American Electric Power Company, Inc. v. United States

GILMAN, J., delivered the opinion of the court, in which COLE, J., joined. DAVID A. NELSON, J. (pp. 745 - 747), delivered a separate concurring opinion.

OPINION

GILMAN, Circuit Judge.

American Electric Power Company, Inc. (AEP) implemented a corporate-owned life insurance (COLI) plan in 1990 that caused it to purchase policies on the lives of over 20,000 of its employees. In 1996, AEP deducted from its federal income tax approximately $66 million in interest due on loans made against the policies. The IRS disallowed the deduction and assessed AEP with an additional $25 million in taxes for the year in question. AEP paid the additional tax under protest and then sued for a refund in federal district court. After a six-week bench trial, the district court granted judgment in favor of the government. For the reasons set forth below, we AFFIRM the judgment of the district court.

I. BACKGROUND

A. Factual background

AEP purchased a COLI plan offered by Mutual Benefit Life Insurance Company (MBL) in 1990. (By virtue of a reinsurance transaction, Hartford Life Insurance Company replaced MBL as the primary insurer in 1993. For simplicity, both companies will hereafter be referred to as MBL.) The plan consisted of life insurance policies on over 20,000 AEP employees. AEP was the named beneficiary on all of these policies. Each policy had a fixed annual premium of $16,667, the significance of which is explained in Part II.A. below.

The plan utilized an intricate funding mechanism. On the first day of each of the first three policy years, two financial transfers took place simultaneously. The first was the total premium paid by AEP (an amount in excess of $330 million), which MBL credited to the “policy value.” (“Policy value” is a term of art utilized by the parties to denote the cumulative gross value of all the insurance policies within the COLI plan.) The other was a loan extended to AEP by MBL, with the policy value used as collateral, in an amount equal to more than 90 percent of the gross premium. In the first year, for example, this “netting transaction” resulted in AEP actually paying MBL the greatly reduced sum of $23.5 million.

A different netting transaction occurred on the first day of each of the next four policy years. AEP paid the full premium, but MBL applied only about 5 percent of the premium to increase the policy value. It took the remainder as an “expense charge.” MBL then returned to AEP approximately 95 percent of the amount taken as an expense charge as a ‘loading dividend.” AEP also paid the accrued loan interest, but withdrew from the policy value nearly the same amount. Thus, as in years one through three, the cash actually paid by AEP to MBL in years four through seven was only a small percentage of the stated cost of the premiums.

The net equity of AEP’s COLI plan was based upon the policy value, minus policy loans and accrued interest. AEP’s plan operated so that whatever cash was paid in at the beginning of the year was used up by the year’s end. At the end of each year, therefore, the COLI plan had a net equity of zero.

The" plan was also designed to ensure “mortality neutrality.” This meant that neither AEP nor MBL reasonably expected to profit over the life of the plan because of the death of AEP employees. To achieve this equilibrium, “cost-of-insurance charges” increased if more employees died than were expected, and, if experience proved the opposite, the insurance company paid AEP “mortality dividends.”

The COLI plan afforded AEP the opportunity to annually select from a menu of options the interest rate that it would pay MBL on the policy loans. AEP always picked one of the highest interest-rate choices, never the lowest. For example, the policy-loan interest rate chosen for the first year was 3.1.88 percent, even though AEP could have selected a rate as low as 7.4 percent. AEP had two incentives to choose the higher rates: (1) the higher rates increased its tax deduction for interest paid, and (2) the COLI plan provided that the true cost to AEP was only the one-percent differential between what AEP paid MBL as interest on the policy loans and what MBL credited back to AEP in the form of “interest” on the policy value.

The interest-rate component of AEP’s COLI plan was only one of many provisions designed to exploit certain aspects of the Internal Revenue Code (IRC) then in effect, which are discussed in Part II.A. below. AEP fully understood that the plan would generate positive cash flow in every policy year only if it could deduct from its income taxes the interest on the policy loans. It decided to implement the COLI plan based upon this expectation. In 1996, however, the IRS refused to allow the claimed deduction. AEP therefore paid the additional tax under protest and then instituted this action to recover its alleged overpayment.

