Five insights on the decumulation phase

25 September 2018 | Webinars


 Remove  Save


How do retirees behave in the decumulation phase? In this webinar, Cynthia Pagliaro, senior research analyst at the Vanguard Center for Investor Reseach, speaks to Alice Shepherd, investment writer, about five insights on retiree behaviour.

This is an edited transcript of their conversation.

Alice Shepherd: Broadly speaking, retirement has two distinct phases, accumulation, and decumulation. Following the end of compulsory annuitisation, what do we know about how retirees behave in the decumulation phase? The assumption is that they gradually run down their savings, but the reality may be more complex. My name is Alice Shepherd, I have with me here Cynthia Pagliaro, senior research analyst at the Vanguard Center for Investor Research.

The Center conducts and sponsors research into investor behaviour and decision-making Cynthia is going to share with us five insights on the decumulation phase. Cynthia, welcome to our monthly adviser webinar, and thank you very much for being here. If you're having trouble with the quality of the audio, there should be a window on your screen giving you instructions on how to dial in directly on your phone, and this should help to give you a nice, clear sound.

For our live listeners, you should soon see a box into which you can type questions; we have a large audience today, so, if we don't get to you personally, we will definitely follow up. The call will last approximately thirty minutes, and the webinar earns participants a half hour's worth of CPD points; you will need to request your certificate by contacting your usual Vanguard representative, or by writing to, or by phoning 0800 917 5508.

Now we've covered the logistics, I'm going to hand over to Cynthia, who is going to take us through her research; I'll try not to interrupt too much, but I may interject now and again with questions, if that's okay. So, Cynthia, over to you.

Cynthia Pagliaro: Great, thank you Alice; good afternoon everyone. I'm really delighted to be here with you today to share a series of insights on retirement, or what we now call the decumulation phase of life. And this presentation revolves around really one simple question, and that is, are retirees behaving the way that we think they are?

And in the Center for Investor Research, we're a behavioural science unit, and one of the important distinctions we draw, is that between normative and positive behaviour. So normative behaviour is what you should do; positive behaviour is what you actually do, and this phase of life is really a perfect case study to understand the distinction between the two.

And that's because many of our mental models of retiree behaviour are based on normative statements, and that is what we think retirees ought to be doing.

So, today I'm going to provide you with a different view of retiree behaviour, and that's insights based on academic, and industry research into exactly what people do in retirement. Now, whilst most of the data is from the US, people are people, so I expect you may observe similar behaviour among retirees in the UK, despite the difference in our systems.

So, for today's talk, I'll start with a bit of motivation, then we'll look at five key insights, and briefly mention some implications for advisers.

So, as I said, we have some mental models about retirees and how they behave, and probably the first and most profound in our minds, is this idea of the life cycle of retirement savings. There is typically an accumulation phase, and that coincides with the working years where people typically save for retirement, then, there's a very specific date, that is the date of retirement, when households stop saving and they begin to regularly draw down their assets to meet their retirement needs.

Another model we have, is that we have this idea that traditional pensions have largely disappeared from the landscape; now again, I'm talking about the United States here, because in the US we've known a near 40–year transition, from defined benefit to defined contribution schemes.

So, we now assume that retirees will no longer receive regular pension payments, and they really need to figure out for themselves how to generate spending from their DC accounts, to support their lifetime income.

And lastly, another important mental model we have is a very traditional actuarial model, and this model is really inspired by the world of traditional pensions. Now, the data that I'm showing you here, it's for certain government employees in the US, and trust me, they do very well in retirement.

So, typically what they'll do, is they'll earn a household income, a regular wage, and then they will replace a portion of that income, typically 70–80%, with regular streams of income; so, they'll get regular payments from their pension, they'll also probably receive a social security, or money from our state pension system, as well as investment income.

The point here is there are regular sources of income, and these streams are very predictable, and they also grow over time in the face of inflation.

Alice Shepherd: Cynthia, looking at this chart, it's obviously a US chart, how would we be able to extrapolate from the US: how would it change when we're talking about UK investors?

Cynthia Pagliaro: So, what I would say, is for someone who receives income from a traditional pension, I suspect this picture will probably look very similar, and we're talking about high replacement rates here of 70–80%, so I suspect that it would probably be pretty reflective of somebody who's probably a more affluent retiree.

