TAM 200115002: Standards for insurers' unpaid loss reserves

Plain-English summary

A group of property and casualty insurers disputed IRS adjustments to estimated unpaid losses, including asbestos and environmental claims. Chief Counsel advised that estimates must be based on loss events that have already occurred and must be fair and reasonable using information available when the estimates were made. Uncertainty about asbestos and environmental liabilities alone does not make a reserve nondeductible; estimates for actual unpaid losses can qualify if their amounts are fair and reasonable. The IRS should test the reserves through historical development rather than later claim results, and the taxpayer bears the burden of showing an adjustment is wrong. Greater reliance on actuarial techniques did not itself change the insurers' accounting method, assuming the earlier estimates also concerned actual unpaid losses.

Ruling snapshot

• Question: What standards govern unpaid loss reserves, asbestos and environmental claims, and a change in estimation technique?
• Outcome: Advice given, with deductibility dependent on actual losses and reasonable estimates
• Key authorities: IRC § 832; Treas. Reg. §§ 1.832-4(a)(5), 1.832-4(b); Rev. Proc. 75-56

Full text (IRS public release)

INTERNAL REVENUE SERVICE
NATIONAL OFFICE TECHNICAL ADVICE MEMORANDUM December 21, 2000

Number: 200115002
Release Date: 4/13/2001
Index (UIL) No.: 832.06-02
CASE MIS No.: TAM-112573-00/CC:FIP:B4

Taxpayer’s Name: .
Taxpayer’s Address:

Taxpayer’s Identification No:
Years Involved:

LEGEND:

Taxpayers = .
Year 1 =
Year 2 =
Year 3 =
Year 4 =
Year 5 =
Date 1 =
State A =
A
=
B
=
C
=
D
=
$a =
$b =
$c =
$d =
$e =
$f =
$g =
$h =
$i =
$j =
$k =
$l =
$m =
$n =
 2
TAM-112573-00

$o =
$p =
$q =
$r =
$s =
$t =

ISSUES:

I. What is the National Office position for the deduction of “losses incurred” with regard to the standards and procedures reflected in Treas. Reg. § 1.832-4 and Rev. Proc. 75- 56, 1975-2 C.B. 596?

1. Do the rules set out in Treas. Reg. § 1.832-4(a)(5)1 and 1.832-4(b) reflect different standards for determining the amount of a property and casualty (P&C) insurance company’s allowable loss reserve and if so, what are the reasons for those different standards?

2. What is the meaning of the phrase “must represent actual unpaid losses as nearly as its is possible to ascertain them,” as reflected in Treas. Reg.
§ 1.832-4(a)(5)?

3. What does the term “fair” mean in the “fair and reasonable” standard?

4. What is the meaning of the phrase “will be required to pay,” as reflected in Treas. Reg. § 1.832-4(b)?

5. What is the appropriate standard of proof in cases regarding adjustments to unpaid losses?

6. Does Rev. Proc. 75-56 prescribe a testing methodology that the Service must utilize to determine whether the reserves that the taxpayer established were fair and reasonable?

7. Does Rev. Proc. 75-56 require the Service to hold that loss reserves for a particular year are fair and reasonable if subsequent actual loss payments plus remaining reserves for that year exceed the original reserves?

II. Are contingent or anticipated asbestos and environmental (A&E) losses allowable for federal income tax purposes?

1
Effective January 5, 2000, Treas. Reg. § 1.832-4(a)(5) has been redesignated as Treas. Reg. § 1.832-4(a)(14).
 3
TAM-112573-00

III. Whether the change in Year 5 from a non-actuarial to an actuarial calculation of unpaid A&E losses constitutes a change in method of accounting?

CONCLUSIONS:
I.
1. Treasury Reg. § 1.832-4(a)(5) and -4(b) do not set forth disparate standards. The standard set forth in Treas. Reg. § 1.832-4(a)(5) generally applies to the method in which the taxpayer estimates unpaid losses, whereas the standard set forth in Treas. Reg. § 1.832-4(b) focuses on what is required before the Service may make an adjustment. Thus, where a reserve is fair and reasonable in amount, it is assumed that the taxpayer has based the reserve on “actual unpaid losses as nearly as it is possible to ascertain them.”

2. The phrase “must represent actual unpaid losses as nearly as it is possible to ascertain them,” as used in Treas. Reg. § .1832-4(a)(5), merely directs taxpayers to base their estimates of unpaid loss reserves on losses that have actually been incurred, and prohibits taxpayers from artificially inflating reserves by making estimates in excess of a more likely estimate. Since the Service may only adjust a taxpayer’s estimate of unpaid losses under Treas. Reg. § 1.832-4(b) if that estimate is not “fair and reasonable,” the relevant standard is “fair and reasonable.”

3. In the context of Treas. Reg. § 1.832-4(b), the terms “fair” and “reasonable” are closely related, if not synonymous. In addition, there is no authority that sets forth “fairness” as a requirement separate from “reasonableness” under Treas. Reg. § 1.832-4(b). Thus, the term “fair,” added to “reasonable,” does not mean “more than reasonable.”

4. The use of the term “will,” rather than “might” in Treas. Reg. § 1.832-4(b) refers to the requirement that taxpayers base their estimates of unpaid losses on actual losses, not losses that might occur in the future. The phrase “will be required to pay” does not impose a requirement that taxpayers’ estimates of unpaid losses represent the “most accurate” estimate.

