Report Prepared

for the

New Zealand Business Roundtable

Government-guaranteed Financial Institutions:

Policy Issues Relating to the

National Provident Fund

Government Superannuation Fund

and Public Trust Office

June 1997

ACKNOWLEDGMENTS

The principal authors of this report are Bryce Wilkinson and Susan Begg. The authors are grateful to the officers of the Government Superannuation Fund, the National Provident Fund and the Public Trust Office and to many other individuals for comments on earlier drafts. However, the authors alone are responsible for the views expressed and any remaining errors and omissions.

First published in 1997 by New Zealand Business Roundtable

PO Box 10-147, The Terrace, Wellington, New Zealand

ISBN 1-877148-27-X

© 1997 edition: New Zealand Business Roundtable

© Text: First NZ Capital Limited

Production by Daphne Brasell Associates Ltd, Wellington

Printed by Astra DPS, Wellington

Contents

Executive Summary

1. Successive governments have reduced the Crown's involvement in the provision of financial services by privatising and restructuring government financial institutions and deregulating financial markets. Notwithstanding the reforms to date, the Crown remains responsible for the operations of the National Provident Fund (NPF), the Government Superannuation Fund (GSF) and the Public Trust Office (PTO). Fiscal risks accompany such responsibilities.

2. The comprehensive restructuring of the National Provident Fund has curtailed fiscal risks relating to this organisation. The government has closed the fund to new members, ensured that the NPF operates under clear legislation and is subject to a well-constructed trust deed, provided for the minister of finance to appoint board members, and imposed reporting requirements on its board. However, some risks remain since various benefits are government guaranteed and the Crown remains responsible for the NPF.

3. As a consequence, the government continues to have a responsibility for monitoring the NPF's performance in relation to matters affecting its cost structures (which include service quality issues), its asset allocation decisions (particularly where these affect taxpayer risks) and member benefits.

4. The government should consider opportunities for further cost savings, particularly in respect of the DBP Annuitants Scheme (DBP meaning 'defined benefit plan') where any excess costs fall on the taxpayer. The reasons why the costs being attributed to this scheme are so much higher than those being incurred by the GSF should be investigated.

5. The Crown and its advisers should also be alert to any opportunities to reduce taxpayer exposure when circumstances change. For example, from time to time a change in circumstances may give rise to uncertainty as to how existing contractual terms apply to the new situation, or the change in circumstances may provide an opportunity to modify schemes to improve outcomes for taxpayers, members and employers.

6. The full or partial amalgamation of the GSF and the NPF is a further cost-saving possibility. Other possibilities include reducing the number of fund managers employed by the NPF and reviewing cost allocations and service quality in the light of the differences in cost structures between the GSF and the NPF.

7. The most radical option for further reducing fiscal risks would involve the Crown paying a credible third party to assume the Crown's obligations - for a consideration which would be determined by competitive tender. This could be particularly difficult where the purchaser would in effect be underwriting existing employer obligations. However, the Crown's commitments under the DBP Annuitants Scheme do not fall into this category.

8. The government has reformed the Government Superannuation Fund by closing the scheme to new members, contracting out scheme administration, clarifying the roles and accountability of the members of the superannuation board, the superintendent of GSF and the minister, and improving accountability by requiring the GSF to disclose specified information. Large risks remain with the taxpayer.

9. Given the open-ended nature of the government guarantee of the GSF, the Crown needs to be vigilant in monitoring events and resisting rule amendments and interpretations which could transfer wealth from the Crown to scheme beneficiaries.

10. The GSF continues to provide investment management and custodial services internally. The amount being spent on these services, at 0.03 percent of the GSF's $3.2 billion net assets (3 basis points), is minute from a commercial perspective and presumably reflects a relatively simplified operation. Contracting out of these services should be investigated, but any savings are likely to be minor in relation to the scale of the GSF's operations.

11. The GSF also appears to have abnormally low scheme administration and head office costs per member. The major component of these costs has been determined by competitive tender. Their low level therefore also presumably reflects a comparably simplified operation.

12. The current practice of keeping the GSF's assets very low in relation to the actuarial liability, and investing them in debt and fixed interest securities, preserves an asset allocation which is dramatically different from that of a conventional defined-benefit fund. The fact that pension funds in North America and Europe tend to have significant equity weightings indicates that equity investments should be regarded as suitable unless compelling grounds exist to the contrary. Furthermore, the current allocation also differs dramatically from that of the NPF's government-guaranteed DBP Annuitants Scheme. There may also be a case for lengthening the average duration of the GSF's fixed interest investments if so doing would better hedge taxpayers against future shocks to the actuarial deficit, or enhance the returns generated by the scheme. Deciding on an optimal asset allocation would be complex, but it is hard to see a compelling case for the current government-issued investment guidelines which prevent the GSF from investing in equities.

13. A number of options for changing the GSF's investment strategy exist: the GSF could be required to invest only in Crown debt (this would shift responsibility for all meaningful asset allocation decisions to the New Zealand Debt Management Office (DMO)); the GSF's fund managers could be allowed to manage the fund actively around a notional benchmark portfolio; the $3 billion fund could be invested in existing asset classes so as to minimise the volatility in the actuarial deficit; investments in higher yielding assets could be allowed subject to the same objective; or the GSF could be fully funded and operated like a traditional funded superannuation scheme or the comparable DBP schemes run by the NPF. In our view there is a plausible case for funding the GSF more fully and allowing it to adopt an asset portfolio which has an equity component for growth and a duration which better matches the duration of its liabilities.

14. A complementary option would be to amalgamate the GSF and the NPF. Integrating the GSF into the NPF might facilitate a move to a more fully-funded and conventionally-invested trustee structure. An issue here would be to ensure that taxpayers were not penalised as a result of the NPF's higher cost structure.

15. A more radical option for reducing the Crown's fiscal risks would be for it to pay a credible third party to assume the Crown's obligations - for a consideration which would be determined by competitive tender. This would be a major undertaking and would require very careful consideration of the degree to which taxpayer and member interests had been truly protected. One option might see the private sector party's obligations take the form of a contract for differences. This option is more attractive the less confident the government is about which investment mix would be in taxpayers' best interests. It could eliminate the need for further consideration of the investment mix and full-funding issues just canvassed. These would become a private sector responsibility. This approach may require the GSF to split its schemes into contributor and annuitant schemes.

16. In our view, there are no public policy reasons for the Crown to own, sponsor or guarantee a trustee corporation such as the Public Trust Office. Competing trustee companies, the legal profession and private individuals provide competitive trustee services without the benefit of a government guarantee, and the PTO would be likely to continue to service most of its clients without government backing.

17. Our (brief) examination of the PTO's social and statutory roles suggests that none of the 'social' services it provides, nor its statutory functions, require a state-backed trustee corporation. Instead, services that serve a useful purpose could be purchased from the PTO or other organisations on a commercial basis or provided by other government departments.

18. Despite the strong case for reform of the PTO, a number of impediments to restructuring exist. These include:

• conflicting legal opinions as to current ownership of the PTO;
• the extent of legislative change required to remove the special treatment of the PTO;
• the desirability of reviewing the trustee companies legislation; and
• a lack of political will.

19. Given these constraints, the current approach by officials - of improving the accountability of the PTO within the current statutory arrangements, while continuing to examine other options for reform - appears sensible.

20. In the longer term, the PTO should be established as a trustee company under the legislation that governs trustee companies, with the shares distributed to customers or the company sold through a tender. The proceeds could either be retained by the government or distributed (in whole or part) directly to clients or to a trust to provide benefits to past and current clients.

21. In any event, the government should remove the guarantee of the PTO's common fund. The government could possibly require a change in the company's name to reflect the severing of its links with the government if there is a significant risk that the term 'public' carries connotations of government ownership and if the benefits of the change outweigh the loss of a name with wide customer recognition.

Chapter 1 Introduction

1.1 The Crown's involvement in the provision of financial services can take many forms including: direct ownership or other funding; control over board appointments and other aspects of their operations; and conferral of special status through selective legislation, regulation and Crown underwriting or guarantees. Such activities expose taxpayers to financial risks in the event of financial failure or litigation. Successive governments have reduced the Crown's involvement in the provision of financial services by privatising and restructuring Crown financial institutions and deregulating financial markets.

1.2 Notwithstanding the reforms to date, the Crown retains a significant involvement in the operations of the National Provident Fund (NPF), the Government Superannuation Fund (GSF) and the Public Trust Office (PTO). This involvement is risky for taxpayers. The purpose of this report is to review current policies in relation to these institutions.

1.3 The NPF, and to a lesser extent the GSF, have already been comprehensively restructured, with their schemes now being closed to new members. The PTO, however, has been relatively untouched by the reform programme to date.

1.4 Chapters 2 and 3 of this report review the few policy issues that remain in relation to the NPF and GSF respectively. Chapter 4 considers the ownership and governance issues that relate to the PTO. Recommendations are presented in Chapter 5. The Executive Summary at the beginning of the report presents our conclusions.

Chapter 2 The National Provident Fund

2.1 Background

2.1.1 The NPF is a true superannuation fund. At 31 March 1996 it had $2.7 billion in accumulated net assets. It has 75,000 contributors and 700 employers and at 31 March 1996 was paying pensions to 27,000 annuitants.

2.1.2 The NPF provides seven defined contribution schemes, seven defined benefit schemes and two cash accumulation schemes, i.e. 16 in total. At 31 March 1996, the distribution of net assets across these three groups of schemes was 43 percent, 47 percent and 10 percent respectively.

2.1.3 We illustrate the nature of these schemes with four examples, key details for which are summarised in Table 1. The Pension National Scheme is the largest of the 16 schemes. At 31 March 1996 this defined contribution scheme had 70,580 members (of whom 21,018 were contributors, 39,671 were inactive, with no known contact addresses, and 9,891 were pensioners). Its net assets at 31 March 1996 were $867.4 million.

2.1.4 At 31 March 1996, the second largest scheme was the DBP Annuitants Scheme with 8,356 members (all pensioners as its name implies) and net assets of $602.9 million. (DBP meaning 'defined benefit plan'.) This scheme's actuarial deficit was estimated to be $429 million at 31 March 1995. This is the only major actuarial deficit in the NPF's schemes. In the Minister of Finance's 12 September 1996 Pre-Election Economic and Fiscal Update, the Crown's liability to the NPF in respect of the government guarantee was estimated to be $429 million on 30 June 1996.

2.1.5 The third largest scheme is the closely related DBP Contributors Scheme. It had net assets of $489.5 million at 31 March 1996 and 4,084 members, all of whom were contributors as its name implies. Membership is falling markedly with 801 persons transferring out of the scheme during the financial year. Of these, 236 transferred to the DBP Annuitants Scheme.

2.1.6 The largest cash accumulation scheme is the Lump Sum Cash Accumulation Scheme. At 31 March 1996 it had $219.2 million in net assets and 18,270 members, only 47 of whom were pensioners, the rest being contributors. The funds are all invested in a fixed interest portfolio with a weighted average duration equal to that of the First NZ Capital New Zealand Government Stock Index. The significance of tax and expenses is indicated by the fact that the investment return on the scheme during the year to 31 March 1996 was 4.44 percent, against a 7.5 percent gross return. Operating expenses in 1995/96, at $2.5 million, were 1.26 percent of the average of opening and closing net assets.

2.1.7 The NPF has the legal characteristics of a financial mutual. The National Provident Fund Restructuring Act 1990 makes the NPF's Board responsible for managing members' funds. The payment of benefits by the schemes is guaranteed by the Crown.

2.1.8 Many factors contributed to the impetus for fundamental reform of the National Provident Fund which was embodied in the National Provident Fund Restructuring Act 1990. These factors included: the reduced role of central government in local government funding; the increased focus by governments on better managing Crown balance sheet risks; the collapse of the Development Finance Corporation (DFC); and evident weaknesses in the NPF's governance arrangements at that time.

2.1.9 The NPF's schemes were closed to new members from 1 April 1991. This reduced fiscal risks. In addition, the Board was reconstituted as a Board of Trustees, new management practices were imposed and a competitive framework was put in place for scheme administration and asset management.

2.1.10 The National Provident Fund now consists of a Board of Trustees which is serviced by a small executive office currently comprising a chief executive and four other staff.

2.1.11 The Board of Trustees manages the schemes and their accumulated funds in the best interests of scheme members, taking into account the interests of the Crown as guarantor of scheme benefits. All six members of the Board are appointed by the minister of finance. None are NPF employees, but the chief executive does attend Board meetings. The Board meets monthly. It reports separately on each scheme. The full set of scheme reports is sent to the minister of finance. It also reports regularly to scheme members and produces a Provident Review newsletter.

2.1.12 The executive office is responsible for reporting, monitoring and advising the Board on its contracts for the necessary scheme administration, investment management and custodial services. This role includes reviewing the asset class allocations for each scheme. The investment managers take these allocations as given. Other activities include protecting the scheme's interests and discussions with the government about possible scheme modifications.

2.1.13 These arrangements make the minister of finance accountable for the performance of the Board of Trustees. The Treasury advises the minister of finance and is therefore responsible for monitoring the performance of the Board. The Board is responsible for the performance of the NPF's executive office. This encompasses the quality of its recommendations on asset allocation decisions. (This structure arises from the unique historical development of the NPF, and is unusual in that asset allocation benchmarks would normally be approved by the (employer) sponsor of a pension plan rather than by the trustees.) This structure also makes The Treasury responsible for monitoring the NPF's asset allocation benchmarks and investment strategies.

2.1.14 As a result of competitive tenders, Jacques Martin New Zealand has day-to-day responsibility for administering the 16 schemes. State Street New Zealand Limited, an international global custodian, provides custodial services. These include accounting for, and reporting on, the investments of the funds. Twelve local and overseas fund managers share responsibility for managing the accumulative funds of the combined investment portfolios - the Global Asset Trust.