Alice Shepherd: Great, thank you.

Cynthia Pagliaro: You're welcome. So, these are some of the conventional models of post-retirement income that we have, and yet actual retiree behaviour, that positive behaviour I was referring to earlier, actually bears little resemblance to these beliefs.

So, in a variety of studies, the evidence suggests that one, retirees actually do not spend down their assets in retirement. Guaranteed income is still a prominent source of income for many; many households actually continue to safe into retirement, and yes, while individuals strive to take a regular income from financial accounts, they also take many ad hoc and unpredictable withdrawals from their savings, and this really completely violates any norm like that predictable 4% spending rule we always think about.

And finally, there's this dawning recognition that the reason for this behaviour, is really motivated by a profound concern about long-term or old age care. So, the risk of needing to pay out of pocket for expensive care, even though it's decades off for many younger retirees, really dominates decisions in a powerful way.

So, let's look at evidence for each of these points; so, first, let's look at the fact that retirees really aren't spending down their assets; so, let's look at the change in the wealth of among three household groups. So, long the X axis you see we're looking at people from age 65 to 90, and we're looking at three population tertiles.

So, the blue line is the richest third, the red line in the middle is the middle third, and the green line is the bottom third, and I do want to mention these values are in real terms, so the lines are increasing at various stages ahead of inflation.

And what's interesting to see, is these lines are actually increasing and not decreasing; so, we would expect perhaps to see affluent households see the largest increase, but what is surprising, is that even the middle and low income tertiles continue to basically maintain or slightly increase real wealth as they age, and we call this the wealth decumulation puzzle.

So, all economic theory would suggest that all three of these lines should be declining with time, but they're actually heading in the other direction. And this isn't an unusual finding or specific to this one research paper – there are a number of other academic studies in the US that have found similar results.

In fact, this isn't just limited to the US; there are other studies in other countries that find similar behaviour, and I would point to one study of Norway, and Norwegian households that actually observe increasing real wealth over time as well, so it's truly a global puzzle.

The second point that I'll make, is guaranteed income is actually still very prominent among retirees, and it's an important benefit for many; so, for this and the next few slides, I'm actually going to reference some research we did, work we did several years ago, and that was a survey of about 3000 retirement-age households, not necessarily Vanguard clients, that had about 100,000 US dollars in financial assets.

And one of the aspects of this survey, was to document all the sources of retirement wealth these households had, and from this information, we were able to categorise these households based on their predominant wealth holding, and I do want to mention this does exclude housing wealth.

So, here you see these eight wealth profiles; so, the two blue profiles represent what we consider traditional retirement investors, who have large, or their predominant sources of income come from guaranteed sources, either social security, or our state pension, or they actually rely on traditional pensions.

Then we have two green groups, and these are what we call new retirement investors; again, that transition to defined contribution occurred 40 years ago, so, we see a rather large bucket of retirement investors whose predominant source would be either defined contribution assets, or personal pensions like individual retirement accounts, and we also have taxable investors.

And then you'll see a scattering of several other smaller groups; the important point of this slide, is noting the size of that pensioner box; a full quarter of all of these households still rely on traditional pensions as a primary source of income, so that DB to DC transition was not as severe as we thought.

Alice Shepherd: Cynthia, I wonder, would you be able to comment on what that diagram might look like if we were talking about UK investors?

Cynthia Pagliaro: Oh yes, sure, Alice; so, if I were to look at this chart for the UK, knowing that pension assets still are still pretty much dominated by defined benefit schemes, I would suspect that that pensioner box would actually be much larger in the UK. And again, we're talking about wealthier retirement-age households here, so I do think that box would be bigger.

And when I looked at the asset and wealth survey from this year, I did notice that household wealth for retirees in the UK is predominantly dominated by pension.

Alice Shepherd: Okay, thank you.

Cynthia Pagliaro: You're welcome; okay, moving on to our third point; savings behaviour actually continues into retirement; so, it's not something that actually stops, it really does continue later in life. So, from that same study, we asked investors to illustrate how they take their income and spend it and save it, and here we have all of those eight wealth profiles lumped together, and it's a wonderful picture of the activity, though it does look a bit daunting.

