Towards a smarter default Account-Based Pension

The Actuaries Institute’s Rule of Thumb Working Group produced a paper ‘Spend your decennial age’ for the Actuaries Summit, in which the authors proposed the following withdrawal rates from the beginning of year account-based pension (‘ABP’) balance for Australians entitled to receive the full or part age pensions (subject to the statutory minimum):

“Take the first digit (say x) of your age: withdraw x% from your ABP, unless your asset testable assets are within the Asset Testing Range, in which case withdraw (x+2)%.”

These withdrawal rates were derived from utility-optimising criteria and a series of simplifying assumptions.

Australian Super has issued a ‘Smart Default’ ABP, from which the default withdrawal rate is 6% below age 80.  All credit to Aussie Super for producing a product where the default is higher than the statutory minimum.  The Rule of Thumb paper reverse engineered the statutory minimum basis and found that the minimum withdrawal rates were consistent with the inverse of term annuity values assuming those values were calculated at 2.5% real interest and using a term certain equal to the 95th survival percentile based on ALT 2010-12.  That basis seems fine for a statutory minimum, but withdrawals on that basis should not be expected to be ‘optimal’.

The rule of thumb developed by the Working Group was designed to be simple and easily understood and remembered by ABP pensioners.  But the point of this article is to suggest that ABP providers could offer some drawdown guidance other than the statutory minimum through a slightly tweaked ‘default’ withdrawal rate, using a more complex approach with the objective of producing an even better member outcome.

Those tweaks might include the following.

1 – Age banding


The simple version of the ‘Rule of Thumb’ implies a 17% increase in the amount of pension payable when a 69 year old reaches age 70 (assuming the beginning of year balance of their ABP is the same).  It seems preferable to adopt a rule where there are no significant inherent discontinuities in the drawdown amount.  If the provider were doing the calculations, instead of using the ‘first digit’ x of someone’s age, the ‘default’ ABP drawdown rate could be {(age-5)/10}%.  Effectively over time, the total drawdowns will be (approximately) the same, but without the discontinuity when the account holder reaches ‘big 0’ ages.  Alternatively, providers could go further by using the most detailed and comprehensive drawdown rule in Appendix A of the Working Group’s paper, given that they have all the information they need to perform the calculation.

2 – Asset banding

The adjustment in the above rule of thumb that allows for optimising age pension entitlements is a rational response to the current age pension means tests, but a drawdown rule that means someone whose assets are $1 more than the lower Asset Testing Threshold (or $1 less than the Upper Asset Testing Threshold) will draw down MUCH more than someone with $1 less than the Lower Asset Testing Threshold (or $1 more than the Upper Threshold) is not ideal.  But Appendix A of the paper quotes the optimal drawdown rates for asset bands of $10,000.  So a provider could tap into those more detailed optimal rates to produce an ABP with default drawdown rates that are appropriate to the actual means testable assets held by the ABP holder (though this would require obtaining information from the ABP holder about their other means testable assets).

3 – Market value smoothing

Finally, the Rule of Thumb paper assumed that the drawdown would be strictly proportional to the beginning of year balance of the ABP.  However, pensioners will not favour sharp changes in drawdown payments (especially sharp decreases).  This may not be a huge issue if the asset allocation backing the ABP is not very volatile but is likely to be more important where more aggressive asset allocations have been adopted.  There are ways of smoothing the drawdown to mitigate the effect of choppy markets.  For example, an ABP provider could follow the example of the Yale Endowment fund, and calculate the appropriate drawdown for a given year as a weighted average of:

  1. the previous year’s drawdown, indexed to inflation; and
  2. the appropriate percentage of the beginning of year value of the fund.


The Yale fund applies an 80% weight to the previous year’s drawdown and a 20% weight to the calculated drawdown based on the current market value of the fund.

With the above smoothing approach, will the resulting drawdowns as suggested by the Rule of Thumb Working Group provide a constant level of real income?  Sadly, no.  However, the aim is not to provide a constant real income, but to adjust to changing circumstances to provide guidance on appropriate levels of drawdown over time.

I have looked at the history of real returns on the Australian All Ordinaries Index, and find that historically, the worst ever beginning of year retirement date was 1 January 1970.  If a retiree had retired on that date, and taken the recommended ‘rule of thumb’ drawdowns from their ABP, then the terrible real returns in the early 1970s would have had such an impact that in 1975, the value of the retiree’s ABP drawdown would have been only one third of what the retiree drew in the first year.

However, noting that the rule of thumb was intended for those entitled to full or part age pension, the apparently diabolical reduction in the real value of the ABP would have represented (even for the most seriously impacted asset level) only 25% of the retiree’s first year income (allowing for the age pension as well as the ABP). 

So the worst thing that has ever happened in Australian history to someone notionally adopting the rule of thumb guidance is that if they had held an asset allocation MUCH more aggressive than most retirees adopt, their total spendable income in the single worst projection year would have fallen by a quarter.  It seems highly likely that this downside impact would have been markedly less for a retiree with a more balanced asset allocation.

So when will we see smarter Account-Based Pensions?

CPD: Actuaries Institute Members can claim two CPD points for every hour of reading articles on Actuaries Digital.