Trial; Mutual insurance company — participating policyholders; Demutualization; Gain on sale of stock received in exchange for ownership rights; Basis allocation rule — Treas. Reg. § 1.61-6; "Open transaction" doctrine — Burnet v. Logan; Limited exception to Treas. Reg. § 1.61-6 where impossible or impractical to value property; Pierce, Inaja Land and Warren construed; Mutual ownership rights found not to be susceptible of valuation; Expert reports on valuation; "Open transaction" exception to Treas. Reg. § 1.61-6 applicable; Refund entitled.

OPINION

FRANCIS ALLEGRA, Judge

Since its infancy, the Federal income tax law has provided that gross income includes gains derived from dealings in property and that such gains generally equal the amount realized less the seller's cost basis in the property sold. Though clear in principle, these rules are not always easily applied — particularly, where the property sold was first acquired, for a lump sum, as part of a larger assemblage, and, especially, where the values of the individual components of that grouping are not readily ascertainable. For generations, courts faced with the scenario just-described have grappled with two possibilities: to treat the property sold as having little or no cost basis, so that most or all the sale proceeds are taxable, or to treat the property as sharing the cost basis of the entire bundle, such that no gain is realized until all the capital represented by that basis is recovered. These are among the possible outcomes in this tax refund suit, which involves insurance policy rights that were acquired as an indivisible package, but then separated and sold as part of a demutualization of the insurance provider.

I. FINDINGS OF FACT

Trial in this case was conducted in Phoenix, Arizona. Based on the record at trial, including the parties' joint stipulations, the court finds as follows:

Prior to 2000, Sun Life Assurance Company (Sun Life) was a Canadian mutual life insurance and financial services company that conducted business in Canada, the United States and other countries. A mutual insurance company has no shareholders, but instead is owned by its participating policyholders, which possess both ownership rights, such as voting and distribution rights, as well as the more typical contractual insurance rights. Their voting rights differ from those possessed by traditional shareholders in that each policyholder has but a single vote, regardless of how many policies it owns or the amounts thereof. Once the mutual company pays its claims and operating expenses, the profits belong to the policyholders. Typically, some of those profits are returned to the policyholders as dividends, which reduce premium payments, while the remainder is retained as surplus, often accumulating from year to year. Payment of such policy dividends is largely at the discretion of the board elected by the participating policyholders. The ultimate goal of this arrangement is to provide insurance at the lowest possible cost.

On June 28, 1990, the Seymour P. Nagan Irrevocable Trust (the Trust) purchased a life insurance policy from Sun Life on Seymour Nagan and Gloria Hagan. The policy was for $500,000, with annual premiums at $19,763.76 per year. Under this "participating policy," plaintiff's ownership rights included the ability —

to vote on matters submitted to participating policy holders . . . to participate in the distribution of profits of Sun Life of Canada from all its businesses, to participate in any distribution of demutualization benefits, and in the unlikely event of a liquidation if Sun Life of Canada were ever to become insolvent, to participate in the distribution of any remaining surplus after satisfaction of all obligations.

Plaintiff's right to receive distribution of profits took the form of an annual dividend representing the amount, if any, of profits not retained in surplus. These ownership rights could not be sold separate from the policy and were terminated when the policy ended.

On January 27, 1998, the Sun Life Board (the Board) requested the insurer's management to develop a plan to convert the company into a publicly-traded stock company through a so-called "demutualization." On September 28, 1999, the Board voted to recommend that the policyholders approve the demutualization. It perceived that the conversion would permit the reorganized company to provide stock options to its employees, offer more diversified products and obtain, more readily, capital financing for its businesses, including those unrelated to providing insurance.

On October 29, 1999, Sun Life proposed a plan to its policyholders to demutualize. Under the plan, the policyholders would retain their insurance coverage at premiums that would be unaffected by the demutualization, but would receive shares of stock in a new holding company, Sun Life of Canada Holding Corp. (Financial Services), which would become the corporate parent of Sun Life. Those shares were to be exchanged for the ownership rights possessed by the participating policyholders, with approximately 20 percent of the shares being allocated to compensate for the loss of voting control and the remaining 80 percent of the shares being allocated to compensate for the loss of other ownership rights, including the right to receive a liquidating distribution. Under the plan, eligible policyholders — those that had policies in force as of January 27, 1998 — did not have to take stock in exchange for their shares. Those in the United States, for example, could elect to sell the shares issued in connection with a planned initial public offering, an option referred to as the "cash election." If the policyholder took this election, it would receive an amount determined "by multiplying the number of Financial Services Shares sold . . . by the Initial Share Price at which the number of Financial Services Shares are sold in connection with the initial public offering." Policyholders were informed as to how many shares they would be issued in a "share allocation statement."

