The courts have taken varying approaches to determining the basis of stock that is received by an insurance policyholder in exchange for the policyholder's surrender of membership rights in a mutual insurance company, in a "demutualization" transaction. While this may seem to be a narrow and abstruse question, the approaches taken by the courts may have application in other areas of the tax law affecting analogous transactions.
Most recently, the Court of Appeals for the Ninth Circuit, in Dorrance v. United States,1 reversed the district court decisions in that case,2 agreed with the government position that the policyholder's basis in the stock did not include any part of the premiums paid by the policyholder for insurance, and concluded that the entire proceeds from the subsequent sale of the stock by the policyholder constituted gain.
Mutual insurance companies are owned by policyholders, rather than by shareholders. Policyholders who purchase insurance from a mutual company typically obtain, without payment of additional consideration beyond stated premiums, "membership rights," including voting rights and the right to participate in the distribution of surplus in the event of the dissolution of the company. Ordinarily the membership rights in a mutual insurance company have no significant value: each policyholder typically has only one vote, regardless of the amount of insurance purchased, and the record before the Court of Appeals included expert testimony indicating that none of the testifying experts could recall a dissolution of a solvent mutual insurance company that resulted in distributions to its members.
By the 1990's, it became clear that an insurer's doing business as a stock company rather than as a mutual company could result in significant advantages with respect to, for example, raising additional capital and diversification. Changes in state law were then made to facilitate "demutualization" (that is, conversion from a mutual company to a stock company). As a condition to the approval of such transactions, however, state insurance regulators required that mutual policyholders be compensated for the surrender of their voting rights and rights to surplus through the receipt of stock in the stock company or cash in connection with the demutualization transaction. The IRS issued a series of private letter rulings confirming that the receipt of stock would generally be nontaxable, on the theory that the demutualizations qualified as "reorganizations" for income tax purposes.
Bennett and Jacquelynn Dorrance (the Dorrances) purchased in 1996, through a trust formed by them, life insurance policies with an aggregate face value of approximately $88 million from five mutual insurance companies. The policies were purchased to fund the anticipated payment of estate taxes upon the Dorrances' death.
Upon the demutualization of each of the five insurers in 2000 and 2001, stock was issued to the Dorrances in exchange for the surrender of their membership rights in the mutual company, and their insurance policies were continued in force by the resulting stock company. The companies informed their policyholders that the demutualization transactions were believed to be nontaxable, but that, consistent with the IRS position as expressed in published rulings, the policyholders would have a basis of zero in the shares received. The shares issued in respect of the Dorrances' insurance policies had an aggregate value of $1.8 million at the time of receipt.
The shares were sold in 2003 for $2.2 million, and the full proceeds from the sales were reported as capital gain. In 2007, however, the Dorrances filed a claim for refund, arguing that the cash received for the shares should be treated as a partial return of insurance premiums paid in respect of the insurance policies to which the membership rights were attributable, such that no tax was owed with respect to the sale. The IRS did not allow the claim, and the Dorrances filed suit for a refund in 2009.
The Dorrances were successful in the district court, which rejected the government's position that, because the insurance premiums were paid for insurance coverage and not for membership rights, the Dorrances had no basis in the stock received in the demutualizations. The district court also rejected an argument made by the Dorrances based on the "open transaction" doctrine.3 (Such an argument had been accepted by the Court of Federal Claims in a similar case where the shares received in a demutualization transaction had been sold immediately for cash (pursuant to a "cash option" provided by the insurance company).4)
The argument was that, taking into account that the value of the membership rights could not be ascertained at the time those rights were acquired (that is, when the insurance policies were issued), and, therefore, that an allocation of basis could not be made based on relative values of the insurance coverage and the membership rights at that time, it was appropriate under the open transaction doctrine to treat the cash received as a partial recovery of the taxpayer's overall cost based upon the insurance premiums paid, and as not includible in income where the cash received was not in excess of the full premium cost of the policy.
Having rejected the absolutist positions urged by each side, the district court concluded, in its first decision in the matter (in 2012), that some amount had been paid by the Dorrances for their membership rights. These membership rights had been received under policies for which premiums had been paid, and those policies had included both insurance policy rights and membership rights. Therefore, further proceedings were necessary to determine an appropriate allocation of basis.
The district court further concluded, in its second decision (in 2013), that the basis of the membership rights could be determined as the value (at the time of each demutualization) of the stock that was received in exchange for the surrender of (i) voting rights and (ii) rights to the insurance company surplus that were attributable to the policyholder's deemed past contributions to surplus. The value of the stock that was received in exchange for these rights was determined by reference to calculations under which the insurance companies had calculated the amounts of stock to be issued to the policyholders in the demutualizations.
