UPWARD REVALUATION OF LIFE INSURANCE POLICY RESERVES IS LIMITED TO TERM POLICIES IN FORCE FOR MORE THAN 15 YEARS.

MEMORANDUM

The numerous stipulations of the parties in this matter are hereby adopted in full as the factual basis for the conclusions of the court. Except where necessary to comprehend this court's conclusions of law, these stipulations will not be repeated. Any other necessary findings of fact will be placed in the corresponding discussion of law to which that finding is relevant. Additionally, the court's analysis and conclusions of law will follow the same six- issue organizational structure used by the parties. Furthermore, obvious legal conclusions assumed or granted by both parties will generally not be restated within this opinion.

ISSUES 1 AND 2

For purposes of this decision, Issues 1 and 2 may be analyzed together as one issue. While the two issues technically present different types of insurance contracts, both present the same issue of statutory construction which this court views to be dispositive. Basically, the dispute arises over an Internal Revenue Code provision which authorizes, for tax purposes, a revaluation of reserves by life insurance companies which compute reserves on a preliminary term basis rather than through the use of a net level method. This provision, Internal Revenue Code section 818(c) provides as follows:

(c) Life Insurance Reserves Computed on Preliminary Term Basis. -- For purposes of this part (other than section 801), at the election of the taxpayer the amount taken into account as life insurance reserves with respect to contracts for which such reserves are computed on a preliminary term basis may be determined on either of the following bases:

(1) Exact revaluation -- As if the reserves for all such contracts had been computed on a net level premium rate for both the preliminary term basis and the net level premium basis.

(2) Approximate revaluation -- The amount computed without regard to this subsection --

(A) increased by $21 per $1,000 of insurance in force (other than term insurance) under such contracts, less 2.1 percent of reserves under such contracts, and

(B) increased by $5 per $1,000 of term insurance in force under such contracts which at the time of issuance cover a period of more than 15 years, less 0.5 percent of reserves under such contracts.

Having established that subsections A and B are part of the same approximate revaluation formula, it then becomes important to note that only application of the second component of the formula is directly in dispute. This is the component which authorizes an upward revaluation of reserves set aside for term insurance. The plaintiff contends that section 818(c)(2)(B) authorizes the upward adjustment for all term insurance reserves, regardless of the time the term insurance was to be in force, as long as the entire insurance policy as a whole contains some insurance in force for a period greater than 15 years. The government, on the other hand, argues that subsection B authorizes only the upward adjustment for term insurance reserves which were computed for term insurance in force for more than 15 years. In other words, the plaintiff believes the 15-year period mentioned in subsection B refers to the life insurance policies as a whole, while the defendant asserts the period refers to just the term insurance portion of the life insurance policy. The problem arises because the statutory language is somewhat ambiguous concerning whether the word "insurance" or the word "contracts" is modified by the clause "which at the time of issuance cover a period of more than 15 years." Narrowed down even further, the debate can be stated as whether the subject for the verb "cover" is "insurance" or "contract" because whichever noun is the subject for "cover" is clearly the noun modified by the time period.

A first impression/commonsense reading of the statute seems to more naturally convey the idea that the term insurance, and not the entire contract, is being described by the portion of the sentence which discusses something covering a period of more than 15 years. One aspect of the statutory scheme which helps to convey this impression was alluded to earlier when the court noted that subsections A and B of section 818(c)(a) are separate components of one single approximate revaluation formula. The breakdown of the formula into a segment dealing with term insurance and a segment dealing with "other than term" insurance reveals a legislative intent to treat the term insurance component of a policy separately from the "other than term" component. The fact that a single policy contained both term and "other than term" components did not prevent Congress from breaking down the single "combination" policy into its separate components for purposes of tax computations. Yet, to some extent, the plaintiff's position is one of arguing that a combination policy should not and cannot be broken down into its separate components for purposes of revaluing reserves computed on a preliminary term basis.

Theoretically, a combination policy containing a whole life component and a less-than-15-year-term component could be issued as two separate policies. And if such were done, the plaintiff would not even dream of contending that the reserves set aside under the less- than-15-year-term policy should be revalued pursuant to section 818(c)(2). However, because of the mere fact that in this case the two policies were combined into one, the plaintiff suddenly contends that the two types of insurance cannot be considered separately. Accordingly, the plaintiff also argues that the term insurance reserves can be revalued, despite their origin being in term insurance of 15 years or less, simply because the term insurance was issued as one part of a contract which did last for more than 15 years.

To the court, this distinction appears blatantly artificial and as nothing more than a clever way to get around the intended limits of the revaluing procedure. Congress' division of the revaluing formula into term and "other than term" components suggests the 15- year period mentioned in the term component of the formula was intended to be a limitation on what term insurance reserves could be revalued. It was not intended to be a loophole through which insurance companies could enhance term reserve revaluations simply by piggybacking term insurance onto other policies which last more than 15 years.

Nevertheless, the court must still recognize that a very technical grammatical examination of the sentence supports the proposition that the drafters were referring to whole contracts of which at least part covers a period of more than 15 years, not just term insurance in force for a period of more than 15 years. If the drafters intended the 15-year qualification to apply to term insurance, proper grammar would have required the use of the singular verb "covers," rather than the plural "cover," since insurance is a singular noun. On the other hand, since contracts is a plural noun, the statute's use of the verb "cover" was grammatically correct if the legislature was intending the 15-year period to qualify only the contract or policy as a whole without regard to the length of the term insurance itself (despite the fact that it is only the reserves associated with the term insurance which would be revalued in accordance with subsection B).