2.1.15 Head office expenses in relation to the NPF's schemes totalled $2.5 million in 1995/96. Administrative expenses (Jacques Martin New Zealand) totalled $12.8 million. The sum of these two amounts represented an average cost of $150 per member (comprising 75,000 contributors and 27,000 annuitants). Total operating expenses, which include administrative expenses, audit and legal fees, actuarial fees and bank and sundry charges, totalled $16.8 million or $165 per member. These exclude investment management and custodial fees which are deducted directly from the earnings of each of the unit funds, with only the net earnings figure being disclosed.

2.1.16 In 1993 the Board commissioned Frank Russell to review the NPF's asset allocations. The main objectives of the review were to assist the Board to implement policies which would:

(i) maximise returns to members at an acceptable level of risk;

(ii) maximise the Board's ability to meet its obligation to credit each year a minimum earnings rate of 4 percent per annum to the total credits of members of defined contribution and cash accumulation schemes; and

(iii) ensure that the risk to the government, as guarantor of benefits payable, does not increase.

2.1.17 The NPF's obligation in respect of the 4 percent minimum earnings rate has markedly affected its asset allocation policy. Table 2 illustrates this point. Where this obligation does not apply, the NPF typically adopts the following asset allocation benchmark: domestic fixed interest 20 percent, property 15 percent, New Zealand equities 12.5 percent, and overseas equities 52.5 percent. Where the obligation does apply, the asset allocation benchmark is either 60 or 70 percent in cash and domestic fixed interest, with corresponding allocations of 7.5 or 5 percent in each of property and New Zealand equities and 25 or 20 percent in overseas equities.

2.1.18 The minimum annual requirement therefore forces very high combined cash and fixed interest weighting in the portfolio. Only the largest scheme, the contributors' section of the Pension National Scheme, and the very small National Superannuation Scheme for Farm Workers have the 70 percent weighting in cash and fixed interest. The 100 percent fixed interest allocation for the Lump Sum Cash Accumulation Scheme, noted in paragraph 2.1.6 above, is an exception.

2.1.19 The unconstrained asset allocation (second row in Table 2) can be compared with local fund manager benchmarks. For example, Jacques Martin New Zealand report each quarter on the performance and asset allocation decisions of wholesale superannuation funds managers. Their March 1996 Investment Update surveyed 20 superannuation funds (encompassing 12 fund managers) and reported that on average they had 43 percent in cash and bonds, 11 percent in property and 47 percent in equities. On this comparison, the NPF's benchmark is slightly more aggressive than the reported benchmark allocation for a growth fund, of 25 percent in cash and bonds, 20 percent in property and 55 percent in equities.

2.1.20 The NPF's equity investments have a relatively high overseas component. For example, according to the same survey, the average allocation to offshore markets by the surveyed superannuation fund managers was 24 cents of each dollar invested in equities. This was very

close to the reported benchmark figure of 25 percent for a growth fund. In contrast, the comparable figure for the NPF is 73 cents (in the case of the 20 percent/7.5 percent weights) or 83 cents (in the case of the 25 percent/5 percent weights). For a different perspective, note that a fully diversified (market) global equity portfolio would have over 98 cents in each dollar invested outside New Zealand. The 24 cents figure for the surveyed local fund managers is up from 19 cents one year earlier, which probably indicates that local fund managers are slowly moving to more diversified portfolios. Remarkably, Jacques Martin New Zealand could report in the same issue that in 1994 four of the surveyed fund managers had no North American or European shares - despite their weightings of 43.6 percent and 28.2 percent respectively in the Morgan Stanley Capital International World Index.

2.1.21 Since the NPF's schemes are independent, the NPF does not have any target for the overall distribution to asset classes of its $2.7 billion of net assets. The overall asset mix has no operational significance for the Board. It is simply an outcome which reflects the asset allocation decisions made for each scheme and the amounts members have invested in each scheme. However, analysts of the superannuation industry are sometimes interested in the overall asset allocation across diverse schemes. From this perspective it may be useful to note that at 31 March 1996, the Global Asset Trust was invested in the following unit funds: 42.5 percent in cash and fixed interest; 46.8 percent in equities; and 10.7 percent in property. Remarkably, according to Jacques Martin New Zealand, the average for surveyed fund managers at 31 March 1996 was the same, at 43 percent, 47 percent and 11 percent respectively.

2.2 Crown Risks and Monitoring Responsibilities

2.2.1 The Crown guarantees benefits under all the NPF's schemes. It also guarantees investments and interest thereon deposited with the NPF Board prior to 1 April 1991. However, the reforms have done much to reduce the Crown's future risks by: closing the fund to new members; ensuring that the NPF is operating under clear legislation and subject to well-constructed trust deeds; providing for the minister of finance to appoint all NPF board positions; and imposing reporting requirements on the board which allow The Treasury and others to monitor its activities.

2.2.2 As a result, in many cases the ongoing risks to the Crown appear to be low, as is indicated in paragraphs 2.2.3 to 2.2.8 below. In other cases some risks clearly remain.

2.2.3 Defined Contribution Schemes. In the case of defined contribution schemes, the risks of disappointing investment returns typically lie with contributors. In such circumstances, the risks to the Crown should be low.

2.2.4 The DPB Contributors Scheme. The risks to the Crown from actuarial deficits arising from obligations to existing contributors are reduced by the provisions in section 44 of the National Provident Fund Restructuring Act 1990. These provide for employer contributions, in respect of the DBP Contributors Scheme and the Aircrew Scheme, to be increased or decreased so as to achieve actuarial balance. Actuaries value the liabilities each year and determine a range for the required future employer contribution rate for actuarial balance. This flexibility reduces Crown risks.

2.2.5 At 31 March 1995, there was an estimated $53 million actuarial deficit in the DBP Contributors Scheme ñ up from $33 million the previous year. The Board's actuary determined that employers would need to contribute at a rate of 157 - 201 percent of contributors' contributions in order to achieve actuarial balance. Since the current contribution rate of 170 percent was within that range, the Board did not recommend a change in that rate. Under section 44 of the National Provident Fund Restructuring Act 1990, the minister of finance could direct the Board to increase the contributions made by a corporate employer to the scheme, notwithstanding the absence of a Board recommendation to do so. Currently, the Crown's confidence that this deficit can be covered by employer contributions is reflected in the absence in the Crown Financial Statements of any provision for liability for this deficit.

2.2.6 Guarantee of investments and interest. The Crown guarantees investments and interest thereon deposited with the NPF prior to 1 April 1991. These guarantees relate to sinking funds deposited with the Board in its capacity as Sinking Fund Commissioner. On 7 December 1993, the Board entered an arrangement with the New Zealand Debt Management Office. As explained in note 6 to the accounts of the DBP Annuitants Scheme for 1995/96, this arrangement has the effect of transferring any risks associated with these deposits to the Debt Management Office.

2.2.7 Claims by non-members. As the DFC collapse illustrated, the Crown can be vulnerable to claims by non-members, such as third-party creditors, in certain circumstances. However, these risks should be small in respect of the NPF's unlevered investments in cash, government bonds, and shares in the sharemarket. Losses in value in such investments should not involve any third parties.

2.2.8 The Crown indemnity. The Government's Pre-Election Economic and Fiscal Update of 12 September 1996 also lists a provision at 30 June 1996 and beyond of $96 million by way of indemnity to the NPF. We understand that this is a technical provision arising out of arrangements put in place at the time of the collapse of the DFC and does not imply that the Crown has a net fiscal exposure of this amount. No issues appear to arise here.

2.2.9 The Actuarial Deficit in the DBP Annuitants Scheme. The Crown Financial Statements provide in full for an actuarial deficit of $429 million at 31 March 1995 in the non-contributory DBP Annuitants Scheme. The annuitants in this defined benefit scheme receive a retirement benefit which is guaranteed by the Crown. The benefit is calculated according to a formula based on the member's final average salary, years of membership and an age-related pension factor.

2.2.10 The Crown is fully exposed to any changes in the value of this scheme's assets relative to its actuarially determined liabilities. However, the NPF and the Crown can probably do little to affect fluctuations in the actuarial value of the scheme's liabilities. These fluctuations will arise inter alia from revised actuarial assumptions about life expectancies, real interest rates (the pensions are inflation-linked) and the decisions made by superannuitants in relation to capitalisation options.

2.2.11 However, the Crown might be able to influence some factors which could affect the scheme's liability. These include the scheme's asset allocation choices, since these affect investment returns, the costs associated with running the scheme and the possibility that from time to time decisions may have to be made about interpretations of rules or variations in the rules. The Crown therefore has reason to monitor these matters relatively closely. We briefly look in turn at each of these aspects.

2.2.12 The asset allocation benchmark for the DBP Annuitants Scheme has 80 percent of funds in equities and properties. As noted in section 2.1, this is relatively high. If anything, there is an argument that a lower rather than higher ratio would be more appropriate for an annuitants' scheme than for a contributors' scheme for time horizon and liquidity reasons. However, since the NPF's scheme is government guaranteed, the relevance of a more conservative allocation is reduced. Any review of this issue should start with the Frank Russell report. Such a review does not appear to warrant a high priority. The Board reviews the asset allocation benchmarks for all the NPF's schemes regularly every three to five years.

2.2.13 The cost structures in the DBP Annuitants Scheme reflect the NPF's overall cost structures and the costs which are specific to the scheme. The major portion of the NPF's costs across all schemes result from competitive tenders which were carried out in 1991. In this case they should represent fair value for the quality of the services provided. However, it remains a puzzle as to why the costs associated with the DBP Annuitants and Contributors Schemes are so much higher than those of the GSF's schemes (refer to paragraph 3.5.6 below). There may be issues here of timing differences, service quality or the allocation of common costs for the Crown to consider since it is in effect bearing the costs of service quality for both the NPF's DBP schemes and the GSF's schemes. The cost differential raises the possibility that service quality may differ markedly between the NPF's and GSF's schemes. Differences in service quality would raise a question as to which quality level represented the better balance between taxpayer and member interests.

2.2.14 A second issue of cost structure relates to the possible scope for reducing investment management costs. The NPF does not publish information on its investment management and custodial expenses. These are deducted from the earnings of each of the NPF's unit funds, with only the net figure being published. The NPF has appointed its investment managers by competitive tender. Its costs should therefore be comparable to private sector best-practice norms. However, it is fair to ask if the multiplicity of 12 fund managers is unduly raising costs by reducing economies of scale and increasing the monitoring burdens of head office. Costs might be reduced if the funds were spread over fewer fund managers. A smaller number of fund managers would reduce monitoring and reporting costs. Indeed, the NPF has reduced appreciably the number of external fund managers it is using. A countervailing argument is that reducing the number of managers may raise prudential management issues.

2.2.15 A third possibility for cost reductions would be to amalgamate the NPF and the GSF, or at least to amalgamate the defined benefit schemes. This option is discussed in section 3.5 below.

2.2.16 In any long-lived scheme, opportunities are likely to arise from time to time for modifying or reinterpreting the rules of the scheme in ways which could transfer wealth between the affected parties. At such times there are risks and opportunities for the Crown, scheme members, and contributing employers. Where the schemes are operating under a government guarantee, all those with a pecuniary interest in the schemes may favour amendments or interpretations which affect taxpayers adversely. Unless those who represent taxpayer interests are vigilant, and have access to information about what is happening, or have properly contracted with an agent who does have the incentive, information and authority to act, there is the risk of taxpayer loss from an open-ended government guarantee.

2.2.17 The NPF's board is a board of trustees subject to the acts of parliament, most notably the National Provident Fund Restructuring Act 1990. As such, it cannot and should not be made responsible for representing taxpayer interests exclusively. It cannot encourage members to take decisions that are not in members' interests. This means that The Treasury has to retain a monitoring role to ensure that the taxpayers' interests are being protected.

2.2.18 The 4 percent per annum minimum. The NPF is obliged, in respect of all defined contribution and cash accumulation schemes, to credit members each year with a minimum earnings rate of 4 percent, after tax and the payment of expenses. With lower interest rates this return may not be readily achievable in any given year. For example, in 1994/95 none of the 10 defined benefit and cash accumulation schemes were able to provide their required 4 percent minimum return without calling on reserves.

2.2.19 However, the Crown's exposure under this arrangement is only likely to be material should investment returns, net of expenses and taxes, be less than 4 percent per annum for a prolonged period. This is because the Crown is not required to make good any deficiency in the earnings rate in any one year. Instead, the NPF may be able to restore its reserves by crediting members with at least the 4 percent minimum, but less than the full earnings rate in the following year or years. As explained in the Provident Update, December 1994, the NPF's policy is to build up reserves, when the net earnings rate exceeds 4 percent, to 10 percent of a scheme's net assets.

2.2.20 The 4 percent per annum minimum requirement is troublesome for the asset allocation decision when bond yields are low and when reserves are low or negative. Positive reserves allow the return declared on members' total credits to exceed the return from funds actually invested, since these reserves can be used to supplement members' total credits. Conversely, the more negative the reserves become, the higher the required return from invested funds and the less likely it is that prime fixed-interest investments will be able to produce this return. A higher weighting in equities would increase the likelihood of achieving the 4 percent return in any period. On the other hand, investing more heavily in risky assets increases the potential for major loss in the event of a sharemarket collapse. Increasing equity weightings as reserves become negative is clearly a risky strategy.

2.2.21 The NPF is very conscious of these difficulties and is giving consideration to proposals for easing the fundamental constraint imposed by the 4 percent per annum minimum return requirement. Any such proposals could presumably be achieved only through legislation. This might be difficult to achieve. Furthermore, a major fall in share prices could easily create negative reserves if equity weightings were significant, causing the difficulties noted in the previous paragraph.

