|dc.description.abstract||Based on agency theory, this thesis investigates the link between agency costs and earnings management (EM) using data from listed public companies in China from 1999 to 2014. In this thesis, I examine whether EM reduces agency costs in China, where companies suffer from agency problems. Using static and dynamic models to test the agency costs/EM nexus, I find a significant and positive relationship between agency costs and EM based on a static model that suggests opportunistic EM in China. These results are consistent with suggestions in the literature that EM can be used opportunistically by managers with regard to agency costs (e.g., Dechow & Sloan, 1991; Guidry, Leone, & Rock, 1999; Healy, 1985; Holthausen, Larcker, & Sloan, 1995). However, I find an insignificant relationship between agency costs and EM using a dynamic model that takes into account the endogeneity issue. Therefore, my results suggest that engagement in EM has no significant influence on agency costs in China. Furthermore, board size, board independence, firm size, leverage and CEO duality, which are conventionally thought to be important in explaining agency costs, do not have a significant impact. The results add support to the growing literature on the relevance of endogeneity issues in corporate governance studies, since failing to take these into account can lead to spurious results.
This thesis examines both principal-agent and principal-principal agency costs. The concept of traditional principal-agent agency conflict caused by the separation of ownership and management was conceived in a developed market context. However, in emerging economies, principal-principal agency conflicts are the major concerns in corporate governance. Emerging economies are characterized by weak legal protection, high information asymmetry, and concentrated ownership structures, which provide the economic and institutional roots for principal-principal agency conflicts.
Based on agency theory, Jiraporn, Miller, Yoon, and Kim (2008) relate agency costs to the extent of EM, and find the practice of EM, on average, reduces agency conflicts in the U.S. In emerging economies, with weak institutional environments and highly concentrated ownership structures, managers become affiliated to the dominant shareholders, and principal-principal agency costs become prevalent. As in other emerging economies, the Chinese market is also subject to characteristics such as highly concentrated ownership structures, weak protection for minority shareholders and uneven legal enforcement. Therefore, this thesis accommodates principal-agent and principal-principal agency costs in relation to EM in China.
The endogeneity issue was the major concern when conducting this study. We add new empirical evidence to support the growing literature on concerns relating to endogeneity issues in corporate governance studies. Consistent with the prior studies of Wintoki, Linck, and Netter (2012) and Schultz, Tan, and Walsh (2010), I show that failing to take endogeneity issues into account can lead to spurious results; therefore, caution must be exerted in making policy implications based on empirical results that fail to address endogeneity issues. I expect my results to be of great interest to academics involved in researching corporate governance topics that have inherited dynamic natures and endogeneity issues.
As agency cost is a fundamental problem in corporate governance worldwide, this study is expected to be of interest to regulatory and supervisory authorities, investors, and financial analysts. The findings are inconsistent with what is implied by Agency theory, suggesting that corporate governance in China is different; therefore conventional western market theory may not fully incorporate the corporate governance dilemmas prevailing in emerging economies. In agreement with researchers such as Young, Ahlstrom, and Bruton (2004), I suggest that creative solutions need to be explored by emerging economies to resolve their particular agency conflicts in their specific institutional contexts, which indicates fruitful avenues for future research.||