducing his estate tax liability. Moreover, had he lived for several more years, the size of his estate would have continued to diminish, leaving the 1970 Family Trusts with an ever-decreasing estate tax obligation. * * * [Id. at 394-395.]

The Court of Claims concluded: “Thus, plaintiffs’ inability to reasonably estimate the amount of tax, if any, to be paid from the * * * [trusts] made it proper to compute the gift tax on the basis of the full value of the trust assets. Robinette v. Helvering, 318 U.S. 184, 188-89 (1943).” Id. at 395.⁴⁹

B. Discussion

The specific question before us is whether to treat as part of the sale proceeds (consideration) received by each petitioner on the transfer of the gifted interest any amount on account of the children’s obligation pursuant to the assignment agreement to pay the 2035 tax that would be occasioned by the death of that petitioner within 3 years of the valuation date. We have not faced that specific question before.⁵⁰ Neither Armstrong Trust v. United States, supra, nor Murray v. United States, supra, is binding on us, and, indeed, the facts of both cases are somewhat different from the facts before us today. Armstrong Trust did not involve a specific assumption of the potential 2035 tax as a condition of the underlying gift (as is the case here); rather, the donees were statutorily liable for any 2035 tax under section 6324(a)(2). In Murray, unlike the instant case, the obligation was not limited to the “gross-up” tax of (original) section 2035(c), with its preordained inclusion amount and accompanying 3-year window of inclusion (indeed, that provision and the unified gift and estate tax system had yet to be enacted). Nevertheless, we agree with what we believe to be the basis of those two opinions, i.e., that, in advance of the death of a person, no recognized method exists for approximating the burden of the estate tax with a sufficient degree of certitude to be effective for Federal gift tax purposes. See also Estate of Armstrong v. Commissioner, 119 T.C. 220, 230 (2002).

Petitioners’ computation of the mortality-adjusted present value of the children’s obligation to pay the 2035 tax does nothing more than demonstrate that, if one assumes a fixed dollar amount to be paid, contingent on a person of an assumed age not surviving a 3-year period, one can use mortality tables and interest assumptions to calculate the amount that (without any loading charge) an insurance company might demand to bear the risk that the assumed amount has to be paid. However, the dollar amount of a potential liability to pay the 2035 tax is by no means fixed; rather, such amount depends on factors that are subject to change, including estate tax rates and exemption amounts (not to mention the continued existence of the estate tax itself⁵¹). For that reason alone, we conclude that petitioners are not entitled to treat the mortality-adjusted present values as sale proceeds (consideration received) for purposes of determining the amounts of their respective gifts at issue.⁵² See Robinette v. Helvering, 318 U.S. 184, 188-189 (1943) (donor’s reversionary interest, contingent not only on donor outliving 30-year old daughter, but also on the failure of any issue of the daughter to attain the age of 21 years, is disregarded as an offset in determining the value of the gift; actuarial science cannot establish the probability of whether the daughter would marry and have children).

Our conclusion is further buttressed by broader considerations of Federal gift tax law. Under the “estate depletion” theory of the gift tax, it is the benefit to the donor in money or money’s worth, rather than the detriment to the donee, that determines the existence and amount of any consideration offset (sale proceeds) in the context of an otherwise gratuitous transfer. See Commissioner v. Wemyss, 324 U.S. 303, 307-308 (1945); 2 Paul, Federal Estate and Gift Taxation 1114-1115 (1942). When a donee agrees to pay the gift tax liability resulting from a gift, the benefit to the donor in money or money’s worth is readily apparent and ascertainable, since the donor is relieved of an immediate and definite liability to pay such tax. If that donee further agrees to pay the potential 2035 tax that may result from the gift, then any benefit in money or money’s worth from the arrangement arguably would accrue to the benefit of the donor’s estate (and the beneficiaries thereof) rather than the donor. The donor in that situation might receive peace of mind, but that is not the type of tangible benefit required to invoke net gift principles.

