Spending Typically Declines as We Age

The most common assumption of retirement spending strategies is that real (inflation-adjusted) spending from savings will be flat throughout retirement, yet most studies of actual retiree household expenditures show that constant real spending is atypical. For most retirees, expenditures decline pretty consistently as we age.

Two of my favorite studies on this topic are David Blanchett's Estimating the True Cost of Retirement (2013, PDF) and Sudipto Banerjee's Expenditure Patterns of Older Americans, 2001-2009 (2012, PDF). The results of the studies are quite similar – not surprising since they used the same databases – but each provides unique information.

Blanchett christened his findings the "retirement spending smile", though be forewarned that if you Google "Blanchett smile", you will find a multitude of photos of Cate Blanchett's lovely face with poor David nowhere to be found. (It wasn't a terrible disappointment.)

Following is a chart of the "smile" from Blanchett (2013). (A quick note: you can double-click any chart in my posts to see a larger version. Also, while burnt orange text indicates a link to another website, yellow text indicates a mouse-over. Hover your mouse over the link for more information.)

There are three things I should note about the chart. First, the term "Experience" labeling the y-axis is an "auto-incorrect" for "Expenditures." Second, the smaller smile was added because of limited sample sizes for some tests. Pay more attention to the 30-year smile. My third point is a larger issue.

I suspect that some readers interpret the spending smile as showing that spending is high in early retirement, becomes lower until age 75 and then returns to nearly the level of early retirement near age 90, but this is not a graph of total annual spending. It is a graph of the annual real change in consumption for a typical retiree. In other words, it shows a decrease (and very rarely an increase) in spending at say, age 61 compared to age 60. It shows not the change of spending, but the rate of that change.

The rate of the decrease changes throughout retirement, but because these rates are nearly always negative (below the zero percent line on the y-axis in Blanchett's chart above), spending constantly decreases, but at different speeds. Banerjee shows the data in terms of total spending instead of the rate of annual change in spending and this point is more clear in his chart:

Reconstructing annual total expenditures from Blanchett's annual rate of change data for a retiree with a $100K annual spending target, we see a chart below that is similar to Banerjee's.

Mathematically speaking, the Banerjee curve is an annual spending function and the Blanchett smile curve is the derivative of the spending function. Banerjee shows the spending curve for a typical retiree while Blanchett shows the acceleration of that curve. Both show that expenditures generally decline with age, as have earlier studies. Blanchett additionally shows that expenditures drop more rapidly each year of early retirement and drop more slowly each year of late retirement, but both show that the amount of spending almost always declines.

Medical expenses late in life can increase expenditures significantly, but both studies appear to show that even when medical expenses do increase expenditures at older ages, they are lower than early retirement spending in real dollars.

The Banerjee chart and the Blanchett annual expenditure chart are not identical. Banerjee shows a steeper decline. Part of the reason for this may be, as Blanchett suggests, that he scrubbed the data to eliminate data points that seemed unreasonable, while Banerjee appears to have used the entire dataset.

Another reason is that Blanchett shows that rates of spending decline vary for undersavers and oversavers, while Banerjee provides a single rate of decline for all households.  Regardless, both studies find that typical retiree expenditures decline as we age. They do not remain constant in real dollars as spending strategies generally assume.

Why is this important? It should be obvious that when we try to estimate an amount of our savings that we can safely spend in the current year we must make some assumption about our future spending patterns. Spending strategies assume that our expenditures in real dollars will remain flat throughout retirement. If our actual spending will increase over time, we can safely spend less in the current year than these strategies predict, and the reverse is true if our expenditures will actually decline after we retire.

A 30-year retirement with level real spending of $100,000 a year would cost about $2.4M if we discount future expenses at 2%. Assuming Blanchett's findings for a retiree with a spending target of $100,000 a year, the same retirement would cost about $2.1M. Using the Banerjee 2012 finding that expenditures tend to decline about 2% annually, that retirement would cost only about $1.8M.

The following chart shows the expected annual spending and cost of an initial $100K annual retirement using all three assumptions:

Future spending is difficult to predict with any accuracy, but a spending strategy that assumes flat real spending throughout retirement, as nearly all do, will underspend early in retirement if the retiree's expenditures decline over time as Blanchett, Banerjee and several other researchers believe they commonly do. In these examples, Blanchett predicts a 12.5% less expensive retirement and Banerjee forecasts 25% less. From another perspective, that means a worker would need to save 12.5% or 25% less to fund retirement.

To quote Blanchett, "While many retirement income models use a fixed time period (e.g., 30 years) to estimate the duration of retirement, modeling the cost over the expected lifetime of the household, along with incorporating the actual spending curve, results in a required account balance at retirement that can be 20% less than the amount required using traditional models."