On December 15, 1999, the Board certified that the demutualization plan had been approved by the eligible policyholders. In early March of 2000, Sun Life began its initial public offerings and received various regulatory approvals to proceed with the demutualization. On May 19, 2000, in response to a request from the company, the Internal Revenue Service (IRS) issued Private Letter Ruling 200020048, which dealt with various tax aspects of the demutualization. In that ruling, the IRS noted that the aforementioned ownership rights "cannot be obtained by any purchase separate from an insurance contract issued by [Sun Life]." It ruled that, under section 354(a)(1) of the Internal Revenue Code of 1986 (26 U.S.C.) (the Code), "[n]o gain or loss will be recognized by the Eligible Policyholders on the deemed exchange of their Ownership Rights solely for Company stock." It further opined that the "basis of the Company stock deemed received by the Eligible Policyholders in the exchange will be the same as the basis of the Ownership rights surrendered in exchange for such Company Stock," that is, "zero." The IRS did not rule on the tax treatment to be afforded the cash received in lieu of shares exchanged for ownership rights.

When the demutualization took effect, plaintiff received 3,892 shares of Financial Services stock in exchange for its voting and liquidation rights. Opting for the "cash election," plaintiff permitted Sun Life to sell those shares on the open market for $31,759.00. It reported this amount, unreduced by any basis adjustment, on its federal income tax return for 2000 and paid the resulting tax of $5,725.00. On February 11, 2004, plaintiff filed a timely claim seeking a refund of its money, and, upon the denial of that claim, filed the instant suit. On March 14, 2005, plaintiff filed a motion for partial summary judgment; on December 20, 2005, following the completion of discovery, defendant filed a cross-motion for summary judgment. On May 2, 2006, the case was reassigned to the undersigned. After a referral for alternative dispute resolution did not lead to a settlement, the court, on November 15, 2006, denied the pending dispositive motions. It found that the proceeds from the sale of the Financial Services stock could not be deemed a distribution by Sun Life of a policy dividend, or the equivalent thereof, so as to be excluded from gross income as a return of capital under the annuity rules of section 72 of the Code. The court then concluded that it could not resolve, as a matter of law, plaintiff's claim that no capital gain was realized on the sale of the Financial Services Stock because the proceeds were offset by plaintiff's basis in the stock, finding that the claim presented material questions of fact that required a trial.

Trial in this case began on June 18, 2007. At trial, the parties' expert witnesses assigned dramatically different values to the basis of the ownership rights. Plaintiff's expert, Eugene Cole, testified that he could not form an opinion as to the fair market value of the ownership rights because he found the ownership rights to be inextricably tied to the policy; in his view, the ownership rights added value to the policy but never had a separate value. Defendant's expert, Mark Penny, determined that the fair market value of the ownership rights was zero. He emphasized that none of the premiums were specifically dedicated to acquiring the ownership rights, that there was no available market for the ownership rights, and that it was highly unlikely, at the time the policy was acquired, that a demutualization would occur. The latter assertion was also made by defendant's expert on the insurance industry, James Reiskytl.

II. DISCUSSION

We begin with common ground. Section 61(a)(3) of the Code provides that gross income includes "[g]ains derived from dealings in property." Section 1001(a) indicates that "[t]he gain from the sale or other disposition of property shall be the excess of the amount realized therefrom over the adjusted basis provided in section 1011 for determining gain." This "language provides a straightforward test for realization" of income, the Supreme Court has stated, to wit, "to realize a gain or loss in the value of property, the taxpayer must engage in a 'sale or other disposition of [the] property." Cottage Sav. Ass'n v. Comm'r of Internal Revenue, 499 U.S. 554, 559 (1991); see also Phil. Park Amusement Co. v. United States, 126 F. Supp. 184, 187-88 (Ct.Cl. 1954). Section 1011(a) states that "adjusted basis" is the basis determined under section 1012, with adjustments not herein relevant, which the latter section generally sets as "the cost of such property." See United States v. Hill, 506 U.S. 546, 554-55 (1993).

The rules become a bit more complicated when a taxpayer transfers only a portion of an asset previously-acquired. Then, the basis of the latter asset generally must be apportioned between the portions disposed of and retained. Treas. Reg. § 1.61-6(a) provides —

When a part of a larger property is sold, the cost or other basis of the entire property shall be equitably apportioned among the several parts, and the gain realized or loss sustained on the part of the entire property sold is the difference between the selling price and the cost or other basis allocated to such part. The sale of each part is treated as a separate transaction and gain or loss shall be computed separately on each part. Thus gain or loss shall be determined at the time of sale of each part and not deferred until the entire property has been disposed of.

Under this regulation, "where property is acquired for a lump sum and interests therein are subsequently disposed of separately, in order to compute the gain or loss from each disposition an allocation or apportionment of the cost or other basis to the several units must be made." Fasken v. Comm'r of Internal Revenue, 71 T.C. 650, 656-57 (1979), acq. 1979-2 C.B. 1; see also Gladden v. Comm'r of Internal Revenue, 262 F.3d 851, 853 (9^th Cir. 2001) ("This regulation tells us that when property is acquired in a lump-sum purchase but then divided and sold off in parts, the cost basis of the property should generally be allocated over the several parts."). This apportionment is done by dividing the cost basis of the larger property among its components in proportion to their fair market values at the time they were acquired.