5. In cases involving unpaid loss estimates, taxpayers bear the burden of proving by a preponderance of the evidence that the Service’s adjustment is incorrect. Taxpayers may accomplish this by establishing that their reserve estimate was fair and reasonable in accordance with Treas. Reg.
§ 1.832-4(b). The reference in Treas. Reg. § 1.832-4(b) to “the opinion of the district director” does not create an abuse of discretion standard.
Similarly, the Tax Court’s holding in Vinson & Elkins v. Commissioner, 99 T.C. 9 (1992), aff’d, 7 F.3d 1235 (5th Cir. 1993), does not create the presumption that a P&C actuary’s estimate of unpaid losses is correct 4
TAM-112573-00
unless proven by the Service to be “substantially unreasonable.” Lastly, taxpayers cannot meet their burden of establishing that their reserves are reasonable by simply showing that the amount of unpaid losses claimed on their return is identical to the amount reflected on their annual statement.2

6. Rev. Proc. 75-56 does not set forth a particular testing methodology.

7. Since the reasonableness of a taxpayer’s estimate of unpaid losses depends upon the information available to the taxpayer when the estimate is made, the use of a hindsight analysis is not appropriate under Treas.
Reg. § 1.832-4(b) and Rev. Proc. 75-56. Rather, the Service should test reserves on the basis of a historical development analysis.

II. Although Taxpayers’ unpaid losses attributable to their A&E liabilities are inherently uncertain, that fact alone does not establish that the reserves are non-deductible contingency reserves. If Taxpayers based their estimates of A&E losses on actual unpaid losses, rather than on anticipated losses arising from future loss events, Taxpayers’ estimates are properly includible in their reserves for unpaid losses if the estimates are fair and reasonable in amount.

III. Taxpayers’ enhanced reliance on actuarial techniques in Year 5 does not result in a change in its method of accounting, assuming that Taxpayers were basing their estimates on actual unpaid losses prior to Year 5. We note, however, that if a taxpayer changes the manner in which it estimates actual unpaid losses, the Service may make an adjustment if the change yields a result that is not fair and reasonable.

FACTS:

The years at issue are Years 2 through 5. Taxpayers are comprised of a holding company and its property and casualty insurance subsidiaries (“the subsidiaries”). The subsidiaries are domestic companies that are regulated by State A. Among the lines written by the subsidiaries are: (1) auto bodily injury; (2) commercial multiple peril; (3) workers compensation; (4) general liability; and (5) homeowners. Taxpayers’ internal actuarial department prepared yearly reports indicating estimates of the subsidiaries’ unpaid losses for the purpose of annual statement reporting. For the years in issue, the actuarial department’s reports and Taxpayers’ annual statements reflected year-end estimates of the subsidiaries’ combined unpaid losses and loss adjustment expenses (LAE) as follows:

2
The term “annual statement” refers to the financial statement form prescribed by the National Association of Insurance Commissioners (NAIC). 5 TAM-112573-00

Year Point estimate Range of Estimate on Difference (%) between low
per report (in estimates per Annual Statement and high of estimate range
millions) report (in millions) (in millions) per report and Annual
Statement estimate

Year 1

Year 2

Year 3

Year 4

In addition to the reviews performed by Taxpayers’ internal actuaries, Taxpayers’ unpaid loss estimates have been reviewed by several independent actuarial consulting firms. The recommended unpaid loss estimates set forth in these reviews, in comparison to the estimates set forth in Taxpayers’ annual statement, are as follows:

     Actuarial              Analysis performed                          Recommendation

firm

1. A Reserve review for end of Year 2 limited Estimated reserve $a less than
   to certain key lines Annual Statement estimate

2. B Reserve review for end of Year 2 Estimated reserve $b less than annual
   performed in conjunction with Taxpayers’ statement estimate
   financial audit

3. B Reserve review for end of Year 3 Estimated reserve $c less than annual
   performed in conjunction with Taxpayers’ statement estimate
   financial audit

4. B Reserve review for end of Year 4 Estimated reserve $d less than annual
   performed in conjunction with Taxpayers’ statement estimate
   financial audit

5. C Reserve review for end of Year 4, Estimated reserve $e greater than
   performed in conjunction with State A’s annual statement estimate
   insurance department’s triennial financial
   examination

6. B Reserve review for end of Year 5 Estimated reserve $f less than annual
   performed in conjunction with Taxpayers’ statement estimate
   financial audit

7. D Reserve review for end of Years 2 Estimated reserve $g less than annual
   through 5, prepared for the Service in statement estimate for Year 2, $h less
   conjunction with the examination of for Year 3, $i less for Year 4, and $j
   Taxpayers’ returns less for Year 5

Consistent with D’s conclusions described above, the Revenue Agent has proposed adjustments to Taxpayers’ unpaid loss reserves for Years 2 through 5 in the amounts of
$g, $h, $i, and $j, respectively.
 6
TAM-112573-00

In addition to the liabilities described above, during the years in issue Taxpayers were potentially liable for A&E claims attributable to policies written prior to Year 1. This is because parties insured by Taxpayers prior to Year 1 are subject to joint, retroactive liability pursuant to the Comprehensive Environmental Response, Conservation, and Liability Act of 1980 (CERCLA), presently codified at 42 U.S.C. §§ 9601 et seq. (2000). As relevant here, CERCLA imposes joint, several, strict, and retroactive liability on all potentially responsible parties (PRPs) for the costs of remediating hazardous waste sites, the most polluted of which are placed on the National Priorities List (NPL). PRPs may include generators, transporters, storage facilities, treatment facilities, owners of the site land, operators of the site, and lenders. According to the Environmental Protection Agency’s (EPA) internet website, there were 1,229 sites listed on the NPL as of December 8, 2000. See http://epa.gov//superfund/sites/query/queryhtm/npltotal.htm.

Since PRPs generally seek indemnification from their general liability insurers for their CERCLA remediation costs, Taxpayers in this case are potentially subject to retroactive liability for claims arising under CERCLA. Particularly, Taxpayers are potentially subject to claims covered by policies in effect during the 1960's and 1970's. Environmental claims may be made against multiple policies and insurers because pollutants are often released over many years, and the EPA periodically updates the NPL to add or remove sites. Accordingly, Taxpayers’ liabilities arising from CERCLA during the years in issue were difficult to ascertain.

