MEMORANDUM FINDINGS OF FACT AND OPINION

THORNTON, Judge: Respondent determined a $ 998,508 deficiency in petitioner's 1996 Federal gift tax. The issue for decision is the fair market value of interests in a family limited partnership that petitioner transferred in 1996.

Unless otherwise indicated, all section references are to the Internal Revenue Code as amended, and all Rule references are to the Tax Court Rules of Practice and Procedure.

             FINDINGS OF FACT

The parties have stipulated many of the facts, which we incorporate by this reference. When petitioner filed her petition, she resided in Fruitport, Michigan.

On October 20, 1995, petitioner and her daughter, Clarajane Middlecamp (Clarajane), formed, pursuant to Georgia law, the Lappo Family Limited Partnership (the partnership). On April 19, 1996, petitioner and Clarajane conveyed to it a portfolio of marketable2003 Tax Ct. Memo LEXIS 257">*258 securities (principally municipal bonds) and certain parcels of Michigan real estate that were subject to a long-term lease. ¹

After these initial capital contributions, petitioner's and Clarajane's respective partnership interests were as follows:

The allocation of initial partnership interests was based on the December 31, 1995, market value2003 Tax Ct. Memo LEXIS 257">*259 of the assets contributed to the partnership. The appraised market value of the real estate was $ 1,860,000. The market value of the securities was $ 1,318,609.

1. The April 19, 1996, Gifts

On April 19, 1996, petitioner transferred a 69.4815368- percent limited partnership interest, representing the following gifts: a 66.80917-percent limited partnership interest to Clarajane as Trustee of the Lappo Generation Trust; and a 0.6680917-percent interest to each of her four grandchildren, Seth R. Middlecamp, Lisa Middlecamp-Silky, Wendy Thomas, and Alyson Middlecamp.

2. The July 2, 1996, Gift

On July 2, 1996, petitioner gave her remaining 29.2184632- percent limited partnership interest to Clarajane in her individual capacity.

Consequently, after these gifts, the partnership interests were as follows: 2003 Tax Ct. Memo LEXIS 257">*260

Under the partnership agreement, the general partners have exclusive authority to manage the operations and affairs of the partnership and to make all decisions regarding its business, including the distribution of cash. No partner can: (a) Withdraw any part of her capital or receive any distributions from the partnership except as provided for in the partnership agreement; (b) demand or receive any assets other than cash in return for her capital interest; or (c) be paid interest on any capital contributed to or accumulated in the partnership. Withdrawal of a general partner will cause dissolution of the partnership unless there is at least one other general partner to carry on the partnership's2003 Tax Ct. Memo LEXIS 257">*261 business or unless all remaining partners agree to continue the partnership and to appoint one or more general partners.

In general, the limited partners have no rights to participate in managing or controlling the partnership's business. A limited partner may assign his or her interest, but such assignment will not dissolve the partnership or entitle the assignee to become a partner or to exercise any rights as a partner (unless the general partners give their prior consent); rather, the assignee will be entitled only to receive distributions to which the assignor would have been entitled.

Under the partnership agreement, the partnership may purchase all, but not less than all, of a limited partner's interest upon the limited partner's death or upon any transfer of the limited partner's interest by operation of law. The purchase price will be the fair market value as agreed upon by the limited partner and the partnership or else as determined by appraisal.

If a limited partner undertakes to sell his or her interest, the partnership has the right of first refusal. If the partnership elects to buy the limited partner's entire interest, the price to be paid will be the price set forth2003 Tax Ct. Memo LEXIS 257">*262 in the selling partner's original proposal, less 15 percent.

The partnership is to continue until December 31, 2045, unless it is dissolved sooner by consent of all the partners, by the withdrawal of a general partner (in the absence of another general partner to carry on the partnership's business), or by entry of a decree of judicial dissolution.

On April 11, 1997, petitioner filed a Federal gift tax return, reporting her April 19, 1996, gifts of limited partnership interests, which she valued at $ 1,040,000. With this return she remitted $ 153,000 of reported gift tax liability. On February 6, 1998, petitioner filed an amended Federal gift tax return, reporting her July 2, 1996, gift to Clarajane of a 29.218462-percent limited partnership interest which had been omitted from her original 1996 gift tax return. Petitioner valued the July 2, 1996, gift at $ 423,871 and remitted an additional $ 177,265 of gift taxes.

