Pension scheme investment policies
A hard copy of this report summary can be obtained by contacting Paul Noakes [E-Mail: Paul.Noakes@dwp.gsi.gov.uk] or by writing to him at the 'Social Research Division, Department for Work and Pensions, 4th Floor, Adelphi, 1-11 John Adam Street, London WC2N 6HT'.
Research Report No. 82
By C Pratten and S Stachell
This report examines the decisions trustees make about pension scheme investment. Depth interviews were carried out with trustees of 48 self-administered private sector occupational pension schemes, between November 1996 and February 1997. The study, by the Department of Applied Economics at the University of Cambridge, was commissioned by the Department of Social Security (DSS). The research provides detailed baseline information about what was happening prior to the implementation of the Pensions Act 1995. The Act introduced a number of legal requirements on trustees concerning the management and funding of occupational pension schemes. The changes came into effect from April 1997. The main findings are:
- The objectives of the trustees' investment policies were to achieve a good return on the assets of the scheme so as to provide attractive pensions and to limit the costs of the schemes to the sponsoring employers; and to avoid risking a significant reduction in the value of the schemes assets and incomes.
- Trustees sought to achieve these aims by: appointing expert advisers and fund managers with successful track records and monitoring their performance; adopting what they perceive as cautious investment policies; and providing guidelines and benchmarks for fund managers for investing the schemes' assets.
- The five main influences on asset allocation identified by trustees were:
- equities had provided the highest investment returns in the past and were expected to so in future;
- the asset allocation of other schemes - most schemes kept close to the CAPS (Combined ActuVerdana Performance Services) or WM (World Markets) averages ;
- the financial state of the scheme - schemes with surpluses had greater freedom to allocate funds to equities, which were expected to have the highest returns but were seen to be the asset class exposed to most risk;
- the maturity of the scheme - less mature schemes (with a higher ratio of active members to pensioners and deferred pensioners) could take a longer perspective and hold a higher proportion of assets in equities. However, many schemes had surpluses which meant that maturity was not a dominating consideration;
- the decisions of fund managers - who determine asset allocation, usually within ranges set by trustees.
- Trustees had made changes in the past five years to improve their schemes' performance rather than as a result of the Pensions Act. Most trustees did not expect to make any major changes to their investment policies in the next three years mainly because they were satisfied with their schemes' current arrangements and performance.
Introduction
This report examines the decisions trustees make about pension scheme investment. Depth interviews were carried out with trustees of 48 self-administered private sector occupational pension schemes, between November 1996 and February 1997. The study, by the Department of Applied Economics at the University of Cambridge, was commissioned by the Department of Social Security (DSS). The research provides detailed baseline information about what was happening prior to the implementation of the Pensions Act 1995. The Act introduced a number of legal requirements on trustees concerning the management and funding of occupational pension schemes. The changes came into effect from April 1997.
The research examined: how trustees responsible for pension scheme investment make their decisions, the factors they take into account and why; their relationship with fund managers, actuaries and the sponsoring employer; whether their investment decisions had changed or would change in the future, and if so, the reasons for any changes.
The investment background
Most of the schemes included in the study were soundly financed with a surplus of assets over estimated liabilities. Generally, the surplus of assets over liabilities indicated by the Minimum Funding Requirement (MFR) calculations were greater than those based on actuVerdana calculations.
The investment background to the study was that the actual returns on investments obtained by most pension schemes have exceeded prior actuVerdana expectations over the period since 1980 because of the buoyant performance of equities in nominal and, more important, in real terms.
The objectives of the investment policies of trustees were to achieve a good return on the assets of the schemes so as to provide attractive pensions and to limit the costs of the scheme to the sponsoring employers; and to avoid risking a significant reduction in the value of schemes' assets and income.
Trustees reported that they seek to resolve the conflicting objectives of achieving a high rate of return and putting assets at risk by: appointing expert advisers; appointing fund managers with a reputation for, and record of, successful investment and then monitoring their performance; adopting what they perceive as cautious investment policies; limiting and constraining the investment policies of the fund managers and providing them with guidelines and benchmarks to ensure the assets of the schemes are invested in a range of different classes of assets and that investments in equities are distributed over a wide range of companies.
How trustees determine and manage investment policies
Generally, the largest schemes with assets of £100m or more had investment sub-committees of the boards of trustees. The extent to which decision-taking was delegated to these sub-committees varied but, where they existed, most investment sub-committees had a good deal of delegated authority for deciding investment policy. The trustees of smaller schemes with less than £20m of assets tended to function without formal investment sub-committees.
The trustees reported that they had been able to reach consensus on matters relating to investment strategy by discussion and that there had been few sources of major conflicts or disagreements between trustees. There were two main elements of investment decision making by trustees: the appointment of advisers and the fund managers; and deciding the allocation of assets between asset classes.
Generally, trustees were not involved in stock selection, which was left to fund managers.
All of the sample schemes had access to actuVerdana advice, either from firms of consulting actuaries or from firms of administrators. Generally, the advisers to the trustees were appointed by the trustees. At some larger schemes with assets of £20m or more, the actuaries who provided investment advice attended all the meetings of trustees at which investment matters were considered. The advice of actuaries was important for informing trustees whether their policies conformed to laws, rules and the practice of other pension schemes. The valuations and papers from actuaries, administrators and other advisers were important sources of information for trustees considering investment policy.
