How to avoid a splitting headache.

Dividing pension-plan assets during a sale is often confusing and potentially costly. The trick is to get your assumptions right.

It's just like a scene from a long-running play. The exhausted negotiators for a corporate sale have one more wrinkle to iron out: splitting the pension benefits to be transferred from the seller's plan to the buyer's. But there's a twist in the action. No one knows how much of the assets to transfer along with the benefits. As the other players wait in the wings, the negotiators cast about frantically for guidance.

Many business divestitures or acquisitions involve the transfer of employees from the seller's pension plan to the buyer's plan. When the time comes to divide pension liabilities and plan assets, a process called a spinoff, the basis for determining the amount of assets can be the most bitterly contested section of the sale agreement. Unfortunately, the negotiating teams may be unaware of the effect of legal restrictions on the transfer amount, so the final calculations could topple the most carefully crafted compromise.

Current regulations aren't much help. The Employee Retirement Income Security Act and the Internal Revenue Code provide only theoretical guidelines. As a result, it's usually up to the actuary for the seller's plan to declare the transfer in compliance with the law and the IRC. Therefore, actuarial assumptions are the tools for enlarging or shrinking an asset transfer, and selecting them requires expertise in spinoffs, discretion and cautious advocacy.

Pension-plan spinoffs are governed by Sections 208 of ERISA and 414(1) of the IRC. These require you to ensure that participants in both the spinoff group and the remaining group are as well-protected if the plan were to terminate after the spinoff as they would be if the plan had done so immediately before. If the original plan is fully funded on a termination basis, both segments must remain that way after the spinoff. In a corporate sale, surplus assets can be allocated with total discretion, subject to any special restrictions in the plan document.

TREAD CAREFULLY

The quandary is ascertaining how much money your company needs to fully fund the hypothetical plan termination. A financial executive might expect that the accumulated-benefit obligation (roughly equivalent to the value of accrued benefits) under SFAS 87 would be appropriate, while an ERISA attorney might seek the answer in the regulations under IRC 414(1). However, following either rule literally could be not only wrong but costly. An intended asset transfer of $20 million can readily become $12 million or $26 million, depending on the actuarial assumptions.

As a guide to how far you can go in selecting both legal and financially favorable assumptions, let's look at a case scenario. Assume that your company sponsors a well-funded plan with $100 million of assets and generous early-retirement subsidies. The plan covers primarily nonretirees. Twenty percent of the participants (representing an identical demographic slice of the entire plan) are being transferred to the plan of an independent buyer. How do you determine the minimum and maximum amount of assets that you can spin off?

Since the objective is simulating a plan termination, the first step is calculating what benefits the plan would pay when terminated, excluding any Pension Benefit Guaranty Corp. top-ups (the amount PBGC would kick in if a company couldn't fulfill its entire pension-benefit obligation). All accrued benefits as of the spinoff date, as well as all early retirement subsidies, annuity forms and other rights and features attached to these benefits would be included. Other benefits, such as death or disability provisions and subsidies that might become payable due to a future plant shutdown, can create tremendous additional liabilities. These benefits can be removed from the plan before a future termination. But unless they've already been removed before the spinoff, it may be difficult to ignore them in the spinoff calculations. Assume, however, that the two parties agree on this aspect.

The minimum "bullet-proof" asset transfer is the amount a financially sound insurance company would charge to annuitize the spinoff benefits, because an annuity purchase would be necessary if the plan were to terminate. Both participating and nonparticipating annuities are acceptable to PBGC. However, because the regulations under IRC 414(1) permit the actuary to step into an annuity carrier's shoes and to estimate what its cost might be, neither the plan sponsor nor the annuity carrier has to undergo this shopping expedition. Since the plan doesn't actually terminate at the spinoff date, you could even consider the cost of a structured bond settlement as a reasonable proxy for an annuity purchase.

Anyone may select the assumptions and methods used to determine the total amount of assets to be transferred (both the minimum amount and any additional amounts on which the parties agree), provided the enrolled actuary certifies that the company has met the legal requirements and any restrictions in the plan document have been honored. The actuary starts by selecting the appropriate actuarial assumptions on lifespan (mortality table), retirement age, interest discount, rates of employee turnover and/or disability. These same types of assumptions are used to determine IRS minimum annual funding requirements and for SFAS 87 calculations, although the levels used for funding and expense usually would not be selected to simulate an annuity purchase. It would also be possible to adopt the demographic assumptions used for these calculations but to substitute a Treasury bond rate of appropriate duration as the interest rate.

AN UNSAFE HARBOR

Even more confusion arises because of the regulations under IRC 414(1), which offer as a safe harbor those assumptions that the PBGC uses as of the date of spinoff. The problem with this approach is that the assumptions historically have been quite conservative and don't recognize employee turnover. Also, since the promulgation of the 1981 regulations, all PBGC rules have been completely revamped without a concurrent revision in these regulations. Interestingly, the PBGC itself has just replaced the assumptions with a new basis, effective as of November 1, 1993. Moreover, the IRS now considers only the PBGC interest and mortality assumptions, used together, satisfactory for its requirements. It no longer deems PBGC retirement-age assumptions automatically reasonable.

Therefore, as a plan sponsor, your company is left with the actuary's professional judgment as the only protection against an illegally large or small transfer amount. You can take solace in recent court cases that have reaffirmed that there are a whole range of reasonable assumptions. However, you must recognize that under ERISA, the enrolled actuary's duty is to act in the interests of the plan participants.

The table on page 28 shows asset-transfer amounts for an April 1993 spinoff from a $100-million plan. While a different date would change the magnitude of the asset transfer, the relative positions and conclusions would remain the same. All of the possible asset amounts have been used or considered for recent spinoffs. While the interest-discount assumption usually attracts the most attention, it's often the retirement-age assumption that has the most financial impact and is the most difficult to select.

Also, as assumptions move to the more conservative end of the spectrum, a plan's surplus can evaporate, and the plan may even become underfunded. Note that this occurs in the first example shown, where the remaining 80 percent of the participants would be left with only $74 million of assets -- an illegal outcome because it leaves the remaining participants with less than the minimum amount of assets required under IRC 414(1). And when interest rates change rapidly, a specific interest rate negotiated in advance of the spinoff may prove unacceptable by the time the transfer date rolls around. The use of the SFAS 87 discount rate may seem a simple solution, since it must track current annuity or bond markets. However, many plan sponsors are using rates that favor year-to-year stability over close tracking of these markets.

Assumptions that actuaries have used in the past may no longer be TABULAR DATA OMITTED appropriate for this calculation. For example, future retirement patterns may differ significantly from the plan's past experience, because of business restructuring, differences in pension-plan provisions and retiree medical availability. Plans with employee contributions or entitlements to surplus present serious issues of equity. Underfunded plans involve a different set of questions. Consider, for example, the 1986 Pension Protection Act, which prohibits a solvent plan sponsor from terminating an underfunded plan. Inexpertly drafted or confusing sales agreements can cause thunderous disputes concerning which assumptions the parties even intended.

As actuarial assumptions come under increasing scrutiny from IRS audits, Financial Accounting Standards Board requirements and the growing sophistication of the financial community, corporate negotiators have found their actuaries can be valuable allies, if they discuss pension issues frankly and in advance. A creative actuary's experience can frequently help break through a negotiating impasse. An actuary called to the scene too late, however, may bear the bad news that the intended spinoff runs afoul of professional judgment, actuarial standards and legal compliance.

Ms. Riebold is a managing director in the New York office of William M. Mercer Inc.