Of course, for this formula to work, one must be able to derive the fair market values of the component parts of the larger property. The regulations presume these values are obtainable, stating that "only in rare and extraordinary cases will property be considered to have no fair market value." Treas. Reg. § 1.1001-1(a); see also Likins-Foster Honolulu Corp. v. Comm'r of Internal Revenue, 840 F.2d 642, 650 (9^th Cir. 1988). But, what if, despite this regulatory bravado, it proves impractical or impossible to derive the values needed for the basis apportionment formula, at least without engaging in undue speculation? Does that mean that none of the basis of the originally-acquired property is allocable to the part disposed of or that all of it is allocable thereto until exhausted? These questions, of course, beg a deeper inquiry as to how, if at all, Treas. Reg. § 1.61-6 applies in such circumstances — whether, for example, conditions not immediately apparent, perhaps those lying in the substructure of the income tax, serve to delimit the regulation? The parties vigorously dispute whether this is the case, with defendant arguing that the regulation, by its terms, is controlling, and plaintiff asseverating that the regulation, in the circumstances of this case, is inapposite. Deciding who is correct requires the court to study the evolution of the regulation, particularly with reference to the concepts of income realization and return of capital, as they have metamorphosed over time.

A.

While the earliest Revenue Acts defined income to include "gains from sales or dealings in property," see Revenue Act of 1916, ch. 463, § 2(a), 39 Stat. 756, 757 (1916); Tariff Act of 1913, ch. 16, § II(B), 38 Stat. 114, 167-68 (1913), neither they, nor the supporting Treasury Regulations, provided much guidance on how to calculate such gains. Revenue laws in 1918 and 1921 conditioned the realization of income on the receipt of property with a "fair market value, if any." Early regulations interpreted this statutory language as conditioning the occurrence of a taxable event on the receipt of property with a cash equivalency, stating that to "complete or close a transaction from which income may be realized," there must be a "change into the equivalent of cash." Shortly after these regulations were promulgated, the Treasury Department, in 1921, issued the progenitor of Treas. Reg. § 1.61-6. That regulation, Treas. Reg. 45, art. 43 (1921), dealt with the subdivision of real estate into lots and provided:

Where a tract of land is purchased with a view to dividing it into lots or parcels of ground to be sold . . . the cost . . . shall be equitably apportioned to the several lots or parcels . . . to the end that any gain derived from the sale of any such lots or parcels which constitute taxable income may be returned as income for the year in which the sale was made. This rule contemplates that there will be a measure of gain or loss on every lot or parcel sold, and not that the capital invested in the entire tract shall be extinguished before any taxable income shall be returned. . . .

Other regulations promulgated around this same time took a different tack, however. They recognized that apportioning a basis among assets acquired as a bundle might, in some situations, prove impractical, requiring income recognition to be deferred until the original cost of the whole bundle was recovered. One of these, Treas. Reg. 45, art. 39 (1921), applied to common stock "received as a bonus with the purchase of preferred stock or bonds." It provided, generally, for the apportionment of basis between the various securities purchased, but indicated that "if that should be impracticable in any case, no profit on any subsequent sale of any part of the stock or securities will be realized until out of the proceeds of sales shall have been recovered the total cost." See also Treas. Reg. 62, art. 39 (1922). Similarly, Treas. Reg. 45, art. 1567 (1921), which dealt with the non-taxable exchanges, provided that where a taxpayer received two kinds of property in such an exchange, the cost of the property originally-possessed had to be apportioned among the new properties. Id. But, "[i]f no fair apportionment is practicable," the regulation continued, "no profit on any subsequent sale of any part of the property received in exchange is realized until out of the proceeds of sale shall have been recovered the entire cost of the original property." Id.; see also Treas. Reg. 62, art. 1567 (1922); Green v. Comm'r of Internal Revenue, 33 B.T.A. 824, 828 (1935) (discussing the evolution of this regulation).

The use of "cash equivalency" principles to govern the realization of income soon proved unworkable. See 64 Cong. Rec. 2851 (1923) (stmt. of Rep. Green); Hearings Before the S. Finance Comm., 67^th Cong. 199 (1921) (stmt. of Dr. T.S. Adams, Tax Advisor, Treas. Dept.). This led Congress, in 1924, largely to abandon these principles in favor of enacting the predecessor of section 1001(a) of the Code and, with it, the concept of "amount realized" — defined, as it is today, to include the fair market value of property other than money or money equivalents received in a transaction. See Revenue Act of 1924, Pub.L. No. 68-176, § 202(c), 43 Stat. 253, 255 (1924); see also Campbell v. United States, 661 F.2d 209, 216 (Ct.Cl. 1981); Warren Jones Co. v. Comm'r of Internal Revenue, 524 F.2d 788, 791-92 (9^th Cir. 1975). The accompanying Committee Reports criticized prior law as being "so indefinite that it can not be applied with accuracy, nor consistency," H.R. Rep. No. 68-179, at 13 (1924), reprinted in J.S. Seidman, Legislative History Of Federal Income Tax Laws 1938-1861, at 686 (1938), leading to "[g]reat difficulty . . . in administering" the law, S. Rep. No. 68-398, at 13-14 (1924), reprinted in Seidman, supra, at 686-87. See also Kwall, supra, at 994. The implication was clear — Congress desired more certainty in determining the timing and amount of the gains realized upon sales or exchanges. See Bradley T. Borden, "Reverse Like-Kind Exchanges: A Principled Approach," 20 Va. Tax Rev. 659, 665-66 (2001).