Taxpayers established unpaid loss reserves for A&E liabilities separately from the unpaid loss reserves described above. Taxpayers’ actuarial reports explain that Taxpayers established reserves for A&E liabilities as part of “Special Situation Reserves.” Due to the uncertain nature of A&E liabilities throughout the insurance industry during Years 2 through 4, there was no generally-accepted actuarial method for calculating unpaid loss reserves for such liabilities. Consequently, Taxpayers did not employ a generally-accepted actuarial method in making estimates for Years 2 through 4. Taxpayers estimated unpaid loss reserves pertaining to their A&E liabilities, net of reinsurance, in the amounts of $k for Year 2, $l for Year 3, and $m for Year 4.

A performed an actuarial review of Taxpayers’ unpaid loss reserves for A&E liabilities as of Date1, Year 5. In its report, A recommended that Taxpayers increase its unpaid loss reserve for A&E liabilities by between $n and $o. In estimating Taxpayers’ liabilities, A employed three methods: (1) a model analysis, which calculated ultimate losses by applying a simulation model based on a sample group of insureds to Taxpayers’ actual claims data; (2) a loss development factor projection, which calculated ultimate losses by applying an assumed development pattern to Taxpayers’ paid or incurred losses; and (3) a market share analysis, which calculated Taxpayers’ ultimate losses by attributing to Taxpayers, on the basis of market share, a portion of the entire industry’s estimated ultimate liability. Relying upon A’s analysis, Taxpayers established unpaid loss reserves for A&E liabilities at the end of Year 5 in the amount of
$p, reflecting an increase from Year 4 in the amount of $q (i.e., $p less $m). 7 TAM-112573-00

During the course of the examination of Taxpayers’ returns, D accepted as reasonable Taxpayers’ A&E reserves for Years 2 to 4, and performed an analysis of Taxpayers’ A&E reserves for Year 5. D concluded that Taxpayers’ reserves for Year 5 should have been reduced by the amount of $r to $s; thus, of Taxpayers’ $q increase in its A&E reserves for Year 5, D accepted $t (i.e., $q less $r) as reasonable. D reasoned that a $t increase was reasonable because that amount represented the midpoint of A’s recommendation, less losses actually paid.

LAW AND ANALYSIS:

Property and casualty insurers must include in gross income the amount of their underwriting income as provided in I.R.C. § 832(b), computed on the basis of the underwriting and investment exhibit of the NAIC annual statement. § 832(b)(1)(A). Section 832(b)(3) defines the term “underwriting income” as the premiums earned on insurance contracts during the taxable year less losses incurred and expenses incurred. With respect to the term “losses incurred,” § 832(b)(5) provides that a property and casualty insurance company is entitled to reduce gross income for the taxable year to the extent that its estimated unpaid losses exceed its estimated unpaid losses for the previous taxable year. Conversely, § 832(b)(5) provides that a property and casualty insurance company must increase gross income for the taxable year to the extent that its estimated unpaid losses are less than its estimated unpaid losses with respect to the previous taxable year.

The deduction for unpaid losses is not subject to cash or accrual accounting rules. See United States v. General Dynamics Corp., 481 U.S. 239, 246-247 (1987). Rather, unpaid losses are an estimate, made at the close of the taxable year, of the insurer’s liability for claims that it will be required to pay in future years. Western Cas. & Sur. Co. v. Commissioner, 65 T.C. 897, 917 (1976), aff’d. on another issue, 571 F.2d 514 (10th Cir. 1978). In estimating its unpaid losses, an insurance company must comply with Treas. Reg. § 1.832-4(a)(5), which requires:

In computing "losses incurred" the determination of unpaid losses at the close of each year must represent actual unpaid losses as nearly as it is possible to ascertain them.

Regarding whether an insurer has improperly overstated its unpaid losses for tax purposes, Treas. Reg. § 1.832-4(b) provides:

Every insurance company to which this section applies must be prepared to establish to the satisfaction of the district director that the part of the deduction for "losses incurred" which represents unpaid losses at the close of the taxable year comprises only actual unpaid losses. ...These losses must be stated in amounts which, based upon the facts in each case and the company's experience with similar cases, represent a fair and reasonable estimate of the amount the company will be required to 8
TAM-112573-00
pay. Amounts included in, or added to, the estimates of unpaid losses which, in the opinion of the district director, are in excess of a fair and reasonable estimate will be disallowed as a deduction. The district director may require any insurance company to submit such detailed information with respect to its actual experience as is deemed necessary to establish the reasonableness of the deduction for "losses incurred."

Rev. Proc. 75-56 provides that “the standard of reasonableness” concerning the computation of unpaid losses “will be that set forth in sections 1.832-4(a)(5) and 1.832- 4(b) ....” Rev. Proc. 75-56, § 3.01, 1975-2 C.B. 596. Whether a taxpayer’s estimate of unpaid losses is “fair and reasonable” is essentially a valuation issue and a question of fact. Hanover Ins. Co. v. Commissioner, 69 T.C. 260, 270 (1977) aff’d, 598 F.2d 1211 (1st Cir. 1979).

I. What is the National Office position for the deduction of “losses incurred” with regard to the standards and procedures reflected in Treas. Reg. § 1.832-4 and Rev. Proc. 75- 56, 1975-56, 1975-2 C.B. 596?

1. Do the rules set out in Treas. Reg. § 1.832-4(a)(5) and 1.832-4(b) reflect different standards for determining the amount of a P&C company’s allowable loss reserve, and if so, what are the reasons for those different standards?

Treasury Reg. § 1.832-4(a)(5), which requires that taxpayers’ estimate unpaid losses at the close of each year “represent actual unpaid losses as nearly as it is possible to ascertain them,” seems to set forth a more stringent standard than Treas. Reg. § 1.832- 4(b), which requires that estimates of unpaid losses be stated in amounts which “represent a fair and reasonable estimate of the amount the company will be required to pay.” Treasury Reg. § 1.832-4(a)(5) appears to require that insurers’ estimates of actual unpaid losses to be as accurate as possible, whereas Treas. Reg. § 1.832-4(b) merely requires that such estimates be fair and reasonable.

