based on the special facts of each case. No appellate court, no matter how ingenious, can devise a simple, mechanical formula which will reveal the “correct” characterization of the transaction at issue in every instance. As the Court held in Cumberland, “[i]t is for the trial court, upon consideration of the entire transaction, to determine the factual category in which a particular transaction belongs. Here as in the Court Holding Co. case we accept the ultimate findings of fact of the trial tribunal.” 70 S.Ct. at 282-83.
In the instant case, the district court concluded that the government’s Court Holding theory was “inapplicable to the factual situation presented in this case,” finding that the distribution of the option to Son occurred before a sale of the property was even contemplated,11 that Son was personally involved to a substantial extent in the development activity which led to the increased value of the Piedmont-Old Ivy properties,12 and that Son negotiated with Pope & Carter in his own behalf with respect to his interest in the properties.13 These findings must be accepted unless clearly erroneous. Hines v. United States, supra, 477 F.2d at 1071 n. II.14 Our review of the record convinces us that there is ample evidence to support the court’s factual determinations. We therefore affirm the court’s judgment that this transaction was not stricken by Court Holding’s selling virus.
In so holding we are not unaware of certain doubts raised by the government concerning the economic reality of the transactions between Corporation and Son. The government correctly notes that the corporation bore all the risks with respect to the zoning action, which was necessary to make the purchase a profitable venture. Son did not actually exercise his option and acquire title to the property until favorable zoning was obtained and the sale to Pope & Carter became a virtual certainty. Furthermore, Corporation actively participated in the negotiations with Pope & Carter which led to the sale of the Piedmont-Old Ivy properties. Son never expended a single cent on the transaction, either in acquiring the option or in exercising it. Indeed, the note Son gave Corporation as payment for the properties was modified shortly after its execution to allow a pass through to Corporation of Pope & Carter’s payments to Son. The government further argues that, as a matter of law, the options granted by Corporation to Son were merely continuing [872]*872offers to sell which did not transfer a present property interest to Son. Were this the case, no property interest would have been transferred from Corporation to Son until after the agreement with Pope & Carter was negotiated and virtually consummated by Corporation. To the extent Son participated in the negotiations before the sale from Father to Son, his activities would have been on behalf of Corporation since Son would have had no interest of his own. In such circumstances, a finding that the sale of Son’s interest to Pope & Carter’s assignee was made by Son, rather than by Corporation, would be difficult, if not impossible, to support.
The critical link in this chain of reasoning by the government is its legal argument that the option is invalid under state law for lack of consideration. We disagree. In Smith v. Wheeler, 283 Ga. 166, 210 S.E.2d 702 (1974), the Supreme Court of Georgia held that an option to buy real estate was binding on the optionor even though the optionee failed to deliver the one dollar recited in the contract as the sole consideration for the option. The Court reasoned that the recital to pay a nominal consideration gives rise to an implied promise to pay which can be enforced by the other party. The government attempts to distinguish Smith on the ground that the instant case involves an intra-family transaction while in Smith the Georgia Court was confronted with an arm’s-length agreement struck by unrelated individuals. We can discern no basis for such a distinction under state law. Related individuals are free to enter into binding contracts, and we have found no Georgia cases holding that different rules govern the validity and enforceability of a contract where related individuals are involved. Here the option agreement was formalized in writing and signed by both parties. Under Georgia law, the recitation of nominal consideration having been paid by the optionee is sufficient to bind the optionor.15
[873]*873Notwithstanding our rejection of the government’s legal position that a grant of an option by a closely held corporation to a shareholder for nominal consideration cannot transfer a property interest, we concede that the factors discussed by the government raise doubts concerning the district court’s factual characterization of the transaction.
Nonetheless, there is other evidence in the record which supports the district court’s findings and conclusions. Under the standard applied by this Court in Hines v. United States, supra, the critical factual question is whether Corporation “actively participated in the transaction that produced the income to be imputed.” 477 F.2d at 1069 (emphasis supplied). Application of this standard in the case at bar is complicated by the fact that both Son and Corporation had interests in the property. In the more typical situation, such as that presented in Hines, the entire property is transferred to the shareholder. In such a case, any participation by the corporation in negotiating the sale of the property after that property is distributed to the shareholder evidences a sale by the corporation, rather than the shareholder. Here, however, Corporation retained a one-half interest in the property. It is absurd to suggest that income may be imputed to the corporation simply because it negotiated with third parties concerning this disposition of its own interest in the property. Instead, the crucial question becomes whether the Corporation “actively participated” in the sale of Son’s share of the property. This is a question of fact, and proper deference must be given to the findings of the district court. The district court found in this regard that Father negotiated with Pope & Carter “on behalf of the Corporation” and that Son negotiated “on his own behalf.” See footnote 13 supra. This finding is itself supported by the record and is further supported by two subsidiary findings indicating the reality of Son’s interest in the property.
The court found that at the time the option was granted, Father and Son intended to use the property themselves “for a motel.” This finding, supported by evidence in the record, indicates that no sale of the property was contemplated when Son received the property interest reflected by the option. This, in turn, supports the conclusion that the option was not merely a subterfuge designed to conceal Corporation’s subsequent sale of the entire property, but instead was intended to pass an interest to Son which Son, and not Corporation, later disposed of in the sale to Pope & Carter.
