Appendix H. Financial implications

MILITARY SUPERANNUATION REVIEW - ACTUARIAL REPORT

TABLE OF CONTENTS

2 PROPOSED NEW ACCUMULATION ARRANGEMENT, PRIORITY CHANGES TO MSBS AND DFRDB AND TRANSFER TERMS FROM MSBS AND DFRDB 2

3 DATA 3

4 CURRENT POSITION 4

5 TRANSFER ASSUMPTIONS 6

6 IMMEDIATE IMPACTS OF THE TRANSFER OFFER 9

7 POSITION AFTER TRANSFER OFFER 11

8 COMMENTS ON THE PROJECTIONS INCORPORATING
PROPOSED ARRANGEMENTS 14

9 CLOSURE OF MSBS TO NEW MEMBERS WITH NO
TRANSFER OFFER 15

10 AWOTE INDEXATION OF DFRDB PENSIONS 17

11 SUPERANNUATION GUARANTEE ARRANGEMENTS –
ORDINARY TIME EARNINGS 19

INTRODUCTION

The Australian Government Actuary (AGA) has been asked to prepare a brief report on the costs associated with changes to military superannuation arrangements as recommended by the Military Superannuation Review Team (Review Team). This report has been requested by the Review Team for inclusion in their report putting forward their recommended changes.

The time allocated to the Review Team to consult and recommend changes was fairly tight. We have provided the Review Team with information and costings as required over the last three months to assist the Review Team in forming an in principle approach and then reaching a final position. This report does not seek to cover the full range of advice that was provided. Rather, it summarises most of the elements of our advice which are relevant to the final position reached by the review team. As costings have been provided over a period of time, the costs provided in this report do not tie in exactly with the final recommendations. However, the differences are minor and commentary has been provided on the extent of the differences.

I believe that the advice that we have provided to the Review Team is such that the Review Team can make properly informed decisions about their recommendations. However, it should be noted that given the fairly tight time frame allocated to the Review Team, the Review Team has concentrated on the major structural changes, benefits and conditions that would apply to the vast majority of Defence personnel. If a decision is made to proceed with some or all of the Review Team’s recommendations, many decisions on the detailed implementation of the recommendations will still need to be made. Examples include the treatment of individuals that have been subject to a Family Law superannuation split and the detailed design, implementation and pricing of the ‘actuarially fair’ pension option.

The costings were based on data as at 30 June 2006 and it was assumed that changes were implemented as at 30 June 2006. It is not known what date, if any, will be the implementation date for some, or all, of the recommended changes. However, given that a decision needs to be made on major structural changes followed by many decisions on detailed implementation which would then be followed by changes to legislation, changes to administrative systems, preparation of individual transfer offer material and associated communication material, I consider an implementation date of 1 January 2009 optimistic but achievable if everything goes smoothly. The main features disclosed by our costings would still be present if the implementation date were in 2009. We have included further commentary in the report on a later implementation date.

While this report will form part of the Review Team’s report, its presence should not be taken as our endorsement of the Review Team’s recommendations or otherwise. The recommendations of the Review Team are theirs and theirs alone. We know that the Review Team has consulted with many parties in forming their recommendations and these parties include the AGA.

PROPOSED NEW ACCUMULATION ARRANGEMENT, PRIORITY CHANGES TO MSBS AND DFRDB AND TRANSFER TERMS FROM MSBS AND DFRDB

- Proposed new Accumulation Arrangement

A brief summary of the main elements of the proposed new accumulation arrangement is set out below:

• Employer contributions

16% of superannuation salary for the first 6 years
23% of superannuation salary for the next 9 years
28% of superannuation salary thereafter.

Superannuation salary is the same salary definition as is currently used for MSBS.

In addition to the contributions referred to above, the employer:

• will cover the administration costs of the new accumulation arrangement;

• provide separate death and invalidity cover.

Members will be able to choose the level (if any) of their contributions. There will be a default contribution rate of 5% of superannuation salary for those who do not nominate a contribution level.

- Priority Changes to DFRDB and MSBS from the Superannuation Review

The Review Team recommended a central package of what it termed priority changes. This package does not include any changes to the DFRDB. There is, however, a lower ranking recommendation to improve indexation of DFRDB (including DFRB) pensions. This has not been included in our main costings but has been considered separately.

With the MSBS, there are no substantial priority changes to benefits and conditions. The first significant priority change contemplated is to abolish the MBLs. The costings assume this improvement applies retrospectively to serving members at the implementation date.

