| Summary: | 碩士 === 國立中興大學 === 會計學研究所 === 100 === A recent stream of literature shows that family control is central in most countries of the world, but little research exists regarding family control and accounting conservatism. The primary objective of this study is to examine the association between family control and accounting conservatism including unconditional conservatism and conditional conservatism. Furthermore, this study investigates whether family control has impacts on accounting conservatism after controlled the effect of the board of director characteristics, measured by board independence, board sizes and duality. This study measures unconditional conservatism by using Givoly and Hayn’s (2000) accrual-based model and Beaver and Ryan’s (2000) market-value based model. In regard to conditional conservatism this study uses Basu’s (1997) asymmetric timelines of earnings model.
Using a sample of firms listed in Taiwan Stock Exchange over the 2007-2009 period this study performs the fixed-effect model and year-by-year annual regression to control unobservable firm characteristic and reduce the heterogeneity in different years and different firms. After controlling standard economic determinants of the level of accounting conservatism, e.g., institutional ownership, firm size, growth opportunities, cash flow, leverage, litigation risk and audit quality, the result shows that a negative relation between family control and unconditional conservatism, indicating family control structure is inclined to adopt less conservative accounting policies. This sudy also finds no significantly positive association between family control and asymmetric timeliness of earnings. However, the greater the proportion of independent directors of the company is, the board is more capable of overseeing the managements and diminishing window dressing the finance report. As a result, independent directors contribute to the early release of bad news, and would be able to respond bad news promptly to be conditional conservatism.
|
|---|