B. Procedural background

A bench trial commenced in October of 2000 and concluded two months later. Both sides filed proposed findings of fact and post-trial memoranda. In February of 2001, the district court granted judgment in favor of the United States. The court determined that the COLI plan was an economic sham. It also held that the dividends in the fourth through seventh years of the plan, among other aspects, were shams in fact. This timely appeal followed.

II. ANALYSIS

A. Whether the COLI plan was an economic sham

Generally, a taxpayer may deduct “all interest paid or accrued within the taxable year on indebtedness.” IRC § 163(a). But special rules apply to life insurance. In 1990, IRC § 264(a)(3) provided that, with certain exceptions, no deduction was permitted for “any amount paid or accrued on indebtedness incurred or continued to purchase or carry life insurance ... pursuant to a plan of purchase which contemplates the systematic direct or indirect borrowing of part or all of the increases in the cash value of such contract....” Section 264(c)(1) delineated the exception relevant to AEP’s COLI plan: a deduction was allowed “if no part of 4 of the annual premiums due during the 7-year period (beginning with the date the first premium on the contract to which such plan relates was paid) is paid under such plan by means of indebtedness.” In any case, § 264(a)(4) capped the interest that could be deducted on indebtedness incurred to purchase life insurance at $50,000 of indebtedness per insured life.

AEP’s COLI plan on its face fits neatly within the 4-of-7 safe harbor rule. Only in the first three years of the plan were premiums financed through policy loans. And because the premiums were fixed at $16,667, the policy loans through the first three years were precisely matched to the $50,000 of indebtedness per insured life on which the interest was deductible ($16,667 x 3 years = $50,001).

But when “it is patent that there [is] nothing of substance to be realized by [the taxpayer] from [a] transaction beyond a tax deduction,” the deduction is not allowed despite the transaction’s formal compliance with Code provisions. Knetsch v. United States, 364 U.S. 361, 366, 81 S.Ct. 132, 5 L.Ed.2d 128 (1960) (holding that the purchase of several annuity bonds was substantively a sham, and should therefore be disregarded in determining the validity of claimed income tax deductions, where the premiums were paid by loans secured by the bonds and additional borrowing reduced the annuity each bond would pay from tens of thousands to a pittance). This is known as the economic sham, or sham-in-substance, doctrine. In this circuit,

[t]he proper standard in determining if a transaction is a sham is whether the transaction has any practicable economic effects other than the creation of income tax losses. A taxpayer’s subjective business purpose and the transaction’s objective economic substance may be relevant to this inquiry.

Rose v. Comm’r, 868 F.2d 851, 853 (6th Cir.1989) (citations omitted).

The district court decided that AEP’s COLI plan was an economic sham. Its findings of fact on this issue are reviewed under the “clearly erroneous” standard. Rink v. Comm’r, 47 F.3d 168, 172 (6th Cir.1995). But the standard of review for the ultimate conclusion — sham or no sham — is not as clear. The government asserts that the ultimate conclusion is also subject to the clearly erroneous standard of review, citing Ratliff v. Commissioner, 865 F.2d 97, 98 (6th Cir.1989) (stating that “[t]here is one issue presented in this case: whether the Tax Court erred in concluding that the straddle transactions engaged in by the taxpayers were ‘shams in substance,’ ” and then asserting that “the determination of the Tax Court is binding on the appellate court unless clearly erroneous”).

This court, however, has subsequently read Ratliff as meaning that “the legal standard applied by the Tax Court and its legal conclusions based upon its findings of fact are reviewed de novo.” Smith v. Comm’r, 937 F.2d 1089, 1096 (6th Cir.1991) (citing Ratliff, 865 F.2d at 98). Relying on Smith, AEP contends that de novo review of the district court’s ultimate determination is appropriate. Beyond their arguments and the above citations, the parties do not provide any additional guidance concerning the proper standard of review.

