Is there a positive relationship between superannuation fund costs and returns?

Economic Papers (Economic Society of Australia), Sept, 2003 by Michael E. Drew, Jon D. Stanford

1 Introduction

The efficient and effective transformation of savings into retirement income rests with superannuation fund managers. (1) Apart from the collection, pooling and investment of contributor funds, superannuation funds also facilitate the flow of capital from surplus units to deficit units in modern capitalist economies. In Australia, retail superannuation funds manage around one-third of retirement savings, or AUD 155 billion (APRA 2001), the largest distribution to any single fund-type. (2)

The defining feature of the retail superannuation management industry is its flexibility. This flexibility takes two forms. First, retail funds provide choice in terms of mandates and investment styles. For instance, multi-sector funds engage in both asset allocation and selection (e.g. a growth fund, with 70% allocation to growth assets such as shares and property, with 30% in income producing assets). Alternatively, contributors may wish to invest through a specialist single-sector manager providing dedicated asset selection services (e.g. a domestic equities fund, minimum of 80% in domestic shares and a maximum of 20% in domestic fixed interest securities). Second, retail funds provide greater scope for making voluntary contributions (Productivity Commission 2001).

The Productivity Commission (2001) defines retail funds as public offer superannuation funds that members join by purchasing investment units or policies that are sold through intermediaries such as financial planners. (3) The definition is appropriate for this study, with one amendment. The retail fund market includes two sub-sets, funds for individual investors (retail funds), and, funds designed for the professional investors (wholesale funds). The distinction between the two markets can be found by the minimum investment requirement on a per fund basis and cost structure. A typical 'retail' fund has a minimum entry amount of AUD 2,000 and levies an average annual management expense ratio of around 1.95%. In addition, these funds may charge up to 5% of the contribution as an entry load, with a 3% exit load. This is in contrast to a wholesale fund with an initial investment amount of, say, AUD 250,000 with an annual management expense ratio of 0.75%. In addition, wholesale funds have no entry or exit loads. As the average superannuation balance per member in Australia is AUD 59,400 (APRA 2001), we focus our analysis in this study to the individual investor sub-set of the market. (4) In a complete superannuation choice framework, retail funds for individual investors would be the market most accessible to Australians. (5)

So how has the Australia's retail superannuation management industry performed in terms of risks taken and costs incurred? Alternatively, from the financial economics perspective, we ask has the industry generated alpha? Warwick (2000) notes that an investment manager is said to generate alpha ([alpha]) under two conditions. First, alpha is generated if investment returns exceed an appropriate benchmark, if the risk taken to achieve the return is similar to that of the benchmark. Second, alpha is generated if managers' returns are equivalent to an appropriate benchmark, if the risk taken to achieve the return is less than that of the benchmark.

While previous studies have undertaken superannuation fund performance analyses using risk-based measures (Drew and Stanford 2001a, 2002, 2003b) and others have considered the issue of superannuation costs (Bateman 2002, Rice and McEwin 2002), this study links risk and cost to examine quality aspects, or the alpha generation capabilities, of Australia's retail superannuation management industry. (6) Specifically, we are interested in the following questions: do retail funds deliver returns commensurate with the risks taken? and, finally, do retail funds that impose higher costs on contributors deliver superior returns?

2 Sample

In answering these questions, we investigate a sample of funds undertaking specialist asset selection in domestic equities. Australia, like most other OECD nations, has a home-bias in the equities asset class, with the domestic equities accounting for 44% of fund holdings compared to its international proportion of global market capitalisation at around 1.5% (APRA 2001). Given the long-term investment horizon of retirement savings, the performance of domestic equities is the central driver for wealth creation. (7)

Morningstar Research Pty Ltd, an independent measurement service in Australia, provided monthly return observations (net of management fees, excluding entry and exit loads) for every retail superannuation fund classified as 'Retail superannuation fund Australian equity--general', from January 1991 through December 1999. The sample of 148 funds is complete in the sense that it contains all of the funds with no missing data and was maintained by the same independent data collection agency throughout the period (8).

The sample contains 3 distinct cohorts exist within the retail classification: open-end, closed-end and non-surviving. The retail open-end cohort consists of superannuation funds that are structured to accept investments from individuals. These funds are pooled and invested by a fund manager in a portfolio of general Australian equities. A typical retail fund requires a minimum initial investment of AUD 2,000, with minimum monthly contributions of AUD 100. Retail open-end funds allow investors to buy and sell at a unit price based on the appraised value of total assets. Investors can leave and enter at any time and assets may be continually added to the fund. A total of 68 retail open-end funds are investigated in this study.