The second significant priority change is to the salary definition used for death, invalidity A and invalidity B benefits. The change is that these benefits will be calculated using final superannuation salary rather than final average superannuation salary. This change has not been included in any costings that incorporate changes. If this change went ahead there would be a very small increase in accrued (and unfunded) MSBS liabilities and an increase in the MSBS notional employer cost of about 0.2% to 0.3% of MSBS superannuation salaries. Given that few MSBS contributors are likely to remain after the transfer offer, the impact on the overall notional employer cost is likely to be small.

- Transfer Arrangements for MSBS and DFRDB

Non-pensioner members of DFRDB and MSBS would be offered a one off opportunity to transfer to the new accumulation arrangement.

A brief summary of the recommended transfer terms for MSBS members is set out below:

- Preserved Members

The transfer value would be the nominal (or face) value of the Preserved Benefit. This would include any member component and the employer component.

- Contributors

The transfer value for contributors would be the nominal (or face) value of the accrued benefits in MSBS. That is, it would be:

• Member account; plus

• Accrued employer multiple X Final Average Salary.

In brief, the conceptual approach taken for DFRDB contributors is to assume a ‘de facto’ transfer to MSBS in 1991 followed by a transfer to the accumulation arrangement. This results in a transfer value of:

• Member contributions plus interest at the 10 year Commonwealth Bond rate; plus

• Accrued MSBS multiple X Final Average Salary (FAS).

The accrued MSBS employer multiple would be calculated assuming the individual had transferred to the MSBS, that is, 18% for the first 7 years, 23% for the next 13 years and 28% thereafter.

Note that the terms would be different where special situations are involved, for example, for re-entered recipients or where benefits have been subject to a Family Law split.

DATA

The data used for the costings is data as at 30 June 2006. Rough checks against the data used for the Long Term Cost Report as at 30 June 2005, along with Defence payroll information and ComSuper pension payroll information at 30 June 2006 indicate that it is suitable for the current costings.

A summary of the data as at 30 June 2006 is set out below:

Contributors	Number	Superannuation Salary $m
MSBS	45,151	2,442
DFRDB	6,265	452
Preserved Members		Nominal Value of Benefits $m
MSBS	61,818	3,647
Pensioners		Annual Pension $m
MSBS	5,915	105
DFRDB	56,557	1,150

CURRENT POSITION

We have prepared projections to 2045 on the same assumptions as were used for the Long Term Cost Report as at 30 June 2005. The projected outlays and unfunded liabilities have been calculated in the same way as for the Long Term Cost Report as at 30 June 2005. In particular, we have assumed unaltered MBLs for MSBS. These projections assume that combined contributory membership of DFRDB and MSBS stays constant. That is, every contributor exit from DFRDB and MSBS is replaced by a new member of MSBS. A copy of the Long Term Cost Report as at 30 June 2005 can be accessed at www.aga.gov.au/publications.

The table below shows projected Commonwealth cash outlays and the notional employer cost for each year until 2009/10 and then for every fifth year thereafter. The notional employer cost is the cost as a percentage of total superannuation salaries of benefits accruing each year to serving Defence Force personnel.

PROJECTED COMMONWEALTH OUTLAYS AND NOTIONAL EMPLOYER COSTS (CURRENT ARRANGEMENTS)

Year	MSBS ($’m)	DFRDB ($’m)	Total ($’m)	Notional Employer Cost %
2006/07	264	1,314	1,578	25.6
2007/08	252	1,347	1,599	25.8
2008/09	276	1,384	1,659	25.7
2009/10	292	1,416	1,708	25.7
2014/15	469	1,532	2,001	25.2
2019/20	752	1,608	2,360	25.1
2024/25	1,226	1,636	2,861	25.1
2029/30	1,801	1,621	3,423	25.2
2034/35	2,576	1,577	4,153	25.3
2039/40	3,607	1,473	5,079	25.3
2044/45	4,769	1,304	6,074	25.4

* Note that components may not add up to total due to rounding.

It should be noted that the projections are in nominal dollars. That is, they have not been discounted to give a 2006 value.

The table below shows projected unfunded liabilities for each year to 2010 and every fifth year thereafter.

