An annuity is a reasonable solution to an obvious problem. You don’t know how long you will live in retirement, but an annuity provider can estimate this, take a risk, and sell you an income that lasts the rest of your life. In practise this hasn’t represented good value for money, so I welcomed the budget announcement to give retirees more choice in investing or disposing of their own money. Yesterday Janice Turner argued for the Collective Defined Contribution pension model as another solution to the same problem. I have my doubts about CDC and I want to suggest another alternative.
The purpose of an annuity is to ensure that you have an income for the rest of your life. If you don’t purchase one, and draw down prudently on your pension pot you would base the amount you draw on some estimate of how long you expect to live. And when you die there is likely to be money left, which you can bequeath, but is in a sense wasted in terms of providing yourself a pension. Or maybe not.
Suppose I place a bet with my neighbour, who is retiring on the same day as me, that I will outlive him, to the value of what is left in our pension pots when the first death occurs (the lesser of the two amounts). That bet would give the survivor an extra lump of cash to fund their longer retirement. No gilts, no actuaries, no profits, though we do have to trust each other not to murder.
There are two elements to the risk that annuity providers take:
- there is the risk that the individual will live longer than the average – insurance is good at protecting us from this sort of risk because these risks all average out; and
- there is the risk that life expectancies in general will rise more than expected – this is a big risk for insurers because they can’t average it against other claimants, they have to stand it with their own assets and so they will charge a big risk premium
From the point of view of the individual, the risks are more or less reversed. The variation between me and my neighbour is very likely, statistically speaking, to be considerably greater than any improvement in life expectancy over the next few years. The important risk from my point of view is the one that is cheap to insure, not the one I would be paying the biggest risk premium insuring against if I bought an annuity.
The solution is to turn my bet with my neighbour into a low cost, simple and comprehensible financial arrangement that protects me (unlike drawdown) from my own longevity, but not from increases in life expectancy in general. For want of a better name I will call it a mutunnuity: pensioners are mutually annuitising for each other, near enough.
If 1000 people of roughly the same age and health status all put £50,000 into a mutunuitty when they retire, that £50m is used to pay retirement incomes to those 1000 people using a prudent draw down calculation. As each of the participants dies, this leaves a greater share of the income for the survivors to balance the declining value of the pot.
This is not the fixed income of an annuity, but you would have to be extremely unlucky for it to drop below annuity levels. The chance of 30 people all living longer than average is 1 in a billion, so in a mununnuity of 100 or 1000 pensioners you are very well protected against this kind of bad luck.
Mutunnuities would allow, if desired, investment in shares for a higher return, at some risk – and this is surely the point of low interest rates, to encourage investment in real business.
The CDC pension model has some of these features, but is a complex feat of financial engineering that risks introducing new kinds of intergenerational unfairness. The mutunnuity is simple and requires no risk management from the provider, and is therefore very cheap to operate. It eliminates the greatest part of the longevity risk in a transparently fair way. Almost all the benefits of income drawdown combined with almost all the benefits of an annuity.
* Joe Otten was the candidate for Sheffield Heeley in June 2017 and Doncaster North in December 2019 and is a councillor in Sheffield.
Follow @libdemvoice.org on Bluesky
Like us on Facebook
Subscribe to our feed
Sign-up for our daily email digest
Back on the campaign trail
1963: London's summer of the Hoverbus
Have resignations led to a rethink?
Steve J Smith
nigel hunter
Chloe
Abrial Jerram
30 Comments
This proposal is merely an annuity without any safeguards for the “policy holders”. There is a third, hugely significant risk that insurance companies shoulder & that is the class of financial risks (the two biggest being inflation & investment returns). This class of risks will far outweigh the impact of (lighter) mortality, which is why one sees the annuity rates closely tied to predictions for future interest rates (this is the principle reason for their decline, not mortality changes). It is not clear that any attention has been given to this class of risks and how this type of fund would be any better for the fund holders than an insurance company guaranteeing their income.
A fairer summary of this proposal is “almost all the benefits of income drawdown combined with almost none of the benefits of any annuity”.
John, the issue is how much it is worth insuring against future investment risks, and that would determine the balance of gilts and equities in the mutunnuity. It is a little arbitrary at the moment that a day before retirement it considered fine to have all your pot in shares, and the day after to have it all in fixed income. Whatever your risk appetite, a more gradual transition would be reasonable.