How does this impact our retirement plan? Clearly our future spending trend assumption has a significant impact on both how much we need to save and how much we can "safely" spend in the current year. Unfortunately, like assuming many other critical retirement unknowns such as future market returns and the length of our own retirement, choosing a future spending assumption is both critical and challenging.

In my next post, Retirement Spending Assumptions and Net Worth, I'll explore these two papers to see what they tell us about how we should choose.

What Academics Say

I've meant to write a post about academic papers for some time and I received an email recently that gives me a great opportunity.

“I generally think highly of [Dr. Redacted’s] work. . . But I’ve been puzzled by why his results sometimes weren’t more thorough – they are sometimes like abstract points . . . They can’t be applied without significantly more refinement and sometimes can lead to a wrong conclusion.”

Exactly.

But the problem, I believe, is not with the academic research – which is as it should be – but with our expectations and our over-reading of the conclusions. Sometimes, we just miss the point. Other times, we try too hard to convert theory into practical advice. And lastly, we're not the intended audience.

Let's look at those one point at a time.

It seems that many people unfamiliar with how academic research works are looking for a study that shows the one way to plan for retirement that is better than all previous recommendations and is the single best way to plan. As they say on police dramas, that ain't gonna happen.

Research is often an “experiment” that provides new information that we should consider and perhaps use to modify our previous beliefs to some extent. Or as someone said, “science is the process of continually improving your answer.” If you expect the next big academic paper to provide “the answer” that supersedes everything we used to think that we knew was correct, your expectations are too high. A good piece of research simply improves on the previous answer. It rarely replaces it.

So, is research sometimes "like abstract points"? Yes, often. One definition of abstract is “theoretical” and most academic research is theoretical, that is “concerned with or involving the theory of a subject or area of study rather than its practical application.”

Molding theory into practical application is the next challenge. Sustainable withdrawal rates research is, I think, an excellent example. Most studies use constant-dollar withdrawals, a spherical cow that is useful for research but problematic in practical application. No rational retiree is going to continue to spend the same amount every year once she observes that she is going broke. The technique is useful to understand the process of sequence of returns risk, but it is way too simple a model to predict outcomes for an actual retiree.

As a highly-respected retirement researcher once told me, “Bengen did an outstanding job showing that sequence risk exists, but then trying to identify a safe withdrawal rate was a fool's errand.”

Unfortunately, Money magazine and other popular media outlets translated this research directly into practical advice for several years before the 2007-2009 market crash convinced them that people really need to spend less money when they become less wealthy. The risk is not so much that you will deplete your savings with SWR as that you will irretrievably lose your lifestyle.

The comment above that research papers can sometimes lead to wrong conclusions is spot on but, again, this is more a problem with our conclusions than with the research. Here's an example. Wade Pfau and Michael Kitces caused a stir (a sign of good research) by writing a paper entitled, Reducing Retirement Risk with a Rising Equity Glide Path. Their research suggests that “rising equity glide paths from conservative starting points can achieve superior results.” One well-known retirement expert immediately tweeted something to the effect of, “Great idea - let's put all 95-year olds 100% in stocks.”

That wasn't the point of the research, but the effects were predictable. Clients asked if I thought they should be following a rising glide path and invest more in equities when they are older. I told them that they should decide that when they are older – I have no idea what their financial situation will be when they are 95.

Wade Pfau agreed, telling me that a custom asset allocation will always be better. He believed their findings simply suggest that if we knew nothing about a client other than his or her age, or if the retiree were only willing to do the absolute minimum of planning, a rising glide path would be the best bet.

He added that this is probably the best strategy for a target-date mutual fund that needs to serve a broad range of retirement investing needs. He also felt that the most interesting finding of the research is that you can sometimes “reduce both the probability of failure and the magnitude of failure for client portfolios” by increasing market risk. Not that everyone should adopt a rising glide path upon retirement and stick with it come what may, though this is what many concluded.

Sometimes, we just miss the point. David Blanchett wrote a paper that inspired the term “the Blanchett Smile” (download PDF). Many readers understood his findings to suggest that expenses begin to decline at retirement, bottom out about the mid-point, and then increase until death, in the shape of a smile. The “smile” Blanchett found, however, showed the acceleration of spending, not the amount of spending. The amount always trended downward, at about 2% or so per year for appropriately-spending retirees, throughout retirement. Only the acceleration of the downward trend (it's first derivative) eventually turned upward.

Misreading academic papers isn't a problem only among do-it-yourself retirees. After I wrote about this at Advisor Perspectives, a top-notch retirement adviser privately thanked me. He said that he had been present when Blanchett presented the paper and that he completely missed this point, as did, he believed, everyone sitting near him.

My last point is an important one: we are not the intended audience for academic papers. They are written primarily for other academics, who will review the papers before publication. Consequently, they tend to make modest claims that can be strongly supported by the evidence provided. Claims more like “increasing portfolio risk has the potential to actually reduce both the probability of failure and the magnitude of failure for client portfolios” rather than “everyone should follow a rising glide path.”

