CCA 200629028: Proposed nuclear decommissioning coverage is not insurance

Plain-English summary

The IRS notified a field office that two taxpayers withdrew a request after receiving an adverse view of proposed nuclear plant decommissioning contracts. A proposed subsidiary would reimburse operators for a portion of decommissioning costs. Chief Counsel reasoned that decommissioning is inevitable once a plant operates, so the contracts did not cover uncertainty about whether a loss would occur. The uncertainty concerned how much and when operators would spend, which the advice classified as business and estimation risk rather than insurance risk. It concluded that the proposed contracts would not be insurance for federal income tax purposes.

Ruling snapshot

• Question: Would contracts covering nuclear plant decommissioning costs be insurance for federal tax purposes?
• Outcome: Advice given after withdrawal, the contracts would not qualify.
• Key authorities: IRC § 831; Rev. Rul. 68-27; Rev. Rul. 89-96.

Full text (IRS public release)

Office of Chief Counsel

Memorandum
Internal Revenue Service

Number: 200629028
Release Date: 7/21/2006
CC:FIP:B04:JEGLOVER
Third Party Communication: None
POSTN-119663-06
Date of Communication: Not Applicable

UILC: 831.00-00

date: April 14, 2006
to: Keith Jones
Industry Director (Natural Resources & Construction)

from: Donald J. Drees, Jr.
Senior Technician Reviewer, Branch 4
(Financial Institutions & Products)

subject: Notification of Withdrawn Letter Ruling Request

LEGEND

Taxpayer 1 = --------------------------------------------------------------------- Taxpayer 2 = --------------------------------------------- NEWCO = ---------
A = ------------
Number 1 = ---

Pursuant to § 7.07(2)(a) of Rev. Proc. 2005-1, 2005-1 I.R.B. 1, 26, this is to notify you that Taxpayer 1 and Taxpayer 2 have withdrawn a letter ruling request after we reached conclusions adverse to those requested and to provide you our view on an issue raised in the request.

Taxpayers requested a ruling that proposed contracts to be issued by NEWCO would constitute insurance for federal income tax purposes. We concluded that the contracts would not.
POSTN-119663-06
2

FACTS:

In order to operate a nuclear power plant, a license from the United States Nuclear Regulatory Commission (NRC), among other government instrumentalities, is required. 42 U.S.C. § 2131 (2005). This license has an initial term of 40 years and is renewable, 42 U.S.C. § 2133(c); 10 C.F.R. §§ 54.31(b) and (d).

The obligation to decommission an operational plant is concomitant with the privilege of operating it. Since the Atomic Energy Act of 1946, licensees have been required to satisfy safety standards to protect life and property. Atomic Energy Act of 1946, Pub. L. No. 79-585, § 7(c), 1946 U.S.C.C.S. 722, 731.1 Since the Atomic Energy Act of 1954, the terms and conditions of a license are subject to amendment, revision, or modification by rules and regulations issued in accordance with the Act. Atomic Energy Act of 1954, Pub. L. No. 83-703, § 187, 1954 U.S.C.C.A.N. 1076, 1098 (codified at 42 U.S.C. § 2237). Since 1961 the NRC (through its predecessor) has had the authority to require applicants seeking to terminate a license to describe the
proposed procedures for disposal of radioactive material, decontamination of the site, and other procedures, to provide reasonable assurance that the dismantling of the facility and disposal of the component parts will be performed in accordance with the regulations in this chapter and will not be inimical to the common defense and security or to the health and safety of the public.

26 Fed. Reg. 9546-47 (1961). See also H.R. Conf. Rep. No. 861, 98th Cong., 2d Sess. at 877 (1984), 1984-3 (Vol. 2) C.B. 1, 131 (“Generally, under Federal and State laws, utilities that operate nuclear power plants are obligated to decommission the plants at the end of their useful lives.”). Finally, in 1988 the NRC promulgated final regulations “to provide specific requirements for the decommissioning of nuclear facilities.” 53 Fed. Reg. 24018 (1988).

A licensee must periodically report to the NRC evidence of assurance that funds will be available to decommission the plant. 10 C.F.R. § 50.75(f). A licensee must report to the NRC the permanent cessation of operation of the plant and generally must complete decommissioning within 60 years thereafter. 10 C.F.R. § 50.82(a).

