Menu
Tax Notes logo

Amicus Dow Chemical Argues Loading Dividends and Partial Withdrawals Not Functional Equivalents in AEP COLI Case

MAR. 15, 2002

American Electric Power Co. Inc., et al. v. United States

DATED MAR. 15, 2002
DOCUMENT ATTRIBUTES
  • Case Name
    AMERICAN ELECTRIC POWER COMPANY, INC. AND AFFILIATED CORPORATIONS AND AEP SERVICE CORPORATION, Appellants, v. UNITED STATES OF AMERICA, Appellee.
  • Court
    United States Court of Appeals for the Sixth Circuit
  • Docket
    No. 01-3495
  • Authors
    Magee, John B.
    Stark, Richard C.
    Madan, Rajiv
    Goldman, Gerald
    Gilfeather, Valorie
  • Institutional Authors
    McKee Nelson LLP
    The Dow Chemical Co.
  • Cross-Reference
    American Electric Power, Inc., et al. v. United States; 87 AFTR2d

    Par. 2001-529; No. C2-99-724 (20 Feb 2001)(For a summary, see Tax

    Notes, Feb. 26, 2001, p. 1198; for the full text, see Doc 2001-5282

    (26 original pages) or 2001 TNT 36-8 Database 'Tax Notes Today 2001', View '(Number'.)
  • Code Sections
  • Subject Area/Tax Topics
  • Industry Groups
    Insurance
  • Jurisdictions
  • Language
    English
  • Tax Analysts Document Number
    Doc 2002-10049 (37 original pages)
  • Tax Analysts Electronic Citation
    2002 TNT 86-78

American Electric Power Co. Inc., et al. v. United States

 

IN THE UNITED STATES COURT OF APPEALS

 

FOR THE SIXTH CIRCUIT

 

 

On Appeal from the United States District Court

 

For the Southern District of Ohio, Eastern Division

 

 

THE BRIEF OF THE DOW CHEMICAL COMPANY AMICUS CURIAE IN

 

SUPPORT OF NEITHER PARTY AND NEITHER AFFIRMANCE NOR REVERSAL

 

 

Valorie A. Gilfeather

 

The Dow Chemical Company

 

2030 Dow Center

 

Midland, MI 48675

 

(989) 636-6400

 

 

Of Counsel

 

 

John B.Magee

 

(Lead Counsel)

 

Richard C. Stark

 

Raj Maden

 

Gerald Goldman, Of Counsel

 

McKee Nelson LLP

 

1919 M Street, N.W.

 

Washington, D.C. 20036

 

(202) 775-1880

 

 

Counsel for

 

The Dow Chemical Company

 

UNITED STATES COURT OF APPEALS

 

FOR THE SIXTH CIRCUIT

 

 

(This statement should be placed immediately preceding the table of contents in the brief of the party. See copy of 6th Cir. R. 26.1 on page 2 of this form.)

 

AMERICAN ELECTRIC POWER, INC., ET AL.,

 

Plaintiffs-Appellants

 

v.

 

UNITED STATES OF AMERICA,

 

Defendant-Appellee

 

 

DISCLOSURE OF CORPORATE AFFILIATIONS

 

AND FINANCIAL INTEREST

 

 

Pursuant to 6th Cir. R. 26.1, The Dow Chemical Company makes the following disclosure:

1. Is said party a subsidiary or affiliate of a publicly owned corporation? No

 

If the answer is YES, list below the identity of the parent corporation or affiliate and the relationship between it and the named party:

 

2. Is there a publicly owned corporation, not a party to the appeal, that has a financial interest in the outcome? No
If the answer is YES, list the identity of such corporation and the nature of the financial interest:
______________________

 

(Signature of Counsel)

 

Date: March 15, 2002

 

TABLE OF CONTENTS

 

 

STATEMENT OF DISCLOSURE OF CORPORATE

AFFILIATIONS AND FINANCIAL INTEREST

TABLE OF CONTENTS

TABLE OF AUTHORITIES

IDENTITY AND INTEREST OF AMICUS CURIAE AND SOURCE OF AUTHORITY TO FILE

STATEMENT OF FACTS

ARGUMENT

 

I. CONTRARY TO THE POSITIONS OF THE PARTIES IN THIS APPEAL, LOADING DIVIDENDS AND PARTIAL WITHDRAWALS ARE NOT FUNCTIONAL EQUIVALENTS WHEN USED TO PAY POLICY PREMIUMS IN SIMULTANEOUS NETTING TRANSACTIONS

 

A. The "4 Of 7" Safe Harbor Of IRC § 264 Permits The Shamming Of Loading Dividends, But Not Partial Withdrawals

B. Partial Withdrawals Netted Against Premium Obligations Have Real Consequences That Cannot Be Ignored, Whatever The Proper Treatment of Loading Dividends May Be

II. THE DISTRICT COURT ERRED IN HOLDING THAT THE "4 OF 7" SAFE HARBOR OF IRC § 264 REQUIRES THAT ANNUAL PREMIUMS BE LEVEL

III. THE DISTRICT COURT ERRED IN TESTING AEP'S COLI PROGRAM FOR "MORTALITY NEUTRALITY" RATHER THAN THE TRANSFER OF MORTALITY RISK

IV. THE DISTRICT COURT CORRECTLY HELD THAT POLICY LOANS MADE WITHOUT UNENCUMBERED CASH VALUE PRE-DATING THE LOANS ARE NOT FACTUAL SHAMS

CONCLUSION

CERTIFICATE OF COMPLIANCE

 

TABLE OF AUTHORITIES

 

 

CASES

American Electric Power Inc. v. United States, 136 F. Supp. 2d 762 (S.D. Ohio 2001), appeal pending 6th Cir. No. 01-3495

Campbell v. Cen-Tex, Inc. 377 F.2d 688 (5th Cir. 1967)

Comerica Bank, N.A. v. United States, 93 F.3d 225 (6th Cir. 1996)

Consumer Prod. Safety Commission v. GTE Sylvania, Inc., 447 U.S. 102 (1980)

