Analysis MySuper vs Choice

Analysis MySuper vs Choice Australian Institute of Superannuation Trustees 11 September 2018 SYDNEY MELBOURNE ABN 35 003 186 883 Level 1 Level 20 AFSL 239 191 2 Martin Place Sydney NSW 2000 303 Collins Street Melbourne VIC 3000 P +61 2 9293 3700 P +61 3 8621 4100 F +61 2 9233 5847 F +61 3 8621 4111 www.ricewarner.com

Analysis-MySuper vs Choice Table of Contents 1. Executive Summary...2 1.1 Overview...2 1.2 The Data...3 1.3 Summary of Findings...3 2. Background...6 2.1 Introduction...6 2.2 The Data...6 3. Member Demographics and Market Share...9 4. Projection Scenarios... 12 4.1 Funds Under Management... 12 4.2 Impact of Fees... 13 4.3 Impact of Asset Allocations... 13 4.4 Retirement Savings... 14 5. Comparison of fees... 16 6. Analysis of Performance... 18 6.1 Investment performance for Retail and PTM members... 18 6.2 Investment performance by age... 18 6.3 Investment performance by balance... 19 6.4 Investment Performance by Asset Allocation... 20 6.5 Investment Performance by Gender... 22 Appendix A... 23 This report constitutes a Statement of Advice as defined under the Financial Services Reform Act. It is provided by Rice Warner Pty Ltd. which holds Australian Financial Services Licence number 239 191. This report should not be distributed, in whole or in part, without Rice Warner s prior written consent. September 2018/376038_5 Page 1 of 23

1. Executive Summary One of the prime drivers of superannuation value has been the 5% return provided above inflation for 25 years or more by MySuper products and their predecessors, Balanced Funds. There is a general view that members who choose to invest their superannuation assets outside MySuper products should achieve a better outcome as they are tailoring their investment decisions to their personal circumstances and financial goals. However, our analysis shows that this is not generally the case. This report shows that higher returns and lower fees of MySuper products could continue to have a material impact on the superannuation market over the next 10 years. Our modelling shows that, if all members of APRA regulated funds had fees and returns consistent with the average MySuper product, superannuation assets would be $52.5 billion higher (in today s dollars) in 10 years time (compared to our baseline projections). The economic impact of Choice members being $52.5 billion worse off in a decade s time would be significant. $52.5 billion equates to 3% of GDP (at 30 June 2017) and is more than an entire year of Age Pension outlays. 1.1 Overview The Australian Institute of Superannuation Trustees () has commissioned this report to compare fees and outcomes for MySuper and Choice products for the Retail and Profit-to-Member (PTM) market segments. Specifically, is interested in assessing the impact of fees and asset allocation of MySuper and Choice products on net member outcomes and industry size. MySuper products were developed to create a range of easily comparable, relatively simple products, which would focus competition on costs and net investment returns. commissioned this report to investigate the extent to which this has been successful to date, and therefore likely to significantly improve retirement outcomes compared to other superannuation products (i.e. Choice). Projections scenarios in this report show that projected Superannuation Funds Under Management (FUM) would be higher if fees and asset allocations for everyone in an APRA regulated fund were set at levels consistent with the average MySuper product. This result is broadly driven by lower assumed fees and higher returns for MySuper products. Asset allocation assumptions for MySuper and Choice products have been estimated using information from Rice Warner s Super Insights Database which utilises the actual choice decisions of over 10 million members. Assumed fee rates for average MySuper and Choice products are preliminary analysis from Rice Warner s Superannuation Fees Report 2018. In addition to this projections analysis, we have included analysis of APRA s Annual Superannuation Statistics from 2017 to show how the market share of Profit-to-Member (PTM) funds has increased over the past 10 years. Analysis of APRA data also highlights the distribution of member accounts by age and account balance for the PTM and Retail sectors. Analysis of performance outcomes of MySuper and Choice products for both Retail and PTM market segments for the three years to 2017 is also included in the report. Analysis of Self-Managed Superannuation Funds (SMSFs) is not included in this report. September 2018/376038_5 Page 2 of 23