Nevertheless, this court is quite aware of the fact that proficiency in English grammar is not a prerequisite to election to Congress. For that matter, perhaps fortunately, neither is complete expertise in this area a prerequisite to nomination to the bench. Regardless, the point is that courts must not place so much weight on grammatical technicalities that the commonsense construction of a statute is forgotten. Similarly, and even more importantly, the courts must guard against using grammatical technicalities to derail the legislative intent behind a statute. Ultimately, legislative intent is preeminent as long as that intent does not go beyond the bounds of the Constitution. In fact, the plain language of a statute receives such focused attention only because the law presumes an idea conveyed by the plain language of a statute is the idea intended to be conveyed by the legislature. The same type of presumption lies beneath grammatical inquiries. These presumptions are in turn based on the assumption that legislators generally use proper grammar and generally intend for their words to express what those words are ordinarily assumed to express.

As basic as these principles may be, they also point to the weakness of the plaintiff's desperate reliance upon grammatical technicalities in this case. While it may be true that a technically correct grammatical version of the statute would have used the verb "covers" if the legislature had intended to convey the idea advanced by the defendant, this court is mindful of the fact that people very often use plural predicates with a singular subject, especially when a plural noun appears between the subject and predicate. In fact, in situations such as these, the proper grammatical rule may be honored more in the breach than in the performance. Thus, in this case the assumption that the legislative body would use proper grammar is suspect. Furthermore, even without resort to analysis of the purpose and function of the revaluation formulas as a whole, the statute as it appears may very easily be taken to mean what the defendant asserts it means. Accordingly, the court is hesitant to put extreme weight upon the plaintiff's very technical and grammatically based argument. Rather, the court will focus more on the intent behind the statute and a determination of which statutory construction best advances that intent.

To some extent, legislative intent has already been discussed. As pointed out earlier, the basic design of the statutory formula for revaluing reserves suggests that Congress intended "combination" policies to be broken down into their separate components for purposes of revaluation. This, in turn, indicates that the 15-year period mentioned in the formula component dealing with term insurance was intended to be a limitation on what term insurance reserves could be revalued. Beyond this point, a deeper understanding of the legislative intent underlying section 818(c) depends upon a greater understanding of the role of reserves in tax calculations in general.

In simplest essence, increases in reserves are treated as an expense which may be deducted from income, thereby reducing an insurance company's income tax. Reserves, however, can be computed in two different ways. Some companies use what is known as the "net level method." Others, such as the plaintiff, use the "preliminary term method." Under the net level method, a uniform portion of the policy premium is set aside for reserves each year. Under the preliminary term method, on the other hand, reserves for the first year are computed in a manner which recognizes and accounts for the higher first year expenses on a policy. This results in a lower reserve amount for the first year, and then a higher but uniform amount for all subsequent years covered by the policy. In other words, a lesser portion of the first year's premium paid by the insured is used for reserve purposes than in subsequent years. At first, then, the preliminary term method results in a lower reserve amount than does use of the net level method. Thus, the preliminary term method would also at first result in a smaller deduction in income and a larger tax. Eventually, of course, amounts not added to the reserve in the first year under the preliminary term method are made up by accumulating larger reserves in subsequent years of the policy, and at some point, the amount ultimately reserved under either method would equalize. Nevertheless, the fact remains that an insurance company using the preliminary term method starts out with a tax disadvantage in relation to companies which use the net level method.

Legislative history and the basic function of Code section 818(c) indicates that the statute was enacted to relieve preliminary term method companies from this tax disadvantage by allowing them to revalue their reserves for tax purposes. As the plain language of the provision indicates, section 818(c) allows preliminary term method companies to revalue reserves exactly as if they were using the net level method. Alternatively, the company can revalue reserves using a formula which approximates what the reserves would be if the net level method had been used. This formula is set out in section 818(c)(2), and the application of this formula is the heart of the controversy on Issues 1 and 2. While legislative history does not detail exactly how the formula should work in specific situations, it does make clear that the intent was to place preliminary term method companies on equal status with net level method companies as far as taxes are concerned.

H.R. Report No. 34, 86th Cong., 1st Sess., (1959), 1959-2 C.B. 736, 748, states in part as follows:

ELECTION FOR LIFE INSURANCE RESERVES COMPUTED ON PRELIMINARY TERM BASIS. -- Some life insurance companies compute their life insurance reserves on what is called a preliminary term basis. The effect of this is to take the full agents' commissions (which are larger in the initial period of a life insurance contract) out of amounts which would otherwise be added to reserves during the first year of a contract and to add correspondingly larger amounts to reserves in later years. The effect of this is to work a hardship on insurance companies using preliminary term reserves as compared with those which use ordinary reserves, since the policy and other contract liability deduction depends on the size of the reserves. Moreover, additions to the reserves, deductible under phase 2, also would in some cases be smaller. To avoid this result, life insurance companies which have computed their reserves on a preliminary term basis are permitted to recompute their reserves on a net level premium basis. This can be done either by an exact revaluation of the reserves to a net level premium basis or by approximating this result under a formula set forth in the bill.

Furthermore, this report from the House of Representatives is closely paralleled by a report from the Senate. With language very similar to the House Report, S. Rep. No. 291, 86th Cong., 1st Sess., (1959), 1959-2 C.V. 770, 792, states in pertinent part as follows:

ELECTION FOR LIFE INSURANCE RESERVES COMPUTED ON PRELIMINARY TERM BASIS. -- Some life insurance companies compute their life insurance reserves on what is called a preliminary term basis. The effect of this is to take the full agents' commissions (which are larger in the initial period of a life insurance contract) out of amounts which would otherwise be added to reserves during the first year of the contract and to add correspondingly larger amounts to reserves in later years. The effect of this is to work a hardship on insurance companies using preliminary term reserves as compared with those which use ordinary reserves, since the policy and other contract liability requirements which determine the policyholders' and life insurance company's shares of investment income depend on the size of the reserves. Moreover, premium additions to reserves, deductible under phase 2, also would in some cases be smaller. To avoid this result, life insurance companies which have computed their reserves on a preliminary term basis are permitted to recompute their reserves on a net level premium basis. This can be done either by an exact revaluation of the reserves to a net level premium basis or by approximating this result under a formula set forth in the bill.