Gitlitz v. Commissioner, 531 U.S. 206 (2001)

Golsen v. Commissioner, 54 T.C. 742 (1970)

Golsen v. United States, 80-2 USTC ¶ 9741 (Ct. Cl. 1980)

Hanover Bank v. Commissioner, 369 U.S. 672 (1962)

Helvering v. Le Gierse, 312 U.S. 531 (1941)

Horn v. Commissioner, 968 F.2d 1229 (D.C. Cir. 1992)

IES Indus., Inc. v. United States, 253 F.3d 350 (8th Cir. 2001)

In re CM Holdings, 254 B.R. 578 (D.C. Del. 2000), appeal pending, 3d Cir. No. 00-3875

Lerman v. Commissioner, 939 F.2d 44 (3rd Cir. 1991)

Rose v. Commissioner, 868 F.2d 851 (6th Cir. 1989)

Sears, Roebuck and Co. v. Commissioner, 972 F.2d 858 (7th Cir. 1992)

Shirar v. Commissioner, 916 F.2d 1414 (9th Cir. 1990), rev'g 54 T.C.M. (CCH) 698 (1987)

Tax Analysts v. Internal Revenue Serv., 117 F.3d 607 (D.C. CM 1997)

United Parcel Service of America, Inc. v. Commissioner, 254 F.3d 1014 (11th Cir. 2001)

United States v. Consumer Life Insurance Co., 430 U.S. 725 (1977)

Walls v. Waste Resource Corp., 823 F.2d 977 (6th Cir. 1987)

Winn-Dixie Stores v. Commissioner, 113 T.C. 254 (1999), aff'd, 254 F.3d 1313 (11th Cir. 2001)

Woodson-Tenent Labs., Inc. v. United States, 454 F.2d 637 (6th Cir. 1972)

STATUTES

26 U.S.C. (Internal Revenue Code)

I.R.C. § 264

IRC § 803

IRC § 805

IRC § 808

IRC § 7702

IRC § 7702A

Health Insurance Portability and Accountability Act of 1996, Pub. L. No. 104-191, § 501, 110 Stat. 1936, 2090 (1996)

LEGISLATIVE HISTORY

H.R. 4389, 101st Cong., 2d Sess. § 1(b) (1990)

S. 2722, 101st Cong., 2d Sess. § 1(b) (1990)

H.R. Rep. No. 88-749 (1963)

S. Rep. No. 88-830 (1964)

Pub. L. 88-272, § 215(a), 78 Stat. 19, 55

Staff of the Jt. Comm. on Internal Revenue Taxation, Comm. Print, 88th Cong., 1st Sess. 61-64 (1963)

TREASURY REGULATIONS AND ADMINISTRATIVE PRONOUNCEMENTS

Section 1.264-4

PLR 9812005 (Jan. 22, 1998)

1998 FSA LEXIS 412 (Feb. 23, 1998)

OTHER AUTHORITIES

C. A. Williams, Jr., M. L. Smith, & P. C. Young, Risk Management and Insurance 86 n.3 (7th ed. 1995)

Ernst & Young LLP, 1 Federal Income Taxation of Life Companies § 8.02[1] (2d ed. 2001)

Paul A. Samuelson, "Risk and Uncertainty: A Fallacy of Large Numbers," Scientia (1963), reprinted in 1 Collected Scientific Papers of Paul A. Samuelson 153-58 (J. E. Stigletz ed. 1966)

William Feller, 1 An Introduction to Probability Theory and Its Applications 251 (John Wiley & Sons ed., 3rd ed. 1968)

 

IDENTITY AND INTEREST OF AMICUS CURIAE AND SOURCE OF

 

AUTHORITY TO FILE

 

 

[1] The Dow Chemical Company is a Delaware corporation with its principal place of business in Midland, Michigan. Dow and an affiliated group of companies are plaintiffs in a suit against the United States in the U.S. District Court for the Northern Division of the Eastern District of Michigan (Case No.: CV 10331 BC Judge David Lawson) for the refund of approximately $11 million in federal income taxes for tax years 1989-91. This suit involves the deductibility of policy loan interest under broad-based corporate-owned life insurance ("broad-based COLI") and thus raises issues related to those presented in this appeal. As a result, Dow is vitally interested in this appeal.

[2] Dow relies on leave of court for this brief to be filed.

 

STATEMENT OF FACTS

 

 

[3] The sham doctrines applied by the court below require a searching inquiry into the facts to determine whether the transactions under review occurred and, if so, had any practicable economic consequences apart from their tax effects. So that this Court does not unintentionally address issues on this appeal that were not developed in the findings below, it is important to explain that not all broad-based COLI programs are alike.

[4] According to the findings below, AEP's program (1) adopted borrowing and cash value crediting rates in excess of industry standards (136 F. Supp. 2d 762, at 774, 790); (2) dated the first-year loans earlier than originally stipulated in certain "prepayment agreements" with the insurer (id. at 781-82); (3) envisioned maximum borrowing and cash value withdrawals to achieve "zero net equity" (that is, zero unencumbered cash value) at the end of each policy year (id. at 777, 787); (4) at some point involved "an experience rating system to 'fine tune' the matching of death benefits to insurance charges" (id. at 788); and (5) relied on innovative "loading dividends" to pay premiums in policy years four through seven (id. at 782-84).

[5] In contrast, Dow's programs (1) used standard industry loan and cash value crediting rates; (2) dated the first-year loans as initially understood with the insurer: (3) contemplated substantial cash investment over the life of the policies by capping loans and withdrawals so as to create positive net equity; (4) transferred mortality risk to the insurer, without any agreement at any point to true up the cost of insurance and death benefits; and (5) used standard partial withdrawals of cash value to pay premiums in years four through seven.

 

ARGUMENT

 

 

I. CONTRARY TO THE POSITIONS OF THE PARTIES IN THIS APPEAL, LOADING DIVIDENDS AND PARTIAL WITHDRAWALS ARE NOT FUNCTIONAL EQUIVALENTS WHEN USED TO PAY POLICY PREMIUMS IN SIMULTANEOUS NETTING TRANSACTIONS.

