Author: Tan, Monica Guo Sze
Date published: July 1, 2011
Australian superannuation funds are controlled by trustees. The recent Super System Review commissioned by the Commonwealth Government pointed out that almost all issues in the superannuation system can be linked to trustee governance.1 Superannuation fund trustees have the duty of safeguarding members' interests and mitigating investment managers' and financial intermediaries' self-fulfilling behaviours. Trustees' expertise and knowledge are very important 2 since it is impossible to require every superannuation fund member to have adequate financial literacy (Cooper 2010).
The current management structure of Australian superannuation funds involves multiple agents each holding various sets of information. Trustees and fund managers are the prominent agents in superannuation funds. Trustees provide oversight of members' contributions and monitor the performance of internal and external agents. It is the duty of trustees to minimise the gap between the superannuation fund objectives and other agents' objectives through clear mandates and communication policies, effective monitoring and reviews. We argue that trustees, not management, are the main holders of information about superannuation funds' operations (except in terms of investment strategies and processes).
The governance of Australian superannuation funds rests upon the 'prudent man rule' according to which trustees are expected to exercise skill, duty of care, and diligence.3 This rule is less stringent than the 'prudent business person' rule applied to corporate directors by general law (Donald 2009) and allows more flexibility in the requirements and expectations of trustees' skills and expertise. This does not necessarily imply that trustees have fewer skills and less expertise compared with corporate directors. However, the relevant policy requirements set by superannuation funds are currently deficient (Sy et al. 2008). In terms of disclosure, there is no clear requirement for the scope of information that is to be included in a superannuation fund's annual report, nor is there any specific standard for these annual reports. Our survey of all of the 2009 industry superannuation funds annual reports revealed inconsistent disclosure practices with regard to the breadth and depth of information being published. A key indicator of a superannuation fund's corporate governance is the transparency with which it is managed and the accountability of its agents.
While corporate governance literature provides an important framework for examining trustee governance, there are stark differences between corporate boards and trustee boards. Corporate directors are elected by shareholders to safeguard their interests and maximise the value of their share capital. Superannuation fund boards of trustees may be comprised of natural person4 trustees or companies that provide trustee services.5 The agency relationship is more complex with companies acting as trustees (Donald 2009). In this paper, we refer to both natural person trustees and directors of trustee companies as trustees.
Corporate directors are driven by financial gains while trustees have a different set of complex goals and objectives. The type of funds trustees are associated with influences their objectives and decision-making framework. For example, industry superannuation funds are not-for-profit organisations, while retail funds focus on generating capital appreciation for the shareholders of the parent financial institutions and returns for their fund members. Hence, the success of a corporation is mainly measured in terms of financial gains and capital appreciation, while superannuation fund trustees may prioritise legal compliance alongside members' interests (Gupta et al. 2007). The compensation package of corporate directors often includes shares and options that are designed to align their interests with those of the shareholders. Consistent with APRA's study (Sy et al. 2008), our data shows that some trustees do not receive any fee for their services. Individuals' motivation to undertake trustee positions may be very different from those of the corporate directors including, for example, a desire to develop their reputation and gain experience within the superannuation funds industry; it is within this context that we develop our hypotheses below.
Literature and hypothesis development
Disclosure can signal the values and development of a firm, product, or worker (Stiglitz 1979), and it is especially important in mitigating shirking behaviours in markets filled with dispersed ownership structures (Jensen and Meckling 1976). Management influences accounting information, information type and disclosure frequency based on their self-interest and other constraints (Watts and Zimmerman 1978; Leftwich et al. 1981). This underpins the hypothesis developed below.
Governance practices, including disclosure, are influenced by the regulatory environment in which a firm operates: the more stringent the regulation, the better the practices observed (Gillan 2006). Where there is no clear legal requirement and standard for disclosure, we argue that voluntary disclosure is dependent on the trustee structure and trustees' background:
The model shows that the degree of disclosure (DIS) for each superannuation fund depends on a function of nominator type (NOM), the educational background of trustees (EDU),6 board independence (IND), outside directorship (OUT), board size (BS), trustee compensation measured by trustees' fees (DC), and fund size in terms of value of funds under management and membership base (Size). The following two main hypotheses are postulated:
Hypothesis 1: Voluntary disclosure is dependent on the trustee structure.