(Be even more cautious of "pseudo-academic papers" that are not peer-reviewed and thoroughly cited.)

I'm not suggesting that you stop reading academic papers on retirement finance. I am suggesting that you understand their intent and use considerable care in trying to apply their findings to your own personal situation. Treat them as another piece of evidence and weigh them accordingly.

(Interestingly, I've found that the retirement field is not the only one where academic research has a strong following among lay readers. I found a website where non-scientists express great interest in cosmic physics. I find that encouraging.)

While academic papers aren't intended for a broad audience, the authors of those papers often write books, blogs and columns for the press that are. Wade Pfau, for example, does an excellent job of explaining his research to a broader audience of advisers and even do-it-yourselfers at several places, including newsletters to which you can subscribe, the Advisor Perspectives website, and his own Retirement Researcher blog. William Bernstein has written many brilliant books and over the years has intentionally made them more and more accessible.

Dr. Moshe Milevsky is another excellent writer. While I don't recommend most people tackle his papers about lifetime probability of ruin and the reciprocal gamma distribution (it kept me up nights until my tenth reading), I just finished the second edition of Pensionize Your Nest Egg, co-written by Milevsky and certified financial planner, Alexandra Macqueen, and find it quite readable. (Macqueen clearly helps balance Milevsky's inner quant.) It provides a strong argument for the circumstances under which and to what extent one should employ annuities. Spoiler: if you have a lot of wealth or a little, they may be less effective than for those in between.

What I Learned in 2015: Old Guys Rule

My blog posts are largely a mechanism for me to think aloud about retirement finance and to receive helpful feedback from my readers. I’m learning, not teaching, and I learned a lot in 2015.

I published 44 posts on this blog over the past year averaging about 958 words per post, or a little more than 42,000 words. I also co-authored a paper on portfolio ruin with my son and daughter, which I expect to publish early in 2016, that adds another 4,000 words (many of which I didn’t know before we wrote the paper), plus a few posts at Advisor Perspectives. I don’t even want to estimate the number of lines of R code and Mathematica code I wrote this year and there were a few presentations at conferences. Let's just ballpark it at 50,000 words.

(A classic novel is typically 80 to 100 thousand words; Fahrenheit 451 is a little over 46,000.)

That’s a lot of words for a hobby, a lot of lattes and a lot of learning.

My audience is diverse and knowledgeable and teaches me a lot, often by simply asking the right questions. Nearly 86% of them log on from the US, 4% log on in the UK and 1.8% from Canada. The surprise, however, is that 3% of my readers log on from Ukraine and nearly 2% from Russia. I extend a heartfelt Дякую! to the former and Спасибо! to the latter.

(I really hope I got that right. A Russian-speaking friend confirmed one, but my only Ukrainian friend moved to South Carolina last summer.)

Here are a few of the important things I learned in 2015.

There is no dispute between Jeremy Siegel and Zvi Bodie about stocks becoming safer the longer you hold them. They don’t.

I also learned that people have a difficult time giving up beliefs about finance. I still have planners argue that stocks become safer. Spending only dividends is not a valuable retirement income strategy and risk of ruin is primarily useful only as a research tool. A bond ladder held to maturity and a bond fund are not the same thing. A ladder of TIPS bonds held to maturity is essentially cash.

My son’s potato casserole is outstanding, but should be baked in a disposable dish. (I wash the dishes at our house.)

I learned that time segmentation (“bucket”) strategies can’t be depended upon to avoid a bad sequence of returns. They might, but it isn’t a sure thing.

I received an award for a paper in Indianapolis this summer. I had not visited the city for decades and I learned that it is still incredibly flat. I also learned that a small replica of Rodin’s “The Thinker” in your carry-on looks like a bunny rabbit on the x-ray screen of a TSA employee. ("Are you sure it's not a bunny rabbit? It really looks like a bunny rabbit.")

Game theory can be a useful way to think about retirement strategies.

I learned that even decaffeinated coffee after 3 pm can impact my evening’s sleep. My wife insisted that I test this theory and, much to my chagrin, she was proven correct. This dramatically altered my afternoon writing strategy at Caffe Driade.

Retirement income systems may be chaotic and virtually impossible to predict except in equilibrium, so chaos theory is another useful way to view retirement finance.

I learned that co-authoring a research paper with a son you taught to play basketball and who taught you how to play Super Mario World, and a daughter you taught to fish and who danced with you at her wedding is one of the coolest things that you will ever do.

Even with a basic assumption that our portfolio will return 5% with a standard deviation of 12%, the range of reasonable likely outcomes is too broad to effectively choose among spending rates and asset allocations.

Retirement income models probably involve a lot more uncertainty than most people assume.

