(dissenting).
I respectfully dissent.
The Tax Court concedes, in its reasoning in this case, that the expirations are intangible capital assets1 capable of depreciation under § 167(a) of the Internal Revenue Code of 1954 and accompanying regulations. The Tax Court, however, disallowed all deductions claimed by the taxpayer because of two findings: first, the expirations were inextricably linked with elements of goodwill such that they had no determinable value in themselves; and second, the taxpayer had not proved that the expira-tions had a limited useful life, determinable with “reasonable accuracy.” An examination of the record indicates to me that these findings were clearly erroneous. Also, it appears that, in part, they were based on erroneous conclusions of law.
As to the first finding, the Tax Court listed four elements that had been purchased along with the expirations, and to which it found the expirations were inextricably linked: (1) the 2400 noncommercial accounts, (2) the covenant not to compete, (3) the employment of former shareholders and employees of Stokes, together with their use in soliciting accounts, and (4) the use of the Stokes name for nine months on Marsh & McLennan stationery, indicating that Stokes was the predecessor of Marsh & McLennan.
There is evidence that these elements existed, but there is little, if any, evidence that any of the $69,500. premium involved in this case was paid for the purchase of these elements. Marsh & McLennan was interested in purchasing the expirations; that it also procured other elements to make its use of the ex-pirations more effective does not weaken this conclusion. The Tax Court relies upon the testimony of the Chairman of Marsh & McLennan. He testified that Marsh & McLennan felt that the purchase of Stokes was the best way to procure the commercial accounts involved here (71a), but he said that this result was possible only because of the purchase of the expirations: “Most companies do buy door openers [the expirations] and this is what we bought in this ease. We did not buy personnel to any extent * * *.” (71a-72a).2
Mr. Regan, the manager of Marsh & McLennan’s New York office, was the person most directly responsible for the sale. He made clear that Marsh & Mc-Lennan, in order to effectuate its interest in acquiring the commercial accounts (80a, 102a, 106a), was interested in purchasing the expirations (81a), and would have purchased the expirations alone if that were possible (107a).
“Q. What are you most interested in acquiring ?
“A. Well, what I said before, what you call expirations * * *.
«•**■* * *
“Q. Then you are buying the expira-tions plus the net worth of the business?
“A. Yes.” (81a)
This testimony was corroborated by another Marsh & McLennan vice-president and three independent experts, who all testified to the effect that large commercial buyers are without sentiment, that they purchase from the broker who can give them the best coverage for the best price — and that the primary method of convincing a prospective buyer that your organization can give him the best service is by offering a policy fitted to his needs. The only way to offer such a policy is to have the information contained in the expirations records. The vice-president testified that personal contact is irrelevant, because large commercial “insurance buyers are without *672sentiment.” (142a-143a) The first expert said that it is “the usual custom” of insurance buyers to listen to the company with the expirations (153a), although these buyers also look to the present broker who also knows their needs (152a). These companies are interested in service, not personal contacts. (156a-57a) The second expert testified that the main assets a company is buying in a purchase such as this are the expirations (158a), “that is the only thing of value.” (159a; see 161a) The third' expert also testified that a buyer seeks expirations, not personnel or the good will of a firm. (167a, 169a) The one Government witness, who had no experience in the specialized field of tailoring commercial coverage to the individual needs of large corporations (221a), did not contradict this testimony.
Not only was there little, if any, evidence to indicate that the $69,500., allocated by Marsh & McLennan to the six commercial expirations it acquired (see 257a), was for the purchase of anything other than the expirations, but also there is positive evidence or law undercutting the conclusion that any of the four elements listed by the Tax Court as “goodwill” were purchased or could be considered goodwill.
