The Optimal Use of Capital
The level of capital needed to support an income stream is going
to depend on the length of life of the pensioner and their spouse,
the level of earnings achieved and the chosen income stream level
/ increase profile. Let's examine each of these.
It is common in Australia to allow for the probability of dying
in accordance with the population tables published by the Australian
Government Actuary. The most recent table is Australian Life Table
2000-2. The main table is a snapshot of mortality of Australians
over the three year period centred on the 2001 Census. The Australian
Government Actuary has also provided information on mortality improvement
trends.
We strongly recommend using in calculations a methodology that
allows for mortality improvements. This is far more likely to be
a better estimate of anticipated outcomes than the snapshot view.
The consistency of past improvements and the likelihood of further
improvements can be gauged from the following examples of life table
estimates over the past century.
Further years of anticipated life:
| Table |
Male At 0yrs |
Male at 65 yrs |
Female at 0 yrs |
Female at 65yrs |
| 1901-10 | 55.20 | 11.31 | 58.84 | 12.88 |
| 1920-22 | 59.15 | 12.01 | 63.31 | 13.60 |
| 1932-34 | 63.48 | 12.40 | 67.14 | 14.15 |
| 1946-48 | 66.07 | 12.25 | 70.63 | 14.44 |
| 1960-62 | 67.92 | 12.47 | 74.18 | 15.68 |
| 1970-72 | 68.10 | 12.37 | 74.80 | 16.09 |
| 1980-82 | 71.23 | 13.80 | 78.27 | 18.00 |
| 1990-92 | 74.32 | 15.41 | 80.39 | 19.26 |
| 2000-02 | 77.64 | 17.70 | 82.87 | 21.15 |
| With 25yr Trend in 2001 | 92.90 | 19.93 | 95.70 | 23.52 |
We have a calculator to help you with this aspect of deciding for how many years the capital of the pensioner/s will need to last.

A male aged 61 with a spouse aged 58 has a 9.80% chance of reaching
age 100. However, there is a 24.4% probability his wife will be
alive and a 31.8% chance that at least one will be alive.
The government and its legislation place too much emphasis on the
survival probabilities of the younger party and not sufficient on
at least one of them being alive. An older male is not likely to
outlive a younger female spouse - but he might. This contingency
requires an extra capital provision over that of the younger life's
capital requirements.
None of the account based pension calculators below are prescriptive
in the economic assumptions - they are generic tools - you control the
values chosen. However, for our statutory certificate
work on defined pensions, these are the economic assumptions I and
my colleagues are currently using:
| Strategy Description |
%Growth Assets |
Nominal Rate |
CPI |
Real Rate of Return |
70% Margin Allowance |
| Stable | 0-20% | 5.5% | 2.5% | 3.0% | 0.25% |
| Conservative | 20-40% | 6.0% | 2.5% | 3.5% | 0.50% |
| Balanced | 40-60% | 7.0% | 2.5% | 4.5% | 0.75% |
| Growth | 60-80% | 7.5% | 2.5% | 5.0% | 1.00% |
| High Growth | 80-100% | 8.5% | 2.5% | 6.0% | 1.50% |
These rates are net of an expense allowance.
For more information see the 2007/8
SMSF Computational Support Pack.
Account Based Income Streams
Currently we have 6 types of income streams available -
two account based forms (allocated pensions and market linked pensions)
and four defined designs (lifetime complying, life expectancy, flexi
lifetime and flexi term).
The account based income streams are the most straightforward pensions
and may be the only ones available after 30 June 2005. I feel from
the simplicity of design there needs to be a good reason to choose
a defined pension since they may be less tax efficient, require
actuarial certificates annually and have greater complexity eg treatment
of reserves on death.
However some of the reasons why defined pensions are useful include
total control over income stream profile and RBL compression. Slower
income emergence may have tax advantages and not all people see
the requirements for reserves that probably will not be needed as
a disadvantage - it's a resource that can be used for
dependants.
On our web site you will find a calculator that will help with
account based income streams. However, lets in general terms examine
the issue of a suitable design.
With the minimum allocated pension drawdown - capital lasts
an appropriate length of time - but has an inappropriate income
drawdown shape. The level of income in the latter years is very
low.
With the maximum drawdown the shape of the income stream is wrong
and the capital is depleted too quickly.
If we change to a targeted amount between the minimum and maximum
specified then we can create an appropriate income profile -
but the capital is probably still used up too quickly.
Hopefully one of the changes to come out of the current Treasury
pension review will be altered pension minimum and maximum factors
that overcome this problem.
Allocated pensions will not give us access to the pension RBL.
Can these problems be solved with the market linked pension - also known
as a TAP or growth pension?
The pensioner has no control over the individual year's pension drawings -
they will be a fixed amount and could jump around is the market value of
assets are volatile. The first worked example will show what happens to
the level of pension payments if all the monies are placed in a market-linked
contract. It is possible, for example, that the pensioner may prefer a level
of pension increase lower than the inflation rate to either take lifestyle
needs into account and to make the assets last longer than possible under
the market-liked design. It is logical to allocate more resources to the
early active years of retirement rather than the later years. The second
example shows how this can be improved if the flexibility of a combination
of allocated and a growth pension is chosen.