Obviously, these two explanations by the respective houses of Congress are virtually identical in language and completely identical in concept. There can be no doubt that the whole intent of the statute was simply to put preliminary term method companies on an even par with net level method companies when it comes to the question of calculating income tax. Thus, when this court attempts to discern the correct application of the approximate revaluation formula in section 818(c)(2) to this case it should seek to apply the formula in a manner which best advances the intent of Congress to equalize insurance companies for tax purposes regardless of whether they use the preliminary term of uniform net level method of computing reserves.

This court cannot believe Congress intended the subsection B revaluation to be available for all term insurance reserves regardless of the time the term insurance was in force as long as the term insurance was part of a broader policy which included whole life insurance. To hold otherwise and to adopt the plaintiff's position would be a license for insurance companies to design a policy with a small whole life element of more than 15 years and a large term element of 15 years or less. The whole life element would then act as a tail which would wag the entire contract into a deduction far greater than that envisioned as necessary to approximate deductions available if the net level method had been used.

If a company issued a whole life policy and a separate term insurance policy covering a period of less than 15 years, it is undisputed that the reserves computed for the whole life insurance could be revalued in accordance with subsection A, but that the term insurance reserves could not be revalued. Furthermore, it is clear that this outcome is presumed by Congress to result in a reserve approximately the same as if the reserves for the two policies had been computed using the net level method rather than the preliminary term method. Yet, under the plaintiff's approach, merely stapling the two policies together would allow the company to revalue both the whole life and the term reserves, thus resulting in a reserve amount for tax purposes which is higher than the amount Congress believed approximated the net level amount. If policies were designed correctly, the disparity could be quite large. This may be demonstrated by an extreme, although admittedly not realistic, example. If a company issued a $1,000 whole life policy and a separate $500,000 term policy of less than 15 years, obviously only a small amount of reserves, the amount set aside for the $1,000 whole life policy, could be revalued. Yet, by stapling the policies together, according to the plaintiff's position, all the reserves could be upwardly revalued.

This position just does not make sense. Congress had no duty to allow any revaluation in the first place. Nevertheless, the body decided to allow revaluation in order to allow preliminary term method companies to be placed on an even par with net level method companies, at least insofar as taxes are concerned. This goal was supposedly accomplished by allowing a revaluation in accordance with the section 818(c)(2) formula. It is hard to believe, however, that Congress also intended the revaluation to be much greater in those situations where the insurance company fortuitously, or perhaps designedly, stapled two policies together. Such would do nothing to accomplish the objective of approximating net level method reserve results. The plaintiff's position is simply one of trying to take undue advantage of an opportunity which exists in the first place only by the grace of Congress. The bottom line is that the plaintiff seeks to gain a much higher revaluation than the approximation of net level reserve amounts intended by Congress. The court cannot allow such an attempt to succeed. ¹

Additionally, the court can find no logical reason for the length of the underlying policy as a whole to be relevant to an attempt to approximate term insurance reserves calculated under the preliminary term method with term insurance reserves calculated under the net level method. In this process of approximation, the length of other elements of insurance which might be included in the policy make no difference. When reserves are actuarially computed, reserves for term insurance obviously have to be computed with different formulas and figures than reserves computed for "other than term" insurance. Accordingly, the approximation formula set forth in section 818(c)(2) provides for a different type of adjustment for term reserves than for "other than term" reserves. This being so, the court fails to see how it can make any difference to the approximation attempt for term insurance reserves whether the "other than term" component of the policy is any certain length of time. The very structure of the approximate revaluation formula treats term reserves and "other than term" reserves separately. Thus, for purposes of revaluing the term reserves, it should not matter that the "other than term" portion of the policy is of any certain length. On the contrary, only the length of the term insurance period is relevant to an attempt to revalue term reserves in accordance with what the reserve amount would be if a different method of reserve computation had been used. Accordingly, the court concludes that only reserves for term insurance in force for more than 15 years may be revalued.

This same ultimate conclusion could be reached even if the court determined that the word "contracts" in section 818(c)(2)(B), rather than the word "insurance," was the subject for the predicate "covers." Given the statutory structure and policy reasons for dividing an insurance policy into its term and "other than term" components, each separate term and "other than term" component of an entire policy could easily be construed as a separate "contract" within the meaning of the code provision. This is in essence the holding of the tax court in National Savings Life Insurance Co. v. Commissioner, 84 T.C. 509, 543 (1985). Accordingly, regardless of whether "insurance" or "contracts" is qualified by the 15-year time period, this court concludes that Congress definitely intended to limit the revaluation of term insurance reserves to those reserves which were computed for term insurance in force for more than 15 years. The significance of the persuasive authority of the National Savings Life case is not found primarily in its grammatical analysis, but rather in the more fundamental proposition that "combination" policies can and should be broken down into their separate components for revaluation purposes. This is the same proposition this court gleaned from the basic rationale of the revaluation formula. Hence, the plaintiff will not be allowed to revalue the reserves at stake in Issues 1 and 2.

