dissenting.

I think the judgment should be affirmed. To reverse it is to disregard a rule respecting the separate entity o" corporations having basis in logic and practicality and *481which has long been observed in the administration of the revenue acts.

Since the inception of the system of federal income taxation, capital gains have been taxed and certain capital losses have been allowed as credits against such gains. In order that this system might be practical it has been necessary to select some event as the criterion of realization of gain or loss. The revenue laws have selected the time of the closing of a capital transaction as the occasion! for reckoning gain or loss on a capital asset. A typical1 method of closure is a sale of the asset.

As the sale is voluntarily made by the taxpayer, his determination when he shall sell affects hiss capital gain or loss. He, therefore, in a sense, controls the question. whether, in a given taxable year, he must pay tax on a realized gain or may claim credit for a realized loss. Of course such a sale must be bona fide and title must pass absolutely. In the present instance the sale and transfer were such, and, as the Circuit Court of Appeals held, jhere was' not a scintilla of evidence to the contrary for the' jury’s consideration. A taxpayer who pretends he has made a sale when in fact he has a secret agreement which leaves him stillr for all practical purposes, the owner of the thing sold, is but committing a fraud upon the revenue. . ■

If the sale is bona fide, if title in fact passes irrevocably to another, that other takes as his basis, in .reckoning his gain and loss, the price he paid for the asset; and upon his future disposition of it there will be a new reckoning of gain or loss, with respect to such disposition. Here, if Innisfail either sold to the respondent or to a third party it would have to reckon gain or loss on the sale. If it distributed the asset in liquidation the respondent would be subject to a tax liability on the receipt of his.dividend. The sole question, then, is whether, as matter of law, a ' bona fide and absolute sale to a wholly owned corporation *482can constitute a completed transaction, determining a loss.

The problem ás to how a sale to a corporation wholly owned or wholly controlled by. an individual taxpayer is to be treated is not a new one. The existence of such corporations and the dealings between them and their stockholder or stockholders have long been understood. Congress was not ignorant of the problem.1 At the outset Congress might well have adopted the policy that a sale by the stockholder tó the corpóration, or vice versa, should be disregarded, and the stockholder treated as in effect the owner of the capital asset imtil its sale to a stranger. On the other hand, it would be a practical policy to recognize the separate entity of the corporation, tó treat a transfer at current value for adequate considr eration occurring between it and its sole stockholder as. closing a transaction for the purpose of reckoning either «gain or loss, and then to tax the vendee upon his or its gain or loss upon a subsequent transfer by comparison,, of the basis on which the asset was acquired and the amount realized on final disposition by the vendee. In fact, the latter course was adopted and was consistently followed until 1934 when Congress dealt with the subject.

This court, speaking by Mr. Justice Holmes, said, in Klein v. Board of Supervisors, 282 U. S. 19, 24: “ . . . But it leads nowhere to call a corporation a fiction. If it is a fiction it is a fiction created by law with intent. that it should be acted on as if true. The corporation is *483a person and its ownership is a nonconductor that makes it impossible to attribute an interest in its property to its members.”

In this view assets received on the liquidation of a one-man corporation constitute taxable income to the sole stockholder.2 Likewise, losses sustained by a corporation wholly owned by one individual may. not be reported and claimed in the individual tax return of the latter.3 And the sole stockholder and his controlled corporation may not tack successive periods of ownership to make up the two years required for an asset to become, within the meaning of the statute, a capital asset.4

This court has found that a taxable gain was realized in a case where a wholly owned corporation sold securities to its sole stockholder.5 Every element appearing in that case is paralleled here, as a comparison of the facts stated in the opinions in the two cases will demonstrate. This court said, in the earlier case, referring to the corporation: “The fact that it had only one stockholder seems of no legal significance,” and held the corporation a separate taxable entity. It is now said, however, that there is no inequity in not applying the same rule to losses as to gains because the taxpayer who exercises the option to conduct a portion of his business through the instrumentality of a wholly owned corporation does so in the full knowledge that, if he does, gains shown on sales by him to the corporation will be taxed whereas losses oh such sales will not be allowed as deductions. As hereafter will be' shown, this is now true in virtue of the amendment *484embodied in the Revenue Act of 1934 but it was not true as the law stood before the adoption of that amendment.