„ There is no evidence that any of the $69,500. premium (as opposed to the entire purchase price) was allocated to the 2400 non-commercial accounts that came with the Stokes purchase. Marsh & McLennan specialized in large commercial accounts; when it did take personal accounts of the same type as the 2400 cited by the Tax Court, it did so as a favor to some large account:
“Our business is commercial lines. We do have a small amount of personal insurance and the persons that we have this personal insurance with, they are usually in some sound business * * *. But our business is more large commercial accounts and certainly I, personally, abhor building anything on a personal account. * * * The commission would develop, it would cost us about $500 to help handle it.” (46a-47a)
Mr. Regan affirmatively testified that Marsh & McLennan was not interested in Stokes’ personal accounts; the only reason these accounts were purchased was because it was the only way to acquire the expirations, “this is part of the package we had to buy.” (84a-85a) These affirmative disclaimers of interest in the personal accounts were unre-butted, and, indeed, the Tax Court found that Marsh & McLennan “placed a value upon the 5 or 6 expirations which it was primarily interested in obtaining and added such value to such net worth of Stokes. The amount of $69,550.78 represents such calculated value * * (256a-57a) (emphasis added).3
*673The covenant not to compete was obtained to insure that the value of the exclusive use of the expirations to Marsh & McLennan would not be diminished 4 Where a covenant not to compete is procured to protect another element involved in the purchase, it is non-severable from that other element. See, e. g., Estate of Masquelette v. Commissioner of Internal Revenue, 239 F.2d 322 (5th Cir. 1956).5
There is no evidence that Marsh & McLennan was interested particularly in the personal services of former Stokes employees. The contract provision that former Stokes employees be employed was a condition of sale imposed by Stokes, not Marsh & McLennan. As the Tax Court found:
“[The stockholders of Stokes] were interested in selling the business because they felt that Stokes was too small and localized an organization, * * * and because they felt that the sale to another organization would provide Stokes’ staff, which included themselves, with continued employment.” (255a)
Because of this purpose, Marsh & Mc-Lennan was obligated to hire Stokes’ personnel, but “[n]one of Stokes’ personnel were bound by contract to remain employed by [Marsh & McLennan] for any specified time.”6 (258a)
*674The fourth factor was the use of the Stokes name for a short time after the sale to identify Marsh & McLennan as Stokes’ successor. From the testimony of the expert witnesses, it is clear that no goodwill existed in the name of Stokes’ firm as to the six commercial accounts here involved. Furthermore, courts have consistently held that where the name of the seller is used only to identify the purchaser as its successor, even where the firm name may be an asset of goodwill, no purchase of goodwill is contemplated. E. g., Manhattan Co. of Virginia, Inc., 50 T.C. 78, 81-82 (1968); Savings Assurance Agency, Inc., T.C. Memo 1963-52, 22 T.C.M. 200, 201 (1963).
Marsh & McLennan did prove that the expirations had a limited life determinable with reasonable accuracy. It hired statistical experts who testified that the average life of a commercial account was 17 years. This evidence was uncon-tradicted, other than by the Government witness who, having no qualifications in the principles of accounting or statistics, was not qualified to testify on this point. The Tax Court disregarded this evidence on the theory that “such projection indicated that some of such accounts would continue on after the 17 year period for a long and indefinite life.” (264a) The majority affirms on the theory that, because the analysis was based on similar but not identical accounts, the Tax Court was free to disregard the evidence. Either of these theories would prevent most, if not all, depreciation deductions. By its essence, an average useful life must contain the lives of assets expiring after the average as well as before the average. And because a determination of useful life is a projection, it cannot be based on the lives of the assets sought to be depreciated; it must, by necessity, look to the lives of other assets. The Regulation requires only “reasonable accuracy,” not exactness. Average length of customer association, where it can be proved, has always been accepted in customer list cases. E. g., Commissioner of Internal Revenue v. Seaboard Fin. Co., 367 F.2d 646, 653 (9th Cir. 1968); Manhattan Co. of Virginia, Inc., 50 T.C. 78, 83, 93 (1963); Vaaler Ins., Inc. v. United States, 68-1 U.S.T.C. ¶ 9183 (D.N.D. 1968). That the evidence be statistical makes it no less reliable in determining the useful life with reasonable accuracy. See, e. g., Super Food Serv., Inc. v. Commissioner of Internal Revenue, 416 F.2d 1236 (7th Cir. 1969), distinguishing Commissioner v. Indiana Broadcasting Corp., 350 F.2d 580 (7th Cir. 1965); Rev.Rul. 68-545, 1968-42 I.R.B. 6.
I would reverse and remand the case to the Tax Court for further proceedings not inconsistent with this opinion and subject to the right of the Tax Court to hold a supplemental trial for production of expert or other evidence if it finds such a trial is necessary for determination of the factual issues. See Sink, “The Unused Power of a Federal Judge to Call His Own Expert Witness,” 29 So. Cal.L.Rev. 195 (1956); Rule 7-06 of Proposed Rules of Evidence for the United States District Courts (March 1969 Draft of Committee on Rules of Practice and Procedure of the Judicial Conference of the United States).
