Governance of pension funds: interlocking and compensation of Australian superannuation fund boards

This thesis documents the incidence and determinants of board interlocking (where directors simultaneously sit on multiple boards) in pension funds and examines its effect on fund performance. It also investigates the determinants of pension fund director compensation. The motivation to examine these issues stems from the distinctive interlocking and compensation practices in pension funds. Data on a sample of 249 Australian pension funds from 2004 to 2011 is collected from fund documents and government publications. A quasi-natural experiment is employed to address endogeneity in the interlocking-performance relation. Quantile regressions are used to deal with heterogeneity in the determinants of pension fund director compensation. The findings indicate that board interlocking is more likely in pension funds which require similar resources/information, are smaller (in total assets), have smaller boards and have poor past returns. The quasi-natural experiment reveals that an exogenous reduction in interlocking results in an increase in fund returns and a decrease in fund expenses. Moreover, there is a differential impact of the level and type of pension fund board interlocking on fund performance. Overall, these results support the busyness proposition that interlocked directors may be ineffective monitors and associated with poor fund performance. Quantile regression results reveal that board monitoring quality (in funds that pay less director compensation), fund complexity (except in funds that pay high director compensation) and fund performance (particularly when performance is measured with fund expenses) are determinants of pension fund director compensation. Overall, these findings suggest that pension fund director compensation is structured in line with resource dependence theory and agency theory expectations. This research makes several contributions to the board interlocking, compensation and pension fund governance literature. For the board interlocking literature, this is the first examination of non-corporate board interlocking whereas prior studies have focused on corporate boards (Ferris and Jagannathan, 2001; Kiel and Nicholson, 2006; Ong, Wan and Ong, 2003). This study thus provides insights into the practice and theories of interlocking which the extant literature has not considered. For the compensation literature, this is the first examination of the determinants of pension fund director compensation whereas prior studies have focused on corporate director compensation (Andreas, Rapp and Wolff, 2012; Bryan, Hwang, Klein and Lilien, 2000; Cordeiro, Veliyath and Eramus, 2000; Fedaseyeu, Linck and Wagner, 2014). It thus provides a broader understanding of the factors which influence director compensation. For the pension fund governance literature, (Benson, Hutchinson and Sriram, 2011; Clark, 2004; Hess and Impavido, 2003; Kowalewski, 2012; Tan and Cam, 2011) it reveals that board interlocking is an important factor in the pension fund governance-performance relation. The findings have practical implications for pension funds in assessing performance, appointing directors and in developing director compensation policies; for policy makers in consulting with industry and government bodies to monitor these practices; for pension fund regulators by suggesting that future regulation may need to recognise the intricacies of pension fund board interlocking and that director compensation policies should be standardised; and for industry bodies in refining their guidelines on interlocking and compensation best practices.