Hypothesis 2: Trustees' backgrounds have a relationship with voluntary disclosure.
Trustee Nomination (NOM) is an exogenous variable. Given that the nature of the relationship between the nominator and trustees could vary, the level of information shared between them may also differ. We argue that the types of association between nominator and trustees affect the level of insider information access, selection of skills and expertise and, hence, influence disclosure practices.
Trustees' Educational Background (EDU): In corporate governance, the role of a board of directors includes providing assurance to shareholders and potential investors due to the expectation that the directors will provide valuable insights as a result of their financial expertise, and facilitate effective monitoring (Jeanjean and Stolowy 2009). Trustees who have higher financial proficiency tend to handle complex issues and evolving trends better (Gupta et al. 2007). We argue that trustees with higher levels of education (or who choose to disclose their educational background) are more aware of governance practices and development; they are also more likely to be driven to develop positive personal reputation in the market for future gains.
Independent Trustees (IND) and Outside Directorships (OUT): Outside (independent) directors are negatively related to disclosure due to the substitution effect between the two (Eng and Mak 2003). In addition, the degree of information asymmetry between outside directors and firm insiders, such as the CEO, affects how the former could contribute to firm value; the high cost of information will reduce the value of outside directors to a firm (Duchin et al. 2010). Independent nonexecutive directors are also associated with the provision of comprehensive information in mandatory financial disclosures (Chen and Jaggi 2000). The degree of board independence has a positive correlation with the board's financial expertise, which suggests a complementary relationship between the two (Jeanjean and Stolowy 2009). Trustees who hold multiple outside directorships may benefit from a diversity of knowledge, information and resources, and are likely to be associated with better disclosure of information about trustee structure and composition since their networking essentially capitalises on their reputations.
Board Size (BS): Board size may be affected by fund size, fund growth rate and the merger activities of funds. We argue that as trustee board size grows, there will be an increased diversification of trustees' expertise, a reduction in monitoring costs by trustees, and an improved balance of representation of members' interests. This leads to improved disclosure policies and practices. In addition, larger boards may tap into better resources and information for the selection of asset consultants and fund managers.
Control Variables: Trustee compensation and fund size
Trustee compensation indicates if a trustee is providing services on the basis of competitive incentives. Based on the signalling hypothesis, this is an important control for identifying whether trustees provide disclosure in order to increase their future earnings and/or improve their future prospects in the labour market. The size and complexity of a firm may lead to variations in board structure due to the variances in monitoring costs and information complexity (Boone et al. 2007). Hence, firm size affects voluntary disclosure via its effects on governance variables (Kelton and Yang). Therefore, we argue that fund size may have an impact on disclosure.
Members rely on product disclosure statements (PDS) and annual reports for information on their superannuation funds' structure, operation and performance. We followed this path and collected governance variables by downloading all annual reports, PDS and trustee information published by industry superannuation funds on their websites. We then conducted searches using the internet search engine Google and two databases - Business Who's Who in Australia and Company 360 - to collect information about trustees' educational background and outside directorships. We have only collected data for the financial year 2009 because the information on the governance of superannuation funds is limited and past years' records are usually not maintained on websites. Since nomination of trustees for each type of superannuation funds is exogenous, we do not expect any endogeneity issue between trustee governance variables and disclosure level. We obtained industry superannuation funds financial data (trustee fee and fund size) from APRA for the period 2004-09. Due to the limited availability of data and our focus on trustees, we define fund voluntary disclosure through the level of disclosure about trustees' profiles. Specifically, voluntary disclosure is a product of:
= the percentage of trustees who disclose the field of their formal education
x the percentage of trustees who disclose the level of their completed education
x the percentage of trustees who disclose their outside directorship(s)
We measure trustee nomination by the type of nominator as reported by the superannuation funds. The types of nominator include independent party, member, employer, industry association bodies and professional trustee company. Trustees' educational backgrounds are measured by the product of (1) educational completion level (i.e. diploma or school-leavers, undergraduate and postgraduate levels); (2) specialised fields (i.e. business and economics, arts and law, science, and others); and (3) whether a trustee has financial, governance or superannuation industry management expertise. For the third measure, we use trustees' professional memberships (CPA, CA, FAICD, CFA, CFP, ASFA, Finsia and AIST)7 as a proxy of their expertise and to determine access to resources and information relevant to the superannuation funds industry.
