All the facts in this case were stipulated below, and were incorporated by the Tax Court in its detailed opinion reported at 1956, 25 T.C. 903. We shall therefore set forth only those facts essential to a discussion of the legal issues before us on appeal.

The petitioners, a group of affiliated companies, appeal from decisions of the Tax Court finding deficiencies in corporate income taxes for the calendar years 1948 and 1949 in the respective amounts of $79,497.05 and $78,123.69. The American Water Works Company, the principal petitioner (hereinafter referred to as "taxpayer"), was incorporated in 1936 as a holding company for the stock of many water utility companies. Only four of those companies, however, are involved in this appeal. They are the Texarkana Water Corporation ("Texarkana"), the City Water Company of Chattanooga ("Chattanooga"), the Greenwich Water System ("Greenwich"), and the Cohasset Water Company ("Cohasset").

The deficiencies resulted from a reduction by the Commissioner of the basis claimed by the petitioners to be the basis of stock held by the parent corporation in its subsidiaries. This reduction produced a corresponding increase in capital gains realized in 1948 and 1949 from the sale of the stock of two of the subsidiaries (Texarkana and Cohasset), and from capital distributions in excess of the basis of stock held in a third subsidiary (Chattanooga).

1. Texarkana. In 1947 taxpayer acquired all the stock of Texarkana in a nontaxable transfer. It held that stock at the same basis at which the stock was held by its transferor. The transferor had acquired 4,000 shares of Texarkana in 1926 and 1,200 newly issued shares in 1941. In 1948 taxpayer sold its entire interest in Texarkana.

Taxpayer and its transferor filed consolidated income tax returns that included Texarkana for the years 1926 through 1933 and for the years 1942 through 1948, but not for the years 1934 through 1941. During both the years for which consolidated returns were filed and the years for which consolidated returns were not filed, taxpayer and its transferor received substantial capital distributions from Texarkana. A total of $168,714.47 was distributed out of capital on the original 4,000 shares, and $28,322.27 on the 1,200 shares issued in 1941. The Commissioner applied the distributions of $168,714.47 to the unadjusted basis ($84,558.54) of the 4,000 shares, thereby reducing that basis to zero. He applied the distributions of $28,322.27 to the unadjusted basis ($115,755.55) of the 1,200 shares, reducing such basis in that amount.

In 1929 to 1933, for which years the taxpayer's transferor had filed consolidated returns that included Texarkana, the latter company sustained net losses totaling $36,997.12. These losses had been allowed as deductions in arriving at the net income of the taxpayer's transferor and its affiliated companies for those years. Since the basis of the original 4,000 shares had already been reduced to zero, the Commissioner reduced the basis of the 1,200 shares issued in 1941 by $36,997.12, the full amount of the losses sustained by Texarkana from 1929 to 1933.

2. Chattanooga. In 1947 taxpayer acquired all the stock of Chattanooga in a nontaxable transfer. That stock had the same basis to taxpayer as it had to the transferor. The unadjusted basis of this stock to taxpayer was $1,102,497.50, of which $641,452.50 represented the unadjusted basis of 15,000 shares acquired by the taxpayer's transferor in 1914, and $461,045 represented the unadjusted basis of the remaining 5,000 shares acquired by the transferor after 1914. Chattanooga was not a member of the affiliated group for which taxpayer filed consolidated returns for the years 1948 and 1949, but was a member of the affiliated group for which taxpayer's transferor filed consolidated returns for 1933 and prior years.

Prior to 1948, a total of $690,042.73 had been distributed out of capital by Chattanooga to taxpayer's transferor on the 15,000 shares both in years when consolidated returns were filed and in years when separate returns were filed. The Commissioner applied these distributions against taxpayer's basis ($641,452.50) for the 15,000 shares, reducing that basis to zero. Distributions to taxpayer of $19,622.59 in 1948 and $25,265.45 in 1949 on the 15,000 shares, being in excess of the basis of those shares, were determined by the Commissioner to be taxable as capital gains to taxpayer in those years.

3. Greenwich and Cohasset. Greenwich and Cohasset were members of the affiliated group for which consolidated returns were filed by the petitioner or its transferor in the years for which such returns were permitted. In 1929 Greenwich acquired all the stock of Cohasset, which it sold in 1949. In the years prior to 1949 for which consolidated returns including Greenwich and Cohasset were filed, Greenwich received distributions from Cohasset in the amount of $17,686.27. In the years when no consolidated returns were filed, Greenwich received capital distributions of $10,450.72 from Cohasset. The Commissioner applied the total amount of these distributions ($28,136.99) in reduction of the basis of the stock sold by Greenwich in 1949, thereby increasing in an equal amount the capital gain realized on that sale.