. See, e. g., Commissioner of Internal Revenue v. Killian, 314 F.2d 852 (5th Cir. 1963).
. It is noted that this witness was not directly involved in the purchase of Stokes (73a).
. The Tax Court’s reliance in part on the “indivisible asset” rule, as expounded by such casos as Boe v. Commissioner of Internal Revenue, 307 F.2d 339 (9th Cir. 1962), aff’g. 35 T.C. 720 (1961), was improper. As clarified in Commissioner of Internal Revenue, v. Seaboard Fin. Co., 367 F.2d 646, 652 (9th Cir. 1966), that rule was applied in Boa to prevent a taxpayer from claiming that a certain amount of the purchase price could be allocated as a basis for depre-ciable assets, where there was “simply bargaining between the parties as to what would be paid for the business as a whole.” The court continued: “We indicated in Boe, however, that the result would be different if * * * the individual contracts had been separately appraised.” Id. at 652-653. In the instant case, the taxpayer did value the expira-tions it was interested in acquiring, as the Tax Court found (256a-57a), and no part of the $69,500. thereby allocated could be held under the “indivisible asset” rule to apply to goodwill.
That the five or six contracts were not individually evaluated, as was the case in Seaboard Finance, supra, is of no significance in this regard. In Manhattan Co. of Virginia, Inc., 50 T.C. 78 (1968), a decision reviewed by the entire court (Atkins, J., dissenting), the taxpayer purchased two customer lists containing approximately 4500 names. Because the taxpayer did not evaluate the individual value of each name, the court disallowed its deduction for annual customer loss, taken by the taxpayer on the assumption that each name was of equal value. The *673“indivisible asset” rule prevented the taxpayer from claiming that it had bargained for each name on an individual basis. However, the court did allow deductions for depreciation of the value assigned to the list as a whole, holding that past experience of the taxpayer had shown that the average life of customers obtained through purchase of the list was five years. The indivisible asset rule does not, therefore, prevent depreciation deductions if customer information can be assigned a value separate from goodwill, as here, and where the average useful life of the information as a whole can be estimated by calculating the average length of customer association. Also see Western Mortgage Corp. v. United States, 308 F.Supp. 333 (C.D.Cal. 10/30/69).
. Although the Tax Court made no specific finding as to this covenant not to compete, it did note that “[ujsually a covenant not to compete is obtained from the selling brokerage firm to prevent the seller from using the insurance expiration information and to keep such information from competitors”. (253a).
. It is noted that the covenant itself is a capital asset capable of depreciation, not goodwill. E. g., John T. Fletcher, T.C. Memo 1965-273, 24 T.C.M. 1489 (1965) ; Merle P. Brooks, 36 T.C. 1128 (1961).
. The courts and the Commissioner have consistently held that the personal qualifications of employees of a seller, where the employees are not required by contract to continue their employment with the seller or the buyer, are not an element of goodwill. The landmark case in this area is D. K. MacDonald, 3 T.C. 720 (1944). In that case, the sole owner of an insurance brokerage business dissolved the corporation, distributing all its assets to himself for the purpose of continuing the business as a partnership. He claimed no capital gain, since the assets of the business were less than its total liabilities. The Commissioner assessed a deficiency, based on his determination that the business had approximately $100,000. in goodwill, because of the “personal ability, business acquaintanceship, and other individualistic qualities of D. K. MacDonald.” The Tax Court held that these personal qualifications did not represent marketable goodwill, and thus disallowed the deficiency:
“We find no authority which holds that an individual’s personal ability is part of the assets of a corporation by which he is employed where * * * the corporation does not have a right by contract or otherwise to the future services of that individual.” Id. at 727.
This opinion’s reasoning has been consistently followed. E. g., Donal A. Carty, 38 T.C. 46, 59, n. 11 (1962) ; Rev.Rul. 60-301, 1960-2 Cum.Bull. 15. If the seller does not have marketable goodwill because the success of the firm is based on the personal qualifications of its employees, any gain the seller receives must be in exchange for something other than goodwill. It would seem, therefore, that the buyer does not receive goodwill, as that asset is contemplated by Treasury Regulation § 1.167 (a)-3.