Jeffrey I. Okerlund ("Jeffrey") and Lorrie Schwan-Okerlund ("Lorrie"); David J. Schwan ("David") and Diane C. Schwan ("Diane"); Mark D. Schwan ("Mark"); and Paul M. Schwan ("Paul") and Christine H.M. Weigel-Schwan ("Christine") (collectively, "Plaintiffs") appeal the August 23, 2002 ruling of the United States Court of Federal Claims that the Plaintiffs' tax liabilities should be assessed based on a valuation of $24.36 per share of minority stock in Schwan's Sales Enterprises, Inc. ("SSE") as of December 31, 1992. Okerlund v. United States, 53 Fed. Cl. 341 (2002). Because the Court of Federal Claims conducted its valuation analysis correctly, we affirm.

Marvin Schwan ("Marvin") founded SSE in Marshall, Minnesota, in 1948 as a milk processing operation. SSE subsequently expanded, first to become a cafe and an ice cream manufacturer, and eventually into a profitable provider of both home and retail delivery of selected frozen foods. By the early 1990s, SSE had become a sizable producer, wholesaler, and retailer of food products employing about 16,500 people. SSE's operations included both international and domestic sales, totaling approximately $1.95 billion in 1992 and $2.3 billion in 1994. As of 1992, SSE's top management team consisted of President Marvin Schwan, a Tax Accountant, a Director of Marketing, a General Counsel, a Controller, a National Sales Manager, and a Manufacturing Manager.

SSE is a privately held company capitalized with two classes of stock, voting (0.02% of SSE's equity capital) and nonvoting (99.8% of the equity capital). The stock is not listed on any stock exchange and is not publicly traded. Under the company's by-laws, SSE has the right-of-first refusal on all sales or transfers of common stocks held by certain stockholders, based on a fair market value transfer price.

In December 1992, SSE nonvoting stock split 100 to 1, increasing the number of nonvoting shares from 385,550 to 38,555,000. In addition, Marvin, SSE's President, purchased 3,755 newly issued voting shares. He contributed 2,485 of these shares to the preexisting Marvin M. Schwan Revocable Trust and the remaining 1,270 shares to a newly created Great Great Grandchildren's Trust ("3G Trust"). As of the December 31, 1992, valuation date, a total of 7,610 shares of voting shares (including the recently issued 2,485) and 38,555,000 nonvoting shares were issued and outstanding.

Marvin died unexpectedly on May 9, 1993, at the age of sixty-four, leaving SSE without its founder and President. At the time of Marvin's death, 5,076 voting shares and 25,910,100 non-voting shares were held in the Marvin M. Schwan Revocable Trust. In accordance with Marvin Schwan's estate plan, a charitable foundation known as the King's Foundation received the shares owned by this revocable trust, which comprised two-thirds of outstanding SSE stock. SSE redeemed the King's Foundation shares pursuant to a February 4, 1993, Amended Redemption Agreement.

The Plaintiffs are Marvin's four children and three of their spouses. On December 31, 1992, the Plaintiffs established separate trusts for the primary benefit of their respective children, using SSE shares that Marvin had previously distributed to his children. Plaintiffs Lorrie, David, and Paul each gifted 50,000 shares of nonvoting stock to their respective trusts, splitting their gifts with their spouses pursuant to 26 U.S.C. § 2513. Plaintiff Mark gifted 25,000 shares of SSE.

In June 1993, the Plaintiffs employed a firm called Business Valuation Consultants (also known as "Gray") to value SSE's shares. Gray valued SSE at $24.03 per share as of the December 31, 1992 gift date. Based on this valuation, each plaintiff filed a gift tax return reporting a gift of $600,750, a unified credit of $192,800, a Generation-Skipping Tax (GST) exemption of $600,750, and a tax of $277.

In 1996, in the midst of litigation unrelated to the present tax matter, the Plaintiffs retained Willamette Management Associates ("Willamette") to appraise the company's value.1 Though that litigation's settlement was based on a 1997 value of $26.00 per share, the Willamette appraisal valued SSE at $17.40 per share as of December 31, 1992.