The court’s determination that Son, rather than Corporation, negotiated with respect to Son’s interest in the property is further supported by evidence that Son was personally involved to a substantial extent in the development activity which led to the increased value of the properties. The district court found that these efforts were attributable to Son’s belief that he had an interest in the property.16 The fact that Son had previously treated the option as constituting a personal property interest and had relied on this personal interest in developing the Piedmont-Old Ivy tracts strongly suggests that when Son later participated in the negotiations with Pope & Carter, he also did so in his own behalf. The prior treatment of a transaction by the parties is a relevant consideration, which in this case, supports the finding that Son, and [874]*874not Corporation, negotiated with Pope & Carter with respect to Son’s interest in the property.
In sum, we are not persuaded that the transactions at issue must necessarily be characterized as a sham, as less than bona fide, or as, “in substance,” a sale by Corporation. There was evidence in the record to support the findings that Son received his option interests before a sale was contemplated, that the option was treated by all the parties as a meaningful property interest, and that the option was the impetus for Son’s personal efforts which significantly contributed to the appreciation in value of the property. These findings, plus other evidence in the record, supports the findings that Son, not Father, negotiated with Pope & Carter in his own behalf with respect to the sale of Son’s interest in the property and that Corporation did not actively participate in the sale of Son’s interest. While perhaps we might have characterized these transactions differently had we been responsible for the initial factual determination, in our role as an appellate court we may not impinge on the district court’s fact-finding prerogatives. The factors outlined above provide a sufficient basis for the court’s conclusion that the sale of Son’s interest in the property was not the sale of Corporation. Accordingly, we affirm the court’s decision not to impute income from that sale to Corporation under the Court Holding rationale.
We wish to make clear that we have not ignored the government’s argument that this decision will enable closely held corporations to avoid Court Holding by granting options to their shareholders on every piece of corporate property, thereby avoiding corporate tax if the property is later sold. This argument, however, does not persuade us that the result urged by the government is required in this case. In this area of tax law, each case will necessarily turn on its own particular facts. Evidence that a corporation has distributed hundreds of options to its shareholders would clearly be relevant to the district court’s evaluation of an option transaction before it. We emphasize that where the district court correctly applies the proper legal standard, the characterization of a transaction as, in substance, a sale by the corporation or a sale by the shareholder rests on findings of fact which are within the province of the district court. The district court’s determinations must be accepted on appeal where the factual findings are not clearly erroneous. To reiterate, “[i]t is for the trial court, upon consideration of the entire transaction, to determine the factual category in which a particular transaction belongs.” United States v. Cumberland Public Service Co., supra, 70 S.Ct. at 282-83. Nothing in today’s opinion is meant to preclude the trial court from imputing income to a corporation where the record indicates that, in reality, the option at issue is merely a sham device for transmogrifying what is truly a corporate sale into a sale nominally made by the shareholders.
At a more fundamental level, the government’s argument fails to recognize the limited scope of the Court Holding doctrine. The Supreme Court in Cumberland affirmed the factfinder’s characterization of the transaction at issue even though that transaction was designed solely to avoid double taxation of corporate gains. In Hines v. United States, supra, we held that income need not be imputed even though the transfer was made “in anticipation of a sale by the shareholders, and with no valid business purpose aside from motives of tax avoidance.” 477 F.2d at 1069, 1070. As the Supreme Court stated in Cumberland:
Congress having determined that different tax consequences shall flow from different methods by which the shareholders of a closely held corporation may dispose of corporation property, we accept its mandate.
70 S.Ct. at 282.
Thus where corporate property is distributed to a shareholder pursuant to a valid option, and the sale of that property is, in reality, negotiated and consummated by the shareholder rather than by the corporation, the courts are not permitted to impute the [875]*875income from that sale to the corporation. Just as we may not inoculate transactions infected by the Court Holding selling virus, we may not allow that virus to reach epidemic proportions in response to the government’s lamentations that its coffers are ailing and ill-nourished. Such decisions are best left to Congress.
III. The Option: Arm’s-Length Transaction or Constructive Dividend?
Our conclusion that income from the sale of Son’s interest in the Piedmont-Old Ivy property is not imputable to Corporation under Court Holding requires us to examine the government’s alternative theory that the ascertainable value of property conferred by Corporation on Son without adequate consideration is taxable to Father as a constructive dividend.17 The district court substantially agreed with this theory and held that Father received a constructive dividend when Corporation granted the option to Son in 1966 and amended it in 1967. However, finding that the value of the option was not ascertainable in 1966 or 1967, the court deferred recognition of the dividend and liability for the tax until Son exercised his option in 1968. The court then valued the option with reference to the consideration paid by Pope & Carter for monthly extensions of its option.
Our review of the court’s holding involves several district inquiries. In subsection III(A) we discuss why, as a general matter, the district court was correct in refusing to assume at once that a transaction between a corporation and the sole shareholder’s son was an arm’s-length transaction. We then consider in subsection III(B) whether constructive dividend treatment was warranted in this case. Finally, in subsection III(C) we examine the district court’s valuation of the constructive dividend.