The court notes the plaintiff's attempt to inject state law into the equation and deems it irrelevant. State requirements for setting aside reserves have nothing to do with whether and how those reserves should be revalued for tax purposes. The plaintiff has complained that it is not capable of dividing the policies into their separate components and then revaluing reserves based upon those components. The court finds this complaint to be ridiculous. The revaluation is simply a process of computations on paper for tax purposes. It in no way hinders compliance with state law. Moreover, the government certainly appears to have had no trouble figuring out a way to revalue only that portion of reserves which stem from term insurance in effect for more than 15 years. Thus, the court cannot accept the plaintiff's claim that it cannot do the same. Nevertheless, even if such a breakdown were impossible, the plaintiff would still have no grounds to complain. Congress had no obligation to provide any revaluation, but out of kindness, or perhaps more cynically, out of lobbying pressure, Congress decided to grant preliminary term method companies the option of revaluation. In granting this revaluation option, Congress was free to design the formula and conditions as it saw fit. If the plaintiff cannot then bring itself within the confines of that formula and conditions, it must simply be content with reserve amounts which are not revalued at all. However, this court is extremely confident that the plaintiff will suddenly be able to comply with the constraints of the revaluation formula as articulated above when it becomes clear that the courts will not allow the complete revaluation sought by the plaintiff.

ISSUE 3

The court considers the third issue to be a much closer question than the first two. This third issue is whether amounts set aside as reserves for certain "certified insurability riders" to its life insurance and endowment policies are truly reserves. Obviously, if they are not reserves, the plaintiff cannot use these amounts in its calculation of reserve increases and consequential deductions in income for tax purposes. To qualify as reserves, the set-aside amounts must meet three separate conditions set forth in section 801(b) of the Internal Revenue Code. First, they must be computed or estimated on the basis of recognized mortality or morbidity tables and assumed rates of interest. Second, they must be set aside to liquidate future, unaccrued claims arising from certain insurance contracts. Third, they must be required by law to be maintained. See also Treas. Reg. 1.801-4(a). To the extent an amount claimed as a reserve fails to meet any one of the above criteria, it fails to constitute a life insurance reserve.

Only the first criterion is at issue in this case. The government concedes that state law required reserves to be set aside for the relevant policies and that the plaintiff had indeed set aside the amount for the purpose of liquidating future, unaccrued claims arising from those same insurance policies. However, the government contends that the amounts were not based upon recognized mortality tables. Accordingly, the government also claims that the amounts cannot be considered reserves.

Importantly, however, not all of the amounts reserved to liquidate future unaccrued claims incurred under the Certified Insurability Riders are under attack. The amounts can be divided into three separate components: (1) reserves to liquidate the inherent higher mortality risk of persons who have already exercised Riders and who are covered under policies issued as a result of that exercise, (2) reserves to fund liabilities under the term insurance in effect for the 90-day period prior to each option exercise date and (3) reserves to fund the mortality risk borne by the plaintiff under the unexercised options. The government accepts the second and third components as properly computed reserves, but rejects the first as not being computed on the basis of recognized mortality or morbidity tables and assumed rates of interest. However, since the first component is computed directly from the third component, which the government concedes to be actuarially proper, the court concludes the first component must also be proper.

When the plaintiff first issued a group of base policies which also contained Certified Insurability Riders, reserves were set aside not only for liquidating the liabilities which would accrue on the base policy, but also for the extra-mortality risk that would be associated with the unexercised options. Of course at that time it was impossible to know which policyholders would exercise their options when the time of decision arrived, but it was obvious that there would be an extra-mortality risk associated with the options since people with impaired health at the time of the option date would be more likely to exercise the option. In essence, the system worked as follows. From the time a base policy was issued until the option exercise date, a portion of the premiums for each policy would be set aside to liquidate the increased mortality risk that would be inherent in the yet unknown number of rider policies which would be issued in the future at the exercise date.

Of course, some of the base policies would never have their rider exercised. Yet, a portion of those base policy premiums would nevertheless be set aside for the increased risk associated with the riders that would eventually be exercised. This was necessary because of the obvious fact that it was impossible to know at that time which riders would be exercised. If the company had been omniscient, it could have set aside a larger portion of the base policy premiums from only those base policies which would have their rider exercised in order to cover the extra-mortality risk of the riders to be exercised in the future. It would then not have to set aside any portion of the premium on the base policies which would not have their rider exercised (at least not for the purpose of liquidating the extra-mortality risk associated with the riders). However, given that omniscience was not a corporate asset, the company was forced to simply set aside a portion of the premiums from all of the base policies. Of course, since the "base" was therefore larger, this also allowed for a smaller portion of each of the base policy reserves to be set aside in order to equal the same amount which would result from using a larger premium portion from only those base policies which would have their rider exercised. Significantly, the government accepts this calculated amount as a proper actuarially computed reserve amount for the extra-mortality risk associated with the riders until such time as the exercise date actually occurs.

At the time of the exercise date, the plaintiff's procedure was simply to take the amounts previously set aside from all the base policy premiums for the extramortality risk of the yet unexercised riders, and reclassify them as the reserves for the extra-mortality risk of the riders now that they were exercised. This certainly appears logical. After all, if it is accepted that the amount was properly computed to cover the extra-mortality risk of the riders before they were exercised, why is it not still the proper amount to cover that same risk now that the riders have been exercised?

In essence, what the government seeks is a "fine tuning" of the amount now that the exact number of exercised riders is known. The court, however, concludes that such fine tuning is not necessary in order to meet the definition of life insurance reserves. The debated segment of the definition simply requires computation based on acceptable actuarial data and principles. The government concedes that the amount in issue was indeed at one time (up until the rider exercise date) computed on acceptable actuarial data and principles. Furthermore, the risk for which that acceptable amount was originally computed is the same risk for which that amount is set aside at the time the government suddenly objects. The only difference is that the quantity of the risk is better known at the time of the government's objection. While the court recognizes some logic behind the government's position it nevertheless concludes that this difference is just not enough to exclude the amount from meeting the basic definition of life insurance reserves. Ultimately, the fact remains that the amount was computed upon data and principles which the government concedes to be actuarially proper. This is all the definition requires. There is no additional requirement of "fine tuning" every time some new information is available.