Smith does seem to adopt a strained reading of Ratliff, and “[a] panel of this [e]ourt cannot overrule the decision of another panel,” Salmi v. Sec’y of Health & Human Servs., 774 F.2d 685, 689 (6th Cir.1985). But the standard of review announced in Smith is clearly correct, given that the Supreme Court has stated that “[t]he general characterization of a transaction for tax purposes is a question of law....” Frank Lyon Co. v. United States, 435 U.S. 561, 581 n. 16, 98 S.Ct. 1291, 55 L.Ed.2d 550 (1978). Because Ratliff did not specifically hold that the ultimate question of whether a transaction is a sham is to be reviewed under the clearly erroneous standard (although one of its parenthetical citations did so state), and because Frank Lyon indicates that review should be de novo, we will follow Smith. The district court’s conclusion that AEP’s COLI plan was a sham is therefore subject to de novo review.

We now turn to the merits of the case. “The main nontax benefits insurance plans generally offer are mortality gains to the beneficiary, who does not pay tax on proceeds, and interest-free inside build-up [the accrual of interest on the policy value].” In re CM Holdings, 301 F.3d 96, 103 (3d Cir.2002). AEP was unable to realize either benefit on its COLI plan, however, because the plan was mortality-neutral and had zero net equity at the end of each policy year.

The district court found that, absent the policy-loan interest deductions, AEP would lose a substantial amount of money on its COLI plan; but with the deductions, AEP stood to gain positive cash flow every year. Am. Elec. Power Inc. v. United States, 136 F.Supp.2d 762, 787 (S.D.Ohio 2001). Moreover, the district court found that “the opportunity for tax arbitrage ... was the real motivation for AEP’s purchase of the COLI policies.” Id. at 786. The COLI plan thus bears the hallmarks of an economic sham.

AEP contends, however, that Woodson-Tenent Laboratories, Inc. v. United States, 454 F.2d 637 (6th Cir.1972), dictates a contrary conclusion. We disagree. This court in Woodson-Tenent found that the two corporate-owned life insurance policies on key employees had economic substance. Under those policies, the company could have realized a substantial mortality gain if its key employees died prematurely. In contrast, the mortality-neutral design of AEP’s COLI plan eliminated this possibility. Woodson-Tenent is therefore inappo-site.

AEP also relies upon United States v. Consumer Life Insurance Co., 430 U.S. 725, 97 S.Ct. 1440, 52 L.Ed.2d 4 (1977). In Consumer Life, the Supreme Court held that a reinsurance agreement between two insurance companies, which included a provision that the reinsurer would eventually recoup the amount it paid out on claims through “experience refunds” from the primary insurer, had economic substance. Id. at 737-38, 97 S.Ct. 1440. We agree, but simply find a world of difference between having a “mortality neutrality” provision or its equivalent in a reinsurance agreement between two insurance companies as opposed to having such a provision in an agreement between an insurance company and the insured. Mortality gains to the beneficiary, after all, are one of the main nontax benefits, if not the main non-tax benefit, that motivates consumers to purchase life insurance policies. In re CM Holdings, 301 F.3d at 103. We therefore find AEP’s reliance on Consumer Life unpersuasive.

Similar reasoning leads us to reject AEP’s argument that because “each individual COLI policy transfers risk and has economic substance,” the COLI plan as a whole must have economic substance. Each individual COLI policy did in fact transfer risk and thereby have economic substance, as was the case in Woodson-Tenent Mortality neutrality, however, operated on the plan as a whole, ensuring that whatever transfer of risk occurred at the level of the individual policy ceased to exist for the overall plan.

