盈余管理计量模型检验——来自中国上市公司的证据

Testing Earnings Management Measurement Model:Empirical Evidence from Chinese Listed Companies

运用2001年-2008年沪深上市公司的财务报告数据,以修正经营利润计算的经营性应计利润为基础,检验了只考虑管理当局相机抉择动机、只考虑经济基础动机和综合考虑两种动机的十类盈余管理计量模型揭示盈余管理的能力。通过随机游走检验和犯第Ⅰ、Ⅱ类错误检验,研究发现:(1)没有一个模型能够完全揭示盈余管理;(2)控制业绩影响非常重要;(3)考虑当期经营活动现金流量符号和公司业绩的修正FLOS模型Ⅲ整体表现最好,修正FLOS模型Ⅰ、考虑业绩影响的修正Jones模型和考虑业绩影响的Jones模型也是比较好的模型。

Using the year 2001 to 2008 financial data of listed companies in Shanghai Stock Market and Shenzhen Stock Market, based on operating accruals calculated by modified operating earnings, this paper tests earnings management detecting power of ten earnings management measurement models from different perspectives of management contingent selection incentives only, economic fundamental incentives only and both the two incentives. Following random walk test and error type I and II test, we conclude that ： （ 1 ） no model perfectly detects earnings management; （2） controlling performance factor is critical ; （3） the best model is the modified FLOS III considering the cooperate performance and dummy variable reflecting the sign of net cash flows from operating activities in current period. Other accepted models include the modified FLOS I , the modified Jones model controlling performance effect, and the Jones model controlling performance effect.