“Decommissioning” is defined as

1 This requirement was repeated in §§ 103(b) and 104(d) of the Atomic Energy Act of 1954, Pub. L. No. 83-703, 1954 U.S.C.A.A.N. 1076, 1098-99 (codified at 42 U.S.C. §§ 2133(b) and 2134(d)).
POSTN-119663-06
3
to remove a facility or site safely from service and reduce residual radioactivity to a level that permits (1) release of the property for unrestricted use and termination of the license; or (2) release of the property under restricted conditions and termination of the license.

10 C.F.R. § 50.2. The variables that influence the ultimate cost of decommissioning fall into four categories:

1. time: when do operations at the plant cease;
2. extent of actual contamination;
3. changes in regulatory requirements for decommissioning (e.g., standards and procedures); and
4. the economic conditions at the time of decommissioning (e.g., the cost of labor and supplies).

Although the costs of decommissioning cannot be presently known with certainty, industry experience provides sufficient data to build reliable models of the timing and amount of such costs. Accordingly, the costs of decommissioning a given plant can be estimated; the certification of assurance of adequate funding required by the NRC may be based on an estimate. E.g., 10 C.F.R. § 50.75(b)(4).

Congress enacted § 468A of the Internal Revenue Code as part of The Deficit Reduction Act of 1984, Pub. L. No. 98-369, 98 Stat. 494, 604-06, 1984-3 (Vol. 1) C.B. 1, 112-114. This section permits an eligible taxpayer, defined by § 1.468A-1(b)(1), to establish a Nuclear Decommissioning Reserve Fund (a.k.a. “qualified decommissioning trust”, or “QDT”) which is to function as “[a] segregated reserve dedicated exclusively for the payment of nuclear decommissioning costs, taxes on fund income, and management costs of the fund.” H.R. Conf. Rep. No. 861, 98th Cong., 2d Sess. at 877, 1984-3 (Vol. 2) C.B. at 131-32. In 2005, the Congress amended § 468A to provide an additional tax incentive for plant owner/licensees to fund QDTs. To supplement a QDT, an eligible taxpayer can establish a “nonqualified decommissioning trust fund” (a.k.a. a “NQDT”). The NQDT is not governed by § 468A and is commonly treated as a grantor trust.

In the situation presented by Taxpayers, Taxpayer 2 is presently responsible for the decommissioning of an operating nuclear power plant(s). Taxpayer 1 was to form NEWCO as a wholly owned subsidiary. NEWCO was to issue a contract covering decommissioning costs incurred by entities such as Taxpayer 2. NEWCO would issue Taxpayer 2 a contract only if it could issue contracts covering at least Number 1 power plants, no one such plant accounting for more than 10% of the total risk NEWCO would assume. For the payment of a specified charge, each NEWCO contract would have
POSTN-119663-06
4

indemnified each contractholder for a fixed percentage of the costs incurred to decommission a specified plant, in excess of a ‘deductible’ amount, up to a fixed limit.2

The contract would entitle the contractholder to reimbursement for incurred decommissioning costs regardless of when or for what reason decommissioning occurred. The beneficiary of the NEWCO contract will be the decommissioning trust (i.e., the QDT or NQDT as appropriate) or as otherwise designated.

Estimates of the cost to be incurred to decommission the specified plant were included in the formula to have been used to determine the specified charge for each NEWCO contract. Each NEWCO contract provided that the charge would be reduced if subsequent developments reduced the assurance amount required by regulators.3 The specified charge would not have been increased if subsequent developments increased the assurance amount.

The ‘deductible’ and coverage limit of the NEWCO contracts depended on the timing of decommissioning. If operation of the specified plant ceased (such that decommissioning began) at, or prior to, the expiration of the current license period, the ‘deductible’ and coverage limit were certain amounts. If operation of the specified plant continued beyond the current license period (i.e., under a renewed license), the ‘deductible’ and coverage limit were different certain amounts.4 However, the aggregate amount of NEWCO’s liability was capped: the NEWCO contract provided that the resources available to pay claims on NEWCO contracts were premiums, surplus, and reinsurance recoverable. In the event claims for covered decommissioning costs against NEWCO exceeded these resources, the amount payable to any one contractholder would have been limited to an amount determined by reference to a ratio of the contractholder’s decommissioning costs to the aggregate decommissioning costs of all contractholders. The contractholder would have to return to NEWCO the excess.

