Full opinion text
TABLE OF CONTENTS Page I. INTRODUCTION.1160 II. PARTIAL TERMINATION CLAIMS.1161 A. General.1161 B. Vertical Partial Termination.1163 1. Significant Number or Percent.1163 2. What Number or Percent is Significant?.1164 3. How is the Number or Percent Calculated?.1164 a. Vested or Non-Vested Terminations.1164 b. Employees Who Transferred to a Successor Plan.1165 c. Turnover Rate .1166 d. Time Period.1167 e. The Relevant Number and Percent.1167 4. Significant Corporate Event.1169 5. The Number and Percent are Significant.1170 6. The IRS Determination.1170 C. Horizontal Partial Termination.1172 III. REMEDY FOR PARTIAL TERMINATION.1178 A. The Gulf Pension Plan.1179 B. Remedy Under the Gulf Pension Plan Upon a Partial Termination.1180 1. Standard of Review.1181 2. The Legally Correct Interpretation.1182 a. Uniformity of Construction .1182 b. Fair Reading of § 10.A.2.1182 (1) Does § 10.A.2 Apply When the Plan is Overfunded at the Time of a Partial Termination?.1182 (2) Does § 18.d of the Chevron Retirement Plan Bar Plaintiffs’ Claims to Surplus Gulf Plan Assets?.1183 (a) Validity of § 18.d Under ERISA’s Exclusive Benefit Rule_1184 (b) Validity of the 1986 Plan Merger Under § 208 of ERISA.... 1185 (c) Validity of § 18.d Under the A & B Plan, the CRP and the SAP.1185 (i) The A & B Plan.1186 (ii) The CRP.1190 (iii) The SAP.1194 (3) Can Surplus CRP and SAP Assets be Distributed Under § 10.A.2 Upon a Partial Termination?.1196 (4) Consequences of a Fair Reading of § 10.A.2.1198 c. Unanticipated Costs .1198 3. Abuse of Discretion.1198 a. Internal Consistency. 1198 b. Relevant Regulations.1199 e. Factual Background and Inference of Lack of Good Faith.1199 d. Conclusion.1201 IV. TERMINATION OF THE CRP AND SAP AS WASTING TRUSTS.1201 V. FIDUCIARY CLAIMS .1205 A. Pension Plan Expenses.1205 B. Self-Dealing by Chevron.1208 C. SRAP.1211 D. AVIS .1212 E. Defendants’ Promises to Set Aside Gulf Plan Assets for Plaintiffs’ Benefit.1213 F. Other Alleged Fiduciary Breaches.1214 VI. CONCLUSION.1214 OPINION LAKE, District Judge. I. INTRODUCTION This is a consolidated class action brought by more than 40,000 former participants in the Pension Plan of Gulf Oil Corporation. Defendants are Chevron Corporation, Gulf Oil Corporation, the Chevron Corporation Retirement Plan, the Pension Plan of Gulf Oil Corporation, the Benefits Committee of the Pension Plan of Gulf Oil Corporation and each of its members, and the Pension Committee of the Pension Plan of Gulf Oil Corporation and each of its members. To understand the case some appreciation of the demise of Gulf Oil Corporation is necessary. On January 1, 1982, Gulf had 29,706 employees covered under the Gulf Pension Plan and was one of the largest integrated oil companies in the United States. Concerned that Gulfs share price did not reflect its true value, Gulfs management began a plan to streamline the company that included substantial reductions in the number of employees. By the end of 1983 Gulf had reduced its work force to 23,054 active Gulf Plan participants, but its share price had still not risen substantially. In January of 1984 Gulf management learned that a hostile tender offer was imminent from a group led by T. Boone Pickens. Gulf sought to interest several other large oil companies, including Chevron Corporation, in a friendly merger to avoid the Pickens’ takeover attempt. In February of 1984 Gulf and Chevron announced a merger, and a merger agreement was signed in March of 1984. For the next year the companies operated independently under a standstill agreement while Gulf divested itself of certain assets required by the FTC and a number of combined Gulf-Chevron working groups determined how to integrate the two companies and their pension and other employee benefit plans. Because Gulf and Chevron were in the same business it became apparent that a number of employees would be redundant in the merged company. Chevron also decided to sell parts of Gulf to help pay the debt it had incurred in making the acquisition. By July 1, 1986, when the Gulf Pension Plan and the Chevron Annuity Plan were merged into the Chevron Retirement Plan, 13,545 former Gulf Pension Plan participants remained on Chevron’s payroll. All of the Gulf employees who left between January 1, 1982, and June 30, 1986, were covered by Gulf employee benefit programs, including pension plans. This action was filed by plaintiffs, Dean Borst, et al., in November 1986. In April of 1987, plaintiffs, Harry Back, et al., filed a similar action in the United States District Court for the Western District of Pennsylvania. On the motion of the defendants, the Back action was transferred to this Court and consolidated with the Borst action. The Court preliminarily certified the case as a class action on November 9, 1987, and appointed the individual plaintiffs as class representatives. After several hearings, on February 26, 1990, the Court certified the consolidated action under Fed. R.Civ.P. 23(b)(2) and defined the main class as (1) All participants in the Pension Plan of Gulf Oil Corporation or any predecessor plan (sometimes abbreviated as the “Gulf Plan” or “Plan”) who terminated employment for any reason after December 31, 1981, and before July 1, 1986, with Gulf Oil Corporation or its successors or affiliates, or any subsidiaries that had adopted the Gulf Plan (“Gulf”). (2) All Gulf Plan participants who were accruing benefits under the Gulf Plan as of June 30, 1986. (3) All Gulf Plan participants who terminated Gulf employment prior to January 1, 1982, and who were receiving a pension or entitled to an immediate or deferred pension or a refund of accumulated employee contributions under the Gulf Plan as of June 30, 1986. (4) All Gulf Plan participants who terminated employment with Gulf prior to January 1, 1982, and after December 31,1975, who were not entitled to any pension benefit under the Gulf Plan at the time of such termination (other than a refund of accumulated employee contributions) but who, if they had been reemployed by Gulf as of June 30, 1986, would have been entitled under the ERISA break-in-service rules to credit for prior service under the Gulf Plan. (5) All spouses, joint annuitants, or other plan beneficiaries of any deceased Gulf Plan participants described in the foregoing categories. (6) All alternate payees under qualified domestic relations orders of separated or divorced Gulf Plan participants described in the foregoing categories. The Court also certified a divestiture sub-class defined as All members of the main class who were offered employment by any of the following purchasers of assets applicable to the operation in which they were employed by Gulf: Sohio/BP (“Sohio”), Cumberland Farms, Champion Energy, or Ther-mex. On January 4, 1990, the Court granted defendants’ motion to strike plaintiffs’ request for a jury trial, and granted defendants’ motion to dismiss most of plaintiffs’ common law claims as preempted by the Employee Retirement Income Security Act of 1974, 88 Stat. 829, as amended, 29 U.S.C. §§ 1001, et seq. (“ERISA”). The parties agree that the Court has jurisdiction over the remaining claims pursuant to § 502(e) and (f) of ERISA, 29 U.S.C. § 1132(e) and (f). Plaintiffs went to trial seeking relief under ERISA and the language of various benefit plans for four types of wrongs allegedly committed by defendants. First, plaintiffs claimed that defendants’ actions resulted in a partial termination of the Gulf Pension Plan and its predecessor plans that entitled plaintiffs to relief provided by ERISA and the Plan. Alternatively, plaintiffs claimed that two of the plans had served their purposes and their trusts should be terminated as “wasting trusts” under common law. Second, former Gulf employees who were transferred to Sohio, Cumberland Farms, and Champion Energy as a part of divestitures of former Gulf operations in 1985 and 1986 sought early retirement benefits under the Gulf Pension Plan.