PROJECTED UNFUNDED LIABILITIES (CURRENT ARRANGEMENTS)

Year ending 30 June	MSBS ($’m)	DFRDB ($’m)	Total ($’m)
2006	9,637	23,623	33,261
2007	10,548	23,824	34,372
2008	11,569	23,997	35,566
2009	12,669	24,130	36,799
2010	13,864	24,223	38,088
2015	21,149	24,166	45,315
2020	30,877	23,305	54,182
2025	43,161	21,790	64,951
2030	58,057	19,701	77,758
2035	76,216	17,077	93,293
2040	97,645	14,019	111,664
2045	122,906	10,748	133,654

* Note that components may not add up to total due to rounding.

The projections of cash flows and unfunded liabilities are similar to those from the Long Term Cost Report as at 30 June 2005. This is to be expected.

TRANSFER ASSUMPTIONS

The take up rates of the transfer offer are subject to significant uncertainty. Past experience with the closure of DFRDB and its civilian equivalent, the CSS, where there were options to transfer to MSBS and its civilian equivalent, the PSS indicated that most individuals made sensible decisions about whether to transfer or remain. That is, most individuals made the decision that was to their advantage. However, it appears that there were some individuals who, with the benefit of hindsight, made decisions that were not to their eventual advantage. Some individuals also seem to have made poor decisions, possibly related to apathy or mistrust of the Commonwealth’s motives.

In setting transfer assumptions we have had regard to the following:

• MSBS pension option. The ability to convert the MSBS employer component into pension on very favourable terms is an argument for remaining in MSBS.

• Member contributions. The new accumulation arrangement does not require the payment of member contributions whereas MSBS and DFRDB do. If no member contributions are paid there would be a significant increase in take home pay and this is an argument for transferring.

• MSBS Preserved Benefit. The current MSBS preserved benefit is widely disliked because the majority of the benefit is only indexed in line with CPI and it cannot be transferred to another superannuation scheme. Given that most MSBS contributors leave well before age 55 and hence will be preserved beneficiaries for an extended period, this is likely to result in more transfers.

• Future investment returns. The estimated cost of the new accumulation arrangement is at a similar level to the assessed cost of the MSBS. The assessed cost of the MSBS was derived using an interest rate of 6% per annum. To the extent that future investment returns could be expected to be greater than 6% per annum, the expected benefits from the new accumulation arrangement could be expected to be higher. There is a reasonable expectation that future investment returns will be higher than 6% per annum and this will make transferring more likely to be attractive. The recent history of very good investment returns may play a part in individuals’ decisions.

• Death and invalidity cover. The new arrangements have a very different structure from the current arrangements. Existing contributors may have trouble understanding all the implications of the new arrangements and may treat them with suspicion. This makes transferring less attractive.

• Tax status. With the new ‘simpler super’ arrangements, benefits from the new arrangements will be tax free after age 60 and this may be attractive. However, it should be noted that the unfunded part of the transfer will be subject to 15% contribution tax on transfer.

• Inertia and suspicion of Commonwealth’s motives. These factors will be present.

• Comprehension and community norms. The accumulation part of the arrangement is easier to understand. Accumulation arrangements are now the community norm and the new arrangements are well in excess of the 9% standard employer contribution rate. This may make transferring more attractive.

• Different accrual patterns. The jumps in contribution rates occur earlier with the new accumulation arrangement than the corresponding jumps in the MSBS accrual rates. This may make transferring more attractive.

- Contributor Assumptions

The decision to transfer or remain will depend on individual circumstances and preferences. However, overall, I believe that the transfer offer will be very attractive to younger MSBS contributors. The transfer offer will be far less attractive to older contributors intending to take the MSBS pension.

I believe that the transfer option for DFRDB contributors will only be attractive in very limited circumstances and there will be a very low take up rate.

We have prepared costings on two different sets of assumptions for MSBS. The first assumes that decisions are mainly based on rational factors and that inertia and suspicion only plays a small part in the process. This results in a high level of transfer. These (high transfer) assumptions are set out below:

Age	Other Ranks			Officers
	Short %	Medium %	Long %	Short %	Medium %	Long %
To 24	90	90	90	90	90	90
25 – 29	90	90	90	90	90	90
30 – 34	90	90	90	90	90	90
35 – 39	85	85	75	75	75	50
40 – 44	60	60	50	40	40	40
45 – 49	40	40	40	25	25	25
50 or more	40	40	40	25	25	25

Short service is up to 10 years; medium service 10 to 18 years; long service is 18 years or more

Following the transfer offer any individual remaining in the MSBS is assumed to convert all the employer component into pension on eventual retirement at age 55.

