Title: Looking for Earnings Management in Corporate Defined Benefit Pension Plans
Abstract: Introduction Firms that provide a defined benefit pension plan to their employees can potentially use the pension plan to engage in earnings management. The accounting rules that govern pension plans allow the firm to smooth the returns by choosing an assumed rate of return to measure plan earnings. This expected rate of not actual returns, figures into the calculation of net pension expense, which in turn impacts the calculation of the firm's net earnings. There exists the possibility that managers will choose the assumed return opportunistically in order to reach certain earnings goals. The research on earnings management is extensive and spans multiple disciplines, including accounting, finance and economics. Much of the research concerns the use of accruals to manage earnings. Bergstresser, Desai and Rauh (2006) present the most recent and widely cited study of earnings management carried out through the defined benefit pension plan --specifically, the choice of the long-run expected rate of return (ERR) on pension assets. Using data from 1992-2002, the authors report that firms with the most to gain from manipulating pension plan assumptions are firms with large pension funds relative to net income. Bergstresser, Desai and Rauh find that these firms choose higher long-run expected rates of return for their pension plans, which improve reported earnings and thus infer earnings management. In the current study, I test the panel data models of Bergstresser, Desai and Rauh with more recent data from 2003-2011. Since the work of Bergstresser, Desai and Rauh, there have been changes to accounting rules that are designed to increase transparency, which ought to reduce the extent of earnings management. Using the Bergstresser, Desai and Rauh model, I find no evidence that earnings management has declined. I also find that pension fund size can explain a good portion of the variation in ERR. Because large pension funds are better able to allocate assets to investment categories, such as alternative classes that provide higher long-run expected returns, they can justify a higher ERR. A pension fund that favors alternative asset classes, such as private equity and hedge funds, provides some justification of a higher ERR and should not be construed as earnings management. Although we find, within a cross-section, that firms with larger pension funds choose higher ERR values, there is also considerable persistence in ERR over time within a firm. ERR is a sluggish variable that some firms leave unchanged from one year to the next. Such persistence in ERR suggests the influence of time invariant factors that can be captured in the model with firm fixed effects. These fixed effects control for unobserved firm heterogeneity. I find that controlling for this heterogeneity across the firms removes all the statistical significance of earnings management. The fixed effects also renders insignificant the effect of fund size, leaving the model with little ability to explain the determinants of ERR and thus no ability to identify managers who opportunistically set ERR. Earnings Management Using Pension Plan Assumptions Firms must account for their defined-benefit pension fund activities in their financial accounts, and the pension fund can have a significant impact on the firm's income statement. In the course of a year, the plan incurs costs equal to the expected benefits earned by employees during the year. These costs tend to be quite smooth since they are determined by slow moving variables such as compensation increases and life expectancy. As such, the costs do not add much variability to the firm's profits. Pension returns, on the other hand, show significant variation over time given the volatility in financial markets. The Financial Accounting Standards Board (FASB) regulations, which govern the accounting for pensions, allow firms to smooth these returns by using an expected long-run rate of return, instead of the actual return on plan assets, as a measure of the income earned by the fund. …
Publication Year: 2016
Publication Date: 2016-01-01
Language: en
Type: article
Access and Citation
Cited By Count: 1
AI Researcher Chatbot
Get quick answers to your questions about the article from our AI researcher chatbot