C. Conclusion

Because petitioners have failed to show that their computation of the value of the children’s obligation to pay any 2035 tax is reliable, we do not accept it as establishing any proceeds received by petitioners and reducing the value of the net gifts made by them.

VIII. Conclusion

The fair market value of the gifted interest on the date of the gift was $9,883,832 ($4,941,916 for each half thereof). Petitioners are entitled to an aggregate charitable contribution deduction under section 2522 for the transfer to CFT in the amount of $415,019 ($207,510 apiece).

To reflect the foregoing,

Decision will be entered under Rule 155.

Reviewed by the Court.

Wells, Cohen, Swift, Gerber, Colvin, Gale, and Thornton, JJ., agree with this majority opinion.

1 Under the terms of the partnership agreement, a class A limited partnership interest does not carry with it a “Percentage Interest” (as that term is defined in the partnership agreement) in MIL.

2 The children’s counsel had also represented petitioners in connection with the transaction. However, petitioners were not involved in the allocation of the gifted interest among the assignees pursuant to the confirmation agreement.

3 Those figures also reflect cash gifts of $10 by each petitioner to nominally fund the trusts.

4 Consistent with that argument, petitioners have preserved their right to claim an increased charitable contribution deduction under sec. 170 on an amended income tax return for 1996. Petitioners’ 1996 income tax liability is not before us.

5 The parties have not informed us of their basis for stipulating that respondent bears the burden of proof. The burden of proof is normally on petitioner. See Rule 142(a)(1). Under certain circumstances, the burden of proof can be shifted to the Commissioner. See sec. 7491; Rule 142(a)(2). We assume (without deciding) that the conditions necessary to shift the burden to respondent have been satisfied.

6 For purposes of this report, we use the term “assignee interest” (assignee interest) to signify the interest held by an assignee of a partnership interest who has not been admitted as a partner in the partnership.

7 The partnership agreement provides that the term “partnership interest” means the interest in the partnership representing any partner’s right to receive distributions from the partnership and to receive allocations of partnership profit and loss. The statutory definition is similarly worded. See Tex. Rev. Civ. Stat. Ann. art. 6132a-l, sec. 1.02(11) (Vernon 2001) (the Act) and accompanying bar committee comment; see also id. sec. 7.02(a)(1), (3), and (4) (assignment of partnership interest entitles the assignee to distributions and allocations, but the assignor continues to be a partner and to have the power to exercise any rights or powers of a partner, until the assignee becomes a partner).

8 To use the terminology favored by the parties, we shall value the gifted interest as an as-signee interest.

9 Relying on Morrissey v. Commissioner, 243 F.3d 1145, 1148 (9th Cir. 2001), revg. Estate of Kaufman v. Commissioner, T.C. Memo. 1999-119, and on Estate of Smith v. Commissioner, T.C. Memo. 1999-368, petitioners contend that the confirmation agreement is conclusive proof of the value of the gifted interest because such agreement was an arm’s-length transaction that was the “functional equivalent of an actual sale.” We disagree. Suffice it to say that, in the long run, it is against the economic interest of a charitable organization to look a gift horse in the mouth.

10 Although the cited cases involved the estate tax, it is well settled that the estate tax and the gift tax, being in pari materia, should be construed together. See, e.g., Shepherd v. Commissioner, 283 F.3d 1258, 1262 n.7 (11th Cir. 2002) (citing Harris v. Commissioner, 340 U.S. 106, 107 (1950)), affg. 115 T.C. 376 (2000).

11 Mr. Frazier asserts that a 41.167-percent assignee interest in MIL (i.e., one-half of the gifted interest) has no more proportionate value than a 1-percent assignee interest therein, and respondent’s expert does not dispute that assertion.

12 We note that the class A limited partner’s sole economic interest in MIL consists of a guaranteed payment for the use of such partner’s (nominal) capital. This case does not require us to determine whether the class A limited partner is a partner of MIL for Federal tax purposes.