I learned that academic papers on retirement finance are often misinterpreted, not just by retirees, but by financial planners, as well. Examples include "retirement spending looks like a smile" (the rate of annual spending change looks like a smile, but actual spending typically declines throughout retirement) and “retirees should increase their asset allocation as they age” (it depends – a custom asset allocation plan is always preferable).

I took the GMAT 30 years ago to get my MBA and again this past year. Perhaps most fun of all, I learned in 2015 that I can still outperform 4 out of 5 young whippersnappers taking the exam.

Old guys rule.

Here’s to an educational 2016.

---

In my next post, I'll explain Why Retirees Go Broke.

Retirement Spending Assumptions and Net Worth

In my last post, Spending Typically Declines as We Age, I reviewed the results of research by David Blanchett (PDF) and Sudipto Banerjee (PDF) that shows expenditures in retirement typically decline as we age. Most retirement spending strategies assume, as I noted in that post, that future real spending will remain constant throughout retirement.

The amount that we can safely spend from retirement savings in the current year depends heavily on the assumptions we make about future spending trends. If our future spending needs will decline, spending rules that assume constant real spending will be unnecessarily conservative and, of course, if future spending will increase, those spending rules will recommend spending that may not be sustainable.

What assumptions should our retirement plan make about future spending? The two papers I reference offer some clues.

First, Banerjee reports that spending in retirement increased for only 16% of the households in the data he studied, while it declined for 66% of them. It is significantly more likely that your expenditures will decline as you age, but they might not, so our retirement plans should also consider worst case outcomes.

We could assume a worst case, that expenditures will increase perhaps 1% per year on average, but that would significantly increase the predicted cost of retirement. If we assume we will live 30 years or more, that our market returns will be quite conservative by historical standards and that our expenses will grow in retirement, we will quickly realize that hardly anyone could afford that retirement. Making lots of conservative assumptions doesn't make a very good plan.

Blanchett offers additional insights by segmenting the data based on the level of annual spending relative to net worth. He creates four categories of consumption: low spenders with high net worth, low spenders with low net worth, high spenders with high net worth and high spenders with low net worth. By figuring out into which group you best fit, you may be able to narrow the field of spending assumptions for your plan.

The dividing line for high and low spenders was $30,000 per year and the hurdle for high net worth was $400,000 in Blanchett's study. These are the median values for his data sample, not for the population of retirees. In other words, more than $30,000 of annual spending made households in the data sample high spenders relative to other households in the sample, but that amount wouldn't make you a high spender relative to all other retirees in the U.S. today. The breakpoints for the larger population of retirees would likely be much higher. Blanchett is showing that expenditures in retirement depend on the relationship between annual spending and net worth; he is not claiming that these are the dividing lines for all retirees.

Blanchett notes that two of these groups, Low Spending, Low Net Worth and High Spending, High Net Worth retirees, consume efficiently (green), while the other two groups consume too much (red) or too little (yellow).

Following is a diagram from Blanchett's paper plotting these four segments. Please note the very important point, as I explained in Spending Typically Declines as We Age, that these graphs show annual rate of change in spending and not annual spending, itself. With the exception of Low Spending, High Net Worth households (the red squares on Panel B) almost all of the annual changes in expenditures are negative, meaning spending declines throughout retirement for the other three groups. (Double-click the chart for a larger image.)

Notice that the following graphs of annual spending are quite different than Blanchett's graphs of annual spending change above. Because some readers have mistaken the Blanchett "smile" rate-of-annual-change graphs for annual expenditures graphs, I provide both in the examples below.

In fact, Blanchett's paper shows this in his Figure 7, though most of that paper addresses annual spending change and not annual real dollar spending. My graphs will show typical real dollar annual spending that is derived from the rate-of-change graph to its right. I place the spending function on the left because I believe that information will be more meaningful to most of my readers, and I switched Panels A and B in Blanchett's Figure 7 for consistency – the charts on the left always show annual spending.

Let me be very clear about this. The Blanchett smile curves, like Panel A above, show how quickly typical spending changes each year. The spending curves, like Panel B above, show how much spending changes in real dollars and, in most cases, spending goes steadily downward throughout retirement, like the curves in Panel B.

Now, let's look at Blanchett's four spending/net worth classifications.

Low Spending, Low Net Worth households likely spend a large portion of their budget on non-discretionary expenses with little opportunity for reducing expenses later in retirement. We usually give up some discretionary items later in retirement, like extensive travel and sports, and these households have fewer of those to eliminate, so expenditures don't decline a lot with age.

High Spending, High Net Worth households also consume efficiently and will likely see greater declines in spending than Low Spending, Low Net Worth households, because they will have more discretionary spending to eliminate as they age.

Low Spending, High Net Worth households appear to have the highest probability of increased expenditures throughout retirement, but they can afford it. They are underspending. A likely cause for such an increase in expenditures, in fact, is recognition over time that they have the resources to spend more.