We note at the outset that our journey through the constructive dividend area of tax law is a difficult one, often with no clear precedents to guide our way. In seeking to divine congressional intent, we must view the Code as a coherent whole, gleaning whatever rules and policies it may offer for deciding the issues before us. We attack these issues with a full arsenal of judicial weapons, including our own precedents, the Treasury Regulations, the Code sections which apply directly and those which indicate congressional purpose by analogy.
A. Judicial Scrutiny of Transactions Between Closely Held Corporations and Their Shareholders: A General Discussion
We begin by considering taxpayers’ contention that the option to Son was granted by Corporation at a fair price in an arm’s-length transaction. According to taxpayers, no tax consequences should attach either to the grant or to the exercise of the option. In taxpayers’ view, Son’s option should be treated no differently than the option granted to Pope & Carter. We conclude that this argument is untenable. In order to elucidate our views on the oft-perplexing problem of the conferral of benefits on a controlling shareholder by a closely held corporation, we begin with first principles.
Distributions of “property” made by a corporation to a shareholder with respect to its stock are dividends to the extent of the corporation’s current and accumulated earnings and profits, 26 U.S.C. §§ 301(a)(c), 316(a),18 and must included by the share[876]*876holder in gross income. 26 U.S.C. §§ 301(a), 301(c)(1).19
For purposes of corporate distributions, the Code defines “property” as
money, securities, and other property; except that such term does not include stock in the corporation making the distribution (or rights to acquire such stock).
26 U.S.C. § 317(a). In accordance with this broad definition of property, dividends may be “in cash or in kind, and may also result when the corporation makes a ‘bargain sale’ of its property to the shareholder at less than fair market value. When there is a ‘bargain sale,’ the shareholder receives a dividend in the amount of the difference between the fair market value and the price paid for the corporate property . . . .” Green v. United States, 460 F.2d 412, 419 (5th Cir. 1972). As the Supreme Court has recognized, “a sale, if for substantially less than the value of the property sold, may be as effective a means of distributing profits among stockholders as the formal declaration of a dividend.” Palmer v. Commissioner, 302 U.S. 63, 58 S.Ct. 67, 82 L.Ed. 50 (1937). No such formal declaration is required for a shareholder to be charged with a constructive dividend. Crosby v. United States, 496 F.2d 1384 (5th Cir. 1974).
Two other guiding principles require mention at this stage. It is clear that the intent of the parties does not govern the characterization of a distribution. Instead, courts have looked to the economic effect of the transaction at issue in determining the existence of a dividend. See id; Loftin and Woodard, Inc. v. United States, 577 F.2d 1206, at 1214 (5th Cir. 1978). It is also well settled that a dividend does not escape taxation “simply because it fails to pass through the hands of the particular taxpayer when . . . the dividend is diverted at the behest of the shareholder into the hands of others.” Green v. United States, supra, 460 F.2d at 419, quoting Sammons v. United States, 433 F.2d 728, 730 (5th Cir. 1970), cert. denied, 402 U.S. 945, 91 S.Ct. 1621, 29 L.Ed.2d 113 (1971).
In the Green case, supra, this court reviewed a district court’s judgment that, as a matter of law, a sale of property by a closely held corporation to the children of the corporation’s controlling shareholder did not constitute a constructive dividend. The government had contended that the property was sold to the children at less than fair market value. We reversed the district court, holding that if the corporation did in fact consummate a transaction with favorable consequences for a controlling shareholder’s immediate family, “only in the most extraordinary circumstances” would it be possible to conclude that the shareholder had not heavily influenced the corporation’s decision. For that reason, we held that the sale was a constructive dividend. 460 F.2d at 420-21.
Likewise, in Crosby v. United States, 496 F.2d 1384 (5th Cir. 1974), this court closely examined a transaction to determine whether constructive dividend treatment was warranted. There the corporation made improvements to its stockholder’s land after the stockholder had conveyed the land by deed to the corporation. The court determined, however, that the only property right which the corporation had acquired through the transfer of the deed was the right to sell the property. An investigation of the substance of this transaction convinced the court that the underlying purpose of many of the improvements was to benefit the stockholder. We therefore remanded the case for a determination of “which improvements were for a corporate purpose and which were not.” Those that were not, we said, were to be treated as constructive dividends. As we recently remarked in Loftin and Woodard, Inc., supra, at 1216,
“[h]ad the Crosby court merely accepted this transaction at face value, it would have overlooked the existence of an economic benefit conferred by the corporation without expectation of repayment [877]*877and, as a result, overlooked a potential constructive dividend.”
This comment is particularly apposite to the instant case. See also Commissioner v. Gordon, 391 U.S. 83, 88 S.Ct. 1517, 1521, 20 L.Ed.2d 448 (1968); Hardin v. United States, 461 F.2d 865 (5th Cir. 1972); Honigman v. Commissioner, 466 F.2d 69 (6th Cir. 1972). See generally Commissioner v. LoBue, 351 U.S. 243, 76 S.Ct. 800, 803, 100 L.Ed. 1142 (1956) (sale of corporation’s own stock to employee at a bargain price pursuant to an employee stock option did not constitute an arm’s-length transaction and employee realized taxable gain when he purchased the stock).
