The Liability Nature of Unfunded Pension Obligations since ERISA
从理论和实证上研究了ERISA前后固定收益养老金计划未备基金义务的负债性质，发现ERISA及PBGC保险实际上降低了其负债性质，并提出了模型修正。
The liability nature of the unfunded obligations of defined benefit pension plans is examined theoretically and empirically in post-ERISA and pre-ERISA periods. The theoretical model suggests that if the unfunded obligations of a firm's pension plan are considered by the market as liabilities, these obligations should be positively related to the systematic risk of the sponsoring firm's stock. The empirical analysis indicates that ERISA and the insurance provided by the PBGC actually decreased the liability nature of unfunded pension obligations. The authors suggest modifications to the theoretical model and extend the model to include consideration of the Single Employer Pension Plan Amendment Act. The nature of unfunded pension obligations has increased in importance since the passage of the Employee Retirement Income Security Act (ERISA) of 1974. Prior to ERISA, unfunded pension obligations were not corporate liabilities. The legal claim of beneficiaries upon termination of a plan was limited to the assets of the pension plan. ERISA created a complex plan termination insurance system for defined benefit pension plans which exposes an employer to financial liability if a plan is terminated and there are unfunded liabilities for vested benefits. Viewing pension obligations as liabilities, the Financial Accounting Standards Board (FASB) recently issued standards1 which will require to report a liability on the balance sheet in certain circumstances. Rodger G. Holland is Assistant Professor of Accounting at Florida State University. He earned the Ph.D. at Ohio State University where he was a Coopers & Lybrand fellow. Holland has published in Issues in Accounting Education and the Journal of the American Taxation Association. Nancy A. Sutton is Assistant Professor of Risk Management and Insurance at Florida State University. She earned the PH.D. at University of Georgia and is a CLU. The authors appreciate the helpful suggestions of an anonymous reviewer on an earlier draft of this paper. 'Financial Account Standards Board [19] and [20]. Casual analysis would suggest that ERISA increased the liability nature of unfunded pension obligations. It forced minimal funding, required vesting within 15 years, and held the employer liable for unfunded benefits up to 30 percent of the employer's net worth. This content downloaded from 157.55.39.69 on Sun, 06 Nov 2016 04:13:55 UTC All use subject to http://about.jstor.org/terms Unfunded Pension Obligations 33 ERISA created the Pension Benefit Guarantee Corporation (PBGC) to assure payment of vested benefits (up to substantial maximums)2 of defined benefit pension plans and authorized the PBGC to collect insurance premiums and the contingent employer liability established by the law. The original act limited the contingent employer liability to 30 percent of the employer's net worth. The liability which exceeded this amount was assumed by the PBGC and funded by a $1 per participant annual premium. The rate for multiemployer plans was initially set at $.50 because such plans were considered safer than single employer plans. Although this was increased to $2.60 in 1978, the premiums proved inadequate for the liability assumed by the PBGC and was recently increased to $8.50. The PBGC has generally been in a deficit position since it began operations in 1974. The traditional accounting view suggests that unfunded pension obligations are perceived by the market as long-term debts and should be treated as economic liabilities.3 However, unfunded pension obligations may not be perceived by the market as long-term debts of a firm for three interrelated reasons. First, some of the liability is transferable to the PBGC. Second, the flat insurance rate charged by the PBGC may give an insurance bargain to the firm that maximizes the difference between unfunded pension obligations and the fixed cost of the insurance. Third, to the extent that the insured event (plan termination) could be manipulated in a manner that was profitable for the firm, the market may not have perceived pension obligations as long-term debts of a firm. In addition to changing the per participant annual premium, recent modifications have changed the liability and termination provisions. These provisions are discussed in a separate section entitled Effects of Modifications to ERISA. Firms with equal numbers of employees are required to pay equal insurance premiums to the PBGC regardless of their funding strategy or the financial strength of the pension plan. Therefore, (particularly those whose unfunded pension obligation exceeded the maximum amount that could be collected by the PBGC) were indirectly given a financial incentive to minimally fund a plan since the plan could be terminated at some future time under circumstances where the PBGC assumed payment of unfunded pension obligations. The original insurance provided by the PBGC created an incentive for to terminate a plan any time the unfunded pension obligations exceeded 30 percent of the firm's net worth. An options pricing framework has been used by several authors for analyzing the optimal financial strategy for with defined benefit pension plans.4 Sharpe [49] reasoned that the optimal corporate policy, given the fixed premium charged by the PBGC, is the one that maximizes the difference 2The maximum is the lesser of 100 percent of the average wages of the worker during his or her five highest-earnings years or a dollar maximum that varies with the Social Security taxable wage base. This dollar maximum was initially set at $750 per month and had increased to $1,857.95 per month by 1987. 'See Stone [52] and Werner and Kostolansky [57] for a discussion of this view. 4The options pricing framework has been used by Bicksler and Chen [9], Bulow [11], Lengetieg, et al. [31], Sharpe [49] and Treynor [55]. This content downloaded from 157.55.39.69 on Sun, 06 Nov 2016 04:13:55 UTC All use subject to http://about.jstor.org/terms 34 The Journal of Risk and Insurance between the value of the insurance (the insured unfunded pension obligations less the contingent employer liability) and its fixed cost (the PBGC premium). Elaborating on this, Bicksler and Chen [9] concluded that the value of the insurance decreases as the value of pension assets increases, and conversely, the value of the insurance increases as the value of pension assets decreases. Marcus [35] has shown that the liabilities of the PBGC can be extremely sensitive to a firm's funding policy and some problem firms may derive considerable value from the pension insurance. Some authors have suggested that the laws which defined the insured event as plan termination allowed the plan sponsor to manipulate plan termination and recoup excess plan assets for general corporate purposes.5 In the typical scenario, split their pension plans into two parts, one for active employees and one for retirees, and placed all surplus assets into the retiree plan. The plan was then terminated and the entire surplus recaptured. Pension assets in excess of what it cost to pay employees for their accumulated benefits could be recaptured by the company for general corporate use. Approximately one-fourth of the losses incurred by the PBGC have occurred from plan terminations of ongoing employers who are not in bankruptcy proceedings or being liquidated. Over half of the recent losses absorbed by the PBGC have been in situations where the company was in reorganization proceedings, and the PBGC has only recovered 6 percent of the losses in these cases.6 It is the purpose of this study to examine the economic nature of unfunded pension obligations (UPO) of defined benefit pension plans. Specifically, the relationship of unfunded vested benefits and unfunded past or prior service cost to systematic risk is examined both theoretically and empirically (using beta as a surrogate for systematic risk). The study is based on a theoretical model that shows that if these pension costs are considered by the market as liabilities, these costs should be positively related to the systematic risk of the sponsoring firm's stock. That is, which have larger amounts of UPO should have higher betas. The information content of these pension costs measures is examined empirically in the post-ERISA and pre-ERISA periods. The authors suggest modifications to the model to explain the empirical findings of the post-ERISA period and the implications of these findings are discussed. The empirical results across both time periods is summarized and modifications to ERISA, along with their expected results, are presented. The final section summarizes the paper and presents conclusions.