Schemes designed to deal with longevity risk


Australia’s large super funds are building better products to provide income in retirement for their members. In part, this reflects policy initiatives such as innovative income streams, but some funds are actively considering their retirement offer ahead of the potential requirement to offer each member a CIPR (Comprehensive Income Product for Retirement).

The key element of a CIPR is to manage longevity risk. This can’t be done if the only option is an account-based pension (ABP), which the majority of superannuation pensions are currently based on. While a partial investment in an annuity can provide the longevity risk management, there are other options for funds to use a collective income stream alongside the ABP.

What exactly is a collective income stream?

In short, a collective income scheme is one in which:

  • members have no individual account (i.e. ownership of capital) in the scheme.
  • the liability of the employer sponsor(s) to contribute is both certain and limited.
  • there is a retirement income target, but no concrete promise (this of course could be made more secure (e.g. by derivatives) or guaranteed by a third party, but without recourse to the sponsors).
  • longevity risk is spread across the pool.
  • investment risk is spread across the pool.

These schemes are sometimes called ‘group self-annuitisation schemes’ (or GSAs) but the definitions have blurred since GSAs were first described by some Australian academics. There are key differences between the various collective schemes in their degree of flexibility and the approach to managing retiree risks. These factors include:

Flexibility Risk Management
– Entry point (pre/post retirement)
– Choice and ability to change
– Access to capital
– Payment of residual capital (estate)
– Timing of contributions
– Timing of payments
– Mortality pooling
– Market risk protections
– Diversification (asset allocation)
– Guarantees and capital protection
– Inflation protection

There are three key benefits from using a GSA (or other collective income schemes):

1. Pooling idiosyncratic longevity risk

There are two forms of longevity risk. One risk is related to how long everyone will live, and will change with medical improvements and lifestyle changes etc. (systematic longevity risk). The other risk is that some people will live considerably longer than the average (idiosyncratic longevity risk).

GSAs pool idiosyncratic longevity risk. Pools of retirees (in the same age cohort) tend to have a more reliable distribution of ages at death, particularly as the pool becomes larger. When planning for 10,000 retirees, the law of large numbers will start to see quite a predictable distribution of lifespans around the mean and hence the risk is effectively diversified away.

Because it has no resources beyond what is in the pool, a GSA arrangement is still exposed to systematic longevity risk. This form of risk, if it unfolds, will be borne directly by the GSA-funded retiree in the form of a reduced income.

2. Mortality credits

A mortality credit is the higher payment that is available to someone who contributes their capital to a longevity pool, where participants are only entitled to payments while they are alive. Those who live beyond the actuarial life expectancy of the pool benefit from the contributions of those who die earlier.

Leading annuity expert, Moshe A. Milevsky (2006), describes it as a process where the capital and interest of the deceased member is ‘lost’ to that person and their beneficiaries. It is then ‘gained’ by the surviving members of the pool. The remaining value of the notional capital of the deceased is spread across all members to help support their lifetime income payments.

As the life expectancy is an average, approximately half of the members of the pool will die before reaching the expected average and will not benefit further. The remaining value of their notional capital is then available to support the remaining liabilities in the pool. These mortality credits are distributed ex-ante by the scheme in setting its targeted payment rates. In other words, mortality credits enable the income paid to the member to be higher than the combined total of the partial return of capital and projected asset returns of the scheme comprised in each payment.

Mortality credits provide a form of return not directly linked to the capital markets.

3. Reduced (or no) capital costs

Pooling of longevity risk (both idiosyncratic and systematic) is available through a lifetime annuity offered by a life insurance company. These products also remove the market risks from the retiree and pay a guaranteed income. In order to secure these payments, the shareholders of the life insurance company provide capital as a buffer to protect the retiree. This capital is at risk to the shareholders and needs a sufficient return. The guaranteed payments to the annuitant are set so that what remains from the returns on the total asset pool provides the expected return to shareholders. If these expectations turn out to be wrong, the losses are borne by the shareholders, who, in the worst case, would be called on to provide even more capital under powers given to APRA in 2012.

The logic for GSAs is that by not using capital buffers or guarantees, they will be able to avoid the cost of the capital or the insurance afforded by the guarantee and thereby increase the retirement income able to be distributed to members. The flip side of this argument is that a guarantee has a value in the defensive or ‘safety-first’ part of the portfolio and not having a guarantee is a weakness, rather than a strength.

The development of better retirement outcomes for Australians is likely to see growing use of GSAs and other collective income streams. This will require solutions to some of the more technical aspects, such as operating a GSA over risky assets with a need for surplus/deficit management or highly volatile income streams.

There is also a regulatory concern over the disclosure of the GSA target. Without a guarantee, there can be no real promise of income in retirement. How will retirees be able to distinguish between alternative structures that might target different incomes from the same asset mix? At least with a guaranteed product, the income can be relied on. The additional capital backing the promise provides this security for the retiree.

There is no magic pudding in retirement. A GSA scheme can share investment risk between one member or generation of retirees and another, but it can’t reduce it overall. If one GSA member takes less investment risk, another member is taking more. Pooling does reduce the idiosyncratic mortality risk, but as with idiosyncratic market risk under the capital asset pricing model, this is an unrewarded risk. Removing it alone does not increase total returns to the pool.


Jeremy Cooper is Chairman, Retirement Income, at Challenger, a sponsor of Cuffelinks. For more articles and papers from Challenger, please click here.

Print Friendly, PDF & Email

, , , ,

5 Responses to Schemes designed to deal with longevity risk

  1. David Bell November 29, 2018 at 3:51 PM #

    Jeremy – it is a credit to your integrity and desire for a better retirement system that you present such a balanced overview of collective income schemes.

    For the benefit of some of the other respondents, there is daylight between these schemes and the original tontine schemes – the one area of commonality is that those who live longer do better overall (the detail of how that occurs is quite different across the two structures).

    The design of collective income schemes can vary greatly, as you have rightly pointed out. I personally see benefit in exploring the concept of DGSA’s (deferred group self-annuitisation). They do not require a large portion of member’s savings at retirement, and don’t over-insure against longevity risk given the existence of the Age Pension. A deferred solution, whether guaranteed or not, is something that fund trustees should consider on behalf of their members, especially if they take a whole-of-life view and consider the potential for cognitive decline at later ages.

    Cheers, Dave

  2. James November 26, 2018 at 5:44 PM #

    Isn’t this just describing a Tontine or a variation?

    • SMSF Trustee November 26, 2018 at 8:24 PM #

      Are you trying to start a pillow fight, James?

      Yes, it is a tontine. There’s not that much that’s new under the sun and retirement income ideas have been a subject that many minds have grappled with for centuries.

      Getting Australians interested will be the challenge.

  3. Albert Uo November 25, 2018 at 1:39 PM #

    It sounds to me that the GSA is almost identical to the current Govt. Pension Scheme. Might as well get the Govt to run this eg. the S’pore EPF scheme

  4. Frank November 23, 2018 at 8:35 AM #

    Good to see developments in this area, as risk pooling seems the only solution to longevity risk when interest rates on long-term bonds are around 2%. In real terms, that’s going backwards, so you can’t buy a secure government bond stream to finance your latter years.

Leave a Comment:

This site uses Akismet to reduce spam. Learn how your comment data is processed.

Register for our free weekly newsletter

New registrations receive free copies of our special investment ebooks.