- Third, former Gulf employees who were transferred to Champion Energy, Cumberland Farms, and Thermex incident to divestitures sought benefits under a severance plan that existed -from February 1, 1984, to February 1, 1986 (Plan 728). Finally, plaintiffs sought damages for a number of alleged breaches of fiduciary duty during the corporate merger and the integration of the two pension plans. On November 8, 1990, after hearing 16 days of testimony, the Court ordered the attorneys in charge for the class and defendants and the CEO of Chevron to meet and determine if some or all of the claims could be settled. The following week the parties announced a proposed settlement of the plaintiffs’ claims for early retirement benefits and severance benefits under Plan 728. As part of the settlement Chevron also agreed to vest in their accrued benefits under the Gulf Pension Plan all members of the class whose employment with Gulf was terminated between January 1, 1984, and June 30,1986, without a vested pension benefit, and plaintiffs agreed to dismiss the partial termination claims of former Gulf employees who were terminated between January 1, 1982, and December 31, 1983. After notice to the class, the Court held a hearing on the partial settlement on January 25, 1991, and approved it. The Court will now address the plaintiffs’ remaining claims. II. PARTIAL TERMINATION CLAIMS A. General An understanding of the partial termination claims requires an appreciation of some basic features of pension plans. An employee normally does not have an unconditional right to benefits provided by a defined pension benefit plan until the employee has worked for the employer for some number of years, as determined by the plan. If the employee ends his employment before he completes this period, he forfeits his right to pension plan benefits. If he meets the plan’s time-in-employment criteria, he becomes vested with the right to receive some minimum pension benefit, as defined by the plan, even if he should later end his employment. An employer makes contributions to a defined benefit pension plan based on its estimate of the amount of money needed by the plan to pay for current and future liabilities. For a number of reasons, a plan can achieve a surplus over what is needed to fund current and future liabilities. If a surplus is created and the plan terminates, 1.e., ceases to exist, the surplus may revert to the employer if allowed by the plan. One of the benefits an employer receives by making contributions to a qualified ERISA pension plan is the right to deduct contributions to the plan for federal income tax purposes. As a condition for this favorable tax status, § 401(a)(7) of the Internal Revenue Code (the “Code”) provides that “[a] trust shall not constitute a qualified trust under this section unless the plan of which such trust is a part satisfies the requirements of § 411 (relating to minimum vesting standards).” Section 411(d)(3) of the Code, which was added by ERISA, requires that all qualified plans must provide for vesting (i.e., non-forfeita-bility) of benefits of employees affected by a plan termination or a partial plan termination. The purpose of Code § 411(d)(3) is to prevent the forfeiture of terminated employees’ benefits that have not yet vested and to prevent an employer from reaping a windfall by “mak[ing] deductible contributions on which he would enjoy a tax-free buildup of income, then terminate the plan and have all amounts revert back to him.” Tipton & Kalmbach, Inc. v. Commissioner, 83 T.C. 154, 160 (1984); see United Steelworkers of America v. Harris & Sons Steel Co., 706 F.2d 1289, 1298 (3d Cir.1983). Unlike a complete termination, a plan continues after a partial termination. The genesis of the partial termination concept was apparently a 1954 article by the IRS’s chief pension expert, Isidore Goodman, who explained that “[a]t times, it is easier to devour an object with several bites than it is to attempt to swallow it in one gulp. So it is with a plan which is chopped down little by little until it becomes merely an empty shell.” 32 TAXES 48, 53 (January 1954). Two kinds of plan-related activities can result in a partial termination: (1) The number of participants can be reduced, thereby causing the number of non-vested employees who forfeit their pension benefits to be greater than would otherwise be anticipated under actuarial formulas, or (2) future benefit accruals can be reduced, thereby increasing the potential for a reversion of plan assets to the employer upon termination of the plan. Commentators have referred to these two scenarios as “vertical” and “horizontal” partial terminations. Stuart M. Lewis, Partial Terminations of Qualified Retirement Plans—An Evolving Doctrine, 13 COMP. PLAN J. (BNA) 223 (1985). This case involves claims that both types of partial terminations occurred. Even though Chevron agreed, as part of the partial settlement, to vest all plaintiffs who were terminated during the period when .plaintiffs now allege that a vertical partial termination occurred (January 1, 1984, through June 30, 1986), the Court must nevertheless address plaintiffs’ partial termination claims because the remedy for a horizontal partial termination is to vest class members who were employed on July 1, 1986, the date of that alleged partial termination, and because plaintiffs seek relief in addition to vesting if they are successful on their vertical or horizontal partial termination claims. Although an employer may declare whether a partial termination has occurred, that determination is accorded no deference by a court; the question is one of law for the court to decide de novo. Anderson v. Emergency Medicine Associates, 860 F.2d 987, 990 (10th Cir.1988). Neither ERISA nor the Internal Revenue Code defines a partial termination. Aside from ease law, most of the authority addressing this question is found in Treasury Regulations, IRS revenue rulings, and the IRS Plan Termination Handbook. These sources offer analytical guidelines, but again do not define a partial termination. Although decisions affecting pension plans must conform to a vast array of detailed statutory and regulatory mandates, no precise guidance is provided on whether a change in plan members or benefits results in a partial termination. Instead, these questions are to be answered by a case-specific factual analysis. See Treas.Reg. § 1.411(d)-2(b)(1) (1977). B. Vertical Partial Termination The vertical partial termination rule is found within the “facts and circumstances” test of Treas.Reg. § 1.411(d)-2(b)(l). General rule. Whether or not a partial termination of a qualified plan occurs (and the time of such event) shall be determined by the Commissioner with regard to all the facts and circumstances in a particular case. Such facts and circumstances include: the exclusion, by reason of a plan amendment or severance by the employer, of a group of employees who have previously been covered by the