Ways to improve DC pensions
International policy best practice holds lessons for UK defined contribution pensions
- The UK regulator has sought to improve at-retirement choices by improving information provision
- Lessons can be drawn from other countries, such as Chile’s mandatory national annuity brokerage
- Danish variable-income products pool longevity risk
- Some Canadian funds create large economies of scale, to the benefit of the membership
When the UK government ended the quasi-compulsory system of pension annuities in 2015 it set in train a revolution in the way individuals acquired retirement income. But immediately the outlines of the policy challenge became clear: how to combine choices at retirement with the security of a reliable income?
The regulator of retail pension arrangements in the UK – the Financial Conduct Authority (FCA) – has identified expensive and opaque retirement investment products for sale in a market defined by the much greater information which accrues to sellers than buyers. The FCA retains the right to introduce a charge cap if matters do not improve (there is a 75bps price cap for DC workplace schemes in the accumulation phase) and is promoting retirement investment pathways which will simplify the complex options retirees now confront.
The FCA continues to place faith in the same kind of information remedies that failed to improve outcomes in the annuities market. More fundamentally, it is far from clear why the regulator should be more concerned about poor choices within product types, rather than between products types. It is highly unlikely that most savers are well-equipped to select between annuities and drawdown.
This is where the new mass workplace multi-employer schemes that have emerged with auto-enrolment come in. They are developing products that will combine income drawdown and an annuity. The annuities will be bulk-bought from the wider market and the best wholesale price passed onto members, thereby mimicking what Chile has introduced on a national basis.
The merit of the annuity element of the retirement product is that it ensures that members have longevity protection and that they cannot run out of money in later life. This is potentially much more important in the UK, where the state pension pre-tax is worth less than 30% of the average wage, compared with, say Denmark, where the state pension is worth 82%.
Most people will not have a diversified set of investments and will need protection against market vicissitudes for a core element of their retirement. The cost of a guarantee will be lowest if the individual pools the risk of living beyond the average age with other savers. In other words, the retirement product must at least protect against both investment risk and longevity risk.
The most common way of providing a guaranteed income in DC retirement is via an annuity. Purchasing such a product used to be mandatory for those considered to be mass-market purchasers in the UK and who were retiring with a DC pension. Pension freedoms removed the obligation. Annuities had become unpopular in the UK. This was partly because low interest rates reduced the income insurers could offer purchasers in exchange for their lump sum.
It was also partly because the market offered lower rates to those savers who stayed with the vertically integrated pension schemes with whom they had also saved – and this group comprised most savers. In the UK, the FCA has sought to tackle both discriminatory pricing and lower prices generally through successive attempts at increasing the quality of information made available to savers from their pension scheme.
International comparative studies indicate that Chilean retirees get the best value for money when they purchase an annuity. Since 2002, Chile has broken the direct link between the pension scheme and the saver by requiring that annuity purchasers use a mandatory and neutral national brokerage system, Sistema de Consultas y Ofertas de Montos de Pension SA or SCOMP. It provides prospective annuitants with the best three annuity offers in ranked order. Chile has been able to harness the market to work in the annuitant’s interest.
This is in contrast to the UK’s historical approach, which enabled vertically integrated savings and annuity providers to profit from the information assymetries which always exist in this space between sellers and buyers. The vast majority of UK annuitants continue to default to the annuity offered by the insurer with whom they have saved, even when it comprises poorer value. The Chilean national brokerage system also applies to drawdown.
In Denmark, annuities refer to products which have guarantees and also lifelong income products which do not have guarantees. In the case of the latter, if the value underlying investments decline, then the ‘annuity’ income will fall until the value of the underlying investments recovers. The Danish product is unlike UK drawdown because it is not an individualised product. Members pool longevity risk and this is why it is considered an annuity. In the UK, we would now probably refer to such a product as collective DC. Danish ‘collective DC’ (CDC) may or may not include risk-sharing across generations, depending on the scheme, but this would typically not be the case. If in the future CDC were to be more widely deployed in the UK, it is probably as a Danish-style ‘annuity’.