True, on some occasions, proper actuarial principles would require use of the new information. However, the court concludes that in this case the simple shift or reclassification of the amounts at the time of the exercise date was actuarially sound. If more actuarial data had been available at that time, perhaps proper actuarial principles would have required integration of the new information and the mortality data relating to riders of the type issued. However, at the time in question, the Certified Insurability Riders were a relatively new type of policy benefit in the the insurance industry. Accordingly, there was little or no reliable data on the mortality rates associated with the exercised riders. Consequently, the fact that a point arises at which the plaintiff could determine the exact number of riders which would be exercised was of little significance. Without more actuarial data, then not available, that fact did not shed significant light on the quantity of the extra-mortality risk for which reserves were needed. Accordingly, the plaintiff's method of using the pre-exercise reserves for the extra-mortality risk as the reserves for the extra- mortality risk after date of exercise was actuarially proper. The government's position on Issue 3 is rejected.

ISSUE 4

The fourth issue is also a question of whether certain amounts set aside as reserves are truly reserves within the meaning of the tax code. This time, however, the question is not close. Neither can it be resolved in favor of the plaintiff. As mentioned in the discussion of the third issue, one of the three criteria for an amount to qualify as reserves is that it must be "set aside to mature or liquidate, either by payment or reinsurance, future unaccrued claims arising from life insurance . . . contracts." 26 U.S.C. section 801(b). Despite what may have been the best intentions of the plaintiff, the "Phoenix Mutual Reserves" are not really reserves because of failure to meet this criterion.

The whole issue arises out of two retrocession reinsurance contracts with Phoenix Mutual Life Insurance Company (hereinafter "Phoenix"). Pursuant to those contracts, the plaintiff agreed to reinsure, on a "pass through" basis, a portion of insurance written by Equity Funding Life Insurance Company (hereinafter "Equity") which had been reinsured by Phoenix. The reinsurance contract with Phoenix required the plaintiff to establish reserves for the policies based on information furnished by Equity through Phoenix to the plaintiff. Accordingly, the plaintiff dutifully set aside amounts as "reserves" for what it thought was the underlying insurance policies.

Unbeknownst to the plaintiff, however, some of the insurance policies supposedly issued by Equity, reinsured by Phoenix, and subsequently reinsured by the plaintiff were actually fictitious policies. Thus, the amounts set aside as "reserves" for those "policies" would and could never be used to "mature or liquidate . . . claims arising from life insurance . . . contracts" because there were no real policies from which claims could arise. With no policies in force, there could be no liabilities or potential liabilities to which the "reserves" could apply. Accordingly, the amounts set aside as "reserves" for those fictitious policies do not meet the section 801(b) definition of reserves and therefore cannot be treated as reserves for tax purposes. The intent behind setting aside the amounts is irrelevant. It may be true that the plaintiff intended the amounts to be set aside for the purpose of liquidating future liabilities or claims arising from life insurance policies, but that does not change the all-important fact that there were no policies and that therefore there could be no policy liabilities for the amounts to liquidate.

This conclusion is also supported by the basic idea behind the role reserves play in tax computations. Allowing income deductions for reserve increases is basically a way of spreading over a period of time something which would otherwise be a sudden expense. In other words, for tax purposes, the reserve increase deduction serves as a mechanism for the expense of paying policy claims to be treated as an expense which accrues over time rather than simply as a large expense which the company pays only at the time a claim is filed. In the case of fictitious policies, however, the ultimate expense of paying a claim on a policy never occurs. Consequently, it makes no sense for a company to be able to take the tax benefit of allowing the "expense" to accrue over time. After all, there really is no expense to accrue.

ISSUE 5

The fifth issue is an outgrowth of the first two. It having been determined that the government properly disallowed the reserves disputed in Issues 1 and 2, the parties now debate the proper treatment of those disallowed reserves. More specifically, the parties disagree concerning the proper way for the government to recapture the tax lost as a result of the excessive reserve computations discussed in Issues 1 and 2. Excessive reserve computations naturally result in excessive reserve increases, and since the latter may be deducted from income, this also results in a smaller than proper tax being paid.

Increases in reserves are computed simply by subtracting a year's opening reserves from that same tax year's closing reserves. Thus, if only the tax year had involved excessive reserve computations, recapturing the lost tax would only involve removing the excessive reserves from both the opening and closing amounts, subtracting the opening from the closing to arrive at a proper reserve increase, deducting that proper increase from income, arriving at the proper tax on that income, and then assessing the taxpayer for the deficiency between the paid tax and the newly computed proper tax. Unfortunately, this case is not that simple. Rather, this case involves the recapture of taxes from years which cannot simply be reopened and recomputed as just discussed. The parties apparently agree that recapture is appropriate; they just disagree on how to perform that recapture.

The government's proffered method of recapture completes the task in one year. Since the earliest tax year subject to reopening and recomputation is 1970, the government contends the proper method of accounting for all the earlier excessive reserves is to include the improper reserves in the 1970 opening reserves while excluding all improper reserves from the 1970 closing amounts. This would result in an IRS recapture in 1970 of all the improper deductions stemming from the establishment of improper Issues 1 and 2 reserves and the consequential improper increase of those reserves. While not objecting to a type of tax recovery, the plaintiff contends that the recovery should be spread over a period of 10 years.

The plaintiff's argument for a 10-year spread of the recovery is based on Internal Revenue Code section 810(d)(1) and Regulations 1.810-2(c)(2) and 1.810-3(a). Code section 810(d)(1) provides as follows:

(d) Adjustment for change in computing reserves.-

(1) In general. -- If the basis for determining any item referred to in subsection (c) as of the close of any taxable year differs from the basis for such determination as of the close of the preceding taxable year, then so much of the difference between --

(A) the amount of the item at the close of the taxable year, computed on the new basis, and

(B) the amount of the item at the close of the taxable year, computed on the old basis, as is attributable to contracts issued before the taxable year shall be taken into account for purposes of this subpart as follows:

(i) if the amount determined under subparagraph (A) exceeds the amount determined under subparagraph (B), 1/10 of such excess shall be taken into account, for each of the succeeding 10 taxable years, as a net increase to which section 809(d)(2) applies; or

(ii) if the amount determined under subparagraph (B) exceeds the amount determined under subparagraph (A), 1/10 of such excess shall be taken into account for each of the 10 succeeding taxable years, as a net decrease to which section 809(c)(2) applies.

