The retirement income system

Preliminary assessment


The retirement phase of superannuation is underdeveloped and does not meet the risk management needs of many retirees.

Individuals are justifiably concerned about outliving their savings. As it is currently designed, the retirement income system does not adequately enable many individuals to manage their income and risks in retirement effectively.

During the accumulation phase, employers make Superannuation Guarantee contributions automatically on behalf of employees, with defaults applying to those who are less engaged with the system. This framework effectively guides individuals through the accumulation phase. However, such a framework does not exist for the retirement phase. Instead, retirees make critical once-in-a-lifetime decisions regarding when and how to draw down their savings over the remainder of their lives. Individuals also have to manage the investment, inflation and longevity risks associated with these decisions. Many, but not all, individuals are unprepared for these important decisions.

The lack of effective risk management, particularly longevity risk management, is a major weakness of Australia’s retirement income system.

Assessment of current products

As discussed in the Framework section of this chapter, retirees require income products that deliver three main features: income, risk management and flexibility. No product provides all of these features (Table 8.1). Account-based pensions are more flexible than annuities but provide much less risk management. Thus, current retirees benefit from two of the three features they need, but the majority do not effectively manage risk.

Table 8.1: Features of retirement income products

This table describes the income, flexibility and risk management characteristics of both account-based pensions and lifetimes annuities. Account-based pensions are more flexible than annuities but provide much less risk management.

One important feature of a retirement income product is the expected income it delivers during retirement from a given accumulated balance. In this report, ‘income efficiency’ is defined as the expected present value of this income as a percentage of the purchase price of the product.

A lifetime annuity has an estimated income efficiency of 76 per cent for an average 65-year-old male (for reasons discussed later in this chapter).16 The income efficiency of an account-based pension that is drawn down at the minimum rate for a 65-year-old male is around 70 per cent.17 The remaining funds in an account-based pension will typically be left to beneficiaries.

There are trade-offs between income efficiency, flexibility and risk management. For example, a higher drawdown rate will increase the income efficiency of an account-based pension, as a larger portion of the pension is expected to be paid as income during the life of the retiree, but will increase the risk of outliving savings.

Longevity risk

Most individuals risk outliving their savings. Uncertainty around the length of time in retirement makes it difficult for people to manage their wealth efficiently. For example, although the life expectancy of a 65-year-old female today is about 89 years, 10 per cent of 65-year-old females will die before they reach 77 years and 10 per cent will live past 100 years (Chart 8.3).

Chart 8.3: Distribution of cohort life expectancy for a 65-year-old female

This chart shows the distribution of cohort life expectancy for a 65 year-old female. 10 per cent of 65 year-old females are expected to die before they reach 77, and 10 per cent are expected to live past 100.

Source: Australian Government Actuary.18

This uncertainty makes it difficult for people to choose the ‘right’ rate at which to draw down their wealth from an account-based pension. Retirees risk either exhausting their retirement savings prematurely (and falling back on the Age Pension), or drawing down their benefits too conservatively and having a lower standard of living in retirement.

Managing longevity risk on an individual basis can lead to a dynamically inefficient allocation of resources. To reduce the risk of outliving their assets, individuals need a higher savings rate during their working life to target a given level of consumption in retirement. This can lead to inefficiently low consumption during an individual’s working life. Alternatively, insufficient provisioning for retirement leads to inefficiently low levels of consumption in retirement.

In Australia, the total cost of this inefficiency is likely to be substantial, given both the large number of people with account-based pensions who draw down their funds at the minimum rate,and the large proportion of older Australians receiving the Age Pension.19 The cost is largely borne by retirees and, to a lesser extent, the Government through the total cost of tax concessions and Age Pension payments.

A large body of evidence, including from surveys, supports the assertion that people value longevity risk protection. Results from one survey suggest that more than 90 per cent of Australians over the age of 50 believe that ’money that lasts my lifetime’ is somewhat important or very important.20 Over half of respondents to another survey were either worried or extremely worried about outliving their savings.21 When asked to identify the single most important feature in a retirement income product, twice as many members in the accumulation phase identified ’income that lasts a lifetime‘ as identified the second most popular response.

Lump sums

Several submissions refer to Australia’s (real or perceived) ‘lump sum culture’ and call for a shift towards income streams and longevity risk management in retirement.

In general, superannuation funds seek to maximise a member’s lump sum balance at retirement, rather than target a level of income in retirement. This is a common feature of defined contribution schemes globally.

Around half of superannuation benefits in retirement are currently paid as lump sums.22 Close to half of retirees take a lump sum only — having never received an income stream payment.23 For people with small superannuation balances, taking the entirety of their benefits as a lump sum may be an optimal strategy because the income stream generated from a small balance is negligible and has relatively high costs and no tax advantages. However, as more people retire with higher account balances, there has been a gradual trend towards purchasing income stream products (Chart 8.4).

Chart 8.4: Superannuation benefit payments — lump sums and pensions

This chart shows the amount of superannuation assets drawdown as lump sums, and income streams, over the period 2003 to 2013. Around half of benefits are paid as lump sums, half as income streams.

