Presentation on theme: "A Discussion on Directorate Composition and CPA Firm Dismissal — Empirical Evidence From China ’ s Securities Market ( 李弢 薛祖云） Discussant: Tianyu Zhang."— Presentation transcript:
A Discussion on Directorate Composition and CPA Firm Dismissal — Empirical Evidence From China ’ s Securities Market ( 李弢 薛祖云） Discussant: Tianyu Zhang Chinese University of Hong Kong
What we can learn from this paper? The relationship between the composition of board and CPA dismissal Compare a sample of 142 public firms, which switch CPA firm, and a controlling sample constructed according to two- digit SIC and total assets.
Findings When chairman comes from largest shareholder, the more internal director sitting on board, the more likely that CPA firm is dismissed Chairman turnover increases the probability of CPA dismissal Existence of Auditing Committee decrease the probability CPA dismissal
Comments The paper appears to be a replication of U.S. study (or studies) Contribution: prove a validness of U.S. phenomena in China emerging market My comments Institutional effect Hypothesis Interpretation/definition of variables Future
Comments-- institutions Development of CPA firms is closely-related to government behaviors even after the de-linkage in 1998. Lack rule of law in China market Role of Auditor in China market? Monitoring role? Collusive role? Can its quality be adopted as a credible communication mechanism?/ do minority shareholders care about auditing quality? Is there any alternative channel to communicate with market? How to build reputation for auditing quality?
Comments-- institutions Decision rights allocation between management and shareholder Who will play the dominant role in deciding hiring (or dismissing) CPA? Role of largest shareholder? (identity, incentive) Will government interference affect CPA market? What’s the role of board in the corporate governance? (Chen, Fan and Wong, 2005)
Comments-- Hypothesis Theoretical term: Key person control? When key person the turns to be chairman then the author has H1 and H2. Lack a strong theoretical link: what’s the benefit and cost for key person to dismiss shareholder? For H1, why firm turns out to be more likely to dismisses CPA within year 2002 and 2003? Any changes in the board composition? If no significant change, what’s wrong with the CPA they originally chose ? Any other driven force for dismissal of CPA? H2 is not clear. Should understand Chairman turnover first? Because key person does not set auditing committee, then they have H3. Why?
Comments-- Hypothesis Serious endogeneity problem: Board structure and choice of CPA may be simultaneously determined: controlling shareholder may dismiss a good auditor and put more internal director on board for asset stripping Chairman turnover and dismissal of CPA may be both because of poor performance of firms. Auditing committee and choice of CPA
What’s the economic meaning of auditor dismissal? Opinion shopping? Signaling? Good Bad / Bad Good Dismissed or quit? Can market see through it? Comments-- interpretation
Independent variables Inside: interaction between ratio of internal directors in board and a dummy variable indicating whether chairman comes from largest shareholder. The author interpret this result as: when chairman comes from largest shareholder, more internal director, more likely the dismissal of CPA. Why not the other way: Given ratio of internal director, the chairman from largest shareholder will be more likely to dismiss CPA?
Comments — technical points Inside is not significant in univariate test but significant in multi variables test— co-linearity? Sample construction Dismissal of CPA may be exaggerated Actual dismissal: 10% Research Sample: 50%
Comments — minor points Some terminology I don’t understand … 外部审计的独立性则难以保证， 投资者 的意志 就有可能被淡化，从而使他们的利 益受损 … … 中国证券市场出现了股权高度集中性、国 有性和股权割裂相伴而生的现象 …
Comments — future improvement Integrate institutions into the analysis Fill the theoretical gap Use change in board structure, if there is, to diminish endogeneity problems. As realized by the author, expand the observation period Reconstruct the sample (the result may disappear)