 

[6] AEP argues that the loading dividends it received were real, in part because they were the functional equivalent of partial withdrawals, which are real. AEP Br. at 19, 54-57. The government, in response, argues that there are indeed no significant practical differences between loading dividends and partial withdrawals, but both of them are factual shams. U.S. Br. at 44-46.

[7] For two independent reasons, both parties are wrong in equating loading dividends and partial withdrawals. First, under the "4 of 7" safe harbor of section 264 of the Internal Revenue Code ("IRC § 264"), one can ignore as shams premiums paid using the proceeds of policy benefits only if the parties to the insurance contract have agreed that the real premiums are the premiums stated in the contract less the benefits received. The parties to a policy that provides for the payment of artificially inflated loading dividends as a premium rebate make such an agreement, but the parties to a policy that merely permits the policyholder to make partial withdrawals to pay premiums do not.

[8] In addition, the test for a factual sham in a simultaneous netting transaction is not mere circularity of cash flow, but rather whether each leg of the transaction has separate and real economic effects that the other leg does not cancel out. Loading dividends used to pay premiums do not pass this test, but partial withdrawals do.1

A. The "4 Of 7" Safe Harbor Of IRC § 264 Permits The Shamming Of Loading Dividends, But Not Partial Withdrawals.
[9] The "4 of 7" safe harbor of IRC § 264 states that the disallowance of interest deductions on systematic borrowing to pay premiums does not apply:

 

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 [of purchase] by means of indebtedness . . . .

 

(Emphasis added.)2The clear meaning of "the annual premiums due," evidence of Congressional intent, administrative promulgations by the IRS, and practice in the life insurance industry all show that the "4 of 7" test applies on the basis of the agreed-upon consideration for the policy, not the agreed-upon consideration less policy benefits like traditional, nonguaranteed dividends or partial withdrawals.

[10] Anyone familiar with life insurance knows that traditional dividends and withdrawals are policy benefits like death benefits and that "the annual premiums due" are the contractual consideration charged for those benefits, however they are used. Indeed, the Internal Revenue Code itself distinguishes between premiums and benefits by including in an insurance company's income "[t]he gross amount of premiums" (IRC § 803(a)(1)(A)), while allowing a deduction of "claims," "policyholder dividends," and other "benefits" (IRC §§ 805(a)(1) and (3)), which include withdrawals.3 The annual premiums due" for purposes of the "4 of 7" rule cannot be the contract premiums reduced by death benefits, dividends, withdrawals, or any other benefit, even if the benefit is used to pay premiums. The "annual premiums due" instead are simply the consideration on which the parties have agreed for the policyholder to be eligible to receive death benefits and dividends, make withdrawals, and take advantage of other policy benefits.

[11] The legislative history of the "4 of 7" test confirms this interpretation. A description of the provision prepared by the staff of the Joint Committee on Internal Revenue Taxation in 1963, while the legislation was under consideration, concretely demonstrates that taxpayers can pay premiums with policy benefits other than loans without running afoul of the "4 of 7" rule. The description provides two examples of compliance with the "4 of 7" safe harbor when the taxpayer paid the annual premiums in policy years two through seven by means of both cash and an allocation of traditional annual dividends declared by the insurer. See Staff of the Jt. Comm. on Internal Revenue Taxation, Comm. Print 88th Cong., 1st Sess. 61-64 (1963). 4 Moreover, the language of the report underscores the fundamental difference between "the annual premiums due" and the net amount paid by the insured: the total amount the insured must pay over to the insurance company each year . . . is determined by taking the annual premium decreased by the amount [of annual dividends] in column 2." Id. at 62 (emphasis added).5

[12] The government's own regulations reflect a similar understanding that the statutory language "the annual premiums due" means "the annual premiums due," "the stated annual premiums due on a contract," the "annual gross premium[s]," or "the aggregate of premiums due for the year [when premiums are not annual]." Treas. Reg. §§ 1.264-4(b) Example, -4(c)(1)(ii), -4(d)(1), and -4(d)(iv), Example 1. Nothing the regulations even remotely suggests that "the annual premiums due" are the agreed-upon premiums less policy benefits if those benefits are used to pay the premiums.

[13] Finally, subsequent legislative developments also corroborate this interpretation of section 264.6 In 1990, Congresswoman Kennelly and Senator Pryor sought to amend section 264 to restrict the use of dividends and withdrawals to pay premiums under the "4-of-7" safe harbor. This proposed legislation would have amended section 264 to provide that:

 

the payment of more than 25 percent of any annual premium by the direct or indirect application of any policy dividend, distribution, or surrender proceeds shall be deemed a payment under a plan . . . by means of indebtedness.

 

H.R. 4389, 101st Cong., 2d Sess. § 1(b) (1990); S. 2722, 101st Cong,. 2d Sess. § 1(b) (1990). This legislation would have reduced, but not eliminated, the taxpayer's right, consistently with the "4 of 7" rule, to use dividends and withdrawals to pay premiums. By choosing not to pass even this limited incursion into the policyholder's right of access to internal policy funds, Congress confirmed that using traditional dividends and cash withdrawals to pay premiums is compatible with the "4 of 7" safe harbor.

[14] The recognition that one may use policy benefits to pay premiums does not require the blind endorsement of a rebate of inflated premiums in the guise of a policy benefit. According to the findings below, the loading dividends purported to be traditional dividends, but in actuality amounted to a guaranteed refund of premium made simultaneously with the premium payment. 136 F. Supp. 2d at 782-83.