This graph also demonstrates the key point that increasing expenditures don't necessarily mean that retirement is getting more expensive and decreasing expenditures don't mean it is getting less expensive. They mean that retirees are spending more or less. Expenditures, the subject of this analysis, are not the same as expenses. Sometimes expenditures change because retirees have to spend less and sometimes it is because they can spend more.

High Spending, Low Net Worth households also consume inefficiently and they are likely to recognize as they age that their level of spending is unsustainable. This realization will eventually lead to declining expenditures later in retirement, as can be seen in the following chart.

The next chart combines all four annual spending curves for comparison. The two curves in the middle result from efficient consumption. Inefficient consumption produces the two extremes.

Your personal ratio of annual spending to net worth should suggest whether your own spending is more likely to rise, fall, or remain somewhat constant throughout retirement.

To summarize this information, Banerjee tells us that two-thirds of retirees will experience declining expenditures as they age and only 16% will see increased spending. Blanchett tells us that spending will likely increase for Low Spending, High Net Worth households as they realize they are able to spend more as they age, as it will likely decrease for High Spending, Low Net Worth households as they realize they are running out of savings. Among households that consume efficiently, those with High Spending and High Net Worth are more likely to spend less later in retirement because they will have more discretionary expenses to "age out of" than will Low Spending, Low Net Worth households.

Some expenditures change for reasons that have little to do with how much annual spending a retiree targets or how wealthy they are. Some changes are the result of aging. Health care expenses tend to increase as we get older but we also become less active and other expenses decline. I recently read that international air travel declines among septuagenarians and domestic air travel declines among octogenarians. I suspect the sales of bungee-jumping and rock-climbing gear decline in those market segments, as well. I spend less on hair care expenses.

These two studies deal with typical retiree spending patterns and assume that expenditures will follow some trend, rising, constant or declining, throughout retirement. They don't, however, deal with the most likely scenario for an individual household, irregular net spending.

Both income and expenses in retirement are likely to vary significantly over time. Income will vary, for instance, as Social Security benefits ramp up for retired couples and more income needs to be withdrawn from savings early in retirement. There may also be large planned expenses later in retirement, like college for a child or grandchild. Net spending is the important consideration, the difference between annual income and annual expenses. Irregular net spending from savings might look like the red bars in the following chart:

These irregular net spending years, whether they are caused by changing income or changing expenses, must be considered when calculating a sustainable amount to spend in the current year. Like steadily rising or steadily declining expenditures, spending rules assume flat future spending and don't accommodate irregular net spending very well.

It is helpful to know that expenses typically decline in real terms throughout retirement, but yours may not. You need to plan for expected spending declines but be prepared for a worse case. Like portfolio returns, total expenditures in retirement are unpredictable.

So, most retirement income strategies assume constant spending throughout retirement and most retirement expenditure studies show that constant spending isn't the norm. What's a retiree supposed to do with that?

In my next post, Spending Rules That Fit the Patterns of Retirement, and Some That Don't, I'll explore spending strategies in light of future spending expectations.

The Non-Existent Law of Averages

There’s an ancient joke about a statistician who drowns wading across a river that is, on average, three feet deep. I have often found myself wading rivers to fly fish with water up to my knees, only to step into an unseen hole that drenched me to the neck. I know that my belongings might get dunked even if I stash them in my cap.

Funny stories, perhaps, but I find that hardly anyone understands even the most basic concepts of statistics and probability. That includes professionals who should understand at least the basics, like doctors and financial planners.

That lack of understanding includes the most basic statistic, the average, or mean. For example, you frequently hear the expression “law of averages”, though there isn’t one. Look it up. No such thing.

You also hear about average market returns. Those do exist, but they may not mean what you think.

In order to understand averages, you need to be able to differentiate long-term probabilities from one-time events. The difference is demonstrated by actuarial forecasts.

If we gather 1,000 people of the same gender and health and age in a room, actuaries can tell you how many of those people will still be alive in 10 years, 20 years, 30 years, or longer with amazing accuracy — but they can’t tell you which individuals in the room will still be alive.

This is useful information for an insurance company. Long-term averages are accurate enough for them to make bets that will result in profits from selling life insurance policies. They will guess wrong sometimes and right sometimes but they get to make lots of bets and on balance they will come out ahead.

However, if you are an individual in that room, probabilities can’t tell you how long you will live. The insurance company gets many bets, but you get just one. Your life is a one-time event. 

50% of the people in that room will live to their life expectancy or longer, but you either will or you won't. You can't have 50% of each outcome.

Perhaps a fourth of the people in the room will actually live past age 90, but no one can know whether they are in that group or not. Consequently, everyone in the room should plan for the possibility of living past 90 because if they bet on living a shorter life, they may run out of money if they're wrong.

Actuaries might tell us that the life expectancy of everyone in the room is age 78. That's the age at which they predict half the rooms occupants will still be alive and half will be dead.