plan_ (emphasis added) Perhaps because this rule gives so little guidance, it has spawned a number of complex issues, few of which have been resolved by controlling authority in this Circuit, and most of which are present in this case. Before addressing the issues, the Court will first describe them. 1. Significant Number or Percent IRS revenue rulings suggest that an employer-initiated permanent reduction of either a significant number or a significant percent of employees from a plan as part of a major corporate event may constitute a partial termination. Rev.Rul. 81-27, 1981-1 C.B. 228 (termination of 95 of 165 employees in connection with the closing of one of two divisions was a significant number that resulted in a partial termination); Rev.Rul. 73-284, 1973-2 C.B. 139 (when 12 of the 15 employees (80%) covered by a qualified plan declined to transfer to a new location and were terminated, a partial termination occurred because a significant percent of the employees were excluded from participating in the plan); Rev.Rul. 72-439, 1972-2 C.B. 223 (the significant percent test was met when 120 of the plan’s 170 participants (71%) became ineligible to participate in future employer contributions). Although courts have given lip service to the significant number test, they have been reluctant to use this test as a basis for holding that a partial termination has or has not occurred. See Tipton, 83 T.C. at 160 n. 5 (“[Since 34% and 51% are significant,] [w]e need not and do not decide whether a partial termination would occur where a significant number of participants but not a significant percent, are excluded from participation in a plan.”); Ehm v. Phillips Petroleum Co., 583 F.Supp. 1113, 1115-16 (D.Kan.1984) (since 415 terminated employees comprised only 2.5% of the plan participants, the court found no partial termination under the significant percent test and declined to apply the significant number test). No court has found a partial termination under the significant number test. In this case plaintiffs argue that the loss of approximately 10,000 Gulf Plan participants between January 1, 1984, and June 30, 1986, satisfies both the significant percent and the significant number tests. Plaintiffs concede that the only reported instance in which the significant number test has been used to find a partial termination involved the discontinuance of an employer’s business at a particular location with attendant layoffs. Rev.Rul. 81-27, 1981-1 C.B. 228. Neither the Court nor the parties have uncovered any instance in which the significant number test has been applied to a case such as the present one in which a large number of company-wide terminations is alleged to have resulted in a partial termination. 2.What Number or Percent is Significant? Problems arise in applying the significant number and the significant percent tests when the facts are not as manifest as those in the cases and revenue rulings cited above. The Internal Revenue Service has stated, and some courts have found persuasive, that an employer-initiated turnover rate in excess of 20% of a plan’s participants might be significant if it is coupled with other factors, such as the closing of a plant or division. These authorities state that termination of a lesser percent of plan participants, especially on a company-wide basis, could only be significant if the plaintiffs present evidence of egregious factors such as evasion of pension obligations or prohibited discrimination in favor of highly compensated employees. See Weil v. Retirement Plan Administrative Committee of the Terson Co., 913 F.2d 1045, 1052 (2d Cir.1990) (“Weil II”); Plan Termination Handbook §§ 252(8), 252(9)(d) and 252(10). 3.How is the Number or Percent Calculated? The significant percent and significant number analyses in this ease are further complicated by disagreements concerning: 1. whether terminated vested as well as non-vested employees should be considered in calculating the percent and number; 2. whether employees who transferred to a successor employer, and whose accrued pension benefits were transferred to a successor plan, should be counted; 3. whether defendants should be entitled to exclude from the number of terminations normal turnovers, what turnovers are “normal,” and who has the burden of proof of establishing the normal turnover rate; and 4. whether the percent and number of terminations should be calculated on an annual basis, as defendants argue, or whether the percent and number of terminations occurring from a significant corporate event over a two-and-one-half-year period will satisfy the tests, as plaintiffs argue. Because the number and percent of terminated Gulf employees must first be quantified before their significance can be judged, the Court will address these issues first. a. Vested or Non-Vested Terminations Many of the former Gulf employees who were terminated between January 1, 1984, and June 30, 1986, were vested under the Gulf Pension Plan. Defendants argue that these employees should not be considered in calculating either the percent or number of terminations. The partial termination rule is designed to prevent forfeiture of pension benefits that have not yet vested and to prevent a windfall to the employer through a reversion of money on which the employer has paid no federal income taxes. Including vested terminees does not further the first policy since they forfeit no benefits. Furthermore, Congress has taken a more direct approach to remedy the potential for a tax-free corporate windfall. When a pension plan is terminated, any assets that revert to the employer are not only included in the employer’s gross income, they are also subject to an additional excise tax. The Revenue Reconciliation Act of 1990 raised the excise tax on employer reversions from 15% to 20% and imposed a 50% excise tax unless the employer transfers part of the reversion to a replacement plan or provides more favorable benefits. The Court therefore concludes that in applying both the significant percent and the significant number tests, only non-vested terminations are relevant. However, out of an abundance of caution, the Court has calculated terminations both with and without vested terminees and concludes that the differences in percentages and numbers of terminations are not significant enough to affect the outcome of the vertical partial termination claim. b. Employees Who Transferred to a Successor Plan In 1985 Gulf sold certain refining and marketing facilities to Sohio, and Chevron sold the merged company’s Northeast and Ohio marketing assets to Cumberland Farms. In both divestitures the asset purchase agreements obligated the buyers to establish defined benefit pension plans with terms substantially similar to those of the Gulf Pension Plan. Incident to both divestitures Gulf and Chevron agreed to transfer pension assets in amounts at least equal to the accrued benefits of the transferring employees. Defendants argue that former Gulf employees who transferred to Sohio and Cumberland Farms should not be included as terminees for purposes of the vertical partial termination analysis. In Morales v. Pan American Life Ins. Co., 718 F.Supp. 1297, 1302 (E.D.La.1989), aff'd, 914 F.2d 83 (5th Cir.1990), the employer, PALIC, closed its Medicare Division on December 31, 1984. During the preceding months PALIC transferred 30 of the 159 Medicare Division employees to other jobs in PALIC. In calculating the percentage