[15] Premiums paid by partial withdrawals have fundamental differences. First, these premiums contain none of the fictitious charges inserted into the premiums in the loading dividend policies; in other words, they are the real charge for the benefits provided by the policies. Put another way, the insurer includes benefits like partial withdrawal options in the pricing of the policy along with expenses and profit. Second, while loading dividends are a novel departure from industry practice, partial withdrawals are strongly rooted in the tradition of providing access to cash value in whole life contracts through change provisions, the surrender of paid-up additions, and the withdrawal of accumulated dividends. Moreover, more than two decades ago, traditional whole life insurance evolved into universal life, under which cash value inside the policy functions much as a savings account from which the policyholder can freely withdraw. No actuarial, contractual, or legal requirement has ever limited withdrawals to previously accumulated cash value. To the contrary, policyholders have long had the legal right to make withdrawals months, weeks, or seconds after depositing the cash in the policy that is to be withdrawn.

[16] Finally, unlike loading dividends that are "virtually guaranteed" by the carrier, partial withdrawals are an option that the policyholder is free to exercise or not in its discretion. Even when the insurer knows that the policyholder is contemplating withdrawals to pay premiums, the insurer has no assurance, let alone a contractual right to require, that the policyholder will do so. It remains entirely possible that, as a result of altered circumstances (such as a change in interest rates) or simply a change in mind, the tax advantages afforded inside buildup will cause the policyholder to refrain from making withdrawals to pay premiums. For all these reasons, partial withdrawals, unlike loading dividends are not agreed-upon premium rebates.

[17] A comparison of the IRS's treatment of the "4 of 7" rule in a 1998 technical advice memorandum ("TAM")and a 1998 Field Service Advice ("FSA") confirms this analysis. 7 The TAM held that the "annual premiums due" in circumstances essentially identical to those in this appeal were the annual premium on the specifications page of the policy reduced by the loading dividends" because that was the consideration actually agreed upon by the parties:

 

Neither Taxpayer nor Insurer intended that the gross annual premium specified in each COLI policy actually be paid during each of the first seven contract years. Rather, it was contemplated from the outset that these contractually prearranged loading dividends would reduce the premiums actually required to be paid in years in which no borrowing occurred.

 

Private Letter Ruling 9812005, at 114-15 (Jan. 22, 1998).

[18] At the same time, the FSA rejected "the possibility" of applying the TAM argument to reduce the "annual premiums due" under section 264 by partial withdrawals used to pay premiums. The FSA conceded that "[t]he analysis in the TAM and the statute and regulations cited therein do not readily lend themselves to an argument that 'annual premiums due' should be reduced by the withdrawals . . . ." 1998 FSA LEXIS 412 (Feb. 23, 1998) (emphasis added). Apparently for this reason, the government in the first broad-based COLI case ever tried conceded that the taxpayer had complied with the "4 of 7" test despite the fact that the taxpayer had availed itself of its partial withdrawal rights to pay the premiums due in policy years four through seven after having borrowed to pay the first three premiums. See Winn-Dixie Stores v. Comm'r, 113 T.C. 254, at. e.g., 260, 291 (1999) aff'd 254 F.3d 1313 (11th Cir. 2001). On the appeal of that case, the Eleventh Circuit added its imprimatur to that concession by stating that the taxpayer's "loans fell within the 4-of-7 exception, all agree." 254 F.3d at 1315.

[19] It follows that the statute itself precludes the shamming of premiums by withdrawals even though it may allow the shamming of premiums paid through loading dividends of the type described the opinion below. Unlike loading dividends, partial withdrawals are not artificially preengineered premium rebates that may be set off against the stated premiums to determine the agreed upon consideration for the contract. As a result, unlike the case with a loading dividend policy, "the annual premiums due" remain level for purposes of the "4 of 7" safe harbor when partial withdrawals are used to pay premiums during the first seven policy years.

 

B. Partial Withdrawals Netted Against Premium Obligations Have Real Consequences That Cannot Be Ignored, Whatever The Proper Treatment of Loading Dividends May Be.

 

[20] When one pays premiums with partial withdrawals, the legs of the transaction have separate and real economic consequences that do not cancel each other out. When one "pays" premiums with loading dividends, the legs of the transaction do not have such consequences. This is a separate reason why partial withdrawals are significantly different from loading dividends.

[21] Simultaneously netting cash flows is a well-established business practice in the life insurance industry no less than in business generally, and the simultaneous offset of cash flows against one another cannot mean that neither occurred. Indeed, the Internal Revenue Code demonstrates that netting by itself provides no evidence of insubstantiality of premium payments. IRC § 808, for example, provides insurers a deduction for policyholder dividends, including dividends that are credited toward premium payments. Section 808(e)(2) states: "For purposes of this part, any policyholder dividend which . . . reduces the premium otherwise required to be paid, shall be treated as paid to the policyholder [and hence eligible for deduction] and returned by the policyholder to the company as a premium.

[22] The mere circularity of cash flow cannot be the test for whether a simultaneous netting transaction is a factual sham, or every transaction involving circular cash flow would be a sham. In the dividend illustration just given, the circular cash flow does not by itself mean that the insurer did not pay the policyholder the dividend or that the policyholder did not pay the premium in full. Recognition of the premium payment is necessary to prevent the policy from going into non-forfeiture status, and the full premium, less the insurer's expenses and profit, is credited to policy cash value. At the same time, the dividend reduces the insurer's divisible surplus, while providing the policyholder a non-guaranteed improved cash flow. The reality or unreality of these transactions obviously depends on something other than the circularity of cash flow -- namely, whether the legs of the transaction have separate and real economic effects that do not cancel each other out.

[23] Based on the findings below, the loading dividends did not satisfy this test of reality because they were considered to be nothing more than a promised, simultaneous rebate of premiums, having no independent economic significance. Payment of the difference between the nominal premium stated in the policy and the net premium after rebate did not secure any additional benefit for the taxpayer, impose any additional risk or obligation on the insurer, or otherwise make any economic difference at all. This increment did not even incur a state premium tax because of the deduction typically allowed for dividends.

[24] In contrast, partial withdrawals used to pay premiums entail separate and real economic consequences for both the premiums being paid and the withdrawals being made. The full premium is subject to state premium tax. At the same time, the partial withdrawal results in an immediate reduction in death benefits, unless the insurer agrees to defer the reduction in return for a fee. The reality of the partial withdrawals and premium payments in the Winn-Dixie case discussed above was so obvious that the Tax Court did not have to undertake any analysis to determine that they "actually occurred." 113 T.C. at 278-79.