1.2 The Data In this report, we utilise data published by APRA supplemented by Rice Warner s Super Insights research, which involves the collection of annual statement data pertaining to over 10 million members per annum. Further information on the dataset can be found in Section 2.2. 1.3 Summary of Findings Member demographics and historical market share of PTM and Retail funds (taken from publicly available APRA data) are included in this report to provide context for the projections scenario modelling. Projections scenarios in this report show the impact MySuper products would have on FUM, fees, and earnings if everyone in an APRA regulated fund were to experience fees and returns consistent with MySuper products over the next ten years. These results are compared to projected outcomes if the whole superannuation market were to experience returns and fees consistent with Choice products over the next 10 years. Assumed fees and asset allocation assumptions for MySuper and Choice products have been estimated using Rice Warner s Super Insights Research, preliminary analysis from Rice Warner s Superannuation Fees Report 2018 and publicly available APRA statistics. We have also included a comparison of fees and returns for MySuper and Choice products for PTM and Retail sectors. 1.3.1 Member demographics and market share APRA s Annual Superannuation Statistics from 2017 shows that the market share of PTM funds has increased in the last 10 years. A relatively high proportion of individuals aged over 65 hold their superannuation assets in Retail funds. Those with high superannuation balances are even more likely to hold their assets in Retail funds. Generally, older members with higher balances will tend to move into Choice products. As these members have generally higher average balances, their impact on overall fees and earnings of Choice products is fairly significant. This impact of fees and returns is reflected in our projections scenarios. 1.3.2 Projection Scenarios We have modelled two hypothetical projection scenarios in this report: One scenario sets fees and asset allocations for all funds consistent with the fees and asset allocations of an average MySuper product (the MySuper scenario). The second scenario sets fees and asset allocations consistent with average Choice products (the Choice scenario). Overall, projected FUM is $108.0 billion (in today s dollars) higher by 2027 in the Mysuper scenario relative to the Choice scenario. Further, while member based fees are higher for MySuper ($73 p.a.) than Choice ($63 p.a.) products, lower asset based fees found in MySuper products offset this and drive a projected $18.2 billion (cumulatively) decrease in fees under the MySuper scenario, relative to the Choice scenario. September 2018/376038_5 Page 3 of 23

Lower investment returns in Choice products (relative to MySuper products) results in projected earnings that are $146.1 billion higher (cumulatively) over the 10-year period under the MySuper scenario, relative to the Choice scenario. 1.3.3 Comparison of fees Rice Warner s Superannuation Fees Report 2018 provides information on fees paid by fund type and whether an investment choice has been made. Preliminary results from analysis in the report show that: Total expenses as a proportion of assets are 0.70% for PTM funds and 1.40% for Retail Funds. This includes 0.03% in advice fees for PTM funds, and 0.46% for advice fees for Retail funds. For the whole industry, total fees (excluding advice fees) as a proportion of total assets are 0.78% for MySuper members and 0.83% for Choice members. MySuper asset based fee rates (excluding advice fees) of 0.59% are generally lower than those of Choice products (0.76%). However, MySuper accounts generally have smaller balances than Choice accounts. The dollar fee plus the asset based fee for MySuper products results in similar levels of total fees as a proportion of total assets for Choice and MySuper products. This gap in asset based fees suggests that, as members with higher average balances move into Choice products, the fees paid by these members would increase significantly. 1.3.4 Analysis of Performance Analysis of performance for Retail and PTM MySuper and Choice members for the three years to June 2017 shows: Retail Choice members have achieved lower returns over the period. Investment earnings for Choice members over 65 are lower than younger age cohorts and lower than MySuper members over 65. The strongest performing investment option over the three years to 30 June 2017 were High Growth options. However, MySuper options tended to outperform Growth options over this period. Though some of the results in this report are impacted by differences in asset allocation for different membership classes, the Productivity Commission s draft report into the efficiency and competitiveness of the superannuation industry analysed the investment performance of the Retail and PTM funds and MySuper and Choice products against a series of benchmarks (i.e. controlling for asset allocation). This report arrived at some similar conclusions. A copy of this report can be found on their website. 1.3.5 Conclusion Overall: Projected FUM is $108.0 billion (in today s dollars) higher by 2027 if the members of all APRA funds were in MySuper products rather than Choice products. This can be broken down into: Choice members would be $52.5 billion worse off had they stayed invested in the average Choice product rather than a MySuper product, assuming the differences continue. MySuper members would be $55.6 billion better off by not selecting the average Choice product. High fees notably in retail Choice funds - are also key factors behind the underperformance. September 2018/376038_5 Page 4 of 23