If everyone in the room planned to pay for a retirement only until their average life expectancy, half of the people in the room would die broke. Averages turn out to be a very poor way to develop your retirement plan. Planning for the worst case works better.

Long-term stock market returns are another example. The S&P 500 grew at an annual rate of 9% from 1928 to 2008, but people who began investing in 1928 only earned about 7.9% over the next 30 years, while people who began investing in 1942 averaged 13.6% a year.

That’s a huge difference over 30 years. At 7.9%, $1 grows to $13,579 in 30 years. At 13.6%, a dollar grows to $40,280.

So, yes, had you invested a dollar in 1928 and held onto that investment for eighty years (and achieved market returns), you would have earned an average 9% a year — and you would be very old. In reality, you will only have about 30 years to invest for retirement. 

Depending on when your 30 years of investing ended (which depends largely on when you were born), however, you might have earned as little as 7.9% or as much as 13.6%. You probably wouldn't have received the actual average of 9%.

Your investing life is a one-time event. The market returns you receive may be much more or much less than the long term average.

Investors have been convinced by the marketing of brokerage firms that if you just stick with the market long enough, you will end up with an 8% to 10% return. Just keep the faith in the bad times, stay invested, and you will eventually get back to that 8%.

But that simply isn’t true. Your investment career is a one-time event and your investing results may be very different than the returns of a hypothetical person who might have been investing in the stock market for the past 80 years.

Be realistic. You could earn a lot in the stock market and you could lose a lot, or somewhere in between. Hanging on in bad markets is no guarantee that you will eventually be just fine.

So, what does this mean (pun intended) for retirement planning? It means that you have to plan for those holes in the river I sometimes step into. Have a plan that keeps your head above water if you live to 90, or if the stock market gods don’t smile on you.

You have to plan for the worst case, not the average.

In my next post, I'll explain why you probably won't earn "market average" returns.

But What If You Do?

People nearing retirement age often tell me that they plan to claim their Social Security benefits at the earliest possible age of 62. More often than not they’ll explain that they would have to live 15 years or so, to age 77 or 78 depending on individual circumstances, to come out ahead. Someone told them that is their "break-even age" for Social Security retirement benefits.

They go on to explain that they don’t expect to live that long, though for the life of me I can’t figure out how a reasonably healthy person would know how long he or she will live, or why they would be willing to bet a lot of money on it.

At a baseball game last summer, one of my buddies looked around and asked, "What's the break-even age for Social Security benefits?"

A partner from a CPA firm answered, "Around 78-80 depending on your specific situation."

"Huh!" he grunted, turning back to the game. "I'm not gonna live to 80."

So I asked him, “But what if you do?”

Consider Fred and Ethel, a married couple in their 60’s. Fred would be eligible for Social Security benefits of $1,735 monthly if he received them at age 62, or $2,300 if he waited until age 66. If he could postpone receiving benefits until age 70, he would receive $3,030 a month for the rest of his life, a whopping 75% more than if he claimed at the earliest age.

Fred plans to claim at age 62 because he thinks he won’t live a long life. He’ll come out ahead in lifetime benefits if he claims early and doesn’t live past age 78 or so.

But what if he does?

The following chart shows Fred's total lifetime payments from Social Security retirement benefits if he lives to various ages and claims either at age 62, 66 or 70. For example, if he claims at age 62 and dies at age 76, he will receive payments totaling about $312,000 over 14 years. Should he claim at age 66 and die at 76, he would receive about $304,000.

To simplify the chart, let's divide it into three sections, or three possible future scenarios for Fred.

First, look at the middle section when Fred is around his break-even age of 78. If he dies within a few years either side of age 78, all three claiming ages work out about the same for him, so his benefit-claiming age decision wouldn't have a big impact either way.

Next, let's look at the early retirement years preceding the middle section. In these years, Fred will either have a short-term and urgent need for his benefits because he has limited income and savings, or he will have the option to transfer some early retirement standard of living to late retirement. He would do this by postponing his claiming age. That would give him less income in early retirement and more in late retirement, should he live that long.

Regardless, if Fred doesn't live past early retirement he is clearly better off claiming early, but his spouse may not be. Ethel's survivors benefits will be 100% of Fred's retirement benefit at death.¹ If she survives Fred, she will live the rest of her life with the lower benefit that Fred chose for her by claiming his own retirement benefits early.

If Fred lives just a few years after retiring he will receive more total benefits if he claims early, but the opposite is true if he lives a long life. So, which should he bet on?

A short retirement is relatively less expensive than a long one, and often a lot less expensive. A short retirement requires less savings. It may be possible to return to work if you run low on funds early in retirement.

A long retirement is far worse from a financial perspective. It costs a lot more. You're too old to go back to work. You've probably depleted most or all of your savings. You'll need some of that early retirement standard of living more in your old age than you needed it in early retirement.