Nevertheless, Treas. Reg. §§ 1.832-4(a)(5) and -4(b) do not represent different standards. The apparent disparity between Treas. Reg. § 1.832-4(a)(5) and Treas. Reg. § 1.832-4(b) can be reconciled by exploring the background of each provision. The relevant language in the predecessor to Treas. Reg. § 1.832-4(a)(5) was contained in a regulation promulgated on March 1, 1943. See T.D. 5236 (March 1, 1943). During this time, the NAIC required insurers to establish unpaid loss reserves equal to the greater of two separately-calculated reserves: (1) a case-based reserve representing the aggregate reserves for specific claims estimated by the insurer’s claims adjusters or; (2) a formula reserve representing a specified percentage of the insurer’s premium volume. See Charles W. Tye, The Convention Form and Insurance Company Tax Problems, 6 Tax Law Rev. 245, 245-246 (1951). Prior to promulgation of T.D. 5236, the Service had successfully litigated its position that insurers were only entitled to deduct unpaid losses that were calculated on the case method, thereby preventing 9 TAM-112573-00

insurers from using the formula method for tax purposes. See, e.g., Pacific Employers Ins. Co. v. Commissioner, 33 B.T.A. 501 (1935), aff’d 89 F.2d 186 (9th Cir. 1937); American Title Co. v. Commissioner, 29 B.T.A. 479 (1933) aff’d 76 F.2d 332 (3d Cir. 1935). Accordingly, the language presently contained in Treas. Reg. § 1.832-4(a)(5) was initially included in the regulations in an attempt to emphasize the Service’s longstanding position that insurers were not entitled to use the formula method for tax purposes.3

Although the Service in promulgating T.D. 5236 expressly restricted taxpayers to the case method, the requirement that “unpaid losses must represent actual unpaid losses as nearly as it is possible to ascertain them,” standing alone, appeared to set forth an exact standard with respect to an estimated item that was inherently uncertain. Accordingly, T.D. 5236 raised additional questions concerning the manner in which the Service would determine whether case-based reserves were overstated. See Charles W. Tye, Federal Taxation of Insurance Companies and Their Problems, 21 Taxes 594, 616 (November 1943) (opining that “[t]he [Service], if it is to try and treat the computation of ‘unpaid losses’ as an exact science on a case basis should give the companies more of a guide to their intention ... than to merely state that ‘unpaid losses must represent actual unpaid losses as nearly as it is possible to ascertain them’.”). The Service addressed these concerns on July 1, 1944 by promulgating the predecessor to Treas. Reg. § 1.832-4(b), which authorized the Service to make adjustments to reserves that it deemed impermissibly excessive, i.e., not “fair and reasonable.” See T.D. 5387 (July 1, 1944). In this regard, accompanying issuance of T.D. 5387 was Comm. Mim. R.A. No. 1366, which set forth a rule of thumb for auditing agents to use in determining whether estimates of unpaid losses were reasonable.4 Specifically, it directed agents to make adjustments to loss reserves for certain lines of insurance only if the average of the preceding five years’ estimated losses exceeded 115 percent of the average one year development of those estimates. See Rev. Proc. 75-56.

3
Despite promulgation of T.D. 5236 in 1943, taxpayers and the Service continued to dispute whether formula reserves were appropriate for tax purposes. The Tax Court eventually permitted the use of formula reserves, thereby reversing its decisions in Pacific Employers and American Title. See Columbia Cas. Co. v. Commissioner, 7 T.C.M. (CCH) 282 (1948). The dispute was eventually resolved in 1950 when the NAIC no longer required insurers to take formula reserves into account in computing income for annual statement purposes. See Tye, Convention Form, at 252-253.
4
Relevant portions of Comm. Mim. R.A. No. 1366 are set forth in Hanover Ins.
Co. v. Commissioner, 598 F.2d 1211, 1214 n.7 (1st Cir. 1979). Due to technological advancements within the insurance industry, the Service in Rev. Proc. 75-56 announced that it would no longer rely upon the guidelines set forth in Comm. Mim. R.A. No. 1366. 10 TAM-112573-00

Accordingly, the standard set forth in Treas. Reg. § 1.832-4(a)(5) generally applies to the method in which the taxpayer estimates unpaid losses, whereas the standard set forth in Treas. Reg. § 1.832-4(b) focuses on what is required before the Service may make an adjustment. Treasury Reg. § 1.832-4(a)(5) directs the taxpayer to base its estimate of unpaid loss reserves on losses that have actually been incurred, and prohibits taxpayers from artificially inflating reserves by making estimates in excess of a more likely estimate. Given that estimates of unpaid losses are inherently uncertain, however, Treas. Reg. § 1.832-4(b) contemplates that reasonable parties may disagree as to what constitutes the most accurate estimate of an insurer’s unpaid losses. Therefore, Treas. Reg. § 1.832-4(b) provides that the Service may only adjust a taxpayer’s loss reserve if it is in excess of a “fair and reasonable” amount, based upon the facts in each case and the taxpayer's experience with similar cases. Thus, where a reserve is fair and reasonable in amount, it is assumed that the taxpayer has based the reserve on “actual unpaid losses as nearly as it is possible to ascertain them.”

For the foregoing reasons, we conclude that Treas. Reg. § 1.832-4(a)(5) and -4(b) do not set forth disparate standards.

2. What is the meaning of the phrase “must represent actual unpaid losses as nearly as it is possible to ascertain them,” as reflected in Treas. Reg.
§ 1.832-4(a)(5)?