2.2.22 Outright Sale. As is discussed in more detail in section 3.6 below, the government could also consider paying a reputable private sector party to assume responsibility for meeting the Crown's commitments in relation to the NPF's schemes. This party would have to be a credible primary guarantor of benefits. Unless scheme members allowed the Crown to contract out of its obligation, the Crown would presumably remain the residual guarantor.

2.2.23 Since it is difficult to envisage who might take over the obligations of existing employers, it may be easier to achieve this objective for schemes which are not exposed to employer risk. The DBP Annuitants Scheme is one such example.

2.3 Concluding Comments

2.3.1 The comprehensive restructuring of the NPF and the closing of schemes to new members have contained the government's future exposure in regard to this institution. However, the government remains responsible for the NPF's performance as an institution and guarantees the benefits payable by the schemes.

2.3.2 As a consequence, the government continues to have a responsibility to monitor the NPF's performance in relation to matters affecting its cost structures (which include cost allocations and service quality issues), its asset allocation decisions (particularly where these affect taxpayer risks) and member benefits.

2.3.3 As part of this responsibility, the government should consider opportunities for further cost savings, particularly in relation to the DBP Annuitants Scheme where any excess costs fall on the taxpayer. The full or partial amalgamation of the GSF and the NPF is one possibility. Other possibilities include reducing the number of fund managers employed (although any savings here need to be balanced against the arguably less diversified prudential risks) and reviewing service quality in the light of the differences in cost structures between the GSF and the NPF.

2.3.4 In respect to member benefits, changing circumstances may from time to time give rise to uncertainty as to how existing contractual terms apply to the new situation, or the change in circumstances may provide an opportunity to modify schemes to improve outcomes for taxpayers, members and employers. The Crown and its advisers should be alert to any such opportunities to reduce taxpayer exposure.

2.3.5 The most radical option for further reducing fiscal risks would be for the Crown to sell to credible third parties its responsibility for meeting its commitments.

Chapter 3 The Government Superannuation Fund

3.1 Background

3.1.1 The Government Superannuation Fund Act 1956 governs the management and administration of the schemes which comprise the GSF, as well as its investments. It defines the minister of finance to be the responsible minister. By convention the minister is referred to as the minister in charge of the GSF. The Government Superannuation Fund Department is currently responsible for the administration of the Act. The department is a department of state which is listed in the first schedule to the State Sector Act 1988. Its chief executive must manage the department in accordance with the Government Superannuation Fund Act 1956, the Public Finance Act 1989 and other applicable legislation.

3.1.2 In meeting its obligations with respect to the GSF, the government purchases outputs aimed at:

· ensuring its contractual and statutory obligations to members are met; and

· minimising the Crown's liability.

3.1.3 The Crown's appropriations fund the following classes of activities:

· superintendent the role of advising the minister;

· investment management;

· schemes management monitoring of the contract for schemes administration and the superintendent's exercise of prudent discretion under the principal Act; and

· scheme administration payments of benefits, and collection of contributions.

3.1.4 For the 1996/97 fiscal year, the appropriations for these four activities were $148,000, $948,000, $761,000 and $3,167,000 respectively. The total appropriation for the Government Superannuation Fund Department's output classes was $1.857 million (the sum of the first three appropriations listed).

3.1.5 Unlike the NPF, the GSF cannot be likened to a financial mutual. Nor does it necessarily meet the definition of a superannuation scheme in the Superannuation Schemes Act 1989. Members of the GSF's schemes do not have the protection of trustee law. The main asset of the GSF is the government's legislated obligation to meet all its future pension obligations.

3.1.6 The GSF comprises eight defined benefit schemes. At 30 June 1996, the GSF's net assets were $3.144 billion and the unfunded liability was estimated to be $8.328 billion. Subsidy payments from central government net of residential withholding tax totalled $528.2 million in 1995/96, compared to $434.9 million in 1994/95. The number of beneficiaries fell by 0.3 percent to 45,610 in the year to June 1996. The number of contributors fell 8.8 percent to 44,458. In 1994/95 contributors exceeded beneficiaries.

3.1.7 Reflecting the fact that the Crown guarantees the payment of benefits under the GSF's schemes, the unfunded liability is recorded in the Crown's Financial Statements as a pension liability. Members of the GSF have minimal reason to be concerned with the GSF's costs or the investment performance of the $3.1 billion of net assets in the scheme since their benefits are defined and guaranteed regardless of costs and performance.

3.1.8 The overriding risk issue for the government is how to minimise the fiscal costs arising from the GSF's obligations and associated activities. In this respect the situation is very similar to the Crown's obligation for the NPF's defined benefit schemes. Existing risks are usefully set out in section 5 of the GSF's December 1996 post-election briefing. Risks relate to future employer contributions, investment returns, demographic and other factors which affect benefit entitlements on retirement, the exercise of the superintendent's discretion, and the efficacy of scheme administration and management.

3.1.9 Concerns about fiscal risk and conflicts of priority and purpose embodied in the GSF's earlier governance arrangements led to significant changes in the 1990s. In particular:

· all the GSF's schemes were closed to new members from 30 June 1992, except for the Cook Islands, Niue and Tokelau Public Service schemes which were closed to new members from 21 October 1995; and

· new administrative arrangements were made allowing the liability to be better identified and the administration of benefits to be contracted out.

3.1.10 From 1 March 1996, the GSF contracted out the function of administering the schemes to National Mutual, following a competitive tender process. This function was the major portion of the GSF's annual expenditure. Contracting it out markedly reduced the Government Superannuation Fund Department's role as a provider.

3.1.11 Governance issues were clarified with the passage in mid 1995 of an amendment to the Government Superannuation Fund Act 1956. The amendment:

· redefined the Government Superannuation Board, making it solely an appeals board;

· redefined the superintendent of the GSF to be the chief executive of the department responsible for administering the principal Act;

· specified that the minister responsible for the GSF must be the minister of finance;

· provided that the minister of finance would appoint the five members of the appeals board - of these, one or two members would represent contributors and one would represent beneficiaries; and

· made the minister of finance responsible for determining and laying before Parliament the investment guidelines for the GSF.

3.1.12 This reform appears to have removed the conflicts of priority and purpose embodied in the earlier structure which comprised a minister, a board, a department and a review by Ombudsman. At the time of the review, in 1993, the board comprised 13 appointees. Prior to the reforms, the chief executive was also the superintendent and reported to the minister and the board, with each party having different priorities.

3.1.13 The 1995 amendments should improve accountability, control and performance assessment. It is now absolutely clear that all administrative decisions concerning the GSF are the responsibility of a chief executive who has no responsibility to favour the interests of beneficiaries and/or contributors relative to taxpayers. Beneficiaries' interests are protected by the contractual nature of the scheme and by their ability to appeal to the appeals board against administrative decisions.

3.1.14 Transparency has also been improved since, in addition to the last requirement listed in paragraph 3.1.11 above, the Government Superannuation Fund Amendment Act 1995 requires:

· the GSF's superintendent to maintain an up-to-date published statement of the policies in the exercise of his or her discretionary powers in matters affecting contributors' and beneficiaries' entitlements;

· the GSF to provide each contributor, upon request in writing, with an estimate of that contributor's benefits and/or the GSF's annual report without fee, where such requests are made no more frequently than once every 12 months;

· the GSF to publish an annual report; and

· the government actuary to publicly report on the financial position of the GSF no less frequently than three-yearly.

3.1.15 The following guidelines for the investment of the GSF's assets were promulgated on 17 December 1995 by the Minister of Finance in accordance with the requirements of the legislation:

That the assets of the Government Superannuation Fund be invested in fixed interest and other debt securities (such as Government or Government owned and part owned and local government bodies' issues and corporate investment grade debt securities); indexed securities; synthetics (such as swaps, futures, options and forward rate agreements) where these are used for hedging purposes only; and GSF member housing mortgages where these meet portfolio requirements.

The Fund's assets may be lent on a fully secured basis for participation in the New Zealand debt market, within the constraints of the investment guidelines.

3.1.16 Reflecting historical practice and the Minister of Finance's current directive, the assets in the GSF are currently entirely invested in fixed interest and debt securities. At 30 June 1996, the GSF's $3.1 billion net assets were invested in government stock, government guaranteed debt and other Crown debt (64.5 percent), corporate and other debt bonds (15 percent), short-term investments (16 percent), local authority stock (4 percent) and mortgages (0.5 percent). We understand that while the GSF measures its performance against the First NZ Capital New Zealand Government Stock Index, it does not seek to match any given bond market index passively. Instead it manages the trading portion of its portfolio actively to outperform the index.

3.2 Outstanding Issues: Overview

3.2.1 As noted in paragraph 3.1.8, the GSF exposes the taxpayer to major costs and risks.

3.2.2 Some options for reducing these risks may involve material changes in structure ñ for example, moving to a more fully funded investment management operation, merging in part or in full with the NPF, or an outright tender of the Crown's commitments to retirees.

3.2.3 One option, within the current legislative structure, would be for the GSF to contract out the provision of investment management services. This option is discussed in section 3.3 below.

3.2.4 The GSF's and the NPF's defined benefit schemes differ markedly in respect of the degree of funding and therefore the size of the actuarial deficit. The actuarial deficit in the DBP Annuitants Scheme is somewhat less than its net assets, whereas that in the GSF is four times its net assets.

3.2.5 The minister of finance's investment guidelines for the GSF require it to adopt a dramatically different asset allocation from that of a conventional defined-benefit superannuation fund. Some information on the asset allocations adopted by local institutional fund managers is provided in paragraph 2.1.19. In Europe, the Global Investor, November 1990, reported that the top 100 pension funds held typically something of the order of 60 percent of their funds in equities. In the United States, equity weightings of the order of 25 - 35 percent appear to be typical.

3.2.6 Further, the GSF's asset allocation benchmark differs markedly from those applying to the NPF's defined benefit schemes. The latter have 65 percent allocated to equities, 15 percent to property and 20 percent to domestic fixed interest and cash. Both the NPF's DBP Contributors and Annuitants Schemes, which bear the closest relationship to the GSF's scheme, have this weighting. These allocations contrast sharply with the 100 percent weighting of the GSF's investments in cash and fixed interest.

3.2.7 The issue of the optimal asset allocation for the GSF's investments is likely to be closely related to the issue of the optimal level for the unfunded liability. These issues are considered in section 3.4.

3.2.8 As there is with the NPF, there is an important ongoing need for adequate monitoring of decisions from a taxpayer perspective. Risks to the Crown exist given the open-ended government guarantee and the scope for wealth transfers through rule amendments, interpretations and reinterpretations. For example, the scheme is always open to lobbying by existing members for rule changes or rule clarifications which expand scheme benefits relative to contributions. The Crown needs to be vigilant in representing taxpayer interests in such situations. The superintendent of the GSF (who is also the chief executive of the department) can exercise some discretion, applying the criteria of members' interest and Crown interest to ensure any changes or new policies are fair and equitable and, as far as possible, do not increase the Crown's costs.

3.2.9 In addition, the size of the actuarial deficit may be altered by rule changes which affect members' decisions, e.g. in respect of withdrawal before retirement, taking a lump sum on retirement and the like. It may be possible for the Crown to negotiate or implement rule changes or rule interpretations which generate benefits for the Crown and for members.

3.2.10 The Crown can protect itself against the risk that opportunities to reduce the actuarial deficit will be inadequately considered or pursued, through its ability to appoint the superintendent and to determine the superintendent's contract with the minister of finance. Reporting requirements also allow The Treasury to monitor the superintendent's activities from a taxpayer perspective.

3.2.11 Compared with the NPF, there is greater ability for the Crown to contract with the superintendent to protect taxpayer interests. The GSF's superintendent and chief executive does not have trustee responsibilities to scheme members, unlike the NPF board and consequently its chief executive. Instead, the GSF appeals board exists to protect members. While members of the GSF may be more vulnerable in these respects, the Crown has an obligation to meet its commitments to members.

3.2.12 Institutional reforms could also reduce costs to the taxpayer and/or fiscal risks. Section 3.5 considers the scope for amalgamating the GSF and the NPF. Section 3.6 considers the radical option of reducing Crown risks by selling the actuarial liability.

3.3 Contracting Out of Investment Management and Custodial Services

3.3.1 The Government Superannuation Fund Amendment Act 1995 permits investment management and custodial services to be contracted out from 30 June 1996. However, no such action has been announced yet.

3.3.2 The $948,000 voted for investment management (and custodial) services is extremely small from a commercial perspective, being only 0.03 percent of the GSF's $3.1 billion net assets. This is
3 basis points. A comparable figure is not readily available for the NPF schemes since the costs of each fund manager and the custodian are netted out from reported investment income. Even the world's major fund managers (who should benefit the most from economies of scale) incur larger expenses. For example, in 1995 the average annual expense ratio amongst US mutual funds was 1.1 percent, according to Lipper Analytical Services. Index funds have lower expense ratios. The average expense ratio for all S&P 500 index funds in the USA was 0.45 percent in 1994, according to Gruber. He also reports that "bond index funds, small stock index funds, growth index funds, and value index funds were available with expense ratios below 30 basis points" (0.30 percent). The cost of the New Zealand Stock Exchange's passive, indexed, Top 10 Fund is 0.4 percent for retail amounts. It could be appreciably lower for major amounts.

3.3.3 The extremely low cost of the GSF's investment management services reflects many factors, including its different regulatory environment and output mix. For example, it does not have to market its investment services, compete on the basis of relative performance or handle the level of 'deposits' and 'withdrawals' of a commercial mutual fund manager. Nor does it have the flexibility to make major asset allocation decisions. For such reasons, its cost of operation is arguably not directly comparable to those cited above.

3.3.4 While the provision of its existing services should be contracted out if it is economic to do so, any cost savings are likely to be minor in relation to the scale of the GSF's operation, because the dollar amounts are relatively small. Broadening the asset mix would materially raise investment management costs, and expected returns.

