By John Southall, head of strategic research at L&G Asset Management
Collective Defined Contribution schemes (CDC) have attracted significant interest as a different way to provide retirement income. They can be a good solution for some members but, like any pension arrangement, involve trade-offs. While CDC offers genuine benefits, it also brings risks, some that are not widely appreciated. Understanding these is essential to assessing both whole-of-life CDC and its newer retirement-only variant, R-CDC.
Higher expected pensions
CDC schemes can offer higher expected pensions than other arrangements, although expected outcomes are only part of the story.
The main reason for an uplift is straightforward: more investment risk. A whole-of-life CDC scheme might invest around 90% of its assets in equities, for example. That higher allocation to growth assets increases expected returns and expected outcomes. This is simply a reward for greater uncertainty.
Pooling longevity risk can also increase expected pensions but also doesn’t come for free. Members who die earlier help support the pensions of those who live longer. Death benefits could still be provided, but only at the expense of some of the gains from longevity pooling.
Does smoothing reduce uncertainty?
Higher expected pensions are only one part of the story. Equally important is how CDC schemes manage the resulting uncertainty. On this front, CDC proponents may underline a key feature of CDC schemes: they smooth short-term changes in pension levels.
With no employer required to cover funding shortfalls, benefits must adjust when experience differs from expectations. Schemes typically do this by changing the scheme’s ‘indexation rate’ – the expected rate at which members’ pensions increase each year – to keep liabilities aligned with assets.
This smooths year-to-year pension changes, but it does not remove risk. Instead, members experience it through changes in future pension increases rather than immediate reductions in income.
Even small differences compound over time. For example, if pensions increase by 2% a year instead of 3% a year, pension payments are around 25% lower after 30 years. Many members may prefer this smoother approach because gradual changes are easier to plan for than sudden reductions in income. But the scheme cannot reduce investment risk by changing its payment policy. Rather, there is a trade-off: more stable short-term pension levels but greater long-term uncertainty in payments. Describing investment risk as “shared” in CDC, as if it disappears within the collective, is misleading. Smoothing shifts uncertainty towards longer-term payments and, consequently, from older to younger members.
CDC myths
There are other misconceptions in CDC. Below we list three other ‘myths’:
Myth 1: CDC reduces timing and sequence risk
Some argue that CDC schemes reduce timing and sequence risks. But this isn’t clear-cut. The smoothing mechanism creates an implicit de-risking glidepath because younger members are more exposed to changes in expected future pension increases, which compound for longer. In retirement, for example, members still face sequence risk because strong returns early in retirement lead to better outcomes than strong returns later.
Myth 2: CDC is a Ponzi scheme
Members already in a CDC scheme will typically want younger members to keep joining so they can transfer risk to them. However, CDC’s transparent rules mean it is not a Ponzi scheme, even if the implications of those rules for risk-adjusted outcomes are not widely understood.
Myth 3: CDC is irreparably unfair to young members
Intergenerational fairness is one of the most persistent concerns raised about CDC. Younger members must wait longer for benefits to be paid and bear more risk due to the payment smoothing mechanism. If these, and potential other factors, are not allowed for then young members may suffer a bad deal. This is a legitimate concern. However, in principle these factors can be allowed for via the terms on which benefits are exchanged. In other words, younger members could be compensated through more favourable exchange terms[1].
How Retirement-only CDC is different
There are two main types of CDC schemes. Whole-of-life CDC covers both accumulation and decumulation phases whereas Retirement-only CDC (R-CDC) applies only from the point of retirement using an existing pension pot. In the case of R-CDC, members make a single contribution at the point of retirement.
Although the expected uplift for R-CDC is smaller than for whole-of-life CDC, it remains material and most of the attractions of CDC remain. Many retirees will appreciate a ‘do-it-for-me’ approach that removes tricky investment and spending decisions. R-CDC pools longevity risk when it matters most, while continuing to invest in growth assets. Other structures, such as tontines or growth-linked annuities, could offer similar advantages, but neither currently appears to have a viable route to market.
Intergenerational issues are also far more modest in R-CDC than in whole-of-life CDC. This is because exchange only happens at retirement, timespans are shorter, and the investment strategy is typically more cautious. Even if a member turns out to be the last entrant, our calculations suggest their benefits are only around 5% less valuable than if the pipeline of new entrants had continued[2].
On the downside, our modelling indicates considerable risk of a relatively low pension in old age under R-CDC due to investment risk taking combined with smoothing. R-CDC also faces potential selection challenges relating to health and life expectancy, although underwriting ought to at least partially address this.
CDC is neither the miracle sometimes claimed by its supporters nor the disaster portrayed by its critics. Understanding the trade-offs versus other pension solutions is more important than focusing on headline pension increases alone.
[1] It turns out that the same techniques used to value financial derivatives can be used to place a fair value on uncertain future pension payments.
[2] Assumes that asset relative to liabilities have 8% pa volatility and the benefit adjustment mechanism involves using changes in the indexation rate when possible, but one-off scaling adjustments if the indexation rate falls below 0% pa nominal or above 5% pa nominal.