So, let's start in that middle bar; so, that middle bar shows all the sources of income; so, for convenience, we've grouped everything together so the green bars represent cornerstone accounts like retirement assets, or mutual fund accounts, or traditional savings. The blue are the guaranteed income sources; white are wages, and the orange is other income sources, and interestingly, affluent households do have a lot of miscellaneous income sources.

Now, to the left we see how that money gets spent, typically on routine or discretionary expenses, and to the right we see how that money is saved. The important point here, is to note that whilst 69% of money is spent, a full 31% of money is actually saved, and this is a really large and surprising number, and it was certainly evidence to us, that savings actually continues into retirement.

Now turning to our fourth point, withdrawals from financial accounts are modest and ad hoc. So, this is really the other side of the coin; we know investors are not spending all the money they get from guaranteed income, nor are they spending all the withdrawals they take from financial accounts. And when they do withdraw from their financial accounts, it turns out they generally take very small amounts, and then an important proportion of these withdrawals are not regular income streams, but actually ad hoc withdrawals.

So, looking at that same survey, this chart actually illustrates how modest withdrawals from financial accounts really are; that is true at least today, while people are still receiving some type of pension income.

So, if we start with the blue line, we could see at the very left, 22% of households in our study withdrew nothing from financial accounts. I mean zero, that's zero. If we move to the next blue bar, we see 31% withdrew somewhere between 1% and 3%, and then an additional 14, somewhere between 3% and 5%.

Now remember, our households save a portion of withdrawals, and they reinvest the savings elsewhere; so, the green bars actually reflect true spending, and what's important to note here, is a full 30% of people in this group spent nothing, nothing from their financial accounts. Another 36% spent between 1% and 3%, and another 12% between 3% and 5%.

So, nearly eight in ten households have spending rates of 5% or less. Now obviously, at the right we see individuals with very high withdrawal and spending rates; now our study looked at only a one-year snapshot, so, we don't know whether these households who are really spending at unsustainable rates, continue to do so, though we do have evidence as I showed earlier, that they're not completely liquidating their assets.

Now, the other aspects of withdrawals that I mentioned, is they can also be ad hoc; so, here I'm showing you withdrawals from various types of financial accounts, both systematic in blue, and ad hoc in green. So, clearly to the left, certain types of accounts we would expect to see high levels of regular withdrawals, so, from annuities, or defined contribution accounts, but as we start to move to the middle, we see that the percentage of ad hoc withdrawals is actually increasing.

So, if we look at mutual fund accounts, we see there's a considerable percentage of money that's taken on an ad hoc basis, and certainly increases as we move to more liquid accounts.

So, this is a good point to remember, the distinction between normative and positive behaviour; if you were to ask someone what retirees typically do, that person might reply well they set up regular withdrawal plans from their savings, usually something using a 4% rule.

And this description is not positive behaviour, but normative; while we certainly encourage investors to establish regular income streams, the real-world investors don't necessarily comply and ad hoc withdrawals are more common than we expect.

Alice Shepherd: Cynthia, I noticed on the previous slide that quite a high percentage have a withdrawal amount of 0% - why is that?

Cynthia Pagliaro: Yeah, you're correct; that is actually a very high percent, and actually your question, Alice, is a really good segway to the last of our five insights. So, the first four of our insights really highlighted this idea of unexpected precautionary savings in retirement; so, we're seeing that even in decumulation, investors are still saving for a rainy day.

And there's probably a number of reasons for people, why people do that, but increasingly research is pointing to the fact that people are trying to prepare for unpredictable costs of long-term care in retirement.

So, our fifth insight really speaks to this point that long-term care is emerging as the central point to the decumulation phase of life. So, now we all know about longevity and aging – everybody is living longer. In fact, if you're a little girl born today in a rich country, you'll probably live to be a hundred years old, but in addition to that, we also know that more affluent individuals are living longer.

So, on the left-hand side of this graph, it's showing life expectancy at age 50 by income quintile for Americans born in 1930; so, you can see for those with the lowest lifetime incomes, their life expectancy after the age of fifty was about 26.6 years, and for the highest paid, it was about 31.7 years; that's a spread of about five years.