of employees who were involuntarily terminated, the Court held that “[t]he transferred employees should not be included with the number of employees involuntarily terminated as they remained PALIC employees and Plan participants ...” The Court is persuaded that the reasoning of Morales is also applicable to this case. Although the Sohio and Cumberland Farms transferees did not remain Gulf employees or members of the Gulf Pension Plan, assets from the Gulf Pension Plan sufficient to cover all of the transferred employees’ accrued benefits under the Gulf Plan were transferred to the Sohio and Cumberland Farms’ plans. For the reasons discussed above in connection with the vested, non-vested issue, neither of the policies underlying the partial termination rule would be served by including these employees as “affected employees” for purposes of determining whether a vertical partial termination occurred. In fact, to count such employees could deter employers from taking steps to protect employees in similar situations by potentially penalizing an employer whose terminated employees transfer to a successor employer that is obligated to continue the same or similar benefits and to whom plan assets have been transferred. The Court does not believe that either Congress or the IRS intended such a result. Cf. IRS Gen.Couns. Mem. 39,824 (August 27, 1990). c. Turnover Rate Since one of the purposes of the partial termination rule is to deter employers from terminating non-vested employees, it is generally agreed that in calculating the number or percentage reduction of plan members, only involuntary, employer-initiated, terminations should be considered. E.g., Anderson, 860 F.2d at 990. In cases involving relatively few employees it may be possible to resolve this issue by determining the facts surrounding each termination. More often, however, as exemplified by this case, it is not feasible to analyze individual terminations. To address this problem, the IRS Plan Termination Handbook allows an employer to exclude its normal turnover rate in determining whether the reduction in employees is significant. Section 252 provides in part: (6) ... The facts and circumstances must be considered in each case and may include the extent to which terminated employees are replaced, and the normal turnover rate in a base period. The base period ordinarily should be a set of consecutive plan years (at least two) from which the normal turnover rate can be determined, and should reflect a period of normal business operations rather than one of unusual growth or reduction. Generally, the plan years selected should be those immediately preceding the period in question. (7) The turnover rate is determined by dividing the employer-initiated terminations by the sum of the total participants at the start of the period and the participants added during the period. Employer-initiated terminations are generally all terminations other than those attributable to death, disability retirements and retirement at normal age. In certain situations, the employer may be able to prove that other terminations were also not employer-initiated. Defendants seek to exclude from consideration those Gulf employees who retired and terminated voluntarily between January 1, 1984, and June 30, 1986. Since defendants recognize that many of these departures, however labeled in defendants’ employment records, may have been prompted by the impending closings or divestitures of Gulf facilities and across-the-board reductions-in-force, defendants have calculated what they contend was Gulf’s “normal turnover rate” before the onset of the troubles described at the beginning of this opinion. Using the years 1978-1981 as the base period, Defendants’ Ex. 10 purports to calculate a “normal” Gulf turnover rate of 8.59% and a “normal” Gulf retirement rate of 2.43%. There are several problems with defendants’ 8.59% “normal” turnover rate. Although 1978 through 1981 was a period of normal business operations by Gulf, it was also a period when many new employees were hired each year. In contrast, during the period from January 1, 1984, through June 30, 1986, very few new employees were hired by Gulf. From 1978 through 1981 at least 3,500 new employees were hired each year; there were only 1,084 new hires in 1984, 331 in 1985, and 130 during the first six months of 1986. (Defendants’ Ex. 11) This distinction lessens the reliability of the 1978-1981 era as a base period. Plaintiffs’ actuarial expert, Mr. William A. Dreher, testified that fewer long-term employees are terminated either voluntarily or involuntarily. Thus, when an employer is continuously hiring large numbers of new employees, the turnover rate will be higher than when the work force is contracting. It is therefore not reasonable to assume that the 1978-1981 turnover rate would have continued in the 1984-1986 era. Plan Termination Handbook § 252(7) and Examination Tip (4)(a) and case law place the burden on the employer to prove that terminations are voluntary. See Morales, 718 F.Supp. at 1302-03. Although defendants tendered a blizzard of numbers, they presented no evidence that the 8.59% normal turnover rate would have continued in the 1984 through 1986 period. Absent such evidence, and considering the dramatic reduction in new hires and the various programs undertaken by Gulf and Chevron to reduce the work force during this period, the Court concludes that defendants have not proved either that there was a normal turnover rate from 1984 through June of 1986, or that the 8.59% rate urged by defendants represents such a rate. The Court will, however, exclude from its calculation of affected employees the 2.43% normal retirement rate reflected on Defendants’ Ex. 10. Retirement, unlike termination, is not a function of the rate of new hires. Under the Gulf Pension Plan any employee who had 75 points, which generally meant at least 20 years of employment, could retire. Logically, some retirements would have occurred from 1984 through June of 1986 even if Gulf and Chevron had not introduced new initiatives to reduce the size of the work force, and this rate would be unaffected by the low new-hire rate during this period. The Court is not persuaded by plaintiffs’ argument, based on § 252(7) of the Plan Termination Handbook, that only normal retirement at age 65 should be considered. Section 252(7) states that this is the general rule, but permits proof by the employer to the contrary. The evidence showed that Gulf employees could receive non-discounted early retirement at age 60, and that eligible Gulf employees could and did retire before age 60 with reduced benefits long before the Gulf/Chevron merger. However, since retirements were accelerated by the initiatives of Gulf and Chevron to reduce their work force, the Court will also treat the 2.43% normal retirement rate as a ceiling, and retirements in excess of that rate will be considered to be employer-initiated. In calculating the number of affected employees under the significant number and significant percent tests, the Court will exclude for each year a number of employees equal to the lesser of the actual number of retirements or 2.43% of Gulf Plan participants at the end of the year. d. Time Period Defendants argue that in determining whether a vertical partial termination occurred the Court should consider only the number of employees excluded within a single plan year. The Court does not find this position to be