The second set assumes that inertia and suspicion plays a greater role in the process. This results in a lower level of transfer. These (low transfer) assumptions are set out below:

Age	Other Ranks			Officers
	Short %	Medium %	Long %	Short %	Medium %	Long %
To 24	70	70	70	70	70	70
25 – 29	70	70	70	70	70	70
30 – 34	70	70	70	70	70	70
35 – 39	65	65	60	55	55	45
40 – 44	50	50	45	30	30	30
45 – 49	30	30	30	20	20	20
50 or more	30	30	30	20	20	20

Short service is up to 10 years; medium service 10 to 18 years; long service is 18 years or more

Following the transfer offer any individual remaining in the MSBS who was aged 40 or more at the transfer offer date is assumed to convert 90% of the employer component into pension on eventual retirement at age 55 and take the remainder in lump sum form. An individual remaining in the MSBS who was aged 39 or less at the transfer date is assumed to convert 60% of the employer component into pension.

- DFRDB Contributors

I believe that the transfer option will be unattractive to virtually all DFRDB contributors. We have assumed that only 5% of DFRDB contributors transfer. It definitely would not surprise me if this assumption turned out to be too high.

- MSBS Preserved Members

As with MSBS contributors, we have used two sets of assumptions. The first set is deliberately conservative, assuming sensible decision making, and results in high transfer levels which leads to higher costs. This set of assumptions is set out below:

Transfer Assumptions (High)
Age	Other Ranks %	Officers %
34 and under	100	100
35 – 39	100	100
40 – 44	100	100
45 – 49	30	20
50 and over	30	20

All of those that remain in the MSBS are assumed to take the employer component in pension form on retirement at age 55.

The second set of assumptions assumes considerable inertia and lower take up rates. These assumptions are set out below:

Transfer Assumptions (Low)
Age	Other Ranks %	Officers %
34 and under	65	65
35 – 39	60	45
40 – 44	45	30
45 – 49	30	20
50 and over	30	20

The assumed pension take up rates for the employer component for those who remain in MSBS are:

- Transfer age 39 or less 60% pension, 40% lump sum.
- Transfer age 40 or more 90% pension, 10% lump sum.

IMMEDIATE IMPACTS OF THE TRANSFER OFFER

If the transfer offer had taken place as at 30 June 2006, then, on the assumptions made, there would be a number of significant immediate impacts which are set out below:

- Membership

Membership numbers in MSBS and DFRDB would be expected to fall.

		High Transfer Assumptions		Low Transfer Assumptions
	Number at 30 June 2006	Transfer	Remain	Transfer	Remain
MSBS contributors	45,151	37,424	7,727	29,121	16,030
MSBS preserveds	61,818	57,878	3,940	35,448	26,370
DFRDB contributors	6,265	313	5,952	313	5,952

- Commonwealth Cash Expenditure

Significant Commonwealth cash expenditure would be required because the DFRDB and MSBS are largely unfunded. This is particularly the case in respect of the MSBS where transfer option will be attractive for many individuals.

	High Transfer Assumptions	Low Transfer Assumptions
	Unfunded Component $bn	Unfunded Component $bn
MSBS contributors	2.5	1.9
MSBS preserveds	2.5	1.5
DFRDB contributors	0.1	0.1
Total	5.1	3.5

Thus, immediate additional Commonwealth cash expenditure associated with the transfer is estimated to be of the order of $4bn to $5bn. Note that the Commonwealth would get a taxation receipt associated with this due to the 15% tax on the unfunded component of the transfer value when it is paid.

- Accrued (unfunded) Liability Impacts

The immediate impact on the accrued liability (and unfunded liability once a decision is made but before funding for the transfer values occurs) is as follows:

	High Transfer Assumptions	Low Transfer Assumptions
	Increase in Accrued Liability $m	Increase in Accrued Liability $m
MSBS contributors	-483	-448
MSBS preserveds	378	268
DFRDB contributors	-60	-60

Overall, there is a decrease in accrued liabilities of around $200m.

Some care needs to be taken in interpreting these figures. Other things being equal, it would be expected that if it is attractive to transfer then the benefit being paid on transfer would be higher than the actuarial value of the existing accrued benefit. That is, individuals would be expected to choose the option which benefits them most, which is normally the one that costs the employer most. Thus, the total liability would be expected to be higher after transfer. For MSBS preserved beneficiaries, this is the result. For MSBS contributors, however, the liability has reduced.