Source: APRA.24

The ability to use superannuation lump sums to extinguish debt can encourage higher pre-retirement consumption and borrowing:25

  • Approximately 44 per cent of retirees who take a lump sum use it to pay off housing and other debts, to purchase a home, or make home improvements.
  • A further 28 per cent use their lump sum to repay a vehicle or holiday loan or to purchase a holiday or new vehicle.

The number of households entering retirement with debt, particularly a mortgage, is increasing.

Age Pension means-test

The Age Pension means-test can distort financial decisions made before and during retirement.

Butler, Peijnenburg and Staubli concluded that a pension means-test can substantially reduce demand for longevity insurance products.26 The Age Pension in Australia is the primary source of longevity insurance and is contributing to the low demand for market-based products that provide such insurance.

Similarly, Hulley, McKibbin, Pedersen and Thorp found that means-tests for the public pension encourage eligible and near-eligible retirees to decumulate assets faster and choose riskier portfolios, especially early in retirement.27, 28 This distorts the amount of risk eligible and near-eligible retirees expose themselves to and shifts longevity risk to the Government.

Stakeholders have suggested that some individuals with small- and medium-sized asset balances tend to structure their affairs around the Age Pension means-test. Indeed, maximising access to the Age Pension is (understandably) a central feature of financial advice for retirees. Mercer’s submission illustrates the reasons for this by looking at the case of ‘Fred’29:

Fred is a single (home-owner) retiree with a $500,000 account based pension. He is drawing down the minimum 5 per cent of the account each year. By spending $100,000 of his superannuation account on an overseas trip, Fred increases his age pension by $150 a fortnight ($3,900 a year).

If he continues to draw down the minimum allowed from his account based pension, his withdrawals reduce by $5,000 a year meaning he is only $1,100 a year (or $21 per week) worse off after his $100,000 overseas trip.

There is no empirical evidence of this behaviour on a large scale. However, a significant portion of superannuation savings are being depleted before reaching Age Pension age. On average, around one-third of superannuation assets are withdrawn by the time an individual reaches the Age Pension eligibility age.30 Around one-quarter of people with a superannuation balance at age 55 have depleted their balance by age 70.31 Aligning incentives with desired outcomes would tend to reduce the longevity risk borne by the Government. It would also improve the capacity of future Governments to provide the Age Pension to the many Australians who will need it in retirement.

Defaults cease at retirement

Many individuals are not required to interact with their superannuation fund during their working lives. Employers make Superannuation Guarantee contributions automatically on behalf of employees. If employees do not nominate a fund or investment option, contributions are made to a default fund chosen by the employer, and investment decisions are made by professionals.

These superannuation defaults cease at retirement. Retirees make once-in-a-lifetime critical decisions about how to manage their assets, ideally to deliver an income stream and to manage effectively the associated investment, inflation and longevity risks. The consequences of a poor decision can be severe and costly to rectify; individuals typically have limited or no capacity to top up their funds after they retire. Benefits remain in a superannuation fund until an active decision is made. Assets held in a superannuation account are not used to provide income, and the returns on these assets are not exempt from tax.

A number of studies have reported that Australians are unprepared for the financial decisions they need to make as they approach retirement. They know neither how much to save for retirement nor how to create income from their accumulated balance: most people in their 50s and 60s have not planned the main aspects of their retirement,32 and only one-third of accumulating superannuation members have heard of lifetime annuities.33 As discussed in the Consumer outcomes chapter, this highlights the importance of high-quality financial advice for retirees and those near retirement.

Other countries with defined contribution schemes face similar challenges to those in Australia. The prestigious Squam Lake Working Group found that defined contribution schemes “place much greater burdens on consumers to make good financial decisions. There is widespread concern that many households are not up to the task”.34

Low demand for longevity-protected products

There is low demand in Australia for products that provide protection against longevity risk. Annuities are currently the only market-based products that provide longevity insurance.

Academics have struggled to solve the ‘annuity puzzle’ — the fact that while annuities deliver desirable characteristics, demand for annuities is very low.35, 36, 37 A range of behavioural biases have been found to discourage people from purchasing products with longevity protection:

  • Annuities are perceived to be risky gambles rather than insurance.
  • Individuals underestimate their life expectancy.
  • Retirees want the flexibility to meet unforeseen cash requirements and leave their residual assets as a bequest.
  • Individuals undervalue future consumption relative to current consumption.
  • Annuities are not perceived to deliver value for money.

A range of factors reduce the attractiveness of annuity pricing and increase the attractiveness of account-based pensions.

Life insurance companies impose various margins on annuities that increase their price for a given income stream. These companies must use significant capital to fund their assets and minimise the risk of failure. They must cover administrative costs and provide profits to shareholders.