Into this evolving legal environment was born the so-called "open transaction" doctrine, an accouchement traced to Burnet v. Logan, 283 U.S. 404 (1931). In that case, Mrs. Logan sold stock of a closely-held corporation which assets included stock in a second corporation that owned a mine lease. Id. at 409. She and the other shareholders, which included her mother, exchanged the stock for cash and a stream of annual payments corresponding to the amount of iron ore extracted from the mine. The IRS argued that, at the time of the sale, the right to receive the mining royalties could be estimated based upon the amount of reserves at the mine and that the transaction should be taxed based upon the value of that estimate. Id. at 412. The Supreme Court demurred, holding that Mrs. Logan was entitled to recoup her capital investment in the stock before paying income tax based on the supposed market value of the mineral payments. It reasoned:

As annual payments on account of extracted ore come in, they can be readily apportioned first as return of capital and later as profit. The liability for income tax ultimately can be fairly determined without resort to mere estimates, assumptions, and speculations. When the profit, if any, is actually realized, the taxpayer will be required to respond. The consideration for the sale was $2,200,000 in cash and the promise of future money payments wholly contingent upon facts and circumstances not possible to foretell with anything like fair certainty. The promise was in no proper sense equivalent to cash. It had no ascertainable fair market value. The transaction was not a closed one. . . . She properly demanded the return of her capital investment before assessment of any taxable profit based on conjecture.

As viewed by the Logan Court, then, the income tax law did not resolve every doubt in favor of taxation — irreducible values could exist in that world, with the effect of postponing the recognition of income. In the years that followed, the predecessor regulations to Treas. Reg. § 1.61-6 and the "open transaction" doctrine developed like a pavane — intertwined in theory, but rarely touching in the decisional law. A dozen years after Logan, in Pierce v. United States, 49 F. Supp. 324 (Ct.Cl. 1943), it was not the taxpayer, but the United States, that claimed that a transaction was still open. In that case, the First National Bank of the City of New York, in order to give its stockholders the benefits of investments in securities that could not then be lawfully held by a bank, organized a separate company, First Security Company, to invest in such securities. Each of the certificates of stock in the bank was endorsed with a statement that the stockholder had an interest in the dividends or profits, and, in case of dissolution, in the distribution of capital of the Security Company, ratable with its interest in the bank. Id. at 329. Via this arrangement, the shareholders also had limited control over the Security Company, albeit control exercised through the votes of the holders of two-thirds of the bank stock. Neither the bank stock by itself, nor the interest represented by the endorsement, could be transferred separately from the other. Between 1928 and 1932, the plaintiffs' testator bought thirty-five shares of the bank stock with the endorsements. The Banking Act of 1933, however, banned the securities arrangement used by the bank, causing the Security Company to be dissolved; transferable interests in the proceeds of the dissolution were issued to the bank stockholders and the endorsements were removed from the stockholders' certificates of bank stock. The plaintiffs' testator received his interest in the proceeds of the dissolution on December 6, 1933, and promptly sold them on January 29, 1934, allegedly at a loss, on account of which they sought a refund of income taxes.

The United States contended that —

the sale by plaintiffs' testator of the declarations of interest in the dissolution of the Security Company may not be treated separately as showing a loss, since his interest in the Security Company was acquired in combination with his stock in the bank, and the answer to the question whether a loss or profit resulted from the transaction cannot be had until the bank stock is sold, so that it may be known how much the combined investment has sold for.

The focus of our inquiry next shifts to Inaja Land Co., Ltd. v. Comm'r of Internal Revenue, 9 T.C. 727 (1947), acq. 1948-1 C.B. 2, an "open transaction" case much debated by the parties here. There, the taxpayer owned about 1,200 acres of land on the banks of a river that it had purchased for $61,000. The land was used for fishing and for grazing. In 1934, the City of Los Angeles began altering the flow of the water in the river; ultimately it paid the taxpayer $50,000 for a perpetual easement to allow water to flow over the land toward the city. A tax dispute arose over the treatment of this money. The Tax Court found that the amount received constituted proceeds from the disposition of an interest in real property, that is, the easement. It concluded, however, that it would be wholly impracticable and impossible to apportion a cost basis to the easement involved because the easement could not be described by metes and bounds as the flow of water was likely to change over time and was not predictable. Id. at 735. Citing Logan, the Tax Court reasoned that "[a]pportionment with reasonable accuracy of the amount received not being possible, and this amount being less than petitioner's cost basis for the property, it can not be determined that petitioner has, in fact, realized gain in any amount." 9 T.C. at 736. It concluded that "[a]pplying the rule . . ., no portion of the payment in question should be considered as income, but the full amount must be treated as a return of capital and applied in reduction of petitioner's cost basis." Id.