For the whole industry, the average total fees (excluding advice fees) as a proportion of total assets are 0.78% for MySuper members and 0.83% for Choice members. All Australians, but particularly future taxpayers, have a stake in our super system operating at maximum efficiency. The economic impact of Choice members being $52.5 billion worse off in a decade s time would be significant. Australia s compulsory super system was designed to help reduce Government expenditure on the Age Pension. Reducing the potential super savings pool by $52.5 billion puts added pressure on Government s Pension bill. $52.5 billion equates to 3% of GDP (at 30 June 2017) and is more than an entire year of Age Pension outlays. This report was prepared and peer reviewed for Australian Institute of Superannuation Trustees by the following consultants. Prepared by Alun Stevens Senior Consultant Martin Stevenson Consultant Telephone: (03) 8621 4105 Telephone: (02) 9293 3722 alun.stevens@ricewarner.com martin.stevenson@ricewarner.com Richard Dunn Consultant Telephone: (02) 9293 3713 richard.dunn@ricewarner.com Peer Reviewed by Michael Rice CEO & Authorised Representative Telephone: (02) 9293 3707 michael.rice@ricewarner.com 11 September 2018 September 2018/376038_5 Page 5 of 23

2. Background 2.1 Introduction As part of their 2018 research agenda, the Australian Institute of Superannuation Trustees has commissioned this report to compare MySuper and Choice products for the PTM and Retail market segments. This report provides a sensitivity analysis of projected retirement income outcomes if the whole superannuation market experiences fees and returns consistent with MySuper products over the next ten years. These results are compared to projected outcomes if the whole superannuation market experiences return and fees consistent with Choice products over the next ten years. Member demographics and historical market share of PTM and Retail funds (taken from publicly available APRA data) are included in this report to provide context for the projections scenario modelling. Assumed fees and asset allocation assumptions for MySuper and Choice products have been estimated using Rice Warner s Super Insights Research, Rice Warner s Superannuation Fees Report and publicly available APRA statistics. A comparison of fees paid for Choice and MySuper members in Retail and PTM funds is also included, along with a comparison of historical performance for the three years to 30 June 2017 for MySuper and Choice members in PTM and Retail funds. The report illustrates the distribution of outcomes for MySuper and Choice products by age, gender, member balance and investment decision. 2.2 The Data 2.2.1 APRA Data In addition to the Super Insights data, we have also examined statistics published by APRA. The APRA data provides a market wide benchmark that complements the depth of the Super Insights data. We have examined the following sources for this report: Annual Fund Level Superannuation Statistics - containing detailed profile and structure, financial performance and financial position, conditions of release, fees and membership information for APRA-regulated superannuation funds. Quarterly MySuper Statistics containing data on MySuper products. The report contains information on the product profile, product dashboard measures, asset allocation targets and ranges, investment performance, fees disclosed for MySuper products, or where relevant, for the lifecycle stages underlying MySuper products with a lifecycle investment strategy, and MySuper dashboard website URLs. 2.2.2 Super Insights data Super Insights is a yearly data analysis project undertaken by Rice Warner which involves the collection of anonymised details for more than 10 million member accounts per annum. The database is the only one of its kind for superannuation members and supplements our other research in forming the basis for many of our insights into the superannuation market. September 2018/376038_5 Page 6 of 23

The data used for this study is de-identified member data at 30 June 2014. The data requirements are substantially aligned with the data used for benefit statements at the end of each financial year. The data used for Super Insights covers the following aspects of the market: demographics investments contributions insurance fees employers pensions exits. The sample is drawn from a variety of large and small funds that represent all market sectors (industry, retail and some public sector and corporate funds). This data allows us to provide detailed analysis across several key areas: Analysis of membership profile by different segments. Quantifying and comparing differences between members across gender, ages and geographic area. Understanding inflows of funds from contributions and roll-ins. Investment strategy participation in default strategies versus choice investments, and how choice members differ from default members. Insurance analysis take up of cover and average levels of cover along with analysis of insurance needs. Employer analysis including average number of employers, average employer size, distribution of employers. Pensions demographic and behavioural analysis of pension members including investment decisions, drawdown rates and balance analysis. Exits analysis identification of key areas of member leakage from funds in absolute terms and the dynamics of the exiting membership. 2.2.3 Caveats and investment performance Although the Super Insights data is extremely detailed and represents a significant sample of the superannuation industry, there are the following limitations. Overall, the quality of the sample is determined by the participant funds and despite the large overall participation in the study, there is still a bias in the data we have collected. For instance, the sample is skewed by the largest funds in the study. Also, members who may currently be in smaller, higher cost funds are less likely to be a part of the sample. Data on fees and investment performance have been drawn separately from Rice Warner s Superannuation Product database. September 2018/376038_5 Page 7 of 23

