Do superannuation funds manage disbursements and risk efficiently in generating returns? New evidence

ABSTRACT

Superannuation funds (SFs) offer pension programmes in Australia. Generally, employers must pay money in to an SF account in the name of their employees and employees have the freedom to select an SF of their choice. In this paper, we determine how efficiently SFs manage disbursements (costs, fees and expenses) and risk in generating returns to investors from 2017 to 2019. We introduce two measures (disbursement utilization and risk utilization) under the data envelopment analysis (DEA) methodological framework to compare disbursements and risk management performance in relation to overall management performance. The average disbursements utilization is 0.26 and risk utilization is 0.94. On average, more than 80% reduction in disbursements is required to gain disbursement efficiency whereas the average risk reduction required to gain risk-efficiency is less than 20%. No individual SF is disbursement efficient in all five sampled years. Generally, SFs manage disbursements poorly compared to risk in return generation. Variation in fees charged by managed funds is an industry-wide concern. Therefore, how funds manage disbursements is an important consideration for investors. The evidence highlights SF managers the need to pay attention to disbursements management in their pursuit of excellence in overall fund management performance.

KEYWORDS:

• Data envelopment analysis
• superannuation funds
• disbursement efficiency
• risk efficiency
• overall efficiency

JEL codes:

• C44
• c67
• g23

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

¹ https://www.statista.com/statistics/255518/mutual-fund-assets-held-by-investment-companies-in-the-united-states/.

² Australia has a government-sponsored means-tested age pension scheme. However, superannuation funds form the backbone of the Australian pension system.

³ Common fees charged by SFs are administration fees, investment management fees, performance fees and others such as advice fee and switching fee. In a study of 52 corporate funds, Nguyen, Tan, and Cam (2012) report that fees vary according to their governance practices and fund performance and fund governance are not related. Tan and Cam (2015) report a similar finding in not-for-profit pension funds. This is not specific to Australian SFs. Fees vary from fund to fund and from country to country as well (Khorana, Servaes, and Tufane 2008).

⁴ DEA results depends on the choice of input and output measures, model specification and optimization metric.

⁵ https://www.abs.gov.au/ausstats/[email protected]/7d12b0f6763c78caca257061001cc588/09130f63d6f4e0b6ca2572ea00154b73.

⁶ https://www.apra.gov.au/news-and-publications/apra-2020-2024-corporate-plan.

⁷ https://www.apra.gov.au/news-and-publications/apra-releases-superannuation-statistics-for-june-2020.

⁸ Corporate funds that are set up for the employees of a specific corporation. Earlier, workers of a specific industry patronized industry funds. Industry funds are now open to a majority of Australian workers and hence have become public-offer funds. Financial institutions such as banks and insurance companies manage retail SFs. Retail funds distribute profits to shareholders whereas industry funds return profit to members. Range of investment options offered in retail funds is generally higher than that in industry funds. Public sector super funds are generally only available to employees in the public sector.

⁹ Choi and Murthi (2001) point out that MFs generally operate at different scales (increasing, constant, or decreasing), which may affect portfolio performance regardless of managerial skill. Therefore, it is important to control for scale effect in evaluating performance.

¹⁰ https://www.apra.gov.au/annual-fund-level-superannuation-statistics.

¹¹ Retail funds include eligible rollover funds (ERFs) that manage small and inactive SF balances. ERFs offer few investment options.

¹² As robustness check of the results to presence of outliers, we repeated the analysis by removing all SFs whose observations lie outside three standard deviations from the mean. We repeated the analysis only with funds common to all five years as well. In both cases, our overall conclusions with reference to disbursement and risk management performance remain unchanged.

¹³ According to our modelling framework, these three funds will be overall efficient as well. See Table 6.

¹⁴ For example, in 2019, the number of investment options offered by SFs vary between 1 and 3928 with median at 17.

¹⁵ An efficiency score obtained assuming constant returns to scale (CRS) is ‘pure’ technical efficiency. CRS efficiency reflects managerial performance. A DMU operates under CRS if increase/decrease of its inputs by a certain proportion results in increase/decrease in its outputs by the same proportion. For example, if the inputs double then outputs also must double under CRS technology. If production technology exhibits constant returns to scale, $\sum _{j=1}^J{\lambda }_j=1$, $\sum _{j=1}^J{\mu }_j=1$ and $\sum _{j=1}^J{\xi }_j=1$ must be removed from models (1)-(3).