Now, if you look at the same data for men born in 1960, life expectancy at the bottom fifth, has slipped a bit, to 26.1 years, but yet among the highest earners, life expectancy has sped ahead now to nearly 39 years.

So, that's a difference of 13 years; so, the key point here is not everyone is living longer, but many people are living longer, and certainly the higher your lifetime income and your socio-economic status, the greater the gains in life expectancy. And each increase in life expectancy is a good thing; I mean we're all happy about that, but it does come with some very important risks, and one of those risks, is cognitive decline and dementia.

So, here I'm showing you two charts, data from both North America and the United Kingdom. So, on the left-hand side, let's look at the prevalence of dementia in North America; and what's interesting to point out is the risk of dementia actually doubles every five years.

So, between the ages of 60 and 64, the risk is less than 1%, but from 65 to 69, it rises to nearly 2%, and five years later, almost to 4%. So, by the age of 85, the risk of dementia is actually three in ten, and this by the way is full-fledged clinical dementia – it doesn't even include the mild cognitive impairment numbers that will actually sit on top of this.

Now the statistics are similar in the UK, although this is displayed in a different format; what's important, though, to think about and remember here, is incidence of dementia increases with age, and we also know that people are living longer; so, the combination of those two factors is really creating this surge in the number, or incidence of dementia that we're going to observe in the future.

So, if you take this risk of life expectancy increasing, plus this rising incidence of dementia with age, you're really confronted in the end with the risk of long-term care, or in the UK, it's referred to as social care, but the great dilemma for many investors, is this incidence, or this long-term care risk, is highly asymmetric; many will not need any care at all, but a small but significant portion of people will.

 So, what I'm showing here, is not incidence of nursing-home care, but the incidence of needing any help with care in retirement. Now, there are five common, what we call activities of daily living, things like bathing, dressing and eating, where people may require care, and the care is not necessarily the result of cognitive decline, but could also be driven by other medical conditions.

So, let's look at the risk of someone needing help, particularly a 65-year-old otherwise healthy female. So, 40% of these females will require no assistance through their lives, and that's a pretty significant number, yet at the other extreme, we see more than a quarter will require five or more years of help, and another 2% to 4% will actually require help too; and here help is again the help with those activities of daily living, and this help can come in the form of family support or visiting nurses, social service, and finally institutions like assisted living, or nursing homes.

So, to sum it all up, it's really not a flip of the coin; some people are going to need a lot of help, but a small but significant portion will need a lot of help, excuse me, and some won't need any at all.

Now the risk of this really becomes salient to older households as they themselves confront health issues, or they see spouses, friends, or even aging parents need long-term care.

So, in the US, a very limited amount of long-term care is actually covered by Medicare, which is our Federal Health Insurance programme for elderly people. And typically, they pay for what's considered rehabilitative care; so, think about if someone had a broken hip and they needed therapy, but custodial care, and that is help with daily living.

First, people would have to pay out of their private savings, and it's exceedingly expensive in the US; typically, it would cost about a 150,000 US dollars a year, and then only after personal assets are spent down, are people able to qualify for Medicaid, which is another Federal programme that really provides assistance to lower-income, disabled, and elderly people.

And research by one of my Vanguard colleagues suggests that it's this aversion to running out of money and going on public assistance like Medicaid that is really motivating this intense precautionary savings that we've seen among retirees, and this point is confirmed by a number of other recent studies as well.

So, retirees aren't just saving because they're worried in general about the future, or inflation, or market shocks, they're actually saving out of a desire to have resources for better private care later in life; so, we now recognise long-term care risk as a central theme in the decumulation phase

So, I realise we've only had a short time together; so, this has been a really high-level overview of that decumulation phase, again, with the focus not on those normative standards that we set for retirees, but on actual real-world behaviour.

So, I hope you will agree first and foremost that the decumulation phase is much more complex than we previously imagined, very nuanced as well, so, it's really time for us to abandon our simple mental models of what we think is happening at this stage of life.