2.2.4 Unable to examine SMSFs We have not undertaken any analysis regarding the differences in fees or investments for members in SMSFs for this report. 2.2.5 Fees analysis The fees assumptions used for this report have been taken from a preliminary assessment of Rice Warner s Superannuation Fees Report 2018. As such, assumptions on superannuation fees may differ from those published in the final version of Superannuation Fees Report 2018. September 2018/376038_5 Page 8 of 23

% of total market share MySuper vs Choice 3. Member Demographics and Market Share The following analysis examines the member demographics and historical market share of Retail and PTM funds. This analysis is based on APRA s 2017 Annual Fund Level Statistics. Estimates of MySuper and Choice assets by sector as at 30 June 2017 is also included in this section. This information has been taken from Rice Warner s 2017 Superannuation Market Projections Report. This provides useful context for the remainder of the report which is based on Rice Warmer s Superannuation Fees Report, Super Insights data and market projections modelling. Graph 1 shows that the market share of PTM funds (by value of assets) has increased over time. Graph 1. Historical market share of Profit-to-Member and Retail Funds 100 90 80 70 60 50 40 30 20 10 0 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 Year Retail PTM Source: APRA annual fund level statistics. Graph 2 shows that the proportion of people holding their superannuation in Retail funds rises significantly in the early years of employment, is fairly stable through the remainder of employment and then rises again above age 65. September 2018/376038_5 Page 9 of 23

% of total accounts % of total accounts MySuper vs Choice Graph 2. Distribution of member accounts by age 100 90 80 70 60 50 40 30 20 10 0 <25 25 to 34 35 to 44 45 to 49 50 to 54 55 to 59 60 to 64 >65 Age Retail PTM Source: APRA annual fund level statistics. Graph 3 shows a relatively high proportion of those with large balances (over $1 million) hold retail funds. Graph 3. Distribution of members accounts by balance 100 90 80 70 60 50 40 30 20 10 0 < $1,000 $1,000 to $24,999 $25,000 to $49,999 $50,000 to $99,999 $100,000 to $199,999 $200,000 to $499,999 $500,000 to $999,999 > $1,000,000 Balance Retail PTM Source: APRA Annual Superannuation Statistics. September 2018/376038_5 Page 10 of 23

Graph 2 and Graph 3 confirm that people with higher balances are opting for Retail Choice products above age 65. Analysis of our Super Insights database confirms that people with higher balances are opting for more defensive asset allocations in a Choice environment as they approach retirement. The split of Choice and MySuper by market segment as at 30 June 2017 is shown in Table 1. The portion of PTM FUM invested in MySuper products is substantially higher than the equivalent portion for Retail funds. Across both segments, approximately 40% of FUM in the accumulation phase was invested in MySuper products. Table 1. Estimated MySuper and Choice assets at 30 June 2017 Market segment Pre-Retirement Retirement Total Choice MySuper Choice All MySuper ($m) ($m) (%) ($m) ($m) (%) Not-for-Profit Funds Corporate Funds 44,490 21,125 32.2 10,576 76,191 27.7 Industry Funds 159,713 349,826 68.7 63,189 572,727 61.1 Public Sector Funds 167,511 120,232 41.8 57,703 345,445 34.8 Subtotal 371,714 491,183 56.9 131,467 994,364 49.4 Commercial Funds Employer Master Trusts 57,506-160,991 Personal Superannuation 249,885 103,482 24.9-249,885 16.4 Eligible Rollover Funds 4,400-4,400 Commercial Post Retirement Products - - - 216,298 216,298 - Subtotal 311,791 103,482 24.9 216,298 631,572 16.4 Self-Managed Super Funds Self-Managed Super Funds 217,507 - - 481,118 698,625 - Subtotal 217,507 - - 481,118 698,625 - Total superannuation market 901,012 594,666 39.8 828,883 2,324,561* 25.6 *Excludes unallocated reserves. Source: Rice Warner s Superannuation Market Projections Report 2017. September 2018/376038_5 Page 11 of 23