In Pierce and Inaja Land, then, the courts made short shrift of basis allocations that lacked a rational foundation. The First Circuit would reach the same result in Warren v. Comm'r of Internal Revenue, 193 F.2d 996 (1^st Cir. 1952), another case involving hybrid securities. There, the taxpayer purchased preferred stock of a business trust and received, as an endorsement on the stock, a guaranty of the payment of a liquidating dividend. Upon the liquidation of the trust in 1939, the taxpayer exchanged his shares of stock for cash and a claim, evidenced by certificates, against the guarantor of the liquidating dividend. The Commissioner asserted that the latter certificates had a minimal basis and that considerable income was generated when they were sold in 1944. Id. at 999. The Tax Court agreed — but, the First Circuit did not. Citing the regulations that would become Treas. Reg. § 1.61-6, the latter court stated that "[n]ormally when a taxpayer acquires an aggregate of assets for a single purchase price, on subsequent sale of any portion he must allocate a part of the price he originally paid to the portion being sold on the basis of its proportionate value at the time of purchase so that gain or loss on the partial sale can be determined. Id. at 1001. But, "[i]n some situations," the court noted, "where at the time of the acquisition of the aggregation there was no separate market for the different parts of the aggregate, rational apportionment of the purchase price between the several elements purchased cannot be made." Id. Recognizing that this principle derived from the bonus stock regulation which did not apply, by its terms, to the situation presented, the court, nonetheless, observed that "if the regulation enunciates a sound rule, as unquestionably it does, a similar principle ought to govern analogous situations where the price paid for a bundle of assets cannot be allocated among them on a rational basis." Id. Remanding the case to the Tax Court, the First Circuit concluded that if "[i]t is wholly impracticable to make such an allocation of the purchase price, proper tax treatment would be to treat the cash disbursement upon liquidation in 1939 as a return of capital going to reduced basis, and to recognize no loss until the last part of the package, the guaranty, was sold in 1944." Id.

While all these cases were percolating through the system, the Treasury Department periodically reissued the regulations dealing with the sale of real property in lots and bonus stock. These iterations, however, were triggered by the passage of new revenue acts and reflected nothing new by way of substance. See Treas. Reg. 118, § 39.22(a)-8 (1953) (bonus stock); id. at § 39.22(a)-11 (sale of real property in lots); Treas. Reg. 111, § 29.22(a)-8 (1943) (bonus stock); id. at § 29.22(a)-11 (sale of real property in lots); Treas. Reg. 103, § 19.22(a)-8 (1940) (bonus stock); id. at § 19.22(a)-11 (sale of real property in lots); Treas. Reg. 94, art. 22(a)-8 (1936) (bonus stock); id. at art. 22(a)-11 (sale of real property in lots); Treas. Reg. 86, art. 22(a)-8 (1935) (bonus stock); id. at art. 22(a)-11 (sale of real property in lots). During the later 1940s and early 1950s, the Commissioner sometimes saw fit to argue that the apportionment principles reflected in the regulation dealing with real property lots ought to apply, by analogy, to other forms of real and personal property — with varying levels of success. See, e.g., Comm'r of Internal Revenue v. Cedar Park Cemetery Ass'n, 183 F.2d 553, 557 (7^th Cir. 1950); Atwell v. Comm'r of Internal Revenue, 17 T.C. 1374, 1379-80 (1952); W.D. Haden v. Comm'r of Internal Revenue, 5 T.C.M. (CCH) 250 (1946). In an apparent effort to lessen the need to make such extending analogies, the Treasury Department, in 1957, promulgated Treas. Reg. § 1.61-6 — patterned after the provision previously applied to real property lots, but applicable now to most forms of property. See Computation of Taxable Income, 22 Fed. Reg. 9419, 9422 (Nov. 26, 1957).

Nevertheless, even after this new and broader regulation was promulgated, both taxpayers and the Commissioner continued to invoke the "open transaction" doctrine — and, at times, did so successfully. Sometimes, as in Logan, the doctrine was pressed by taxpayers claiming that no gain should be realized upon the sale of a portion of a given property until the basis of the entire original property acquired was recovered. In other instances, defendant invoked the doctrine in seeking either to: (i) postpone income to years in which the statute of limitations on assessments was still open; or (ii) argue that no loss should be deducted upon the sale or exchange of a portion of a property until all the interests comprising the property have been sold or exchanged. So things remained until the scope of the "open transaction" doctrine was tapered by the passage of several amendments to the Code. Principal among these was section 453, enacted by the Installment Sales Revision Act of 1980, Pub.L. No. 96-471, 94 Stat. 2247, which provided a new method of reporting gains on an installment basis, to be applied, unless the taxpayer elects out. See 26 U.S.C. § 453(d). Yet, the legislative history of this section confirms that Congress envisioned that the "open transaction" doctrine would still be available, albeit, to use the words in one report, in "rare and extraordinary" circumstances. And it bears noting that none of these statutes directly addressed the impact of the doctrine on basis allocations — the form of the doctrine pertinent here. Accordingly, while these statutes undoubtedly narrowed the scope of the doctrine, they did not defenestrate it — the doctrine survives, albeit in more limited form, but with its basic rationale unscathed, leaving the courts to apply it as appropriate. See Gladden v. Comm'r of Internal Revenue, 262 F.3d 851, 855 (9^th Cir. 2001) (recognizing the continuing viability of the doctrine); Davis, 210 F.3d at 1348 (same).

B.