The second observation I'm going to make, is that planning for retirement is emerging not necessarily as just a wealth problem, but as a health and wealth problem. And here what I mean, is not necessarily planning for the cost of healthcare, but actually using your health state to help plan; so, considering your health status, medical risk, understanding your risks of disease, of dementia.

Certainly, if you're healthy and independent, you're going to face one set of costs, but if you're not, you're going to face a much higher set of costs. And today, most health planning actually takes place in a doctor's office, and over time that issue, those issues are actually going to start drifting into a financial planner's office.

So, my third point would be that the future of advice is really going to point to a closer integration of these life and health events, into the standard investment and wealth-planning events. So, we can expect longevity estimates become more personalised; we're going to have better measure of those long-term care risks based on health information.

We would like to know, and people are going to contract a particular disease because it's going to impact life expectancy, and affect how long we plan, and also our risk of long-term care exposure, which is going to affect how we're going to spend.

So, in the world of advice, classic financial planning questions are gradually being automated; so, as a result, the world of financial advice is really being transformed, and it's not just about money management, but on the central health and wealth trade-offs that our clients confront every day in their retirement lives.

So, I thank you very much for your attention, and hope you enjoyed this presentation.

Alice Shepherd: Thank you so much, Cynthia – that was great; so, we've come to the end of the presentation part of the webinar, but I think we may now have some questions from the audience. One question is do we think that these models will change as younger people retire with more DC schemes?

Cynthia Pagliaro: So, yes, yes, absolutely; I think the models will change; again, as I say, much of the data that I shared with you today is based on retirees today, but as we see the composition of wealth change, we will most certainly see these pictures change as well.

Alice Shepherd: Great; and I know you've touched on this, but how might an adviser use these insights when working with clients?

Cynthia Pagliaro: So, as I mentioned, I think the integration of these health issues is certainly going to be something that an adviser's going to need to think about in the future, but I think a key takeaway from this presentation today first and foremost, is to recognise that retirees don't necessarily behave the way that we expect them to.

So, I think those conversations have to be more detailed; I mean, people are not going to walk in and say I just want to create a regular income stream, you know, people are going to want to save money for a rainy day; people are going to want to consider their long-term care risks, and they're going to want to continue to save; so, an adviser's going to need to address all of those needs.

Alice Shepherd: Thank you so much; so, we have heard here that the decumulation phase differs from expectations in some significant ways; retirees don't spend as much as we think they do, and they continue to save even after retirement with long-term care costs in mind.

It's been a fascinating presentation, thank you so much; if you're interested in further research, Cynthia and her group recently produced How the UK Saves, a piece of work that was done in partnership with Nest Inside to examine the investment behaviour of UK pension savers since the introduction of automatic enrolment, the largest study into retirement savings ever undertaken in the UK. If you'd like to find it, you can find it on the UK Adviser website, or please get in touch with your Vanguard representative.

If you'd like to sign up for one of our CPD accredited workshops, please do get in touch with your Vanguard representative either directly, or by calling 0800 917 5508, or write to, and use the same contact details to claim your CPD credits for this webinar.

We'll be back next month, but this time on the fourth Tuesday of the month at 2 PM, Tuesday 23rd of October; in October, we'll be speaking to Kunal Mehta about the right way to invest in credit; the reason we're moving the webinar, is because of our symposium, which takes place on the 16th of October in Leeds, and the 17th of October in London. If you haven't yet signed up for the symposium, don't miss out; the theme is the Advantage of Advice, and you can find out more and register by going to

That's us for today; thank you so much, Cynthia, and thank you to all our listeners.

Investment risk information

The value of investments, and the income from them, may fall or rise and investors may get back less than they invested.

Important information

This document is directed at professional investors and should not be distributed to, or relied upon by, retail investors. It is designed for use by, and is directed only at, persons resident in the UK.

The opinions expressed in this material are those of the individual speakers and may not be representative of Vanguard Asset Management, Limited.

This material was produced by Vanguard Asset Management, Limited. It is for educational purposes only and is not a recommendation or solicitation to buy or sell investments.

Issued by Vanguard Asset Management, Limited which is authorised and regulated in the UK by the Financial Conduct Authority.

© 2018 Vanguard Asset Management, Limited. All rights reserved


 Remove  Save