Consider the following situation for a 100% of assets in a market linked pension:
| Name: | A. Stute |
| Date of Birth: | 9/10/1942 |
| Sex | Male |
| Pension Commencement Date: | 1/1/2005 |
| Account Balance: | $595,000 |
| Reversionary | Yes |
| Spouse Date of Birth | 1/8/1945 |
| Assumed Earning Rate: | 7.5% |

The use of a market linked pension on its own causes the capital to run out too quickly.
The use of an allocated pension in conjunction with the market linked pension will
allow a few more years to be funded before the money runs out. This together with
a rollover along the way to further extend the term (assuming it doesn't generate
an RBL creep issue), may be a solution in many cases.

In conclusion we often can produce reasonably appropriate outcomes. They may be
a little shorter duration of capital than is desirable, but they
can be reasonable solutions, especially if some AP assets are held
back to extend the term. However, they certainly don't allow the
same control over capital usage as a defined pension. If the government
does not allow defined pensions in SMSFs after 30 June 2005, hopefully
they will fix the current problems with account based pensions.
Defined Benefit Income Streams
These pensions:
- 1.06(2) Life Time Complying
- 1.06(6) Flexi Life and Flexi Term; and
- 1.06(7) Life Expectancy Pension
may not be available after 30 June 2005.
The main reasons why they may be considered are:
- Direct control over capital usage and shape of the income streams
- To achieve RBL compression
- Smooth run of payments
- Possible tax deferral reasons on large amounts where marginal
tax rates outweigh the drag of taxation on the reserves.
They have disadvantages. They require annual actuarial certificates
for adequacy and tax exemption percentage calculation. Since 15%
tax is paid on reserves and surplus they can, but not always, be
slightly less tax effective. They will tend to leave assets in reserves
on death, which is messy.
The future is uncertain. The best estimate value of pension liabilities
can be calculated as the product of:
- A projection of the payment, e.g. pension plus anticipated
yearly increases
- Probability of occurrence, e.g. probability of the pensioner
being alive, or at least one of the pensioners being alive or a
particular pensioner being alive, etc.
- Discount factor - to bring back to a present value
at the calculation date.
Actuarial mathematics has developed an elegant set of notations
and approaches to work out these "commutation factors".
Since we are dealing with stochastic concepts - there are
ranges of possible answers with associated probabilities -
some answers are plausible and some are not. Different actuaries
will choose assumptions which will result in different best estimates.
The criticism of the lack of a single answer disappears when one
realizes that what is required is an understanding of the variability
of the possible outcomes. This is important for ongoing management
of the arrangement. Actuarial assumptions don't change the
real world - pension cheques, expenses, investment earnings,
etc do that. What they do is provide a monitoring system. The assumptions
need to be realistic because at some stage the real world will demand
recognition about what is the true state of affairs.
Calculating 70% adequacy answers are more difficult. With a pensioner
and a reversionary pensioner, thousands of different age combinations
at death could occur. One way of producing the results is a simulation
projection program. Random numbers are generated to simulate the
operation of the arrangements and the program is run tens of thousands
of times or even millions of times. The resulting empirical distribution
of outcomes are then tabulated and observed. Very complex arrangements
can be allowed for. A self managed super fund may have two pensions,
but the two pensions are not statistically independent. If the first
pensioner dies, the reversion on the second pension may disappear.
Offsets between longevity risks and investment risks can be allowed
for. Simulation programs also have disadvantages. The programs can
take a long time to run. You can run the program twice on the same
data and get slightly different answers. In order to get closer
answers each time, it is necessary to increase the number of iterations
- and then the results take even longer to run. In the sample
calculators we make available, we have adopted an alternative approach.
By dividing outcomes into one year cells, even when you have two
lives to consider, the maximum number of cells to calculate is no
more than 6 - 10,000. That can be done in the blink of an
eye and even take different pension reversion levels into account.
A quick but slightly conservative result is generated.
Our website has calculators which help with best estimate and APRA adequacy
reserve calculations.
For more information please see the 2007/8
SMSF Computational Support Pack.
RBL Strategies
You will notice that the web site has two calculators that will help you with
RBL compression exercises. The first is the RBL capital value of a lifetime pension.
This is currently written in Visual Basic. We find advisors prefer Excel applications.
They are easier to load.
The current RBL application will look as follows:
From the above answer a reduction can be made for undeducted contributions,
post 1994 invalidity and concessional component.
Enter Name, Date of Birth, Pension Start Date and Pension amount p.a.
If the pension is a single life, click the "Below 50%" option in the Level
of Reversion section. A 100% reversion is the other usual choice. Notice
that due to these factors being constructed in a time of high inflation, a
very different answer will result from the use of an indexation/pension
increase rate of CPI versus 3% p.a. fixed.
