Two essays on managerial discretion in pension accounting under SFAS 87.

摘要

Pension accounting has received considerable attention recently from the popular press and accounting regulators, and much of this attention has focused on whether managers use discretion in pension accounting and funding to opportunistically manage reported earnings. These concerns culminated when the SEC announced its investigation into firms' pension accounting practices in October 2004. Concurrent with these concerns are calls from accounting academics to focus earnings management studies on specific accruals in order to understand exactly how managers manipulate earnings and whether this type of manipulation affects resource allocation.;This dissertation addresses both the concerns of the popular press and accounting regulators and the call from accounting academics and investigates pension accounting as an earnings management tool. Specifically, this dissertation is composed of two studies that examine the effects of managerial discretion and pension accounting. The first study investigates whether managers use the discretion afforded them under SFAS 87 to meet earnings targets. The second study explores whether investors and analysts "see through" the discretion exercised by managers under SFAS 87.;In Study 1, I find evidence consistent with the notion that managers make aggressive discount rate assumptions that increase the likelihood that the firm's annual earnings will meet analysts' forecasts. In addition, I find that the reduction in pension expense that results when managers make larger than expected contributions to their pension funds also significantly increases the likelihood of meeting analysts' forecasts.;In Study 2, I investigate whether investors are misled by the pension-related earnings management documented in Study 1 by investigating whether there are predictable patterns in returns related to the extent of managed pension cost. The results of the tests performed in Study 2 indicate that there appears to be no detectable mispricing associated with my empirical proxies for the managed components of pension cost. These findings corroborate prior evidence regarding the market's efficiency with respect to pension accounting information. I conclude that while managers use the discretion afforded them under pension accounting to help them meet the analyst forecast, investors do not appear to be misled by this discretion.