The corporate Taxpayer appeals from an adverse judgment after a nonjury trial of its tax refund suit. Taxpayer is both an insured and a stockholder of a credit insurance company. The immediate question is whether distributions to the Taxpayer by the insurance company from so-called earnings on insurance written for the Taxpayer are entitled to the 85% dividends received credit under § 26(b) (1) of the 1939 Code. 26 U.S.C.A. § 26(b) (1). If they are, then a very substantial tax benefit results ostensibly from the unique form in which the transaction is cast. As one may lawfully yearn for and achieve tax savings and may therefor utilize and exploit available legitimate and beneficial means, the question becomes really the troublesome one whether the transaction passes muster under the accepted substance over form tax principle.1

At the outset it is well to point out that the case has been tried and the findings of fact by the District Court are binding unless clearly erroneous, F.R.Civ. P. 52(a), 28 U.S.C.A. The ultimate issue determined by the Trial Court and presented here for review is: were these really dividends representing corporate earnings? Or were they really a mere return of premiums?

In a capsule this is what brought about the problem. Taxpayer is an installment credit finance company which buys installment credit contracts from used car dealers at a discount. To meet competition from other similar finance concerns, Taxpayer obtained a group2 life, health and accident policy from Consolidated General Life Insurance Company insuring the initial borrowers for Taxpayer's account. It paid Consolidated its standard premium. Taxpayer simultaneously became an owner of special classified preferred stock of Consolidated. Taxpayer was the sole owner of Class "A" preferred stock. No other person could acquire Class "A" preferred stock. The preferred stock, implemented by a contract, called for payment of quarterly fixed dividends out of earnings. Of great importance here, it also provided for quarterly special dividends in the amount of the profit "earned upon the Class `A' insurance business" of the particular stockholder. This was a way of providing that credit insurance written for the particular classified preferred stockholder would be separately treated in ascertaining profits. The "earnings" or "profits" were to be determined by deducting from the total premiums paid by the classified preferred stockholder the sum of (a) losses actually paid on such policy and (b) 10% of the premium for administrative overhead and management. For the fiscal tax year involved, the premium payments to Consolidated totaled $62,358.99. After deducting (a) losses actually incurred aggregating but $3,734.37 and (b) the 10% overhead charge, Consolidated paid as "dividends" $54,725 to Taxpayer.3

The Taxpayer deducted the full total premiums paid ($62,358.99) as ordinary and necessary business expense on the ground that this had been paid without refund or subsequent reduction. It then took the 85% dividends received credit on the total of $54,725. The result was that Taxpayer got a tax benefit of approximately $26,000 because the savings in insurance costs came through the form of corporate dividends.4

The Commissioner disallowed the dividend received credit.5 The District Court denied Taxpayer relief on the alternative grounds that the transaction was a sham and, if not, the payments by Consolidated were not out of corporate earnings and hence did not constitute dividends under § 26(b) (1). 26 U.S. C.A. § 26(b) (1) [IRC 1939]. Oral argument was a convincing demonstration that the second reason was but a part of the first. For if the transaction is really what it purports to be, then there were sufficient "earnings" out of which distributions could be, and were, made so as to constitute dividends. If not, the payment was not from "earnings" and the distribution was not in legal effect a dividend.

In assaying the record to test the Trial Court's decision on this one critical issue, it is not our function to sift and weigh and here recite the evidence which the Court heard. It is sufficient that we highlight some of those factors which, within his fact finding province, justified the inferences reached.

Of crucial importance was the fact that the arrangement was designed and executed as a package deal. The impetus seems to have come largely from Consolidated. It was successful in selling this arrangement to nine other finance companies who became holders of like preferred stock for Classes "B" through "J" with similar contracts for quarterly payment of profits on the insurance written for each. There was virtually no risk to Taxpayer as a stockholder since the principal stockholders and the active management of Consolidated, a closely held Texas corporation, bound themselves (a) to continue their ownership and management and (b) to repurchase the stock at par.6

Just as there was little, if any, real risk in the "investment" of $25,000 for the Class "A" preferred stock, so it is plain that it was never intended that that "investment" would produce the returns which were either contemplated or obtained. To put it another way, the purpose of the stock was not to acquire an ownership in a business whose earnings would be distributed to its stockholders. It was a means by which savings in insurance costs would be returned to the assured obstensibly as dividends,7 not return of premiums.