Sure, if you want to buy insurance against inflation, by all means go for an index linked annuity. Then your little-more-than-zero income will go up with inflation.
You are right that I didn’t mention inflation and investment return risks explicitly, but they are both in the category of expensive-to-insure, and pensioners are likely to be better off standing those risks, and insuring only against their own longevity.
Mortality risk is a major issue for final salary (defined benefit) schemes because the provider is not only estimating life expectancy at the age a person retires, but the future life expectancies of all those who have yet to retire, bearing in mind the formula for calculating the pension is exactly the same whether the person is 20 or 60. When buying an annuity at, say 65, the insurance company only has to estimate the life expectancy of everyone else retiring at that same moment in time. If life expectancy has increased for those retiring 10 years later, the annuity rate can therefore be adjusted downwards. This is not possible with a final salary scheme, and so the sponsoring employer has to make up the difference by pumping more money into the scheme. This is of course why there are now so few private sector schemes with future accrual, let alone open to new members. On the other hand the public sector continues to enjoy the benefit of this type of pension on the basis that future governments will simply put up taxes, or cut other services, to meet any shortfall.
Yes. They are expensive to insure and that is why insurance companies charge a lot to do so. These products aren’t anywhere near so bad value as some attempt to make out. If these companies were scamming the public, we’d see insurers being taken to task by the competition authorities – are they being so investigated or condemned?
@ Graham
Not for much longer. Most, if not all, public sector schemes are moving to career average schemes shortly. And you are right that a company scheme has a different risk profile compared to insurance companies. Even so, the financial risks are the most volatile and influential on the costs of pension provision.
Hi Joe, from an adviser’s point of view when it comes to retirement I am making looking at risk and reward. If someone doesn’t want the low returns of an annuity, but also doesn’t want the relatively high risk of a shares based drawdown plan, then a low cost bespoke hybrid can be constructed by having a drawdown plan invested into bonds. There are also plenty of hybrid annuities that usually guarantee a minimum level of income, with the rest depending on stock-market performance. These are more expensive, but they work when it comes to matching a client’s risk profile.
I think your idea could be suggested to the private sector to see if they think there would be demand for it and the state could have a hand in it too, but I don’t think it is a major policy concern because there is already a lot of hybrid options available.
Having said that, I really appreciate the kind of “industrial activism” public and private partnership suggested here, as Peter Mandelson would have called it. There is a roll for the state, it just has to be balanced. I would recommend the state’s involvement to be possibly greater protection for the consumer in case of insurance companies going bust, or simply tweaking the state pension to improve it again.
Advisers can also try to predict market conditions, but it is risky and at the lowest levels I think advisers are best not trying to be too clever :). If an adviser is too late in predicting interest rate rises the client’s pension pot will fall, and even though annuity prices would also fall, the client has paid higher advice fees for nothing. I think this is relevant to the whole argument of “waiting for annuity prices to improve”.
I think Joe Otten’s idea is worth further consideration. I share his doubts about Collective Defined Contribution, which seems to me as horribly opaque as with-profits life policies.
I thought the reason why annuities are so expensive is because they have to be backed by index-linked gilts, of which there is a severe shortage. This is necessary because, as John Broggio says, annuities protect against inflation and against variability in investment returns. If customers were willing to accept variability in investment returns, the fund could be invested in shares and that would automatically protect against inflation.
A problem with the Mutunnuity as described is the assumption that all members in one cohort would have the same health expectations. As with annuities, there would have to be some recognition of impaired life in the form of higher income, but it’s not impossible.
More thought needs to be given to how the Mutunnuity works when only a few members of a cohort remain, and ultimately only one. That is when the fund might run out to soon, and, if it doesn’t, who would inherit what is left? The CDC at least avoids that problem.
As an alternative, perhaps the government could issue much larger amounts of index-linked gilts and so make annuities cheaper.
Nicholas, it’s a fair point, and I wanted to spare the detail in the original description. When a scheme gets down to its last few members, as presently defined there would be considerable volatility in incomes, with the direction depending on whether the last few members were dying faster or slower than expected. There is also the problem of what to do with the rest of the pot when the last member dies. As it stands it would be part of that member’s estate.