13 For purposes of determining MIL’s NAV, Mr. Frazier does not apply discounts to the real estate limited partnership interests that McCord Brothers Partnership contributed to MIL upon formation. Mr. Frazier did, however, discount the value of those real estate limited partnership interests by 57.75 percent for purposes of valuing McCord Brothers Partnership’s capital contribution to MIL and determining the MIL partners’ percentage interests in MIL. This case does not require us to address the gift tax consequences, if any, of the initial capital contributions to, and allocation of percentage interests in, MIL.

14 Unlike a shareholder of an open-end fund, a shareholder of a closed end fund cannot, at will, by tendering his shares to the fund for repurchase, obtain the liquidation value of his investment (i.e., his pro rata share of the fund’s NAV). For that reason, a share of a closed end fund typically trades at a discount relative to its pro rata share of the fund’s NAV. Since, according to the expert witnesses, that discount has no marketability element, it is, to some extent, considered reflective of a minority interest discount.

15 As noted earlier, a shareholder of an open-end fund generally can obtain the liquidation value of his investment (i.e., his pro rata share of the fund’s NAV) by tendering his shares to the fund for repurchase. It stands to reason that, to the extent the conversion of a closed end fund to open-end status is imminent, the share price of such fund will tend to approach the fund’s NAV per share.

16 In his direct testimony, Dr. Bajaj states that the two excluded funds “could not be identified in Morningstar Principia dataset as of December 31, 1996”. Since Mr. Frazier excludes those funds from his sample as well, we similarly exclude them from consideration.

17 A dual purpose fund has both income shares and capital shares. At a set expiration date, the fund redeems all income shares, and the capital shareholders then vote to either liquidate the fund or convert it to open-end status.

18 Dr. Bajaj may have excluded Liberty All Star Growth Fund from his sample due to the lack of NAV information with respect to such fund as of the valuation date. For purposes of our analysis, we utilize the fund’s NAV and price data as of Jan. 5, 1996, which is in the record.

19 For instance, less regulation implies lower compliance costs, which seemingly would offset, at least to some extent, any pricing effect of relatively lax investor protections.

20 Although MIL inherited any unrealized gain with respect to assets contributed by the initial MIL partners, see sec. Y23, the portion of such precontribution gain otherwise allocable to a subsequent purchaser of an interest in MIL, see sec. 1.704-3(a)(7), Income Tax Regs., generally would be eliminated if the general partners of MIL were to agree to make a timely sec. 754 election with respect to MIL, see secs. 754, 743(b); sec. 1.755-l(b)(l)(ii), Income Tax Regs. The same would be true with respect to any postcontribution unrealized appreciation with respect to MIL’S assets.

21 In their reports, Mr. Frazier and Dr. Bajaj determine the average, but not the weighted average, of the discounts with respect to the equity funds in their respective samples. We follow the same approach here.

22 In his direct testimony, Dr. Bajaj states that the six excluded funds “could not be identified in Morningstar Principia dataset as of December 31,1996”.

23 Mr. Frazier’s regression analysis produced R-squared calculations of 0.29 for the Morningstar rating (management quality) variable and 0.01 for the fund size variable. Elsewhere in his direct testimony, Mr. Frazier indicates that an R-squared calculation of 0.34 is “relatively low”, leading to the conclusion of “no clear correlation” between the variable in question and the level of sample fund discounts.

24 In their reports, Mr. Frazier and Dr. Bajaj determine the average, but not the weighted average, of the discounts with respect to the bond funds in their respective samples. We follow the same approach here.

25 Dr. Bajaj limits his real estate analysis to MIL’s real estate partnership interests. We address the minority interest discount factor for MIL’s direct real estate holdings infra in sec. V.C.2.d.

26 A real estate investment trust is a tax-favored vehicle through which numerous investors can pool their resources to invest in real estate. See secs. 856-859.

27 Cf. Estate of Desmond v. Commissioner, T.C. Memo. 1999-76 (approving the use of the market approach for valuing an operating business based on two guideline companies in the same —as opposed to similar — line of business).