I regard part of the reduction shown for MSBS contributors as being a genuine feature. This is because many individuals will make sensible decisions to transfer based on factors other than actuarially assessed costs of the two alternatives.

I also regard part of the reduction as being what could be argued to be a change in ‘accounting treatment’. The current actuarial valuation method for the MSBS calculates accrued employer benefits on a pro rata basis. To explain this, consider an individual who currently has 5 year membership and a retirement benefit that is expected to be payable after 20 years membership. The MSBS employer multiple after 20 years is calculated as

7 x 18% + 13 x 23% = 4.25 x Final Average Salary

Using the pro rata approach the accrued amount is:

When calculating the transfer value, a scheme multiple accrual approach is used. Under this approach the transfer value for the employer component would be calculated as

5 x 18% = 0.90 x Final Average Salary

POSITION AFTER TRANSFER OFFER

We have prepared projections to 2045 following the changes.

The projected unfunded liabilities are set out below:

PROJECTED UNFUNDED ACCRUED MSBS AND DFRDB LIABILITIES (NEW ARRANGEMENTS)

	High Transfer Assumptions			Low Transfer Assumptions
Year ending 30 June	MSBS ($’m)	DFRDB ($’m)	Total ($’m)	MSBS ($’m)	DFRDB ($’m)	Total ($’m)
2006	4,523	23,417	27,940	6,042	23,417	29,460
2007	4,774	23,607	28,380	6,454	23,607	30,060
2008	5,043	23,769	28,812	6,891	23,769	30,661
2009	5,314	23,893	29,207	7,335	23,893	31,228
2010	5,589	23,978	29,567	7,790	23,978	31,768
2015	6,907	23,892	30,799	10,089	23,892	33,981
2020	8,014	23,019	31,033	12,312	23,019	35,330
2025	8,826	21,500	30,326	13,873	21,500	35,373
2030	9,275	19,412	28,687	14,764	19,412	34,176
2035	9,377	16,798	26,174	15,095	16,798	31,893
2040	9,050	13,762	22,811	14,568	13,762	28,330
2045	8,180	10,526	18,706	13,337	10,526	23,863

* Note that components may not add up to total due to rounding.

The projected Commonwealth cash flows are set out below for the high transfer assumptions. The notional employer cost as a percentage of total Defence superannuation salaries is also given.

PROJECTED COMMONWEALTH CASHFLOWS AND NOTIONAL EMPLOYER COSTS (NEW ARRANGEMENTS – TRANSFER OPTION) HIGH TRANSFER ASSUMPTIONS

Year	MSBS ($’m)	DFRDB ($’m)	New Accumulation ($’m)**	Total ($’m)	Notional Employer Cost %
2006/07	154*	1,306*	488	1,949*	26.6
2007/08	146	1,339	527	2,012	26.5
2008/09	153	1,374	571	2,097	26.3
2009/10	159	1,405	616	2,181	26.1
2014/15	222	1,518	859	2,599	25.3
2019/20	306	1,592	1,125	3,024	25.1
2024/25	401	1,619	1,410	3,430	25.1
2029/30	491	1,604	1,733	3,828	25.1
2034/35	564	1,558	2,116	4,238	25.1
2039/40	638	1,452	2,572	4,663	25.0
2044/45	697	1,283	3,130	5,110	25.0

* 2006/07 figures do not include transfer values of $5.1 billion.

** includes an allowance of 3.8% of superannuation salary for new accumulation members for death and invalidity cover. (MSBS and DFRDB figures include death and invalidity benefits.)

Note that components may not add up to total due to rounding.

The projected Commonwealth cash flows are set out below for the low transfer assumptions together with the notional employer cost as a percentage of total Defence superannuation salaries.

PROJECTED COMMONWEALTH CASHFLOWS AND NOTIONAL EMPLOYER COSTS (NEW ARRANGEMENTS – TRANSFER OPTION) LOW TRANSFER ASSUMPTIONS

Year	MSBS ($’m)	DFRDB ($’m)	New Accumulation ($’m)**	Total ($’m)	Notional Employer Cost %
2006/07	178*	1,306*	382	1,866*	26.0
2007/08	168	1,339	422	1,928	25.9
2008/09	177	1,374	466	2,016	25.6
2009/10	183	1,405	513	2,101	25.3
2014/15	256	1,518	770	2,545	24.7
2019/20	372	1,592	1,053	3,018	24.6
2024/25	627	1,619	1,359	3,606	24.8
2029/30	741	1,604	1,705	4,050	25.0
2034/35	892	1,558	2,103	4,552	25.0
2039/40	1,037	1,452	2,570	5,059	25.0
2044/45	1,052	1,283	3,130	5,465	25.0

* 2006/07 figures do not include transfer values of $3.5 billion.