NEWCO’s obligation to pay each contractholder was certain because of the obligation to decommission each plant.

2 Taxpayers describe the NEWCO contracts as involving “coinsurance” or “quota share”. Harvey W. Rubin, DICTIONARY OF INSURANCE TERMS 2d ed. (Barron’s 1991) defines “coinsurance” as being a characteristic of property insurance where claims are limited to an amount equal to the actual loss multiplied by the ratio of insurance carried over insurance required, Id., at 74, and defines “quota share reinsurance” as “automatic reinsurance which requires the insurer to transfer and the reinsurer to accept a given percentage of every risk within a defined category of business … The percentage is constant throughout and applies to premiums and losses alike”, Id., at 337. An example of NEWCO’s coinsurance/quota share mechanism: assume a contract has a ‘deductible’ of $50x, a limit of $500x, and a coinsurance/quota percentage of 50. If contractholder incurs $600x of covered decommissioning costs, NEWCO will pay $275x (($600x – ‘deductible’ of $50x) X 50%). 3 Such a reduction would result in a refund if the specified charge had been paid in full or, if the specified charge was to be paid in installments, in a reduction of future installments due. 4 If the contractholder breached its obligations under the contract or cancelled the contract, the coverage limit was limited to the consideration paid less any distributions.
POSTN-119663-06
5

NEWCO would have complied with A’s applicable regulatory requirements.

LAW AND ANALYSIS:

Law

Neither the Code nor the regulations thereunder define the terms “insurance” or “insurance contract.” The bedrock for evaluating whether an arrangement constitutes insurance is Helvering v. Le Gierse, 312 U.S. 531, 539 (1941), in which the Court stated that “historically and commonly insurance involves risk - shifting and risk – distributing” in “a transaction which involve[s] an actual ‘insurance risk’ at the time the transaction was executed.” Insurance has been described as “involv[ing] a contract, whereby, for adequate consideration, one party agrees to indemnify another against loss arising from certain specified contingencies or perils…[I]t is contractual security against possible anticipated loss.” Epmeir v. United States, 199 F.2d 508, 509-10 (7th Cir. 1952). Cases analyzing “captive insurance” arrangements have distilled the concept of “insurance” for federal income tax purposes to three elements, applied consistently with principles of federal income taxation:5 1) involvement of an insurance risk; 2) shifting and distribution of that risk; and 3) insurance in its commonly accepted sense. See, e.g., AMERCO, Inc. v. Commissioner, 979 F.2d 162, 164-65 (9th Cir. 1992), aff’g 96 T.C. 18 (1991).

For an arrangement to constitute insurance, the “insured” must have an insurable interest in the subject matter of the policy at the time the policy is written. See, e.g., Greer v. United States, 408 F.2d 631, 636 (6th Cir. 1969) (“The law is well settled that the subject matter of an insurance policy must be in existence when the policy is written and the insured must have an insurable interest in the subject matter of the policy.”) The risk transferred must be risk of economic loss. Allied Fidelity Corp. v. Commissioner, 572 F.2d 1190, 1193 (7th Cir.), cert. denied, 439 U.S. 835 (1978). The risk must contemplate the fortuitous occurrence of a stated contingency, Commissioner v. Treganowan, 183 F.2d 288, 290-91 (2d Cir.), cert. denied, 340 U.S. 853 (1950) and must not be merely an investment risk. Le Gierse, 312 U.S. at 542; Rev. Rul. 89-96, 1989-2 C.B. 114.

Risk shifting occurs when a person facing the possibility of economic loss transfers some or all of the financial consequences of the potential loss to the insurer such that a loss by the insured does not affect the insured because the loss is offset by a payment from the insurer. Clougherty Packing Co. v. Commissioner, 811 F.2d 1297, 1300 (9th Cir. 1987); Rev. Rul. 2005-40.

