Banks Listed On The Official Market Of The Stock Exchange Of Mauritius Accounting Essay
The whole population is composed of banks listed on the official market of the Stock Exchange of Mauritius, i.e. the Mauritius Commercial Bank Ltd and the State Bank of Mauritius. We have chosen this sample since all such companies shall comply with the Code of Corporate Governance. In case of non-compliance, these enterprises shall disclose and explain the reasons for their non-compliance.
3.2 Data collection
Data can be classified as being either primary or secondary data. Primary data represent that are “collected specifically for the purpose of the investigation at hand”. It can be collected through observations, experiment, surveys, focus group and questionnaires. On the other hand, secondary data are published data that have been presented in some kind of report or publications. For this study, we will base ourselves on secondary data.
Secondary data can be either obtained from both external and internal sources. External data may come from government sources like census of population, investment, manufacturing retail trade, Commercial research houses: student’s dissertation and other research reports etc. Whereas internal data may come from budgets, sales figures, profit and loss statement, debtors list, database of customers and other stakeholders etc. To be more precised, internal data will be collected through annual reports and quarterly financial statement of banks for the period of analysis 2008-2010.
3.3 Data analysis
Qualitative tools are based on content analysis, among other things and are presented in non-numerical format. They allow the researcher to gain a very deep insight into the topic that he is investigating; they are not suited for all types of studies. The quantitative tools, on the other hand, for data analysis generally borrowed from the physical sciences, in that they are structured in such a way so as to guarantee (as far as possible), objectivity and reliability (Creswell, 2003). In addition to that, the quantitative tools are objective and straightforward and, so, are ideal for testing the validity of certain hypotheses. Based on our objective, we will make use of quantitative tools.
3.5 Measurement instruments
The dependent variable
The firm’s risk management activity is measured by the delta percentage (Delta%) defined as the delta of the risk management portfolio held by the firm divided by its expected production. The Delta% is measured at the quarter end.
For evaluating the impact of the board and audit committee independence and financial knowledge on corporate hedging, we will go through the tables below.
Table 2 - measures of board structure
Table 3 - measures of audit committee structure
Table 4 - measures of risk monitoring committee structure
Note: We categorize directors as unrelated if they are independent of the firm’s management and free from any interest or relationship that could conceivably affect their ability to act in the best interests of the firm, other than interests arising from shareholdings.
The checklist for appraising the board, audit committee and risk monitoring committee is given in appendix 1.
We make use of various control variables for risk management. As in Graham and Rogers (2002), we use the percentage of shares held by institutions (%inst) as a proxy for information asymmetry. Moreover, Tufano (1996); Pertersen and Thiagarajan (2000) find that managerial risk aversion is an important determinant of the risk management policy in the gold mining industry. Therefore, the number of the firm’s common shares held by the CEO (CEO_CS), the value of options held by the CEO (ValCEO_op) and the CEO age (CEO_age) will be employed in this research. The two first variables capture the Smith and Stulz (1985) argument that compensation packages leading to a concave (convex) function between the managers’ expected utility and the firm’s value encourage managers to hedge more (less). CEO_age is expected to be negatively related to corporate hedging because older CEOs have a smaller fraction of their revenues and human capital tied to the firm’s value and therefore are less risk averse. However, Tufano (1996) suggest that older CEOs facing imminent retirement might prefer reducing fluctuations in the firm’s value and hence hedge more extensively. Bear in mind that the CEO age could also indicate his experience.
The firm’s financial distress costs are also be controlled since they are considered as an incentive for firms to increase their hedging ratio. Our proxy for the firm’s financial distress costs is leverage (Leverage) measured as the book value of the long-term debt divided by the firm’s market value.
3.4 Regression model
To test our hypotheses we make use of the following regression equation:
For H1- The independence and financial knowledge of the board has a positive impact on risk management:
Delta% = α0 + α1CEO_CS + α2ValCEO_OP + α3Leverage + α4Explo + α6Indbor + α7Finbor + α8%inst + α9CEO_age + ui + Ɛit
For H2- The independence and financial knowledge of the audit committee has a positive impact on risk management:
Delta% = α0 + α1CEO_CS + α2ValCEO_OP + α3Leverage + α4Explo + α6Indaud+ α7Finbaud + α8%inst + α9CEO_age + ui + Ɛit
For H3-The independence and financial knowledge of the risk monitoring committee has a positive impact on risk management:
Delta% = α0 + α1CEO_CS + α2ValCEO_OP + α3Leverage + α4Explo + α6Indrisk+ α7Finbrisk + α8%inst + α9CEO_age + ui + Ɛit
For H4- the board, audit committee and risk monitoring structure altogether have a positive impact on risk management:
Delta% = α0 + α1CEO_CS + α2ValCEO_OP + α3Leverage + α4Explo + α6govindexbor+ α7govindexaud + α8govindexrisk + α8%inst + α9CEO_age + ui + Ɛit
Delta% = risk management
CEO_CS = common shares held by the CEO
ValCEO_OP = value of options held by the CEO
Leverage = leverage
Explo = investment opportunities
%inst = percentage of shares held by institutions
CEO_age = CEO age
govindexbor , govindexaud/govindexrisk = refer to appendix 3
Indbor, Indaud/ Indrisk = proxies for board, audit committee and risk monitoring committee independence
Finbor, Finbaud /Finbrisk= proxies for board, audit committee and risk monitoring committee financial knowledge
3.5 Regression Analysis Method
For this project, all our independent variables are quarterly measured. However, the information on the board, audit committee and risk monitoring committee and on the background of their members is published on an annual base. Therefore, we assume that the firm’s corporate governance will remain constant between two consecutive general annual meetings since directors are usually elected, for at least a one year term, at the annual general meeting.
For example, if the firm’s fiscal year end for 1997 is December 31st, and the general annual meeting is held on May 28th, 1997, we suppose that the general annual meeting is held in the second quarter of fiscal year 1997. And that the next annual general meeting occurs in the second quarter of 1998. Therefore, the corporate governance data collected from the 1997 proxy statement is used for the third and fourth quarters of 1997 and the first and second quarters of 1998. Figure 2 summarizes the procedure used to construct our sample.
Audit and board characteristics are supposed to be constant over this period
1998 Annual general meeting
1997 Annual general meeting
Corporate governance variables observed on the 1997 annual general meeting will be used as independent variables for these risk management observations
Figure 2 - Sample constitution procedure
For examining the effect of the board, the audit committee and the risk monitoring characteristics on risk management practices, we use multivariate analysis. Moreover, a Tobit model is also employed to run our regressions in order to account for the censoring of our dependent variable (Delta%).
moreover, opted for the random effect specification because according to Greene (2004) the incidental parameters problem affecting the fixed effect specification does not lead to biased estimates of the slope in the case of a Tobit model, but does cause a downward bias in the estimated standard deviations. Such a problem might lead to erroneous conclusions concerning the statistical significance of the variables used in the regressions.
Need an essay? You can buy essay help from us today!