The parties in this case dispute whether the phrase “as nearly as it is possible” imposes a “most accurate” standard upon taxpayers. Actuarial estimates of unpaid losses are inherently uncertain; thus, different actuaries may arrive at different results in attempting to calculate a “most accurate” estimate. Accordingly, Treas. Reg. § 1.832- 4(b) authorizes the Service to make an adjustment to a taxpayer’s unpaid loss estimate only if the estimate exceeds a fair and reasonable estimate. As explained, infra, the unpaid loss estimate reflected on a taxpayer’s annual statement in some cases may not satisfy the fair and reasonable standard for tax purposes. A taxpayer’s estimate of unpaid losses will typically be considered fair and reasonable for tax purposes if the taxpayer estimates its unpaid losses on the basis of a recognized methodology that is appropriate for its particular line of business, calculates the estimate in accordance with actuarial standards, and properly takes into account its prior experience.

Furthermore, an estimate chosen from a range of reasonable estimates may, in certain cases, satisfy the requirements set forth in the regulations. See Utah Medical Ins. Ass’n v. Commissioner, T.C. Memo. 1998-458. Since different actuaries or different actuarial techniques may at different results in attempting to arrive at a “most accurate” estimate, the language “as nearly as it is possible to ascertain them” does not preclude the use of ranges. Whether a range is appropriate will depend upon the facts of each case; for example, a range may be warranted with regard to more volatile lines of insurance. Similarly, the breadth of such a range will also depend upon the facts of each case. 11 TAM-112573-00

Accordingly, the phrase “must represent actual unpaid losses as nearly as it is possible to ascertain them,” as used in Treas. Reg. § 1.832-4(a)(5), does not impose upon taxpayers a “most accurate” standard in terms of estimating unpaid losses. That language merely directs the taxpayer to base its estimate of unpaid losses on losses that have actually been incurred, and prohibits taxpayers from artificially inflating reserves by making estimates in excess of a more likely estimate. Since the Service may only adjust a taxpayer’s estimate of unpaid losses under Treas. Reg. § 1.832-4(b) if that estimate is not “fair and reasonable,” the proper standard is “fair and reasonable.”

3. What does the term “fair” mean in the “fair and reasonable” standard?

The parties dispute whether the term “fair,” added to the term “reasonable,” imposes upon taxpayers a “higher than reasonable” standard with respect to estimates of unpaid losses. In the context of Treas. Reg. § 1.832-4(b), the two terms are closely related, if not synonymous.5 If there is any distinction between the two terms, it is that “fair” refers to something impartial and free from bias, while “reasonable” refers to something governed in accordance with reason or sound thinking, or within the bounds of common sense. See American Heritage Dictionary at 486, 1031 (Houghton Mifflin Co., 2d College Ed. 1982).

With respect to unpaid loss estimates, an estimate that is “reasonable” could typically be characterized as “fair,” and vice versa. If, for example, a taxpayer’s unpaid loss estimate is based on sound reason, then the estimate is not biased against the Service. In this regard, the authorities concerning estimates of unpaid losses do not refer to the “fairness” of the reserve separately from the “reasonableness” of the reserve. See, e.g., Treas. Reg. § 1.832-4(b) (providing that the Service may require insurers to furnish documentation “as is deemed necessary to establish the reasonableness of the deduction for losses incurred”); Rev. Proc. 75-56 (referring to “the standard of reasonableness in computing unpaid losses”).

Accordingly, the term “fair,” added to “reasonable,” does not mean “more than reasonable.”

4. What is the meaning of the phrase “will be required to pay,” as reflected in Treas. Reg. § 1.832-4(b)?

The parties dispute whether the provision in Treas. Reg. § 1.832-4(b) that estimates in unpaid losses “represent a fair and reasonable estimate of the amount the company will be required to pay,” imposes a requirement that taxpayers’ estimates of unpaid losses represent the “most accurate” estimate of unpaid losses. In particular, the parties dispute whether the regulation’s use of the phrase “will be required to pay,” as opposed

5
The American Heritage Dictionary at 1031 (Houghton Mifflin Co., 2d College Ed. 1982) defines “reasonable” with reference to “fair” as follows: “4. Not excessive or extreme; fair: reasonable prices.” 12 TAM-112573-00

to the phrase “might be required to pay,” prevents taxpayers from using anything other than the “most accurate” estimate.

As discussed, supra, the proper standard for the Service to use in adjusting a taxpayer’s unpaid loss reserve is “fair and reasonable.” Since estimates of unpaid losses are inherently uncertain, taxpayers cannot be sure what amount they will be required to pay at the time that they make the estimate. Thus, the term “will” in the regulation does not require that the taxpayer be absolutely certain that its estimate will prove to be correct. The standard “fair and reasonable” accounts for this uncertainty. The regulation’s use of the term “will,” rather than “might,” refers to the requirement that taxpayers base their estimates of unpaid losses on actual losses, not losses that might occur in the future.

Accordingly, the phrase “will be required to pay” does not impose a requirement that taxpayers’ estimates of unpaid losses represent the “most accurate” estimate.

5. What is the appropriate standard of proof in cases regarding adjustments to unpaid losses?

The parties dispute whether the language in Treas. Reg. § 1.832-4(b) that permits the Service to adjust unpaid loss estimates “which, in the opinion of the district director,” exceed a fair and reasonable estimate, means that the taxpayer can successfully challenge the Service’s adjustment only by showing that the adjustment constituted an abuse of discretion. Alternatively, the parties dispute whether Taxpayers in this case can successfully challenge the Service’s adjustments by proving that Taxpayers’ estimates were fair and reasonable, rather than by proving that the Service’s adjustments were not fair and reasonable.