But it would be possible to convert to annuities for the last member, or even for the last 5, or 10, or 20 members. This would give a guaranteed income to those who would otherwise experience the most volatility. Another possibility would be to consilidate with another mutunnuity also down to its last 20 or so members. A third would be that leftover pots on the death of the last member go into a fund to prop up mutunnuities where the last few member were dying slower than expected. This would be complicated to do fairly because all funds in the scheme would have to operate with similar levels of prudence (to earn the same protection).
It would also be quite reasonable to have higher rates of state pension kick in at 85, 105 etc, in recognition of the difficulty there is in funding such a long retirement.
You are right that not everybody has the same life expectancy. In principle it would be possible to accommodate differing people in the same scheme, but that would rely on actuarial calculations that we can otherwise avoid. Similarly people could invest differing amounts, but there would have to be some mathematics to ensure payouts remained fair. (It’s not quite as simple as getting a payout in proportion to your stake, because the last two members stand to get the whole remaining pot however much they put in.)
Joe’s ‘winner takes all’ proposal is an interesting idea but it is a lot more difficult than it might seem at first glance.
First, everyone in the scheme has to be about the same age. Then they all have to be health screened to ensure that, so far as is possible, they all have the same life expectancy. So quite a few are likely to be rejected at that point. The final group is likely to comprise quite healthy middle class individuals who are prepared to take the financial gamble. However, mortality rates are calculated from data on a vast range of people, so the chances of a carefully selected group of health screened individuals living longer than currently predicted mortality rates are actually quite high. And what happens to the people too sick to be included?
This also doesn’t solve an additional problem around current DC scheme design: that every individual has to make their own financial decisions on how their own pot of money is invested.
Collective DC has many advantages: key to it is the collective nature of it. Individuals don’t have to make their own investment decisions – this is done for the entire collective pension pot by the scheme’s trustees, which cuts out more fees. And the scheme can pay pensions out of the scheme’s assets instead of people having to go away and buy one – yet more fees saved.
Simon McGrath misunderstands this: he considers the current drawdown option to be the same thing but, like the current annuities, it involves going to the financial services industry and purchasing such a product. While he may be keen to give away some of his pension pot to the city in fees and profits, it would be far better for trust-based schemes to provide pensions themselves at cost.
And in a CDC scheme you don’t exclude people because they are too healthy or not healthy enough. This trend of recent years towards extreme individualism has produced a situation whereby it is in the individual’s financial interest to take up smoking and drinking half a bottle of whisky a day when they’re nearing retirement in order to obtain a better DC pension. And even the mutunnuity proposal carries the warning that it’s in your fellow members’ financial interest that you die as quickly as possible!
Janice, first of all Joe’s idea is not a “winner takes all” proposal, which is the kind of language that irritated me about your article. Secondly, if you are so against going to the City for insurance then you are going to be going to the Treasury instead. Why don’t you make this clear?
Someone who is more of a “historian of business” than I am might be able to give a fuller explanation, but I do wonder whether what Joe is suggesting isn’t rather like the “Society for Equitable Assurances on Lives and Survivorships”, founded in 1762. For over 200 years that particular (Mutual) Society worked very well.
Simon, that’s Equitable Life, right?. Not really. Yes we’re talking about a mutual concept. Equitable Life killed itself off by selling products to customers as if it weren’t a mutual, making promises to some customers it could only keep by robbing the pockets of others. Mutuality made this worse because there was no shareholder capital to fall back on.
The provider of a mutunnuity by contrast is not taking on any risk, but simply enabling the biggest risk to pensioners to be shared by a group of pensioners. And by sharing, largely eliminated.. The provider is in a sense irrelevant and this is a mutual arrangement between pensioners.
Possibly Equitable life started out with this kind of product, and it is easy to see how it might have evolved into what it ended up as. The temptation is always to add financial engineering, to add “guarantees”, and regulation, and marketing may push things in this direction. But Equitable shows us that was a mistake.
Products without cash-defined “guarantees” may be difficult to sell, (though the annuity guarantee isn’t available when a pension is first taken out, so the guarantee doesn’t exist at the most important point anyway) and I am trying to address this by offering a comprehensible product. The cash return isn’t guaranteed, but you know at any point how big the pot is, and what share of it you own, as long as you stay alive. You can’t say the same with most financial smoke and mirrors products.