In the notice of deficiency, issued June 19, 2001, respondent determined that petitioner's 1996 gifts of partnership interests should be increased from the originally reported $ 1,040,000 to $ 3,137,287, resulting2003 Tax Ct. Memo LEXIS 257">*263 in a $ 998,508 gift tax deficiency as determined by reference to petitioner's originally filed gift tax return. ² Respondent credited the additional $ 177,265 that petitioner had paid with her amended gift tax return as an advance payment of the gift tax deficiency so determined.

                OPINION

The only issue remaining in dispute is the fair market values of the limited partnership interests that petitioner transferred during 1996. The parties2003 Tax Ct. Memo LEXIS 257">*264 generally agree that these fair market values should be determined by reference to the net asset values (NAVs) of the partnership's assets (i.e., its real estate holdings and marketable securities portfolio), reduced by minority interest and marketability discounts. The parties agree on the NAVs of the partnership's assets. ³ They disagree on the size of the applicable minority interest and marketability discounts. Petitioner bears the burden of proof. See Rule 142(a). ⁴

In support of their positions, each party relies on an expert opinion. We evaluate expert opinions in light of all the evidence in the record and may accept or reject expert testimony, in whole or in part, according to our own judgment. Helvering v. Natl. Grocery Co., 304 U.S. 282">304 U.S. 282, 304 U.S. 282">295, 82 L. Ed. 1346">82 L. Ed. 1346, 58 S. Ct. 932">58 S. Ct. 932 (1938); Shepherd v. Commissioner, 115 T.C. 376">115 T.C. 376 (2000), affd. 283 F.3d 1258">283 F.3d 1258 (11th Cir. 2002); Estate of Newhouse v. Commissioner, 94 T.C. 193">94 T.C. 193, 94 T.C. 193">217 (1990). "The persuasiveness of an expert's opinion depends largely upon the disclosed facts on which it is based." Estate of Davis v. Commissioner, 110 T.C. 530">110 T.C. 530, 110 T.C. 530">538 (1998). We may be selective in our use of any part of an expert's opinion. Id.

1. Petitioner's Expert

Petitioner's expert, Robert P. Oliver (Mr. Oliver), is an accredited appraiser who has been with Management Planning, Inc. (MPI), since 1975 and has served as its president since 1996. MPI has been in the business of preparing financial analyses of closely held businesses and in evaluating securities of such businesses since 1939. Mr. Oliver is an author and speaker on valuation and related topics.

In2003 Tax Ct. Memo LEXIS 257">*266 his direct testimony, Mr. Oliver concluded that a 7.5- percent minority interest discount is appropriate with respect to the marketable securities component of the partnership interests. With respect to the real estate component, he concluded that a 35-percent minority interest discount should apply to petitioner's April 19, 1996, gifts, and a 30-percent minority interest discount to her July 2, 1996, gift. In addition, Mr. Oliver concluded that there should be a 35-percent marketability discount for each gift.

2. Respondent's Expert

Respondent's expert, Dr. Alan C. Shapiro (Dr. Shapiro), is the Ivadelle and Theodore Johnson Professor of Banking and Finance and past chairman of the Department of Finance and Business Economics, Marshall School of Business, University of Southern California (USC). He is also an outside director of LECG, LLC, an economic and financial consulting company. Prior to joining USC in 1978, he taught at the Wharton School of Business at the University of Pennsylvania and has been a visiting professor at Yale, the University of California at Los Angeles, the University of British Columbia, and the Stockholm School of Economics. He is the author of numerous2003 Tax Ct. Memo LEXIS 257">*267 academic articles and books on corporate finance.

In his direct testimony, Dr. Shapiro concluded that the partnership interests should be valued to reflect an 8.5-percent minority interest discount and an 8.3-percent marketability discount.

In estimating the fair market value of a noncontrolling interest in a closely held business entity, it may be appropriate to decrease NAV to reflect lack of control inherent in the interest. See, e.g., Estate of Andrews v. Commissioner, 79 T.C. 938">79 T.C. 938, 79 T.C. 938">953 (1982).