These expansive principles governing the treatment of constructive dividends reflect an awareness by Congress and the courts of the ingenious methods which have been devised by owners of closely held corporations trying to escape taxes at the corporate or shareholder level. “Instances of ‘constructive’ or ‘disguised’ distributions are commonly encountered in the context of closely held corporations whose dealings with their stockholders are, more often than not, characterized by informality.” B. Bittker & J. Eustice, Federal Income Taxation of Corporations and Shareholders 7-24 (3d ed. 1971). As ’ a consequence, courts have generally examined dealings between a closely held corporation and its shareholders or relatives of shareholders with a jaundiced eye and microscopic vision. Such transactions are simply not entitled to the presumption that they are conducted at arm’s length. Indeed, they call for special scrutiny. We suggest that the metaphor of a corporation and its sole shareholder walking arm-in-arm is more descriptive of reality than the vision of the corporation and shareholder holding each other apart at arm’s length.20
The district court, then, was properly skeptical when taxpayers insisted that Corporation and Son dealt at arm’s length. Both law and common sense require us not to presume that the option granted to Son was entitled to the same treatment as, for example, the Pope & Carter option, which all concede was negotiated at arm’s length. We now proceed to an examination of Son’s option to determine whether the district court properly held Father taxable on a constructive dividend in the circumstances of this case.
B. The Grant of the Option as a Constructive Dividend
Our inquiry is in two parts. First, we must focus on whether, in fact, the grant of the option conferred a benefit on Son by distributing valuable rights without consideration.21 If we determine that the option conferred such a benefit, we must then examine whether the resulting corporate distribution constituted a constructive dividend to Father. “[W]hether or not a corpo[878]*878rate distribution is a dividend or something else, such as a gift, compensation for services, repayment of a loan ... or payment for property purchased, presents a question of fact to be determined in each case.” Hardin v. United States, 461 F.2d 865, 872 (5th Cir. 1972), quoting Lengsfield v. Commissioner, 241 F.2d 508, 510 (5th Cir. 1957).
1. The Option: Valuable Rights or Fair Bargain?
Our initial task of valuing the option to determine whether the corporation distributed valuable rights to Son is greatly complicated by the unsettled state of the law in this area. Neither the Code nor the applicable case law provides clear guidance. Taxpayers maintain' that under the Supreme Court’s decision in Palmer v. Commissioner, 302 U.S. 63, 58 S.Ct. 67, 82 L.Ed. 50 (1937), there is no dividend if the exercise price of the option represents the fair market value of the underlying property at the time the option is issued, even if the value of the underlying property has appreciated greatly by the time the option is exercised. Taxpayers argue that the $100,-000 exercise price of the option on the combined properties was a reasonable valuation of the fair market value of those properties at the time the option was granted to Son.22 Thus, according to taxpayers, no valuable rights constituting a dividend were distributed by Corporation.
The government responds first that taxpayers’ argument rests on an erroneous reading of Palmer, and, second, that the 1954 Code has modified this aspect of the Palmer decision in any event. We agree with the government that the position contended for by taxpayers is justified neither by good sense 23 nor by the applicable case law.
The Supreme Court’s decision in Palmer has been a controversial one. In that case, a corporation distributed options, exercisable for 15 days, to its shareholders, entitling those shareholders to purchase from the [879]*879corporation portfolio shares in a second corporation. On the date distribution of the option was authorized, the exercise price of the options equalled the market value of the portfolio shares. By the time the options were issued, eight days later, the market value of the portfolio shares had increased considerably.
The Court held first that under the Revenue Act of 1928, a shareholder who is granted an option to purchase portfolio shares is not subject to income tax on the receipt of the option, even though the option itself is valuable. “The mere issue of rights to subscribe and their receipt by stockholders, is not a dividend. No distribution of corporate assets or diminution of the net worth of the corporation results in any practical sense.” 58 S.Ct. at 70. The Court then considered whether the shareholders would be taxable upon exercise of their option rights when a spread developed between the option price and the market value of the shares. The Court held that the exercise of these rights does not result in a dividend if the circumstances demonstrate “that distribution of corporate assets, effected by the sale, was not intended to be a means of distributing earnings, and that the price when fixed represented the fair market value of the property to be distributed.” Id. at 71. Such a distribution “is not converted into a dividend by the mere circumstance that later, at the time of their delivery to the stockholders, [the assets] have a higher value.” Id.
We recognize that this language appears to support taxpayers’ contention that because the exercise price of Son’s option equalled the fair market value of the property when the option was granted, see footnote 22, supra, no dividend should result from either the grant or exercise of the option. The continuing vitality of this aspect of the Palmer decision under the 1954 Code, however, is far from self-evident. The commentators generally have been critical of the reasoning of Palmer and most have concluded that a different result is mandated by the 1954 Code. See e. g., Carlson, Taxation of “Taxable” Stock Rights: The Strange Persistence of Palmer v. Commissioner, 23 Tax.L.Rev. 129 (1968); Whiteside, Income Tax Consequences of Distributions of Stock Rights to Shareholders, 66 Yale L.J. 1016, 1026-27 (1957); Comment, 51 Calif.L.Rev. 146, 150-51 (1963). See also, Rev.Rul. 70-521 (A distribution by a corporation of short-term rights to acquire stock in another corporation is a section 301 distribution within the meaning of section 317. The value of the dividend equals the fair market value of the rights, rather than the spread between the value of the stock and the exercise price). We need not resolve this question, however, because a closer reading of Palmer reveals that the Court’s holding in that case does not reach the transaction now before us. In Palmer the stock rights were extended to shareholders of a publicly held corporation and were exercisable for only fifteen days. The option price, when fixed eight days before the rights were issued, reasonably reflected the stock’s current market value. As the Court observed, this was the “only feasible method by which a corporation of large membership can effect a sale of its assets to stockholders.” Id. at 71.
