Pension Funding and Properties of the Analyst Information Environment

退休金會計與分析師資訊環境

退休金是十分複雜的一項會計項目，其中包含許多估計的資料與數據，文獻指出當分析師對此複雜的退休金會計項目做預測時，會有較高預測誤差，因此本研究擬探討存在此複雜與不確定性特性的退休金會計對分析師蒐集私有資訊行為之影響。以美國上市公司1988年至2008年樣本分析後，本研究發現一般而言雖然公司退休金之提撥程度不會影響分析師取得私有資訊之誘因，然而，當認列在資產負債表內的退休金資訊與附註揭露在表外之退休金資訊差異大時，表示資訊缺乏一致性，分析師在此情況下對提撥不足或過度提撥程度愈大的公司做預測時，會納入更多私有資訊，而導致盈餘預測一致性的下降，更進一步造成分析師平均盈餘預測誤差顯著的下降。

Pension funding is a complex accounting issue that includes pension calculation along with a large number of actuarial assumptions. There should be greater analysts' earnings forecasts errors for firms with higher pension complexity. Therefore, this paper explores the relationship between pension accounting and the acquisition of private information by analysts. Using US firms over the 1988 to 2008 period as our sample, we find that, on average, pension funding does not affect analyst motivation to acquire more private information. However, the large discrepancies between amounts recognized on balance sheets and actual pension funding disclosed in financial statements represent a lack of consistency in information. For highly inconsistent cases, analysts incorporate more private information into their earnings forecasts for firms with severely underfunded or overfunded pensions, which results in a decline in analyst forecast consensus. In addition, since reliance on private information increases, a greater proportion of each individual forecast error is idiosyncratic. This increased reliance reduces errors in mean forecasts of upcoming earnings.