One of the perennial arguments between industry super funds, public or retail funds, financial planners and advisers is which is cheaper and provides better performance.
Now the industry regulator says that retail funds are more expensive and produce lower returns, according to a report on the website of the Australian Prudential Regulation Authority this week.
"The results show that for balanced/growth style investment options, default investment options of large retail funds have earned lower net returns relative to comparable investment options from the not-for-profit sector over the study period, and these lower net returns are not due to differing account balances or differing asset allocation.
"The evidence indicates that part of the net retail underperformance is due to embedded fees that are already incorporated by the investment vehicles used by these funds at the gross return level, rather than poor investment manager skill.
"Retail fund expenses, explicit and embedded, lower the net earnings of the retail sector relative to the not-for-profit sector," APRA said in a summary of the report (The full report is here).
In the past studies by APRA, the key super regulator, and others have found persistently lower net returns for retail funds relative to corporate, industry and public sector funds.
(Evidence has been accumulating in recent years that there exist systematic differences in investment returns between different types of superannuation funds. Several studies (e.g., Coleman, Esho and Wong, 2006, APRA, 2007, Drew, 2003, Langford, Faff, Marisetty, 2006) have found that retail funds tend to earn lower net returns on average than other fund types).
But the retail sector has rejected these and argued that they have different measurement levels and people invest for differing reasons.
So APRA did another survey and released details this week.
"This study examines the investment performance, asset allocation, fees, investment expenses and taxes of large corporate, industry, public sector and retail superannuation funds from 1 July 2001 to 30 June 2006," the regulator said.
APRA says that the large superannuation trustees subject to APRA supervision can be broadly divided into not-for-profit trustees and retail trustees.
"On the one hand, not-for-profit trustees represent corporate, public sector and industry funds, and on the other hand, retail funds are typically part of larger banking, insurance or funds management groups.
"These trustees face identical statutory requirements: to look after their members’ best interests.
"They also have access to similar expertise in funds management and administration.
"Under the auspices of the Council of Financial Regulators, APRA undertook to investigate the reasons for this finding.
"As such, this study further explores potential drivers for return differences.
"Differences in net returns between any two superannuation funds or groups of superannuation funds have four possible explanations: asset allocation, investment performance relative to asset allocation benchmarks, expenses (including fees), and taxes.
"APRA surveyed trustees to gather more detailed data on funds than was available in the standard APRA data collection.
"This survey enabled APRA to examine returns for each of the four candidate causes of difference in return"
Here are excerpts:
This study examines average superannuation fund performance across the four sectors, and the components of these returns on a comparable basis.
To focus on comparable returns, we examine the returns of a particular investment option for each fund, in addition to the total fund return.
We use the default investment option for each fund as this investment option represents the return that a fund member would earn if she did not make an active choice.
Next we use asset allocation information to calculate a benchmark for each investment option and focus our comparisons on performance relative to this benchmark.
We decompose the net returns of the default investment options into four components: (1) asset allocation; (2) investment manager skill in security selection and active investing; (3) costs and (4) taxes, and examine these components in turn.
We also repeat this exercise using a common representative investor for all funds with a focus on fees charged to investors and net returns.
Our main findings about the drivers of returns are as follows:
First, there are very few statistically significant differences in returns between corporate, public sector, and industry funds.
Retail funds, by contrast, sometimes displayed significant differences when compared to the other fund types.
In this paper, we are often able to simplify an analysis of difference to “not-for-profit vs. retail”, rather than an examination of each of the four fund types, because the three not-for-profit fund types are statistically nearly indistinguishable over the five year survey period.
Second, for the investment option provided by each fund, one quarter of the retail funds had conservative asset allocations, whereas almost all not-for-profit funds provided us with an investment option that is a balanced option tilted towards growth assets.
Despite this difference in asset allocation, benchmark returns over the data period are not statistically different across fund types.
Third, in examining gross return performance relative to passive ben