[25] For this reason as well, the Court should reject the assertion by the parties to this appeal that partial withdrawals and loading dividends are functional equivalents.

 

II. THE DISTRICT COURT ERRED IN HOLDING THAT THE "4 OF 7" SAFE HARBOR OF IRC § 264 REQUIRES THAT ANNUAL PREMIUMS BE LEVEL.

 

[26] The court below held that "[i]mplicit in the four out of seven safe harbor rule is the requirement that the annual premiums must remain level for the first seven years of the policy." 136 F. Supp. 2d at 783. The court based this conclusion entirely on a concession by AEP and the holding to that effect in In re CM Holdings, 254 B.R. 578, 645-46 (D.C. Del. 2000), appeal pending, 3d Cir. No. 00-3875, where the court also found that the policyholder had not argued the point.8 This ruling does not withstand even minimal analysis.

[27] Section 264 contains no requirement that the taxpayer pay level premiums during the first seven policy years. To the contrary, the statute explicitly envisions both substantial and insubstantial premium increases that are consistent with the "4 of 7" test. The statute states:

 

if there is a substantial increase in the premiums on a contract, a new 7-year period . . . [for applying the "4 of 7" test] with respect to such contract shall commence on the date the first such increased premium is paid.

 

IRC § 264(d) (emphasis added). Congress adopted this safeguard "to prevent avoidance of this provision by taking out a contract with very low premiums for the first 4 years, with the premiums [paid by indebtedness] being substantially greater thereafter. . . ." H.R. Rep. No. 88-749, at 62 (1963) & S. Rep. No. 88-830, at 79 (1964). Congress did not at the same time adopt a rule for declining premiums because section 264 in another provision already disallowed deduction of interest on indebtedness to acquire "a single premium" policy -- that is, a contract under which "substantially all the premiums on the contract are paid within a period of 4 years from the date on which the contract is purchased . . . ." IRC §§ 264(a)(2), (b)(1).9 Obviously, these provisions reflect Congress' understanding that policies might provide for a wide range of premium levels without the variation in and of itself requiring disallowance of the taxpayer's interest deductions.

[28] Government regulations bolster this conclusion. Treas. Reg. § 1.264-4(c)(1)(ii) provides (emphasis added):

 

if the stated annual premiums due on a contract vary in amount, borrowing in connection with any premium, the amount of which exceeds the amount of any other premium, on such contract may be considered borrowing to pay premiums for more than one year.

 

This regulation purports to authorize the IRS to attribute borrowing in one year to multiple years as it wishes, when premium levels vary on a policy. Nothing on the face of section 264 or in the legislative history supports such unbridled discretion. But even if one assumes that the regulation is valid, the use of permissive, rather than mandatory language (i.e., "shall be considered"), establishes that level premiums are not a required element of the statutory "4 of 7" rule. See also Treas. Reg. §§ 1.264-4(d)(1)(i).

[29] In CM Holdings, the trial judge insisted that the level premium requirement was necessary to prevent the policyholder from circumventing section 264 "by taking out loans to pay large premiums in the first three years, and then paying very small cash premiums in the next four years." 254 B.R. at 646. But Congress could easily have framed the "4 of 7" test to impose a level premium requirement. It plainly did not do so, and now there is no license to rewrite the statute. See Gitlitz v. Commissioner, 531 U.S. 206, 219-20 (2001) (following the "plain text" of the Code, the Supreme Court refused even to address the IRS Commissioner's "policy concern" that the taxpayers would enjoy "a 'double [tax] windfall"').

[30] Furthermore, section 264 need not be interpreted to impose a level premium requirement in order to accomplish its purposes. The provisions on substantial premium increases and single premium policies already limit the taxpayer's ability to front-load or back-load premiums against which large loans might be taken.10 It also would be illogical from a tax policy perspective to read a level premium requirement into section 264, at least in the case of partial withdrawals. That would be tantamount to requiring policyholders to retain a greater investment in tax- advantaged inside buildup -- which would serve only to decrease tax revenue.11 Thus, not only is there no level premium requirement in section 264, but to import one on the grounds urged by the district judge in CM Holdings would be bad policy.

 

III. THE DISTRICT COURT ERRED IN TESTING AEP'S COLI PROGRAM FOR "MORTALITY NEUTRALITY" RATHER THAN THE TRANSFER OF MORTALITY RISK.

 

[31] The district judge recognized that mortality protection could imbue a life insurance policy with economic substance independently of the accumulation of inside buildup. But he erred in holding that the relevant mortality feature of a sham policy is "mortality neutrality" rather than the absence of transfer of mortality risk. See 136 F. Supp. 2d at 787.

[32] Mortality neutrality is a novel test of economic substance. According to the court below, it means that "it [is] anticipated that over the life of the plan, the death benefits [will] equal COI [i.e., cost of insurance] charges, minus [the insurer's] modest profit on COI." 136 F. Supp. 2d at 787 (emphasis added). Read literally, this cannot be the standard for whether the mortality protection of a life insurance policy is an economic sham because every policy would flunk it. Premiums always are set so that the policyholder, on an expected actuarial basis, pays more in premiums than it receives in benefits over the duration of the plan. "Insurance companies indeed do not make a habit of issuing policies whose premiums do not exceed the claims anticipated. . . ."United Parcel Service of America, Inc. v. Comm'r, 254 F.3d 1014, 1018 (11th Cir ,2001). Otherwise, they would go out of business.