** includes an allowance of 3.8% of superannuation salary for new accumulation members for death and invalidity cover. (MSBS and DFRDB figures include death and invalidity benefits.)

Note that components may not add up to total due to rounding.

The tables above incorporate an allowance of 3.8% of superannuation salaries for the new accumulation arrangement for the employer provided separate death and invalidity cover. It is intended that the invalidity cover arrangements would be integrated with workers compensation arrangements. The 3.8% allowance is the cost of current MSBS death and invalidity insurance cover if it applies to the combined MSBS and DFRDB contributors.

We tried to estimate the costs of the new combined superannuation and workers compensation cover compared to the existing combined superannuation and workers compensation cover. We found that the data available was not sufficiently reliable to enable robust costings to be made. However, rough reasonableness type checks and general logic suggest that the combined costs of the new arrangements would be less than the combined cost of the existing arrangements. The difference could be of the order of about 0.5% of superannuation salaries for the individuals covered by the new arrangements but is highly uncertain. We thus believe that the use of a 3.8% allowance is conservative and probably slightly overstates costs for comparison purposes.

The details of the death and invalidity cover arrangements are still to be decided, including the financing mechanism. For the projections we have assumed that the equivalent element to the existing MSBS cover is pre-funded using an employer contribution of 3.8% of superannuation salaries of the individuals in the new accumulation arrangement. If the arrangement is not pre-funded (that is, operates on an unfunded or pay-as-you-go basis) then Commonwealth cash outlays will be lower, particularly in the early part of the projection period, but Commonwealth unfunded liabilities will be higher.

COMMENTS ON THE PROJECTIONS INCORPORATING PROPOSED NEW ARRANGEMENTS

Note that the projections have only looked at the direct costs of the schemes. We have not attempted to incorporate tax and social security effects. These are very difficult to incorporate as they depend on individual circumstances, tax rates and social security rules over a long time frame. However, given that the Commonwealth has broadly tried to maintain consistency of tax impacts and social security impacts between unfunded and funded superannuation arrangements, it would be expected that tax and social security effects should roughly even out in the long term between the two types of arrangements. Accordingly, considering the notional employer costs of the respective arrangements is probably the best cost comparison available.

The new arrangements have little effect on projected employer cashflows and unfunded liabilities for DFRDB. This is mainly because most of these are locked in as they are attributed to current pensioners who will not be offered any option to transfer. I expect the transfer offer made to DFRDB contributors will not appear attractive and I anticipate a low take up rate of the offer.

There is a substantial reduction in projected unfunded liabilities for MSBS. This is partly due to the initial transfer. Going forward, the level of ongoing accruals for MSBS contributors is also greatly reduced as there will be far fewer contributors following transfer. The remaining MSBS contributors will gradually exit over the next thirty to forty years. The employer cashflows for MSBS are noticeably reduced in the short term with the reductions becoming much more significant as time goes on.

Overall employer cashflows are much higher. This starts with the initial transfers of about $4bn to $5bn. In the near to medium term, employer cashflows are about 15% to 25% larger than currently which is about $300m to $400m per annum in the near term. This relative increase in employer cashflows only starts reducing after about 15 years and it takes about 30 to 35 years before the annual employer cashflows are similar to those projected for the current arrangements. Thereafter, annual employer cashflows are lower than for the current arrangements.

Overall, when the accrued liability position and projections of notional employer cost are taken into account, I am of the view that the assessed cost of the current arrangements and the proposed new arrangements are not dissimilar.

The figures in this report have been based on data as at 30 June 2006 and assume that any transfers from MSBS and DFRDB happen on 30 June 2006. I believe that they are appropriate to enable decisions to be made. In practice, the transfer will not take place for some time, quite possibly not until 2009. The key features disclosed by the projections using a transfer date of 30 June 2006 will still be present if the transfer date were in 2009. These key features are:

• the notional employer cost under the proposed new arrangements is similar to the current arrangements. The current annual superannuation cost to Defence for accruing benefits is of the order of $750m (26% of $2.9bn superannuation salaries);

• unfunded liabilities initially fall somewhat, but decline dramatically towards the end of the projection period;

• annual Commonwealth cash expenditure is about 15% to 25% higher for about 15 years, with the margin reducing to be about breakeven after 30 to 35 years before moving to some savings relative to the current position after that;

• accrued liabilities will be of the same order; and

• there is a large amount of immediate Commonwealth cash expenditure associated with the transfer.