1. Marketable Securities

As previously indicated, with respect to the marketable securities component of the partnership interests, petitioner's expert recommends a 7.5-percent minority interest discount, whereas respondent's expert recommends an 8.5-percent minority interest discount. The parties agree that the difference is not significant. At trial, petitioner agreed to use respondent's slightly higher net asset values for the marketable securities. Out of fairness to petitioner, we also use Dr. Shapiro's slightly higher 8.5-percent minority interest discount rate.

2. Real Estate

a. Selection of Guideline Companies

Both experts agree that2003 Tax Ct. Memo LEXIS 257">*268 publicly traded real estate investment trusts (REITs) provide an appropriate starting point for determining the minority interest discount with regard to the partnership's real estate holdings. They disagree, however, on which REITs should be used for comparison and on the analysis of the REITs data.

i. Petitioner's Expert

To select his guideline companies, Mr. Oliver started with over 400 REITs and real estate companies listed in Moody's Bank and Finance Manual. From this initial group of over 400, he eliminated all but seven companies (three REITS and four real estate companies) as being insufficiently comparable to the partnership. ⁵ For instance, he eliminated numerous companies that did not report current appraisals of their real estate assets. He also eliminated other companies that he considered to have investment characteristics dissimilar to the partnership.

2003 Tax Ct. Memo LEXIS 257">*269 We are not persuaded that Mr. Oliver's guideline group is sufficiently large or made up of companies sufficiently comparable to the partnership. "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'." McCord v. Commissioner, 120 T.C. 358">120 T.C. 358, 120 T.C. 358">384 (2003) (quoting Estate of Heck v. Commissioner, T.C. Memo. 2002-34). For the relevant period, the four real estate companies included in Mr. Oliver's guideline group appear dissimilar to the partnership in fundamental ways that might be expected to skew the price-to-NAV discounts of the guideline companies upward. ⁶

2003 Tax Ct. Memo LEXIS 257">*270 ii. Respondent's Expert

Dr. Shapiro started with the "core coverage group" of 62 real estate companies as reported by Green Street Advisors, Inc. (Green Street). ⁷ He eliminated 10 companies that were not REITs or that had what he believed were "substantially different investment characteristics" from the partnership. The record does not reveal the identities of the 52 REITs included in Dr. Shapiro's guideline group. Petitioner does not dispute, however, that they are all REITs -- a consideration of some significance, given that the parties agree that REITs are an appropriate starting point for determining the minority interest discount. Moreover, we believe the size of Dr. Shapiro's sample was sufficiently large to make tolerable any dissimilarities between the partnership and the REITs in his guideline group. See McCord v. Commissioner, supra at 385.

2003 Tax Ct. Memo LEXIS 257">*271 Petitioner complains that Dr. Shapiro's guideline group is inappropriate because Green Street derived NAVs using a valuation method different from that used to value the partnership's real estate. The valuation method used by Green Street, however, appears similar to that used to value the partnership's real estate. ⁸ In any event, even if the valuation methods are not identical, insofar as each method is reliable and unbiased (and petitioner does not contend that either is not), each might be expected to produce reasonable valuations so as to provide a meaningful basis for comparing share prices to net asset values.

2003 Tax Ct. Memo LEXIS 257">*272 b. Price-to-NAV Discounts

i. Petitioner's Expert

To determine the NAVs of his seven guideline companies, Mr. Oliver reviewed their reported book values and then made what his expert report tersely describes as "certain adjustments" to adjust these book values upward to reflect "appraised values disclosed by management." Comparing these NAVs to quoted share prices, Mr. Oliver concluded that the median price-to-NAV discount for the guideline companies was 29.3 percent on April 19, 1996, and 20.3 percent on July 2, 1996.