It remains to trace, more precisely, the contours of this exception — a task complicated by the fact that the "open transaction" doctrine has "flowered into various rather disparate branches." Lykken, 42 Okla. L. Rev. at 581; see also Mertens, supra, at § 5:15; Bittker Lokken, supra, at ¶ 41.6.1. Some decisions that rely on Logan premise "open transaction" treatment on contingencies that impact the value of the compensation received and focus on what amount, if any, should be realized in the year of the sale. The debate in these cases is over income realization. Other cases premise "open transaction" treatment on the inability to separate out the values corresponding to the portions of a previously-acquired asset being sold or retained, and focus on what amount, if any, should be viewed as the basis of the portion sold. The debate in these cases — as here — is on return of capital. Encompassed within this latter category are not only cases in which it is simply impractical or impossible to determine the value of a component of a larger whole, but also, as is illustrated by Pierce, a particular species in which the doctrine was triggered because the part disposed of was, when first acquired, inseparable or indivisible from the part retained. Nonetheless, despite the variety of scions that have been engrafted onto the stock of Logan, it bears repeating that the "open transaction" exception is still limited, "confined in its application to those situations that present elements of value so speculative in character as to prohibit any reasonably based projection of worth." Campbell, 661 F.2d at 215; see also Treas. Reg. § 1.1001-1(a).

C.

We return, at last, to the facts. The experts in this case had markedly divergent views not only as to the value of the ownership rights that were transferred under the demutualization, but even as to whether those rights were susceptible of valuation. Plaintiff's valuation expert, Mr. Cole opined that traditional methods could not be used to value the "ownership rights" associated with the policy because those rights were neither separable nor alienable. While convinced these rights added to the value of the policy, he concluded that, prior to the demutualization of Sun Life, their fair market value, separate from the policy itself, was "not determinable." One of defendant's experts, Mr. Reiskytl, previously worked at a mutual insurance company. He confirmed many of the premises underlying Mr. Cole's opinion. Contrasting the ownership rights of mutual policyholders to those of traditional shareholders, Mr. Reiskytl observed that "[u]nlike shareholder ownership rights that are separate from the contractual rights of the insurance policy, the mutual policyholder's ownership rights are inextricably tied to the underlying insurance contract." (Emphasis in original). He further noted that the policyholder ownership rights could not be "separately purchased, transferred or sold." and that "[t]here is no separately determinable or identifiable price for these ownership rights at the time of purchase of an insurance policy." Yet, in a somewhat self-contradictory fashion, Mr. Reiskytl proceeded to set a value for these rights, specifically concluding, based upon various factors, that they had "no" value. Defendant's other expert, Mr. Penny, also recognized that "the subject ownership rights could not be purchased nor [sic] sold separate from the purchase of an insurance policy." But, he essentially turned a blind eye to this fact in concluding that the value of the ownership rights was "best stated at zero during the 1990 calendar year." To derive this value, Mr. Penny purportedly used cost- and market-based approaches to valuation, based on his view of the cost of replacing asset and its market value, respectively. Yet, he did so, conspicuously, without considering any comparable properties, serving to highlight the fact that there were neither such comparables, nor, for that matter, any market in which the ownership rights or some derivative could be sold.

All the experts subscribed to the same, basic definition of "fair market value" — essentially, the "price at which the property would change hands between a willing buyer and a willing seller, neither being under a compulsion to buy or to sell and both having reasonable knowledge of the relevant facts." United States v. Cartwright, 411 U.S. 546, 551 (1973) (citing Treas. Reg. § 20.2031-1(b)); see also Campbell, 661 F.2d at 221; Union Pacific R. Co., Inc. v. United States, 524 F.2d 1343, 1383 (Ct. Cl. 1975); Treas. Reg. § 1.170A-1(c)(2). Yet, in applying that definition, Mr. Cole believed that the circumstances prevented him from determining a value for the ownership rights, while Mr. Penny saw some of the same circumstances as reasons for setting that value at zero. So what caused those differences? It would appear that the experts parted company in deciding whether the nature of the ownership rights made them "impossible or impractical" to value or simply valueless. And that disagreement, in turn, undoubtedly stemmed from unstated differences as to what is meant by the phrase "impossible or impractical" — a phrase that, despite dozens of "open transaction" cases, has received little in the way of direct definition. That the phrase "impossible or impractical" has largely been left undefined almost undoubtedly derives from the fact that most "open transaction" cases are heavily fact-driven. Nonetheless, a synthesis of the decisional law yields several factors that have proven pivotal in deciding whether a rational basis exists for determining fair market value.

The first of these focuses on the marketability of the asset — both in terms of whether it is separately sellable or alienable and, if so, whether an established or private market exists in which to effectuate that sale. Several cases, among them the decisions in Pierce and Warren, have relied on the fact that an asset is not separately sellable to indicate that it lacks an ascertainable market value, particularly where that inalienability is inherent in the asset itself and not superimposed as a contractual limitation. See Pierce, 49 F. Supp. at 330; Warren, 193 F.2d at 1001. Likewise, courts frequently highlight the absence of any market in which to sell assets of the type at issue as suggesting that the targeted asset is not susceptible to valuation. A second factor focuses on whether there are any proxies that may be used to estimate the needed value — for example, recent sales or exchanges of assets comparable to the one being valued. Because having such comparables is essential to using the comparable sales or market approach to valuation, the absence thereof not only deprives these methods of any utility, but also provides further indication that the value of the asset cannot rationally be ascertained. See Estate of Wiggins, 72 T.C. at 712. In similar vein, valuation methods that depend upon some bartering or exchange convention — that the value of an item received in an arms-length transaction is equal to the value of the item given up — are also unavailing if the taxpayer, via the barter, acquired a group of items of which the object to be valued is only part. Finally, an asset is more likely to be deemed insusceptible of valuation if its value is "contingent upon facts and circumstances not possible to foretell with anything like fair certainty." Logan, 283 U.S. at 413. Contingencies whose impact cannot be reasonably estimated, in particular, frustrate the use of methodologies that derive the value of an asset based upon the present value of an income stream, e.g., the income capitalization or discounted cash flow approaches. While not all contingencies prevent such present value calculations, it remains that "open transaction" treatment has been applied in "those situations that present elements of value so speculative in character as to prohibit any reasonably based projection of worth." Campbell, 661 F.2d at 215.