In July 1996, as a result of this lower valuation, the Plaintiffs filed for a Claim for Refund and Request for Abatement with the IRS, seeking restoration of their respective unified credits in the amount of $59,100, a restoration of their respective GST exemptions in the amount of $165,760 each, and a gift tax refund of $277.

In 1994, several of the Plaintiffs gifted and donated additional SSE shares to various recipients. These transactions led to disputed calculations of the Plaintiffs' 1994 gift taxes and income taxes based upon the disputed value of SSE's shares as of December 31, 1994. The Court of Federal Claims valued the shares at $19.77 as of December 31, 1994, as urged by the Plaintiffs. The Plaintiffs' income and gift taxes for 1994 were then recalculated based on this value, and the government has not appealed the Court of Federal Claims' ruling. For the purposes of this appeal, then, only the December 31, 1992 valuation remains relevant.

Proceedings Before the Court of Federal Claims

The challenge facing both the Court of Federal Claims and this court is the determination of the fair market value of SSE's nonvoting stock as of December 31, 1992. Both the Plaintiffs and the government employed the assistance of valuation experts. The government retained Dr. Herbert T. Spiro, while the Plaintiffs retained Dr. Shannon P. Pratt. Dr. Pratt was the founder of Willamette, the firm that had assessed SSE's market value during the unrelated litigation in 1996.

The Court of Federal Claims evaluated and compared the two experts' valuations carefully and thoughtfully — particularly with respect to the more-contentious December 31, 1992 valuation — and accepted some assessments proposed by each of the experts before reaching its final determination. See Helvering v. Nat'l Grocery Co., 304 U.S. 282, 295, 58 S. Ct. 932, 82 L. Ed. 1346 (1938); Lukens v. Commissioner, 945 F.2d 92, 96 (5th Cir.1991). The Court of Federal Claims approached this challenge within the statutory and regulatory framework established by the Treasury Regulations, 26 C.F.R. § 25.2512-1. Under this framework, property without a well-defined market value is assigned the value that it would obtain in an arms-length market transaction between a buyer and a seller, each of whom is trying to maximize profits. See Estate of Klauss v. Comm'r, T.C. Memo 2000-191, 79 T.C.M. 2177, 2000 WL 823377 (2000). Furthermore, a gift of property is valued as of the date of transfer, 26 U.S.C. § 2512(a), and the value of the gift is determined by the value of the property passing from the donor, not the property value received by the donee. Treas. Reg. § 25.2511-2(a).

Both Dr. Spiro (for the government) and Dr. Pratt (for the Plaintiffs) applied blended methodologies to the valuation task, using both a "Market Approach" and an "Income Approach" to valuation, deriving a weighted average of the two indicated values, and then applying various discount factors.2 The Court of Federal Claims accepted Dr. Spiro's application of the two valuation methodologies and Dr. Pratt's assessment of the appropriate discount factors to arrive at its own conclusion, valuing SSE at $24.36 per share as of December 31, 1992. See Lukens, 945 F.2d at 96.

The Market Approach to valuation rests upon the assumption that similar investments should obtain similar values. The key to this approach is therefore the identification of comparable publicly traded "guideline" companies and the application of those companies' financial fundamentals to those of SSE. Dr. Spiro and Dr. Pratt agreed on the propriety of this method and upon the categories of relevant financial factors. Their primary disagreements lay in the weight assigned to different factors and in the relationship between SSE's specifics (e.g., leadership and leverage) to those of the guideline companies. After calculating market pricing multiples based upon guideline company quoted trading prices and financial data, both experts adjusted for risk factors that affected the company as a whole; Dr. Pratt's assessment of these risks was far more pessimistic than was Dr. Spiro's.

The Income Approach stems from the well-accepted proposition that the value of a corporation is given by a risk-adjusted discounted cash flow analysis of its future profit streams. The application of this approach, as both experts explained, requires the determination of a suitable discount rate, the projection of future cash flows and profits, and the identification of relevant risk factors. Once again, the Court of Federal Claims found Dr. Spiro's application of this methodology more reliable than Dr. Pratt's.