Mr. Oliver then considered a number of factors that he said differentiated the partnership from his seven guideline companies. On the one hand, he concluded, the partnership had a "much stronger" financial position than the guideline companies, which would indicate a relatively smaller price-to-NAV discount for the partnership interests. On the other hand, he concluded, certain factors augured for a deeper price-to-NAV discount: The partnership had "very small" real estate holdings compared to the guideline companies; it was dependent on just one primary tenant; and as a newly formed entity, it lacked a track record of operations. The net effect of such factors, Mr. 2003 Tax Ct. Memo LEXIS 257">*273 Oliver concluded, was that a minority interest investor would value the partnership's real estate component at a deeper discount than the guideline median price-to-NAV discount. In the final analysis, he concluded that the appropriate minority interest discount was 35 percent for petitioner's April 19, 1996, gifts of partnership interests and 30 percent for petitioner's July 2, 1996, gift.

Mr. Oliver has inadequately explained how he derived the NAVs that are critical to his computation of the price-to-NAV discounts. The record does not adequately reflect the management disclosures that led to Mr. Oliver's upward adjustments of the companies' reported book values, with a directly corresponding upward effect on his price-to-NAV discount computations. ⁹ Moreover, Mr. Oliver has failed adequately to explain the apparent volatility in his recommended price-to-NAV discounts over less than 3 months (decreasing from 29.3 percent on April 19, 1996, to 20.3 percent on July 2, 1996). ¹⁰ It seems most likely that the volatility results from the small size of his sample and the inclusion of entities that are insufficiently comparable to the partnership. ¹¹

2003 Tax Ct. Memo LEXIS 257">*274 Moreover, we are unconvinced of the appropriateness of the upward adjustments Mr. Oliver made to this volatile guideline company data to account for factors specific to the partnership. He does not explain how he quantified each factor, how the factors were netted, or why the net effect should result in an upward adjustment to the median guideline company discounts, rather than a downward adjustment or a wash. ¹² Nor does Mr. Oliver explain why this fixed set of factors should result in a 5.7-percent upward adjustment for petitioner's April 1996 gifts but a 9.7-percent (70-percent larger) upward adjustment for another gift less than 3 months later. It seems most likely that Mr. Oliver's upward adjustments are, to some extent, plug numbers used to justify his ultimate, very round minority interest discount figures of 35 percent and 30 percent for April 19, 1996, and July 2, 1996, respectively.

2003 Tax Ct. Memo LEXIS 257">*275 Mr. Oliver opined that the reasonableness of his recommended minority interest discounts was confirmed by reference to the average 36-percent price-to-NAV discount that he calculated for a select group of 14 publicly registered, nonpublicly traded real estate limited partnerships (RELPs). ¹³ The record provides, however, an insufficient basis for us to make an informed judgment as to the partnership's comparability to these 14 RELPs or as to the reliability of the methods used to determine NAVs and market prices for these 14 RELPs. ¹⁴ In addition, petitioner concedes that the trading volume of RELPs, which do not trade on organized stock exchanges, is very low. Consequently, Mr. Oliver's reliance on the published RELP market prices seems questionable.

2003 Tax Ct. Memo LEXIS 257">*276 ii. Respondent's Expert

Dr. Shapiro compared Green Street-reported market prices and NAVs to conclude that for the relevant period, his 52 guideline REITs traded at a 4.8-percent median price-to-NAV premium. To be conservative and to account for the difference that the partnership, unlike REITs, is not obligated to pay large and regular distributions to its interest holders, Dr. Shapiro looked below the median, to the 15th percentile, and began with an .8-percent discount as of March 25, 1996, and a 1.48-percent premium as of June 25, 1996.

Dr. Shapiro concluded that these 15th-percentile REITs data should be adjusted downward in petitioner's favor to remove a liquidity premium that is inherent in REITs; i.e., a premium that arises because REIT interests, unlike the assets underlying them, are publicly traded in reasonably liquid markets. To gauge the size of this inherent liquidity premium, Dr. Shapiro referred to an academic study of private placement discounts for a period ending just before the valuation dates for the subject partnership interests. Bajaj et al., "Firm Value and Marketability Discounts", 27 J. Corp. L. 89 (2001) (hereinafter the Bajaj study). On2003 Tax Ct. Memo LEXIS 257">*277 the basis of the Bajaj study, Dr. Shapiro concluded that, for the relevant time period, liquid assets such as REITs were trading at a premium of about 7.5 percent over illiquid assets such as the partnership interests. Subtracting this 7.5-percent liquidity premium from the previously indicated 15th-percentile REITs data, he concluded that the real estate component of the partnership interests should be valued to reflect minority interest discounts of 8.3 percent (-. 8 minus 7.5) and 6 percent (1.48 minus 7.5), as of April 19, 1996, and July 2, 1996, respectively.