The Court of Appeals in Dorrance, noting that the Dorrances had the burden of establishing their basis in the stock that was sold, and concluding that they had not met that burden, reversed.
The evidence suggested to the Court of Appeals that, apart from the extraordinary context of the demutualization, membership rights had little if any value, because the voting right was limited to one vote per policyholder, and none of the testifying experts could recall any instance of a liquidation of a solvent mutual insurance company in which holders of membership rights received distributions.
It was also noted that, from a mutual company's perspective, there was no cost in providing membership rights (apart from the minimal administrative expense associated with a policyholder vote), that the rights were not transferable, and that the policyholder would not, in the context of termination of an insurance policy, receive anything in exchange for the surrender of those rights.
The Court of Appeals also focused on the manner in which basis in an insurance policy is determined under the Internal Revenue Code. For example, upon receipt of the cash surrender value for a policy, a policyholder will generally not have income if the amount received does not exceed the policyholder's "investment" in the insurance policy. This "investment" includes the aggregate amount of premiums paid (less any amounts previously received under the policy).5 The court observed that a taxpayer "can't have it both ways" – that is, that treatment of the full amount of the premiums as an investment in the contract (the insurance and cash surrender features of the policy) for other tax purposes is inconsistent with allocating a portion of the cost attributable to those premiums to stock received in exchange for membership rights.
Further, there was no indication that the policyholder paid more in premiums because of the associated membership rights. In fact, one of the Dorrances testified at trial that it was his understanding that he would pay less for an insurance policy from a mutual company than from a stock company. It was also noted that the policies were continued after the demutualization with no reduction in the cost of the premiums, notwithstanding the termination of the membership rights.
Taking into account all of the above, the Court of Appeals concluded that the Dorrances had paid nothing for the membership rights and therefore had no cost basis in the rights; and that the district court had erred in concluding that there was basis in the membership rights without focusing on this critical step in the basis analysis. The lower court also erred, according to the appellate court, in estimating basis by reference to the value of the stock at the time of demutualization, rather than by calculating basis at the time the membership rights arose through acquisition of the policies.6 Consequently, the district court's determination as to basis could not stand, and the government prevailed in its assertion that the Dorrances had no basis in the shares received in the demutualization and were not entitled to any refund.
One judge of the Court of Appeals dissented. He criticized the majority for failure to adequately reconcile their assertion that the Dorrances paid nothing for their membership rights, with the circumstance that the state government insurance regulators clearly viewed those rights as having substantial value in determining the terms of the demutualizations.
The conclusion of the Court of Appeals in Dorrance does not seem surprising, and the court's opinion underscores a point that perhaps has been implicit in the sparse authority concerning allocation of basis in the context of the determination of gain from a partial disposition of property: that is, that where a portion of a larger property is sold, the cost basis of the part that is sold should generally be determined by reference to an allocation of the initial cost of the entire property that is based on the values of the various portions of the property at the time the larger property was acquired.7
As a corollary, it may obviously be helpful, in applying this general rule to determine gain or loss in the context of a disposition of a portion of a larger property, if the records of the earlier acquisition contain as much information as possible not only as to the overall cost of the property acquired but also as to the relative values of the various components or other portions of such property.
1. Docket Nos. 13-16548, 13-16635 (December 9, 2015), reversing Dorrance v. United States, 111 AFTR 2d 2013-1280 (D. Ct. Arizona).
2. The decisions of the district court in Dorrance in 2012 and 2013 were discussed by the authors in an earlier article in this column. See E. Pisem and D. Kahen, Diverse Approaches to Allocation of Basis in Demutualizations, NYLJ, April 18, 2013.
3. See Burnet v. Logan, 283 U.S. 404 (1931). But see Treas. Reg. § 1.1001-1(a) (appearing to limit the application of the open transaction doctrine to "rare and extraordinary cases" where the property received in an exchange is so speculative in nature as to be considered to have no fair market value).
4. Fisher v. United States, 102 AFTR 2d 2008-5608, aff'd without opinion, 105 AFTR 2d 2010-357 (Fed. Cir. 2009). The district court in Dorrance noted that another district court, in a context where the stock was similarly sold several years after the demutualization, had declined to apply the open transaction doctrine, taking into account that the value of the stock received was readily determinable in the year in which it was received. Reuben v. United States, 111 AFTR 2d 2013-620 (C.D. Calif. 2013).
5. See IRC § 72(e).
6. The Court of Appeals opinion further states in a footnote that, in light of the decision of the court, "it is unnecessary to address" whether the open transaction doctrine (as applied by Fisher) was applicable.
7. See, e.g., Treas. Reg. § 1.61-6(a).
Originally published in The New York Law Journal.
The content of this article is intended to provide a general guide to the subject matter. Specialist advice should be sought about your specific circumstances.