Running out of money in late retirement is the worst-case scenario and a good financial planner would insist on taking the worst-case scenario off the table.

So, Fred believes that he and Ethel won't live a long life, but what if they do?

They would be much worse off claiming benefits early and living a long time than claiming later and living an average lifetime or less.
 
Break-even ages would be interesting if they showed that you would have to live to 105 to break even (everyone would claim early), or if you only had to live to 64 (everyone would wait), but they don’t. They show that you must live to an age that you might reasonably expect to attain and that you might not. 

Consequently, I find break-even ages relatively useless.

How about investing your retirement savings in the stock market?

The “4% Rule” strategy for investing retirement savings in stocks says we have a 95% chance of not outliving our savings. That’s great if you don’t end up in the 5% group of retirees who go broke, but what if you do?

Those web-based retirement calculators show you what will probably happen to most people. They don't show what might happen to you.

Anticipating the worst-case financial outcomes and planning to avoid them, even at some cost of our retirement standard of living, is what we call risk management. Break-even analysis isn’t risk management, nor is betting on future stock market returns. It’s betting that everything will work out fine. It’s playing the odds that we (and our spouses) won’t live longer than an average lifetime, that inflation won’t run away, and that the stock market gods will smile on us.

Perhaps you and your spouse won't live longer than average lives. Maybe inflation will remain below 3%. Maybe you won't experience a long-term bear market after you retire.

But what if you do?

---

¹ Assuming Fred has begun receiving those benefits, otherwise it will equal 100% of his retirement benefit at full retirement age.

Unraveling Retirement Strategies: Floor-and-Upside (An Update)

When a reader recently quoted me from a four-year-old post, I realized I needed to update a few of them. I've learned some things in the past few years and, like many of us who research retirement finance, my thinking has "evolved."

(My wife would undoubtedly question both claims, but I refer specifically to retirement finance matters at present.)

Sometimes reading an old post is like seeing a photo of yourself from 1975 with mutton chop sideburns a la Neil Young and saying out loud, "What was I thinking?"

One such post is Unraveling Retirement Strategies: Floor-and-Upside[1], from February 6, 2014. I've made a few changes to reflect my updated perspective and to incorporate some of the excellent reader comments that post attracted. Specifically, I've become more enamored with annuitites and less with TIPS ladders.

The floor-and-upside strategy for financing retirement is sometimes referred to as “safety first” and derives from The Theory of Life-Cycle Saving and Investing[1].

The basic idea behind floor-and-upside is that a retiree devotes some of her retirement funding assets to building a lifetime stream of income and the remainder to an investment portfolio to provide liquidity and the possibility of increasing wealth over time.

It's important to note that growth of the upside portfolio isn't guaranteed and, in fact, the "upside" investment portfolio may shrink over time or even be depleted prematurely (an "upside portfolio" also has downside). The "floor" is a safety net that will provide income should the upside portfolio fail. The rest of your retirement plan should ensure that having to live off the floor income alone is very unlikely.

Assets suitable for constructing the floor portfolio include Social Security retirement benefits, life-contingent annuities, and pensions. A TIPS bond ladder is not guaranteed to last a lifetime but it is conceivable that one could be built for 35 years, for instance, that would be highly likely to outlast a joint lifetime. Whether or not the cost of the ladder would be acceptable is another matter.

You could build a really simple floor-and-upside strategy by using part of your retirement savings to buy a life annuity to guarantee a certain amount of income for as long as you live and then investing whatever is left of your savings in an S&P 500 index fund.

In fact, since most Americans are eligible for Social Security retirement benefits, most Americans have a "floor." Those who also have some savings to invest in retirement, therefore, have a floor-and-upside strategy. Social Security benefits alone, however, may not provide as much floor as desired.

Deciding how much floor income you should build into your plan is sometimes easy. I've had clients say, "I don't care about upside potential, I'm not as impressed with my husband's investing skills as he is. I just want a check in the same amount monthly for as long as I live." This is a person who wants nothing but floor.

I also have had clients and readers say, "I believe in my investing skills and that the market will always eventually go up throughout my lifetime." Since some suggest that they should invest their Social Security benefits, too, I assume there is a group of retirees who want no floor at all.

In between these extremes, the decision can be more difficult. The best I can recommend is that you imagine that you are 85 and your upside portfolio balance just went to zero, a victim of sequence of returns risk. What is the least amount of income you could have remaining that would not make your life an economic misery? This is the floor level you wish to have.

The next question, of course, is whether you can afford that much floor and your level of wealth may or may not dictate that you choose a lower level. Interest rates are historically low at present so floors are expensive.