AEP alternatively contends that the plan had nontax economic consequences because some of its subsidiaries experienced mortality loss, while others experienced mortality gain. This argument suffers from the same flaw. The economic-sham analysis focuses on the “substance to be realized by [the taxpayer] from [a] transaction,” Knetsch, 364 U.S. at 366, 81 S.Ct. 132 (emphasis added), and the transaction here was AEP’s purchase of the COLI plan. Mortality neutrality and zero net equity operated only at the level of the entire plan. AEP cites Humana Inc. v. Commissioner, 881 F.2d 247 (6th Cir.1989), for the proposition that “differing subsidiary-level experience means the transaction has economic effect.” But Humana is readily distinguishable, partly because of the completely different fact pattern where both the insured and the insurer were wholly owned subsidiaries of the same parent corporation, and partly because the policies at issue in Humana were not mortality-neutral. The case simply has no relevance to the factual circumstances before us.

Another case upon which AEP relies, Sacks v. Commissioner, 69 F.3d 982 (9th Cir.1995), is also easily distinguishable. The court in Sacks reasoned as follows:

Factors which demonstrate that Mr. Sacks’ deal had genuine economic effects, and was not a sham, are that (1) Mr. Sacks’ personal obligation to pay the price was genuine; (2) he paid fair market value; (3) the tax benefits would have existed for someone, either BFS Solar or Mr. Sacks, so the transaction shifted them but did not create them from thin air; (4) the business of putting solar water heaters on homeowners’ roofs was genuine; and (5) the business consequences of a rise or fall in energy prices and solar energy devices were genuinely shifted to Sacks by the transaction.

Id. at 988. In contrast, AEP did nothing more than “show[ ] a deductible expense on paper, without actually suffering any of the ordinary economic consequences of paying the money,” id., because circular netting transactions obviated the obligation to ever actually repay the underlying policy loans. This means that the tax benefits generated by the circular policy-loan interest deductions arose “from thin air.” Id. Furthermore, the business consequences of unanticipated mortality experience were not genuinely assumed by either AEP or MBL because the plan was mortality-neutral.

AEP therefore has no basis to rely on the dictum in Sacks that “[w]here a transaction has economic substance, it does not become a sham merely because it is likely to be unprofitable on a pre-tax basis.” Id. at 991. As the foregoing analysis demonstrates, Sacks’s transaction had genuine economic consequences other than the creation of deductions, whereas AEP’s does not.

Nor do we believe that the Sacks court’s dictum about “pre-tax” profitability means, as AEP urges, that we should evaluate the transaction’s profitability after the challenged deduction has been allowed. To do so would swallow the sham analysis entirely. As the Third Circuit has explained:

The point of the analysis is to remove from consideration the challenged tax deduction, and evaluate the transaction on its merits, to see if it makes sense economically or is mere tax arbitrage. Courts use ‘pre-tax’ as shorthand for this, but they do not imply that the court must imagine a world without taxes, and evaluate the transaction accordingly. Instead, they focus on the abuse of the deductions claimed_Choosing a tax-favored investment vehicle is fine, but engaging in an empty transaction that shuffles payments for the sole purpose of generating a deduction is not.

In re CM Holdings, 301 F.3d 96, 105 (3d Cir.2002). Life insurance is tax-favored, and a company may utilize the tax advantages of a COLI plan so long as the plan has real economic consequences. Such was the case in Woodson-Tenent. But it may not utilize life insurance solely to create tax deductions, as was done in the present case, and expect those deductions to be allowed.

AEP’s final argument as to why the COLI plan was not a sham is based on AEP’s good faith intent to utilize the tax savings from the plan to offset dramatically increased costs due to a mandated accounting change regarding its employees’ medical benefits, rather than having to cover the cost increase with higher utility rates to its customers. The controlling question, however, is whether there is anything of “substance to be realized by [the taxpayer] from [a] transaction beyond a tax deduction.” Knetsch, 364 U.S. at 366, 81 S.Ct. 132.