Difference to BAU ($b) MySuper vs Choice 4. Projection Scenarios To illustrate the impact of fees and asset allocations of MySuper and Choice products on the industry, two scenarios have been analysed using Rice Warner s Retirement Outcomes Model: A MySuper scenario, where fees and asset allocations for all Corporate, Retail, Industry, and Public Sector funds are set at average MySuper levels. A Choice scenario, where fees and asset allocations for all Corporate, Retail, Industry and Public Sector funds are set at average Choice levels. Projections under these scenarios are compared to the BAU projections scenario included in Rice Warner s Superannuation Market Projections Report 2017 (the BAU scenario). The fees and asset allocation assumptions used are shown in Appendix A. 4.1 Funds Under Management The projected impact on aggregate FUM of the MySuper and Choice scenarios is shown in Graph 4. Under the MySuper scenario, FUM is projected to grow steadily (relative to BAU) over the course of 10 years resulting in FUM of $52.5 billion (or 1.5%) higher than the BAU scenario by 2027 (in today s dollars). Under the Choice scenario, FUM is projected to be consistently lower (relative to BAU) over the course of 10 years. By the end of the projection period FUM is in total $55.6 billion (or 1.6%) lower than the BAU scenario (in today s dollars). These results are driven by the assumed lower fees and more growth oriented asset allocations of MySuper products relative to Choice products. Graph 4. Impact of scenarios on FUM relative to BAU scenario ($b) in today s dollars 60 40 20 0-20 -40-60 -80 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 Year All MySuper All Choice September 2018/376038_5 Page 12 of 23

Difference to BAU ($b) MySuper vs Choice 4.2 Impact of Fees The annual impact to fees of both the MySuper and Choice scenarios is show in Graph 5. Under the MySuper scenario, fees are projected to be $3.5 billion (or 17.1%) lower per year than the BAU scenario in 2017. Under the Choice scenario, fees are projected to be $559 million higher per year (or 2.7%) than BAU scenario in 2017. The Rice Warner Retirement Outcomes Model assumes fees converge to a longterm assumption by 2025. Therefore, the annual impact on fees reduces (in absolute terms) over the projection period. Under the MySuper scenario, members are projected to pay cumulatively $18.8 billion less (in today s dollars) in fees over this projection period compared to the BAU scenario. Under the Choice scenario, members are projected to pay cumulatively $617 million less (in today s dollars) in fees over this period compared to the BAU scenario. While annual fees are initially increased relative to BAU under the Choice scenario, they decrease in the long term as the long-run fees assumption is reached due to decreased balances. Similarly, fees under the MySuper scenario are higher than the fees under the Choice scenario once the long-run fees assumption is reached. This is driven by the higher balances for the MySuper scenario at this point. Graph 5. Impact of scenarios on fees relative to BAU scenario ($b) in today's dollars 3 2 1 0-1 -2-3 -4 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 Year All MySuper All Choice 4.3 Impact of Asset Allocations Analysis of Rice Warner s Super Insights data suggests that members invested in the Choice environment tend to have a higher allocation to defensive assets, relative to members in MySuper. This is reflected in the projection results for gross earnings shown in Graph 6. September 2018/376038_5 Page 13 of 23

Difference to BAU ($b) MySuper vs Choice Graph 6. Impact of scenarios on investment earnings relative to BAU ($b) in today's dollars 8 6 4 2 0-2 -4-6 -8-10 -12 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 Year All MySuper All Choice Under the MySuper scenario, earnings in 2017 are $3.1 billion higher than the BAU scenario (equivalent to returns moving from 6.31% to 6.44%). By the end of the projection period, earnings are $6.7 billion higher than the BAU scenario (equivalent to returns moving from 6.54% to 6.63%). Cumulative earnings for the 10-year projection period are $55.8 billion higher (in today s dollars) under the MySuper scenario. Under the Choice scenario, earnings in 2017 are $5.7 billion lower than the BAU scenario (returns falling from 6.31% to 6.07%). By the end of the projection period, earnings are $10.9 billion lower than the BAU scenario (returns falling from 6.54% to 6.33%). Under the Choice scenario, earnings are (in total) $90.3 billion lower (in today s dollars) over the 10-year projection period. 4.4 Retirement Savings 4.4.1 Average Balance Analysis from the Retirement Outcomes Model The impact of MySuper and Choice at a member level has been analysed by comparing average member balances over time under the three scenarios for those who are 15 years old today (see Graph 7). At age 65, the average balance under the MySuper scenario is $11,000 higher (or 1.5%) than the BAU scenario (increasing from $760,000 to $771,000). Under the Choice scenario, the average balance for 65- year old s in 50 years is projected to be $26,000 lower (or 3.4%) than the BAU scenario (from $760,000 to $734,000). These results are indicative of the projected impact of the MySuper and Choice scenarios over the course of a working life. September 2018/376038_5 Page 14 of 23