An important factor in average pension outcomes is contribution levels. UK automatic enrolment combined contributions from employer and employee are relatively low. The figures for the UK are about to rise to 8% of the relevant salary band. In Denmark, contributions typically started lower than in the UK, at less than 1% of salary and were incrementally raised. They are set in collective negotiations, so vary by sector – but are typically now about 12%, with two-thirds paid by the employer.
UK trust-based providers are legally obliged to put members’ interests first. This is not the case for the UK’s retail style contract-based providers. The latter are obliged to treat members’ ‘fairly’. This is an ambiguous term. It is also a principle that is not justiciable by consumers but is enforced (or not) by the FCA.
In practice, none of the main reforms of workplace pensions in the past five years have been initiated by the FCA applying its principles. They have, instead, required legislative intervention. EU law has obliged the FCA to include a requirement in its principles for providers to act in the best interests of the ‘client’ and this is potentially actionable in the courts. However, lawyers are not interpreting this as a fiduciary duty.
In the United States, advisers to whom regulators wanted to attach a fiduciary duty have lobbied instead for a client’s best interest standard. The view appears to be that a client’s best interests can more easily met by the provision of information, whereas a ‘beneficiary’s’ interest requires the provider to be more proactive in the saver’s interest.
The OFT’s 2013 report into UK DC identified the problem of conflicted interests and poor governance leading to poor outcomes. The OFT recommendation that policy should be used to promote “robust independent governance” eventually resulted in independent governance committees1. This model is a first attempt at delivering “robust independent governance” in the retail-style environment.
In Switzerland, as in many other countries, workplace pension schemes can only be run by trustees. Switzerland puts great emphasis on the separation of powers between the independent trustees and the executive functions of the pension scheme.
The vast majority of low and medium earners automatically enrolled into a workplace pension in the UK are in a DC scheme and the majority of those members are with a few larger providers. There is, however, a long tail of small, often individual company schemes. TPR estimates that there are just over 36,000 DC schemes in total2. There are about 2,180 schemes with more than 12 members of which only 80 have more than 5,0003.
There are scale benefits that accrue to pension schemes which lowers costs and increases the returns to members’ savings. These can potentially occur at two levels – at the administration layer and at the investment layer. Self-reporting by small schemes to TPR’s annual surveys continues to demonstrate a struggle to provide adequate governance and assess value for money4.
The total costs at the administration layer do not rise much as extra members are added (for example, the size and cost of operating the board of trustees and advisers to the board can potentially be fairly similar regardless of the size of the scheme), and as a consequence the cost per member drops as each additional person joins the scheme. Indeed, there is no evidence of any real limit to the benefits of scale in this layer (Bikker: 2013)5.
At the investment layer, scale enables pension schemes to negotiate more effectively with asset managers and drive down the costs of the latter. In addition, scale above a certain level allows pension schemes to take some investment services such as infrastructure investment in-house. The costs of doing so can be a third of what financial institutions would charge (Bikker: 2012)6, and, diversification into this kind of less-liquid asset may increase returns to members’ savings.
A 2012 report to the deputy prime minister and minister of finances in Canada found that pension schemes needed to have CAD50bn (€33bn) under management to operate most efficiently.
Chile, Denmark, Switzerland, Canada. Not a football world cup finals group. Instead, nations the UK might look to for inspiration as it seeks to a build a retirement income system which combines the flexibility that the ‘pensions freedoms’ reforms of 2015 introduced with the security and stability of traditional pensions, such as an income for life.
1 webarchive.nationalarchives.gov.uk/20140402194810/http://www.oft.gov.uk/shared_oft/market-studies/oft1505, p.167.
6 Bikker, J. A., O. W. Steenbeek, and F. Torracchi (2012). The impact of scale, complexity, and service quality on the administrative costs of pension funds: A cross-country comparison. Journal of Risk and Insurance 79 (2), 477–514.
Gregg McClymont is director of policy and external affairs at the People’s Pension and Andy Tarrant is an independent public policy adviser