supported by logic or required by authority. Most reported vertical partial termination rulings and decisions did not involve massive corporate restructuring of the scale implemented by Gulf and Chevron. The facts in those decisions were therefore generally limited to a period of less than a year. There is nothing in the language of the rule itself, however, that requires that a significant corporate event occur within a year, and the ending of a calendar or plan year has no intrinsic relevance in making this evaluation. The IRS guidance found in § 252(7) of the Plan Termination Handbook instructs that “[t]he turnover rate is determined by dividing the employer-initiated terminations by the sum of the total participants at the start of the period and the participants added during the period.” (emphasis added) Both the IRS, in its Technical Advice Memorandum Concerning The Employee’s Retirement Plan of A & P Company, and the Second Circuit in Weil I, 750 F.2d at 12, have stated that a series of employer-initiated terminations related to the same event may be considered in determining whether a partial termination has occurred even if those terminations occur over a multi-year period. This Court likewise concludes that, assuming plaintiffs can prove it, a vertical partial termination may occur from a significant corporate event that manifests itself in employer-initiated terminations occurring over the two- and-one-half-year period urged by plaintiffs. e. The Relevant Number and Percent For purposes of the significant number test, the relevant number of terminations is the total number of plan members terminated, less: (1) those who were already vested, (2) those who voluntarily retired, including both discounted and undiscounted early retirees and disability retirees, (3) those who transferred to the Sohio and Cumberland Farms successor plans, and (4) those who died. To calculate the relevant percent the Court will divide the number obtained by the formula in the previous sentence by the sum of the non-vested plan participants at the start of each period and the non-vested participants added during that period. Application of these formulas to the facts yields the following chart, which the Court finds to be established by the evidence. Total 2.5 years 10,947 Jan.-June 1986 1984 1985 2,623 6,588 Total terminations Deductions: (1) vested terminees ^ -7] to 03 CCO CO CD 03 (2) retirements l — 1 OO -a t-H CO ID CO CO ^ (3) transfers to Sohio & Cumberland Farms ID ^ 03 t-CD O O (4) deaths CD t-H O CO 03 ID Total Deductions 798 2,909 920 Total relevant terminations 6,427 1,825 3,679 Non-Vested Members at Beginning of period 12,688 O 03 IQ i — ‘ t-L CO ^ New members added 1,084 id ^ LO O CO r — I CO H Total non-vested members 13,772 CO CO 03 O »D CD y-1 J — 1 CO O Ol Percentage Decrease 13.3% 32.4% 12.1% 45.2% Alternatively, calculating the relevant number and percent by including vested plan participants m both the numerator and the denominator, yields the following chart: 4. Significant Corporate Event The parties differ over the number of significant corporate events that occurred between January 1, 1984, and June 30, 1986. Defendants contend that two distinct corporate events occurred during this period: the merger of Gulf and Chevron in 1984 and 1985, and then a drastic decline in oil prices with resulting layoffs, which began in late 1985 and culminated in 1986. Plaintiffs argue, and the Court finds, that the business decisions by Gulf and Chevron to reduce the work force all resulted in a single corporate event throughout this two-and-one-half-year period. After the merger between Gulf and Chevron was announced in February 1984, Chevron and Gulf began a series of actions designed to reduce their partially redundant work force. In March of 1984 Gulf suspended hiring new employees. In May of 1984 Chevron announced and implemented a severance plan for involuntary terminations occurring between April 27, 1984, and April 26, 1986. (Plan 728; Plaintiffs’ Ex. 289) Later in 1984 Gulf adopted a three-phase Surplus Manpower Reduction Program to reduce merger-related surplus employees. Under the first phase of the program, announced in November 1984, Gulf offered a Voluntary Termination Incentive Program (“VTIP”) to provide sever-anee pay to employees working in locations where Chevron and Gulf had identified a surplus in their work forces. Under VTIP employees were required to come forward during November and December 1984 and volunteer to terminate their employment. Gulf and Chevron required these employees to remain on the job, however, until they were finally released, and most actual VTIP terminations did not occur until the end of 1985. In the second phase of the program Gulf employees were offered transfers or demotions, and were paid severance if they chose not to accept. The third phase involved involuntary terminations. By late November of 1985 this three-phase program had resulted in 4,468 terminations: 923 by voluntary severance, 1,486 by “transfer/demote severance,” and 2,059 by involuntary severance. (Plaintiffs’ Ex. 484 at p. 3) At the same time, and also as a result of the merger, Chevron began selling off various assets of Gulf and Chevron, both to reduce redundancies and to finance the enormous debt it had incurred to acquire Gulf. These divestitures, which eliminated thousands of additional Gulf employees, included the sale of Gulf’s explosives group to Thermex, the sale of Gulf Oil Trading Company to GOTCO N.V., the sale of the Cedar Bayou polypropylene plant to Amoco Chemicals, the donation of the Harmarville Research Center to the University of Pittsburgh, the shut-down and sale of the Gulf headquarters building in Pittsburgh, and a number of other divestitures, including those to Sohio and Cumberland Farms discussed above. Although the various employee reduction programs and divestitures were initiated in 1984 and 1985, some of the affected employees were not actually terminated until the first half of 1986. For example, the shut-down of Gulfs Harmar-ville Research Center and its Pittsburgh headquarters occurred in gradual increments during this two-and-a-half-year period. Chevron argues that in 1986 additional terminations occurred because of the steep decline in oil prices that began in late 1985 and that these terminations were not merger-related and should not be included as part of any “merger-related” significant corporate event. The evidence shows that Chevron’s response to the decline in oil prices was to formulate, in the spring of 1986, a new retirement enhancement program known as the Special Retirement Allowance Program (“SRAP”). However, this program was not even announced to employees until June 1986, and the program did not begin until July 1986. (Plaintiffs’ Ex. 865) The Court finds that none of the employee terminations that occurred during the first six months of 1986 were due to the decline in oil prices that occurred during that era. Instead, the Court finds that those terminations were the vestigial effects of merger-related actions — both across-the-board reductions-in-force and the divestitures — that began in 1984 and 1985. The Court finds that all of the terminations between January 1, 1984, and June 30, 1986, that are reflected in the charts above were part of a single, significant, merger-related corporate event. 