Board size measures the total number of trustees on each board.8 We derived two proxies for board independence: (1) the percentage of independent trustees per board; and (2) a dummy variable that indicates whether there are any independent trustees on each board. Outside directorship is measured by the number of directorships external to the sample superannuation funds that a trustee holds. Trustee compensation is measured in terms of total trustee fee. Fund size is measured in terms of year-end fund value and membership base.
We surveyed 529 trustees from all 66 industry superannuation funds. Our results are comparable to those of Gupta et al. (2007)9 and APRA's survey10 (Sy et al. 2008). With a total of 1,319 trustees in the whole industry, according to APRA's 2008 survey (Sy et al. 2008), our sample consists of approximately 40 per cent of the trustee population. Table 1 shows the descriptive statistics: one set for trustee level, and another set for fund level. Descriptive statistics at the trustee level in Table 1 provide an outlook of Australian industry superannuation fund trustees' background, as well as their funds disclosure practices and committee structures.
An average board has nine trustees, which is consistent with past surveys (Gupta et al. 2007; Sy et al. 2008). The number of outside directorships held by trustees in the industry superannuation funds (2.96 per trustee in Table 2) is slightly lower than the overall industry average of 3.5 (Sy et al. 2008). In terms of board committee, only 14.6 per cent within our sample had disclosed audit, risk and compliance committees, 7 per cent had disclosed finance committees, and 80 per cent had disclosed investment committees. These results were much lower than those in the APRA survey, which found that 76 per cent of all funds had both independent audit and regular self-assessments for regulatory compliance review (Sy et al. 2008); this indicates a poor level of current disclosure practices concerning governance information. An average trustee received $1083.77 in fees in 2009, and $632.17 (on average) in the 2004-09 period. This is significantly lower than the average remuneration of $38,000 in the APRA survey; however, the survey also found that 54 per cent of trustees did not receive any remuneration for their service, or did not declare their fees (Sy et al. 2008).
Overall, voluntary disclosure of trustee board committees is low. Approximately 5.1 per cent of trustees do not disclose their nominators, while over 70 per cent of them do not disclose their educational and professional background. On the other hand, only about 5 per cent of trustees are sitting on the board of funds that do not disclose whether they have independent trustees.
Approximately 44 per cent of trustees in the sample are nominated by industry/professional associations/ unions, while 40 per cent are nominated by members (Table 2). Only 8.51 per cent are nominated by trustees' employers. This compares with 32 per cent of trustees within the industry who are employer-representatives and 20 per cent who are member-representatives (Sy et al. 2008). Nomination is highly transparent as only 5.1 per cent of trustee nomination is not disclosed. Based on disclosed information, almost 15 per cent of trustees have attained undergraduate qualifications; 10.21 per cent have a postgraduate education, and 4.54 per cent are diploma holders or school-leavers. This is significantly lower than the results in APRA's survey (Sy et al. 2008) which indicated that 65 per cent of trustees reported having university degrees, while only 11 per cent reported not holding any formal qualifications.
Trustees in Science and other fields, including Education, Carpentry, Trade Services and Engineering are the group that has the highest levels of disclosure (13.42 per cent), followed by trustees in Arts & Law (7.56 per cent) and Business & Economics (5.48 per cent). In addition, 71.6 per cent of trustees did not disclose whether they have financial expertise, 9.83 per cent of trustees disclosed that they have financial expertise, and 18.5 per cent disclosed that they do not have financial expertise. In comparison, Gupta et al. (2007) found that 40 per cent of their surveyed trustees have a background in finance, but only 14.5 per cent of the 131 respondents have formal investment qualifications. The differences between our findings and those of Gupta et al. (2007) and Sy et al. (2008) provide a strong indication of continued poor disclosure practices.