Janice, nobody would want to reject unhealthy applicants – the problem is the opposite one. If your health is poor, a mutunnuity is bad value for you in the same way, for the same reasons as an annuity. If this idea is adopted, there would be a market for impaired health mutunnuities in the same way as there is for impaired health annuities. These problems are all faced by annuities in much the same way. My comments on CDC are in the thread under your article.
@Joe Otten
Yes indeed, that’s Equitable Life, and I assume that for 200 years, until 1962 at least, it did very well behaving as a true Mutual.
Joe, I’m confused as to how these will provide better value for money than what is currently available in the market place. If the risk is pretty much eliminated then they will just end up providing similar income levels to annuities. Is the idea to just put another competitor / business structure in the market-place in order to hopefully get better value all around?
Eddie, look again at my example bet with my next door neighbour. Can you see that even a mutunnuity of 2 members gives me extra security in retirement compared to income drawdown at the expense only of what is left when I’m gone.
Sharing a risk like this reduces it dramatically. This is one element (the cheap one) of insurance.
Annuities contain other, more expensive elements of insurance, as I explained. Index-linked annuities even more so, which makes them bad value, particularly at the moment.
Hi, yes I can now see, I was a bit confused. So to keep it simple the idea is that you effectively insure your drawdown at the cost of any inheritance. It sounds like it would appeal to people with no dependents, but then again it now sounds awfully similar to a regular annuity without a spouse’s pension. My position on mutuals is I think they are fine as long as they aren’t funded by the taxpayer, which I am worried this would be.
Interesting discussion between knowledgeable people, but what is the role for government, and hence a political party. Very limited I would suggest.
Why can this pot of money be managed so much better than the insurance companies do for their annuitants? I just don’t understand what insurance companies are (all) doing so very wrong if these mutunnuitants can get orders of magnitude more money for the same durations as normal annuitants.
I back the chancellor on this. People should be allowed to spend “their” money as they wish.
There is a philosophical issue behind this which needs do be addressed. Who is welfare for, is it for those who hit hard times and need a cushion for support or is it a service for all (including those on middle/high incomes related to contribution. Those who advocate a contributions based welfare system (and there have been many from all parties doing this recently) are advocating the latter. Whilst this approach may appear fair to some, it is definitely more expensive than the “cushion” option as many will be benefiting including many who could look after themselves.
At a time when public funds are squeezed, we all need to ask the question above. Restricting welfare to taxation based subsidy to the needy will hold down the welfare bill substantially, but will remove the concept of personal contribution, a concept which I find dubious anyway.
Applying this to pensions. The state provides the state pension, a flat rate for all. State involvement should stop here (unless the recipient is disabled and needs extra help) . Whilst encouraging saving for pensions, subsidies and ruls should be minimal. For example I support limiting pension contribution tax relief to the basic rate
A few years ago, I developed a model that was similar to this one, but I also tried to eliminate the element of constantly paying down the fund and hoping that there would be enough in the pot to keep the final survivors going to the end.
I think I succeeded.
People within a certain age band join the scheme, each buying a number of units. The fund pays out income (ie interest or dividends, depending on how it’s invested), but does NOT pay down the original capital.
Whenever the member dies, their units are allocated as bonus units for the remaining members. Each member cashes in their bonus units over a period that they decide (for example, five years); they can vary the period as they grow older. This means that, each year, each member is receiving the income from their original units, plus a proportion of the bonus units, ensuring a reasonably steady income.
When a member dies, and still has some bonus units, those units are cashed in for their nominated beneficiaries or their estate.
The fund is not limited to its original members: the age band rises with time, and new members who are within that age band can continue to join.
However, there will come a day when the final member dies (unless we conquer death, which would be nice but I’m not counting on it). Their remaining bonus units and their original units are cashed in and go to the member’s nominated beneficiaries or their estate, and the fund comes to an end.
The principle is that each member benefits from an increase in their income as other members die, and the member’s capital remains in the fund for the whole of their remaining life, so there’s no risk of it ever running out.
One great advantage over annuities is that ultimately everything is distributed. You don’t end up with another scandal of “orphan assets” that have accumulated in the fund and can be siphoned off by the fund manager because they “don’t belong to the remaining members.”