In tax disputes, including those involving unpaid loss estimates, the burden of proof is typically upon the taxpayer to establish that the Service’s determination was incorrect.6 T.C. Rule 142(a); Hanover, 69 T.C. at 270; Minnesota Lawyers Mut. Ins. Co. v. Commissioner, T.C. Memo. 2000-203. Whether a taxpayer has met this burden is determined by a preponderance of the evidence. Estate of Mellinger v. Commissioner, 112 T.C. 26, 39 (1999). In Hanover, however, the Tax Court in dicta suggested that taxpayers may have an even greater burden in cases involving estimates of unpaid losses, explaining:

Indeed, it might well be argued, in light of the language of [Treas. Reg.
§ 1.832-4(b)], that petitioner bears the burden of showing not only that its unpaid loss reserves were reasonable but also that respondent’s

6
We note that § 7491, enacted by § 3001(a) of the Internal Revenue Service Restructuring & Reform Act of 1998, Pub. L. 105-206, 112 Stat. 685, 726-727, provides that the Secretary bears the burden of proof regarding factual issues in certain cases. The provisions of § 7491 are not at issue in this case. 13 TAM-112573-00
adjustments constituted an abuse of discretion.... However, it is not here held that petitioner bears this double burden.

Hanover, 69 T.C. at 270, n.6. In so doing, the court compared adjustments to unpaid losses with adjustments to bad debt reserves, where an abuse of discretion standard was applicable.7 See Westchester Dev. Co. v. Commissioner, 63 T.C. 198, 211 (1974).

Despite the dicta in Hanover, the deduction for unpaid losses can be distinguished from the deduction for bad debt reserves. As discussed, supra, § 832(b)(1)(A) and (b)(3) requires insurers to deduct unpaid losses in accordance with the annual statement method of accounting, which in turn requires insurers to estimate unpaid losses. In contrast, the deduction for bad debts under § 166 of the 1954 Code provided taxpayers with the choice of deducting reasonable additions to bad debts under a reserve method of accounting, subject to “the discretion of the Secretary.” The deduction for unpaid loss reserves, therefore, is a component of a broader statutory scheme that requires insurers to employ the annual statement method of accounting, whereas the deduction for bad debt reserves under § 166 of the 1954 Code was neither mandatory nor part of a larger statutory scheme.

Therefore, we conclude that the Service’s adjustments to a taxpayer’s unpaid loss estimates are not subject to an abuse of discretion standard. Treas. Reg. § 1.832-4(b) authorizes the Service to make an adjustment to a taxpayer’s estimate of unpaid losses only if the estimate is not fair and reasonable. Although the Service’s determination is presumed correct, the taxpayer can rebut this presumption by establishing by a preponderance of the evidence that its reserve estimate was fair and reasonable. Thus, by establishing that its reserve estimate was fair and reasonable, the taxpayer effectively proves that the Service’s adjustment was improper. This is consistent with the standard employed by the Tax Court subsequent to the issuance of its opinion in Hanover. See, e.g., Minnesota Lawyers, supra; Utah Medical, supra.

Nevertheless, the Service is not required to accord deference to the opinion of the taxpayer’s actuary. In this regard, the Tax Court in Vinson & Elkins v. Commissioner, 99 T.C. 9, 16-17 (1992), aff’d, 7 F.3d 1235 (5th Cir. 1993), in addressing the actuarial assumptions of a defined benefit plan under I.R.C. § 412(c)(3), explained that the Service was only permitted to retroactively challenge an actuary’s assumptions if the assumptions were “substantially unreasonable.” Such actuarial deference is not applicable with respect to estimates of unpaid loss reserves for P&C insurers. First, the court in Vinson & Elkins addressed specific statutory provisions that are not applicable to P&C insurers, and relied upon the unique legislative history underlying those provisions. Second, the United States Court of Appeals for the Fifth Circuit, in affirming the Tax Court’s opinion in Vinson & Elkins, questioned whether the “substantially

7
Repealed in 1986, § 166(c) of the 1954 Code provided that “there shall be allowed (in the discretion of the Secretary) a deduction for a reasonable addition to a reserve for bad debts.” 14 TAM-112573-00

unreasonable” standard employed by the Tax Court was appropriate. 7 F.3d at 1240. Lastly, Treas. Reg. § 1.832-4 authorizes the Service to adjust a taxpayer’s reserves if they are not fair and reasonable in amount; there is no requirement that the reserve be established as “substantially unreasonable” prior to any adjustments.8

We also note that taxpayers in general cannot meet their burden of establishing that the Service’s adjustment is incorrect by simply showing that the amount of unpaid losses claimed on their return is the same as the amount reflected on their annual statement. Treasury Reg. § 1.832-4(b) was fashioned in an attempt to limit the deduction for unpaid losses to estimates that were fair and reasonable in amount. Both the Tax Court and the United States Court of Appeals for the First Circuit, in upholding the validity of Treas. Reg. § 1.832-4(b), have held that a taxpayer’s adherence to the annual statement does not preclude the Service from contesting the reasonableness of the estimates contained therein. See Hanover Ins. Co. v. Commissioner, 598 F.2d 1211, 1217 (1st Cir. 1979), aff’g 65 T.C. 715 (1976).9 The legislative history underlying the Tax Reform Act of 1986, Pub. L. No. 99-514, cites with approval the First Circuit’s opinion in Hanover. Specifically, the House Conference Report contains the following explanation:

The amount of the deduction for losses incurred must be reasonable.
See Reg. §. 1.832-4(b) and Hanover Insurance Co. v. Commissioner, 598 F.2d 1121 [sic] (1st Cir. 1979), cert. denied, 444 U.S. 915. Thus, under present law, the Internal Revenue Service may review, and, if appropriate, adjust the amount of the deduction for unpaid losses and unpaid loss adjustment expenses.

H.R. Conf. Rep. No. 841, 99th Cong., 2d Sess. II-357 (1986), 1986-3 C.B. (Vol. 4) 338, 357.