Dr. Shapiro then compared these results to his own study which suggested that minority interests in holding companies trade at a discount of 8.5 percent relative to controlling interests in holding companies. Adjudging holding companies to be comparable to the partnership in certain respects, he concluded that to be conservative he would apply this more favorable 8.5-percent minority interest discount to the partnership interests.

Dr. Shapiro's study on holding companies is not in evidence. The minimal description of it in his testimony provides an inadequate basis for us to rely upon it in determining the appropriate minority2003 Tax Ct. Memo LEXIS 257">*278 interest discount here.

We agree with Dr. Shapiro that, in order to derive a minority interest discount factor from REIT price-to-NAV data, one must account for the liquidity premium inherent in REIT data prices. See McCord v. Commissioner, 120 T.C. 358">120 T.C. 385. In quantifying that liquidity premium, however, we hesitate to rely on a single academic study -- particularly one that Dr. Shapiro did not participate in preparing and could not elaborate upon first hand. We therefore seek common ground between the Bajaj study and similar studies (the Wruck study and the Hertzel & Smith study) ¹⁵ cited therein.

According to the Bajaj study, the Wruck study found that the average2003 Tax Ct. Memo LEXIS 257">*279 discount observed in unregistered private placements exceeded the average discount observed in registered private placements by 17.6 percentage points. ¹⁶ The differential reported in the Hertzel & Smith study is 13.5 percentage points. ¹⁷ Those figures are consistent with the differential reported in the Bajaj study, 14.09 points. ¹⁸ The average of these three figures is approximately 15 percent, which yields a liquidity premium of 17.6 percent (1/[1 -.15]). ¹⁹

Using a liquidity premium of 17.6 percent, we arrive at minority interest discounts of 18.4 percent (. 8-percent price-to-NAV discount less 17.6-percent liquidity premium) for the April 16, 1996, gifts and 16.12 percent (1.48-percent price-to-NAV premium less 17.6- percent liquidity premium) for the July 2, 1996, gift. Following Dr. Shapiro's lead, we round these figures up slightly to a uniform 19- percent2003 Tax Ct. Memo LEXIS 257">*280 minority interest discount rate, which we shall apply to the real estate component of the partnership interests.

3. Conclusion

As explained above, we find that an 8.5-percent minority interest discount is appropriate in valuing the marketable securities component of the partnership interests and a 19-percent minority interest discount is appropriate in valuing the real estate component of the partnership interests. These minority interest discount factors yield weighted averages of 14.70 percent and 14.49 percent, as of April 19, 1996, and July 2, 1996, respectively. ²⁰ Rounding these weighted averages, we conclude that an overall minority interest discount of 15 percent is appropriate in determining the fair market value of each gifted partnership interest.

The experts agree that private placements of publicly traded stock are an appropriate starting place for determining a marketability discount here. The experts disagree on the appropriate private placements to be considered and what is measured by those comparisons. The experts also disagree on the inferences to be drawn from the partnership's specific characteristics.

1. Empirical Analysis

a. Petitioner's Expert

Mr. Oliver compared private-market prices of unregistered (restricted) shares in public corporations with the public-market prices of unrestricted but otherwise identical shares in the same corporations. ²¹ He attributes the price difference to the restricted shares' lack of marketability.

2003 Tax Ct. Memo LEXIS 257">*282 More particularly, starting with a preexisting MPI study analyzing 197 private transactions in common stocks of actively traded corporations from 1980 through 1995, Mr. Oliver identified a guideline group of 39 transactions in unregistered (restricted) stock. He determined that these 39 transactions occurred at prices reflecting discounts of 5.2 percent to 57.6 percent from the public- market price, with the median discount being 29.3 percent.

Thirteen of the 39 companies in Mr. Oliver's guideline group appear to be high-technology companies and also to have some of the highest discounts, ostensibly reflecting greater risk. ²² We are unpersuaded that these 13 companies are comparable to the partnership. If these 13 companies are removed from Mr. Oliver's guideline group, the median discount of the remaining 26 companies is 19.45 percent.