Risk distribution incorporates the statistical phenomenon known as the law of large numbers. Distributing risk allows the insurer to reduce the possibility that a single costly claim will exceed the amount taken in as a premium and set aside for the

5 These principles include respecting the separateness of corporate entities, the form and substance of the transaction(s), and the relationship between the parties. Sears, Roebuck and Co. v. Commissioner, th
96 T.C. 61, 101-02 (1991), aff’d in part and rev’d in part, 972 F.2d 858 (7 Cir, 1992).
POSTN-119663-06
6

payment of such a claim. Insuring many independent risks in return for numerous premiums serves to distribute risk. By assuming numerous relatively small, independent risks that occur randomly over time, the insurer smoothes out losses to match more closely its receipt of premiums. See Clougherty Packing Co., 811 F.2d at 1300. Risk distribution necessarily entails a pooling of premiums, so a potential insured is not in significant part paying for its own risks. See Humana v. Commissioner, 881 F.2d 247, 257 (6th Cir. 1989), aff’g in part and rev’g in part 88 T.C. 197 (1987); Rev. Rul. 2005-40.

The “commonly accepted sense” of insurance derives from all of the facts surrounding each case, with emphasis on comparing the implementation of the arrangement with that of known insurance. Court opinions identify several nonexclusive factors bearing on this, such as the treatment of an arrangement under the applicable state law, AMERCO, Inc., 96 T.C. at 42; the adequacy of the insurer’s capitalization and utilization of premiums priced at arm’s length, The Harper Group v. Commissioner, 96 T.C. 45, 60 (1991), aff’d 979 F.2d 1341 (9th Cir. 1992); separately maintained funds to pay claims, Ocean Drilling & Exploration Co. v. United States, 24 Cl. Ct. 714, 728 (1991), aff’d per curiam, 988 F.2d 1135 (Fed. Cir. 1993); and the language of the operative agreements and the method of resolving claims, Kidde Indus. Inc. v. Commissioner, 40 Fed. Cl. 42, 51-52 (1997).

It is important to bear in mind that the type of risk contemplated by life insurance is different than that by non-life (a.k.a. property and casualty) insurance:

…..Although all forms of insurance are alike in that they require a combination of many risks into a group, they differ with regard to the perils covered. A peril is a cause of loss, such as fire or windstorm with respect to property or an accident with respect to health. In the nonlife forms of insurance, perils insured against may or may not happen and, in the great majority of cases, do not happen. In life insurance, the event against which protection is granted – death – is an uncertainty for one year, but the probability of death generally increases with age until it becomes a virtual certainty. If a life insurance policy is to protect an insured during the whole of his or her life, an adequate fund must be accumulated to meet a claim that is certain to occur.

Kenneth Black, Jr. and Harold D. Skipper, Jr., LIFE INSURANCE 19-20 (12th ed. 1994).

Insurance is a somewhat nebulous concept. “[T]he subject has no useful, or fixed definition. There is neither a universally accepted definition or concept of ‘insurance’ nor a [sic] exclusive concept or definition that can be persuasively applied in insurance lawyering.” 1 APPLEMAN ON INSURANCE 2d, § 1.3 (2005). While “it seems appropriate that any concept and meaning of insurance be sufficiently broad and
POSTN-119663-06
7

flexible to meet the varying and innovative transactions which humankind perpetually produces”, care must be used to describe insurance because “overbroad definitions are not useful and may cause many commercial relationships erroneously to constitute insurance.” Id. It is commonly understood that insurance is the mechanism to manage the risk of loss from fortuitous events. Insurance is not the mechanism to manage losses that are at least substantially certain to occur, i.e., that are not the result of fortuitous events.6 This principle has various labels, and “embod[ies] the concept that one may not obtain insurance for a loss already in progress, or for a loss that the insured either knows of, planned, intended, or is aware is substantially certain to occur.” 43 Am. Jur. 2d Insurance, § 479 (2005); see also COUCH ON INSURANCE 3d, § 102:8 (1997). Put another way, “[t]he fortuity principle is central to the notion of what constitutes insurance. The insurer will not and should not be asked to provide coverage for a loss that is reasonably certain or expected to occur within the policy period.” 1 APPLEMAN ON INSURANCE 2d, § 1.4. Perhaps better described:

Fortuity is another key element in determining what constitutes insurance for purposes of legal classification. It would be foolhardy for insurance companies to sell insurance that would pay for losses strictly within an insured’s control…This is the point where the concept of fortuity comes into play. Insurance is designed to cover the unforeseen or at least unintentional damages arising from risks encountered in life and business: injuries and damages caused by negligence and other similar conduct where the insured stands to sustain a real and palpable loss (generally pecuniary) as a result of the event for which the insurance has been purchased.