[33] The economic substance doctrine examines "whether the transaction has any practicable economic effects other than the creation of income tax losses." Rose v. Comm'r, 868 F.2d 851, 853 (6th Cir. 1989) (emphasis added). Even if there is a truing up of death benefits and COI at the end of a broad-based COLI program, which typically is decades after its inception, there will be such economic effects as a result of annual mortality variances because actuarially expected mortality is never achieved. Moreover, these variances include substantial absolute deviations from expectation because of the size of the group. "[A]s the size of the pool increases the law of large numbers takes over, and the ratio of actual to expected loss [relative variance] converges on one." Sears, Roebuck and Co. v. Comm'r, 972 F.2d 858, 863 (7th Cir. 1992). But "the absolute size of the expected variance [absolute variance] increases," id., making it a fallacy to believe that "the absolute level of losses becomes more predictable in large groups." C. A. Williams, Jr., M. L. Smith, & P. C. Young, Risk Management and Insurance 86 n.3 (7th ed. 1995). In fact, the opposit, is what occurs. Cf. United Parcel Service, 254 F.3d at 1018 ("Insurance companies . . . do not make a habit of issuing policies whose premiums do not exceed the claims anticipated, but that fact does not imply that insurance companies do not bear risk").12

[34] Protection against fluctuation in mortality is a very important practicable economic effect of broad-based COLI even if there is an arrangement for truing up death benefits and COI by the end of the program. Indeed, the benefit that corporations seek from insurance often is "to spread the costs of casualties over time." Sears, Roebuck and Co., 972 F2d. at 862. In other words, "[b]ad experience concentrated in a single year, which might cause [depressed earnings or] bankruptcy . . . , can be paid for over several years." Id. The transfer of the risk of such "financial variance" (id. at 863) is itself sufficient to qualify a contract as real insurance. See United States v. Consumer Life Insurance Co., 430 U.S. 725, 738 (1977) (sustaining reinsurance treaties against an economic sham attack partly on the ground that "[e]ven though the reinsurer would eventually recapture . . . [any catastrophic] losses, it would be of substantial benefit to the ceding company to spread those payments out over a period of months or years").

[35] In short, the trial court's focus on the relationship between COI and death benefits over the life of the plan ignores the protection afforded by COLI against financial variance. Only when the program provides for full retrospective experience rating by incorporating a mechanism or an agreement for truing up the death benefits and the COI over a relatively short period of time does the mortality feature eliminate risk transfer and turn out to be economically meaningless.

[36] The mortality neutrality test adopted below also ignores other important practicable effects arising from the purchase of a broad-based COLI policy that is less than fully retrospectively experience rated. Because mortality rates shift over time and population variations, no mortality table provides a completely accurate basis for calculating COI. This may work in favor of or against the policyholder. Furthermore, broad-based COLI provides protection against catastrophic loss. In Winn-Dixie Stores, 113 T.C. at 284, the court dismissed the possibility of such loss as "so improbable as to be unrealistic and therefore . . . [of] no economic significance." This dismissal seems cavalier in the wake of the events of September 11, 2001 -- particularly for a company, like Dow, whose employees are engaged in dangerous activities at concentrated locations.

[37] Fundamentally, the error committed below was the adoption of the mortality neutrality test to begin with. The test is novel; it also is completely unnecessary. A well-established test for whether an insurance policy has practicable economic effects apart from the creation of tax deductions already exists. It is the test for real insurance established in Helvering v. Le Gierse, 312 U.S. 531 (1941). The Supreme Court there analyzed whether a life insurance policy purchased by an elderly woman constituted "insurance" that would enable the proceeds to pass tax free to her daughter. The issue was complicated by the fact that the insurer had sold the woman an annuity at the same time as the policy, thereby assuring that the insurer would avoid loss. The Court held that the policy was not insurance on the ground that "insurance involves risk-shifting and risk-distributing," which the woman's policy did not because of the annuity. Id. at 539-42. Thus, if a contract transfers risk from the policy owner to the insurer, then by definition it is a real life insurance policy with real economic effects.

[38] It is not clear from the findings below whether AEP's policies shifted risk. According to the trial judge, the insurer "instituted an experience rating system to 'fine tune' the matching of death benefits to insurance charges . . . ." 136 F. Supp. 2d at 788. If this amounted to full retrospective experience rating, then there was no risk transfer once the system was implemented. Otherwise, there was. See Sears, Roebuck and Co., 972 F2d. at 862 (concession by the IRS that policies less than fully retrospectively rated constitute insurance for tax purposes); see also IES Indus., Inc. v. United States, 253 F.3d 350, 355 (8th Cir. 2001) (a transaction is not "a sham for tax purposes merely because it does not involve excessive risk").

[39] In sum, if AEP's policies were less than fully retrospectively experience rated, they satisfied the test for real insurance and thus the test for economic substance, irrespective of any accumulation of inside buildup.

 

IV. THE DISTRICT COURT CORRECTLY HELD THAT POLICY LOANS MADE WITHOUT UNENCUMBERED CASH VALUE PRE-DATING THE LOANS ARE NOT FACTUAL SHAMS.

 

[40] The district court ruled that the policy loans AEP used to pay premiums in policy years one through three were real obligations. 136 F. Supp. 2d at 780-781. The government urges this Court to overturn this ruling as an alternative basis for sustaining the judgment below. The government principally argues that AEP's loans were factual shams because they were created in a "circular" transaction in which the premiums paid by the loans created the cash value that secured the loans. U.S. Br. at 51. The government's position is simply wrong.13

[41] First, the "4 of 7" safe harbor of section 264 establishes that Congress did not regard policy loans as factual shams, whether or not they are made on the security of cash value pre-dating the loan. Congress adopted the safe harbor to deal with life insurance (known as "minimum deposit insurance") in which "the taxpayer each year borrows all, or a substantial part, of the funds necessary to pay the premium on the policy." H.R. Rep. No. 88-749, at 61 & S. Rep. No. 88-830, at 78 (emphasis added). These loans were (and because of the substantial borrowing each year could only have been) created in the same fashion that the government now decries.14 Yet Congress recognized that the interest incurred on these loans was deductible under existing law.15 Moreover, the "4 of 7" test in no way altered the deductibility of interest on loans, with or without pre-existing net equity, other than to limit borrowing under new policies to three of the first seven policy years.16 Congress could not have made it any plainer that it did not regard loans made without pre-existing unencumbered cash value as shams or intend to preclude the deduction of all interest incurred on them.