5. The Number and Percent are Significant There were 6,427 employer-initiated terminations of non-vested Gulf employees during the single, merger-related, significant corporate event that occurred between January 1, 1984, and June 30, 1986. This represented a 45.2% reduction in the total number of non-vested Plan participants during the period. The Court finds that this number and this percent, and the 8,534 terminations and 34.7% decrease under the alternative analysis, are significant and resulted in a partial termination of the Gulf Pension Plan under both the significant number and the significant percent tests. The Court is led to the conclusion by the magnitude of this reduction in plan membership, by the fact that when these reductions were occurring defendants were planning to reduce the benefits available to those Gulf employees who remained, and by the increased potential for a reversion because of these terminations, which occurred in an atmosphere in which Chevron was considering how to revert surplus Gulf Plan assets for its general corporate use. See Plan Termination Handbook § 252(4), (8)(a), and (10). 6. The IRS Determination Defendants argue that the Court should show deference to a determination by the IRS that no partial terminations of the Gulf Pension Plan occurred. Months after this action was filed, Chevron, on the advice of its trial counsel, requested its San Francisco counsel, Pillsbury, Madison and Sutro (“PM & S”), to seek a determination from the San Francisco Regional Office of the IRS that a partial termination of the Gulf Pension Plan had not occurred. On May 1, 1987, PM & S submitted a letter accompanied by an IRS Form 5300 seeking a determination that the new Chevron Retirement Plan was a qualified plan under ERISA. Although IRS Form 5300 contains a box to check if a plan sponsor requests a determination on the issue of partial termination, Chevron did not check this box (Defendants’ Ex. 18 at p. 6, line 3.a.(iv)), and PM & S’ letter neither requested a determination on the partial termination issue nor submitted any information regarding this issue. In a separate cover letter that accompanied the May 1, 1987, request, PM & S referred to this lawsuit and requested expedited treatment of the application for determination. (Defendants’ Ex. 18 at p. 1) As required by IRS procedures, Chevron gave notice of the May 1, 1987, request to Gulf Plan participants who were still on its payroll and to certain collective bargaining representatives, but gave no notice to former employees who were members of the then alleged class in this action. Since the request did not seek a determination on the partial termination issue, the notice made no mention of that issue. (Defendants’ Ex. 17) On September 2, 1987, PM & S received questions from the IRS regarding the determination request, which PM & S labeled as “very limited and quite innocuous.” (Defendants’ Ex. 15 at pp. 2 and 3) In a September 2, 1987, letter to Chevron, PM & S stated that in addition to the information sought by the IRS, Chevron’s submission “will also include a request that the Service make a determination that there was no partial termination of the Gulf Plan during the years at issue in the Borst litigation. Generally speaking, reviewers like to move active cases along very quickly, so we should receive a prompt response to this filing.” (Defendants’ Ex. 15 at p. 1) True to its word, on October 5, 1987, PM & S hand-delivered to the IRS District Director in San Francisco a letter responding to the IRS’ questions. (Defendants’ Ex. 16) In one paragraph, comprising a third of a page, PM & S also requested a determination that there was no partial termination of the Gulf Pension Plan during the years 1982 through June 30, 1986. (Id. at p. 5.) Attached to the letter were three pages of Chevron calculations showing the percentage reductions. Neither the letter nor accompanying calculations provided any discussion of the events surrounding the numbers presented. PM & S did not notify either present or past members of the Gulf Pension Plan or class counsel in this action of this “modification” of its determination request. Although the head of the PM & S Employee Benefits and Deferred Compensation section admitted that if the partial termination determination request contained in the October 5, 1987, letter had been considered as a separate determination request, a separate notice to interested parties would have been required under IRS regulations, she opined that no notice was required of the October 5, 1987, letter because the request was included in a response for information requested by the IRS. On December 9,1987, the IRS District Director in San Francisco issued a favorable determination to Chevron. (Defendants’ Ex. 19) The determination is one page in length, contains no analysis of issues to which this Court has .devoted over 20 pages, and appears to be largely a form letter. ERISA requires notice to interested parties each time the sponsor of a tax qualified plan requests an IRS determination letter. § 3001(a) of ERISA, 29 U.S.C. § 1201(a). The notice to interested parties must specify procedures by which interested parties (generally participants in the plan and their collective bargaining agents) can obtain copies of the materials filed with the IRS, and can file written comments with the IRS, or request the Department of Labor to file comments with the IRS. Rev.Proc. 80-30, 1980-1 C.B. 685, §§ 6 and 7. The notice to interested parties must also indicate the subject matter of the determination letter request. Id. at § 7.03(4). In the case of a plan termination, including a partial plan termination, former as well as current employees are interested parties who must receive notice. Id. at §§ 3.01(4) and 6.01; Treas.Reg. § 1.7476-l(b)(5) (1976). The purpose of these notice requirements is to afford parties who may have views different from the plan sponsor to make those views known to the IRS before it makes a determination. Chevron’s failure to notify any Gulf Pension Plan participants that it was seeking a determination of the vertical partial termination issue was contrary to the IRS regulations and the underlying policy requiring notice. This failure also violates fundamental fairness and the reasonable claims procedure requirements of § 503 of ERISA, 29 U.S.C. § 1133. Apart from the serious procedural defect in the way Chevron obtained the IRS determination letter, the Court also finds that the IRS determination is due no deference because it evidences no investigation or legal analysis of the facts by the IRS. For both reasons the Court concludes that the IRS determination letter is entitled to no weight, and the Court has given it none. C. Horizontal Partial Termination The horizontal partial termination rule is found in Treas.Reg. § 1.411(d)-2(b)(2). Special rule. If a defined benefit plan ceases or decreases future benefit accruals under the plan, a partial termination shall be deemed to occur if, as a result of such cessation or decrease, a potential reversion to the employer, or employers, maintaining the plan (determined as of the date such cessation or decrease is adopted) is created or increased. If no such reversion is created or increased, a partial termination shall be deemed not to occur by reason of such cessation or decrease. The rule has two elements: a cessation or decrease in future benefit accruals in a defined benefit plan and a resultant creation or increase of a potential reversion to the employer. The rule does not require that an employer actually