Average Balance at Age ($) MySuper vs Choice Graph 7. Projected average balances of those aged 15 years old today in today s dollars 900,000 800,000 700,000 600,000 500,000 400,000 300,000 200,000 100,000 0 15 25 35 45 55 65 Age BAU All MySuper All Choice To further illustrate the impact of MySuper vs Choice products over a person s working life, have commissioned Rice Warner to provide analysis of the impact of MySuper vs Choice products under a range of hypothetical cameo scenarios for individuals who experience different financial circumstances over the course of their life. This cameo analysis is provided in a supplementary report. September 2018/376038_5 Page 15 of 23

5. Comparison of fees Preliminary analysis from Rice Warner s Superannuation Fees Report 2018 provides information on fees paid by industry sector and by those in MySuper and Choice products. Table 2 shows the aggregate number of accounts, total assets, and total expenses for the year to 30 June 2017, split by PTM and Retail sectors. Table 2. Fees and expenses by superannuation segment - Year to 30 June 2017 Sector Accounts Assets Expenses ('000) Profit-to-Members 14,958 994,364 6,924 Retail 12,276 631,572 8,872 Small Funds 1,124 698,625 6,538 Total 28,359 2,324,561 22,334 ($m) A comparison of the average fees (excluding advice fees) paid by MySuper and Choice members is shown in Table 3 and Table 4. For PTM funds, fees (as a percentage of assets) tend to be 0.19% lower for Choice members than for MySuper members. Conversely, fees for Retail Choice members are 0.37% higher than Retail MySuper members. For the whole industry, the rate of fees paid for MySuper and Choice products are similar. Fees for MySuper members are 0.78% of assets whilst fees for Choice members are 0.83% of assets. Asset based fee rates are generally lower for MySuper products than Choice products. However, MySuper members also tend to have smaller balances and so total fees (both dollar and asset fees) as a proportion of total assets are comparable to Choice members. Table 3. Average MySuper fees by sector (excluding advice) for Year to 30 June 2017 Fee Rate Profit-to-Members Retail Whole Industry Operating 0.28 0.34 0.29 Investment Management 0.54 0.32 0.49 Total 0.81 0.66 0.78 Investment management fees are lower for PTM Choice customers than for PTM MySuper customers. This is likely due to a higher proportion PTM Choice members assets being invested in more defensive options with lower fees. Table 4. Average Choice fees by sector (excluding advice) for Year to 30 June 2017 Fee Rate (%) Profit-to-Members Retail Whole Industry Operating 0.13 0.40 0.27 Investment Management 0.49 0.63 0.56 Total 0.62 1.03 0.83 (%) September 2018/376038_5 Page 16 of 23

The projection scenario included in Section 4 illustrating the impact of MySuper on market projections assumes that all products have levels of fees and returns consistent with MySuper products. Therefore, this scenario reflects the dollar fee and asset based fee rate of MySuper products. Applying these MySuper fees assumptions to the larger accounts of Choice members results in overall lower fees for this scenario. September 2018/376038_5 Page 17 of 23

6. Analysis of Performance Rice Warner s Super Insights database was used to analyse investment performance for Retail and PTM members in MySuper and Choice products. Analysis of investment performance for MySuper and Choice members by age category, gender, balance and asset allocation has also been included. Individuals investment options were mapped to net investment earnings data for three years to 30 June 2017. Investment performance was not provided for the four years to 30 June 2017 (since the inception of MySuper) as limited data was available for Retail MySuper products in earlier years. 6.1 Investment performance for Retail and PTM members Table 5 compares the investment returns of MySuper and Choice members for PTM and Retail funds over the three years to June 2017. Table 5. Investment earnings p.a. by segment - three years to 30 June 2017 Segment MySuper Members 8.32 8.00 Choice Members 7.63 5.63 PTM (%) Retail Investment performance of Retail MySuper members has been estimated from APRA quarterly and annual statistics. The remainder of this performance data has been estimated from Rice Warner s Super Insights database. The lower return of Retail Choice members is in part due to the more defensive asset allocation of these members. 6.2 Investment performance by age Graph 8 shows the difference in investment earnings between MySuper and Choice members by age. Investment earnings for Choice members are lower for older age cohorts. This reflects the more defensive asset allocation decisions made by members as they approach retirement. September 2018/376038_5 Page 18 of 23

