davidblake default funds in uk defined contribution pension plans

Default Funds in UK Defined Contribution Pension Plans Alistair Byrne CFA, David Blake, Andrew Cairns and Kevin Dowd∗ Abstract: Most defined contribution (DC) pension plans give their members a degree of choice over the investment strategy for their contributions. Many plans also offer a ‘default’ fund for members unable or unwilling to choose their own investment strategy. We analyse the range of default funds offered by UK ‘stakeholder’ DC plans, which by law must offer a default fund. We find the default funds are typically risky, but also that they vary substantially across providers in their strategic asset allocation and in their use of lifecycle profiles that reduce investment risk as the planned retirement date approaches. We use a stochastic simulation model to demonstrate that the differences can have a significant effect on the distribution of potential pension outcomes experienced by plan members who adopt the default fund as the path of least resistance. We also analyse the fees charged for the default funds. This is a pre-press version of a paper that appears in the Financial Analysts Journal, July/August 2007. http://www.cfapubs.org/doi/abs/10.2469/faj.v63.n4.4748 ∗ Alistair Byrne CFA is a lecturer in finance at the University of Strathclyde, Glasgow UK; David Blake is professor of pension economics at Cass Business School, London UK; Andrew Cairns is professor of financial mathematics at Heriot-Watt University, Edinburgh, UK; and Kevin Dowd is professor of financial risk management at Nottingham University Business School, Nottingham, UK. 1 Default Funds in UK Defined Contribution Pension Plans Defined contribution (DC) pension plans are an increasingly common form of retirement income provision in the US, the UK and many other economies. Most DC plans allow members a degree of choice about how to invest their contributions. Typically, a range of mutual funds is offered in the plan and the member can choose one or more of them in which to invest. Many plans also have a default option that is automatically used if the member does not actively choose a fund. Previous research shows that a large proportion, and often the majority, of employees are inclined to take the ‘path of least resistance’ and passively adopt the default arrangements that exist in their plan. For example, Choi et al. (2002) review US evidence on the tendency for members to accept plan defaults for key features such as the contribution rate and the investment fund. Even though employees are free to opt out of default arrangements, very few actually do. In the plans Choi et al. studied, between 42% and 71% of participants accept the default contribution rate and between 48% and 81% of plan assets are invested in the default fund, which is typically a money market fund. In the UK, consulting firm Hewitt Bacon and Woodrow estimate that more than 80% of members in DC plans accept the default fund choice (Bridgeland, 2002). Default funds do bring a number of benefits, especially if they are well chosen with the needs of the pension plan members in mind. Where plan members have relatively little financial knowledge, default funds simplify the pension saving process, which in turn might raise participation rates. The default fund provides an ‘obvious’ choice for the uninformed member, seemingly endorsed by the sponsoring employer or pension plan provider, and helps them deal with an otherwise complex decision (Madrian and Shea, 2001). However, the tendency of DC pension plan members to accept plan defaults does mean that the provider or plan sponsor’s choice of these defaults has the potential to have a significant impact on the welfare of plan members. Put simply, well-chosen default funds will benefit members, while poorly-chosen defaults will impose a cost on uninformed members. Furthermore, to the extent that there is cross-sectional variation in default funds across pension plans that is not explained by differing membership characteristics, members will face something of a lottery. Financial analysts and planners have an important role to play in 2 helping pension providers and plan sponsors to put appropriate default arrangements in place. In this paper we investigate this issue by analysing the variety of different types of default fund offered by stakeholder DC plans in the UK. We document the range of different approaches in use and provide a quantification of what these differences mean in terms of the potential pension outcomes for plan members. Stakeholder pension plans were introduced in the UK in April 2001 with the aim of providing a simple, carefully regulated and low cost savings product that could improve pension provision amongst low and middle-income employees. In essence, they are personal pension arrangements which operate on a DC basis and are offered by financial institutions. They share most of the features of other DC pension arrangements, for example, in terms of permissible contribution rates, the availability of benefits, and tax treatment. However, stakeholder plans also have a number of specific features intended to make them easy for inexperienced investors to use.1 The feature of interest to us is that the regulations require each plan to have a default fund so that members do not have to make an active choice about how to invest (Statutory Instrument 2000:1403). The requirement to have a default fund and the public availability of data for most plans on the default fund used makes the stakeholder pension market an interesting area in which to study the investment strategies financial institutions offer to ‘uninformed’ pension plan members. The stakeholder market is also a significant part of the UK pensions system. Stakeholder plans are offered by most of the major insurance companies and asset managers in the UK. While they can be sold as retail financial products, they are often used by companies for occupational pension provision. The employer ‘adopts’ a plan provider and its employees can then enrol in the plan. All employers with five or more employees, and who do not provide a qualifying occupational pension plan, must make a stakeholder plan available to their employees, but do not need to contribute to it (Blake 2003). 