In 1921 the Treasury was first called upon to deal with a loss deduction arising out of a sale to a wholly owned corporation. In that year it published Law Opinion 1062.6 It was held that if the sale was bona fide and passed title absolutely to the controlled corporation, even though the sale was made with the intent of reducing the tax liability of the vendor it fell within the provisions of the revenue act concerning the reckoning of gain or loss upon a closed transaction. So far as I am informed, the Treasury followed this rule in administering the various revenue acts for years after it was issued. The' first evidence of a change in its position was the refusal of the Commissioner of Internal Revenue to recognize losses resulting to taxpayers from a bona fide sale of bonds owned by them to a wholly owned corporation at the current market price.7 The-Board of Tax. Appeals sustained the Commissioner, but the Court of Appeals of the District of Columbia .reversed the Board in Jones v. Helvering, 71 F. 2d 214. The decision was rendered April 23, 1934. The Commissioner sought certiorari which was denied October 8, 1934.8 The same result has been reached by three other .Circuit Courts of Appeal.9- The Board of Tax Appeals followed these decisions.10 In the *485.meantime the Circuit Courts of Appeal nad decided numerous cases which are, in principle, indistinguish-. able.11

This court having denied certiorari in Jones v. Helver-ing, supra, decided Gregory v. Helvering, 293 U. S. 465, in the following January. It cited the Jones case with approval, at p. 469, saying: “. . . The legal right of a taxpayer to decrease the amount of what otherwise would be his taxes, or altogether avoid them, by means which the law permits, cannot be doubted.”

So well settled had the judicial interpretation become that the Treasury determined to recommend that Congress amend the statute.12 The result, was the adoption of § 24 (a) (6) of the Revenue Act of 1934.13 The committee reports disclose that Congress thought it necessary to change the statute in order to render nondeductible a loss claimed on a sale to a wholly owned or a controlled corporation.14 Subsequent hearings before the Joint Commission on Tax Evasion and Avoidance, 1937, p. 207, indicate the same understanding on the part of the Bureau of Internal Revenue and of Congress that the rule *486of law in effect prior to the adoption of the amendment in 1934 was changed by that legislation. The amendment lists among items not deductible the following:

“(6) Loss from sales or exchanges of property, directly or indirectly, (A) between members of a family, or (B) except in the case of distributions in liquidation, between an individual and a corporation in which such individual owns, directly or indirectly, more than 50 per centum in value of the outstanding stock. For the purpose of this paragraph — (C) an individual shall be considered as owning the stock owned, directly or indirectly, by his family; and (D) the family of an individual shall include only his brothers and sisters (whether by the whole or half blood), spouse, ancestors, and lineal descendants.”

Plainly, prior to 1934, taxpayers were justified in relying, first, upon the Treasury ruling on the subject and, secondly, upon the uniform decisions of the courts in claiming deductions for losses on sales to controlled corporations. After the passage of the amendment they were on notice that this was no longer permissible.

I turn then to the situation here presented. The claims of this taxpayer, as I have said, had been sustained for prior years by the Board of Tax Appeals.15 The Congress had enacted that subsequent to 1934 the taxpayer coijld not claim such losses. Notwithstanding the earlier decisions of the respondent’s case and those of other taxpayers against the Government’s present contention, the Commissioner of Internal Revenue, after the adoption of the Act of 1934, namely on March 11, 1935, served a notice of deficiency upon the respondent respecting losses . claimed in his return for the year 1932 on sales to Innis-fail. Thus the Treasury repudiated the position it had taken in asking that the law be amended to cover cases of this kind; reversed its position in acquiescing in the *487adjudication of the respondent’s tax liability for earlier years and sought, now that it had obtained an amendment of the law operating prospectively, to reach back into sundry unclosed ones, — this one amongst others, — and to attempt to obtain decisions reversing the settled course of decision. I think this pourt should not lend its aid to the effort.

I am of opinion that where taxpayers have relied upon a long unvarying series of decisions construing and applying a statute, the only appropriate method to change, the rights of the taxpayers is to go to Congress for legislation. In. my view, the resort to Congress, on the one hand, for amendment, and the appeal to the courts, on the other, for a reversal of construction, which, if successful, will operate unjustly and retroactively upon those who have acted in reliance upon oft-reiterated judicial decisions, are wholly inconsistent.