Annuities are also made more costly by adverse selection. Individuals usually have more information about the factors affecting their own life expectancy than retirement product providers. Those expecting to live longer are more likely to buy longevity-protected products. Providers will respond by setting prices appropriate for those likely to live longer lives, which makes the pricing unattractive for others. Challenger estimates that for a 65-year-old male buying an annuity in the current environment, adverse selection lowers indexed annuity payments by around 7 per cent (from around $6,000 annually per $100,000 premium to $5,600).38

Annuity prices are also affected by the level of interest rates when they are purchased. The lower level of interest rates over the past 20 years has reduced the income generated from annuities (and some other investments), although this has also been associated with a reduction in inflation.

Finally, retirees have what they perceive to be attractive alternatives to annuities:

  • Account-based pensions invested in high-yielding equities may offer attractive returns compared to annuities, in part due to the benefits of dividend imputation.
  • The Financial Claims Scheme removes counterparty risk for retirees who save through bank deposit products and term deposits, but not annuities.
  • The Age Pension provides government-backed longevity protection.

16 Ganegoda, A 2007, ‘Explaining the Demand for Life Annuities in the Australian Market’, Centre for pensions and superannuation, discussion paper, May 2007.

17 Estimate provided to the Financial System Inquiry by the Australian Government Actuary, 11 June 2014.

18 Commonwealth of Australia 2009, Australian Life Tables 2005–07, Canberra. Using 25-year mortality improvement factors.

19 As discussed earlier, at least 94 per cent of pension assets are in account-based pensions and most drawdowns are at minimum rates.

20 National Seniors Australia and Challenger 2013, Retirees’ needs and their (in)tolerance for risk, National Seniors Australia, Brisbane.

21 Investment Trends 2013, December 2013 Retirement Income Report. Note: Based on a survey of 5,730 Australians aged 40+.

22 APRA 2013, Annual Superannuation Bulletin, June 2013 (revised 5 February 2014).

23 Rothman, G and Wang, H 2013, ‘Retirement income decisions: Take up and use of Australian lump sums and income streams’, paper presented at the 21st Colloquium of Superannuation Researchers, Sydney, 9–10 July.

24 APRA 2013, Annual Superannuation Bulletin, June (revised 5 February 2014), APRA, Sydney. Lump sum benefit payments are benefit payments paid as a lump sum and includes (but is not limited to) retrenchment, redundancies, resignation and disability benefit payments. This item does not include lump sum rollovers or pension benefit payments. Pension benefit payments refer to benefits paid to members in the form of a pension and include complying pensions, allocated pensions and annuity payments.

25 Rothman, G and Wang, H 2013, ‘Retirement income decisions: Take up and use of Australian lump sums and income streams’, paper presented at the 21st Colloquium of Superannuation Researchers, Sydney, 9–10 July.

26 Butler, M, Peijnenburg, K and Staubli, S 2011 ‘How Much Do Means-Tested Benefits Reduce the Demand for Annuities?’ CESifo Working Paper Series No. 3493, Munich.

27 Hulley, H, McKibbin, R, Pedersen, A, and Thorp, S 2013, ‘Means-Tested Public Pensions, Portfolio Choice and Decumulation in Retirement’, Economic Record, Volume 89, Issue 284, pages 31–51.

28 For people with large balances, there are tax incentives to decumulate assets more slowly. Slower drawdown also reduces the risk of outliving assets. The interaction of these effects differs depending on individual circumstances.

29 Mercer 2014, First round submission to the Financial System Inquiry, page 11.

30 Commonwealth of Australia 2009, Australia’s future tax system — The retirement income system: Report on strategic issues, Canberra.

31 Bray, J 2013, ‘In the red and going grey? Wealth and debt as Australians approach Age Pension eligibility age and retirement’, Crawford School Research Paper No. 04/2013, Canberra.

32 Agnew, J, Bateman, H and Thorp, S 2012, ‘Work, money, lifestyle: Plans of Australian retirees’, ARC Centre of Excellence in Population Ageing Research Working Paper 2012/22.

33 Bateman, H, Eckert, C, Geweke, J, Iskhakov, F, Louviere, J, Satchell, S and Thorp, S 2013, ‘Disengagement: A Partial Solution to the Annuity Puzzle’, UNSW Australian School of Business Research Paper No. 2013ACTL10.

34 Squam Lake Working Group on Financial Regulation 2009, ‘Working paper on the regulation of retirement saving’, Council on Foreign Relations Press, New York.

35 When asked about their worries in retirement, two of the three most common responses of retirees were “outliving retirement savings” and “falls in financial markets”. Source: Investment Trends 2013, December 2013 Retirement Income Report. Note: Based on a survey of 5,730 Australians aged 40+.

36 The seminal paper showing that rational retirees should annuitise their benefits was Yaari, M 1965, ‘Uncertain Lifetime, Life Insurance, and the Theory of the Consumer’, The Review of Economic Studies, Vol. 32, No. 2, April 1965, pages 137-150.

37 Recently, Australian research found that longevity-protected products form part of the optimal portfolios for retirees. Hanewald, K, Piggott, J and Sherris, M 2013, ‘Individual post-retirement longevity risk management under systematic mortality risk’, Insurance: Mathematics and Economics, vol 2013, Elsevier, London, pages 87–97.

38 Challenger 2014, data provided to Financial System Inquiry, 4 June 2014.