28 We note that, while Mr. Frazier questions the composition of Dr. Bajaj’s sample of REITs, he offers no specific suggestions for modifying the sample.

29 As Dr. Bajaj explains his calculation: “If an illiquid security trades at a discount of 7% relative to a liquid asset, this suggests that the liquid asset is trading at a premium of about 7.53% relative to the illiquid asset (1/[1 - 7%]).”

30 Wrack, “Equity Ownership Concentration and Firm Value: Evidence from Private Equity Financings,” 23 J. Fin. Econ. 3 (1989).

31 Mr. Frazier and Dr. Bajaj agree on the percentages of NAV assigned to each asset class.

32 See 17 C.F.R. sec. 230.144(d)(1) (1996). The required holding period was shortened to 1 year in 1997. See 62 Fed. Reg. 9242 (Feb. 28, 1997).

33 Mr. Frazier relied on Mr. Emory’s studies in the 1996 HFBE appraisal report but makes no mention of those studies in either his direct testimony or his rebuttal testimony.

34 Mr. Frazier reviews the following studies (the restricted stock studies):

1. Securities and Exchange Commission, Discounts Involved in Purchases of Common Stock (1966-1969), H.R. Doc. No. 64, Part 5, at 2444-2456 (1971).

2. Silber, “Discounts on Restricted Stock: The Impact of Illiquidity on Stock Prices,” Fin. Analysts J. 60 (July-August 1991).

3. A study only described as “The Standard Research Consultants (SRC) Study”.

4. Hertzel & Smith, “Market Discounts and Shareholder Gains for Placing Equity Privately,” 48 J. Fin. 459 (1993).

35 Indeed, under the heading “Revenues” in Table 31 of Mr. Frazier’s direct report, the entry corresponding to “Willamette” is “NA”.

36 Specifically, one restricted stock study, the Silber study, found that the average dollar size of private placements with discounts in excess of 35 percent was $2.7 million, while the average dollar size of private placements with discounts less than 35 percent was $5.8 million. Even taking into account Mr. Frazier’s suggested minority interest discount of 22 percent, the “dollar size” of each half of the gifted interest was approximately $5.7 million. That would indicate that, based on the Silber study, a discount of less than 35 percent would be appropriate for each half of the gifted interest.

37 We note that such other factors should not include the purchaser’s minority ownership position, if applicable; presumably, a minority interest discount is already reflected in the market price of a share of the issuer.

38 Other than his own study, he refers to the Wrack study, supra note 30, and the Hertzel & Smith study, supra note 34.

39 Both experts operate under the assumption that there will not be a ready market for as-signee interests in family limited partnerships during the remainder of MIL’S 30-year term.

40 Amihud & Mendelson, “Asset Pricing and Bid-Ask Spread,” 17 J. Fin. Econ. 223 (1986).

41 On the valuation date, 65 percent of MIL’s assets consisted of marketable securities and an additional 30 percent consisted of real estate limited partnership interests, subject to well-known and relatively routine appraisal techniques (such as cashflow analysis or market multiple methods).

42 That discount is consistent with the average discount (20.14 percent) observed in the Hertzel & Smith private placement study, supra note 34, the study (other than his own) primarily relied upon by Dr. Bajaj.

43 For ease of computation, we determine the fair market value of a 1-percent interest.

44 If f equals the fair market value of the gifted interest (determined by the assignees (or an arbitrator) based on Federal gift tax valuation principles), and the gifted interest is shown as the 82.33369836-percent class B assignee interest in MIL transferred by petitioners, then, assuming f is equal to or greater than $7,044,933, the products of the following formulas show the percentage assignee interests apportioned to the children (including the trusts), the symphony, and CFT, expressed as x¡, X2, and xs, respectively:

$6,910,933

f x 82.33369836% = Xi

$7,044,933 - 6,910,933

f x 82.33369836% = x2

f- $7,044,933

f X 82.33369836% = *3

45 We suppose that, at least in theory, there might be a difference between (1) petitioners’ and the assignees’ expectation on Jan. 12, 1996 (the valuation date), regarding the value of the portion of the gifted interest passing to CFT and (2) the value of that portion as subsequently determined by the assignees. However, no one has suggested how to value the first quantity or that, on the facts before us, the difference would be significant.