A close examination of the language used throughout the Palmer opinion reveals that the Court viewed the issuance of stock rights, exercisable for only fifteen days, as merely a convenient method of selling corporate property to its shareholders at arm’s length. For example, the Court analogized the transaction to an ordinary sale by a vendor:
When the corporation has committed itself to a sale of its assets to stockholders at present market value, the effect on its balance sheet is the same as in the case of other vendors who in various ways assume the risk of rising prices pending the consummation of the sale.
Id. At another point in the opinion the Court commented that “[a]ny vendor who offers property for sale at a named price similarly carries the burden of risk that the property may increase in value between offer and acceptance. . . . It is an inseparable incident of every sale except those in which conditions admit of payment [880]*880for the property simultaneously with its tender for sale, a procedure which may not be available to a corporation seeking to sell its property to stockholders.” Id.
This characterization of the stock rights at issue in Palmer as merely a reasonable method for selling portfolio shares, with neither the purpose nor the effect of distributing profits, distinguishes Palmer from the case at bar. Here the option remained open for over two years. It was contemplated from the beginning of the transaction that rezoning of the property, which would substantially increase the value of the land, might be obtained. Indeed, the rezoning efforts of Father, Son, and Corporation permeated the entire transaction. Son’s option cannot be compared with the stock rights in Palmer which were designed merely to effectuate an immediate sale of the property. As taxpayers admit in their brief, the obvious reason for granting the option at issue here was to provide Son an opportunity to participate in a real estate venture and reap the benefit of any appreciation in the property. Brief for Appellees-Cross Appellants at 22. The potential for appreciation of the underlying property during the extended option period made the option a valuable commodity, the grant of which was expected to confer a benefit on Son.
Another pre-1954 Code Supreme Court decision in a closely analogous area of the law supports our reading of Palmer. In Commissioner v. Smith, 324 U.S. 177, 65 S.Ct. 591, 89 L.Ed. 830 (1945), a corporation granted an employee an option to purchase certain portfolio shares of stock in another corporation. The exercise price equalled the value of the underlying shares at the time of grant, but the option was not exercised until four years later, when the value of the stock greatly exceeded the exercise price. The Court held that the employee was properly taxed on the spread between the value of the stock at exercise and the exercise price, despite the fact that no such spread existed at the time the option was granted. While the precise question before the Court was whether the appreciation in the stock constituted compensation for personal services of the employee, taxable as income under § 22(a) of the Revenue Act of 1938, the Court’s reasoning is instructive in the constructive dividend area. First, the Court noted that the parties “anticipated” an increase in the market value of the stock. 65 S.Ct. at 592. The Court then stated: “the compensation for [the employee’s] services, which the parties contemplated, plainly was not confined to the mere delivery to [the employee] of an option of no present value, but included the compensation obtainable by the exercise of the option given for that purpose.” Id. at 593 (emphasis supplied).
In both Smith and the case at bar, a corporation granted an option which remained open for an extended period of time. In both cases appreciation in the value of the underlying property was contemplated. Unlike the transaction in Palmer, no immediate sales of the optioned properties were intended. The absence of these features in Palmer may have made it reasonable to value the constructive dividend in that case at the spread between the value of the underlying property and the exercise price of the option when the option was issued. See footnote 26 infra. But here the parties contemplated both that the optioned property would greatly appreciate in value if favorable zoning were obtained and that the exercise of the option would await such appreciation. Under the circumstances, the Court’s reasoning in Smith requires that we consider the duration of the option period and the potential for increase in value of the underlying property along with the exercise price in determining the appropriate tax treatment of the option.
Our emphasis on the extended duration of the option and on the contemplated appreciation in value of the underlying property is fully consistent with the applicable provisions of the 1954 Internal Revenue Code, which, of course, govern the transaction at issue. Unlike the prior revenue laws, subchapter C of the 1954 Code contains elaborate and specific rules governing corporate distributions. As we have [881]*881previously discussed, a distribution of “property” made by a corporation to a shareholder with respect to its stock constitutes a dividend. 26 U.S.C. §§ 301(a)(c), 316(a). The definition of property in section 317(a) is extremely broad; it includes any property except the distributing corporation’s own stock and rights to purchase such stock:
[T]he term “property” means money, securities, and any other property; except that such term does not include stock in the corporation making the distribution (or rights to acquire such stock).