1 Stakeholder plans must have a low level of minimum contributions (£20), no penalties for ceasing or reducing contributions, no penalties for transferring to another arrangement, and total charges were initially capped at 1.0% per annum. From April 2005 providers are allowed to charge a fee of up to 1.5% for each of the first ten years the pension product is held by a customer. After ten years the fee cap reduces to 1.0% (www.hm-treasury.gov.uk). 3 As at May 2006, over 2.7m stakeholder pension accounts had been opened since the launch in 2001 (DWP, 2006) Figures from HM Revenue and Customs (2006) show that 1.5m individuals contributed to stakeholder pension plans during the 2004/05 tax year and that total contributions were £2.4bn. The corresponding figures for employer-sponsored group personal pensions, which are similar to stakeholder plans in many respects including the investment strategies they use, were 1.8m contributing members and £4.0bn of contributions.2 Assets under management in personal and stakeholder plans were estimated to amount to £300bn at the end of 2005, compared to an overall funded pension market, including defined benefit plans, of £1,400bn. (UBS, 2006) Our analysis of the default funds in stakeholder plans finds that they are typically risky, with high equity content, but also that there are substantial differences across funds in terms of their asset allocation and the nature of their lifecycle profiles that automatically switch the member’s pension fund assets to fixed-income investments and/or cash as the planned retirement date approaches.3 These differences mean that an individual employed in one company accepting the default arrangements can end up with a very different investment product from a similar individual who happens to work for a different employer, which has selected a different pension plan provider. We also find that fees vary substantially across the various fund offerings, although many plans do charge at the 1.0% original fee cap. We use a stochastic simulation model to illustrate the distributions of possible pension outcomes that the different fund structures generate for plan members accepting the default arrangements. The results of these simulations suggest that the choice of default fund can 2 We emphasise that although our analysis is based on stakeholder pension plans in the UK, it can be generalised to other DC pension arrangements where there are similar default options. 3 Lifecycle asset allocation profiles are used to attempt to reduce the risk that a fall in equity prices close to the planned retirement date reduces the member’s retirement income. Bodie et al. (1992) argue that if an individual’s human capital (i.e. future labour income) is less risky than equity, then at younger ages this capital will constitute a relatively high proportion of total wealth and thus can be balanced by investing a greater proportion of the individual’s financial wealth in risky assets. As time moves on, the share of wealth accounted for by human capital declines and it makes sense to reduce the risk attached to financial wealth. Furthermore, younger individuals have more scope to increase their work effort to make up for any shortfall generated by losses in financial assets. 4 have a major impact on likely pension outcomes. High equity strategies have the obvious benefit of higher expected pensions in payment. However, this comes at the cost of greater variability and members of plans with equity-based default funds may receive lower pension outcomes than those using more cautious strategies if their retirement happens to coincide with a period of equity market weakness. Our analysis of the actual historical performance of a subset of funds confirms the potential for members to receive substantially different outcomes depending on the nature of the default fund in their pension plan.4 These results are of potential concern, especially in light of the evidence that most members of DC pension plans passively accept the default fund chosen by the plan provider. Unless the different choices of default fund made by different providers are somehow correlated with the characteristics of the members of the different plans – and we know of no evidence to this effect – then plan members face an effective lottery: their choice of investment strategy is driven by the provider’s choice of default fund rather than their own circumstances and attitude to risk. Previous Literature A number of previous studies investigate the effects of alternative investment strategies on the anticipated outcomes of DC pension plans. For example, Booth and Yakoubov (2000) used historical return data from the annual Barclays Capital Equity-Gilt Study for the UK to investigate the retirement income implications of five different investment strategies. They assumed the ‘standard’ fund had a constant 70% equity / 20% bonds / 10% cash mix. This standard fund is combined with four lifecycle strategies – a switch to bonds over the ten years preceding retirement; a switch to cash in the final year before retirement; a switch to cash for the final three years; and a switch to bonds for the final three years. They found limited support for the superiority of lifecycle approaches, and also that an equity-based fund in the ten years preceding retirement ‘stochastically dominates’ the cash– and fixed-income–based strategies – principally because of the higher expected return. Blake et al. (2001) investigated similar issues using the ‘PensionMetrics’ stochastic simulation model. Amongst the asset allocation strategies they investigated were a pension- 4 The results of the analysis of historical performance are available in the online appendix. 