3.4 The Asset Allocation Decision

3.4.1 The GSF is currently reviewing its investment criteria, including its asset allocations. The asset allocation decision depends on the resulting investment objective and the strategy for pursuing this objective. One option would be for the government to simply take control of these funds by requiring them to be fully invested in central government debt, perhaps on a passive basis. The Crown would then be fully responsible for the ultimate investment of those funds. The Debt Management Office would be the Crown's likely agent in this respect. This option appears to be close to the current GSF philosophy.

3.4.2 At the other extreme, the Crown could fully fund the GSF by issuing it with sufficient government stock to eliminate the actuarial deficit. The GSF could then be made responsible for trading this portfolio so as to achieve a more conventional investment portfolio for a defined benefit superannuation fund. We consider some of these options in more detail below.

3.4.3 In this section we consider five alternative strategies for investing the GSF's assets. Four of these assume no change in the actuarial deficit. The fifth relaxes this constraint. The five alternatives comprise:

(i) passively investing the existing $3 billion of assets in the market portfolio of government stock - in effect defeasing the debt from a Crown balance sheet perspective;

(ii) as in (i) but allowing limited departures from the government stock benchmark portfolio for active management purposes;

(iii) investing the existing assets so as to minimise the expected volatility in the actuarial deficit;

(iv) widening the investment classes in which the $3 billion can be invested to include equities and/or property; and

(v) fully funding the GSF with government stock, thereby eliminating the actuarial deficit. The GSF would then be permitted to trade that stock until it achieved an optimal investment portfolio. This would no doubt include a significant investment in equities.

3.4.4 These options represent markedly different approaches. Which is likely to create the most value for society? Value can be created by reducing transaction costs, increasing expected returns for a given level of risk, and/or reducing risk for the same return. The ability of fund managers to add value from risk management activities depends on their capacity (and incentive) to identify and price risks accurately and to manage those risks effectively once identified. Most actively managed funds do not outperform the market index (but this is partly because of transaction cost expenses).

3.4.5 In the private sector, investing in riskier assets in order to increase expected returns does not, in itself, create any value. Such an activity only improves outcomes if the riskier portfolio better satisfies policyholders' risk tolerance preferences and if they have no more efficient way of satisfying these preferences.

3.4.6 Individuals' shares in superannuation funds and in the government's portfolio of assets and liabilities are part of an individual's overall portfolio. Whatever the investment strategy adopted by a superannuation fund, or by the government in respect of its own portfolio, individuals may be able to 'undo' the effects of that strategy on their own overall portfolio by making offsetting changes to the components of the portfolio which are under their full control. This means that an optimal risk-return strategy for components of a portfolio cannot be inferred from a knowledge of individuals' overall risk tolerances if they do indeed have the flexibility to make such offsetting adjustments.

3.4.7 The case for having superannuation fund managers adopt an asset mix which produces a risk-return portfolio that accords with the overall risk-return preferences of the representative individual, is strengthened by the degree to which individuals are constrained from being able to offset costlessly the effects of decisions made by superannuation fund managers or government portfolio managers on the expected risks and returns of their overall portfolio.

3.4.8 In the competitive market place, competition would, and no doubt does, resolve the issue of the optimal asset mix for a defined-benefit superannuation fund. The more risky the asset mix, the lower the expected future employer and employee contributions, and the greater the uncertainty about how far the ex post required combined contribution will deviate from the ex ante expected contribution. The evidence that superannuation funds typically have significant investments in equities (see paragraph 3.2.5) should be treated with respect for this reason.

3.4.9 In the competitive market place, the risk of employer bankruptcy might conceivably affect the market outcome. Risks arise in this situation for employee contributors if the company goes bankrupt when there is a large actuarial deficit in the superannuation fund. Such a deficit might arise if the failing company can raid the company superannuation fund in an attempt to hold off its debtors or if the fund is invested significantly in the company's shares. However, as long as the fund cannot invest in the company's own shares, it is hard to see how the risk of bankruptcy could materially affect industry norms for the proportion of superannuation funds invested in equities.

3.4.10 When benefits are government guaranteed, all investment risks are borne by taxpayers. This raises the possibility that the optimal asset mix for a government-guaranteed scheme might be different from that which would emerge from a competitive market in the absence of such a guarantee.

3.4.11 When taxpayers share in the risks, the asset allocation decision could affect the deadweight costs of taxation. If the GSF's assets are invested at an unsatisfactorily low return, future taxes will have to be commensurately higher. Conversely, investing the GSF's assets in higher risk/higher expected return securities would reduce expected future taxes. This factor might favour higher weightings in riskier investments than would otherwise be warranted.

3.4.12 Deadweight losses could be material. For example, if the equity market risk premium is of the order of 9 percent per annum and the deadweight cost of taxation is a conservative 20 cents in the dollar, the expected deadweight loss from investing $3 billion in debt rather than equities could be $54 million a year. This dwarfs any increase in the GSF's operating expenses which would result from extending its investments into equities.

3.4.13 An offsetting factor is that the greater the uncertainty about future returns, the greater the uncertainty about likely future tax rates. Public finance theory suggests that it is plausible that volatility in tax rates through time and across states of nature could reduce national welfare, other things being equal. The optimal equity weighting would not necessarily therefore be 100 percent, even if it were greater than zero.

3.4.14 A further point is that the government can, in principle, reduce expected future taxes by investing in higher-returning assets independently of the GSF. The Crown could, for example, invest the proceeds from bond sales in higher-returning equities. This should reduce the expected future tax burden. In the extreme, all taxes might be eventually eliminated by such a strategy (if actual returns reflect expected returns).

3.4.15 No doubt there are good reasons why governments around the world do not generally adopt such a high-risk strategy. Problems associated with voting, principal-agent and capture in government are such as to create considerable concern that governments will tend to squander surplus income and/or assets. This is why the New Zealand Business Roundtable opposes the retention of surplus assets by local authorities - the representatives of which argue that such investments are desirable because the income from them reduces rates. The Business Roundtable's argument for divestment has added force if the ownership stake is so large as to make the government a direct investor rather than a passive portfolio investor.

3.4.16 These public choice arguments against a strategy of reducing future taxes by investing in relative risky assets have less weight in the context of the disciplines inherent in a professionally structured, arms-length pension fund. Therefore, we do see a case on these grounds that investment in equities by the GSF could improve national welfare. The most powerful argument in our view is the evidence that superannuation funds tend internationally to have equity weightings.

3.4.17 Another issue which could be relevant to the GSF's optimal investment strategy is the possibility that some asset mixes might hedge the taxpayer against shocks to the fund's liabilities. Specifically, a shock which increases the actuarial liability will be less costly for the taxpayer if it also raises the value of the Crown's assets, of which the GSF's investments are effectively a part. There are at least three reasons why it may be inefficient to leave it to taxpayers to hedge against such risks on an individual basis. The first is that many may lack the information necessary to undertake this task. Secondly, many taxpayers may lack the financial means to hedge. Finally, and most importantly, taxpayer hedging cannot guard against the increased deadweight costs from the greater tax burdens which result when an adverse shock occurs and the Crown has not hedged its own portfolios.

3.4.18 Shocks to the GSF's actuarial liabilities could arise inter alia from:

(i) higher than expected real wage inflation shortly before retirement so that the real starting pension is raised without commensurate increases in pension contributions prior to retirement; and

(ii) lower real interest rates which increase the present value of future inflation-indexed annuity payments.

3.4.19 In addition, shocks to the actuarial deficit could occur if changes in nominal interest rates, with unchanged real interest rates, were associated with changes in the value of the GSF's bond or equity investments. (We would not expect such nominal interest rate shocks to affect the GSF's actuarial liabilities much since these are largely indexed.) Real interest rate shocks could of course also influence the value of the GSF's investments and therefore the actuarial deficit.

3.4.20 Effect (i) increases the already high exposure of the Crown, as a major employer, to real wage increases. It would be difficult to find a financial exposure which could help hedge the Crown's balance sheet against real wage risks, but there is some tentative evidence that equities provide a hedge against real wage increases. Further research into this relationship would need to be undertaken if it is significant to the equity decision.

3.4.21 In respect of point (ii) in paragraph 3.4.18, a drop in real and nominal interest rates would increase the market value of the Crown's public debt liability and the present value of future real pension payments. The Crown may thus be exposed to considerable interest rate risk. Shorter-duration Crown debt would reduce the net exposure. Very long-duration fixed interest investments by the GSF would also reduce the exposure. But long-duration investments may not be available in sufficient quantity to make a material difference.

3.4.22 This brief discussion hints at the complexities involved in determining how different asset mixes might hedge against the risk that real wage and/or interest rate shocks might increase the size of the actuarial deficit. The less confident advisers are that any particular asset mix is a better hedge against shocks to liabilities than alternative mixes, the less relevant is this criterion.

Option (i): The Notional Investment Fund

3.4.23 Option (i), the notional investment fund, could be thought of as an administratively convenient way of reducing the GSF's asset-management activities to a bookkeeping operation. As such it effectively makes the DMO responsible for all asset allocation decisions arising from the existence of the $11 billion GSF liability. This is because the DMO is the agency with prime responsibility for advising on the optimal duration and currency composition of Crown debt.

3.4.24 This option avoids the situation of having two authorities responsible for asset allocation decisions in respect of the GSF - the GSF in respect of the $3 billion and the DMO in respect of the management of the Crown's overall portfolio which includes the unfunded $8 billion liability. Possible coordination problems are therefore avoided and some administration costs may be saved. There may be a case for two authorities if each has superior information about different aspects of the situation and/or if the constraints on the two differ in material ways. In respect of the latter point, the size of the unfunded liability currently constrains, to a significant extent, the GSF's ability to use its assets to hedge shocks to liabilities (refer to the discussion below on option (iii)). However, the DMO would find it very difficult to invest in domestic equities due to the implicit government guarantee which such government investments confer on an investment partner. Investments in global equities would reduce this moral hazard problem - if future governments sustained that asset allocation.

3.4.25 The above discussion is somewhat analogous to the debate in the United States about the degree to which company defined-benefit schemes for employees should be fully funded, with the funds being inaccessible to the company, and about whether the company's treasury operation should control all investments. Given the risk of company bankruptcy and the conflict of interest between shareholders and scheme members in respect of such assets when the company is in financial difficulty, it is necessary to have rules to protect members' interests. The situation differs where the employer is the Crown because, conversely, the Crown cannot escape any liability to pensioners. Any such case for a separate fund in order to protect the interest of members is therefore more a matter of sentiment and perception than of economic substance.

3.4.26 However, there are obvious reasons why a company's treasury operation or the DMO would be unlikely to invest company or taxpayer funds in equities. It is not their core business, they do not have this expertise and they would have to hire it. The activity would distract them from their core tasks of managing company or Crown cash flows, debt and liquidity. Further, there is a risk that mixing the two roles might lead to confusion as to how much risk-taking was appropriate. If investing in equities is appropriate in order to reduce the future cost to the employer of funding pensions, why would it not be appropriate to do this for other purposes as well? One reply might be that borrowing to invest in equities is too risky. But if the company or the Crown is already a borrower, as is likely, then from a consolidated point of view it is already borrowing in order to invest in equities through the pension fund. Another argument is that if shareholders or taxpayers can borrow to invest in equities for themselves, does the company or Crown add any value for them if it does this on their behalf? The company may have particular borrowing, information or tax advantages. If so it will probably be in the investment company business anyway. Otherwise it will probably stick to its knitting and decide not to undertake this activity.

3.4.27 As discussed above, public finance theory provides two in principle arguments for having the Crown invest in diversified asset portfolios including equities. One is that such investments might be useful in hedging against shocks to the tax base - thereby reducing the welfare losses caused by volatility in tax rates. The second is that the higher the Crown's net investment income is, the lower average and marginal tax rates can be and, therefore, the deadweight losses of taxation. Note that taxpayers who invest in equities themselves would not thereby reduce either of these welfare losses. These arguments take us beyond the scope of the review of the GSF, but they do illustrate the degree to which confusion could arise about the DMO's ultimate asset allocation objectives once it got into investing in equities for superannuation fund reasons.

3.4.28 Given the obvious potential dangers from having governments borrow in order to invest in equities, and the difficulties of determining what level is excessive, the most likely outcome of the notional approach is what we are seeing - the DMO will not invest in a diversified portfolio of equities and any adviser aware of the dangers would hesitate to advise it to do so. We consider the potential benefits from having the GSF invest in equities in the discussion below about option (iv).

Option (ii): Active Management Around a Notional Benchmark Portfolio

3.4.29 There would not appear to be any major problem in principle with allowing fund managers to add value from active management. By 'add value' we mean increase the expected return relative to the portfolio benchmark without increasing portfolio risk. The difficulties of doing this should not be underestimated.

3.4.30 This option most resembles the status quo. However, more clarity about the benchmark would be desirable if the status quo is to be maintained. Currently, for example, it seems possible that the GSF could aim to outperform its benchmark simply by having sufficiently more risky, higher yielding assets.

Option (iii): Minimising the Volatility in the Actuarial Deficit

3.4.31 One option would be to seek to invest the $3 billion in net assets so as to minimise the probability of a future blow-out in the actuarial deficit arising from financial shocks. However, investing $3 billion to hedge against shocks to an $11 billion liability would be a heroic, and arguably unachievable, strategy. This is because a $3 billion fund would have to be invested in a highly leveraged manner in order to minimise the volatility of a much larger $8 billion actuarial deficit.

3.4.32 To take a purely illustrative example, suppose that an adverse 100 basis-point interest rate shock would be expected to increase the GSF's liabilities by 9 percent from $11 billion to $12 billion. An asset allocation-based hedging policy designed to prevent such a shock from increasing the actuarial deficit would require the same shock to increase the value of the fund's investments by $1 billion. In other words, a 33 percent gain would have to be achieved on a $3 billion investment portfolio in order to offset a 9 percent loss on the pension liabilities. The duration of a $3 billion fund might have to be of the order of 35 to 40 years to produce such a 33 percent gain. Finding an investment with such a long duration could be very difficult. In contrast, the duration of a $11 billion fund might have to be only a much more feasible 10 years in order to produce the required $1 billion gain.

