Volatility in Superannuation Investments and the Australian Age Pension

1
Volatility in Superannuation Investments and
the Australian Age Pension

• Clare Bellis
• Department of Actuarial Studies,
• Macquarie University
• cbellis_at_efs.mq.edu.au

2

• How may changes in superannuation investment
  strategies affect the proportions of retired
  Australians who receive a part or full age
  pension?

3
Before most members in one strategy Now
choice of strategy, eg cash or
growth Future member choice of fund ?
specialised investment vehicles ? very low level
of diversification possible, ? increased
variation across members
4
The interaction with the Age Pension

• Does the means test give members a put option on
  their superannuation assets
• keep the upside, but collect the age pension if
  the downside occurs?
• It depends on post-retirement income.

5
2 extremes

• People on full age pension, can only lose pension
  if their investments soar
• ? the state should encourage low income earners
  to gamble with their super.
• People on no age pension, cant lose pension but
  can get it if their investments flop
• ? the state should discourage wealthy from
  gambling with their super.

6
Projection modelAssume

• 30 year-old single male
• 35 years of unbroken employment
• SGC 9, taxed at 15
• Salary 50,000 or 100,000 in 2005 dollars.
• Ignore insurance and administration costs.
• Converted into a single life CPI-indexed pension
  at 65 at 17.40 per 1 pa

7
Projection modelAssume

• Age pension 12,535 pa less 40 of income between
  3,172 and 34,508.
• No change in the social security legislation!
• Salary, age pension and the means test limits
  increase by 4 pa.
• Investment returns are independent log normal

8
Base case (balanced)

• 6.5 mean
• 7 standard deviation
• for net investment return.

9
Base CaseResults for a salary of 50,000

• 90 percentile range
• without age pension 16 to 36 of final
  salary.
• including age pension 37 to 49 of final
  salary
• All members receive a part age pension, averaging
  71 of the full pension.

10
Split members into defensive and growth funds

• Suppose
• balanced portfolio 50 defensive 50 well
  diversified growth assets
• net investment returns
• defensive 5 mean/ 1 st dev
• well diversified growth 8 / 13.8
• Members divide into defensive or growth investors.

11
Results for a salary of 50,000, defensive
investors

• 90 percentile range, including age pension 38
  to 39 of final salary
• ? Very little uncertainty, but gave up 10
  upside to reduce downside by 1
• All members receive a part age pension,
  averaging 81 of full pension
• (up from 71 for base case).

12
Results for a salary of 50,000, growth investors

• 90 percentile range, including age pension 35
  to 68 of final salary
• 96 of members receive a part age pension, (and
  4 get none), averaging 60 of full pension
• (down from 71 for base case)

13
Result of dividing the investorson salary of
50,000

• State now pays out 2 different rates to 2 groups
• Combined ave. pension 71, as before (defensive
  investors get more, growth investors get less, on
  average)

14
Base CaseResults for a salary of 100,000

• 90 percentile range including
• age pension 23 to 36 of final salary
• 94 of members receive a part age pension,
  averaging 34 of the full pension.

15
Results for a salary of 100,000, defensive
investors

• 90 percentile range, including age pension 17
  to 19 of final salary
• All members receive a part age pension,
  averaging 51 of full pension
• (up from 34).

16
Results for a salary of 100,000, growth
investors

• 90 percentile range, including age pension 22
  to 66 of final salary
• 65 of members receive a part age pension,
  averaging 25 of the full pension (down from
  34).

17
Result of dividing the investorson salary of
100,000

• State now pays out combined ave. pension 38,
  up 4
• (extra cost for defensive investors not fully
  offset by savings on growth investors)

18
Results of choosing a less diversified growth
portfolio

• Assume 8 mean (as before) but 20 s.d.
• Compared to diversified growth, members gain
  about 16 upside for 3 extra downside (means
  test does not affect upside, dampens downside by
  40) attractive!

19
Results of choosing a less diversified growth
portfolio

• average age pension
• for 50,000, 63 (up 3 on diversified growth)
• ? 72 combined average (up 1 on base)
• for 100,000, 35 (up 10 on diversified
  growth, up1 on base )
• ? 43 combined average (up 9 on base)

20
Results of choosing a poorly performing high risk
portfolio

• Assume 5 mean,
• 20 st dev (as before).
• Not attractive, but quite possible!!

21
Results of choosing a poorly performing high risk
portfolio

• average age pension
• for 50,000, 81 (up 21 on diversified growth,
  up 10 on base)
• ? 81 combined average (up 10 on base)
• for 100,000, 57 (up 32 on diversified
  growth, up 23 on base)
• ? 54 combined average (up 20 on base)

22
Summary of Age pension
50,000 salary 100,000 salary
Base (balanced) 71 34
Half defensive, half growth 71 38
Half defensive, half less-diversified growth 72 43
Half defensive, half poor performing growth 81 54
All poor performing growth 81 57
All defensive 81 51
23
Conclusion

• The means test on the age pension does create a
  put option (on our assumptions).
• Cost to the state increases as choices become
  less efficient OR overly defensive.

24
Areas for further thought

• Allow for tax - could help the states position,
  though tax payable would be low.
• Try a more sophisticated economic model.
• Consider allocated pension instead of lifetime
  annuity (which cuts off volatility at age 65).