dissenting:
For thousands of years, philosophers, theologians, and now physicists, have debated whether the earth was created ex nihilo, i.e., out of nothing. Whatever the answer to that question, there is little doubt that my colleagues in the majority have performed a notable miracle of their own in this case, by creating nothing out of something, i.e., nihil ex aliquo. Let us consider how this miracle was wrought by endeavoring to follow the money.
*489I. The Government’s Conditions to De-mutualization
For what precisely did the Dorrances pay when they purchased policies from the mutual life insurance companies involved in this case? The majority contends that they paid only for a death benefit, the right to surrender the policy for a “cash value,” and annual policyholder dividends representing their share of the company’s “divisible surplus.”
But if, as the majority contends, the Dorrances paid nothing for their membership rights, and did not contribute capital, then why did the several governmental regulators involved require, as a condition of demutualization of each of those insurance companies, that they issue stock to their policyholders to compensate them for the loss of those rights?
Since those who acquired shares in the newly publicly traded insurance companies during the IPO process paid cash for their interests, if the policyholders when the insurance companies were structured as mutual insurance companies had not paid for the surplus they later received in stock, then the value of the distributed shares ought to have remained as the insurance companies’ working capital, and not been gratuitously gifted to policyholders. Neither the regulators nor the IPO investors would have tolerated such a gratuity.
But the stock distribution to the Dor-rances, even if not specifically contemplated at the time they purchased the policies, was no gift. While insurance companies may be powerful, they do not have the power of creation ex nihilo. To the contrary, by the very nature of a mutual insurance company, all of its accumulated value comes from premiums paid by its owners, and the investment of those premiums. That is why, when allocating shares during the demutualization process, the insurance companies relied on a calculation of a fixed component based on the loss of voting rights and a variable component related to past and projected future contributions to surplus.
The majority relies on a statement by a government’s expert: “Some may think that the cash paid out in demutualization comes from the distribution of positive surplus of the mutual company; however, such is not the case. The cash actually comes from new stockholders which subscribe to the IPO.... ” Here, the Dor-rances received stock, not cash. Of course, when they sold the stock, the cash that they obtained from the sale came from the buyers of the stock, and not from the insurance companies’ bank accounts. But that is always true in a stock sale. Of course, that does not mean that all stock sales have a zero basis. Thus, the cited government expert’s testimony is merely a truism. It provides no support for the majority’s conclusion.
II. Accrued Surplus or Not?
Some context is in order. The majority mentions the IPO value of the Dorrances’ stock: $1,794,771. The majority also unworthily mentions the Dorrances’ net worth, which is not relevant to any issue before us. While the majority concedes that the premiums the Dorrances had paid to the insurance companies, which totaled $15,265,608, were “substantial,” the majority is unimpressed by that figure because the face value of the policies was substantially larger than the premium. Of course, that is always the case in insurance. The relevance of the premiums paid to the question before us is that the distributed stock represents only 11.7% of the money the Dorrances had paid the insurance companies. That may not be far from the usual dividends paid on mutual insurance *490policies.1
However, the majority is quick to call that return of a small proportion of funds expended a “windfall.” But while the majority asserts that one insurance company official so characterized the stock distribution, he actually took care to state that “windfall” was the company’s characterization, not his. Moreover, the majority ignores the fact that every other insurance company representative deposed in this case either expressly rejected that characterization, or in one instance, did not know how to answer the question.
The majority credits testimony by the government’s expert that the insurance companies charged the Dorrances premiums that were based solely on the expected costs of providing insurance benefits, using calculations that were “very precise in actuarial circles,” such that “there is just no portion of the premium or charge for membership rights.” That asserted precision is disproved by the existence of a surplus accrued within the insurance company. In fact, the majority elsewhere relies on testimony that, at the time of demu-tualization, “less than 10% of the SunLife surplus was attributable to current policyholders; premiums paid by former policyholders accounted for over 90% of the surplus.”