Id., § 1.3.

One commentator describes an insurable risk as having four elements, among them that the loss must be fortuitous or accidental. Emmett J. Vaughn, FUNDAMENTALS OF RISK AND INSURANCE 28-29 (3d ed. 1982). Vaughn explains that for an event to be fortuitous,

[i]t must not be something that is certain to happen. If the insurance company knows that an event in the future is

6 One treatise notes that “[t]here are many contractual devices, legally valid, by which persons seek assurance and peace of mind regarding future events. These contracts of assurance have distinctive names, such as guaranty, warranty, suretyship, indorsement, pledge, mortgage, conditional sale, indemnity, and insurance. Collectively, they have a common purpose in protecting one against the harmful consequences of untoward future events. But it is frequently difficult to distinguish one from the other. Consequently, it is no facile matter to frame a definition which states accurately and plainly the common features of the enterprises that are generally regarded as subject to ‘insurance’ regulation.” 1 APPLEMAN ON INSURANCE 2d, § 1.3.
POSTN-119663-06
8
inevitable, it also knows that it must collect a premium equal to the certain loss that it must pay, plus an additional amount for the expenses of administering the operation.

Id., at 29.

An insurance agreement is a contract, governed generally by contract law.
Accordingly, an insurer will be liable in contract if the insurer knew of the existing loss at the time the contract was placed in force. See COUCH ON INSURANCE 3d, § 102:8. The question is whether such a contract constitutes insurance for federal income tax purposes.

In Treganowan, 183 F.2d 288, the court of appeals held that a program under which the surviving members of the New York Stock Exchange paid a certain sum to the families of deceased members constituted insurance; this program was in essence a term life insurance policy. The court distinguished the holding of Le Gierse that when the two contracts in that case were put together there was no ‘insurance risk’. The holding [of Le Gierse] really highlights the situation here where the payment is actually conditioned upon death, whenever occurring, in the true terms of insurance. “From an insurance standpoint there is no risk unless there is uncertainty, or, to use a better term, fortuitousness. It may be uncertain whether the risk will materialize in any particular case. Even death may be considered fortuitous, because the time of its occurrence is beyond control.” 8 Ency.Soc.Sc. 95. That fortuitousness, whether we speak of death generally or premature death, as the Tax Court wished to emphasize, seems perfectly embodied here to fit both branches of the Supreme Court’s test.

Treganowan, 183 F.2d at 290-91.

The Tax Court has described “insurance risk”:

Basic to any insurance transaction must be risk. An insured faces some hazard; and insurer accepts a premium and agrees to perform some act if or when the loss event occurs.
If no risk exists, then insurance cannot be present.
“Insurance risk” is required; investment risk is insufficient. If parties structure an apparent insurance transaction so as to effectively eliminate the effect of insurance risk therein,
POSTN-119663-06
9
insurance cannot be present. [Le Gierse] illustrates these points.

AMERCO, Inc., 96 T.C. at 38-39.

Rev. Rul. 68-27, 1968-1 C.B. 315, holds that a medical services contract does not constitute an insurance contract because the predominant portion of the issuer’s expenses were not of a type “other than that which it incurs in providing the medical services through a salaried staff of physicians, nurses, and technicians”. The contract provided for indemnification of the costs for needed services the issuer was unable to provide; “although an element of risk exists, it is predominantly a normal business risk of an organization engaged in furnishing medical services on a fixed-price basis, rather than an insurance risk.”

