Life After Full-time Work Blog

Learn about preparing for life after full-time work through posts from Don's upcoming book.

#85 A Case Study On The Investment Glide Path In Decumulation

Experts make assumptions about our attitude to financial risk in our retirement years. Here’s a case study that suggests that psychology plays a considerable role, regardless of finances.


Our most recent theme has been decumulation, and the investment risks associated with it. A big issue on which experts differ is to decide whether an increasing, level or decreasing equity exposure should apply in Life Two. Another aspect of Life Two we’ve touched on is the “buckets” approach to segregating the pension pot during decumulation, one bucket containing assets focused on safely generating the cash flows required in the early years and the other bucket focused on long-term investment growth.

In this post I’ll describe the essential elements of a case study that touches on both of those aspects. Warning: This is definitely advanced subject matter.


Let’s start with some background on the couple involved.

They think of themselves as being in the Lifestyle Zone (see Post #51 – Their pension pot isn’t large enough to buy an immediate inflation-linked annuity that will provide for their desired lifestyle for as long as they both shall live (as the saying goes). Or maybe it is. If they include their home, then yes it is. But they want to leave that to their children, and meanwhile it gives them a cushion. It would be hugely disruptive psychologically if they had to sell it. So they ignore it in their assessment, knowing it’s their Plan B; meanwhile they’re focused on their liquid assets only.

They understand clearly that they have three different goals: safety, growth and longevity insurance. There are no longevity insurance policies available in their country, so to deal with it they have to take the expensive route of self-insurance, that is, providing for possible survival to old age. In their case, they have ascertained that, for couples of their age, there’s a 10% chance that at least one will still be alive after 30 years. (They are both close to 70 years old.) If they approach that time, they’ll make adjustments [I note that I’ve never discussed this aspect – will do, some time!]; meanwhile, that’s the period over which they’re planning their drawdowns.

To provide for that, and in the hope of increasing their sustainable drawdown, they know they need to take some investment risk, and therefore there are no guarantees.


Behavioral economists sometimes jokingly refer to two kinds of people: Econs (the non-existent theoretical folk who are always totally rational about their behavior) and Humans (the rest of us, for whom emotion invariably plays a part). The couple in our case study are trying to anticipate (as Econs) what their psychological (Human) reactions will be. There’s no guarantee that they’ll actually react as they hope to, when they discover the fact that risk (as anticipated by Econs) has a friend called pain (as experienced by Humans). (That’s one of the reasons I’ve discussed the topic of happiness and psychology on this website, to help couples such as them.)

Before they retired they were advised by experts, who assumed that their risk tolerance would be based on asset volatility and that it would remain constant over time, in absolute terms (meaning that the dollar amount of acceptable loss in any year stays constant, as they age). The experts therefore recommended an increasing equity exposure over time, as the size of the pension pot diminishes. The couple’s actual experience of retirement has caused them to reject those notions.

It’s now clear to them that what affects them is not asset volatility but spending volatility. Why is that?

It’s because, like most retirees, they create and maintain a pool, a reservoir, of liquid assets, like cash or a bank current account, from which they fund their current spending needs (as discussed in Post #71 – This holds more than they need to spend in the near future. Periodically they draw down some of their pension pot and put the proceeds in the reservoir. And then they take money out of the reservoir for spending.

In this way, asset volatility doesn’t have an immediate impact on spending. Forced volatility in their spending is something they fear. Asset volatility causes fluctuations in the drawdown amounts; but the reservoir gives them a cushion against reducing spending. Even if the drawdown is somewhat volatile, spending itself doesn’t have to be nearly so volatile.

And therefore there’s no need for very high risk aversion with its consequent very low equity exposure at the start of their retirement, that would otherwise feel natural.


There’s another reason why they don’t want an increasing equity exposure over time.

They think that their psychological attitude towards risk will change over time, as their desired lifestyle settles down, from the go-go early years to the slow-go settled-down sequel that typifies many retirements. They therefore want as much safety as possible in their investments by the time the older one’s age reaches – oh, pick a number – let’s say 80 (so, in roughly 10 years’ time). It doesn’t matter if, financially, they can afford to take a lot more risk. In addition, they don’t want their adult children to worry about them or their finances, and the children will rest easier if they know that their ageing parents have looked after themselves. That suggests a target of 100% in safety-oriented investments, kicking in after the couple’s “autumn crescendo” is over, as Dr Laura Carstensen beautifully describes the early years of Life Two.


Now let’s think about their other goal: growth. Where will the desired growth come from? From a high initial equity exposure. But that’s dangerous! If they have no significant ceiling on their equity exposure in the early decumulation years, then they are particularly exposed to “sequence of returns risk,” meaning that a few years of bad equity returns in the early part of decumulation could condemn them to permanent regret and a permanently much-lower-than-desired lifestyle forever after.