Under the terms of the preferred stock and the special contract made with Taxpayer, Consolidated was obligated to "pay" quarterly (a) a fixed quarterly dividend if earned and (b) a special dividend in the amount of "the profit earned each quarterly period from its Class `A' insurance business."8

From this so-called, but in fact redeemable, investment of $25,000, the Taxpayer in the first year as a stockholder-policyholder received a return of $54,725. The other special class preferred stockholder shared a like bonanza.9 Such fabulous returns were not due to the managerial skill or the superior judgment of Consolidated. It never paid any dividends to common stockholders during these years. It incurred a net loss in its ordinary life department. It made only a modest profit from investments. And its only net profit for 1953 insurance operations (other than the special class preferred stockholders group policies) was less than $25,000 on premiums earned aggregating nearly $500,000.10

It is equally clear that from Consolidated's point of view it did not consider that it was "earning" any such amounts either for itself or as a conduit for distribution to its stockholders, regular or special, common or preferred. Presumably because it fears that we too may somehow come under the spell, the Taxpayer complains vigorously that in this connection the Government has unduly emphasized one page of a financial exhibit discussed in the lengthy testimony of the Secretary-Comptroller of Consolidated. This sheet shows, for Consolidated's calendar year 1953, the "surplus account." It has a bearing both on the issue of "sham" and the availability of "earnings" out of which to pay for that year $301,272.80 dividends to special classified preferred stock. It shows that there is added to the opening surplus account $11,128.94 as profit from "group creditors preferred plan."11 Taxpayer advances the curious argument that the Court was not entitled to put any value on this. The Court was somehow bound to accept the figures from Consolidated's annual statement filed with the Texas Insurance Board showing its total surplus prior to distribution of these large special dividends to preferred stockholders. The reason advanced was that this "surplus account," note 11, supra, was not a part of the official records of the company. Rather, it was one prepared solely for management, not the stockholders. That to the esoteric mysteries of the double entry system of bookkeeping there is now to be added the double standard is itself an uninspiring plea. If management considered that the real earnings were $11,128.94, we can think of little better authority for the trier of fact to rely upon than that. In this process the trier of fact would be fortified by other portions of that comprehensive financial exhibit headed up "Statement of Financial Condition" of Consolidated. One of the most significant is that identified as "Profit & Loss Statement — Credit Insurance & Group Creditors Plans." A solid page of detailed figures, affirmed by the witness to be correct, reflecting premium income, reserves, claims and commissions paid, and net underwriting income less 24 categories of itemized expenses, it analyzes in exactly the same way the three types of credit insurance written by Consolidated.12 Its treatment of the special class preferred stock group policies is, and must have been to the Trial Judge, most revealing. Beginning with premiums earned totaling $351,494.80, it first deducts claims incurred. Then, to arrive at "Net Underwriting Income," it deducts the special dividends paid to the preferred stockholders. This "Net Underwriting Income" is shown to be $29,832.46. From this figure itemized expenses were deducted aggregating $18,703.52. The final result was the "Net Profit-Group Creditors Preferred Plan" in the sum of $11,128.94.13

This revelation and the obvious difficulties which the responsible executive of Consolidated had in the prolonged examination and cross examination of him concerning the accounting practices followed by the company were relevant factors for the Court to weigh. The Court, in view of them, was not bound to accept the accounting treatment in other annual statements which showed the surplus as an amount which included the sums which were to be disbursed later to preferred stockholders as special dividends.14