Rev. Rul. 89-96 considered a circumstance in which Y “incurred a liability to injured persons the exact amount of which could not be ascertained, but was expected to be substantially in excess of $130x.” Unfortunately, Y had insurance coverage with a limit of $30x; Y was vulnerable to claims in an indeterminable amount predicted to be substantially greater than an additional $100x. Y entered into an agreement with Z in an attempt to cover $100x of this shortfall; Z was fully aware of Y’s circumstances. The “premium” charged Y was an amount that, together with the tax savings claimed by Z, was calculated to yield at least Z’s maximum anticipated liability of $100x by the time claims were liquidated. The ruling concluded that the risk elements borne by Z were a timing risk (that the $100x would have to be paid out earlier than anticipated) and an investment risk (that the actual investment yield would be lower than forecast). The ruling concludes that these risks are not an insurance risk.7

Accordingly, what insurance protects against is “the variability of loss.” Ocean Drilling, 24 Cl. Ct. at 729. The premium charged for the insurance protection is intended to cover the average loss and the insurer assumes “the variability of loss, not the complete loss from an event, such as a hurricane or an accident, since the insured pays a premium that covers the average cost of the complete loss.” Id. S.S. Huebner, et al,

7 See also Beech Aircraft Corp. v. United States, 84-2 U.S.T.C. ¶ 9803 (D. Kan. 1984), aff’d 797 F.2d th
920 (10 Cir. 1986) which involved an arrangement whereby Beech established a Bermuda-licensed company to cover its product liability risk. In the first year of its operation, Beech paid this company $1.5 million for two layers of coverage: losses up to $2 million and $3 million in excess of $12 million. The arrangement was similar for subsequent years. The company had “determined that this premium, when invested during the period of time it would hold the money, would equal the total amount it might pay for any losses occurring under” the contract. Beech argued the arrangement constituted insurance because it covered pure risk, which Beech defined as “the variance of actual results from the expected outcome of hazard.” The uncertainties were whether any accidents would occur and the severity of the associated losses. “[L]osses might have to be paid earlier than expected, reducing investment income. “ Beech acknowledged that “the time of payment risk is traditionally viewed as an investment rather than an insurance risk” and was “unable to identify any insurance authority who agreed with [its] statement that time of payment risk is an insurance risk as opposed to an investment risk.”
POSTN-119663-06
10

explain this concept in PROPERTY AND LIABILITY INSURANCE (4th ed. 1996) at 3-5: what insurance does is allow the substitution of a certain loss (the premium) for an uncertain loss. Instead of reserving for the effect of a potential casualty, in exchange for the premium an insured can shift the risk of the potential casualty to the insurer. The insurer combines this risk with other similar risks such that the law of large numbers yields a more certain outcome.

In fact, if the aggregate of risks combined in a group were sufficient to make the operation of the law of large numbers perfect, removing all uncertainty about the amount of losses that would be experienced during a given period of time, the accumulation of money through premiums….could be limited to the exact amount of the expected losses.

Id., at 4. Working from this more certain outcome, the insurer can charge each entity a rate based on its “proper share of the aggregate cost of insurance”, Id. at 621 (Chapter 38, Ratemaking); Vaughn, supra, at 91-95; that is, a rate based on the probability and severity of loss. Scott E. Harrington and Gregory R. Niehaus, RISK MANAGEMENT AND INSURANCE 134-37 (2004).

Analysis

For the NEWCO contract to have constituted insurance it must, among other things, have transferred an insurance risk. When a nuclear power plant is placed in operation, it is inevitable that the licensee will incur the cost of decommissioning the plant when operation ceases. The obligation to decommission has attached therefore no hazard or fortuity as to the occurrence of decommissioning exists. The licensee does not bear the risk of whether decommissioning will occur – that is inevitable. It is certain that under each of its contracts, NEWCO would have been required to pay benefits; in insurance parlance, it is certain that NEWCO would have paid for each and every risk exposure unit of each and every insured.

Decommissioning is not dissimilar from other types of unavoidable costs incurred in conducting a business for which the business must estimate and make provision. The licensee bears the risk that the method it used to estimate and discount its inevitable decommissioning costs was inaccurate; i.e., that when the time comes to decommission, the licensee will have set aside too little money to enable it to fulfill its obligation. This risk antedated the NEWCO contract. The risk to have been covered by the NEWCO contracts was this risk of inaccurate cost estimation, a business risk akin to a normal business risk Rev. Rul. 68-27 concludes is not an insurance risk and akin to the timing and investment risks Rev. Rul. 89-96 concludes are not insurance risks.

Accordingly, the proposed contracts to be issued by NEWCO would not constitute insurance for federal income tax purposes.
POSTN-119663-06
11

Please contact CC:FIP:B04 on --------------------- if you have any questions.

Pursuant to § 6110(k)(3), this document may not be used or cited as precedent.