. I am of opinion that the courts should not disappoint the well-founded expectation of citizens that, until Congress speaks to the contrary, they may, with confidence, rely upon the uniform judicial interpretation of a statute. The action taken in this case seems to me to make it impossible for a citizen safely to conduct his affairs in reliance upon any settled body of court decisions.

delivered the opinion of the Court. \

Certiorari was allowed1 from the judgment of the Circuit Court of Appeals for the Second Circuit2 on account of an asserted conflict between the decision below and that *474of the Circuit Court of Appeals for the Seventh Circuit in Commissioner v. Griffiths.3

The issue considered here is whether a taxpayer under the circumstances of this case is entitled to deduct a loss arising from the sale of securities to a corporation wholly owned by the taxpayer. The statute involved is § 23 (e) of the Revenue Act of 1932.4

The Innisfail Corporation was wholly owned by the taxpayer, Mr. Smith. It was organized in 1926 under the laws of New Jersey. The officers and directors of the corporation were subordinates of the taxpayer. Its transactions were carried on under his direction and were restricted largely to operations in buying securities from or selling them to the taxpayer. While its accounts were kept completely separate from those of the taxpayer, there is no doubt that Innisfail was his corporate self. As dealings by a corporation offered opportunities for income and estate tax savings, Innisfail was created to gain these advantages for its stockholder. One of its first acts was to take over an option belonging to the taxpayer for the acquisition by exchange of a block of Chrysler common stock. Through mutual transactions in buying and selling securities, and receiving dividends, the balance of accounts between Innisfail and the taxpayer resulted, on December 29, 1932, in an indebtedness from him to Innis-*475fail of nearly $70,000. On that date, as a partial payment on this indebtedness, a number of shares of stock were sold to the corporation by the taxpayer at market. The securities sold had cost the taxpayer more than the price charged to the corporation, and in carrying out the transaction the taxpayer had in mind the tax consequences to himself.

In computing his net taxable income for 1932, the taxpayer deducted as a loss the difference between the cost of these securities and their sale price to his wholly owned corporation. The Commissioner of Internal Revenue ruled against the claim, whereupon respondent paid the tax and brought this suit for refund in the United States District Court for the. Southern District of New York. The case was tried before a jury and the verdict was adverse to the taxpáyer’s claim that the purported sales of these securities to Innisfail marked the realization of loss on their purchase. On appeal the judgment was reversed and the case remanded to the Distrio- Court for a new trial. It was the opinion of the Court of Appeals that the facts as detailed above, as a matter of law, established the transfer of the securities to Innisfail as an event determining loss.

Under § 23 (e) deductions are permitted for losses “sustained during the taxable year.” The loss is sustained when-realized by a completed transaction determining its amount.5 In this case the jury was instructed to find' whether these sales by the taxpayer to Innisfail were actual transfers of property “put of Mr. Smith and into something that existed separata and apart from him” or whether they were to be regarded as simply “a transfer by Mr. Smith’s left hand, being his’individual hand, into his right hand, being his corporate hand, so that in truth and fact there was no transfer at all.” The jury agreed the latter situation existed. There was sufficient evidence *476of the taxpayer’s continued domination and control of the securities, through stock ownership in the Innisfail Corporation, to support this verdict, even though ownership in the securities had passed to the corporation in which the taxpayer was the sole stockholder. Indeed this domination and control is so obvious in a wholly owned corporation as to require a peremptory instruction that no loss in the statutory sense could occur upon a sale by a taxpayer to such an entity.

It is clear an actual corporation existed. Numerous transactions were carried on by it over a period of years. It paid taxes, state and national, franchise and income. But the existence of an actual corporation is only one incident necessary to complete an actual sale to it under • the revenue act. Title, we shall assume, passed to Innis-fail but the taxpayer retained the control. Through the corporate forms he might manipulate as he chose the exer-. cise of shareholder’s • rights in the various corporations, issuers of the securities, and command the disposition of the securities themselves. There is not enough of substance in such a sale finally to determine a loss.

