The Problems with Annuities

20 July 2014

The abolition of compulsory annuitisation in the Chancellor’s March 2014 Budget, although welcome, has led to an outburst in the public debate. However, a couple of important issues have been largely overlooked or misunderstood. Firstly, random variation means a typical pensioner has to worry at the same time about two contradictory problems – either dying before managing to spend their pension pot or, conversely, living so long that their pension pot runs out altogether. Secondly, the concern that annuity prices might rise because only healthy people will buy them is misplaced because the higher prices could still be good value given that these healthy people are likely to live longer!

On 19 March, chatting to a colleague as I walked to my office after a meeting, I flippantly stated my simplistic view that “annuities are evil”. Imagine my delight when I got back to my desk to find that George Osborne had abolished forced annuitisation. Several weeks later, however, I am increasingly concerned about throwing the baby out with the bathwater. Taking a step back, let’s recap on what was wrong with forced annuitisation:

  • Making pensioners take the same income each year restricts their ability to deal with one-off payments, such as a new car, fixing a roof, medical expenses or even a special holiday.
  • Office of National Statistics (ONS) figures show that discretionary pensioner expenditure falls over the years after retirement but annuities fully maintain (or even increase) their value.
  • Too many people bought annuities from their existing provider without shopping around (and didn’t always buy the most appropriate policy to allow for annual increases, dependants’ pensions and so on).
  • Prices were higher for smaller pension pots, where costs swallowed up a disproportionate share.
  • Solvency requirements obliged insurers to invest in extremely safe but low yielding investments. Locking into low yield investments for up to 40 years isn’t necessarily going to generate good returns, which has a negative impact on the pricing of annuities.
  • With gilt yields low and arguably distorted by the Government’s policy of Quantitative Easing (QE), consumers were being forced to convert their pension pots in a single transaction at a time of artificially high prices.
The Budget pretty much solved all these problems at a stroke. Consumers will be able to draw money flexibly to meet their needs and can choose the right time to buy an annuity if they still want one. Providers will have to offer products that people actually want, rather than relying on the inertia of a captive audience. There are, however, a few widely identified problems created by the changes:
  • Moral hazard: Some pensioners may spend recklessly and fall back on the State.
  • Underestimating longevity: Retirees underestimate how long they are likely to live and so may plan to exhaust their capital prematurely (or keep hold of it for too long, “just in case”).
  • Adverse selection: Since unhealthy people won’t buy annuities, the prices will go up.
Having identified these, there are various steps that can be taken to mitigate the potential problems. For example, the Government has already noted the need to look at rules on deliberate deprivation for long-term care and the expected guidance at retirement is now to include an estimate of life expectancy. That said, the initial blithe assumption that all pensioners will manage their retirement finances sensibly always seemed optimistic and the recent analysis of the Australian model suggests there is strong evidence that people will not only spend the money as soon as it becomes available but a lot of us will spend it in advance, paying off our credit cards and mortgages with our pension pots at retirement and only then wonder about what we’re going to live off.

Having identified these, there are various steps that can be taken to mitigate the potential problems. For example, the Government has already noted the need to look at rules on deliberate deprivation for long-term care and the expected guidance at retirement is now to include an estimate of life expectancy. That said, the initial blithe assumption that all pensioners will manage their retirement finances sensibly always seemed optimistic and the recent analysis of the Australian model suggests there is strong evidence that people will not only spend the money as soon as it becomes available but a lot of us will spend it in advance, paying off our credit cards and mortgages with our pension pots at retirement and only then wonder about what we’re going to live off.

There are two other specific aspects of the acknowledged problems with annuities that warrant further comment. Firstly, the restricted investment opportunities available to insurers because of the solvency requirements seem to me to have been by far the biggest problem in reality. Gilt yields are currently below 4% per annum and many commentators believe that growth assets such as equities can provide a return of at least double that level. It is hardly surprising that investing in stocks that generate only half the available return leads to a disappointing level of income, especially given the effect of compound interest over 20 years or more.

The more flexible investment options during decumulation that will now be more available will allow pension scheme members to generate significantly higher yields, with only a marginal reduction in security. The most optimistic estimate I’ve heard of so far was a claim that members could double their retirement income and I’m comfortable that a 20-30% increase is perfectly feasible without any need to dramatically increase risk. Incidentally, this boost to retirement income from increased flexibility seems to me to have removed the primary competitive advantage of Collective Defined Contribution (CDC) schemes in the UK before any such schemes have even got off the ground.

Nevertheless, it’s important to remember that annuities can still be a useful tool in retirement despite the obvious problems with the low yields available. I expect older members may still find annuities very appealing from, say, age 75 or 80, when the lower yield has lower impact as it applies for fewer years and most of the annual retirement income will come from the withdrawal of capital either way. At that stage, the mortality protection and guaranteed income security starts to outweigh the drive to target the highest possible return.

That leads nicely on to my comments on the second acknowledged problem with annuities. Members have been very aggrieved about the poor value they obtain from annuities if they die early, with limited (if any) protection for dependants and an impression of the insurer profiting from their death. However, the real problem here is not poor value but poor understanding. A product that shares mortality risk among consumers so that those who happen to live a long time can still receive the secure, guaranteed level of income must by definition fund that at the expense of those who happen to die earlier. Giving members additional benefits on death just reduces the affordable level of pension for survivors. Besides, the insurer won’t make any mortality profit unless more people than anticipated in the assumptions die sooner than expected. So it’s completely unfair to believe annuities confiscate assets on death!

However, my primary concerns now are where the problems haven’t really been picked up in the public debate.

Demonisation of annuities: The adverse selection argument ignores the fact that higher annuity prices are perfectly fair if only the healthiest people buy them. These people will live longer (on average) and so can still get good value for their personal life expectancy. Smokers still buy life insurance (and I still have comprehensive car insurance despite being a terrible driver), so people do recognise that a higher premium to reflect your own higher risk is reasonable. Unfair public comment will undermine confidence in products that still have a useful role to play.

Random variation: Even if a 65 year old retiree is warned that he is likely to live to 86 rather than 82 as he thought, not many people will live exactly that long whatever we actuaries say (as per Harry Harrison’s short story “Not me, not Amos Cabot!”). He might walk under a bus tomorrow or could still be going strong at 100. How can he possibly plan his expenditure to avoid running out of money if he lives too long without keeping back too much if he doesn’t? This is where an annuity can still be the right answer, sharing the risks with early deaths and subsidising those who live longer, so that everyone gets the reassurance of a secure income for life, however long that turns out to be. The key here might be to balance the negative impact of a constrained investment strategy against the annual mortality rates, with annuitisation at age 75 or 80 to get the best of both worlds. (Sadly even that solution raises a further spectre of pensioners having to make important final decisions after their physical and mental capabilities have already started to decline and some may be struggling to manage their own financial affairs optimally.) At the end of the day, perhaps people won’t buy annuities so early but they might still need them later to deal with the random variation of the Grim Reaper’s calendar.
contact Hugh Nolan
Chief Actuary Hugh_Nolan@jltgroup.com 01727 775137