26 U.S.C. § 317(a). This express exclusion of rights to acquire stock in the distributing corporation strongly suggests that these stock rights would otherwise be included in the definition of property. This in turn further strengthens the argument that rights or options to acquire any other corporate asset constitute “property” within the meaning of section 317(a).24 We conclude that under the 1954 Code, a distribution of an option to acquire corporate property constitutes a distribution of property within the meaning of section 301. This distribution is a dividend if the corporation has sufficient earnings and profits. 26 U.S.C. § 316(a). See Commissioner v. Gordon, 391 U.S. 83, 88 S.Ct. 1517, 1521, 20 L.Ed.2d 448 (1968).
The 1954 Code also specifies the method for determining the value of a distributio'n. Under section 301(b)(1)(A), the amount of the distribution is measured by the fair market value of the property received. Thus, in the case at bar the amount of the distribution should be measured by the fair market value of the option, rather than by the spread between the exercise price and the value of the underlying property at the time of grant.25 See Rev.Rul. 70-521. While that spread in some cases may approximate the fair market value of the option,26 here the duration of the option period and the likelihood of obtaining favorable rezoning must also be considered in determining the option’s fair market value. See Treasury Regulation 1.421-6, discussed infra. Thus the determination of whether the grant of the option conferred a benefit on Son is governed by the value of the option, rather than by the spread between the exercise price and the value of the underlying property.
Applying this standard, we are in complete agreement with the district court’s finding that the rights granted under the option were of great value to Son. The option entitled Son to reap the benefits of the contemplated rezoning without incurring any of the risks associated with the real estate venture. He paid no consideration for his rights. If, during the two and one half year option period the properties appreciated in value, Son would realize a gain. If the value of the property decreased or remained the same, Son would be [882]*882no worse off. Son was truly in a “heads I win, tails I don’t lose” position. This “option privilege” represented valuable property to Son.27 As Treasury Regulation 1.421-6, which addresses an analogous question in the employee stock option area, explains:
[Irrespective of whether there is a right to make an immediate bargain purchase of the property subject to the option, the fair market value of the option includes the value of the option privilege. The option privilege is the opportunity to benefit at any time during the period the option may be exercised from any appreciation during such period in the value of the property subject to the option without risking any capital.
Any doubts that Corporation distributed valuable rights to Son when it granted the option are dispelled by comparing Son’s option with that received by Pope & Carter. Pope & Carter agreed to pay substantial amounts to keep its option open and, in addition, obligated itself to devote reasonable time and effort to obtaining favorable rezoning. In return it received the right to acquire the property for $500,000, a price far in excess of the fair market value of the property at the time the option was granted. Son’s option, on the other hand, remained in effect for a longer period, imposed no obligations on Son, and allowed him to purchase a one-half interest in the property for only $100,000 plus 5V¿ percent interest. Even if we ignore the fact that the amendment to Son’s option, which extended the exercise period for an additional two years without increasing the exercise price, was granted a mere thirteen days before the grant of the Pope & Carter option, see footnote 22 supra, it is apparent that in granting the option, Corporation conferred a benefit on Son by distributing valuable rights.
2. The Corporate Distribution: a Constructive Dividend?
Having concluded that Corporation distributed valuable rights to Son, we can easily dispose of taxpayers’ contention that the grant of the option should not result in a constructive dividend to Father. Whether a corporate distribution is a dividend presents a question of fact to be determined in each case. Hardin v. United States, supra, 461 F.2d at 872. A crucial consideration in characterizing a distribution is whether it primarily served a corporate purpose or a shareholder purpose.28 The [883]*883tasks of weighing the evidence, drawing inferences from the facts, and choosing between conflicting inferences are, of course, part of the function of the district court. We must not disturb that determination unless we are convinced that there is “clear error” in the district court’s findings. Loftin and Woodard, Inc. v. United States, supra, at 1214; Honigman v. Commissioner, 466 F.2d 69, 74 (6th Cir. 1972). See Lengsfield v. Commissioner, 241 F.2d 508, 510 (5th Cir. 1957).29
Here the district court found that the distribution to Son was made “by virtue of Father’s position as sole stockholder of the corporate plaintiff” and therefore resulted in a dividend to Father. The record does not require the contrary finding that a corporate purpose was served by giving Son an option to purchase property at Corporation’s cost when all parties recognized that the value of the property would substantially appreciate if favorable rezoning were obtained. Taxpayers maintain that the option was granted because of Corporation’s interest in obtaining the benefit of Son’s services. The evidence, however, more than adequately supports the conclusion that the grant of the option primarily served the personal interests of Father in his shareholder capacity. Indeed, taxpayers admit that the option was granted to satisfy the moral obligation of Father to compensate Son for taking advantage of a business opportunity Father had personally discouraged Son from pursuing. Taxpayers’ Brief at 7. No legitimate business purpose is served by the distribution of funds to satisfy the personal or moral responsibilities of a shareholder. Hardin v. United States, supra, 461 F.2d at 872-37.
Essentially taxpayers urge this court to draw inferences from the evidence different from those drawn by the district court. This we decline to do under the record in this case. As we held in United States v. Green, supra, where a corporation consummates a transaction with favorable consequences for a controlling shareholder’s immediate family, “only in the most extraordinary circumstances” would it be possible to conclude that the shareholder had not exercised “substantial influence” in causing a diversion of corporate assets, thereby justifying constructive dividend treatment. No such circumstances are present here. 460 F.2d at 420-21.