The majority of the schemes had appointed external fund mangers. The minority of schemes used internal fund mangers. Fund managers were usually appointed after a 'beauty parade' at which presentations were made to trustees or a sub-committee of trustees. Final decisions on the appointment of fund managers were usually taken by the boards of trustees of schemes. There was a considerable turnover of fund managers, half of the defined benefit schemes in the study had appointed or changed fund managers within the past five years.
Trustees of most of the largest schemes with assets of £100m or more set strategic benchmarks for asset allocation and bands around these benchmarks within which fund managers could vary asset allocation. The reports received from fund managers were the principal focus of discussion at most trustee board meetings concerned with investments. Fund managers attended meetings of trustees at least once a year.
Most of the larger schemes with assets of £20m or more set targets for their fund managers, usually these targets were related to the Combined ActuVerdana Performance Services (CAPS) or World Markets (WM) averages. Typically, the targets were to beat the CAPS or WM median by one per cent a year over rolling three year periods and not to fall below the median by more than two per cent in any 12 month period. Apart from setting a performance target for fund managers, the targets signalled to the fund managers the degree of risk they should take. The implication was that the higher the target, the greater was the expected divergence from the average asset allocation by pension funds. A majority of the smaller schemes with assets of less than £20m did not set targets for their fund managers. The main explanation for the smaller schemes not setting targets was that they used pooled funds.
Trustees reported that the principal channel through which the sponsoring employers were involved in, or influenced, the policies of pension schemes, was the inclusion of directors of the sponsoring employers on the boards of trustees. Trustees reported that conflicts of interest between trustees and sponsoring employers concerning investment policies had not occurred.
Asset allocation
The main features of the actual allocation of assets between asset classes by the sample of defined benefit schemes were: uniformly high allocation to equities; the holding of a diversity of assets; large variations between schemes in the holdings of each of the classes of assets, apart from equities, as percentages of total assets.
Pension schemes, on average, invested about 70 per cent of their equity portfolios in UK equities and 30 per cent in the shares of companies based overseas. The principal explanation for this allocation, given by trustees, was the possible effects of a currency mismatch with investment overseas and liabilities for pensions set in sterling if sterling appreciated. There was a sharp divergence of policy on investment in property. The majority of the largest schemes with assets of £100 m or more invested in property. A significant minority of the largest schemes and most of the smaller schemes with assets of less than £20 m did not invest in property. There was a general movement to reduce the proportion of assets allocated to property.
The five main influences on asset allocation which were identified by trustees were that:
- equities have provided the highest investment returns in the past and were expected to do so in the future. This was the main influence explaining the dominant position of equities in the portfolios of pension funds;
- the allocation of assets by other schemes as evidenced by the averages published by CAPS and WM. Most schemes had not strayed very far from the average allocation;
- the financial state of schemes. The existence of widespread and substantial surpluses permitted greater freedom to allocate funds to the class of assets expected to provide the highest returns and traditionally perceived to be the asset class exposed to most risk, equities;
- the maturity of schemes. Less mature schemes could take a longer term perspective and hold a higher proportion of their assets in equities, other things being equal. However, maturity was not a dominating consideration because for many schemes their surpluses removed most constraints, apart from those dictated by prudence, on investment allocation;
- the decisions of fund managers who determine asset allocation usually within ranges set by trustees.
Pooled investment funds were used by many, but not all, of the smaller schemes with assets of less than £20m. The larger schemes with assets of £20m or more made some use of specialised pooled funds for certain types of investment including venture capital and investment in emerging markets. A significant minority of the sample of defined benefit schemes used tracker funds.
Whether a scheme was contracted out of SERPS (the State Earnings Related Pension Scheme) or not, was not a factor which trustees consciously considered when deciding the investment policies of schemes.
Control of risks
Most trustees did not obtain risk statistics. Nevertheless, trustees showed that they were aware of risks and took steps to limit their exposure to risks. They limited investment in any company to three or five per cent of the equity portfolio and they limited the use of derivatives; many schemes, particularly the smaller schemes with assets of less than £20m made no use of derivatives. Investments in emerging markets, venture capital and the business of the sponsoring employer were very limited. When asked about the effects of a crash or collapse of equity prices, the response of most trustees was that '...if markets fall they will recover .....' as they always have in the past.
Changes to investment strategies
The two most common changes to investment strategies made by trustees during the past five years were to change, or appoint additional, fund managers and to reduce investment in property. Most trustees did not expect to make any major changes to their investment strategy over the next three years. The main reason given by trustees for not expecting to make any major changes was that they were satisfied with the current arrangement and performance of their schemes.
Informing members
The principal channel of communication for informing members about the performance of larger pension schemes with assets of £20m or more was a newsletter usually distributed once a year. Newsletters were distributed to active members and to pensioners but many schemes did not distribute them to deferred members, often because they did not have the up-to-date addresses for many of these members. Smaller schemes with assets of less than £20m tended to circulate copies of the annual report and accounts to members. Employee trustees of smaller schemes were another channel for informing members about matters concerning the pension schemes.
Relevant publications
Other publications on pension schemes:
- B Casey, J Hales, N Millward (1996) “Employers' Pension Provision 1994” (DSS Research Report No. 58) London: TSO
- D Dundon-Smith, J Hales, M Chetwynd, A Thomas, J Keegan (1997) “Pension Scheme Inquiries and Disputes ”(DSS Research Report No. 66) London: TSO
- K Bunt, D Howells and M Winterbotham (1998) “Experiences of Occupational Pension Scheme Wind-Up' ”(DSS Research Report No. 75) London: TSO
- K Bunt, M Winterbotham and R Williams (1998) “The Role of Pension Scheme Trustees ”(DSS Research Report No. 81) Leeds: CDS