3.4.33 This example also illustrates the point that the smaller the degree to which the GSF is funded, the more difficult it would be to use those assets to effectively hedge against shocks to its liabilities. In the limit the tail cannot, in practice, wag the dog. Indeed, the above calculation suggests that a $3 billion fund may well be too small to allow the goal discussed in this subsection to be meaningfully achieved using the fund.

3.4.34 Nevertheless, the logic is applicable even if dollar-weighted duration matching is unattainable. The longer the duration of the GSF's assets, the better the hedge against interest rate shocks.

3.4.35 The goal of minimising the volatility of the actuarial deficit could be pursued more readily therefore if the GSF were more fully funded or if the DMO adjusted the duration of Crown net debt so as to obtain the required hedging effect. The first possibility is discussed below as option (v). The second possibility is discussed in option (i). In terms of the above illustrative example, the DMO might seek to shorten the duration of net public debt so that its market value increased by $1 billion less than would otherwise have been optimal. Given the much larger sums associated with gross and net debt, such an adjustment should be feasible.

3.4.36 A more fundamental question is whether the objective of minimising the volatility in the unfunded liability is optimal. An alternative (risk minimising) goal might be for the GSF's assets to be invested so as to assist the Crown to minimise the expected volatility of future tax rates. Given current government expenditure plans, these tax rates depend on the volatility of assessable income, economy-wide. Taxpayers might prefer investments to be chosen so as to hedge against future shocks to national income rather than to hedge against shocks to the GSF's actuarial liabilities.

Option (iv): Permitting Investments in Higher-yielding Assets

3.4.37 As noted above, the conventional argument for investing a proportion of the funds in equities is that this would be expected to reduce required employer-employee contributions or, as in this case, the actuarial deficit in the fund, because equities have higher expected returns than do fixed interest investments. Fully-funded employer-sponsored superannuation funds commonly have significant equity exposures for this reason.

3.4.38 Investing a $3 billion fund in equities rather than risk-free investment would increase the expected return by $270 million in any one year - assuming a market risk premium of 9 percent per annum. This would materially help fund the annual government subsidy which is projected to be $694 million for 1996/97 for the general scheme. The larger the fund's assets the greater would be the effect.

3.4.39 On the other hand, equity investments would raise management expenses appreciably (see paragraph 3.3.2). They would also markedly increase the expected volatility of the actuarial deficit because the annual return from New Zealand equities has an expected standard deviation of around 24 percent per annum against an annualised standard deviation of around 1.6 percent for money market investments and 5 percent for New Zealand government bonds. When low-risk money market interest rates are around 8 percent, there is roughly a one-sixth chance in the next year that investing all the GSF's $3 billion of assets in money market investments would produce an annual return below 6.4 (=8-1.6) percent, i.e. $190 million. On the other hand, if all the assets were invested in equities, there would be a one-sixth chance that the return would be less than minus 7 (=8+9-24) percent, i.e. below minus $210 million. There is no free lunch here - only the long-recognised trade-off between risk and return.

3.4.40 One commentator on an earlier draft of this paper thoughtfully observed on this point that:

One view of the asset allocation guidelines currently in place is that they are more a question of risk management for the GSF Minister and management rather than a sensible long-run asset mix consistent with the liabilities of the fund. The 'no-risk' approach in investment decision making limits the political mileage possible when investment markets perform poorly, as they will from time to time, and with many commentators in that time acting with perfect hindsight after the event.

3.4.41 Proponents of equity investments sometimes argue that they make more sense the longer the time horizon. They assert that this proposition is true because the probability that equity investments will outperform fixed interest investments tends towards unity as the investment period tends to infinity. However, the assertion that the weighting of equities in a portfolio should be greater the longer the time horizon for this reason, is invalid as Nobel Prize winner Paul Samuelson has explained at length. The fallacy arises because the probability of outperforming fixed interest investments is not the only thing which counts for risk averse investors. Such investors will not regard a smaller probability of under-performing as unequivocally a good thing if the size of the possible loss in the event of under-performance also increases with the length of the time horizon.

3.4.42 It is hard to see a compelling case for the current guidelines which prevent the GSF from investing in equities. The level of expertise required would not be great if, for example, it were simply to buy units in a passive index fund. Alternatively, it could contract out more complex approaches to equity investments. While transaction costs are greater for equity investments, the significant equity weightings in superannuation funds around the world indicate that equity investments are an efficient means of funding superannuation liabilities.

Option (v): Fully Funding the GSF

3.4.43 As argued above, the DMO cannot effectively invest in equities for moral hazard reasons and the GSF is constrained from doing so by the largely unfunded nature of the scheme and the minister of finance's investment guidelines for the funded portion. The upshot is that the approach adopted to date by the government does not permit the investment flexibility of a conventional superannuation scheme.

3.4.44 An alternative approach would be to fund the GSF more fully (or completely) by issuing it with additional government stock and allowing the investment manager to trade that stock in the market place, buying equities if permitted to do so. As explained above, equities would provide a higher expected return at the cost of greater uncertainty about the future amount available. As such, equities would reduce the expected actuarial deficit while increasing Crown risk, compared with fixed interest investments. Perhaps more fundamentally, they might also help hedge against financial shocks which could increase the GSF's actuarial deficit, if this is a desirable objective.

3.4.45 It would be important in this debate to ascertain the degree to which equity and other investments by the GSF would materially hedge the Crown against the risks it faces from financial shocks to the actuarial deficit. This could be the subject of further research, but the reality is likely to remain that defined-benefit pension funds which provide fully-indexed benefits expose the plan underwriter to large residual risks.

3.4.46 Members of the GSF should be essentially indifferent about these issues because the Crown guarantee protects them from all risks. The issue is therefore one of the taxpayers' welfare. From a taxpayer perspective the government would be issuing bonds in order to invest in equities through the GSF. On a global basis, private investors would be selling equities to the government, thereby ultimately funding global private sector purchases of New Zealand government bonds. Expected future tax payments would fall, but uncertainty as to the actual level of future tax payments would increase. The amount of such risks would be limited by the overall size of the GSF and by the asset allocation benchmarks it adopted. Arguably it would adopt similar benchmarks to those adopted by the NPF for its defined benefit schemes. Differences in risk characteristics are one factor which might warrant a different asset allocation.

3.4.47 Fully-funding the GSF would not alter Crown liabilities or Crown net worth because the Crown Financial Statements already include the actuarial liability in these aggregates. Presentational issues do not therefore pose a constraint in respect of these aggregates. However, gross and net public debt would rise.

3.4.48 Equity investments made within the confines of a well-delegated superannuation scheme should not cause the same level of concern, as would generalised DMO investments in equities, about confusion of objectives, excessive exposure to risk and the possibility that any gains will, in fact, be used to increase expenditures rather than to reduce the deadweight losses of taxes. Indeed the higher reported public debt ratios could conceivably inhibit spending decisions.

3.4.49 The implementation of this option could well see much of the additional government stock sold to overseas investors and much heavier investments by the GSF in offshore equities than in New Zealand equities. For example, the NPF's investment guidelines for its major defined benefit schemes require $4.20 to be invested in offshore equities for every $1 invested in domestic equities. More relevantly, the New Zealand sharemarket is commonly less than 0.2 percent by value of world sharemarkets, so a fully diversified market portfolio of equities would have a minimal exposure to New Zealand. If the GSF had 5 percent of its liabilities invested in the New Zealand sharemarket it would only own about 2.25 percent of the current value of that market. This should not be a concern as long as its holding was well-diversified and clearly not subject to political direction.

3.4.50 Selling New Zealand government bonds to foreigners in order to invest in global equities would involve some transaction costs but there is no reason to expect it to have any macroeconomic effects (e.g. on interest or exchange rates) as long as it was not done in a way which raised doubts about the government's attitude to foreign investment or its resistance to higher future rates of inflation. A well-managed process should not cause any difficulties.

3.5 Amalgamating the GSF and the NPF

3.5.1 A more fundamental reform would be to amalgamate all or some of the functions of the NPF and the GSF. Currently, for example, both are responsible for the administrative and investment services associated with government-guaranteed defined benefit schemes. Amalgamation would have the potential to reduce costs by realising economies of scale and/or eliminating some duplication. In this section we consider in turn the three areas of: investment management, schemes administration and head office costs.

3.5.2 As noted in section 2, economies of scale in funds management will tend to come at the cost of increased exposure to a smaller group of fund managers. However, a rule which restricts the proportion of total funds that can be put under the control of any single fund manager will be less costly in terms of forgone economies of scale the greater the size of the fund.

3.5.3 The full benefits of any economies of scale are only likely to accrue to fund sponsors or members if management contracts in both schemes are tendered simultaneously. Otherwise the incumbent in the contract not being tendered will be the only bidder who can immediately achieve economies of scale. This would make bidding less competitive than would otherwise be the case.

3.5.4 Amalgamation might also facilitate standardisation of service quality and other parameters where this would be beneficial to and consistent with existing contracts. The marked differences noted in section 2.1 between the cost structures of the NPF and the GSF, at least in respect of head office expenses and scheme administration, are notable in this respect. For example, as noted in paragraph 2.1.15, the NPF's costs in respect of scheme administration and head office expenses totalled $16.8 million in 1995/96, or $165 per member (based on 27,000 pensioners and 75,000 contributors). This is much higher than the GSF's costs for scheme administration. Including related head office expenses, the GSF's costs for 1996/97 totalled $3.928 million, or $44 per member (based on 45,610 pensioners and 44,458 contributors).

3.5.5 We have been unable to document industry norms for such expenses, but, with the assistance of Frank Russell, we have spoken to an industry expert in Australia who suggests that costs per member might normally fall in the A$100-180 range. This information, which should be understood to be informal, imprecise and indicative, suggests that the GSF's costs are abnormally low.

3.5.6 The cost differences are more marked when only defined benefit schemes are compared. The NPF's schemes which are most similar to the GSF's schemes are its DBP Contributors and Annuitants Schemes. The NPF allocates $2.6 million in scheme administration and head office expenses to these two schemes. This is two-thirds of the $3.9 million spent by the GSF on these activities. But the NPF's two schemes have only 14 percent of the membership of the GSF's schemes, at 13,185 versus 94,000. The cost per member of the NPF's schemes is $197 against the $44 for the GSF's schemes.

3.5.7 Since, in each case, costs have largely been determined by competitive tender, we presume that these differences represent the different cost structures of these two schemes as perceived by bidders at the (different) times of the two tenders. They presumably reflect differences in timing, service levels, clientele, head office monitoring, complexity of structures in relation to tax and asset allocation issues and the number of employers, and bidding strategies in the tenders. It will be interesting to see if the cost margin narrows when contracts come up for revision or re-tendering.

3.5.8 One issue for The Treasury to consider appears to be the issue of service quality. Differences in service quality should affect costs. Taxpayers have an interest in service quality if they effectively fund any costs, through the government guarantee and the actuarial deficit, which arise from a higher level of service quality than contractual arrangements with members require.

3.5.9 To the extent that service quality does differ, consideration would have to be given as to which level of service quality should prevail. Contractual commitments aside, it is hard to see why taxpayers would wish to fund a different level of service quality to annuitants who are members of the NPF's DBP Annuitants Scheme than to members of the GSF's schemes.

3.5.10 Should amalgamation at head office level be desirable, a number of factors favour having the NPF take over the GSF (rather than having the GSF take over the NPF). First, the NPF's structure is more in accord with the government's fundamental philosophy of withdrawing the government from activities which are commercial in nature and in which it does not have a comparative advantage. Indeed, the NPF's current structure was designed for this purpose. Second, the GSF has no experience with running defined contribution schemes and cash accumulation schemes, but these are important parts of the NPF. Third, the protections offered by the trustee arrangements embodied in the NPF's structure are no doubt important to members of the NPF's schemes. They could also be useful in distancing any equity portion of the merged funds' investments from the possibility of political interference. These factors would favour having the NPF take over responsibility for contracting out the administration and investment management responsibilities for the GSF's schemes, rather than the other way round.

3.5.11 On the other hand, there are some important offsetting considerations. The departmental structure arguably gives the minister of finance greater control of administrative decisions than would the NPF's trustee structure. This is relevant in that the government is involved as an employer (i.e. plan sponsor) in respect of the GSF's schemes but not the NPF's schemes. It may not be desirable to involve the NPF in a situation which could confuse the Crown's purchase interests with its interests as a guarantor. One option here would be to leave the GSF's superintendent role, and the associated appeal board function, outside any merger. Another concern is that the government would need to be satisfied that the NPF's cost allocation rules did not result in taxpayers bearing a higher burden of its costs than at present. This might happen, for example, if the NPF allocated more of its common costs to the defined benefit plans after any merger and less to its defined contribution plans.

3.5.12 Complete amalgamation would require changes in the NPF's legislation. In addition, the Government Superannuation Fund Act 1956 allows the prime minister to determine which department of state is responsible for the administration of the Act, but the NPF is not a department of state. Amalgamation would allow the Crown, as plan sponsor, to contract with the NPF in respect of asset allocation benchmarks, investment strategy and costs and quality of services offered. Issues which would need to be worked through would include: the continuing role for the Government Superannuation Appeals Board; whether the NPF's board would be given a trustee role in protecting members of the GSF; and the relationship between the superintendent of the GSF and the executive officer of the NPF.

3.5.13 A partial amalgamation might involve merging the administration and investment management of the defined benefit schemes. This would raise the question of the desirability of separating the GSF's schemes into contributors schemes and annuitants schemes as has been done by the NPF.