There are a few features that might be considered problematical. Of course there are – there is no perfect solution:
* Each member’s starting income will be relatively low. It will only build up as other members start to die and will continue to grow as the membership becomes older and loses a higher proportion of the survivors each year.
* By the time the age band hits 100+, there won’t be that many members left and the numbers of bonus units awarded to surviving members each year will become increasingly lumpy; this is addressed by members cashing in their bonus units over a number of years, so that the bonus income is evenly spread.
* the Victorians operated “tontines”, a sort of winner-takes-all bet in which a group of citizens each put a fixed amount into a pot; the funds stayed into the pot until there was only one survivor, who then claimed the whole of the pot. The problem was that basically that this created a significant incentive for the families of other members to hasten the demise of their competitors. This risk is minimised by avoiding the “winner takes all” approach, and by maintaining the anonymity of the membership.
There’s been some comment on the different members’ varying investment needs and attitudes to risk, and on the problem of some members being higher risk than others.
The former is a bit of a red herring: the plan’s investment approach has to be stated up front, and anyone who doesn’t like it can steer clear or choose a different plan. There might even be scope for members to switch between different plans with different investment strategies (but with the same age band).
The latter is pretty straightforward. Anyone with an “impaired life”, ie a reduced life expectancy, will get less benefit from this model – but it may still be worth considering. If an annuity provides a better likely return, they should go for it!
I wonder if there is a better way to achieve Joe’s objective, which I believe is to provide the benefit of income for life without the extra cost of making that income constant (in real terms). How about an annuity invested in shares, property, etc from which the income is adjusted each year as the value of the assets varies?
In a conventional annuity the invariability of real income is ensured by purchasing expensive index-linked fixed income instruments. In my proposal, the annuitant would have to accept that if share values drop their income would drop, and would recover when share prices recover. Dividends usually continue when share prices fall, and so the drop in income would generally be less than the fall in share prices. I think possibly this type of investment-linked annuity may already exist; certainly there are (or used to be) with-profits annuities which would be similar.
If the prospect of a temporary drop in pension income is unattractive, customers could have a combination of conventional and investment-linked annuities, the proportion in each being based on the individual’s trade off between higher income and steadier income.
Joe,
A couple of observations, your assumptions are off but that doesn’t mean that this is not appropriate as a product to offer some people.
The assumptions that are off and giving you a positive skew are:
– The likelihood of those investing will be richer and therefore often have a higher life expectancy.
– Also as a voluntary scheme people with bad health conditions will be unlikely to join so the life expectancy will be higher than the numbers you used.
The assumption that is off and giving you a negative skew:
– You refer to “intergenerational unfairness” that would only apply if it operated in the Equitable Life model that you are keen to differentiate from. If this were a sealed pool of people all retiring at the same time and same age then there would be no intergenerational unfairness.
Your other point of differentiation from EL is that you appear to be talking about a fund structure which allows it to prevent it getting involved in distracting new activity and it should just focus on a single activity, to pay a return. A simple model as you suggest should help keep focus. I suppose this an annuity with more risk but potential for a better return, perhaps a hybrid that would find a market.
I can’t say I would use one as I expect my health on retirement to be poor and therefore not worth getting in collective schemes. I think the market for the impaired health products would be limited as illnesses have such a range.
Nothing to rule out this product though, but as Eddie says there may be existing products on the market that it could struggle to compete with (or it may perform so well as to kill off those products).
In normal conditions an annuity is best because an insurance company takes the longevity risk. Insurance companies have a naturakl hedge against longevity risk if they also write life insurance premiums. So if people live longer and insurance companies pay out more than exepcetd, they will also pay out less on life insurance.
Unfortunately, when long term interest rates are low because of quantative easing, then gilt and bond yields are low and insurance companies can only afford to pay out low annuities. When quantative easing is ended and interest rtaes are back at 3% above inflation, then annuity pay outs will be far higher.
Sorry, genius, but Lorenzo Tonti beat you to this idea by 360 year or so.