The most noticeable change associated with deferral of the transfer date to, say, 2009 is that the immediate cash expenditure will be larger. I estimate that it will have grown to approximately $7bn to $8bn compared to around $5bn as at 30 June 2006 on the high transfer assumption. There are two main reasons for the increase. The first and more significant is that serving MSBS contributors will accrue further unfunded benefits during this period. The second effect is that the transfers will be in 2009 dollar terms not in 2006 dollar terms. That is, inflation means that the amount expressed in nominal dollars will be higher.

CLOSURE OF MSBS TO NEW MEMBERS WITH NO TRANSFER OFFER

For reference, we have prepared projections to 2045 assuming the MSBS was closed to new entrants as at 30 June 2006. Existing members of the MSBS and DFRDB would not be offered the opportunity to transfer to the new accumulation arrangement. That is, these individuals would remain in the MSBS and DFRDB respectively.

The recommended changes to the MSBS MBL arrangements have been incorporated in these projections. The recommended changes to the MSBS salary definition used for death, invalidity A and invalidity B benefits have not been incorporated into these projections. If this latter recommendation is implemented, the cost shown in the projections will be slightly understated.

The projected unfunded liabilities are set out below:

PROJECTED UNFUNDED ACCRUED MSBS AND DFRDB LIABILITIES (CLOSURE OF MSBS TO NEW ENTRANTS – NO TRANSFER OPTION)

Year ending 30 June	MSBS ($’m)	DFRDB ($’m)	Total ($’m)
2006	9,718	23,623	33,342
2007	10,632	23,824	34,456
2008	11,599	23,997	35,596
2009	12,582	24,130	36,712
2010	13,596	24,223	37,820
2015	18,868	24,166	43,034
2020	24,386	23,305	47,691
2025	29,602	21,790	51,392
2030	33,677	19,701	53,378
2035	36,322	17,077	53,399
2040	36,692	14,019	50,711
2045	35,205	10,748	45,953

* Note that components may not add up to total due to rounding.

The projected Commonwealth cash flows and the notional employer costs as a percentage of total Defence superannuation salaries is set out below:

PROJECTED COMMONWEALTH CASHFLOWS AND NOTIONAL EMPLOYER COSTS (CLOSURE OF MSBS TO NEW ENTRANTS – NO TRANSFER OPTION)

Year	MSBS ($’m)	DFRDB ($’m)	New Accumulation ($’m)**	Total ($’m)	Notional Employer Cost %
2006/07	266	1,314	3	1,583	25.7
2007/08	246	1,347	44	1,637	25.1
2008/09	262	1,384	92	1,738	24.5
2009/10	268	1,416	145	1,830	24.1
2014/15	391	1,532	458	2,381	23.6
2019/20	592	1,608	802	3,003	23.8
2024/25	943	1,636	1,184	3,762	24.2
2029/30	1,326	1,621	1,608	4,556	24.7
2034/35	1,791	1,577	2,058	5,426	25.0
2039/40	2,280	1,473	2,561	6,313	25.0
2044/45	2,414	1,304	3,130	6,848	25.0

** includes an allowance of 3.8% of superannuation salary for new accumulation members for death and invalidity cover. (MSBS and DFRDB figures include death and invalidity benefits.)

Note that components may not add up to total due to rounding.

AWOTE INDEXATION OF DFRDB PENSIONS

The Review Team has recommended AWOTE indexation (or something similar) of DFRDB pensions. This is a lower priority recommendation. The main lower priority recommendation is AWOTE indexation of pensions after age 55 but with only CPI indexation applying before age 55. A possible variation is AWOTE indexation of pensions after age 65 but with only CPI indexation applying before age 65. The impacts of these improvements are set out below.

- Unfunded Liability

The following table shows the impact of the various AWOTE indexation proposals if they had been implemented as at 30 June 2006.