Taken together, Dr. Spiro assessed per share values of $41.07 using the income approach and $51.82 using the market approach, prior to discounting for lack of marketability and for lack of voting rights. He then explained that the income approach was the more reliable of the two, and proposed a 70/30 weighted average to arrive at a valuation of $44.29 per share for minority nonvoting SSE stock, prior to discounting.

Dr. Pratt did not specify share prices under the different methods. He did, however, calculate enterprise values of $1.28 billion under the income approach and $1.385 billion under the market approach. He also applied an admittedly less reliable "transaction approach," proposed a weighted average, and derived an enterprise value of approximately $1.34 billion or $34.80 per share, also prior to discounting for lack of marketability and for lack of voting rights.

As noted, the Court of Federal Claims considered both the numbers suggested by the experts and the justifications offered for those numbers, and found Dr. Spiro's application of this methodology to be more reliable than Dr. Pratt's. The Court of Federal Claims thus adopted a valuation of $44.29 before considering the necessary discount factors.

The $44.29 value had to be discounted for lack of marketability. Both experts relied on two sources of empirical data for aid in quantifying the discount factor: (1) discounts on sales of restricted shares of publicly traded companies; and (2) discounts on private transactions prior to initial public offerings (IPOs). Based on these studies and on an examination of the perceived risks facing a potential investor in SSE stock, Dr. Pratt concluded that a 45% discount for lack of marketability was appropriate. Dr. Spiro concluded that a 30% discount was justified.

The Court of Federal Claims found Dr. Pratt's analysis of these factors more persuasive than Dr. Spiro's, but nevertheless adjusted it upward to arrive at a 40% percent discount for lack of marketability as of December 31, 1992. Both experts agreed that a further 5% discount for the lack of voting rights was appropriate. The Court of Federal Claims accepted this assessment, applied a combined discount factor of 45% to its undiscounted valuation of $44.29, and arrived at a valuation of $24.36 per non-voting share of SSE, as of December 31, 1992. The Plaintiffs appeal this determination.

The Court of Federal Claims ordered the parties to reach agreement on the Plaintiffs' tax liabilities, and the refund due to Lorrie and Jeffrey,3 based upon valuations of $24.36 per share as of December 31, 1992, and $19.77 per share as of December 31, 1994. On November 25, 2002, the Court of Federal Claims ordered each of the Plaintiffs to pay $3,053 in 1992 gift tax, plus $3,571.97 in interest due (as of the date of the order); and $165 in 1994 gift tax, plus $142.01 in interest due (as of the date of the order). The Court of Federal Claims also ordered the government to refund to Lorrie and Jeffrey $77,672 against their 1994 income tax, plus $55,112.26 in interest due (as of the date of the order). The Plaintiffs timely appealed. We have jurisdiction to hear this appeal under 28 U.S.C. § 1295(a)(3).

We review decisions of the Court of Federal Claims de novo for errors of law and for clear error on findings of fact. Shelden v. United States, 7 F.3d 1022, 1026 (Fed.Cir.1993). The legal framework that the Court of Federal Claims applies within which to value an asset is a legal question that we review de novo. Krapf v. United States, 977 F.2d 1454, 1458 (Fed.Cir.1992). The Court of Federal Claims' actual determination of an asset's fair market value is a factual question that we review for clear error. Id.

This distinction, and the standard of review that we apply to valuation, are not unique to this court. See, e.g., Lukens, 945 F.2d at 96 ("The tax court's determination of the fair market value of the timeshare units is a finding of fact reviewable under the clearly erroneous standard."); Morris v. Comm'r, 761 F.2d 1195, 1200 (6th Cir.1985) ("determination of the fair market value of certain property... is a factual one and should not be reversed on appeal unless clearly erroneous.").