Approximately 60 per cent of the sample trustees were not associated with complete audit, risk and compliance (ARC) committees on their board. Only 8.13 per cent have disclosed complete ARC, while 32 per cent are associated with non-disclosed ARC committee structures. Only 4.73 per cent of trustees were associated with a financial committee; in contrast, 54.25 per cent were associated with an investment committee on their board.
In this section, we discuss what drives trustees to disclose information about themselves and how their educational and professional backgrounds relate to trustee board committees. We use Model (1) to test our research question on voluntary disclosure with a different proxy of trustee's background and fund size to check for robustness. Table 3 shows the OLS results for EDU and completion level models only.
H1 is weakly supported. We found member-nominated trustees have a negative effect on the disclosure of trustees' information. However, the effect of nomination by association is not significant. This result is not robust in other models. The connection between individuals and their nominators based on implicit long-term network and evaluation may be a more significant factor in the nomination process than selection from an open market based on competitive professional profiles. Board size is negatively related to voluntary disclosure in the completion-level model.
H2 is supported by our results where a higher level of educational background has a positive relationship with voluntary disclosure when EDU index and formal qualification are used. Buseco and Artlaw both contribute positive effects to disclosure (results not reported in table). However, completion level models are not significant. This is in line with the signalling argument and consistent with the results for board independence. In contrast, IND per cent11 is found to be significant in all models, implying that the balance of control between independent trustees and other trustees on a board relies on the number of independent trustees relative to the size of the board. The coefficients for IND percentage in all models are the strongest with a positive impact in all models. This shows that the higher the number of independent trustees, the stronger their voices and influence on other trustees and board matters. The number of outside directorships (OUT) also has a significant but lesser positive impact on voluntary disclosure. The results support our hypothesis that board independence is positively related to voluntary disclosure, based on the notion that part of trustees' future earnings growth is dependent on their current development and promotion of their reputational properties. In addition, independent trustees and trustees who have multiple outside directorships may have more resources with which to obtain up-to-date governance knowledge and practices, and wider experiences in undertaking board duties.
Trustee fees (DC2009 and DC Average) do not appear to have any impact on voluntary disclosure. The relative low compensation level for trustees may have little effect on driving trustee and board performance as a whole. We found that the larger the fund (in terms of fund value and number of members), the higher the level of voluntary disclosure. This supports the argument that larger funds are more capable and resourced to cope with disclosure costs. It could also be argued that larger funds attract a higher level of analyst coverage, which demands a higher level of transparency from the funds.
We surveyed Australian industry superannuation funds to investigate their trustee board structure and composition, trustees' backgrounds and current voluntary disclosure practices. Our findings shed light on understandings of trustee governance and show that current disclosure practices are poor.
We found that more highly qualified trustees are related to better levels of voluntary disclosure of their background profile. In addition, independent trustees and trustees who have a higher number of outside directorships have a strong positive effect on levels of voluntary disclosure. This is due to their higher need for the promotion and development of their reputational properties with anticipation for future personal earnings growth. Our empirical results show strong inter-relationships between these characteristics and confirm the significance of the make-up of trustees' boards for standards of funds governance. These unique results provide important insights for policy makers, regulators and practitioners in the regulation of the industry, and the enforcement of trustee accountability. Another important next step will be to test the impact of trustee governance on fund returns and risk.
Despite the limitations in the availability of information and the impact this had on the research methodologies we could employ, this study provides an important, innovative view of trustee governance that establishes new directions for the field. Further research in this area will enhance the understanding of trustee governance and its impact on fund returns and risk management
We would like to thank Scott Donald for his helpful comments and suggestions. We would also like to thank the Australian Centre for Financial Studies for providing financial support.
1. The Review recommends a Code of Trustee Governance, suggesting that effective trustee practices and structures will lead to enhanced performance in the industry (Cooper 2010).
2. See Myners 2000, Kakabadse et al. 2003, and Clark et al. 2006 for trustee competency, background, roles, and effects.
3. Part 6, Superannuation Industry (Supervision), Act 1993 (SIS Act 1993).
4. There are currently only five large funds with natural person trustees (Cooper 2010), hence we do not differentiate the types of trustees in our examination.