So to enable them to focus on growth, they also want a ladder of safety, a tranche of safe investments initially from which they’ll make their drawdowns in the early retirement years. This is very important psychologically, even though they realise that it makes no financial difference to divide their pot conceptually into drawdown and growth segments: ultimately they have an overall asset allocation and that’s what determines the outcomes.

[I apologize for the varied terminology here. I’ve been using the couple’s terminology. The terminology more frequently used by financial professionals refers to “buckets,” the idea being that there are two buckets for drawdown purposes, a “safety bucket” that’s the same as the couple’s “ladder” or “liquidity reservoir,” and a “growth bucket” that is invested for long-term investment growth.]

They hope they can re-extend the ladder periodically, so that it’ll always be available as a safety measure. [Interpretation: they hope they can periodically sell a part of what’s in their growth bucket, and transfer it into their safety bucket, to preserve the length of time for which their safety bucket protects them from having to sell after a market downturn.]

Warning: dangerous! Yes, they realise that, and what they’re betting on is “mean reversion,” the notion that governments or central banks will manage to intervene and prevent a prolonged equity market downturn. If that doesn’t happen, then their risk will indeed turn to pain. They also realise that governments are notoriously inclined to increase their borrowing to help with the current crisis, and then not fully repay the borrowing when good times come back, so governments find they have less and less ability to deal with the next crisis.

The couple feel they are being as realistic as they can be, given the fact that they’re consciously taking risk. [Discussing the dangers is beyond the scope of this website.]


It turns out that the combined notions of a specific overall horizon (30 years, for this couple), a specific time by which all assets should decline to safety (10 years from now) and a specific length of ladder (they’ve chosen 5 years of spending) are enough to determine a customized glide path for the growth/safety exposures as well as an estimate of sustainable annual drawdown. The couple are pleasantly surprised and relieved that they don’t need to understand the intricacies of investing, and don’t have to respond to artificial questions such as how they might feel when their assets fall 10% in value (which most people have no clue about, and simply make up guessed answers).

They also like the fact that this approach doesn’t treat all couples of the same ages and assets identically. It doesn’t assign them identical asset allocations and drawdowns, ignoring how different their feelings about the future may be.

And they remind themselves of their cushions against forced spending volatility: the liquidity reservoir, the safety ladder, full safety (to the extent such a thing exists) in 10 years, the reduced need for income after the first partner passes away, and ultimately their home. They’re lucky.


I thought you might be interested in some numerical results for this couple. [This goes beyond the capabilities of the calculator on this website — sorry.]

My calculations use a 4% annual equity (growth) real return and a 0% safety real return.

For their combination of inputs (5 years in a safety bucket, 10 years overall before they’re fully in safe assets, and a further 20 years of spending after that), their indicated sustainable (real) drawdown per $1,000,000 of aggregate pension pot is $41,616.

Their initial allocation to the safety bucket is of course 5 years of this drawdown, or $208,080. The remaining $791,920 (almost 80% of their assets) go into the growth bucket.

They hope to transfer assets from the growth bucket to the safety bucket gradually for the first 5 years. After that they need to deplete the growth bucket rapidly, as they want it to be empty in a further 5 years.

Think of that as the base case. How much do the three parameters affect the indicated sustainable drawdown?

Here’s Variation #1. If, instead of 5/10/30, they use 5/15/30 (meaning they’ll be out of equities by age 85 instead of 80), the indicated sustainable drawdown rises to $45,447. The initial allocation to the safety bucket is $227,235 and to the growth bucket $772,765. There’s slightly less initially in growth, but the growth allocation lasts much longer, and that’s why the drawdown is higher than in the base case.

Here’s Variation #2. They use 7/15/30 (so, out of equities by age 85, as in Variation #1, but a greater desire for short-term safety). The indicated sustainable drawdown falls to $43,443. The initial allocation to the safety bucket is $304,101 and to the growth bucket $695,899. It’s the bigger safety bucket that causes the drop in the drawdown, relative to Variation #1.

Finally, Variation #3, in response to a friend who suggested they were being far too cautious. Instead of the 10% point for the longevity horizon, use the 25% point, which suggests roughly 25 years of spending rather than 30. Maintain the goal of being out of growth by age 85 (in 15 years), but place only 3 years (rather than 5) of drawdown in the safety bucket (suggesting, at least historically, a 67% chance of recovery rather than 75%). So, with inputs of 3/15/25, what are the results? An indicated sustainable drawdown of $55,597, dramatically higher than in Variation #1. This arises partly because they only want the pot to support 25 years of spending, and partly because only 3 years of this spending (making $166,791) goes into the safety bucket, and an enormous $833,209 goes into growth initially.