The determinations of the Commissioner raised two basic questions, which were considered by the Tax Court and are now before us on appeal:

(1) whether the basis of stock held by the petitioner in affiliated corporations should be reduced by the amounts of capital distributions made by the issuing corporations to the petitioner or its transferor in years when consolidated returns were filed;

(2) whether the basis of stock held by the petitioner in an affiliated corporation should be reduced by the amount of net operating losses sustained by the issuing corporation and availed of in years when consolidated returns that included the two corporations were filed, but before the particular shares of stock in question were issued for new capital.

The Tax Court answered both questions in the affirmative, thereby upholding the Commissioner's determinations in all respects.

The petitioner on appeal does not dispute the determination that the distributions from capital in years for which no consolidated returns were filed by the stockholder corporation and the issuing corporation should be applied in reduction of basis. Those distributions result in reducing the basis of the 4,000 shares of Texarkana to zero, and thus it is immaterial whether the basis of those shares should be reduced on account of distributions from capital in consolidated return years. Since the total distributions from capital on the 5,000 shares of Chattanooga stock in the years here in issue and prior years do not equal the basis of that stock, it is likewise immaterial whether distributions from capital on those shares in consolidated return years should be applied in reduction of basis. The petitioner contends, however, that capital distributions to affiliated companies in consolidated return years should not be applied in reduction of basis, and therefore the gain on the sale of the 1,200 shares of Texarkana and on the sale of Cohasset stock should not be increased by the amount of those distributions. Nor, according to the petitioner, should distributions from capital in the years 1948 and 1949 on the 15,000 shares of Chattanooga result in taxable gain.

The petitioner's second main contention is that the net losses of $36,997.12 sustained by Texarkana in the years 1929 through 1933 are applicable only in reduction of the basis of the 4,000 shares held by petitioner's transferor during those years, and therefore should not be applied to reduce the basis of the 1,200 shares issued in 1941 and acquired at that time by taxpayer's transferor for new capital.

Under the 1939 Code, an adjustment of basis for transactions occurring during a period for which a consolidated return is filed shall be made "in accordance with regulations" issued by the Commissioner. Int.Rev.Code of 1939, §§ 113(a) (11), 141(b), 26 U.S.C.A. §§ 113 (a) (11), 141(b). Hence the resolution of the two basic issues before us requires a construction of Treasury Regulations 104 which were promulgated pursuant to this statutory authority and have, in effect, the force of law. Cf. American Trans-Ocean Navigation Corp. v. Commissioner, 2 Cir., 1956, 229 F.2d 97.

* These regulations do not contain a provision specifically directed at the first question raised by this litigation. Therefore the Tax Court relied primarily on the following general language of § 23.34 (c) of Regulations 104 to support its conclusion that the basis of stock in an affiliated corporation was to be reduced by the amount of capital distributions made by that corporation during a period for which a consolidated return was filed:

"If the sale [of stock] is made within a period during which the selling corporation is a member of the affiliated group * * * (1) The aggregate bases of all shares of stock of the issuing corporation * * * shall be * * * adjusted in accordance with the Code * *" (emphasis added).

The Tax Court construed this provision to mean that the general sections of the 1939 Code requiring a reduction of basis for a tax-free distribution of capital by a corporation to its shareholders, §§ 113 (b) (1) (D)1 and 115 (d),2 are applicable to a distribution to an affiliated corporation in a consolidated return year.

Taxpayer, on the other hand, contends (1) that if an adjustment for distributions from capital during consolidated return years was intended by the regulations, they would have contained an express provision to that effect; (2) that other sections of Regulations 104 indicate that no such adjustment is to be made.

Taxpayer's first argument warrants little discussion, for the general pattern of the Treasury regulations embodies frequent cross-references to the Code, as well as detailed provisions for tax treatment. Thus the absence of express language requiring an adjustment in stock basis for capital distributions in consolidated return years is unpersuasive if the purpose of § 23.34(c) was to assimilate these distributions to ordinary capital distributions and thereby bring them within the coverage of §§ 113(b) (1) (D) and 115(d) of the 1939 Code.