Logic and experience suggest that the presence vel non of the above factors ought to be reflected in the ability (or inability) of an expert to value an asset reliably using accepted valuation methods. See McCormac, 424 F.2d at 620. The latter methods, of course, are not intended to produce results with talismanic precision, for it is well-accepted that fair market value is "incapable of mathematical precision and implicates methods of judgment." United States v. 1,378 Acres of Land, More or Less, Situate in Vernon County, State of Miss., 794 F.2d 1313, 1318-19 (8^th Cir. 1986). Yet, empirically-speaking, if an expert lacks any rational basis upon which to value an asset, that ought to be strong indication that the asset is insusceptible of valuation. The burden of demonstrating this, of course, lies squarely upon the one invoking the exception — here the plaintiff — which must show that there was not "enough hard information in place from which willing buyers and willing sellers could construct soundly based equations of value." Campbell, 661 F.2d at 215; see also Bernice Patton Testament Trust, 2001 WL 429809, at * 2; Rosenberg v. United States, 3 Cl. Ct. 432, 437 (1983) (P. Miller, J.). Ultimately, it falls to the court to consider the record evidence bearing on the factors listed above, with particular focus given the expert opinions provided, and to determine, as a factual matter, whether plaintiff has proven that a rational basis for establishing the value of the asset in question is lacking.

D.

Notably, Mr. Penny readily admitted that there was neither a market upon which to gauge the value of the ownership rights nor any assets that could be deemed comparable to those rights, so as to allow for accurate application of the market method of valuation. Rather, he, and to a lesser extent, Mr. Reiskytl, set the value of the ownership rights at zero because Sun Life had not incurred any costs in establishing those rights — that is, because prior to the demutualization, Sun Life had neither associated any cost with the ownership rights on its books nor accounted for the rights in pricing its policies. But, it is hard to see how either fact is relevant, let alone dispositive here. There is, to be sure, a cost replacement method for appraising the value of an asset — but that method does not establish value based upon the historical costs incurred by the seller with respect to an asset, but rather relies on the cost to the purchaser of replacing or reproducing the asset. Indeed, various courts have rejected claims that an asset has a value of zero because the entity or individual creating it did not associate specific past costs or future liabilities with that asset. For example, in Piper, supra, the Tax Court considered common stock subscription warrants that were acquired by the taxpayer with other stock as part of a reorganization and later sold. The Tax Court viewed the fact that the corporation had not allocated any costs to the warrants as evidence that the warrants were insusceptible of valuation, noting that "[t]hey were not reflected in the capital account of the corporation and did not represent an absolute equitable ownership therein." 5 T.C. at 1110. This result is consistent with cases suggesting that "open transaction" principles apply where costs are not identifiable to a particular asset, but instead represent investment in the business as a whole — indeed, that argument has been made by the IRS itself in seeking to disallow deductions. See, e.g., Capital Blue Cross v. Comm'r of Internal Revenue, 431 F.3d 117, 125-27 (3d Cir. 2005); see also Drybrough v. Comm'r of Internal Revenue, 45 T.C. 424, 434-35 (1966); Mertens, supra at § 5:16. It would seem, then, that the fact that no specific costs were allocated by Sun Life to the ownership rights merely reflects that those rights related to values associated with the business as a whole — the rights to vote for the entire board, to receive proceeds from the company's liquidation and to receive distributions in the case of a demutualization. That no more specific accounting of these rights is available does not support defendant's case, but rather is further indication that the rights are not susceptible to valuation.

Contrary to defendant's experts, that the future financial benefits associated with the ownership rights here were speculative does not mean that those rights should be valued at zero. If that were true then virtually all of the "open transaction" cases, beginning with Logan itself, should have been decided differently. Indeed, one of the reasons why the Supreme Court held that Mrs. Logan was not required to recognize gain was because "the promise of future money payments [was] wholly contingent upon facts and circumstances not possible to foretell with anything like fair certainty." Logan, 283 U.S. at 413; see also Campbell, 661 F.2d at 215. At all events, many of the economic assumptions made by defendant's experts in diminishing the value of the ownership rights to zero do not withstand scrutiny. Those experts, for example, ascribed no value to the voting rights associated with the policies, even though those rights allowed the participating policyholders to elect the Board, which, in turn, established the policies under which dividends were paid and initiated the demutualization process. Also part and parcel of the opinions expressed by those same experts was the illogical view that no value should be ascribed to the demutualization or liquidation distribution rights — even though those rights entitled the policyholders to share potentially in a surplus that during the period in question exceeded $5.7 billion (Canadian). The notion, moreover, that any distribution of the value of the company could not have been anticipated here because demutualizations such as that conducted by Sun Life were unprecedented — a view unhesitatingly offered by both of defendant's experts — is flatly contradicted by Sun Life's own plan of demutualization. An actuarial report, dated September 28, 1999, that was attached to the plan of demutualization thus stated that "[i]n the last ten years, there have been several demutualizations in the U.S.A., the U.K. Australia, and, most recently, South Africa," adding that "they are precedents." Other sources also indicate that such demutualizations were commonplace by the time the policy in question was purchased. In short, the evidence supports plaintiff's claim that the ownership rights did have value, albeit one that was not derivable.