The Plaintiffs do not appeal the Court of Federal Claims' fundamental legal determinations concerning the appropriate standards for valuation, such as the use of an arms-length transaction as a benchmark, valuation as of date of transfer, or valuation as the property passed from the donor, rather than to the donee. The Court of Federal Claims' framework for valuation rested upon sound legal principles. Valuation of property for which there is no market is determined by considering the company's net worth, earnings potential, capacity to pay dividends, and other relevant factors. Treas. Reg. § 25.2512-2(f). Revenue Ruling 59-60, 1959 WL 12594, which is "widely accepted as setting forth the appropriate criteria to consider in determining fair market value," Estate of Newhouse v. Comm'r, 94 T.C. 193, 217, 1990 WL 17251 (1990), identifies eight such relevant factors: (1) the nature of the business and the history of the enterprise from its inception; (2) the economic outlook in general and the condition and outlook of the specific industry in particular; (3) the book value of the stock and the financial condition of the business; (4) the earning capacity of the company; (5) the dividend-paying capacity; (6) whether or not the enterprise has goodwill or other intangible value; (7) sales of the stock and the size of the block of stock to be valued; and (8) the market price of freely traded stocks of corporations engaged in the same or a similar line of business. In addition, when valuing minority interest stock of a closely-held corporation, the application of certain discounts may be warranted to reflect the stock's lack of marketability and/or lack of voting rights. Cent. Trust Co. v. United States, 158 Ct. Cl. 504, 524, 305 F.2d 393 (1962). Both experts presented valuations consistent with these principles. The Court of Federal Claims' valuation methodology, which drew upon those of the experts, was therefore correct as a matter of law.

The Plaintiffs contend that the Court of Federal Claims' failure to consider SSE's actual earnings results in 1993 and 1994 in assessing the 1992 valuation is a legal error subject to de novo review, while the Court of Federal Claims' underassessment of the significance of one particular risk factor — the potential triggering of the Redemption Agreement — is a clearly erroneous factual determination.

All of the Plaintiffs' arguments on appeal therefore stem from a single source: the confluence of several risk factors that actually occurred during the years 1993 and 1994. The 1992 Willamette Report described several risk factors relevant to understanding the ways that SSE differed from otherwise comparable companies. These risk factors included: (1) reliance on a home delivery route system; (2) thin management ranks; (3) reliance on a key management figure, Marvin Schwan; (4) the risk of food contamination; (5) the competitors' greater human resources; (6) SSE's inability to invest in a national advertising campaign, based on its lack of a nationally recognizable brand name and the demographics of its customer base; (7) less diversity in product offerings than the guideline companies; and (8) the relatively small size of SSE's Board of Directors.

By the end of 1994, at least two of these risks had materialized: Marvin Schwan died on May 9, 1993 and an outbreak of salmonella in October 1994 led to a product recall, a plant closure, and a class action lawsuit. In other words, several events that the appraisers had identified as foreseeable though unlikely risks in 1992 had actually occurred by 1994. These unexpected occurrences led to declines in SSE's sales and income during both 1993 and 1994, and account at least in part for the drop in the company's value between the 1992 and 1994 valuation dates.

The Plaintiffs contend that the actual occurrence of these events suggests that Dr. Spiro underestimated their ex ante probability in his 1992 valuation, and that Dr. Pratt's assessments were therefore more reasonable. As evidence of the effect of these underassessed probabilities, the Plaintiffs point to the disparities between Dr. Spiro's projected revenues and gross profits for 1993 and 1994 and SSE's actual revenues and gross profits for those years. The Plaintiffs find it highly significant that Dr. Spiro overestimated revenues by $112,467,000 (1993) and $252,054,000 (1994), and gross profits by $51,937,000 (1993) and $122,801,000 (1994). The Plaintiffs also challenge Dr. Spiro's assertion that considering actual 1993 and 1994 revenues when determining a 1992 valuation would have been "inappropriate appraisal practice."