The principle that the Service may challenge for tax purposes the amount of unpaid losses reflected on an insurer’s annual statement appeared to have been questioned in Sears, Roebuck & Co. v. Commissioner, 972 F.2d 858 (7th Cir. 1992), where the Seventh Circuit explained:

Generalities about what “[f]ederal tax statutes are concerned with” do not control concrete cases. Section 832 is no ordinary rule. It expressly links federal taxes to the NAIC’s annual statement: ... Section 832(b)(1)(A) requires an insurer to use “the underwriting and investment exhibit of the annual statement as approved by the [NAIC]” to determine its “gross

8
Although the court in Utah Medical cited Vinson & Elkins, the court did not do so for the purpose of adopting a standard of deference for the taxpayer’s actuary.
9
The Tax Court more recently reaffirmed this principle in Minnesota Lawyers Mut. Ins. Co. v. Commissioner, T.C. Memo. 2000-203.
 15
TAM-112573-00
income.” ... Both the “premiums earned” and “losses incurred” go into determining “gross income” -- which is to be “computed on the basis of the underwriting and investment exhibit of the annual statement approved by the [NAIC].” State insurance commissioners’ preferences about reserves thus are not some intrusion on federal tax policy; using their annual statement is federal tax law.

Sears, 972 F.2d at 865-866.

The opinion in Sears, however, did not question the Service’s well-established authority to adjust for tax purposes specific amounts reflected on an insurer’s annual statement. Rather, the issue in Sears concerned the timing of losses included in an insurer’s reserve for unpaid losses attributable to mortgage insurance. The court in Sears, rejecting the Service’s position in that case, concluded that the annual statement method of accounting required mortgage insurers to establish reserves when the borrower had defaulted, and that the Service was bound to accept that methodology. In this case, however, the issue is not whether the annual statement’s method of accounting is applicable for tax purposes; the issue is whether Taxpayers properly estimated a particular item set forth on the annual statement. Thus, the holding in Sears is distinguishable from the holding in Hanover.

Accordingly, in cases involving unpaid loss estimates, taxpayers bear the burden of proving by a preponderance of the evidence that the Service’s adjustment was incorrect. The taxpayer may meet its burden by establishing that its reserve estimate was fair and reasonable in accordance with Treas. Reg. § 1.832-4(b). The reference in Treas. Reg. § 1.832-4(b) to “the opinion of the district director” does not create an abuse of discretion standard. Similarly, the holding in Vinson & Elkins cannot be extended to the context of P&C insurance in an attempt to accord deference to the insurer’s actuary. Lastly, taxpayers cannot meet their burden of establishing that their estimates were fair and reasonable by simply showing that the amount of unpaid losses claimed on their return is identical to the amount reflected on their annual statement.

6. Does Rev. Proc. 75-56 prescribe a testing methodology that the Service must utilize to determine whether the reserves that the taxpayer established were fair and reasonable?

Rev. Proc. 75-56 does not set forth a particular testing methodology. Rather, § 3.01 of Rev. Proc. 75-56 provides that “the standard of reasonableness in computing unpaid losses will be that set forth in sections 1.832-4(a)(5) and 1.832-4(b) of the regulations
....”

7. Does Rev. Proc. 75-56 require the Service to hold that loss reserves for a particular year are fair and reasonable if subsequent actual loss payments plus remaining reserves for that year exceed the original reserves?
 16
TAM-112573-00

The parties’ submissions refer to the use of “subsequent development,” “hindsight,” and “historical development” as a means of testing whether an unpaid loss estimate for a particular year is fair and reasonable. We understand the term “subsequent development” as referring to the amount and timing of actual loss payments made by a company with respect to losses contemplated by a previous year’s unpaid loss estimate. We understand the term “hindsight” as referring to an analysis of the subsequent development of a particular year’s unpaid loss estimate for the purpose of determining whether the estimate proved sufficiently accurate to be considered fair and reasonable. We understand the term “historical development” as referring to an analysis of the subsequent development of prior years’ unpaid loss estimates for the purpose of determining whether the unpaid loss estimate for the present year is fair and reasonable. For example, if a company’s unpaid loss estimate for Year 10 were at issue, the subsequent development of that estimate would be the amount of losses paid by the company in Year 11 and each year thereafter with respect to losses contemplated by the estimate for Year 10. A hindsight analysis would attempt to determine whether the estimate for Year 10 was fair and reasonable by comparing the original Year 10 estimate with the amount subsequently paid with respect to losses contemplated by the original estimate, i.e., amounts paid in years after Year 10. In contrast, a historical development analysis would attempt to determine whether the estimate for Year 10 was fair and reasonable by comparing the unpaid loss estimates made in Years 1 through 9 with the amounts subsequently paid in Years 1 through 10 with respect to losses contemplated by each of those estimates.10

The parties in this case dispute whether the Service may use a hindsight analysis in determining whether a taxpayer’s reserves were fair and reasonable. In general, Treas. Reg. § 1.832-4(b) requires that a taxpayer’s estimate of unpaid losses be based upon the facts in each case and the company's experience with similar cases, and authorizes the district director to require the taxpayer to submit detailed information regarding its actual experience as is deemed necessary to establish the reasonableness of the deduction for "losses incurred.” Thus, whether a taxpayer’s estimate is fair and reasonable depends upon the information the taxpayer had at the time it made the estimate. In determining whether a taxpayer’s estimate of calculating unpaid losses is fair and reasonable, therefore, the Service should consider the historical development of the taxpayer’s unpaid loss estimates for the years prior to the year in issue. Use of a historical development analysis is also consistent with the Service’s goal of addressing the potential abuse associated with excessive unpaid loss estimates; since a taxpayer that has overestimated its unpaid loss deduction for a particular taxable year will typically be required to include any excess portion of the estimate in income in

10
Thus, the 115% “rule of thumb” testing methodology set forth in Comm. Mim.
R.A. No. 1366 and described in Rev. Proc. 75-56 was a historical development analysis, rather than a hindsight analysis. 17 TAM-112573-00

subsequent years, only a consistent pattern of overstating estimates of unpaid losses leads to substantial unwarranted tax deferral.11 See Hanover, 598 F.2d at 1218.