Regulation section 1.810-3(a) then quotes the same provision virtually verbatim. Additionally, Regulation section 1.810-2(c)(2) essentially instructs that decreases in reserves which are attributable to a change in the basis of computing reserves should not be taken into account as an ordinary decrease in reserves, but rather should be taken into account as provided by Code section 810(d) and Reg. section 1.810-3(a).

Thus, the clear message from these statutory and regulatory provisions is that if the company's change from excessive to proper reserve calculations qualifies as a "change in the basis of computing reserves," the resulting difference in the reserve increase should be spread over a period of 10 years. Stated even more particularly, the question is whether the shift to a lessened use of the section 818(c)(2)(B) revaluation term insurance reserves should be classified as a change in the basis of computing reserves for purposes of section 810(d). Of course, it should also be noted that the "lessened use" of the revaluation formula was also compulsory in that it resulted from a denial of the formula's applicability to reserves set aside for term insurance in force less than 15 years regardless of any other insurance which may have been issued in the same contract.

The government's position is that the forced discontinued use of the section 818(c)(2)(B) revaluation for certain portions of reserves is not a change in the basis of computing reserves for purposes of section 810(d)(1). Obviously, this position is contrary to what the plain language of the statutory provision expresses. The original use of the revaluation formula for all term insurance reserves, as long as the policy included some "other than term" insurance or "longer than 15 year" term insurance, was a specific method or basis of computing reserves. Likewise, the use of the section 818(c)(2)(B) revaluation formula for only that portion of reserves attributable to term insurance in force for more than 15 years was another specific method or basis of computing reserves. Thus, the spread rule of section 810(d) would appear to be triggered. Nevertheless, the government appears to contend that any change from an improper basis to a proper basis of computing reserves is not a change in basis of computing reserves for purposes of section 810(d)(1). There is no case law or statutory language to support this limited construction of the phrase "change in the basis of computing reserves," but the position is apparently based on the theory that excessive amounts of "reserves" resulting from improper reserve computations or revaluations are not really reserves and therefore cannot be considered as part of a change in the basis of computing reserves. In other words, improper reserves, or disallowed portions of reserves, must be thrown out of the computation altogether and cannot be considered as part of the reserves which may experience a change in the basis of the way they are computed.

Yet another way of articulating the government's position would be to argue that a change from a "wrongful" basis of computing reserves was not intended to be within the scope of the 10-year spread rule. Only a change from a proper basis of computing reserves to a different proper basis qualifies. A perfect example would be a case of reserve strengthening through the use of a different assumed rate of interest. Excluded, however, would be a change in reserves resulting from an improper and subsequently disallowed revaluation of reserves as in the case at hand.

Finally, the critical distinction necessary to the government's favored result could be the voluntariness or involuntariness of the change in basis of computing reserves. Indeed, the government's brief does make a passing allusion to voluntariness being the dispositive factor. For all practical purposes, of course, this is saying very much the same thing as the above-stated rationale for the government's position. Ordinarily, a forced change in the basis of computing reserves will be the result of the first basis being improper or wrongful. Regardless, if wrongfulness or voluntariness is truly critical, the government's position in this case on this issue would have to be sustained.

The court, however, does not interpret section 810(d) as being limited to voluntary changes in the basis of computing reserves or to changes from proper computations to different proper computations. While not articulated by the government, the court can conceive of legitimate public policy reasons for limiting the spread rule's application to those instances, but the fact remains that there is no indication that such a limitation was intended by either Congress or the regulatory authority. The language speaks simply of changes resulting from a change in the basis of computing reserves. There is no hint that the original basis must be "proper" or that the change must be voluntary. Consequently, interpreting the provisions in accord with the common, ordinary meaning of the language leads to the conclusion that the change in this case was a change in the basis of computing reserves for purposes of section 810(d). At first, the company computed reserves by using the preliminary term method and then revaluing the resulting amount in accordance with section 818. As it turned out, some of the revaluation was improper because certain portions of the amounts calculated through the preliminary term method did not qualify for revaluation. Hence, the company turned to a method of computing reserves which did not revalue those certain portions of the reserve amount. By all normal usages of the words, this was a change in the method of computing reserves. Thus, the spread rule of section 810 should be applied.

Furthermore, even if the plain language of the statute had somehow allowed for a regulatory limitation upon its reach, the Internal Revenue Service's own ruling in another case forecloses this possibility. Revenue Ruling 77-198, 1977-1 C.B. 190 involved a factual situation very similar to the one at hand, and yet the Service expressly ruled that the 10-year spread rule applied. In fact, the change in the present case is even more naturally considered a change in the basis of computing reserves than is the situation considered in the revenue ruling. This is because the result of the changes in the revenue ruling case were more capable of being described as "now reserves" than are the result of the changes in the present case. In other words, the present case is more of a simple tinkering with the computation formula. Yet, as discussed below, the revenue ruling specifically rejected the idea that the change at stake resulted in "new reserves" rather than simply being a change in the method of computing reserves.

The revenue ruling case originated with an IRS declassification of certain reserves from their claimed status of "life insurance reserves." They were declassified on the ground that they "were not computed actuarially so as to qualify as life insurance reserves." In response to the declassification, the taxpayer switched to an acceptable actuarial method of computation. The question then addressed by the revenue ruling was whether this was a change in the basis of computing reserves for purposes of the 10-year spread rule in section 810(d). The unqualified answer was yes. The Internal Revenue Service declared that the reserves were not "new reserves" because they were "not attributable to new or additional benefits on policies in force." Rather, the new amount was simply the result of a change in the basis of computing reserves for the same policy benefits. Accordingly, the adjustment was subject to the spread rule.