attempt to achieve a reversion, only that a potential for reversion exist because of a cessation or decrease of future benefit accruals. The partial termination claim arises because of changes in the benefits that were available to former Gulf employees under the Gulf Pension Plan and those available to them after the Gulf Pension Plan and the Chevron Annuity Plan were merged into the new Chevron Retirement Plan effective July 1, 1986. Plaintiffs argue that the effect of the plan merger was to reduce substantially future benefit accruals to former Gulf employees, thereby increasing the potential for a reversion to Chevron upon any final termination of the merged plan. As evidence of this decrease in future benefit accruals, plaintiffs cite a September 25, 1985, presentation to the Chevron Executive Committee analyzing the prospective plan merger. (Defendants’ Ex. 353) The presentation projected that the Chevron Retirement Plan would achieve a decrease of $102 million in present value liability through net decreases in benefits that would otherwise have been paid to former Gulf employees under the Gulf Pension Plan. (A $162 million increase in present value liability was projected for improved benefits to former Chevron employees under the merged plan.) Another schedule presented to the Executive Committee at the same presentation (Plaintiffs’ Ex. 529) showed that this $102 million decrease was achieved because the increase in the present value of the more favorable lump-sum death benefits to former Gulf employees under the merged plan (+ $30 million) was more than offset by four other changes: (1) undiscounted retirement at age 60 under the Gulf Plan was ended, thus eliminating future accruals of this benefit for service after the July 1, 1986, Plan amendment ( — $17 million); (2) the immediate payment of disability retirement benefits to employees with at least 15 years of service under the Gulf Plan based on the level of compensation and years of service to the date of disability without any early retirement discount was eliminated and substituted with an actuarially discounted benefit ( — $65 million); (3) the 40% post-retirement spousal annuity was frozen, thereby eliminating ' future accruals of this benefit for service after the July 1, 1986, plan amendment ( — $31 million); and (4) the pre-Social Security “bridge” benefits provided for Plan members whose early or disability retirement benefit began before age 62 was restricted to service up to the July 1, 1986, Plan amendment, thus eliminating future accruals of this benefit ( — $19 million). Analyzing the same changes in pension plan benefits, plaintiffs’ actuarial expert, William A. Dreher, independently calculated a reduction in pension liability of $118,-563,000. (Plaintiffs’ Ex. 1255 at p. 14) After considering improvements in SRAP benefits to former Gulf employees under the Chevron Retirement Plan, Dreher testified that the net reduction in future benefit accruals to former Gulf employees was $83,803,000. (Plaintiffs’ Ex. 1254, column 4.b) Dreher testified that on June 30, 1986, the Gulf Pension Plan had fully funded all past and future benefit accruals for present employees and was overfunded on an ongoing Plan basis by $125 million. According to Dreher, the net effect of the Gulf Pension Plan changes enacted by Chevron was to increase substantially the already overfunded status of the Plan, thereby also increasing the amount of any reversion to Chevron were the Chevron Retirement Plan terminated, or, in any event, allowing Chevron to delay the amount or timing of further contributions to the merged plan.. Defendants mount both factual and legal defenses against the alleged horizontal partial termination. At trial Chevron witnesses Neil Darling, who in 1985 was responsible for financial affairs of all Chevron pension plans, and Alex Ross, who was then manager of the Chevron corporate benefits staff, attempted to deflate the size of the $102 million projected decrease and to disavow its importance in several ways. First, Darling sponsored a chart designed to show that the $102 million estimated decrease was inaccurate by $59 million and that the correct estimated decrease was only $43 million. (Defendants’ Ex. 348) When other non-pension plan benefits available under the Chevron Profit Sharing/Savings Plan were also considered, Darling’s chart purported to show that the present value of future benefits to former Gulf employees under the Chevron Retirement Plan actually increased by $32.5 million. Ross reiterated Darling’s contentions and testified that the $102 million decrease was not intended to represent a reduction in benefit accruals to Gulf Plan members. He also testified that because of a recently discovered oversight, the —$102 million figure was unreliable because it failed to include more favorable benefits under the Chevron Annuity Plan which, although carried forward into the merged Chevron Retirement Plan, were not considered in the 1985 presentation to the Executive Committee or in other contemporaneous Chevron documents. The Court is not persuaded by Chevron’s attempt at trial to disassociate itself with figures developed over a number of months and presented by the manager of its corporate benefits staff to the Chevron Executive Committee at the conclusion of extensive planning by Chevron to evaluate the cost and method of merging the Gulf Pension Plan and Chevron Annuity Plan. The Court does not find Defendants’ Ex. 348 to be relevant because some of the offsetting adjustments used to moderate the $102 million decrease to a $43 million decrease, e.g., replacement of lump-sum death benefits with life insurance coverage and replacement of disability retirement benefits with a company-paid, long-term disability insurance plan, do not affect the liability of the Gulf Pension Plan or the Chevron Retirement Plan. Although these changes and the others discussed in the next two paragraphs may be relevant to the general effect of the plan merger on former Gulf employees, they are not relevant in determining whether there was a decrease in the present value of Gulf pension plan liability, which is the issue represented in defendants’ 1985 documents (Plaintiffs’ Ex. 529; Defendants’ Ex. 353), or whether there was a decrease in future benefit accruals under the Chevron Retirement Plan within the meaning of the horizontal partial termination rule, Treas.Reg. § 1.411(d)-2(b)(2). For the same reason the alleged benefit that former Gulf employees now receive under the Chevron Profit Sharing/Savings Plan, which Chevron argues resulted in an increase in Chevron’s liability by $32.5 million, is irrelevant in considering a possible horizontal partial termination of the Gulf Pension Plan. Nor is the Court persuaded by Ross’ testimony that improved benefits to former Gulf employees under the Chevron Retirement Plan should be considered as offsetting any possible effect of the $102 million decrease presented to the Chevron Executive Committee. Defendants’ Ex. 363 lists these additional improvements, some of which are also quantified in Defendants’ Ex. 348. Some of these improvements, such as the ability of a former Gulf employee to retire from Chevron and elect a lump-sum payment instead of an annuity, were not considered to be plan benefits during the plan merger discussions because of their optional nature. (Plaintiffs’ Ex. 212, October 17, 1985, Memo from Carter to Ross) There was even disagreement on