If nothing else, these facts are antithetical to the claim that the stock distributions made with respect to the ownership rights were a "windfall." Defendant's repeated and pejorative use of that term seemingly proceeds from the notion that, as in Orphic hymns, the value associated with the ownership rights here sprung from the aether, somehow sparked by the demutualization itself. As characterized by Mr. Scanlon, one of defendant's witnesses, these rights were "embedded values" that were not "monetized" until the demutualization occurred. But, if there is any meaningful distinction to be made between "embedded values" that were not previously "monetized" and ownership rights that were "impossible or impractical" to value at the time they were first acquired, it is almost certainly one without a difference. Nor is there the slightest support for the suggestion, again made by one of defendant's witnesses, that the allocation of stock here was a "windfall" because it was mandated by Canadian and state regulatory agencies. A silent premise in this argument, of course, is that those agencies acted arbitrarily or at least with considerable largesse in requiring compensation to be paid for the loss of ownership rights that — in defendant's view — were valueless. But, the court is no more inclined to believe this charge, sans any shred of evidence to support it, than it would be to ascribe similar conduct to Congress and Federal agencies. Without any evidence to the contrary, the more logical conclusion is that such agencies, and the legislatures that empowered them, sincerely believed that the ownership rights had value and that the policyholders were entitled to be compensated for their loss. The "windfall" tag, therefore, lacks evidentiary adhesive and does not stick.

At all events, the assertion that the ownership rights here ever had a "zero" value is thoroughly rebuffed by the actuarial study provided by Sun Life to its policyholders with the plan for demutualization. That study focused on whether the stock to be provided in the demutualization adequately compensated those policyholders for the ownership rights that were being relinquished. It recognized, as a first principle, that the stock allocation "should fairly compensate for what policyholders lose in the demutualization; namely, voting control of the insurance company and the right to share in the insurance company's residual value if it is wound up." It noted that the demutualization plan "provided for a fixed allocation of 75 Financial Service Shares to each Eligible Policyholder, regardless of the number of policies held, and for a variable allocation to each Eligible Policy of a number of Financial Services Shares which depends on its Cash Value, the number of years it has been in force and its annual premium." The study stated that it —

regarded the fix allocation as compensation for loss of voting control and the variable allocation as compensation for loss of the right to share in residual value. It is appropriate that the fixed allocation be the same to each Eligible Policyholder, since each has one vote and all the votes should be treated as equal. It is appropriate that the variable allocation differ among Eligible Policyholders because they have different customer attachments to, different financial interest in, and different rights to receive surplus distributions from, Sun Life of Canada.

In concluding that the compensation for the lost ownership rights was "fair," "equitable," and "appropriate," the report cited several other facts that suggest that the ownership rights had value prior to the demutualization, including that: (i) value comparable to that being offered in the Sun Life demutualization had been allocated to voting rights in other prior demutualizations; (ii) the loss of voting control could indirectly impact policy dividends, the payment of which was "largely at the discretion of [Sun Life's] board of directors; and (iii) the "primary historical sources of surplus" for Sun Life had been its "individual participating policies." Finally, and importantly, the report recognized that the distribution of stock was a "zero-sum game," that is to say, that certain policyholders would be harmed if the plan struck the wrong balance between the value of the voting rights and the residual rights. Of course, there would be no such harm — and, concomitantly, no need to strike such a careful balance — if either right properly were characterized as a "windfall."

In sum, based on the record, the court simply cannot credit defendant's "zero" valuation of the ownership rights. The opinions of its experts, "like any other judgment . . . can be no better than the soundness of the reasons that stand in support of them." Fehrs v. United States, 620 F.2d 255, 265 (Ct.Cl. 1980); Campbell, 661 F.2d at 222. And, the court finds that the premises upon which these opinions were based were faulty and contrary to the facts. The record here instead supports the opinion rendered by plaintiff's valuation expert that the value of the ownership rights was not discernible, leading the court to conclude that plaintiff has borne its burden of proof in this case. That the facts in this case parallel strikingly those in several "open transaction" cases involving stock and other forms of securities, among them Pierce (which remains binding precedent in this circuit) and Warren, serves only to confirm that this is an appropriate situation in which to apply the "open transaction" exception to Treas. Reg. § 1.61-6. That being the case, and the amount received by plaintiff being less than its cost basis in the insurance policy as a whole, the court finds that plaintiff, in fact, did not realize any income on the sale of the stock in question and, therefore, is entitled to the requested refund.

III. CONCLUSION

IT IS SO ORDERED.