(A reader once commented that anyone who can afford a floor doesn't need one. Not true. Everyone can afford some level of floor even if it consists only of Social Security benefits. No one suggests that your floor cover 100% of what you hope to spend in retirement. That would indeed be expensive. Floor income should cover food, housing, clothing, and the like, but not the annual European vacations you planned before your upside portfolio confirmed your wife's suspicions about your investing skills.)

Here's an example. Let's say you want to spend $60,000 annually in retirement and your household expects $30,000 from Social Security retirement benefits. Non-discretionary spending totals $48,000 of the $60,000 total. (The floor doesn't have to be your non-discretionary expenses, it can be whatever makes you comfortable, but that's a reasonable starting point.) You have saved a million dollars for retirement.

You need another $18,000 of longevity-protected income. Wade Pfau's Dashboard[3] (or a quick online annuity quote from someone like myabaris.com) tells us that a single-premium income annuity (SPIA) for a 65-year old couple today will generate about a 5.63% payout at today's rates. Divide 18,000 by .056 and you can estimate that you need to annuitize about $320,000 of your savings to generate the safe floor you desire.

Invest the remaining $680,000 in stocks and bonds (I recommend index funds) and you have a floor-and-upside plan with $48,000 of longevity-protected income (from Social Security benefits and the SPIA) in the unlikely event that you prematurely deplete your savings portfolio.

---

An update on Floor-and-Upside Strategies.

[Tweet this]

---

There are a few critical concepts to consider at this point. First, the income from both the floor and upside portfolios assumes normal expenses. There will always be a risk of unpredictable, catastrophic expenses — a lawsuit, medical expenses, a child or grandchild who needs your financial support — that can blow up your retirement plan. Insurance may help and a reserve fund might, too, but there is always the risk that neither will be enough. We're planning only for the expenses that we can predict to some extent.

The second important concept is that, if we depend on a portfolio or a TIPS bond ladder for income, their liquidity is somewhat illusory. Both are sometimes referred to as "fettered" assets because we depend on them for future income. We often can't really spend them on something else. Spend from either of these sources when they're intended to provide future income and we give up all or part of that future income.

An annuity will become worthless at death unless you purchase (often ill-advised) options to prevent that. A TIPS bond ladder or an investment portfolio may have a remaining balance at death but that residual balance will be available to our estate, not to us while we are living.

It is true that an annuity provides less flexibility (more liquidity) than a TIPS ladder but the flexibility of the ladder is limited. A large medical expense can't be paid immediately from an annuity, it's true, but paying it from a TIPS ladder isn't much better. You're paying the expense from the source of funding for future years of retirement. An annuity will always provide more income than a bond ladder so you might be better off using the higher income to pay the large expense over time.

Bottom line, if you suffer a huge uninsured expense in retirement, you have a serious problem with any strategy.

A third important concept is that in many scenarios annuitizing part of your savings will result in a larger estate. This is counter-intuitive. In the above example, many retirees would think, "I just took $320,000 out of any future estate value because the annuity will be worthless when I die."

But, the annuity enables the upside portfolio to be invested more aggressively and lowers the retiree's sequence of return risk by reducing the periodic amount spent from savings. This will often lead to a larger estate than a portfolio-spending strategy alone.

Floor-and-upside is a compromise between using all our savings to buy annuities and investing it all in the stock market. We buy enough annuities to provide a safety net and invest the rest.

Annuities will provide for maximum lifetime consumption but have no value at death. Depending entirely on the stock market will provide more consumption and possibly a residual balance at death if the stock market gods favor us and less spending and a smaller bequest if they don't.

How does this fit into the theory of life-cycle saving and investing? That economic theory suggests that households should prefer reducing their consumption a bit in good times if it will improve consumption a bit in bad times ("consumption-smoothing"). We buy health insurance when we are healthy in good economic times, though it may have no immediate benefit, so we will be able to consume more at times when we are unhealthy and have large medical expenses.

Floor-and-upside gives up some of the stock market gains in the good outcomes to make sure we have a bit more income in worse scenarios with poor market returns.

The floor-and-upside strategy will combine the two such that they provide a safety-net level of lifetime income and an opportunity for more consumption if those stock market gods smile down on us.

If they find us annoying, we'll still have the safety net.

In my next post, I'll describe the Constant-Dollar Spending Strategy (the "4% Rule").

---

REFERRALS

From time to time, I am asked for retirement planner references. In the past week, I received such a request as a blog comment. I would prefer you make the request by emailing me at JDCPlanning@gmail.com.

Also, there are relatively few retirement planners that I know and would trust with my own family and they are scattered around the country. All of them work remotely, primarily via email and phone as I used to, but if you want a planner you can meet in person, it is very unlikely that I will know one near you.

Thanks.

---

REFERENCES

[1] The Retirement Café: Unraveling Retirement Strategies: Floor-and-Upside.



[2] The Theory of Life-Cycle Saving and Investing, Federal Reserve Bank of Boston.



[3] Dashboard, RetirementResearcher.com.