2003 Tax Ct. Memo LEXIS 257">*283 b. Respondent's Expert

Dr. Shapiro again relied primarily on the Bajaj study, which analyzed discounts observed in private placements of registered shares as well as private placements of unregistered (restricted) shares. The Bajaj study concluded that the portion of private placement discounts attributable solely to impaired marketability was 7.2 percent.

We note initially that, in this context, we prefer Dr. Shapiro's "private placement" approach (as embodied in the Bajaj study) over Mr. Oliver's more narrow "restricted stock" approach. See McCord v. Commissioner, supra at 394. Absent further explication of the Bajaj study by Dr. Shapiro, however, and without the benefit of other empirical studies that would tend to validate the conclusions of the Bajaj study, we are unpersuaded that a 7.2-percent discount is an appropriate quantitative starting point for determining the marketability discount applicable to the gifted interests in this case.

Looking instead to the raw data from the Bajaj study, we see that the average discount with respect to its sample of private placements is 22.21 percent. ²³ The Hertzel & Smith study, cited in the Bajaj study, reported an average2003 Tax Ct. Memo LEXIS 257">*284 discount of 20.14 percent with respect to its sample of private placements. ²⁴ Averaging those two figures, we conclude that a 21-percent marketability discount is appropriate before adjustments to incorporate characteristics specific to the partnership. We note that this discount rate is very close to the 19.45-percent median discount rate that we determined using Mr. Oliver's methodology and data, modified as discussed above.

2. Further Adjustments

a. Petitioner's Expert

Mr. Oliver ultimately concluded that the marketability discount applicable to the partnership interests should be 35 percent, or 5.7-percent higher than the 29.3-percent median discount that he determined by reference to his private placement sample. He made this upward adjustment to his recommended marketability discount on the basis2003 Tax Ct. Memo LEXIS 257">*285 of the following considerations: The partnership is closely held with no real prospect of becoming publicly held; the partnership is relatively small and little known; there is no present market for the partnership interests; the partnership agreement requires the partnership to be offered the right of first refusal to purchase, at a 15-percent discount, limited partnership interests; and the partnership agreement permits a transferee of a limited partnership interest to become a substituted limited partner only with the general partners' consent.

b. Respondent's Expert

Dr. Shapiro ultimately concluded that the marketability discount applicable to the partnership interests should be 8.3 percent, or 1.1-percent higher than the 7.2-percent discount that he said was indicated by the Bajaj study. In arriving at this upward adjustment, he considered a number of factors. He acknowledged that because the partnership is not scheduled for dissolution until 2045, the partnership interests are less marketable than restricted securities, thereby justifying some additional amount of discount. As a countervailing consideration, however, he opined that the appraised value of the partnership's real2003 Tax Ct. Memo LEXIS 257">*286 estate already incorporates a lack- of-marketability discount. He also acknowledged provisions of the partnership agreement that generally prevent the assignee of a limited partner's interest from becoming a partner and that require a limited partner who wished to sell his or her entire interest to offer the partnership a right of first refusal at a 15-percent discount. He concluded, however, that these restrictions had little effect on marketability. He observed, for instance, that a limited partner could easily circumvent the 15-percent discount associated with the right of first refusal by selling less than her entire interest.

c. Conclusion

On the basis of all the evidence and using our best judgment, we conclude that a 3-percent upward adjustment in the marketability discount rate (as determined by reference to the previously described empirical studies) is appropriate to incorporate characteristics specific to the partnership. Consequently, we find that a discount for lack of marketability of 24 percent is appropriate in determining the fair market value of each limited partnership interest that petitioner transferred in 1996.

We conclude that for April 19, 1996, the2003 Tax Ct. Memo LEXIS 257">*287 fair market value of the 69.4815368-percent gifted partnership interests is $ 1,417,006, computed as follows:

We conclude that for July 2, 1996, the fair market value of the 29.2184632-percent gifted partnership interest is $ 611,455, computed as follows:

To reflect the foregoing,

Decision will be entered under Rule 155.