Since we hold that there is no clear error in the findings of the district court, we affirm the district court’s conclusion that Father received a constructive dividend from the grant by Corporation to Son of the option on the Piedmont-Old Ivy property.
C. Valuing the Dividend
The final issue presented by this appeal is whether the district court properly valued Father’s constructive dividend. In particular, questions have been raised concerning the district court’s use of the consideration paid by Pope & Carter in valuing Son’s option. To resolve this issue, we must first examine whether the dividend occurred upon the grant of the option or upon its exercise by Son. If we determine that grant of the option constituted a dividend, we must then consider both the district court’s finding that the option had no ascertainable value at the time of grant and the Court’s subsequent valuation of the option at the time of exercise in 1968. If, on the other hand, the dividend arose only upon Son’s exercise of the option, we need only consider the proper valuation of the option at that time.
Prior to the 1954 Code, it was settled that the grant of an option to purchase property of a corporation did not result in a dividend. “ [Distribution of corporate property could take place only on [an option’s] exercise” because “[n]o distribution of corporate assets or diminution of the net worth of the corporation results in any practical sense” when an option is granted. Palmer v. Commissioner, supra, 58 S.Ct. at 70. The adoption of the 1954 Code has prompted [884]*884critical reexamination of this aspect of the Palmer decision. The commentators generally have concluded that under the Code provisions discussed in subsections III(B) supra, the taxable event occurs when valuable rights are first granted to the shareholder. See Carlson, Taxation of “Taxable” Stock Rights: The Strange Persistence of Palmer v. Commissioner, 23 Tax L.J. 129, 140-45 (1968); Whiteside, Income Tax Consequences of Distributions of Stock Rights to Shareholders, 66 Yale L.J. 1016, 1031 (1957); Comment, 51 Calif.L.Rev. 146, ISO-51 (1963). They reason that because an option is “property” within the meaning of section 317(a), a distribution of the option is taxable under section 301 upon the option’s issuance by the Corporation. See also Rev. Rui. 70-521. Nonetheless, the continuing validity of the Palmer principle that no dividend occurs until an option is exercised has not been definitively settled by the courts. Indeed, in Commissioner v. Gordon, 391 U.S. 83, 88 S.Ct. 1517, 20 L.Ed.2d 448 (1968), the Supreme Court explicitly left open the question “[w]hether the actual dividend occurs at the moment when valuable rights are distributed or at the moment when their value is realized through sale or exercise . . ..” 88 S.Ct. at 1521.30
The litigants here agree that under the 1954 Code, the taxable event is the grant of an option, not its exercise. The government qualifies its position by maintaining that if the value of the option is unascertainable at the time of grant, under the “open transaction” doctrine of Burnet v. Logan, supra, its valuation is deferred until the option is exercised. Because the result in the case at bar would be the same whether we viewed the dividend as occurring at the time of grant or at the time of exercise,31 we shall assume, without deciding, that the dividend occurs at the time of grant.32
The district court also assumed that the dividend to Father occurred when the option was granted to Son. The court found, however, that the option had no ascertainable value at the time of grant33 and there[885]*885fore applied the open transaction doctrine of Burnet v. Logan, 283 U.S. 404, 51 S.Ct. 550, 75 L.Ed. 1143 (1931). Taxpayers contend that the open transaction doctrine does not apply to the instant case and that any constructive dividend received by Father occurred either upon the grant of the option in 1966 or upon its amendment in 1967. Because neither of these tax years is before us in this suit, taxpayers conclude that any constructive dividend should go untaxed.
Our examination of the record convinces us that the district court properly applied the open transaction doctrine. When taxable property is received in exchange for other property or in a corporate distribution, the fair market value of the property received usually can be ascertained and the amount of the distribution or the gain from the exchange can be computed and taxed at that time. See Burnet v. Logan, supra; Estate of Meade v. Commissioner, 489 F.2d 161, 163 (5th Cir.), cert. denied, 419 U.S. 882, 95 S.Ct. 147, 42 L.Ed.2d 122 (1974); Dennis v. Commissioner, 473 F.2d 274, 285 (5th Cir. 1973). Treasury Regulation 1.453-6 states that “[o]nly in rare and extraordinary cases does property have no fair market value.”