In other words, despite their asserted actuarial precision, the insurance companies had not been returning via dividend all of the premium surplus. Instead, the surplus accumulated within the companies, where it served the role that any accumulation of capital does. Therefore, the majority errs by stating that “it is well understood that policyholders do not contribute capital to the companies.”2 If not from the policyholders, from whence did that accumulated capital come?
Certainly, the cited testimony raises the question of how much the Dorrances contributed to the surplus. That question was addressed during the demutualization. To determine'the number of shares of stock to issue to each member, the insurance companies applied a formula approved by the government regulators, which included a fixed component and a variable component. According to that formula, 14-25% of each company’s shares were allocated on a fixed basis to shareholders. The variable shares were allocated based on the “contribution-to-surplus” method, which allocated the total shares based on a policyholder’s contribution.
Thus, even if we were to accept the majority’s conclusion that the Dorrances had no basis in the voting aspect of the *491membership rights — remembering that the fixed shares granted solely on that basis were worth $3,164, a minuscule portion of the $1,794,771 of IPO stock at issue — the calculations expressly accounted for their actual contribution to the surplus.
III. “Tax Free Exchange” Is Not a Synonym for “Zero Basis”
The majority also misapplies the concept of a tax-free exchange in stating that “[t]he taxpayer can’t have it both ways — a tax-free exchange with zero basis and then an increased basis upon sale of the stock.”
It is unclear how the Dorrances are trying to “have it both ways.” All that is required for the exchange to be tax-free is for the value received in stock to be the same as the value of the- property exchanged. See 26 U.S.C. § '358(a)(1). In this case, the IRS, citing its own interpretations, opined that the basis should be zero. Whether that interpretation squares with the facts is the very question at issue in this case. By relying in part on the IRS’s interpretation to answer the question, the majority assumes the conclusion.
IV. The District Court’s Sound Calculations
After hearing all of the evidence at trial, the district court determined the Dor-rances’ cost basis by deducting the expected future premium contribution from the IPO value of the stock, yielding a cost basis of $1,078,128. This was the sum of: (1) the IPO value of the fixed shares allocated to the Dorrances ($3,164) and (2) 60% of the IPO value of the variable shares ($1,074,964). The 60% proportion reflected an expert estimate of past contributions by the Dorrances to the life insurance policies; the remaining 40% was an estimate of the policyholders’ future contributions to the policies. Applying this formula, the court found that the Dor-rances were required to pay taxes on $1,170,678, which was their sale proceeds of $2,248,806 less their basis of $1,078,128.
Thus, the district court quite sensibly reduced the basis by an expert’s estimate of the future contribution component of the IPO value, ensuring that the Dor-rances would not underpay the taxes owed. This was a careful analysis using reasonable methodology based on the evidence presented at trial. By contrast, the majority’s contrary conclusions do not follow from the facts. A portion of the assets of the insurance companies clearly came from the premiums paid by the Dorrances, and they had a substantial basis in the stock distributed to them. By contending to the contrary, my colleagues in the majority have created nothing out of something. It’s a miracle!
I respectfully dissent.
. The parties did not identify the dividend rates the policies at issue provided. Data for the Massachusetts Mutual Life Insurance Company, not one of the companies at issue, is publicly available. See Historical Dividend Studies from Massachusetts Mutual Life Insurance Company (2015), available at https:// fieldnet.massmutual.com/public/life/pdfs/li 7954.pdf (last visited Nov. 18, 2015). That data 'shows that a policy purchased after March of 1996 yielded a yearly dividend interest rate of between 8.4% and 7.9% between 1996 and 2003.
. The majority misconstrues government witness Ralph Sayre’s testimony in this regard. Sayre testified that, from the view of a mutual insurance company, "because we don’t have shareholders who have contributed to surplus or contributed capital to withstand [the demand for benefit payments], we’re going to have to charge [the policyholder] a little bit more of that up front. But keep in mind that we will also give it back to you. As our experience unfolds and we realize earnings from that extra charge, or from the use of that extra money, we will return it back to you.” Thus, policyholders do contribute capital — but they are eventually supposed to get it back. The majority believes that it comes back with a basis of zero, which complements the majority's belief that the insurance companies created something out of nothing.