Investment Earnings (%) MySuper vs Choice Graph 8. Investment earnings p.a. by age - three years to 30 June 2017 10.00 9.00 8.00 7.00 6.00 5.00 4.00 3.00 2.00 1.00 0.00 <25 25 to 34 35 to 44 45 to 49 50 to 54 55 to 59 60 to 64 >65 Age MySuper Members Choice Members 6.3 Investment performance by balance Examining investment performance for different balances, shows that returns are lower in the cohort of people with balances between $500,000 and $1 million. These results are likely driven by age. Older members tend to have higher balances and to have more defensive asset allocations as they approach retirement. Conversely, the relatively higher returns of MySuper and Choice members with assets between $1,000 and $500,000 is likely due to the more aggressive asset allocation of younger members as they accumulate higher balances. September 2018/376038_5 Page 19 of 23

Investment Earnings (%) Investment Earnings (%) MySuper vs Choice Graph 9. Investment earnings p.a. by balance - three years to 30 June 2017 10.00 9.00 8.00 7.00 6.00 5.00 4.00 3.00 2.00 1.00 0.00 < $1,000 $1,000 to $24,999 $25,000 to $49,999 $50,000 to $99,999 $100,000 to $199,999 $200,000 to $499,999 $500,000 to $999,999 Over $1m Balance MySuper Members Choice Members 6.4 Investment Performance by Asset Allocation Graph 10 illustrates investment performance for various asset allocations for the three years to June 2017. The strongest performing investment options over the three years to 30 June 2017 were High Growth options. However, MySuper options tended to outperform Growth options over this period. Graph 10. Investment earnings p.a. by investment option - three years to 30 June 2017 12.00 10.00 8.00 6.00 4.00 2.00 0.00 MySuper High Growth Growth Balanced Moderate Conservative Cash Investment Options MySuper Members Choice Members September 2018/376038_5 Page 20 of 23

Investment Earnings (%) MySuper vs Choice As expected, members returns are lower for more defensive options. Graph 11 shows the (unweighted) average of returns by sector. Rice Warner s product database was used to produce this analysis. Over 200 options for each of the option groups were analysed for three year returns to 30 June 2017. On a like-for-like basis, PTM funds tended to outperform Retail funds in High Growth, Moderate, and Conservative options. However, Retail funds outperformed PTM funds in Growth options. Balanced and Cash options tended to perform similarly for both sectors. Note that these are net returns, and some of the difference in returns may be explained by the higher fees generally charged by Retail funds. Graph 11. Investment earnings p.a. by investment option and market segment three years to 30 June 2017 10 9 8 7 6 5 4 3 2 1 0 High Growth Growth Balanced Moderate Conservative Cash Investment Options PTM Retail For the purposes of both analyses, the following asset allocation assumptions have been made: High Growth members invest over 90% in aggressive assets. Growth members invest over 70% in aggressive assets. Balanced members invest over 60% in aggressive assets. Moderate members invest over 40% in aggressive assets. Conservative members invest less than 40% in aggressive assets. Cash members invest over 90% in cash. September 2018/376038_5 Page 21 of 23

6.5 Investment Performance by Gender Investment earnings for male Choice members have generally exceeded returns for female Choice members over the three-year period. This was because men in the sample tended have more aggressive asset allocations than women. Table 6. Investment earnings p.a. by gender - Three years to 30 June 2017 Member Type Females Males MySuper Members 7.94 8.29 Choice Members 6.72 7.16 (%) September 2018/376038_5 Page 22 of 23

Appendix A Table 7. Assumed Member and Asset Fee for MySuper and Choice scenarios Member Fee ($) Asset Fee (%) MySuper 73.08 0.70 Choice 62.62 0.97 Table 8. Assumed MySuper and Choice Asset Allocations Asset Class MySuper Choice (%) Aust Equities 26.8 24.4 Int'l Equities 25.5 21.4 Listed Property 4.4 4.6 Direct Property 4.4 4.6 Aust Fixed Interest 7.0 8.5 Int'l Fixed Interest 7.0 8.5 Cash 4.9 12.9 Other 9.5 7.0 Unlisted Equities 2.1 1.7 Infrastructure 8.3 5.9 September 2018/376038_5 Page 23 of 23