The Government urges that the principle underlying Gregory v. Helvering6 finds expression in the rule calling for a realistic approach to tax situations. As so broad and unchallenged a principle furnishes only a general direction, it is of little value in the solution of tax problems. If, on the other hand, the Gregory case is viewed as a precedent for the disregard of ,a transfer of assets without a business purpose but solely to reduce tax liability, it gives support to the natural conclusion that transactions, which do not vary control or change the flow of economic benefits, are to be dismissed from consideration. There is no illusion about the payment of a tax exaction. Each tax, according'to a legislative plan, raises funds to carry *477on government. The purpose here is to tax earnings and profits less expenses and losses. If one or the other factor in any calculation is unreal, it distorts the liability of the particular taxpayer to the detriment or advantage of the entire tax-paying group.7

The taxpayer cites Burnet v. Commonwealth Improvement Company8 as a precedent for treating the taxpayer and his solely owned corporation as separate entities. In that case the corporation sold stock to the sole stockholder, the Estate of P. A. B. Widener. The transaction showed a book profit and the corporation sought a ruling that a sale to its sole stockholder could not result in a taxable profit. This Court concluded otherwise and held the identity of corporation and taxpayer distinct for purposes of taxation.9 In the Commonwealth Improvement Company case, the taxpayer, for reasons satisfactory to itself voluntarily had chosen to employ the corporation in its operations. A taxpayer is free to adopt such organization for his affairs' as he may choose and having elected to do some business as a corporation, he must accept the tax disadvantages.10

On the other hand, the Government may not be required to acquiesce in the taxpayer’s election of that form for doing business which is most advantageous to him. The Government may look at actualities and upon determination that the form employed for doing business or carrying out the challenged tax event is unreal or a sham may sustain or disregard the effect of the fiction as best serves the purposes of the tax statute. To hold otherwise would permit the schemes of taxpayers to supersede legislation in the determination of the time and *478manner of taxation. It is command of income and its benefits which marks the real owner of property.11

Such a conclusion, urges the respondent, is inconsistent with the prior interpretations of the income tax laws and consequently unfair to him. He points to the decisions of four courts of appeals which have held losses determined by sales to controlled corporations allowable12 and further calls attention to the fact that the Board of Tax Appeals has consistently reached the same conclusion.13 But this judicial and administrative construction has no significance for the respondent. The' Bureau of Internal Revenue has insistently urged since February 18, 1930, the date of the Board of Tax Appeals' decision in Jones v. Helvering,14 that a transfer from a taxpayer to a controlled corporation was ineffective to close a transaction for the determination of loss. Every case cited by respondent in the courts of appeals and before the Board of Tax Appeals found the Government supporting that contention. The Board’s ruling in the Jones case was *479standing unreversed at the time of the transaction here involved, December 29,1932. It was only after the transactions here involved and after the reversal of the Board in the Jones case on April 23, 1934, or this Court’s refusal of certiorari on October 8, 1934, that the Board of Tax Appeals and the courts of appeals, over Government protests, ruled in line with the opinion of the. Court of Appeals of the District of Columbia in the Jones case. If the Bureau’s stand in the Jones case represented a change in administrative practice, there can be no doubt that the change operated validly at least from 1930 on.15 After the Jones defeat the Government sought relief in Congress and after the judgment in Commissioner v. Griffiths, supra, certiorari here on a conflict in principle between circuits. Certainly there was no acquiescence by the Government which would justify the taxpayer in relying upon prior interpretations of the law.16.

Respondent makes the further point that the passage of § 24(a)(6) of the Revenue Act of 193417 which explicitly forbids any deduction for losses determined by sales to corporations controlled by the taxpayer is convincing proof that the law was formerly otherwise. This *480does not follow. At most it is evidence that a later Congress construed the 1932 Act to recognize separable taxable identities between the taxpayer and his wholly owned corporation. As the new provision goes much farther than the former decisions in disregarding transfers between members of the family it may well have been passed to extend as well as clarify the existing rule. The suggestion is not sufficiently persuasive to give vitality to a futile transfer.

The taxpayer has preserved two objections to the district judge’s rulings on the evidence. He claims that evidence as to transactions between the taxpayer and the corporation which took place prior to the sale here involved was remote and highly prejudicial. We thir1' it apparent that this evidence was entirely relevant to the present issue; the history of the taxpayer’s relations with the corporation shed considerable light on the actual effect of the sale in question. The second contention is that the district judge charged the jury to give less effect to the book entries of Smith and the corporation than they were entitled to under the applicable book entry statute.18 The alleged departure from the statute has but dubious support in the record, resting on a single statement of the judge lifted from its context as part of an extended colloquy with counsel. In the circumstances there is no merit in the claim of prejudice to the taxpayer.

The judgment of the Circuit Court of Appeals is reversed and that of the District Court affirmed.

Reversed.