[42] At bottom, the government's error in asserting otherwise stems from its refusal to admit that loans and the premiums paid by them have separate and real economic consequences that do not cancel each other out. Payment of the premiums causes the policy to go into and remain fully in effect; increases the cash value and thus the amount of inside buildup; gives rise to state premium tax liabilities; in the case of increasing death benefit policies, raises the death benefits and the net amount at risk (i.e., the death benefit less the cash value); but, in the case of level death benefit policies, decreases the net amount at risk. Borrowing, on the other hand, reduces the net amount due the policyholder on the death of the insured so long as the loan remains unpaid. Borrowing also gives rise to interest expense, which, if left unpaid, eventually causes the policy to terminate. Furthermore, policy loans can be repaid at any time, thereby stopping the accrual of interest, increasing the net amount due on the death of the insured, and permitting even greater growth in net policy equity.

[43] Finally, the government's exegesis of the case law is misleading. The courts, after a searching inquiry into the facts, overwhelmingly upheld minimum deposit plans against an economic sham attack, which necessarily implies that the loans occurred.17

 

CONCLUSION

 

 

[44] For the foregoing reasons, to the extent that the Court reaches the issues addressed in this brief, it should recognize:

1. Contrary to the argument of both parties, partial withdrawals and loading dividends are not functional equivalents.

2. The "4 of 7" safe harbor of IRC § 264 permits the shamming of premiums paid through policy benefits only if the parties to the insurance contract have agreed that the real premiums are the stated premiums minus the benefits.

3. The test for a factual sham in a simultaneous netting transaction is not circularity of cash flow, but rather whether there are separate and real economic effects to each leg of the transaction that do not cancel each other out.

4. The "4 of 7" safe harbor does not require the payment of level annual premiums.

5. The mortality feature of a life insurance contract that by itself imbues the policy with economic substance is the transfer of mortality risk, not the absence of mortality neutrality.

6. Policy loans made without unencumbered cash value pre-dating the loans are not factual shams.

Respectfully submitted,

 

 

John B. Magee

 

(Lead Counsel)

 

Richard C. Stark

 

Gerald Goldman, Of Counsel

 

McKee Nelson LLP

 

1919 M Street, N.W.

 

Washington, D.C. 20036

 

(202) 775-1880

 

 

Counsel for

 

The Dow Chemical Company

 

 

Valorie A. Gilfeather

 

The Dow Chemical Company

 

2030 Dow Center

 

Midland, MI 48675

 

(989) 636-6400

 

 

Of Counsel

 

CERTIFICATE OF COMPLIANCE

 

 

[45] I certify that this brief complies with the type-volume limitation set forth in Federal Appellate Rules 32(a)(7)(B) and 29(d). The Microsoft Word 2000 program used to prepare this brief states that the brief contains 6,743 words beginning with the section entitled "IDENTITY AND INTEREST OF AMICUS CURIAE AND SOURCE OF AUTHORITY TO FILE" and ending with the section entitled "CONCLUSION."
Gerald Goldman

 

FOOTNOTES

 

 

1Contrary to the suggestion in the government's brief (at 44-45), the court below nowhere considered the reality of partial withdrawals used to pay premiums. In determining that AEP's mechanics for paying the year four through seven premiums were a factual sham whether they were labeled dividends or withdrawals, 136 F. Supp. 2d at 785, the court only addressed the facts of AEP's case, in which the premiums in those years were found to have been artificially inflated and then by agreement simultaneously rebated on the date they were due. See also id. at 785 (discussing the reality of partial withdrawals used to pay interest).

2The "4 of 7" safe harbor, originally codified at IRC § 264(c)(1), now appears at IRC § 264(d)(1).

3See Ernst & Young LLP,1 Federal Income Taxation of Life Insurance Companies § 8.02[1] (2d ed. 2001) ("Benefits incurred include death benefits. . . . surrender benefits, . . . and payments of dividend accumulations.").

4The examples are set out at pp. 63-64 of the staff report, but they incorporate the facts illustrated in the chart at p. 61, including the provision for annual dividends (column (2)).

5This legislative history also demonstrates that, contrary to the government's position in this appeal, Congress could not possibly have intended section 264 to require "the annual premiums due" to be the net premiums paid and also be level during the seven-year test period. If policy benefits used to pay premiums were to be subtracted from the gross premiums stated in the contract to determine "the annual premiums due," the premiums under participating policies routinely would vary in amount, since, as recognized in the legislative history, traditional dividends are regularly credited toward the policyholder's next premium due. Thus, the government's position would deny the typical owner of a participating policy the benefit of the "4 of 7" safe harbor -- a result that Congress could not possibly have wanted. For other discussion regarding level premiums, see pp. 15-19 below.

6 Although subsequent legislative history may furnish a hazardous basis for determining Congressional intent (see Consumer Prod. Safety, Comm'n v. GTE Sulvania, Inc., 447 U.S. 102, 118 n. 13 (1980)), it can provide "useful guidance," Walls v. Waste Resource Corp., 823 F.2d 977, 981 (6th Cir. 1987), and in this case serves to substantiate the obvious meaning of the statute.

7 TAMs, like letter rulings, are internal advice by the IRS National Office concerning application of the tax law to a specific set of facts. Although they are not binding precedent, they may properly be considered in determining the law because they "reveal the interpretation put upon the statute by the agency charged with the responsibility of administering the revenue laws." Hanover Batik v. Commissioner, 369 U.S. 672, 686 (1962); see also Comerica Batik, N.A. v. United States, 93 F.3d 225, 230 (6th Cir. 1996). An FSA, which is internal advice rendered by the National Office of Chief Counsel, also has no precedential value, but is useful in elucidating IRS interpretations of the Code. See Tax Analysts v. Internal Revenue Serv., 117 F.3d 607, 617-18 (D.C. Cir. 1997) (FSAs are "considered statements of the agency's legal position," furthering uniformity "throughout the country significant question of tax law")

8Both the appellant in this case and the appellant in CM Holdings have nevertheless contested the ruling on appeal. See AEP Br. at 62-64; Appellant's Br., In re CM Holdings, 3d Cir. No. 00-3875, at 47-54.