Accordingly, the use of a hindsight analysis is not appropriate under Treas. Reg.
§ 1.832-4(b) and Rev. Proc. 75-56; rather, the Service should test reserves on the basis of a historical development analysis.12

II. Are contingent or anticipated asbestos and environmental losses (A&E) allowable for federal income tax purposes?

Since estimates of unpaid losses can only be comprised of “actual” losses, a taxpayer cannot establish reserves unless the underlying loss event has occurred. See Maryland Deposit Ins. Fund Corp. v. Commissioner, 88 T.C. 1050, 1060 (1987) (estimated reserve for contingent future loss events is not allowable). The parties in this case dispute whether Taxpayers’ unpaid loss reserves for A&E liabilities are essentially “contingency reserves,” i.e., reserves for loss events that have not occurred. The Service in its submission asserts that a portion of Taxpayers’ unpaid loss reserves represents a contingency reserve because Taxpayers cannot establish any connection between the occurrence of insurable events and any liability attributable to that event. Particularly, the Field argues that Taxpayers’ unpaid loss reserves for A&E liabilities are based, in part, upon unknown hazardous waste sites and undiscovered policies.

Although Taxpayers’ A&E liabilities are inherently uncertain, that fact alone does not establish that the reserves are non-deductible contingency reserves. Any unpaid loss estimate, even if attributable to actual loss events, is subject to uncertainties; however, a non-deductible contingency reserve is one that is attributable to events that have not occurred. Taxpayers argue that their A&E unpaid loss reserves are not contingency reserves because the underlying liabilities existed at the time they estimated the reserves. Rather, Taxpayers explain that their estimated A&E liabilities are for losses incurred but not reported (IBNR), which are properly includible in their estimate of unpaid losses. See General Dynamics, 481 U.S. at 246-247. Taxpayers argue that since these estimates are attributable to claims that have not been reported, they are inherently uncertain, but are not subject to the type of contingencies contemplated by the term “contingency reserve.”

11
Therefore, where a taxpayer’s unpaid losses in years prior to the year in issue have consistently proven to be greater than necessary, the taxpayer should be aware that its methodology of calculating unpaid loss estimates may be flawed, or that the data underlying such estimates may be unreliable.
12
We recognize that the court in Hospital Corp. of America v. Commissioner, T.C. Memo. 1997-482, used a hindsight analysis in determining that a taxpayer failed to prove that its estimate of unpaid losses was fair and reasonable. The court, however, also based its conclusion on the taxpayer’s unsupported departure from its actuary’s recommended point estimate. 18 TAM-112573-00

Taxpayers in this case are potentially subject to retroactive liability for claims arising under CERCLA, and sites continue to be added to and removed from the NPL. Therefore, Taxpayers’ liabilities arising from CERCLA during the years in issue were difficult to ascertain; thus, any attempt by Taxpayers’ to estimate such losses may seem speculative in nature. Nevertheless, if Taxpayers based their estimates of A&E losses on actual unpaid losses, rather than on anticipated losses arising from future loss events, Taxpayers’ estimates are properly includible in its reserve for unpaid losses if the estimates are fair and reasonable in amount.13

Thus, the fact that Taxpayers’ A&E unpaid loss liabilities are subject to contingencies does not render estimates of such liabilities nondeductible contingency reserves.

III. Whether the change in Year 5 from a non-actuarial to an actuarial calculation of unpaid A&E losses constitutes a change in method of accounting?

The parties dispute whether Taxpayers changed their method of accounting for A&E reserves in Year 5. Specifically, for several years leading up to Year 4, Taxpayers primarily used managerial judgment to estimate unpaid losses for environmental claims that were incurred but not reported (IBNR). In Year 5, however, A performed an in depth survey to actuarially estimate the taxpayer’s IBNR reserves. The parties dispute whether Taxpayers’ change in Year 5 from a non-actuarial to an actuarial method of estimating its unpaid losses constitutes a change in its method of accounting.

The change described is not a change in Taxpayers’ method of accounting. Assuming that Taxpayers estimated reserves on the basis of actual unpaid losses in Year 5 and in prior years, Taxpayers’ relative reliance on actuarial rather than managerial judgment from one year to the next is simply a change in the manner of calculating the estimate. A similar issue was addressed in PLR 8406001. In that ruling, the taxpayer had been using a flawed “paid-to-paid” method to compute its reserves for unpaid loss adjustment expenses, and changed to a different paid-to-paid method. Citing Treas. Reg. § 1.446- 1(e)(2)(ii)(b) (regarding adjustments in reserve for bad debts), the Service concluded that the taxpayer’s change in the manner in which it computed its reserves is not a change in basis of accounting. The Service reasoned that a reserve method of accounting is an estimation method subject to a standard of reasonableness.

13
The request for technical advice at page 30 seeks our views as to whether certain events regarding A&E losses must occur before Taxpayers are permitted to establish a reserve. In general, Taxpayers are entitled to base their IBNR loss estimate for A&E liabilities upon the occurrence of event that gives rise to the liability. See Sears, 972 F.2d at 867. It is beyond our jurisdiction to determine whether a certain event gives rise to a liability under CERCLA. 19 TAM-112573-00

Accordingly, Taxpayers’ enhanced reliance on actuarial techniques in Year 5 does not result in a change in its method of accounting, assuming that Taxpayers’ were basing their estimates on actual unpaid losses prior to Year 5. We note, however, that if Taxpayers’ change in the manner in which it estimated actual unpaid losses for Year 5 yielded a result that is not fair and reasonable, the Service may make an adjustment on that basis.

CAVEAT(S)
A copy of this technical advice memorandum is to be given to the taxpayer(s). Section 6110(k)(3) of the Code provides that it may not be used or cited as precedent.