the witness stand between Darling, who thought this option was not a benefit, and Ross, who now thinks it is. Both witnesses agreed, however, that it would be very difficult to quantify the effect of this alleged benefit because a former Gulf employee who elected a lump-sum pension payment option would lose the benefit of the 40% annuity his spouse would otherwise be entitled to receive (a benefit based on the employee’s past as well as future service) and would likewise lose the benefit of any future AVIS pension enhancements that Chevron might declare. Ross attempted to demonstrate the effect of five of the plan improvements listed in Defendants’ Ex. 363 (and to refute Dre-her’s testimony) through exhibits showing the effect of the alleged Chevron improvements on five categories of former Gulf employees: Case A — Terminated Vested Benefits, Case B — Disability Benefits, Case C — Death Benefits, Case D — Early Retirement Benefits, and Case E — a different Early Retirement Benefit scenario. (Defendants’ Exs. 361 and 362) Ross admitted, however, that if non-pension welfare benefits and benefits under the Chevron Profit Sharing/Savings Plan were not considered, and the focus was limited solely to improved pension benefits under the Chevron Retirement Plan, former Gulf employees who fell within Cases D and E would be better off under the Gulf Pension Plan. He also admitted that under that scenario it would not be possible to make an accurate comparison of benefits under the two plans for Cases A and B and that only under Case C — Death Benefits, would former Gulf employees be better off under the Chevron Retirement Plan. Ross acknowledged that very few employees would fall under Case C and, more importantly, he testified that none of the five cases compared normal retirement benefits under the two pension plans. The most important factor in the Court’s analysis of defendants’ factual defense, however, is that Chevron’s exhibits and supporting testimony are all admitted after-thoughts developed in the course of this litigation. Chevron documents from 1985 show that Ross and Darling then believed and told the Chevron Executive Committee that the new Chevron Retirement Plan would create a decrease of $102 million in the present value of benefits otherwise payable under the Gulf Pension Plan. Chevron’s litigation position that this 1985 estimate was unreliable because of an oversight in the way it was calculated, or its failure to also address non-pension benefits or additional benefits under the Chevron Retirement Plan, does not ring true. There was abundant evidence besides the documents prepared for the September 25, 1985, Executive Committee presentation that Chevron anticipated and planned for a reduction in Gulf pension benefits. In an August 8, 1985, memo to Ross (Plaintiffs’ Ex. 190), Carter stated that “Gulf participant liability has been reduced $205.2 million while Chevron participant liability has increased $176.7. This reflects the deliber-alization of some Gulf Plan provisions ...” The plan design described in this memo went through several modifications, e.g., Plaintiffs’ Ex. 138, September 9, 1985, memo from Carter to Ross, and ultimately resulted in the benefit structure that achieved the $102 million reduction presented to the Executive Committee on September 25, 1985Í The March 5, 1984, corporate merger agreement between Gulf and Chevron limited to two years Chevron’s obligation to continue pension benefits to former Gulf employees at the pre-merger level. (Plaintiffs’ Ex. 497 at § 6.10) On July 1, 1986, less than three months after the expiration of this two-year period, Chevron merged the Gulf Pension Plan and the Chevron Annuity Plan into the Chevron Retirement Plan, and reduced benefits to former Gulf employees. Section 204(h)(1) of ERISA, 29 U.S.C. § 1054(h)(1), requires a plan administrator to give plan members written notice of any plan amendment that provides “for a significant reduction in the rate of future benefit accruals].” In June of 1986, shortly before the plan merger, Chevron filed a notice under § 204(h)(1) articulating seven ways in which the July 1, 1986, amendments to the Gulf Pension Plan resulted in such reductions of benefit accruals. (Plaintiffs’ Ex. 552) Four of the reductions listed were those quantified in arriving at the $102 million decrease shown on Plaintiffs’ Ex. 529. In assessing defendants’ factual defenses to plaintiffs’ horizontal partial termination claim, the Court finds that defendants’ contemporaneous belief, not a position developed for trial, is most probative of the truth. The contemporaneous evidence establishes that when Chevron was planning to merge the two plans it consistently considered reducing benefits available under the Gulf Pension Plan and ultimately concluded that these reductions would result in a $102 million decrease in Chevron’s pension plan liability to former Gulf employees. Defendants also argue that the Gulf Pension Plan changes quantified in Plaintiffs’ Ex. 529, and in particular the change in disability retirement benefits, cannot be considered in determining if a horizontal partial termination occurred because the horizontal partial termination rule only protects “accrued benefits,” and these four categories of reduced benefits are “non-accrued,” or “ancillary benefits.” This argument is premised upon the contention that the term “benefit accruals” in the rule refers to the rate at which an employee earns an “accrued benefit,” which is defined, in the case of a defined benefit plan, by Code § 411(a)(7)(A)(i) as “the employee’s accrued benefit determined under the plan and ... expressed in the form of an annual benefit commencing at normal retirement age....” According to defendants, none of the four benefits that were reduced meet this definition because they did not have a specified accrual rate and did not commence at normal retirement age. The crux of defendants’ argument is that because Code § 411 “protects only accrued benefits” (Defendants’ Post-Trial Brief Regarding Horizontal and Vertical Partial Termination at p. 3), the horizontal partial termination rule adopted to implement this section of the Code must also be limited in scope to protecting future accruals of accrued benefits. Defendants argue that the rule does not prohibit an employer from prospectively eliminating or reducing ancillary benefits. The Court is not persuaded by this argument for several reasons. Defendants have cited no authority, and the Court has found none, that would so circumscribe Code § 411 or the horizontal partial termination rule. Although § 411 is indeed full of references to “accrued benefits,” the only reference in § 411(d)(3) to benefits, whether accrued or otherwise, occurs in addressing the remedy for a plan termination. Section 411(d)(3) provides that if a termination or partial termination has occurred, the remedy is to vest “all affected employees to benefits accrued to the date of such ... partial termination ... to the extent funded as of such date.... ” Given the repeated use of the term “accrued benefits” in parts of § 411 other than § 411(d)(3), the Court does not conclude that the failure of § 411(d)(3) to define a partial termination by reference to a particular type of benefit, and the failure of Treas.Reg. § 1.411(d)-2(b)(2) to speak in terms of a plan amendment that decreases future accruals of “accrued benefits,” were drafting oversights. ' Section 252 of the Plan Terminatio