Most people DON’T WANT their life to be turned into one big casino game. The idea that something as fundamental as what money you will get from retirement till death should be based on where you placed your chips in this stocks and shares and whatever game is repulsive. At least, it is for most people. For the social elite and well informed, who do not know what suffering from lack of money is like, maybe it seems like fun. For most it doesn’t, it’s just a cause of anxiety. Most people aren’t financial experts and don’t want to be financial experts, and fear for very good reasons (see all those financial products misselling scandals) that all this sort of thing is just a trick to diddle them, it’s a way to create this vast class of salespeople and money-shufflers and advertisers all claiming their cut for doing something which in the end is non-productive.
It is a mark of how out-of-touch the Liberal Democrats now are that people who are influential in their policy making seem obsessed with fancy market ideas, complex games in which we are all forced to turn into wheeler-dealers.
Matthew Huntbach
“Most people DON’T WANT their life to be turned into one big casino game”
So let’s refine that a little more:
People don’t want uncertainty. Agreed for the most part.
People don’t want to risk being rained on if they plan a day out with the family in two months’ time either. Uncertainty is a fact of life, risk exists in everything we do, EVERYTHING.
To tell people that you can and will take away all the uncertainty in their lives is both dishonest and not credible.
The facts are that no system is perfect all will involve some element of risk the best that can be done is to allow people to reduce this as far as practical. Unfunded public pension commitments, “guaranteed” funded returns all have negative consequences. Often voluntary schemes where people get the most control will mean that it is a question of everyone gaining or loosing some at the margin. When you attempt to construct big schemes that claim to remove uncertainty you tend to get big winners and losers depending, determined by when people retire.
Most people I know accept that life is inherently uncertain and may not like it but recognize it. They also know it sometimes is possible to limit the uncertainty.
If you actually try and tell people you can guarantee the future they tend to be very pleased initially but then once you are gone the questions start to emerge, and not only do they not think they are any safer they will be a little more cynical about people like you making, unrealistic claims.
Matthew, yours is the sentiment the annuities market relies on. It sells a certain and low income for a high price, and makes a tidy profit doing so. I’m looking for a better way to manage the risks that life throws at us, thematically, being part of a mutual rather than a ripped-off customer.
To be clear are you saying that it is repulsive that risk exists? Or that it is repulsive not to show gratitude to the corporations offering to take that risk away from me at whatever price?
Yes, everything ought to be just fair, dammit. But I don’t see how your revulsion helps.
Toby,
Yes your idea looks very similar. You are specifying the drawdown rate in terms of interest + bonus units. I didn’t specify the drawdown rate, I think that should be determined by actuarial calculations, to give a reasonably high probability than incomes won’t have to fall, and I’m guessing that it can be done more quickly than interest + bonus units, and certainly more quickly than an individual can prudently draw down.
Interest-only would give pensions as bad as annuities initially, then possibly rising faster than necessary later on as bonus units kick in. I suspect a higher starting point and less increase would be better.
Thanks, Joe.
You refer to “drawdown rate” under my model. At the risk of sounding pedantic, actually, this was income – the cash generated by the investments – not drawdown of capital.
I was trying to scrap the whole idea of drawing down your own capital, which is based on the hope that it will outlast you (but not by too much).
I also wanted to replace the annuity concept, whereby your drawdown risk is pooled with the drawdown risk on thousands of other annuitants.
Instead of drawing down your own capital, you draw down the capital of the other members who have died. So you get the benefit of drawdown without losing any of your own capital, which remains in place at death – probably plus a modest sum for your estate.
Of course, the real losers are all those who die early on, getting little benefit from their original capital which goes to the surviving members. But that’s the nature of pooled risk.
You are quite correct: the initial income from your units will be exactly what it would be if you invested it yourself (let us say, if invested in the FTSE All-Shares Index, 3.5%) less running costs, and depending on the model this may even be less than you would receive from an annuity, but it would grow as other members die.
If I’m an investor with a large investment pot that I want to be able to enjoy fully in my retirement years, that’s very attractive.
If I’m gradually moving into retirement, with a declining income from employment, again, it’s very attractive.
What’s the point of all this? Because each member’s capital remains in the pot until their death, the risk is eliminated. And no need for expensive and cautious actuarial calculations. And no life assurance company looking to scoop up all the leftovers (“orphan assets”) when all the fund members have died.
Perhaps I should also say that, as you’ve indicated elsewhere, that there is room for more than one model. Mutunuities will not suit everyone; neither will annuities, nor my own model.