	Unfunded DFRDB Liability ($bn)	Increase in Unfunded Liability ($bn)
Current arrangement	23.6	0.0
AWOTE indexation from age 65	26.2	2.6
AWOTE indexation from age 55	27.8	4.2

- Notional Employer Contribution Rate

The effect on the notional employer contribution rate from the various AWOTE indexation proposals if they had been implemented as at 30 June 2006 is as follows:

	DFRDB Notional Employer Contribution Rate %	Increase as a Percentage of DFRDB Superannuation Salaries %
Current arrangement	33.5	0.0
AWOTE indexation from age 65	36.7	3.2
AWOTE indexation from age 55	38.9	5.4

If the age 65 AWOTE indexation proposal had been implemented as at 30 June 2006, the additional notional employer cost in 2006/07 would have been about $14m which is about 0.5% of total Defence superannuation salaries. If the age 55 AWOTE indexation proposal had been implemented as at 30 June 2006, the additional notional employer cost in 2006/07 would have been about $23m which is about 0.8% of total Defence superannuation salaries. These costs as a percentage of total Defence superannuation salaries would quickly diminish as DFRDB contributors exited service.

- Additional Cash Expenditure

The chart below shows the expected additional cash expenditure had the age 55 AWOTE changes been introduced with effect from 2006/07.

- Comments on Results

It can be seen from the above that any move to AWOTE indexation for DFRDB is an expensive benefit improvement. While additional Commonwealth cash expenditure in the first few years after any change would be relatively small, the additional amount of cash expenditure would continue to increase for a long time and become very noticeable. Extra cash expenditure would peak in about 40 years time in nominal dollars. The peak extra annual expenditure in nominal dollars would be about $700m (at that stage, around 50% higher than current projected outlays) for the AWOTE from age 55 improvement and about $500m for the AWOTE from age 65 improvement.

A deferral of implementation to 2009/10 would slightly reduce the additional Commonwealth cash expenditure and costs in real terms.

SUPERANNUATION GUARANTEE ARRANGEMENTS – ORDINARY TIME EARNINGS

Defence has to ensure that it complies with the minimum Superannuation Guarantee (SG) requirements in relation to its employees. For the new scheme, this requirement extends to the use of Ordinary Time Earnings (OTE) for the purposes of assessing SG compliance.

At the present time, the existing schemes are deemed to have ‘protected earnings bases’. This means that compliance with SG can be assessed in relation to superannuation salary. From 1 July 2008, protected earnings bases will be removed for all defined benefits funds and compliance will be assessed in relation to Ordinary Time Earnings (OTE). Changes to the SG regulations and, consequently, updating of Institute of Actuaries of Australia Guidance Notes in relation to SG compliance are still to occur.

The use of OTE for SG compliance has the potential to cause significant compliance issues for Defence as well as possibly requiring benefit improvements in some circumstances. This is despite the fact that the overall level of employer superannuation support is, by community standards, generous.

In undertaking these costings, we have assumed that compliance will be achieved without any substantial increase in costs (either direct or administrative costs). The following discussion sets out the issues.

The first issue is that there is no real concept of Ordinary Time for Defence Force personnel. Technically, Defence Force personnel are on duty 24 hours a day.

The second issue is that Defence Force pay and allowances are complex and cover a myriad of situations. We have received advice from the Australian Taxation Office (ATO) that many of the special payments for military personnel which do not fall under the current definition of superannuation salary will form part of OTE. While the ATO’s views may be technically correct, I would regard some of the items included in OTE being far from ordinary from a common sense perspective. For instance, according to ATO advice, deployment allowances paid tax free are regarded as being part of OTE.

The third issue is that special payments can be large and lumpy (only payable for a relatively short period of time or in a single lump sum) and, as a result, they could cause the standard employer contribution to the new accumulation arrangement to be insufficient to meet minimum SG requirements from time to time. In many cases, the top up contributions are likely to be small because there is a maximum salary that is used for measuring SG compliance. For 2007/08 the maximum quarterly earnings base for SG is $35,470. Note that although the individual may have received significantly in excess of the minimum SG requirements for most of the period of employment, those excesses cannot be used to offset the requirement to pay more in financial quarters where large amounts of special payments are made.

Given the complexity of Defence arrangements, it would be very helpful for there to be some form of guidance (which may need to be legislated) to clarify what constitutes OTE for Defence Force personnel. For a sensible outcome, it will be important that this guidance takes a common sense approach. For example, the Government could deem that the schemes automatically meet minimum SG requirements given the overall generous level of benefits provided.

Michael Burt
Fellow of the Institute of Actuaries of Australia
30 July 2007