In those rare circumstances in which no fair market value of the property can be ascertained, however, it is impossible to compute the tax. As a result, the courts have found it necessary to treat the transaction as “open” until the value of the property can be ascertained. See Burnet v. Logan, supra; Estate of Meade, supra, 489 F.2d at 163; Dennis, supra, 473 F.2d at 285. Most eases applying the doctrine involve a sale or exchange of property. See, e. g., Burnet v. Logan, supra. This court has used the open transaction doctrine in taxing a corporate distribution upon liquidation. See Estate of Meade, supra. While we have discovered no cases involving a corporate dividend, we see no reason for treating a dividend of unascertainable value differently from the distribution of an antitrust claim upon liquidation in Estate of Meade or to the receipt of mineral rights in Burnet v. Logan. In all these cases it is equally appropriate to defer taxation until the correct amounts can be accurately determined. Moreover, the open transaction doctrine has been applied by Treasury Regulation 1.421-6 to nonstatutory employee options to purchase stock or other property of the employer. Such an option is taxable upon grant only if it has a “readily ascertainable fair market value” when granted. Otherwise, it is taxable at exercise. Notably, the regulation provides that “[although options may have a value at the time they are granted, that value is ordinarily not readily ascertainable unless the option is actively traded on an established market.” 34
In this tax refund suit, the burden is on the taxpayer to establish that the option had an ascertainable fair market value at the time of grant. See Crosby v. United States, 496 F.2d 1384, 1390 (5th Cir. 1974). Taxpayers introduced evidence concerning the fair market value of the underlying Piedmont-Old Ivy property at the time of grant, but the record contains no evidence concerning the value of the option itself. That value reflects not only the exercise price, but also the duration of the option period and the likelihood of obtaining favorable rezoning. The record simply does not establish a basis for evaluating either of these factors. As was the case in Burnet v. Logan, the value of the option was “contingent upon facts and circumstances not pos[886]*886sible to foretell with anything like fair certainty.” 51 S.Ct. at 552. Under the circumstances, taxpayers failed to satisfy their burden of demonstrating an ascertainable fair market value for the option at the time it was issued, and the district court’s finding must be affirmed.
Since there was no ascertainable value, the district court correctly held that the recognition of dividend income by Father should be deferred under Burnet v. Logan until the option was exercised and its value could be ascertained. The court nonetheless erred in valuing the dividend with reference to the monetary consideration paid by Pope & Carter for its option.
The court concluded that since Pope & Carter was willing to pay $3,000 per month to extend its option, Son’s option to purchase a one-half interest in the property was worth $1,500 per month during the period that both the Piedmont and Old Ivy Properties were subject to the option, and a pro rata amount, $1,312.50 per month, during the period in which the option covered only the Piedmont property. This valuation assumes that Pope & Carter and Son received similar options. Yet the record leaves no doubt that Son was granted a significantly more valuable option. Son’s option had an exercise price of $100,000, while the corresponding exercise price in the Pope & Carter option for a one-half interest in the property was $250,000. Thus at the time Son exercised his rights, Son’s option was worth substantially more than that of Pope & Carter. Furthermore, the option granted to Pope & Carter required it to expend reasonable time and effort to obtain favorable zoning. Son’s option required no such activity, see footnote 21 supra, and thus was also more valuable for this reason. Neither of these factors was taken into account in the district court’s valuation of the option.
On remand the court should be guided by the Supreme Court’s decision in Commissioner v. Gordon, supra. There the Court assumed, without deciding, that the distribution to shareholders of an option to purchase stock in another company resulted in a dividend at the time of exercise. See footnote 25 supra. The Court then held that the dividend should be valued “in the amount of the difference between $16 per share [the exercise price] and the fair market value of a share ... at the moment the rights were exercised.” 88 S.Ct. at 1525 (1968). The Court thus recognized that at any moment, an option is worth at least the difference between the present market value of the underlying property and the exercise price of the option. Treasury Regulation 1.421-6(d) uses this method to value non-statutory employee stock options at the time of exercise.
Thus in the instant case the option should be valued by subtracting the exercise price of approximately $100,000 from the fair market value of the underlying property when the option was exercised on December 6, 1968. The district court determined that the market value of the Piedmont-Old Ivy property was at least $500,000 when Pope & Carter exercised its option on December 27. Because the rezoning application was approved before Son exercised his option on December 6, it is likely that most, if not all, of this appreciation had occurred by the time Son exercised the option. Nonetheless, it is for the fact-finder, and not the reviewing court, to determine the precise value of a one-half interest in the properties on December 6 and the resulting value of the option upon its exercise. We therefore remand the case to the district court for a redetermination of the value of the constructive dividend taxable to Father.35
[887]*887IV. Conclusion
The concepts of imputed income and constructive dividend have been developed to prevent taxpayers from clothing their activities in tax-repellant garb. Both doctrines require courts to look behind the superficial or formal nature of a transaction. Application of these concepts in complicated tax cases involves the difficult task of characterizing elusive transactions.
The appellate courts must, of course, articulate and elaborate the applicable law. But in dealing with the questions presented in this case, it is clear that an understanding of the factual situation and the real nature of the transaction are of primary importance. In this area a district judge is likely to have superior insights. Thus in allocating judicial responsibilities, Congress has entrusted this fact-finding process to the district courts, subject, of course, to appellate review. Here the district court found that Son, rather than Corporation, negotiated and consummated the sale of Son’s interest in the Piedmont-Old Ivy property to Pope & Carter and therefore concluded that income from this sale should not be imputed to Corporation. The court also found that valuable rights were distributed by Corporation to Son without business purposes because of Father’s oppo[888]*888sition as the controlling shareholder of Corporation and therefore concluded that Father received a constructive dividend from Corporation. These factual determinations are supported by the record, and the court’s ultimate conclusions rest on a correct application of the controlling law. These determinations are therefore affirmed. The court, however, failed to apply the proper standard for determining the value of Son’s option at the time of exercise and therefore improperly computed the amount of Father’s constructive dividend. Accordingly, this portion of the case must be remanded for further proceedings consistent with this opinion. The judgment of the district court is AFFIRMED IN PART, REVERSED IN PART, and REMANDED IN PART.