The Court of Federal Claims addressed this challenge in footnote 22 of its opinion:

Plaintiff's reliance on SSE's actual financial results for 1993 and 1994 to cast doubt on defendant's appraisal must be similarly rejected. Although the Federal Circuit in Krapf, 977 F.2d at 1459, held that post-valuation date transactions may provide a "parameter or give rise to an inference of the earlier value," and should not be absolutely barred from consideration, consideration of post-valuation financial results here would be improper. As the government points out, "if such a technique were accepted, there would be little need to resort to the courts in valuation cases. The mere passage of time, together with the accumulation of post-valuation date data, would convert the valuation process form an inexact science to an exact science." Estate of Gillet v. Commissioner, T.C. Memo 1985-394, 50 T.C.M. 636, 645, 1985 WL 15056 (1985).

Okerlund, 53 Fed.Cl. at 348 n. 22. The plaintiffs now contend that this response constitutes reversible legal error. According to the Plaintiffs, first, the Court of Federal Claims was required to consider whether SSE's actual performance during 1993 and 1994 demonstrates that Dr. Spiro's underlying projections were unreasonable, and second, had the Court of Federal Claims performed the required analysis, it would have rejected Dr. Spiro's projections as unreasonable.

Neither of the Plaintiffs' arguments is tenable. As to the Court of Federal Claims' legal requirements, the best that the Plaintiffs can do is point to a number of previous tax cases suggesting that it is permissible for courts to use subsequent events as "sanity checks" on projections and valuations. Even these pointers, however, do not really support the Plaintiffs' position. For example, their citation to Estate of Gilford v. Commissioner, 88 T.C. 38, 52, 1987 WL 49260 (1987), as allowing ex post information, put into context, reads:

In general, property is valued as of the valuation date on the basis of market conditions and facts available on that date without regard to hindsight. However, we have held that postmortem events can be considered by the Court for the "limited purpose" of establishing what the willing buyer and seller's expectations were on the valuation date and whether these expectations were "reasonable and intelligent."

Id.

The Plaintiffs did not point to a single case suggesting that the failure to consider subsequent events constitutes legal error. Furthermore, and contrary to the Plaintiffs' contention, the relationship between Dr. Spiro's projections and SSE's actual performance supports the reasonableness of those projections. Dr. Spiro's projections of both revenues and gross profits were off by about 5% for 1993 and by less than 10% for 1994. Given the number of exogenous low-probability high-risk events that intervened, Dr. Spiro's projections appear to have been highly reliable and based upon reasonable assumptions.

The closest that the Plaintiffs come to precedent supporting their assertion that the Court of Federal Claims was required to consider ex post information lies in our rejection of the Claims Court's proposed bright-line test excluding ex post information. Krapf, 977 F.2d at 1459. In Krapf,

[t]he Claims Court [had] ruled that the valuation of closely held stock must generally be made without reference to events which occur after the date of the donation. As "exceptions" to this general "rule," the Claims Court further stated that post-donation data may be used in valuation only where: (1) no material change of circumstances or conditions in the corporation has occurred between the time of donation and the time of the post-donation evidence, or (2) the post-donation changes could have been foreseen at the time of the donation.

adopt the Claims Court view that there is an exclusionary rule with two exceptions respecting admissibility of evidence of post-donation data and events. The question respecting such evidence may involve its relevancy, i.e., its admissibility, but more usually the question is the evidence's probative value. The post-transaction evidence must always be proffered, of course, in support of finding the value of the stock on the donative date. Evidence of post-transaction events may or may not be relevant depending on what events are sought to be proved. Where there was a series of arms length sales at the same price before and after the donation, such evidence might well give rise to an inference that the gift had the same value and, thus, would be admissible. Such a situation would likely fit into the second "exception" noted by the Claims Court, but we conclude that the "rule" of exclusion is simply the usual rule of relevancy. Further, even if relevant, the evidence may have little probative value. On the other hand, evidence of a post-gift sale price may give rise to an inference of the stock's earlier value, in light of other circumstantial evidence.

Id. at 1459. The valuation of a closely held company is an inexact science (some might say an art), and relevant probative evidence should never be ignored. It would be absurd to rule an arms-length stock sale made moments after a gift of that same stock inadmissible as post-valuation date data — as we noted in Krapf, id. The key to the use of any data in a valuation remains that all evidence must be proffered in support of finding the value of the stock on the donative date. Id.