5 fund-average approach – invested across a range of asset classes in proportions typical of UK occupational pension funds in the late 1990s – and a lifecycle strategy that switches from the pension fund average into a 50% bonds / 50% T-bills portfolio over the final ten years before retirement. They also found that the overall distribution of potential outcomes is very wide. In line with Booth and Yakoubov, they found that a well-diversified, high-equity strategy (i.e. the pension-fund-average strategy) produces the best overall outcomes and that, while the lifecycle strategy avoids some of the worst potential outcomes, it does so by significantly reducing the expected level of pension. A third study, Hibbert and Mowbray (2002), used a stochastic model to investigate the outcomes from a variety of asset allocation strategies (including 100% cash, 100% bonds, and 100% equity asset allocations, and various forms of lifecycle strategy). They too found that the 100% equity strategy produces the highest expected value for the pension annuity, albeit with a wide range of potential outcomes. The lifecycle strategies significantly narrow the range of potential outcomes, but at the expense of reduced expected value, particularly where the lifecycle switch begins 15 years from retirement. Our work differs from the papers discussed above principally in that it focuses directly on the fund structures actually offered as the default in UK stakeholder pension plans. The following section describes these fund structures in detail. Data on Default Funds UK legislation requires stakeholder pension plans to be registered with The Pensions Regulator, which makes the register available to the public. As at December 2006, 45 plans were listed on the register and these plans form the universe for our analysis. Of the 45 plans, 14 are closed to new business, e.g. because of mergers between providers, and so no longer provide public information on their fund structures, leaving 31 plans on which we were able to collect data. This sample, in effect, represents all of the stakeholder plans actively marketed in December 2006. The key variables of interest are the basic asset allocation of the default fund and the nature of the lifecycle profile used by the fund. 6 It is important to stress that the term ‘plan’ here refers to a pension arrangement offered to the marketplace by an insurance company, asset manager or, in some cases, a membership organisation such as a trade union. An employer can ‘adopt’ a plan and offer it to its employees. Thus each of our plans will likely be used by many different employers and groups of employees. Equally, many of the plans are offered on a retail basis and any individual can join, either arranging this themselves or via a financial adviser. In the occupational context, the employer chooses a financial institution to offer a stakeholder pension product to its employees. The choice will be made based on factors such as brand, track record and cost. Each financial institution will have a ‘standard’ default fund that it typically uses when implementing a plan for an employer. It is open for the employer to accept this standard plan or to ask that the financial institution uses another fund as the default for that employer’s employees. An employer might do the latter if it felt the standard default fund was inappropriate, e.g. too risky, for its employees, but our industry contacts suggest few employers actually do so. So, in most cases the financial institution’s choice of default fund prevails. Example: Widgets Inc chooses Byrne Investments to offer a stakeholder pension plan to Widget’s employees. Byrne Investments usually nominates its FTSE 100 Index Tracking Fund as the default for stakeholder schemes. The management of Widgets can either accept the usual default fund from Byrne Investments as the default for their employees, or they can ask Byrne to implement an alternative default for their employees. Widget’s management may be reluctant to override the judgement of the financial institution, even though they know their employees better than Byrne Investments does. With the financial institution’s standard default fund likely to be implemented in most cases, the fact that these stakeholder plans are, in most cases, generic plans offered to the whole marketplace rather than tailored for any specific group of employees suggests that the default funds across plans should be similar. Put simply, the default funds should be suitable for the average employee in the economy who randomly chooses, or is randomly allocated, one of available default funds. However, our data do not show this. In fact, we find substantial variation, both in terms of the basic strategic asset allocation and in terms of the use of lifecycle strategies. 7 Default funds are potentially less important in the retail setting. Where an individual joins a pension plan under the guidance of a financial adviser, it is likely that the adviser will guide the individual towards a fund choice that is consistent with his or her financial circumstances and degree of risk tolerance. In addition, individuals approaching a pension provider directly (i.e. doing business on an execution-only basis) are more likely to be financially knowledgeable and prepared to make their own fund choice. Nonetheless, there is the potential for relatively uninformed consumers to deal directly with the plan provider and be inclined to accept whatever default fund is proposed. Table 1a shows the range of default funds in terms of fund type and style of management. The ‘balanced managed’ type fund, which is typically invested 50% to 60% in UK equities, and 20% to 30% in overseas equities, 10% to 20% in bonds, and up to 5% in cash, is used by 13 of the 31 plans. Most of the balanced managed funds are actively managed, but four use a passive approach. A total of 18 plans offer a 100% equity fund as default - 13 of these are invested globally and five are invested only in domestic UK equities. The most common asset allocation for the global funds is 60% UK equities and 40% (capitalisation-weighted) overseas equities, although 50:50 and 70:30 splits are also in use. The majority of these 100% equity funds use passive management. [Table 1a about here] Since April 2005, all stakeholder default funds have been required to use some form of lifecycle asset allocation profile.5 While all default funds must use a lifecycle approach, Table 1b shows that there is variation in the manner in which providers implement it. The most common structure (involving 13 of the 31 plans) is to start switching from the equity or balanced fund five years prior to retirement, moving progressively to a final year allocation of 75% long-dated bonds and 25% cash. A further 11 plans use the same 75:25 final year allocation, but begin switching between six and ten years prior to retirement. 5 A previous version of this study found that in 2004 prior to the regulation change approximately 50% of stakeholder plans