The same principle calls for application of the spread rule in the instant case. The change in reserve amounts in this case comes not from any new or additional policy benefits, but from a change in the basis of computing reserves for the same benefits. Thus, Rev. Rul. 77-198 supports the plaintiff's position that section 810 should be used. The government's only attempt to distinguish this revenue ruling from the present case is to note that the revenue ruling involved a voluntary change in the basis of computing reserves. The apparent implication from this notation is that involuntary changes in computation of reserves are not to be subject to the spread rule. To this, the court first of all notes that it is not convinced that a great degree of difference in voluntariness exists between the two cases. Both began with a change which was in essence force by the fact that the IRS disallowed a portion of the amounts claimed as life insurance reserves. Second, even if the other case may be deemed a voluntary change and the present case involuntary, the court fails to see any statutory or regulatory authority for placing significance upon this distinction. As discussed earlier, the plain language of the statute and regulations speaks simply of changes in the basis of computing reserves. No mention is made of whether the change is voluntary or involuntary, or from a proper or improper method. Furthermore, the plain language of the provision and the revenue ruling leave no doubt that the elimination of an earlier revaluation for a certain portion of reserve amounts is truly a change in the basis of computing reserves. As such, it should therefore trigger application of the 10-year spread rule.

Finally as if this were not enough, the court notes that the government's argument in its brief actually lends credence to applying the spread rule. The government argues that its single-year recapture of the tax loss is authorized by two separate principles or statutes, each of which is enough by itself, but both of which the government argues are right. ² One of these alleged authorizations is based upon Tax Code section 481 which basically authorizes adjustments in income to account for changes in "method of accounting." Accordingly, the government argues that the discontinued revaluation of certain portions of reserves is a change in method of accounting. The government then goes on to explain a change in the method of accounting as including "changes in the treatment of material items," which also "includes any item involving the proper time for including an item in income or taking a deduction." Treas. Reg. section 1.446-1. Furthermore, as pointed out in North-Central Life Insurance v. Commissioner, 92 T.C. No. 15, (1989), disallowance of a reserve is a question of timing and not merely the disallowance of an improper deduction. Accordingly, the court readily accepts the idea that this case involved a change in method of accounting. But if a change in method of accounting was present, how can it be argued that there was not a change in the method of computing reserves? While the phrases are not exactly synonymous, the fact that a partial discontinuance of a reserve revaluation should be considered a change in method of accounting strongly implies that it should also be considered a change in the method of computing reserves.

Additionally, there is no problem with attributing both of these descriptions to what happened in this case. There need be no conflict between section 481 and the 10-year spread rule of section 810. Code section 481 is simply a much more general provision dealing with recapture of tax income in a broad variety of cases. It is a broad code which generally authorizes recapture. Code section 810, on the other hand, is much more specific and deals with a very narrow and limited type of "change in method of accounting." It in no way contradicts the general rule that there should be recapture of tax loss. It simply provides a more specific manner of recapturing tax loss under one set of particular circumstances in which there was an accounting change, namely circumstances in which there was a change in the method of computing reserves. As usual, the specific controls the general. It is not a contradiction of the general rule. Accordingly, while the government is correct in classifying the change at issue as a change in method of accounting, it is also more specifically a change in the method of computing reserves.

Likewise, the more specific provision of section 810 must have preeminence over the general tax benefit rule, the other authority the government offers for its position. Again, section 810 does nothing to contradict the basic proposition of the tax benefit rule that when an amount deducted from gross income in one taxable year is rediscovered in a later year, the recovery is income in the later year. Code section 810 actually goes hand in hand with this rule. It just provides that the recovery should be spread out over time in certain circumstances. Since the necessary circumstances of a change in basis of computing reserves is present in this case, the spread rule of section 810 must be applied.

ISSUE 6

The sixth issue also involves a disputed application of the 10- year spread rule of Internal Revenue Code section 810(d). More specifically, the question is whether the spread rule should be used to adjust the reserve amount used to calculate "required interest." Required interest is simply the investment yield, or at least the expected investment yield, of reserves. It is one of the amounts which must be computed in order to determine taxable income. Ordinarily, required interest can be computed rather easily in accordance with the simple formula laid out in Code section 809(a)(2). The rate of interest assumed in calculating reserves is multiplied by the mean of the beginning and ending reserves for the taxable year. The difficulty in this case arises from the fact that at some point prior to 1970, the plaintiff engaged in reserve strengthening.

The plaintiff's reserve strengthening was accomplished by assuming a lower rate of interest. The process was therefore a classic example of a "change in the basis of computing reserves." As discussed earlier, section 810(d) provides that if the basis for computing reserves changes in any year, then the amount of any difference attributable to that change is to be taken into account ratably over the succeeding 10 taxable years rather than being taken into account completely in the year of the change. Accordingly, for purchases of computing required interest, the plaintiff used reserve amounts adjusted in accordance with this spread rule. The government, on the other hand, rejected required interest calculations which used this lesser reserve amount. The government asserted that required interest computations must be based on the mean of the actual strengthened reserves at the beginning and end of the taxable years in question, namely 1970 and 1971. Thus, the issue is whether the opening and ending reserves used to compute the 1970 and 1971 means of opening and ending reserves should be the full strengthened reserves or the lesser, ratably calculated reserve amounts allowed under section 810(d) for tax purposes.

