CHAPTER 8 CONCLUSIONS, IMPLICATIONS AND LIMITATIONS . 239
8.2 L IMITATIONS AND RECOMMENDATIONS FOR FUTURE RESEARCH
Despite the abovementioned contributions, this study does have some limitations, many of which may indicate fruitful avenues for future research. First, given that the variables relating to board structure change slowly over time, which has the potential to reduce the efficiency of panel data estimations, the four-year dataset used in this study may render comprehensive explanations of governance–
performance dynamics ineffective. Following Wintoki et al. (2012), the author suggests that datasets covering a longer period of time may enable future research
245
to overcome the highly persistent feature of board structure variables by using data at two-year intervals rather than annually.
Secondly, this research considers only the relationship between gender diversity on the BOD and firm performance. Given the two-tier board structure of listed companies in Vietnam, it might be useful for future research to treat gender diversity on the BOS as a factor driving firms’ profitability. Besides, other observable diversity characteristics of directors, such as age, education, and experience, should be included in future research as long as the relevant data are available. Additionally, this study measures financial performance by employing the commonly-used Tobin’s Q ratio. Although the Tobin’s Q ratio has many advantages compared with other accounting-based metrics, using other alternative accounting-based performance measures could possibly lead to different conclusions. This may highlight the sensitivity of inferences drawn from empirical studies in which the observable variables are usually not the perfect proxies for the true phenomenon of interest. For example, it will be interesting to investigate the potential impact of corporate governance structures on operating performance metrics. The metrics for operating performance may be, among others: sales revenue per employee and fixed-asset turnover ratios which look at how well a firm uses its workers and fixed assets to generate sales, respectively.
Another common operating performance metrics is the operating cycle ratio which measures a firm's ability to convert its inventory into cash.
Moreover, like most previous empirical studies on corporate governance, this study’s sample selection process relies primarily on the availability of data, including firm annual reports and corresponding financial reports. It is likely that
246
the selected firms are the more transparent ones and therefore could actually be well governed and/or better performing firms. If that is the case, this research will suffer from selection bias which hinders interpretation and generalisation.
Thirdly, although this study does find a significant linkage between board gender diversity and the performance of Vietnamese listed companies, the channels through which female directors positively affect financial performance remain unclear. It is argued that if the presence of women on the BOD matters for firm outcomes, then there should be gender-based differences in behaviour and characteristics between female and male directors (Mohan, 2014).
A recent survey conducted by Adams and Funk (2012) confirms that this is indeed the case, i.e., female and male directors differ systematically in their core values and risk attitudes. Such gender-related differences may have an effect on firm performance through several potential channels (Mohan, 2014). For instance, they may enrich the directorial behaviours and improve the directorial task performance of the boards, all of which ultimately lead to better performance (Adams & Ferreira, 2009). Female representation on boards is also likely to add value if the market reacts positively to their presence, since they are usually considered to be free from overconfidence and have a lower tolerance for risk (for a review, see Mohan, 2014). The aforementioned findings imply that the issue here is probably not the gender difference of directors but the gender-based differences in leadership behaviour and style, and personality characteristics (Mohan, 2014; Mohan & Chen, 2004).
In agreement with Mohan (2014), therefore, this study suggests that understanding the personality traits of female directors, such as consensus-building ability,
247
management style, or attitude to risk, is essential to shed light on the potential channels through which director gender matters for firm performance. As suggested by Adams et al. (2011); and Mohan (2014), using an event study approach to investigate market reaction to the appointment of female directors, or a survey approach to grasp how gender-related behavioural differences affect board effectiveness and firm outcomes, may be possible directions for future research.
Additionally, in a recent review article, Mohan (2014) points out that there is empirical evidence that greater gender diversity on boards may promote better opportunities for women to be appointed to top management positions. Female representation on top management teams in turn provides a feminine managerial style (Mohan, 2014) and managerial expertise, and helps to improve managerial task performance (Dezsử & Ross, 2012). If this evidence is considered, female directors may also add value through their contribution in choosing the CEO and motivating women in senior management positions. For this reason, this study suggests that investigating the role of women at top management levels in interaction with the role of women on boards may offer potential for understanding another channel through which female directors add value.
Fourthly, due to the lack of data regarding corporate governance practices, this study, like most prior studies, only concentrates on observable corporate governance structures presented in annual reports, such as CEO duality, board size, or the presence of independent directors. As Mak and Kusnadi (2005) have noted, while establishing a corporate governance structure that meets the corporate governance code is important, it does not necessarily guarantee that the
248
corporate governance structure will operate effectively. For example, a board structured on the basis of international best practices probably does not successfully function if it does not meet regularly, or if the so-called independent directors are not independent in practice (Mak & Kusnadi, 2005).
Indeed, a study undertaken by Chuanrommanee and Swierczek (2007) posits that corporate governance as reported in Singaporean companies’ documents is actually unlikely to reflect their real governance practices. It is noteworthy that the research of Chuanrommanee and Swierczek was conducted in a context in which corporate governance practices are considered to be consistent with international best practice (Chuanrommanee & Swierczek, 2007) and are recognised to be the best corporate governance practices in the Asian region (CLSA, 2010, 2012).
Therefore, it is likely that observable corporate governance variables used in this thesis are not good proxies for real corporate governance practices even if in a near perfect research setting like the Singaporean market. If that is the case, it ultimately makes the estimates problematic. As a consequence, it will be seriously misleading to suggest that corporate governance structures have significant effects on firm performance. For this reason, more comprehensive data in terms of corporate governance practices will foster future research in deeply investigating the substance of corporate governance but not the form.
Finally, this study considers ownership concentration as an effective corporate governance strategy employed by shareholders to influence managerial behaviour, mitigate agency problems, and enhance performance. It should also be noted that the ownership identity may have a potential influence on the aims of the owners and the way they implement their power (Thomsen & Pedersen, 2000). As a
249
consequence, different types of ownership concentration may have different motivations and capability that, in turn, have different impacts on firm performance (Holderness, 2009). However, the lack of data for the Vietnamese market has meant that the role of ownership identity noted by Thomsen and Pedersen (2000) could not be investigated in this thesis. For this reason, in line with Judge (2012); and Munisi et al. (2014), it is desirable for further research to seek to understand how various ownership types (such as managerial ownership, foreign ownership, government ownership, and family ownership) are related to the performance of firms in Singapore, Vietnam and other markets in the Asian region.
As mentioned in Section 3.3 of Chapter 3, the high concentration of ownership by government is one of the key characteristics of corporate governance systems in Singapore (Kimber et al., 2005). A common type of fully or partly state-owned firm in Singapore is the GLCs and according to Ang and Ding (2006), these account for approximately 24% of the stock market’s total capitalisation and control over 10% of the economic output of the country. Unlike state-owned firms in many other countries, the “GLCs are run on a commercial and competitive basis” (Witt, 2012, p. 9). In addition, the GLCs in Singapore have higher valuations and have better corporate governance than a control group of non- GLCs (Ang & Ding, 2006). Keeping the dynamic endogeneity in mind, future research taking the role of government or government-related ownership into consideration should prove fruitful.
250