Taxpayer's second argument, however, merits a more detailed analysis. It is contended that §§ 23.31(d) (2) and 23.38 (a) of Regulations 104, rather than § 23.34(c), govern the treatment of capital distributions during consolidated return years. Insofar as here relevant, § 23.31 (d) (2) reads as follows:

"For a taxable year for which a consolidated return is made * * * [various types of income] * * * shall be computed and determined in the case of each affiliated corporation in the same manner and subject to the same conditions as if a separate return were filed, except * * * (iii) Capital gains and losses, short-term capital gains and losses, long-term capital gains and losses, and the additional capital loss deduction authorized by section 204(c) (5) [26 U.S.C.A. § 204(c) (5)] shall be determined without regard to

"(A) Gains or losses arising in inter-company transactions * * *"

It is clear to us, however, that this section concerns the treatment of "intercompany transactions" only in the computation of the net income of an affiliated group of corporations during a period for which consolidated returns are filed3. Thus taxpayer fails to distinguish between (1) the treatment of capital distributions from the issuing corporation to the stockholder corporation for the purpose of computing their consolidated net income in the year of the intercompany distribution, and (2) the treatment of these same capital distributions for the purpose of determining the basis of the stock of the issuing corporation in the hands of the stockholder corporation when that stock is subsequently sold to an outsider. Indeed, as the Commissioner points out, it is because no gain is recognized for tax purposes on intercompany capital distributions, where the issuing and stockholder corporation elect to file consolidated returns, that § 23.34(c) of Regulations 104 prescribes an equivalent adjustment downward in the basis of the stock "in accordance with the Code," i. e., in accordance with §§ 113(b) (1) (D) and 115(d). A contrary construction of the regulations would accord a taxpayer an unwarranted tax benefit from its election to file consolidated rather than separate returns, if that taxpayer were able to drain off the capital from its subsidiary corporations through capital distributions without a corresponding reduction in the basis of the stock held by it in those corporations. The parent corporation, to take an extreme example, might withdraw its entire original investment from a successful subsidiary at no tax cost, and then sell the shares of that subsidiary to an outsider at the price that the parent originally paid for the stock. Under the petitioner's construction of the regulations, the parent would escape any liability for a capital gains tax because the parent's basis for the stock would be equal to the amount received from the outsider.4 It is clear that Congress intended no such "windfall" when it enacted provisions permitting corporations to file consolidated returns. To the contrary, the intent of Congress was that the income of affiliated corporations should be computed in such a manner as clearly to reflect that income and to prevent avoidance of tax liability. See S.Rep. No. 960, 70th Cong., 1st Sess. 13-15 (1928). As Mr. Justice Stone wrote in reference to revenue acts prior to those here applicable. "Primarily, the consolidated return was to preclude reduction of the total tax payable by the business, viewed as a unit, by redistribution of income or capital among the component corporations by means of intercompany transactions." Burnet v. Aluminum Goods Mfg. Co., 1933, 287 U.S. 544, 547, 53 S. Ct. 227, 229, 77 L. Ed. 484.

Taxpayer further argues that § 23.38 ("Basis of Property") also precludes an adjustment for corporate distributions made to an affiliated company during a consolidated return period. That section, insofar as here relevant, reads:

"(a) General Rule.

"Subject to the provisions of paragraphs (b) and (c) and except as otherwise provided in sections 23.34 and 23.39, the basis during a consolidated return period for determining the gain or loss from the sale or other disposition of property * * shall be determined and adjusted in the same manner as if the corporations were not affiliated (see sections 111 to 115, inclusive) * * *

"(b) Intercompany Transactions.

"The basis prescribed in paragraph (a) shall not be affected by reason of a transfer during a consolidated return period * * * from a member of the affiliated group to another member of such group." (Emphasis added.)

But, as we indicated earlier, an adjustment for capital distributions upon a subsequent sale of stock in an affiliated company is provided for in § 23.34, and is therefore expressly excepted from the coverage of § 23.38(b).

We therefore affirm the Tax Court on the first main question raised by this appeal, and hold that the basis of stock held by the taxpayer in affiliated corporations must be reduced by the amounts of capital distributions made by the issuing corporations to the taxpayer or its transferor in years when consolidated returns were filed.

In the alternative, taxpayer argues that the regulations dealing with basis adjustment under pre-1932 revenue acts should not be construed as authorizing adjustment for capital distributions made during any consolidated return period before 1932. In support of this contention, taxpayer points out that the last sentence of § 113(a) (11) of the 1939 Code was added by the Revenue Code of 1932,5 and argues that the revenue acts in force for years prior to 1932 did not require an adjustment to basis for transactions that took place during consolidated return periods. But the 1932 addition of the pertinent sentence was not designed to curtail the Commissioner's authority to require a reduction of basis for transactions occurring in prior years. Rather, its narrow purpose was to make certain that "corporations which were affiliated and filed consolidated returns for any one or more of the years 1929, 1930, or 1931 can not, by filing separate returns in 1932, avoid the adjustments required by the regulations in force at the time the consolidated returns were filed." S.Rep. No. 665, 72d Cong., 1st Sess. 28 (1932).