If section 810(d) standing alone left doubt on whether its spread rule should be used to adjust reserve amounts being plugged into the required interest formula, this court concludes that Regulation section 1.809-2(d) removed any remaining question. Regulation section 1.809-2(d), which carries the heading of "Required interest defined," first states the same basic formula for computing required interest as is stated in Code section 809(a)(2). It then further provides as follows:

For purposes of computing required interest under section 809(a)(2) and subparagraph (1) of this paragraph, the amount of life insurance reserves taken into account [i.e., the opening and closing reserves of which the mean will be computed and then multiplied by the assumed interest rate] shall be adjusted first as required by section 818(c) (relating to an election with respect to life insurance reserves computed on a preliminary term basis) and then as required by section 806(a) (relating to adjustments for certain changes in reserves and assets) before applying the rate of interest required, or assumed by the taxpayer, thereto. However, in the case of the adjustments required by section 810(d) as a result of a change in the basis of computing reserves, the adjustments to any of the reserves described in section 810(c) shall be taken into account in accordance with the rules prescribed in section 810(d) and section 1.810-3.

Since the reserves in question in this case are section 810(c) reserves, and since section 810(d) and regulation section 1.810-3 articulate the 10-year spread rule already discussed, the above- quoted regulation clearly appears to envision use of the spread rule during required interest computations. Otherwise, the last sentence of the regulation would be rendered meaningless. Without some just cause, regulatory provisions should not be ignored or interpreted into meaninglessness. Yet, this is exactly what the government's approach would accomplish because it does not make the slightest attempt to make any section 810(d) adjustment in reserves to account for that portion which is attributable to a change in the basis of computation. Rather, the government simply computes required interest by multiplying the mean of the actual reserves computed on the new basis by the assumed rate of interest. This failure to consider the instructions of Code section 810() [sic] and Reg. section 1.810-2(d) forces this court to reject the government's position on Issue 6.

Interestingly, the only reason offered by the government for rejecting application of the spread rule is a hypothetical example which supposedly shows how the plaintiff's method of computing required interest "breaks down." The government hypothetical is "a case of reserve strengthening attributable to a change in assumed interest rate from 3-1/2 percent to 3 percent, where the year-end reserve computed on the new basis is $100, and the reserve computed on the old basis is $90." It also assumes "that the reserve as of the end of the immediately following year, computed on the new basis, is $105." According to the government, the plaintiff's position would then be that the mean reserve in the year following the change in basis would be $93. This represents the mean of $90 and $96. The $90 comes from that year's opening reserve of $100 minus the full $10 of the section 810(d) amount. The $96 figure then comes from the closing reserve of $105 minus the $9 of the section 810(d) amount.

The problem with this, according to the government, is that neither the $96 closing reserve nor the $93 mean reserve are computed using the 3 or 3-1/2 percent or any other assumed interest rate. This fact, according to the government, prevents computation of required interest through the formula provided in section 809(a)(2). As discussed earlier, the section 809(a)(2) formula for computing required interest prescribes the multiplication of two components: (1) the rate of interest assumed by the plaintiff in calculating reserves, and (2) the mean of such opening and closing reserves. Here, the government argues that the first component of this formula is missing under the plaintiff's method, thereby rendering a computation of required interest impossible. The first component is said to be missing because allegedly no certain interest rate is assumed by the plaintiff in calculating the reserves.

The government's position is a hypertechnical view of section 809 which overlooks practical realities. True, neither the $96 or the $93 reserve amounts are computed straight from a simple and apparent interest rate. However, the court concludes that they are nevertheless computed through the assumption of an interest rate. Thus, there is no problem in applying the section 809 formula. Originally a 3 percent interest rate is assumed in order to reach a preliminary reserve amount, and then adjustments are made to that amount as a statutorily mandated method of easing the otherwise abrupt impact of a sudden change in assumed interest rate. The final result after the spread-rule adjustment is still a reserve amount which is intimately intertwined with an assumed interest rate, since the preadjustment amount stemmed directly from such a rate. The mere fact that some late tinkering is done with the originally calculated amounts does not alter the reality that the final amount is based upon an assumed interest rate.

In essence, Congress and the Internal Revenue Service simply established a process for computing reserves which has the same effect as making a change in the assumed rate of interest a gradual change rather than abrupt. In other words, the effect would be the same as if Congress had decreed that any change in assumed rates would, for tax purposes, be instituted gradually through certain increasing intermediate rates between the old and new rate over a period of 10 years. Under this latter approach, the government would certainly not even be able to argue that there was no assumed interest rate for calculating the reserve amounts and for section 809 purposes. This court fails to see why the same type of argument should carry the day just because Congress decided to reach this transitional result by tinkering with the reserve amount side of the equation rather than the more complicated interest rate side. Whichever approach is taken, the bottom line is that an assumed interest rate is used to calculate relevant reserve amounts. This is not changed by the fact that the original reserve amounts were subsequently adjusted to ease the transition into the use of a new interest rate. Accordingly, there is no difficulty with using the section 810 10-year spread rule when computing required interest under section 809. Furthermore, since the plain language of the spread rule and of Reg. section 1.809-2(d) indicates that the spread rule should be applied when determining required interest, it shall be in this case. The government's position on Issue 6 is hereby rejected.

An appropriate order will be entered.

                                   L. Clure Morton

In accordance with the memorandum contemporaneously filed, it is hereby ordered that the plaintiff is entitled to a refund of taxes, plus interest, as provided by statute, by reason of prevailing on Issues 3, 5, and 6. The plaintiff shall compute the amount of refund due plaintiff subject to the review and approval of defendent, and subject to the review of this court should the parties disagree. Judgment shall be entered for the plaintiff in the amount of such refund, plus statutory interest.

                                   L. Clure Morton

² This is not a case of the government arguing that one proposition is right, or that if that one is not, then another must be. Rather this is a case of the government arguing that two independent propositions are right regardless of the truth of the other. Thus, the government is definitely asserting that a change in method of accounting is involved, not just that it may be if certain other circumstances are present.