In stating that “[a] taxpayer’s subjective purpose ... may be relevant to this inquiry,” Rose v. Comm’r, 868 F.2d 851, 853 (6th Cir.1989), this court was referring to the taxpayer’s subjective intent with respect to the transaction itself, not its intent with respect to potential uses for the tax savings to be derived from a sham transaction. Illes v. Comm’r, 982 F.2d 163, 166 (6th Cir.1992) (quoting the above statement by the Rose court and explaining that the economic-sham inquiry “consists of an examination of the transaction, not the taxpayer. If the transaction lacks economic substance, then the deduction must be disallowed without regard to the ‘niceties’ of the taxpayer’s intent.”). Money generated by means of abusive tax deductions can always be applied to beneficial causes, but the eventual use of the money thus generated is not part of the economic-sham analysis. In re CM Holdings, 254 B.R. 578, 638-39 (2000).

Both the Third and Eleventh Circuits have concluded that COLI plans very similar to AEP’s were economic shams. In re CM Holdings, 301 F.3d 96, 107 (3d Cir.2002) (“[T]he transaction as a whole lacked economic substance, and thus was an economic sham.”); Winn-Dixie Stores, Inc. v. Comm’r, 254 F.3d 1313, 1317 (11th Cir.2001) (“[T]he broad-based COLI program lacked sufficient economic substance to be respected for tax purposes .... ”), cert. denied, 535 U.S. 986, 122 S.Ct. 1537, 152 L.Ed.2d 464 (2002). We reach the same conclusion in this case.

B. Whether particular aspects of the COLI plan were shams in fact

The district court also held that the dividends in the fourth through seventh years of AEP’s COLI plan, among other aspects, were shams in fact. After disregarding those parts of the plan that it concluded were factual shams, the district court held that the plan did not fit into the 4-of-7 safe harbor rule of IRC § 264. This analysis constituted an alternative basis for the district court’s conclusion that AEP was not entitled to deduct the policy-loan interest purportedly incurred in 1996. AEP challenges the soundness of the court’s sham-in-fact analysis. Because we conclude that the COLI plan as a whole was an economic sham, we need not reach this issue.

We observe, however, that the correctness of the district court’s sham-in-fact analysis is far from clear. “Factual shams are ‘transactions’ that never actually occurred.” In re CM Holdings, 301 F.3d at 108. This court’s decision in Kennedy v. Commissioner, 876 F.2d 1251 (6th Cir.1989), is consistent with this understanding of what constitutes a sham in fact. In Kennedy, the taxpayer tried to claim a deduction for “expenditures paid or incurred during the taxable year for the development of a mine or other natural deposit_” Id. at 1254 (quoting 26 U.S.C. § 616(a)). This court upheld the tax court’s disallowance of the claimed deduction because the underlying transaction was a factual sham: “[T]here is no evidence showing that any amounts invested by the taxpayers were spent on mining development of their plots_ The evidence here supports the trial court’s finding that the investors’ mineral claim leases were fictitious.” Id. at 1254. Similarly, the tax court disallowed certain claimed deductions in Forseth v. Commissioner, 85 T.C. 127, 1985 WL 15375 (1985), where it determined that the underlying precious-metal transactions did not exist: “Petitioners have failed to prove that there was any actual gold or platinum, that there was any real market or trading, or that there was any purpose, other than the avoidance of taxes, for any of the transactions in issue.” Id. at 165. Kennedy and Forseth are examples of shams in fact because the purported transactions in those cases on which the taxpayers claimed deductions had not actually occurred.

In holding that the dividends in years 4-7 were factual shams, the district court below focused on the fact that they were generated by circular, cashless netting transactions. But there is nothing in the record indicating that these transactions did not actually occur. The district court’s holding therefore seems to be extending the factual-sham doctrine beyond its generally accepted definition. In dicta, the Third Circuit has rejected the district court’s conclusion on this point: “The loading dividends of years 4-7 were similar simultaneous netting transactions that ‘actually occurred,’ and are therefore not factual shams.” In re CM Holdings, 301 F.3d at 108.

In sum, the soundness of the district court’s sham-in-fact analysis is questionable. Having concluded, however, that AEP’s COLI plan as a whole lacked economic substance, we decline to decide whether any particular aspects of AEP’s COLI plan were factual shams.

III. CONCLUSION

For all of the reasons set forth above, we AFFIRM the judgment of the district court.