Consider for example a member of a self managed superannuation fund with a
RBL problem - assets are more than the pension RBL as follows:
| Name: | Mr. A. Sample |
| Date of Birth: | 7 October 1944 |
| Marital Status: | Married |
| Spouse Date of Birth: | 7 July 1946 |
| Total Assets: | $1,450,000 |
| Undeducted Contributions: | $150,000 |
| Benefits Previously Taken: | Nil |
| Transitional LS RBL: | Not Applicable |
| Transitional pension RBL: | Not Applicable |
| Investment Strategy: | Growth |
| Pension Commencement: | 1/8/2004 |
We have to ensure that in order to access the pension RBL, that
the value on the RBL basis of the complying pension is at least
half the total RBL value. The problem is that the RBL is different
to the value of assets allocated. A suggested approach is as follows:
- Choose an RBL value a little less than half the assets less the undeducted
contributions with an upper maximum of half $1,238,440 in 2004/5.
- Work out from the RBL Calculator what level of initial pension this is,
for the chosen increase and reversionary attributes.
- Work out from the lifetime complying pension calculator what assets are
needed to satisfy the high probability of adequacy requirement.
- Calculate what remains in the allocated pension component which normally
also included the undeducted contribution.
- Check that the RBL value of the complying lifetime is greater than allocated
pension RBL value. Repeat steps above until desired balance is achieved.
- Work out range of income that can be generated from the allocated
pension assets so that total income range can be presented to
the client.
In essence the task is to work the information shown below in an anticlockwise manner.
The pension amounts are annual before the pro-rata in the first year.
Option: Lifetime, 100% reversion, CPI pension increases:
| |
Lifetime |
Allocated |
| Initial Pension Amount | $34,500 | $40,380 to $80,770 |
| RBL PVF | 17 | - |
| RBL Value | $586,500 | $585,000 |
| Best Estimate Value (Tax) | $571,775 | $735,000 |
| Plus Adequacy Reserves | $124,049 | - |
| Plus Surplus | $19,176 | - |
| Account Value | $715,000 | $735,000 |
The undeducted contributions have been included as part of the
allocated pension.
It would be necessary to look at other combinations of reversions and increases.
A strategy using flexi or 1.06(6) pensions can be used when amounts are only a little
above the lump sum limit and compression to the lower limit is desirable to allow some
commutation rights. Commutations of flexi pensions have a variety of disadvantages and
constraints not examined here.
For more information please see the 2007/8
SMSF Computational Support Pack.
Age Pension Optimisation
The calculations required here include how the 50% "asset
test exempt" value provisions are assessed at outset and over
the term of the income stream, how income streams are treated for
asset deprivation issues and the optimization of the income/asset
tests.
Consider the following example:
| Name: | John Sample |
| Date of Birth: | 1/5/1939 |
| Sex: | Male |
| Spouse Name: | Jill Sample |
| Spouse Date of Birth: | 1/6/1939 |
| Date of Calculation: | 01/10/2004 |
| Total ETP Components: | |
| Undeducted Contributions | $205,000 |
| Post 83 Component | $305,000 |
| Additional Financial Assets: | $70,000 |
| Non Financial Assets: | $20,000 |
| Home Owner (Y/N) | Yes |
| Gifted in past 5 years | Nil |
This couple has $600,000 in assets and so would not be entitled
to an age pension. All SMSF assets are in John's name.
The client prefers (say) a mixture of allocated and market linked
pensions are desired.
This couple can't get the maximum pension because even if
the full SMSF assets are used for a market linked pension, the asset
test would be:
| 50% of Market Linked | $255,000 |
| Plus Financial Assets | $70,000 |
| Plus Non Financial Assets | $20,000 |
| Total | $345,000 |
This is $127,500 above the full pension allowance i.e. $3 x $127.50
= $382.50 p.f. reduction in the pension payment to be received from
Centrelink. Please note these limits are probably now out of date.
For the income test, if the Samples choose the longest duration
i.e. 26 years, then the market link pension will be $22,580 in the
first year. The amount of income that will be counted is:
| p.a. | p.f. (26.07) |
| 22,580 - 510,000 / 26 = | 2,964.62 | 113.72 |
The assessed income is:
| Market Linked | $113.72 p.f. |
| Financial Assets | $106.33 p.f. |
| Total | $220.05 p.f. |
This is $4.05 above the full rate i.e. pension is reduced by approximately
$0.80 p.f. The asset test dominates - the level of age pension
that will be paid is $775.20 - $382.50 = $392.70 p.f.
The focus is often to work out what the age pension entitlement will be in the
immediate future. These SMSF pensions result in the use of capital over a period
of time. Consequently while no immediate entitlement may exist at outset, future
entitlement could be predicted under current rules and a few assumptions. The
other aspect is that the optimal structure may be different to that of the immediate view.
We have commenced work on a calculator to investigate this issue. Please watch our
website for more information.
For more information please see the 2007/8
SMSF Computational Support Pack.
|