5. See Section 2 in Chapter 2 of Part 2 in (Cooper 2010)
6. We also use two other proxies as the EDU index greatly reduces our sample size: (2) completion level (undergraduate and postgraduate dummy variables); and (3) formal qualification (buseco and artlaw dummy variables).
7. PS146 as a proxy was dropped due to significant low disclosure.
8. The CEO-Chairman duality variable is dropped because no incidence is present in the sample, reflecting that the two positions are held by separate individuals in all 66 industry superannuation funds.
9. This study comprised of 48 of the top 200 funds in 2004-05, where half of the sample were corporate funds.
10. This survey included 187 superannuation funds representing the whole population in 2006.
11. The IND dummy is found to be a non-significant factor; therefore, results were not included in the table.
Boone, A. L., Field, L. C., Karpoff, J. M. and Raheja, C. G. 2007, 'The determinants of corporate board size and composition: an empirical analysis', Journal of Financial Economics, vol. 85, pp. 66-101.
Chen, C. J. P. and Jaggi, B. 2000, 'Association between independent non-executive directors, family control and financial disclosures in Hong Kong', Journal of Accounting and Public Policy, vol. 19, pp. 285-310.
Clark, G. L., Caerlewy-Smith, E. and Marshall, J. C. 2006, 'Pension fund trustee competence: decision making in problems relevant to investment practice', Journal of Pension, Economics and Finance, vol. 5, pp. 91-110.
Cooper, J. 2010, Super System Review Final Report, Commonwealth of Australia, Barton.
Donald, M. S. 2009, 'What contribution does trust law make to the regulatory scheme shaping superannuation in Australia?', Australian Prudential Regulation Authority.
Duchin, R., Matsusaka, J. G. and Ozbas, O. 2010, 'When are outside directors effective?', Journal of Financial Economics, vol. 96, pp. 195-214.
Eng, L. L. and Mak, Y. T. 2003, 'Corporate governance and voluntary disclosure', Journal of Accounting and Public Policy, vol. 22, pp. 325-345.
Gillan, S. L. 2006, 'Recent developments in corporate governance: an overview', Journal of Corporate Finance, vol. 12, pp. 381-402.
Gupta, V., Jin, H., Orszag, M. and Piggot, J. 2007, How do Australian superannuation fund trustees perceive their role and effectiveness?, Australian Institute for Population Ageing Research, University of New South Wales.
Jeanjean, T. and Stolowy, H. 2009, 'Determinants of board members' financial expertise - empirical evidence from France', The International Journal of Accounting, doi:10.1016/j.intacc.2009.09.002.
Jensen, M. C. and Meckling, W. H. 1976, 'Theory of the firm: managerial behavior, agency costs and ownership structure', Journal of Financial Economics, vol. 3, pp. 305-360.
Kakabadse, N., Kakabadse, A. and Kouzmin, A. 2003, 'Pension fund trustees: role and contribution', European Management Journal, vol. 21, pp. 376-386.
Kelton, A. S. and Yang, Y-w. 2008, 'The impact of corporate governance on Internet financial reporting', Journal of Accounting and Public Policy, vol. 27, pp. 62-87.
Leftwich, R. W., Watts, R. L. and Zimmerman, J. L. 1981, 'Voluntary corporate disclosure: the case of interim reporting', Journal of Accounting Research, vol. 19, pp. 50-77.
Myners, P. 2000, 'The Myners Review of Institutional Investment for HM Treasury', HM Treasury, United Kingdom.
Stiglitz, J. E. 1979, 'Equilibrium in product markets with imperfect information', The American Economic Review, vol. 69, pp. 339-345.
Sy, W., Inman, C., Esho, N. and Sane, R. 2008, 'Superannuation fund governance: trustee policies and practices', Australian Prudential Regulation Authority, Melbourne.
Watts, R. L. and Zimmerman, J. L. 1978, 'Towards a positive theory of the determination of accounting standards', The Accounting Review, vol. 53, pp. 112-134.
MONICA GUO SZE TAN is a Lecturer in the School of Economics, Finance and Marketing, RMIT University. E-mail: email@example.com.
MARIE-ANNE CAM is a Lecturer in the School of Economics, Finance and Marketing, RMIT University. Email: marie.cam@ rmit.edu.au