Earnings Conservatism: 1995-2005

This thesis tries to reveal the relationship between accounting numbers and regulations arrangement by research on earnings conservatism. As an important reference, accounting earnings, which has played a vital role in the reformation of both accounting systems and capital market, become not only the means of security supervision but also the objectives of standard-setters'efforts. What the thesis works on is to find out how the special regulations background of listed firms affects earnings conservatism.Based on that, my sample consists of all firm-year observations from 1995 to 2005 in the CSMAR database. The paper uses and broaden the view of Basu(1997)model to make four inspections on the changes of earnings conservatism: (1)Does the earnings conservatism vary in different years? (2) Do the negative earnings'changes have greater reversion than the positive one? (3) Does the accounting conservatism be different before 1999/2001 from after 1999/2001? The two years mentioned are important for both security supervision and accounting standards;(4) Does the role of earnings'being the supervision means have impact on conservatism?Empirical results are presented: (1) Earnings conservatism exits as a whole and varies in different years; (2) The negative earnings changes have greater reversion speed than the positive one; (3)Earnings conservatism has been strengthened since 1999 ,especially in 2001 ; (4) Earnings conservatism differentiate in firms which have distinct ROE; I also made inspections on the changes of ROE by earnings distribution histograms, and concluded that the earnings distribution was anisomerous after or before 2001 and closely linked to regulations indexes. Surprisingly, earnings conservatism does not exit before 2001 when I control the influences of supervision, while the different results appear after 2001.These conclusions make clear that, the essence of earnings conservatism is the managers'incentives to satisfy the regulators for the purpose of regulations before 2001, which is different after 2001. With the efforts of regulators and standard-setters, benign earnings conservatism appears after 2001. This thesis develops the research on earnings quality in two aspects: First, more than ten years of samples are selected for the empirical analysis; Second, the regulations background of listed firms are considered for the inspections on the changes of earnings conservatism. My empirical findings are original for the for the post-2001 samples: Earnings management and regulations satisfying cannot completely account for earnings conservatism.These findings could help the regulators or standard-setters get better understanding of the importance of accounting numbers in capital market and their mutual effects so as to make endeavors to foster the market mechanism by which high-quality accounting numbers are provided and demanded.