We therefore reject taxpayer's alternative argument, and hold that adjustments in basis must be made for capital distributions made in consolidated return periods before 1932, as well as after that date.

The petitioner also claims error in the Tax Court's resolution of the second major issue in this case. It was decided below that the basis of stock held by the taxpayer in an affiliated corporation should be reduced by the amount of net operating losses sustained by the issuing corporation and availed of in years when consolidated returns were filed, but before the issuance and acquisition, in return for new capital, of the particular shares of stock in question. Two judges dissented from this conclusion. The relevant language of Regulations 104 governing this question appears in § 23.34 (c) (2), and is as follows:

"(2) From the combined aggregate bases as determined in paragraph (1), there shall be deducted the sum of the losses of such issuing corporation sustained during taxable years for which consolidated income tax returns were made or were required * * * after such corporation became a member of the affiliated group and prior to the sale of the stock to the extent that such losses could not have been availed of by such corporation as a net loss or net operating loss * * *

"(3) The sum of the aggregate bases of all shares of stock, after making the deduction under paragraph (2), shall then be apportioned among the members of the affiliated group which hold stock of the issuing corporation, by allocating to each such member that proportion of the sum of the aggregate bases so reduced which the aggregate basis of the stock in the issuing corporation held by such member bears to the sum of the aggregate bases;

"(4) The aggregate basis as determined under paragraph (3) for each member of the affiliated group shall then be equitably apportioned among the several classes of stock of the issuing corporation held by such member according to the circumstances of the case — ordinarily by allocating to each class of such stock that proportion of the aggregate basis which the basis of each class of such stock held by it at the time of the sale is to the sum of the bases of the several classes of such stock held by it;"

The majority, in support of its conclusion, emphasized that the regulations contain no provision segregating stock according to the time of its acquisition. Moreover, § 23.34(c) (4) expressly provides that the aggregate basis is to be allocated to each class of stock held by a member of the affiliated group "at the time of the sale." Thus the majority reasoned that the time of acquisition of stock was irrelevant, and that losses availed of in consolidated return years which could not be applied in reduction of the basis of stock held during those years were properly applicable to reduce the basis of stock subsequently acquired in the issuing corporation.

With this conclusion we cannot agree, and we hold that these losses cannot be applied to reduce the basis of stock that was issued in return for fresh capital subsequent to the years in which the losses were sustained and availed of by the affiliated group.

The Tax Court majority, in stressing the phrase "at the time of the sale," ignored other language in § 23.34 (c) (4), such as the words "ordinarily," "equitably," and "according to the circumstances of the case," which lead us to believe that a flexible application of these regulations was intended. The purpose of § 23.34(c) (2-4) is clear: to prevent a corporation owning stock in an affiliated company from unevenly apportioning among its existing shares of stock in that company a loss that is sustained by the issuing company and is treated as a deduction in the tax return of the family group. To permit an uneven apportionment of the loss would result in a tax deduction for the year of the loss without a corresponding reduction in the basis of any shares of stock that were held at the time of the loss but were subsequently sold to an outsider. Thus an affiliated shareholder corporation cannot manipulate various lots of stock held by it at the time the loss was sustained so that upon a subsequent sale of certain shares to an outsider the stock sold would have a higher proportionate basis than the stock retained.

But, as the minority pointed out, "the limits of the tax effects of those losses on the basis of the stock of the subsidiary which the parent then owned * * * were fixed" as of the time the losses were availed of in the consolidated returns for those years. Theoretically, the adjustment in stock basis could have taken place at the end of the year for which the losses were taken as deductions, although the adjustment became significant for tax purposes at the time the stock was sold to an outsider.

Under the Tax Court approach, as pointed out by the minority, "the investment of new money is penalized since the investor is required to wipe out past tax benefits which are wholly unrelated to the investment in the new stock." We do not believe that the regulations were intended to deter new financing for any affiliated corporation that had previously sustained losses in excess of the basis of its outstanding stock at the time of those losses. Nor are we persuaded that our holding will accord preferential treatment to affiliated corporations by permitting them to retain allegedly undeserved tax benefits. The tax benefit realized by the group in prior years and here retained by virtue of our refusal to apply the excess undeducted value of former losses to the basis of subsequently issued stock would appear to be an incident of the rule that the tax basis of stock can never be adjusted below zero, or be a minus figure.

Affirmed on the first issue. Reversed and remanded to the Tax Court for proceedings in accordance with this opinion on the second issue.