9Congress specifically provided for the single premium rule to survive unaffected by the systematic borrowing rule. See, e.g., IRC § 264(a)(3); H.R. Rep. No. 88-749, at A61.

10See also IRC §§ 7702 and 7702A (restricting the accumulation of cash value and hence the level of potential borrowing); Treas. Reg. § 1.264-4(d)(1)(ii) (rules for attributing borrowing, to multiple policy years when the taxpayer borrows more in any one year than the premium due in that year).

11The only reason that the government is now able to argue for a level premium requirement without fear of compromising the public fisc is that Congress phased out all interest deductions on broad-based COLI loans in 1996. See Health Insurance Portability and Accountability Act of 1996, Pub. L. No. 104-191, § 501, 110 Stat. 1936, 2090 (1996).

12To illustrate the point, suppose that one wins a dollar if a coin flip comes up heads and loses a dollar if it comes up tails. On the first coin flip, the possible outcomes range from +$1 (a heads) to -$1 (a tails) -- an absolute variance of $2. On the second flip, this variance increases to $4 because the range expands to +$2 (a heads followed by a heads) to -$2 (a tails followed by a tails). Because of the additional outcomes outside the absolute variance on the first flip, the probability of being within that variance on the second flip actually decreases. For a seminal discussion of this subject, see Paul A. Samuelson, "Risk and Uncertainty: A Fallacy of Large Numbers," Scientia (1963), reprinted in 1 Collected Scientific Papers of Paul A. Samuelson 153-58 (J. E. Stigletz ed. 1966) (The Law of Large Numbers "is often given an invalid interpretation. Thus people say an insurance company reduces its risk by increasing the number of ships it insures.") (quotation at 153). See also William Feller, 1 An Introduction to Probability Theory and Its Applications 251 (John Wiley & Sons ed., 3rd ed. 1968) (the insurer "plays a large number of games, but because of the large variance the chance fluctuations are pronounced"). The observations by Samuelson and Feller apply no differently to the owner of a large number of policies than to the insurer.

13Quoting Horn v. Comm'r, 968 F.2d 1229, 1236 n.8 (D.C. Cir. 1992), the government also argues that AEP's loans violated "assumptions of commercial dealing." U.S. Br. at 58. The district court, however, correctly ruled that "shams in fact" are simply "reported transactions" that "never occurred." 136 F. Supp. 2d at 779-80. Horn's statement of the standard for a sham in fact not only was pure dictum, but also was fabricated from whole cloth. The only precedent cited in Horn was Lerman v. Comm'r, 939 F.2d 44, 49 n.6 (3rd Cir. 1991) (dictum), which stated merely that "[a] factual sham is one in which the alleged transactions never actually took place."

14A description of the legislation prepared by the staff of the Joint Committee on Internal Revenue Taxation while the legislation was under consideration lays bare the elements of "minimum deposit insurance." Staff of the Jt. Comm. on Internal Revenue Taxation, supra, at 61. These elements included a pre- tax amount paid annually by the policyholder, consisting of "the total amount the insured, must pay over to the insurance company each year after taking into account the 'loan' of the increase in the cash surrender value, the policyholder dividends, the cost of term insurance, and the interest paid on the loan." Id. at 62 (emphasis added).

15See H.R. Rep. No. 88-749, at 61 ("under present law, no interest deductions are denied where the taxpayer purchases an insurance contract [other than a single-premium policy] with the intention of borrowing the maximum amount . . . each year"); S. Rep. No. 88-830, at 77 (same).

16See Pub. L. 88-272, § 215(a), 78 Stat. 19, 55; H.R. Rep. No. 88-749, at 62 (the legislation "will have no effect on [pre-existing] contracts"); S. Rep. No. 88-830, at 79 (same); see also Staff of the Jt. Comm. on Internal Revenue Taxation, supra, at 63 (example 1).

17See, e.g., Shirar v. Comm'r, 916 F.2d 1414 (9th Cir. 1990), rev'g 54 T.C.M. (CCH) 698 (1987), Woodson-Tenent Labs., Inc. v. United States, 454 F.2d 637 (6th Cir. 1972); Campbell v. Cen-Tex, Inc., 377 F.2d 688 (5th Cir. 1967). Also, compare Golsen v. United States, 80-2 U.S.T.C. paragraph 9741 (Ct. Cl. 1980), with Golsen v. Comm'r, 54 T.C. 742 (1970).

 

END OF FOOTNOTES
DOCUMENT ATTRIBUTES
  • Case Name
    AMERICAN ELECTRIC POWER COMPANY, INC. AND AFFILIATED CORPORATIONS AND AEP SERVICE CORPORATION, Appellants, v. UNITED STATES OF AMERICA, Appellee.
  • Court
    United States Court of Appeals for the Sixth Circuit
  • Docket
    No. 01-3495
  • Authors
    Magee, John B.
    Stark, Richard C.
    Madan, Rajiv
    Goldman, Gerald
    Gilfeather, Valorie
  • Institutional Authors
    McKee Nelson LLP
    The Dow Chemical Co.
  • Cross-Reference
    American Electric Power, Inc., et al. v. United States; 87 AFTR2d

    Par. 2001-529; No. C2-99-724 (20 Feb 2001)(For a summary, see Tax

    Notes, Feb. 26, 2001, p. 1198; for the full text, see Doc 2001-5282

    (26 original pages) or 2001 TNT 36-8 Database 'Tax Notes Today 2001', View '(Number'.)
  • Code Sections
  • Subject Area/Tax Topics
  • Industry Groups
    Insurance
  • Jurisdictions
  • Language
    English
  • Tax Analysts Document Number
    Doc 2002-10049 (37 original pages)
  • Tax Analysts Electronic Citation
    2002 TNT 86-78
Copy RID