46 See, e.g., sec. 1.664 — 2(a.)(l)(iii), Income Tax Regs, (providing that a sum certain may be expressed as a fraction or percentage of the value of property “as finally determined for Federal tax purposes”, but requiring that actual adjusting payments be made if such finally determined fair market value differs from the initially determined value); sec. 20.2055-2(e)(2)(vi)(a), Estate Tax Regs, (similar); sec. 25.2702-3(b)(l)(ii)(B), Gift Tax Regs, (similar); Rev. Proc. 64^-19, 1964-1 C.B. 682 (discussing conditions under which the Federal estate tax marital deduction may be allowed where, under the terms of a will or trust, an executor or trustee is empowered to satisfy a pecuniary bequest or transfer in trust to a decedent’s surviving spouse with assets at their value as finally determined for Federal estate tax purposes).

47 The rule is well established that we may approve a deficiency on the basis of reasons other than those relied on by the Commissioner. See Wilkes-Barre Carriage Co. v. Commissioner, 39 T.C. 839, 845 (1963) (and cases cited therein), affd. 332 F.2d 421 (2d Cir. 1964). Because our conclusion that the valuation clause of the assignment agreement does not achieve the claimed “tax neutralization” effect is based on the language of the assignment agreement, we need not address respondent’s arguments that (1) the formula clause is against public policy, and (2) the transaction should be recast as transfers of cash by petitioners to CFT and the symphony under an integrated transaction theory. We note that the application of respondent’s integrated transaction theory would result in an initial increase in the amount of petitioners’ aggregate taxable .gift of only $90,011 (less than 1 percent), which would be partially offset by the resulting increase in the gift tax liability that the noncharitable donees assumed under the assignment agreement.

48 We note that, under our analysis, the assignee interest received by the symphony is worth more than $134,000. Nevertheless, we do not believe that petitioners have claimed any increased charitable contribution deduction under sec. 2522 on account of the transfer to the symphony. If we are mistaken on that point, petitioners can bring that to our attention (or perhaps petitioners and respondent can deal with it in the Rule 155 computation).

49 In Robinette v. Helvering, 318 U.S. 184, 188-189 (1943), the Supreme Court held that, in computing the value of a gift of a remainder, interest in property, the value (as an offset) of the donor’s contingent reversionary remainder interest was to be disregarded because there was no recognized method of determining its value.

50 Nevertheless, in Estate of Armstrong v. Commissioner, 119 T.C. 220, 230 (2002) (addressing certain Federal estate tax questions with respect to the same gifts in question in Armstrong Trust v. United States, 132 F. Supp. 2d 421 (W.D. Va. 2001), affd. sub nom. Estate of Armstrong v. United States, 277 F.3d 490 (4th Cir. 2001)), we said: “The donee children’s mere conditional promise to pay certain additional gift taxes that decedent might be determined to owe does not reduce the amount of decedent’s gift taxes included in the gross estate under section 2035(c).”

51 See Economic Growth and Tax Relief Reconciliation Act of 2001, Pub. L. 107-16, secs. 501(a), 901(a), 115 Stat. 38, 69, 150 (repealing the estate tax with respect to decedents dying after Dec. 31, 2009, and reinstating same with respect to decedents dying after Dec. 31, 2010).

52 We recognize that, in Harrison v. Commissioner, 17 T.C. 1350, 1354-1355 (1952), we reduced the amount of a gift of a trust remainder by the present value of the trustee’s obligation, under the terms of the trust agreement, to pay the settlor-life beneficiary’s income tax liability attributable to the trust’s income for the remainder of her life: “Federal income taxes have become a permanent and growing part of our economy, and there is no likelihood that such taxes will not continue to be imposed throughout the life expectancy of petitioner.” We do not have occasion today to reconsider that opinion.