By Alex White, Head of ALM Research at Redington
For example, how to account for unpaid but necessary work like childcare in a “get out what you put in” system like DC; or how to get more money invested in infrastructure, as the Mansion House statement advocates for. But even at a more basic level, DC struggles to meet member needs, and one big area of difficulty is longevity.
Taking a step back, life insurance works because longevity is statistically somewhat tenable. We can get decent estimates of how many 75-year-old people will still be alive in 10 years’ time. But we know much less at an individual level.
From a pension perspective, longevity risk comes in two flavours: central and idiosyncratic. In a central case, the population could become more (or less) healthy, we could see new treatments or new diseases, and there’s a general level of uncertainty. This is always present, although tends to be smaller than many people might think over longer periods. Idiosyncratic risk is the risk that, even if there is no dramatic population-wide change, one group randomly happens to live a lot longer (or shorter).
Unsurprisingly, this risk is much larger for an individual than a group. Simulating lives through a mortality table, a group of 10,000 pooled members with a life expectancy of 18-20 years might have a standard deviation of 0.1 years (of life expectancy). With 1000 members, this rises to 0.3 years. With 10 members, it’s 2.4 years. With an individual, it’s 7 years.
It’s hard to emphasise how large that really is. As a simple approximation, if you retire at 65, with a life expectancy of 19 years, there is around a 1-in-3 chance that you will either live for fewer than 12 years or more than 26 years. How do you plan for that? You either spend far less than you saved, or you have a material chance of running out of money. This is an even bigger problem for poorer people who are less able to bear the risk.
The answer is pooling. Idiosyncratic longevity, arguably the largest risk for an individual DC member, is a rounding error in a large pool. DB is one way to pool longevity risk, but not the only one. CDC, decumulation-only CDC, or even DC with longevity pooling can dramatically reduce this risk. These solutions don’t need to be perfect to be a meaningful improvement over the current structure.
This doesn’t come for free though – survivors are, at root, funded from the pots of those who die early, which means this is less viable for those who can afford to leave large inheritances. For much of the population, however, these considerations are secondary to affording day-to-day life; and, without wishing to be too brutal about it, your living costs tend to go down when you die.
The push for increased freedom made a certain amount of sense when rates were very low and annuities would struggle to provide enough income. But with rates having risen, and while newer, more ambitious alternatives are still being planned and designed, I hope to see more and more retirees taking up annuities.
|