The present matter could hardly differ more from the circumstances that we anticipated in Krapf. Here, a number of exogenous shocks befell SSE shortly after the December 31, 1992 valuation date. Dr. Pratt's 1992 Willamette report correctly identified both the reliance on a key executive and the possibility of contamination as risk factors relevant to valuation. All of the proffered valuations incorporate them accordingly: the selection of guideline companies implies that SSE's risk of contamination was about the same as that of its competitors, and the discount attributed to the small management team incorporated the risk to SSE of losing part of that team.

Had SSE gone public or engaged in a number of arms-length transactions between the valuation date and the occurrence of these exogenous events, the prices of those transactions might have been deemed relevant and probative — at least as sanity checks on the assumptions underlying the valuation models. But the occurrence of intervening events, or for that matter of even less probable unforeseen events, reduces the probative value of subsequent transactions. Even an arms-length sale of SSE made after Marvin's death would have represented a valuation of a company that differed in a significant manner from the one valued on December 31, 1992.

Our holding in Krapf was correct. Valuation must always be made as of the donative date relying primarily on ex ante information; ex post data should be used sparingly. As with all evidentiary submissions, however, the critical question is relevance. The closer the profile of the later-date company to that of the valuation-date company, the more likely ex post data are to be relevant (though even in some cases, they may not be). The greater the significance of exogenous or unforeseen events occurring between the valuation date and the date of the proffered evidence, the less likely ex post evidence is to be relevant — even as a sanity check on the assumptions underlying a valuation model. Krapf provides no support for the Plaintiffs' assertion that the Court of Federal Claims erred in excluding SSE's 1993 and 1994 performance from its 1992 valuation.

The Plaintiffs' argument about the Redemption Agreement, a pre-existing agreement built into Marvin Schwan's estate plan that was triggered by his untimely death in 1993, is similarly flawed. According to the Plaintiffs, Dr. Pratt correctly identified this looming, though as yet untriggered, Redemption Agreement as a cloud that placed SSE's entire continued existence at risk, and that should thus have reduced SSE's value greatly even in 1992. Dr. Spiro, on the other hand, viewed this untriggered estate planning device as a much less significant risk. The Plaintiffs assert that the Court of Federal Claims committed reversible error by agreeing with Dr. Spiro that the Redemption Agreement was relevant only as a discount to marketability.

To assess this argument, it is important to understand what the Redemption Agreement required. Marvin Schwan had set up a charitable foundation, the King Foundation, and given it roughly two-thirds of the outstanding SSE stock. Under the Redemption Agreement, SSE was required to redeem this stock upon Marvin's death. As things turned out, SSE was required to raise $869 million to repurchase this stock after Marvin died.4 This repurchase left SSE highly leveraged.

The critical question was how the existence of this Redemption Agreement should have been viewed on December 31, 1992. According to the Court of Federal Claims:

in 1992 the estate plan provisions, although in place, had neither been triggered nor anticipated in the immediate future. In other words, they were prospective concerns rather than actual concerns as of the 1992 valuation date. It is well-established that "valuation of the stock must be made as of the relevant dates without regard to events occurring subsequent to the crucial dates." Bader v. United States, 172 F. Supp. 833, 840 (S.D.Ill.1959). In 1992, the major shareholder risks identified in the Willamette Report, and in Dr. Pratt's testimony, were in place, but had not yet been triggered by Marvin Schwan's death. The difference between potential versus actual deterrents to investment supports a 5 percent disparity between the appropriate discount for lack of marketability in 1992 (40 percent) and in 1994 (45 percent).

The Court of Federal Claims' analysis and conclusion are eminently reasonable, supported by substantial evidence, and certainly do not constitute clear error.

The Plaintiffs point to nothing else in the Court of Federal Claims' analysis that they consider to be erroneous.

Because the Court of Federal Claims applied the correct legal standards to valuation and made no factual errors, both its derivation of share values and the tax calculations based upon those values were correct.

AFFIRMED.

No costs.