3.5.14 Amalgamation would involve costs. While there may be some savings in head office costs, an important question is the extent of economies of scale in the amalgamation of schemes administration or funds management. Weak evidence that the economies of scale in schemes administration may not be major is provided by the fact that Jacques Martin New Zealand did not win the contract for the scheme administration of the GSF.

3.6 Selling the Liability

3.6.1 A radical option would be for the government to pay a reputable and secure third party (that is, one with a high credit rating) to assume responsibility for meeting the Crown's obligations to members of the GSF's schemes. The price would be determined by competitive tender. The successful bidder would be the one who would take over responsibility for paying the pensions and lump sum amounts to current and/or future annuitants for the smallest fee. One can envisage a contractual arrangement in which the payments would be made quarterly and subject to conditions about the maintenance of a target credit rating by the successful party and penalties for failure to meet these conditions. This would be close to a 'contract for differences' arrangement.

3.6.2 To be successful, such a sale process would need to:

• protect the taxpayer adequately against the possibility that the bidder might subsequently default on its commitments;
• satisfy members of the GSF that their interests were protected adequately and/or that they were compensated adequately for any loss of the government guarantee; and
• satisfy potential bidders that any discretion which could be exercised by the superintendent of the scheme would not unduly increase expected scheme costs.

3.6.3 Addressing these problems satisfactorily would not be easy. In respect of the second point, the compensation required for the loss of the guarantee could be material. Perhaps the government guarantee would have to remain in place ñ reduced effectively to a low probability underwriting commitment to members. In respect of the third point, bidders might fear that the Crown's agent would be more likely to exercise discretion in favour of scheme members once the Crown had sold its liability to scheme members.

3.6.4 Nevertheless, if these problems could be addressed satisfactorily, a very considerable risk would be removed from the Crown's balance sheet, and every option should be considered. This option is more attractive the less confident the government is about which investment mix would be in taxpayers' best interests. It could eliminate the need for further consideration of the investment mix and full-funding issues just canvassed. These would become a private sector responsibility.

3.6.5 A more limited and arguably more feasible option would be simply to sell liabilities in respect of members as they become pensioners. This would avoid the complication for bidders of evaluating the risks that: employers might either default on their obligations to make contributions in respect of contributing members or find legal loopholes which reduce those obligations, but not employee benefits. For example, the GSF Department's post-election briefing document identifies risks arising from the fund's exposure to the Cook Islands Government (which owed it $6.532 million at 31 October 1996) and from employers who have been commercialised, become Crown entities, or privatised.

3.6.6 The selling or merging of liabilities in respect of annuitants would require separating the GSF's schemes into contributor and annuitants schemes. This may be a complex process.

3.7 Concluding Comments

3.7.1 The government has reformed the GSF by closing the scheme to new members; contracting out scheme administration; clarifying the roles and accountability of the members of the superannuation board, the superintendent of the GSF and the minister; and improving accountability by requiring the GSF to disclose specified information.

3.7.2 Given the open-ended nature of the government guarantee of the GSF, the Crown needs to be vigilant in monitoring events and resisting rule amendments and interpretations which could transfer wealth from the Crown to scheme beneficiaries.

3.7.3 The GSF continues to provide investment management and custodial services itself. The amount being spent on these services at 0.03 percent of the GSF's $3.2 billion net assets is very small from a commercial perspective and indicates that its activities in this respect are relatively limited. The contracting out of these services should be investigated, but any savings are likely to be minor in relation to the scale of the GSF's schemes.

3.7.4 The current investment strategy is based on an asset allocation which differs markedly from that of a conventional defined benefit fund in Europe or North America. It is also dramatically different from the NPF's counterpart scheme, the DBP Annuitants Scheme. There may be some scope for moving the GSF's asset allocation from predominantly fixed interest investments to other allocations that hedge taxpayers better against future shocks to the actuarial deficit, or that enhance the returns generated by the scheme. Deciding on an optimal asset allocation would be complex, but it is hard to see a compelling case for the current government-issued investment guidelines which prevent the GSF from investing in equities.

3.7.5 A number of options for changing the GSF's investment strategy exist. The GSF could be required to invest solely in government debt, transferring responsibility for investment decisions to the DMO. Alternatively, the GSF's fund managers could be allowed to manage the fund actively around a notional benchmark portfolio; the $3 billion fund could be invested to minimise the volatility in the actuarial deficit; investments in higher yielding assets could be allowed; or the GSF could be fully funded and operated like a traditional funded superannuation scheme.

3.7.6 There is a plausible case for more fully funding the GSF and allowing it to adopt an asset portfolio which has an equity component for growth and a duration which matches more closely the duration of its liabilities.

3.7.7 Consideration should also be given to merging the GSF and the NPF. Having the NPF take over the GSF has some advantages, but there are some offsetting considerations. An issue here would be to ensure that taxpayers were not penalised as a result of the NPF's higher cost structure.

3.7.8 Tendering the obligation to meet the government's commitments to the GSF's pensioners is the most radical option. This option is more attractive the less confident the government is about which investment mix would be in taxpayers' best interests. This approach may require the GSF to split its schemes into contributors and annuitants schemes.

Chapter 4 The Public Trust Office

4.1 Introduction

4.1.1 In this chapter we review the Public Trust Office (PTO). Section 4.2 begins by providing some background information on the PTO. Section 4.3 examines the case for government involvement in the provision of trustee services. Non-neutralities in the treatment of the PTO are examined in section 4.4 and options for future governance of the PTO are discussed in section 4.5. Concluding comments are made in section 4.6.

4.2 Background

4.2.1 The PTO was founded in 1872 to provide an independent perpetual state-backed trustee business at a time when it was believed that such services were not provided adequately by the private sector. It was established under the Public Trust Office Act 1872 (now the Public Trust Office Act 1957) and has as its chief executive the public trustee.

4.2.2 The public trustee is technically a corporation sole giving the office the status of a legal entity which is separate and distinct from the Crown. The PTO is a department within the public service. The chief executive is directly accountable to a minister and has a formal performance agreement which is reviewed in detail by the State Services Commissioner. The minister presents a report and the PTO's annual accounts to parliament each year.

4.2.3 Because the PTO is self-funded and the government's ownership role is uncertain, the PTO has not been subject to the checks, balances and accountability procedures associated with annual budgetary processes. As a result of it being self-financing, the PTO does not have a purchase agreement nor is it subject to the processes for appropriation of public funds. The PTO is not legally required to conform with the reporting requirements of the Public Finance Act 1989, although it has voluntarily adopted planning and reporting processes similar to those applying to Crown entities. The chief executive does not currently have to report to a governing board of trustees or of directors. Some external accountability for investment management is provided by the statutory investment board. The public trustee has also announced his intention to establish an advisory board.

4.2.4 In respect of its role in individual estates, trusts, agencies and some other matters, the public trustee's responsibility (enforceable by the courts if necessary) is directly to the beneficiaries or others having an interest in the particular estate or trust. At this level, the PTO's accountabilities are similar to those of any private trustee or private sector trustee corporation.

4.2.5 There is legal uncertainty as to the PTO's ownership. The Treasury and the PTO have received conflicting legal opinions as to whether the government owns the PTO or the public trustee holds the assets in trust for past and present clients of the PTO.

4.2.6 The PTO is a self-funding trading organisation. Capital requirements are funded from reserves. The PTO had $82 million in reserves built up from retained earnings over the years as at 30 June 1996. The PTO pays taxes on its profits but does not pay dividends.

4.2.7 The government provides a guarantee of the PTO's common fund which is akin to a guaranteed line of credit. The legislation provides the legal authority for an advance which would be repayable. None of the guarantees of the PTO which have existed from time to time has ever been called upon.

4.2.8 The Public Trust Office Act 1957 empowers the public trustee to accept appointment as executor, administrator, trustee, guardian, committee, manager, agent, attorney, liquidator or any other appointment of a fiduciary nature. The appointments may be either alone or joint. The public trustee is responsible for sundry statutory duties under 90 separate statutes.

4.2.9 The client services provided by the PTO include the following:

(i) Personal trust services:

• Executorship/will making: This includes providing advice and consultation in making or revising a will and the management (after death) and distribution of estates not involving long-term trusts;
• Personal trusteeships: Involves the making and management of trusts settled during the settler's lifetime (intervivos trusts), testamentary trusts (established in terms of a will) and compensation trusts (primarily funds held for minors from ACC or Social Welfare).

(ii) Corporate trust services: Corporate trusts cover a range of commercial trust and supervisory roles, normally as a requirement of legislation such as the Securities Act 1978 or the Unit Trusts Act 1960. The PTO has restricted its corporate trustee role to unit trusts.

(iii) Agency services:

• Involuntary agencies: Appointments under the Protection of Personal and Property Rights Act 1988 are deemed involuntary agencies. The PTO's role is to manage all or some of the assets of a person subject to the Act or to audit the annual financial statements prepared by other managers appointed under the Act;
• Voluntary agencies: These include investment administration for significant investment portfolios; management of real property; conveyancing services; tax services; engagements "to take care of things", for example for principals in rest homes; engagements for companies, corporations and departments; general agencies; and agencies for investments in the common fund. The PTO also prepares enduring powers of attorney for its customers and manages individuals' financial affairs under these powers as required.

4.2.10 The PTO's role was originally focused on wills and deceased estates. Its business focus has, in more recent years, shifted to 'living' business which involves helping people to plan their financial affairs. Currently, the PTO manages over 50,000 trusts, estates, agencies and funds. It manages around $3.5 billion of assets and around $1 billion in funds. In these fields the PTO competes with the legal profession, trustee corporations and private individuals. The PTO employs around 500 staff who work in the 50 branches and three service centres.

4.3 Public Policy Objectives

4.3.1 As noted in paragraph 4.2.1, the PTO was originally established as a business at a time when it was believed that trustee services were not provided adequately by the private sector. The original intention of the government guarantee of the common fund was to provide security for the PTO's developing business and to provide comfort to settlers, testators and beneficiaries.

4.3.2 Today, there are five other trustee companies (two of which have a common owner) established in New Zealand under the Trustee Companies Act 1967, an established body of trustee legislation, a self-regulating legal profession and protections provided through the Law Society's fidelity fund and professional indemnity insurance. Trustee companies generally deal with estates of over $200,000. Lawyers provide trustee service for deceased estates (but generally do not act as executors). The PTO provides a full executorship service to a large number of smaller estates so that it currently services parts of the market not currently covered by other players. Given the providers that exist in the market, and the probability that in the absence of government support, the PTO would be likely to continue to service some of its market, the primary policy justification for a state-backed PTO with a government guarantee no longer exists. The PTO may not, however, find it profitable to service some of its smaller clients. The case for continued state involvement on social grounds is examined below.

4.3.3 The PTO has voluntarily assumed the following social service roles: it acts as a trustee of last resort, particularly for low valued estates, and it provides a free will-making service (as long as the PTO is named as executor) which provides the benefit of low levels of intestacy. The PTO has no particular obligations to handle intestate estates.

4.3.4 The PTO acts as a trustee of last resort for low valued estates where the cost of administering the estate may exceed the value of the assets. It estimates that the annual cost of the uncommercial trustee of last resort business which it manages, measured in terms of unrecovered costs and/or lost profit opportunities, could be in the order of $3-5 million annually. The PTO recognises that it might undertake some of this work if commercially driven. Commercial trustee companies also provide a small amount of non-economic trustee of last resort services, although the extent of this is limited by commercial imperatives.

4.3.5 Given the high fixed costs of settling an estate (i.e. the costs of undertaking legally required procedures), settlement can exhaust all the value of small estates. However, there appears to be no public policy reason for the government to subsidise the settlement of low valued estates to ensure that some value remains in the estate. If the objective is to assist low income individuals, then subsidising the management of low valued estates appears to be an inefficient way of achieving this, given the alternative of the tax and welfare systems.

4.3.6 There may be a case for the government to subsidise the trustee of last resort role where the cost of settling an estate exceeds the value of the estate (and settlement would be uneconomic for a commercial organisation), if this is the lowest cost approach to achieving settlement. If there is some benefit from subsidising the settlement of small uneconomic estates, this could be achieved by the government specifically purchasing this service on a competitive basis from the PTO or other providers.

4.3.7 The free will-making service has been credited with reducing intestate deaths in New Zealand. The PTO's offer of free will-making is not necessarily inconsistent with a commercial orientation given the commercial/marketing benefits deriving from this approach.

4.3.8 The social justification for subsidising will-making is relatively weak. The costs and benefits of a decision on whether or not to make a will are largely borne privately with few spillover costs to other members of the community. While having a will increases certainty and may avoid costly, time-consuming disputes, as long as these costs are borne privately there is little reason for government involvement. Thus, individuals' decisions as to whether or not to write a will should be respected.

4.3.9 However, arguably there may be some spillover benefits to reducing intestacy. If a person does not make a will, taxpayers may bear some of the (greater) court costs of settling an intestate estate. If the costs of subsidising will-making are lower than the extra court costs incurred and borne by the government if a person dies intestate, then a case for subsidising will-making could be made (at least subsidisation to the extent of the court costs saved by the government). However, it would be difficult to target support only to those who would otherwise not make a will. Ord Minnett have estimated that providing free wills costs the PTO approximately $1.5-2 million per annum. There is no estimate available of the court costs saved through the prevention of intestate deaths. If the government does wish to encourage will-making, an explicit subsidy could be given to all providers on a competitively neutral basis.

4.3.10 The PTO is required by various statutes to fulfil a number of trustee and other roles. For some of these, government involvement may be justified. Examples include:

• acting as a trustee for children under a number of Acts. The PTO acts in a quasi-parental role, holding sums of money for children and deciding whether to make payments for education, maintenance or advancement;
• reporting to the court on the carrying out of functions by other trustees and managers under the Trustee Act 1956 and Protection of Personal and Property Rights Act 1988; and
• holding unclaimed assets, and allowing transactions to take place where absentee owners cannot be located, under a number of different Acts.

