Research design

To examine the relationship between the firm’s performance and corporate governance mechanisms in Taiwan’s financial holding companies, this section employs a two-stage approach. The first stage is the evaluation of the FHCs managerial performance using a two-stage series model. The second stage employs a regression model to investigate the relation between a governance mechanism and the FHC’s performance. Because the efficiency score measured by the DEA model is distributed between 0 and 1, it becomes a censored dependent variable in the regression model. Therefore, the truncated model is applied as an appropriate regression model (Simar and Wilson, 2007) to determine if governance mechanisms have a significant impact on the FHC’s managerial performance.

The truncated regression model incorporates efficiency scores measured by the two-stage series model as dependent variables and corporate governance mechanisms consisting of

conventional and alternative variables as independent variables. To verify whether the conventional governance mechanisms still have an explanatory effect to identify the agency problems and of alternative governance mechanisms and can provide a better effect for Taiwan’s financial holding companies, three different models based on Equation (4) are used to examine these effects. Model (4a) and Model (4b) consider several conventional variables related to board composition and ownership structure and are used to measure the impact of conventional mechanisms on the FHCs’ performance. In terms of Model (4c), the auxiliary governance variables are listed in the regression model to provide a comprehensive result. Mathematically, the truncated regression model with full variables is expressed as follows: variables of board composition and ownership structure. Board composition related to the variables of independent directors, independent supervisors, control-affiliated board members, and board member’s salary, while ownership structure consists of variables related to board shareholding, blockholder shareholding, controlling shareholding and board pledges. AG_c are auxiliary variables related to understanding the information of internal operation and monitoring mechanism of managerial expropriation including institutional shareholding, divergence of control to cash flow rights, direct control and firm leverage. The remaining terms are control variables including firm size and firm risk, which are measured by a firm’s total assets and firm internal operation risk, respectively. A larger

firm scale might have abundant resources and intellectual capital to adopt aggressive competitive strategies and then achieve better performance. Moreover, a financial holding company with fluctuations in its performance may arise from the results of holding too many risky assets, investments and financial exposure. Therefore, this study controls firm size and firm risk to mitigate their impact on the results of the regression model. Table 6 presents a summary of corporate variables and definitions.

Table 6

Summary of definitions for corporate governance variables

Variables Definitions

Board composition

Independent director (%) Proportion of independent directors divided by the total director number

Independent supervisor (%) Proportion of independent supervisors divided by the total supervisor number

Control-affiliated board member (%) Number of directors and supervisors affiliated with a firm’s ultimate controller divided by the total board number

Board members’ salary (%) Proportion of Board members’ salary divided by pretax income Ownership structure

Board shareholding (%) Shareholding proportion of board members

Blockholder shareholding (%) Shareholding proportion of holders whose percentage is more than 5%

or having the top 10 shareholding, excluding the shareholding of the board members

Controlling shareholding (%) Shareholding proportion of a firm’s ultimate controller, including direct and indirect shareholding

Board pledge (%) Percentage of shareholding that board members pledge for bank loans Auxiliary governance variables

Institutional shareholding (%) Shareholding proportion of institutional investors, including financial, foreign, and domestic institutional investors

Cash flow rights (%) Proportion of cash flow rights owned by a firm’s ultimate controller Divergence of control to cash flow rights (%) The ratio of control rights to cash flow rights

Direct control (%) The proportion of shareholding directly controlled by the ultimate controller divided by the critical control share level owned by the ultimate controller

Firm leverage (%) A firm’s total debt divided by total assets Control variables

Firm size Book value of total assets

Firm risk The firm’s internal operations risk, including credit risk, operation risk and market risk

This study examines the relationship between managerial performance and governance mechanisms of 14 FHCs from 2007 to 2010. Regarding FHCs’ managerial performance, the evaluation has been estimated in previous chapter and employed as the dependent variable

in the regression model. In terms of corporate governance variables, 12 explanatory variables and two control variables, which are available from the Taiwan Economic Journal (TEJ) databank, are used to examine the relationship between FHCs’ managerial performance and governance mechanisms and to identify determinant governance factors which have a significant influence on a firm’s value. Among these variables, eight variables related to board composition and ownership structure are regarded as conventional governance mechanisms, which are employed to verify whether these factors have an explanatory effect in Taiwan financial holding companies. The remaining four auxiliary variables related to the monitoring mechanism of managerial expropriation are regarded as auxiliary governance mechanisms which would be expected to provide fuller explanatory power to explore the relation between managerial performance and governance mechanisms and further information about how investors make use of governance mechanisms to monitor their invested firms. Table 7 offers the relevant statistics of governance variables.

Table 7

Data regarding managerial performance and governance mechanisms

Variables Avg. S.D. Max. Min.

Managerial performance 0.799 0.226 1.000 0.068

Independent director (%) 0.210 0.096 0.364 0.000

Independent supervisor (%) 0.023 0.089 0.500 0.000 Control-affiliated board member (%) 0.590 0.154 0.900 0.308 Board member's salary (%) 0.562 0.477 1.550 0.000

Board shareholding (%) 0.166 0.121 0.566 0.024

Blockholder shareholding (%) 0.208 0.068 0.382 0.087 Controlling shareholding (%) 0.171 0.112 0.438 0.012

Board pledge (%) 0.322 0.331 0.930 0.000

Institutional shareholding (%) 0.164 0.140 0.501 0.004 Divergence of control to cash flow rights (%) 10.253 15.306 64.930 1.840

Direct control (%) 0.851 0.220 1.000 0.029

Firm leverage (%) 0.894 0.094 0.968 0.542

Ln (Firm size) 6.362 0.839 7.577 4.501

Ln (Firm risk) 9.339 0.923 10.756 7.499