More than that, this treatment by management for management was a realistic appraisal of the situation. Consolidated did write insurance. But its position was something less than that of the usual insurer. Consolidated did collect premiums. And after payment of losses incurred, what was left over appeared to be earnings. But these were not earnings which it could use. It could not keep them for it had already bound itself irrevocably to pay them out quarterly — almost as soon as the funds producing them were collected. Taxpayer was an assured. But what it got was something different from usual insurance. It did, of course, gain the competitive advantage from the fact that it had a policy benefiting the initial borrowers. It also had insurance for the full amount of the debts reported under the policy plus the right to schedule more within the cancellation period of the policy. Presumably for the 10% management fee, it also obtained a claims service for the adjustment and settlement of claims. But since the record reflects that all were confident at the time the plan was established that substantial savings would be effected and hence dividends paid, and the record is completely silent that anyone ever contemplated that losses would exceed the prescribed paid in premiums, the policy was more nearly that of a stop-loss coverage. That, of course, was, and could be, valuable, and was and could be quite a legitimate underwriting. But recognizing these factors, when the arrangement is viewed as a whole, it was a method by which to return to Taxpayer all of the premiums paid less 10%, no more and no less. Under the terms of the preferred stock, the fixed $2 dividend was not cumulative. It was due only if earned. As to the special dividends from so-called earnings, see note 8, supra, the loss to Consolidated was to be carried forward and recouped fully before any further dividends were payable. If the claims sustained on this Class A business (confined to Taxpayer's own experience) were less than the premiums paid in, the difference was to be returned as special dividends. On the other hand, if the claims in any quarter were greater than the premiums paid in by Taxpayer, the loss to Consolidated was to be made up out of premiums (and net "earnings" thereon) paid in by Taxpayer in subsequent quarters. Thus, Taxpayer was ultimately to pay for all its losses, whether they were greater or smaller than the premiums paid in during any given period. If it were to escape this consequence, one of two things would have to occur: first, some kind of catastrophic loss which would make it simply unlikely that Taxpayer's premiums could ever reasonably make up the huge deficit in the coming quarters or years; or, second, with a bad loss record, Taxpayer could cancel. This would, it is true, relieve it of the burden of making up the losses, but in the meantime it would have no insurance either real, or limited, imaginary or pretended on subsequent operations.

Although implored earnestly by the Government to do so, we withstand the importunities to declare that the arrangement was a sham because the premiums were so palpably unreasonable and exorbitant that they would be within the reach of scrutiny by the Commissioner. See 4 Mertens, Law of Federal Taxation, §§ 25.05-25.09. The District Court expressly rejected as irrelevant the evidence proffered into the record on reasonableness. Apparently the Court proceeded on the idea that the question of a dividends received credit in no way involved the question of reasonableness of the deductions which were not disallowed by the Commissioner, see note 5, supra. Whatever might be said in behalf of the view that the pervasive issue of sham-or-reality made this a relevant subject for inquiry and evidence, we ought not to allow it to affect the scales where the District Court excluded it from all consideration.15 Reasonableness by its very nature is the evaluation of many subtle factors. If it were essential to a disposition of this case, we would be compelled to require that the Trial Court first determine it on a record intended to comprehend that issue.16

We do not get to the question of reasonableness of the premiums or the reasonableness of the Taxpayer's using this route to insure outstanding indebtedness. For our purposes, we may accept it as reasonable. At the same time we fully concur in the finding of the District Court that this was not what it appeared to be: a stock purchase with earnings from the stockholder's insurance business being returned as corporate dividends. What the Taxpayer did in fact — as it was entitled as a business concern to do — was to purchase insurance and receive back as a return of premiums that part of the premiums not needed for claims and claims administration. That the return was cast in the shape of a corporate dividend was to make it appear to be something other than what it was in fact. It was not earnings. It was the return of an unused, unneeded premium. It was the return17 to the policyholder of money which the insurer bound itself to deliver. "These arrangements" with these fictional dividends are "to us transparent devices." Commissioner of Internal Revenue v. P. G. Lake, Inc., 1958, 356 U.S. 260, 266, 78 S. Ct. 691, 695, 2 L. Ed. 2d 743, 749.

Affirmed.