4.3.11 Although we have not reviewed in any detail the justification for government involvement in these trustee functions, it appears that at least some of them serve a useful purpose. Our discussions with a number of parties suggest that the services that serve a useful purpose could be purchased from the PTO or other organisations on a commercial basis or be provided by other government departments. The PTO also undertakes some quasi-judicial statutory functions which arguably could not be properly contracted out, and may need to be provided by a public body. That body need not be the PTO. Thus, none of the provisions requires a PTO established with a government guarantee.

4.3.12 In summary, the government might wish to ensure that some organisation provides some of the non-commercial services currently provided by the PTO. In our view, rather than provide the services itself, the government should contract on a commercial basis with the supplier that can provide the required services at the lowest cost. If there are quasi-judicial functions that cannot be contracted out, these could be assigned to a government agency. The PTO could act as the government's purchasing agent, or it could compete as a service provider, operating on a competitively neutral basis. If the PTO were to assume the purchaser role, it would have to divest its current activities, and manage its business down. Adoption of the provider role would require removal of the government guarantee and removal of the special legislative provisions that relate to the PTO. The government would then establish a body to undertake the purchaser role, or assign the role to an existing government agency. The public trustee supports the proposition of the PTO being a provider, rather than the purchaser.

4.4 Non-neutralities in the Treatment of the PTO

4.4.1 If the government chooses to establish the PTO as a competitively-neutral service provider, it will have to remove non-neutralities which arise because:

• the PTO's common fund is guaranteed by the Crown. In addition, there is an implicit government guarantee of the PTO's other activities as a result of its special legislative provisions and possibly also arising from the use of the word 'public' in its name; and
• the PTO is subject to the privileges, duties and restrictions imposed in 90 statutes.

Government Guarantee of Common Fund and Implicit Guarantee

4.4.2 As noted above, the government guarantee of the common fund was originally intended to provide security for the PTO's developing business. Given the emergence of a competitive trustee industry, and the advantages of the government obtaining the social services it requires from the PTO on a commercial basis, a state-backed government-guaranteed trustee business is no longer justified.

4.4.3 The government's exposure under the government guarantee has been extended as a result of the PTO expanding beyond its original role of being an impartial executor, trustee and manager of agency business. The PTO now competes with private sector providers in a number of financial markets. The PTO introduced a deposit-taking product, SafeCare, in 1987 for at-call money invested in the government-guaranteed common fund. The government guarantee provided the PTO with a competitive advantage over other investment products. The PTO believes that the success of SafeCare was "more related to good timing together with sound product features and competitive returns rather than the incidental fact that it was subject to the guarantee". Whatever the explanation for the product's success, the establishment of SafeCare resulted in a substantial increase in the common fund and provided an important source of the PTO's funds. SafeCare was closed to new investors in 1995 when the PTO expanded its business into providing off-balance sheet investment fund products structured as Group Investment Funds. Although these are not explicitly guaranteed by the government, arguably the PTO benefits from an implicit guarantee.

4.4.4 The investment positions the PTO has taken in the past, supported by the government guarantee, have allowed the PTO to generate a surplus from the common fund. According to a report prepared for The Treasury, the surplus has been used to cross-subsidise the uneconomic activities undertaken by the PTO. The surplus has arguably allowed the PTO to maintain a nationwide infrastructure and modus operandi without due consideration for the overall efficiency of the organisation.

4.4.5 Although the government guarantee of the PTO does not extend beyond the common fund, it is questionable whether the government would allow the PTO to fail. The use of 'Public' in the PTO's name, the special legislation establishing the organisation, the special duties and functions assigned to it in various statutes and the widespread perception that the organisation is owned by the government, reinforce perceptions that the PTO enjoys a government guarantee of its operations. This implicit government guarantee increases the PTO's competitive advantage and extends the government's risk beyond the explicit guarantee.

4.4.6 To date, the government guarantee has not been called on and the risks imposed on the government by the guarantee appear to be reasonably low as long as the PTO manages its risks effectively, particularly those associated with corporate trusts business. The PTO was exposed to considerable interest rate risk from its common-fund fixed interest investments, particularly during the first half of the 1990s when liabilities to clients reached over $1 billion. The sharp drop in interest rates from 1990 to 1993 saw it do well out of this exposure and a positive interest rate margin. Property revaluations in 1988/89 and realised profits from the sale of fixed interest investments in 1993-94 have contributed significantly to a now substantial reserve which provides a buffer against adverse performance. The year-to-year volatility in addition to reserves illustrates the risks which arose from the PTO's activities during this period. The PTO has been reviewing its operations to increase accountability and has strengthened the monitoring and accountability of its investment decisions (via its investment board).

4.4.7 The PTO has chosen not to enter into the higher risk segments of the corporate trusts markets although there is nothing to prevent it legally doing so. Any expansion of its business into these segments may result in a substantial increase in the risks borne by the government through its implicit guarantee of the business. The PTO has restricted its corporate trustee services to acting as trustee for unit trusts and to similar engagements generally regarded as relatively low risk.

4.4.8 The PTO is exposed to risks associated with its obligations as a trustee and its performance in the past. If its management of trusts in the past has been weak, litigation over performance is possible. However, the PTO believes that, given the nature of its trust and estate business which is generally positioned at the low to middle value end of the market, individual risks of this kind are not significant given its financial position.

4.4.9 Because there is no longer a public policy justification for a government guarantee, the government should at least withdraw the explicit guarantee for new business taken on by the PTO. The public trustee agrees that there is no longer any need for a guarantee of the common fund. The government's exposure to the guarantee could then be managed down over a period of time to ensure acquired legal rights are not impinged upon. More rapid removal of the guarantee would require the government or a new owner to provide some protection to existing clients consistent with the security that existed when they entrusted their business to the PTO. Removing the implicit guarantee of the business will be difficult, unless the government adopts a governance arrangement which clearly separates the PTO from the government.

Legislation

4.4.10 The PTO is governed by its own legislation, the Public Trust Office Act 1957. The other trustee companies are governed by the Trustee Companies Act 1967. The Acts confer similar powers on the trustee organisations with the major differences being the provision in the Public Trust Office Act 1957 that allows the PTO to operate a common fund which is guaranteed by the government, and restrictions which prevent the PTO offering a full range of products to its customers.

4.4.11 The PTO is given a role in over 90 separate pieces of legislation. Most of the provisions simply allow the PTO to do something, but impose no positive duty to do so, or they regulate some aspect of its activity. As discussed in the previous section, some provisions involve regulation of trustee services or quasi-judicial functions where government involvement may be justified. The full list of statutes and brief descriptions of their purpose are listed in the Appendix.

4.4.12 The special provisions relating to the PTO bring disadvantages as well as advantages. The Public Trust Office Act 1957 is perceived by the PTO as being outdated. The PTO is constrained from full competition with the legal profession and cannot expand its product range into some areas (for example, the PTO cannot convert its unitised funds into registered superannuation funds).

4.4.13 Officials have agreed to undertake a review of the statutory functions of the PTO. The review of statutory and commercial functions is expected to be completed by mid-1997.

4.5 Governance Options

4.5.1 The ownership of the PTO is in doubt - with conflicting legal opinions suggesting that either the Crown is the owner or the assets are held in a constructive trust for current and previous owners.

4.5.2 The ownership issue has been complicated by the growth in the balance sheet of the PTO. In the 1980s the net equity in the business was in the order of $5-10 million. This has increased substantially (to over $80 million), with a substantial proportion being earned from the PTO's investment business (the majority of the investment funds were derived from other PTO business such as its deceased estate, agency and trust businesses).

4.5.3 There is a case for the government asserting an ownership interest in the PTO, given its guarantee of the PTO's operations, and its monitoring and oversight relationship with the organisation. Given that the government bore the risks of the PTO's financial services through its guarantee, the government might have a claim for at least a share in the profits generated from this business. On the other hand, it may be difficult for the government to claim ownership by virtue of its guarantee of the common fund since, unlike the situation in respect of State Insurance, the PTO does not recontract with its customers every year.

4.5.4 Past and/or present clients may have an ownership claim on the basis that the surplus income generated by the PTO was generated from the clients (in a sense, the PTO was a type of mutual organisation). A constructive trust may arise because the PTO generated profits using its customers' funds. However, because of the government guarantee, the clients did not bear the residual risk associated with the PTO's operations, i.e. the fact that they would not have borne the costs if losses had been sustained by the PTO weakens the case for them claiming full rights to the profits that were actually generated. If the PTO's assets are deemed to be owned by past and present clients, its internal accounting structures would be unable to ascertain which clients own its accumulated reserves. Even if identification could be made, there would be practical difficulties in locating beneficiaries (e.g. some are dead so the PTO would need to find their beneficiaries). While this might present difficulties in implementing the option of client ownership, a court would be unlikely to reject this option on the grounds of impracticality (i.e. it would not allow the trustee to profit from its deficiencies).

4.5.5 Of course, the options of government ownership or ownership by clients are polar cases, with shared ownership being a possibility. One possible option would be to identify (in broad terms) the part of the reserves that was generated from the government-guaranteed investment products, and assign all or part of that to the government (in recognition of the risks it bore), with clients being assigned rights to the return earned from the PTO's core trustee business.

4.5.6 Given uncertainty over the ownership of the PTO, any action by the Crown to assert ownership would probably be challenged through the courts by the public trustee, or possibly by clients of the PTO. Such a course of action would not guarantee a satisfactory ownership structure, and could result in substantial delays to any reform.

4.5.7 In any case, structural reform of the PTO requires a review of legislation governing the PTO, the trustee companies, and the 90 statutes which contain provisions relating to the PTO. As a result of these constraints, and the relatively low political priority for the reform of the PTO and the trustee companies more generally, officials have decided to work with the PTO to strengthen the reporting and accountability framework of the PTO within the legislative constraints of the status quo. Work is continuing on long-term solutions to the ownership, structure and governance of the PTO.

4.5.8 In this section we briefly review some options for the future governance of the PTO. These all assume that the government moves to explicit funding of social services on a contestable basis.

Modified Status Quo

4.5.9 The funds the PTO manages on a trustee basis are clearly property held in trust in a legal sense. The PTO has extended the range of financial services that it provides to its customers to compete in a number of financial markets with deposit-taking institutions like banks. Neither of these roles sits comfortably with the current structure of the PTO in which the organisation is a government department accountable directly to a minister.

4.5.10 After an extensive review of the statutory and constitutional position of the PTO, officials have concluded that it is not possible, for legal reasons, to directly apply the Crown entity framework to either the public trustee or the PTO.

4.5.11 Instead officials, with the agreement of the minister and the PTO, have developed an administrative arrangement to replicate the reporting and accountability framework applicable to Crown entities under the Public Finance Act 1989. The approach envisages the PTO preparing a corporate plan (proposed as equivalent to the statements of intent required of Crown entities) for the next three years. The plan, and accompanying business plans, would then be subject to consultation leading to agreement between officials, the minister in charge of the PTO and the PTO. A regular cycle of planning and performance evaluation, similar to that applying to Crown entities, would then ensue.

4.5.12 Steps have been taken during the past three years to strengthen the Public Trust Investment Board, which has the function of controlling or advising on investment activities. Investment activities represent the major source of both risk and return for the PTO. A majority of the Board are members with investment and risk management expertise gained in the private sector.

4.5.13 These steps appear likely to introduce greater transparency to the PTO's operations and improved accountability to the government. Removal of the government guarantee would require an amendment of the Public Trust Office Act 1957. Under the above modified status quo, both the explicit and implicit government guarantee would remain.

Break-up or Ring Fencing of Activities

4.5.14 The government's exposure to the PTO's activities could possibly be reduced if the government restricted the PTO's services to the core trustee functions it was originally established to undertake. This would require the PTO to exit from non-core activities such as providing managed investment funds. The management of existing funds could be contracted out, or the funds closed and managed down over time. The public trustee, however, argues that investment is a core part of the PTO's business. Core trustee functions invariably generate funds which require investment; for example, long-term trusts and estates require unitised funds and even short-term estates require pooled short-term funds. Funds held purely as an investment constitute a relatively minor part of the total funds under the PTO's management.

4.5.15 Removing the government guarantee for the PTO's new business, and ring fencing of the guaranteed common fund from non-guaranteed activities, could also assist in reducing the government's exposure to the PTO's activities.

4.5.16 Another option, mentioned earlier, would be to manage the PTO down (including selling aspects of its business) to a role as purchaser of the services the government wishes to fund.

Trustee Company Under Trustee Companies Act 1967

4.5.17 In the long run, an option for reform of the PTO is to place it on the same basis as other trustee companies, i.e. re-establish it as a corporate operating under the same legislation. Trustee companies are governed by the Trustee Companies Act 1967 and the Trustee Act 1956. The Trustee Companies Act 1967 allows trustee companies to administer estates and trusts on a commercial basis. Trustee companies can be appointed as trustees and can hold other fiduciary positions. The Trustee Companies Act 1967 proscribes limits on the fees which trustee companies can charge for their services. Trustee companies can establish group funds and invest trust funds in the group investment fund.

4.5.18 There is some concern that the legislation governing trustee companies is outdated. Ideally, it should be reviewed as part of the process of reforming the PTO.

4.5.19 Under the trustee companies model, the CEO would report to a board of directors appointed by shareholders. The trustee companies model may need to be modified if the government continues to own the shares, or the shares are held by a trust. In both these cases, special accountability provisions, similar to those that apply to SOEs or Crown entities, should be incorporated into the governance structure.

4.5.20 Constituting the PTO as a trustee company under the Trustee Companies Act 1967 would place it on the same footing as its competitors in the trustee business. The PTO supports a move to this model, but the PTO and officials have conflicting legal opinions as to who should own the shares in the trustee company.