In this case study, a couple prefers a falling glide path, for purely psychological reasons. They’re willing to take their highest risk in the early retirement years (provided they have a ladder that enables them to climb to safety over the short term), and rely on safe investments as their lifestyles settle down and they become older.


I have written about retirement planning before and some of that material also relates to topics or issues that are being discussed here. Where relevant I draw on material from three sources: The Retirement Plan Solution (co-authored with Bob Collie and Matt Smith, published by John Wiley & Sons, Inc., 2009), my foreword to Someday Rich (by Timothy Noonan and Matt Smith, also published by Wiley, 2012), and my occasional column The Art of Investment in the FT Money supplement of The Financial Times, published in the UK. I am grateful to the other authors and to The Financial Times for permission to use the material here.

6 Responses to “#85 A Case Study On The Investment Glide Path In Decumulation”

  1. David Hartley says:

    Hi Don

    Australian Catholic Super in Sydney has systemised a version of the bucket approach that shares some of the features that you describe. There is a safe bucket from which income is drawn and this is periodically topped up by two types of transfers. The first is a quarterly transfer of dividends, rents and other income that is earned in the growth portfolio, providing that the safe bucket is not at its specified maximum. The second is periodic transfers of capital in those years in which the return on the growth portfolio exceeds its stated objective, again subject to a maximum amount in the safe bucket.

    An implication of such an approach is that returns in the growth portfolio that are derived from income as opposed to sales of assets to realise capital gains provide a more stable base for transfers. This has implications for the types of shares that are most suited for the growth portfolio compared to relative indifference between returns from income and capital gains in accumulation.

    There are also some option strategies that could be useful; for example the sale of out of the money call options can provide a source of income and “sale” proceeds on option exercise.

    • Don Ezra says:

      Thanks very much, David. I didn’t know that this sort of approach already exists. It’s the most advanced one I’ve heard of. And the notion of using equity dividends (and rents etc) takes it a step further, permitting a slightly higher growth exposure because it automatically generates part of the required cash flow (though, in my thinking, I don’t take it a step further by tilting towards high-dividend equities, because factor tilts start to introduce equity selection criteria). I’ve struggled with the criteria for periodic transfers of capital from the growth to the safe ticket and came up with pragmatic rules of thumb that sound slightly different from the ones you describe, but nonetheless, this is very advanced thinking. Thanks so much for bringing it to the attention of our readers. I wish there could be a general discussion of decumulation strategies.

      • David Hartley says:

        I understand what you are saying about an emphasis on dividend paying equities introducing a factor tilt and I would be concerned about this if my objective was to outperform a standard market index. It is a source of tracking error relative to that index that of course could lead to underperformance or outperformance.

        However in ths specific case an underlying purpose for investing in equities is to harvest dividends that contribute towards an income stream in retirement. Chosen carefully, these dividends per share could be expected to grow over time. Given this purpose for the equity portion of the portfolio, the utility of companies that have no earnings and/or no dividends is diminished. At the same time, outperformance of the standard index becomes a nice-to-have as opposed to a primary objective. So while the factor tilt introduced by dividend paying shares is a source of tracking error it is also a mechanism by which the share portfolio becomes more fit-for-purpose.

        • Don Ezra says:

          Thanks, David. I agree, the purpose is to combine long-term growth with another secure source of cash flow (as in my Post #75 on the reliability of the dividend stream), so as long as the tilt doesn’t greatly shift the characteristics of the rest of the equity portfolio, that’s as good as it gets.

  2. Jeremy Brereton says:

    Great blog Don, lots to think about and model. I would be really interested in your views about the type of assets that should make up the Safety Bucket. Bonds have moved closer to equities and although they have less volatility, they offer no return. Alternative assets within Investment Trusts probably fit in your Growth bucket, so is your Safety bucket just cash?

    • Don Ezra says:

      Thank you! Re the Safety Bucket: its goal is simply to produce the money you need, when you need it. So typically its focus is on guaranteed or predictable amounts at predictable times, rather than money streams that are uncertain. In turn, that tends to imply a sort of fixed (possibly/probably in real terms) income stream, in a laddered way. Cash would fit, though at the sacrifice of some interest return. Interest-bearing bank deposits or short-term bonds also fit. A bond index typically doesn’t fit, as the amounts of payouts typically don’t match the needs. Equity dividends tend to be a pretty reliable source, within limits (see Post #75). You mention that bonds have (virtually) no return these days. That’s a separate issue: the purpose of the Safety Bucket is reliability, not return.

Leave your question or comment