4.5.21 Options for ownership of the shares include the government; the public trustee holding the shares in trust for its clients; a community trust (as a proxy for past and present PTO clients); distribution directly to the PTO's current (and possibly past) clients; sale to the private sector (requiring the government to assert ownership initially); or a combination of the above (e.g. part held by government and on-sold, part distributed to present clients, and part placed in a community trust (as a proxy for past clients)).

4.5.22 Ownership by the government is not desirable in the long run, given the incentive problems associated with government ownership.

4.5.23 Distribution of shares to a trust holding the assets on behalf of the PTO's past and/or present clients would suffer from similar difficulties to government ownership. The ultimate owners ('the community' or 'past and present clients of the PTO') can exercise only a weak, indirect influence on the trust and therefore have few incentives to monitor the organisation or the trust's performance. Employees of the trust do not have a direct investment in the assets and therefore have relatively weak incentives to perform. If the trust is not prepared to sell the assets, the constraints normally imposed by a takeover market would be absent. In the event of difficulties, taxpayers may still bear some liability if the Crown were the settlor of the trust.

4.5.24 A possible advantage of giving the shares to a trust rather than the government is that a trust (depending on its trust deed) could subsequently decide to sell down the PTO without suffering the political ramifications that would be borne by the government. A well-constituted trust, with the objective of maximising benefits to its constituents (whoever they might be) would be prudent to invest in a diversified portfolio of assets, rather than hold a single investment. However, if the trust were established to hold the assets of the PTO, and this objective were written into its trust deed, it would not be obliged to diversify its assets to conform with a prudent person approach.

4.5.25 Distribution of shares to past and present clients suffers from the major difficulty of identifying past clients (many of whom would be dead), or their successors. Distribution of shares to present clients would be a possible option. Distribution of shares to clients has the advantage of making the ownership of the PTO contestable. Parties with particular skills in managing a trustee company have scope to buy shares in the company and gain control of it.

4.5.26 Sale of the company through a contestable tender process would facilitate the rapid transfer of the company to an efficient owner of the organisation. However, political resistance, and the uncertainty over the government's ownership claim to the PTO, make this approach unlikely.

4.5.27 A variant of the latter option would be for the PTO to be sold by competitive tender with the proceeds of the sale being held in trust for past and present clients of the PTO. The objectives of the trust could be determined, and the trust would be free to invest in a diversified portfolio of assets rather than being restricted to a holding in the PTO only.

Financial Mutual

4.5.28 An alternative option would be to establish the PTO as a financial mutual owned by its current (and in effect past) clients. This (as long as it was accompanied by the removal of the government guarantee) would remove government's exposure to the organisation. However, mutuals suffer from incentive and monitoring problems comparable to government or trust ownership. Mutuals are also exposed to capital limitations which might constrain their opportunities in the future. There would also be practical difficulties in establishing a mutual given the wide range of client types - from depositors through to those using the PTO to write wills.

4.5.29 This concern is reinforced by the fact that none of the PTO's major competitors is organised as a mutual (although some of them are owned by mutuals), suggesting that a mutual form may not be optimal. It is also noteworthy that a number of financial mutuals have demutualised, or are in the process of demutualising i.e. they believe that it is worthwhile undertaking this costly process to achieve the benefits of a more normal company structure.

4.5.30 If the PTO is established as a financial mutual, it may be difficult for the government to signal the severing of its links with the organisation. For a number of years at least, an implicit government guarantee and consequent risk to the government might remain.

4.6 Summary

4.6.1 In our view, there are no public policy reasons for the Crown to own, sponsor or guarantee a trustee corporation. Competing trustee companies, the legal profession and private individuals provide competitive trustee services without the benefit of a government guarantee. While the PTO serves clients that are not currently serviced by other providers, the PTO would be likely to continue to service some of this market without government backing.

4.6.2 Our (brief) examination of the PTO's social and statutory roles suggests that none of the 'social' services it provides or its statutory functions requires a state-backed trustee corporation. Instead, services that fulfil a useful purpose could be purchased from the PTO or other organisations on a commercial basis or provided by other government departments.

4.6.3 Despite the case for reform of the PTO being strong, a number of impediments to restructuring exist. These include:

• conflicting legal opinions as to current ownership;
• the extent of legislative change required to remove the special treatment of the PTO;
• the desirability of reviewing the trustee companies legislation; and
• a lack of political will.

4.6.4 Given these constraints, the approach by officials of improving the accountability of the PTO within the current statutory arrangements, while continuing to examine other options for reform, appears sensible.

4.6.5 In the longer term, the PTO should be established as a trustee company under the legislation that governs trustee companies with the shares distributed to customers or sold by tender. The proceeds could be retained by the government or distributed (in whole or part) to a trust to provide benefits to past and current clients.

4.6.6 In any event, the government should remove the guarantee of the PTO's common fund, and could possibly require a change in the company's name to reflect the severing of its links with the government. Given the value that is intrinsic to a company name, the latter should be undertaken only if there is a significant risk that the term 'public' carries a connotation of government ownership.

Chapter 5 Recommendations

5.1 National Provident Fund

5.1.1 Consideration should be given to:

(i) reviewing the reasons why the NPF has greater scheme administration and head office costs than the GSF, particularly in relation to costs which may fall on the taxpayer;

(ii) a full or partial amalgamation between the GSF and the NPF;

(iii) the case for further reducing the number of fund managers;

(iv) reducing fiscal risks by tendering the Crown's obligations, particularly in respect of the DBP Annuitants Scheme.

5.2 Government Superannuation Fund

5.2.1 Consideration should be given to:

(i) contracting out investment management;

(ii) lengthening the duration of the GSF's investments;

(iii) allowing the GSF's assets to be invested in global equities;

(iv) reducing the unfunded liability by increasing the GSF's assets;

(v) providing a much more specific statement, through the minister of finance, of the required risk characteristics of the GSF's assets;

(vi) merging all or part of the NPF and the GSF;

(vii) tendering the Crown's obligations, particularly in respect of annuitants.

5.3 Public Trust Office

5.3.1 Consideration should be given to:

(i) removing the government guarantee from the common fund;

(ii) possibly changing the name of the PTO if the word 'public' is believed to connote government involvement;

(iii) establishing, in the longer run, the PTO as a trustee company;

(iv) divesting any Crown shares in the trustee company; and

(v) in the interim, improving the accountability of the PTO within current statutory arrangements.

Appendix List of Statutes Pertaining to the Public Trust Office

General Enabling Provisions

· Public Trust Office Act 1957

· Public Trust Office Regulations 1958

Administrative Provisions

· Securities Act 1978: exemption from requirements of the Act.

· Residential Tenancies Act 1986, schedule 4: the Act applied to the PTO's residential tenancies.

· Evidence Act 1908, s 5: power to produce in evidence prints from films of any business record or similar document.

· Official Information Act 1982: the Act applies to the PTO as a department, but does not extend to trustee/fiduciary responsibilities.

· Ombudsmen Act 1975: exclusion of ombudsman's jurisdiction on decisions made in trustee capacity (though it is the PTO's policy to cooperate).

· Crown Proceedings Act 1950, s 35: nothing in the Act is to affect liability imposed on the PTO by the Public Trust Office Act 1957.

· State Sector Act 1988: the PTO is a department.

· Public Finance Act 1989, s 2: the PTO is not a department under the Public Finance Act 1989.

Tax

· Goods and Services Tax Act 1985: trustees and executors are required to pay GST.

· Estate and Gift Duties Act 1968, s 35B: exemption for chattels under $6,000.

· Stamp and Cheque Duties Act 1971, s 80: enables bulk payment arrangements.

· Income Tax Act 1994: requirement for the PTO to pay tax.

Executors/Administrators

· Administration Act 1969.

· Secondhand Dealers Act 1963, s 11: special powers in administering estates of secondhand dealers.

· New Zealand Society of Accountants Amendment Act 1958, ss 12 and 12: special powers in administering estates of accountants.

· Government Superannuation Fund Act 1956, s 49: the PTO to receive money for deceased contributor; s 61: the PTO receives benefit where contributor dies without leaving a spouse.

Personal Trustee

· Administration Act 1969.

· Trustee Act 1956, s 46.

· Te Ture Whenua Maori Act 1993.

· Charitable Trusts Act 1957, ss 41 and 51: Public Trustee may be appointed trustee of approved scheme.

Corporate Trustee

· Unit Trusts Act 1960, s 5: the PTO may be appointed as trustee of a unit trust.

· Securities Act 1978, s 48: the PTO may be appointed as trustee or statutory supervisor of debt securities and of participatory securities (e.g. forestry partnerships, retirement villages).

Trustees for Children etc

· Social Security Act 1964, s 61D: trustee of Social Welfare death benefits (repealed, some trusts ongoing).

· Accident Compensation Act 1972, s 84: trustee for compensation payments (repealed, some trusts ongoing).

· Deaths by Accidents Compensation Act 1952, s 11: trustee of damages or compensation payments (repealed, some trusts ongoing).

· Workers Compensation Act 1956, s 55: trustee of damages or compensation payments (repealed, some trusts ongoing).

· Human Rights Act 1993, s 88: duty to receive payment from Proceedings Commissioner on behalf of minor or mentally disturbed person.

· Privacy Act 1993, s 88: duty to receive payment from Proceedings Commissioner on behalf of minor or mentally disordered person.

· Health and Disability Commissioner Act 1994, s 57: duty to receive compensation from Commissioner for a mentally disordered person or a minor.

· Government Superannuation Fund Act 1956, s 47: duty to be trustee for money payable to child.

· Children, Young Persons and Their Families Act 1989, s 469: administration of property of child (deemed subject to a Protection of Personal and Property Rights Act 1988 property order).

· Child Support Act 1991, ss 187-188: appointment as receiver of property charged with the payment of maintenance.

· Minors' Contracts Act 1969, s 9: report to Court on contract to be entered into by minor.

· Family Proceedings Act 1980, s 156: duty to act for disabled persons.

Managers

· Protection of Personal and Property Rights Act 1988.

· High Court Rules 1986/District Courts Rules 1992: institution and defence of proceedings where acting as manager under the Protection of Personal and Property Rights Act 1988.

· Secondhand Dealers Act 1963 and New Zealand Society of Accountants Amendment Act 1958 special powers enabling continuation of business of an accountant or secondhand car dealer whose estate is under management.

Attorney/Agent

· Protection of Personal and Property Rights Act 1988, Part IX: provision for EPAs, remuneration set under Protection of Personal and Property Rights Regulations 1988 and Public Trust Office Regulations 1958.

· Pacific Islands Polynesian Education Foundation Act 1972, s 17: Foundation may engage the PTO as agent.

Financial Services

· Friendly Societies and Credit Unions Act 1982, 58: Public Trustee or trustee companies may hold securities and invest funds for societies.

· Valuation of Land Act 1951, s 31: use of district roll valuations for mortgage advances.

· Family Proceedings Act 1980, s 93: investment of maintenance payments.

· Securities Act 1978: exemption for the PTO.

Unclaimed Assets

· Insolvency Act 1967, s 134: surplus from bankrupt's estate paid to the PTO.

· Companies Act 1955, s330A: receipt from Official Assignee of unclaimed assets of a company.

· Land Act 1948, s 128: lease of unclaimed land which is property of Crown.

Acting for Persons under a Legal Incapacity

· Land Transfer Act 1952, s 20: applications to bring land under the Land Transfer Act 1952 in respect of mentally disordered persons.

· Land Act 1948, s 50A: Public Trustee can be member of community water supply association as representative of mentally disordered person.

· High Court Rules 1986, r 81 and District Courts Rules 1992, r 83: duty to represent person if directed.

· Law Reform Act 1936, s 9A: duty to act as administrator in proceedings claiming damages or compensation against estate of deceased owner.

Absentee Owners

· Property Law Act 1952, s 87: repayment when mortgagee cannot be found.

· Chattels Transfer Act 1924, s 44: discharge of instrument by way of security where grantee absent.

· Rating Powers Act 1988, s 145: receipt of balance of proceeds of sale of a property (for rates arrears) where owner absent.

· Valuation of Land Act 1951, s 35: duty to receive compensation for compulsory acquisition where owner is absent, incapacitated, etc.

· Land Settlement Promotion and Land Acquisition Act 1952, ss 13-15: the PTO to represent absentee claimants and to receive compensation money.

· Crown Minerals Act 1991, s 78, 79: the PTO stands in for absent persons in relation to access arrangements.

· Public Works Act 1981: representation of infants, absentee owners, etc; compensation paid to the PTO where doubt about persons entitled.

· Life Insurance Act 1908, s 69: duty to receive money payable under a policy to a minor or incapacitated person and to apply money as specified.

· Valuation of Land Act 1951, s 35: duty to receive compensation for compulsory acquisition where owner is absent, incapacitated, etc.

Miscellaneous/Commercial

· Education Act 1989, School Trustees Act 1989: devolution of property of Boards.

· Energy Companies Act 1992, s 49: duty to hold property of Board for its successor.

· Health Reforms (Transitional Provisions) Act 1993, s 11: duty to prepare scheme for transfer of asset or modifying terms of trust.

· Enemy Property Act 1951, ss 2-8: Public Trustee is custodian of enemy property.

· Maori Housing Act 1935, s 6: power to accept an order on or an assignment of proceeds of alienation of any land.

· Insurance Companies' Deposits Act 1953: maintenance and investment of insurance companies' deposits.

· Life Insurance Act 1908, s 69: maintenance and investment of life insurance companies' deposits.

· Mutual Insurance Act 1955: maintenance and investment of mutual insurance companies' deposits.

· Companies Act 1955, s 419: maintenance and investment of overseas insurance companies' deposits.

· Trustee Companies Act 1967, s 6: maintenance and investment of trustee companies' deposits.

Quasi-judicial Functions

· Wills Amendment Act 1852: wills for minors between age of 16 and 18.

Audit

· Trustee Act 1956, ss 83A and 83B.

· Protection of Personal and Property Rights Act 1988, s 46.