Using Personalized Lifespan Estimation For Retirement

Shownotes:

• Jeremy Ko: LinkedIn
• ShoreUp Retirement Solutions
• "How to Estimate "The End" Of Retirement", by David Blanchett
• American Academy of Actuaries Longevity Illustrator
• Northwestern Mutual Lifespan Calculator
• University of Connecticut Healthy Life Expectancy Calculator
• Navigating Sensitive Topics With Clients: 3 Tools To Get Them To Open Up About Planning Hurdles, by Meghaan Lurtz

Transcript:

Adam: Hello, and welcome back to the Financial Advisor Technician Podcast. I'm your host, Adam Van Deusen. On today's episode, we're going to discuss one of the major challenges when it comes to retirement planning, which is that clients have unknown lifespans. Because if an advisor could know exactly when a client was going to pass away, they could fine-tune the financial plan to precisely meet the client's goals.

However, in the real world where longevity is uncertain, it can make sense to take a conservative approach when it comes to estimating a particular client's lifespan, as clients and advisors alike are likely to consider the consequences of underestimating a client's lifespan, which could result in the premature exhaustion of their assets, to be more severe than those from overestimating the lifespan, which could result in less lifetime spending than the client might have enjoyed otherwise. Which is why many financial planning software programs use a default age at death for clients that is beyond what actuarial tables might suggest.

That said, given the many planning applications of lifespan assumptions from determining how much a client might need to save for retirement to optimal Social Security claiming strategies, having a better estimate of a particular client's lifespan could lead to better planning outcomes.

So to help us dig deeper into how advisors might approach the challenging issue of estimating a particular client's lifespan, I'm joined today by Jeremy Ko, an economist and the founder of Shore Up Retirement Solutions, to discuss different ways to estimate a particular client's lifespan, how generating personalized lifespan assumptions might be easier and less expensive than many advisors might expect, and how advisors can tactfully raise this sensitive issue with clients.

So welcome, Jeremy, and thanks for joining us on the Financial Advisor Technician Podcast.

Jeremy: Thanks so much, Adam. Really appreciate the opportunity to talk to your audience today.

Standard Approaches To Estimating Clients' Lifespans [2:03]

Adam: To start, perhaps you could provide a rundown of some of the perhaps, you might say, less-personalized approaches to lifespan assumptions, and where they might fall short for advisors and their clients.

Jeremy: Yeah. So, I think that there's three common practices that each present their own problems. And I want to start out by saying we're not trying to castigate, condemn the whole profession. The client intake process is really long and complicated as it is, and I think the last thing that an advisor or planner wants to do is to grab more stuff from the client, as it relates to actuarial lifespan. But the three main approaches that exist, I'd say, are first of all, a one-size-fits-all or a monolithic approach, where the planner or the advisor will assume that the client is going to live to some age, like 90, 95, 100, just to be conservative, without any consideration as to that client's particular characteristics on a personal level. So, things like health attributes, their gender, etc. There was a study done by David Blanchette which found that, at least in Canadian financial plans, and I believe the practice is pretty much equally common here in the U.S., that about 70% of them took a assumed lifespan of 90, and about 20% of them took an assumed lifespan of about 95, with very little variation, even by gender. So, that's one approach. And as you pointed out, Adam, that can present its own problems, even if you're trying to be conservative on the upper end.

And then there's a fairly common approach where planners, advisors will ask their client, "Well, how long do you think you're going to live?" that, on its surface, seems like a reasonable approach. But the problem with that is that there's systematic biases in reporting these lifespan estimates that people exhibit in data. So, to be specific about this, there's something called flatness bias, where people on the younger end of the retiree scale, so, these are people in their 60s and 70s, will typically underestimate the length of time that they're going to live. Whereas people in their 80s and 90s, on the older end of the retiree scale, that they will overestimate the amount of time that they're going to live. So, people kind of fudge for the same number, whether they're really young or really old in on the retiree scale. And so that can present its own problems. And then finally, with the use of technology primarily, people can fall back on standardized lifespan tables that come from Social Security or the Society of Actuaries, and that can give a pretty good read on people's lifespan when it comes to the average person. But when it comes to an individual's own health or demographic characteristics, individual lifespans can vary by something like 10 to 20 years, and it doesn't account for that.

And as you pointed out, if you're overestimating lifespan, then the client, the retiree, is not going to be able to consume as much as they otherwise would, and to be able to enjoy their retirement. If you are underestimating lifespan, then you're increasing the chance that this person's going to run out of money in old age and retirement, the longstanding problem, also known as longevity risk.

Adam: Yeah, I think that makes a lot of sense, and I think the point you made in the beginning is a good one, that you're meeting with a new client, you're not necessarily going to lead with a very sensitive topic of, "Well, how long do you expect to live?" or even trying to dig in further to their health history. So, I do think it makes a lot of sense to perhaps lean on some of these actuarial tables or a more generalized approach.

Factors That Have Been Shown To Influence Lifespan [5:40]

Adam: Now, I know that you've done research yourself on some of the factors that can help determine how long a particular client might live. What were some of those that you found to be useful in that regard?

Jeremy: So, I should say that we're doing ongoing research on this issue, as we speak. I've got a co-author who's working on incorporating more variables into the model that we ran in our analysis from 2024 to 2025. And there, the variables we looked at were really basic health variables. So, people generally have a good read over whether they're healthy, whether they're not, whether they've got some chronic condition or some kind of genetic trait, or some kind of adverse health behavior. And saying that you're in fair or poor health, believe it or not, subtracts about eight to ten years of lifespan for the typical retiree. So, it has a big impact. And the other thing that we threw in there was smoking behavior. So, whether or not people currently smoke. And that has an impact, when controlling for other factors, of about five to seven years on lifespan. And then we've got demographic characteristics in there as well. Gender is a very obvious one, which will increase or decrease lifespan by about two to three years. Obviously, women generally live longer than men do. And you've also got things like college education, which has an impact, a smaller impact, of about anywhere from one to two years on projected lifespan. So, all told, these factors can have an impact of about 10 to 20 years, collectively, on projected lifespan. And again, we're currently doing research to incorporate things like chronic conditions, parental lifespan, and we're going to be releasing that hopefully within the next few months.

Adam: That's interesting. And getting back to the first point, about asking them their subjective health, that's a slight twist on the question of, "Well, how long do you expect to live," right? They might not know actuarial tables or how long they're expected to, but they probably do have a good experience in terms of what their own sort of day-to-day health looks like, which, for the planning process, is also potentially useful when it comes to estimating future medical expenses, I would think. So, it's an interesting twist. And we'll talk later on in the conversation about how advisors might sort of bring up this conversation over the course of meeting with a new or existing client.

Adam: So, Jeremy, we've talked about some of the factors that you found, that you found were quite, that really contributed to more accurate lifespan estimates. Were there any sort of characteristics of individuals that were…you might have thought would have been impactful, but turns out might not have been?

Jeremy: Yeah. So, we were actually quite surprised that the demographic variable of race really didn't have a whole lot of impact on projected lifespan, when you control for other factors. So, specifically, the way that we characterize race in our data is we included whether or not the person identified as being Black, and whether or not they identified as being of other race, other than Caucasian in the U.S. And what we found is that these variables really didn't have statistically significant impact on projected lifespan, when controlling for other variables. So, I'm certainly not denying that race, in sort of a coarse way, is related to lifespan. So, if you look at 2019 data, for example, male Native Americans have a projected lifespan of about 69 years, whereas Asian females have a projected lifespan in this country about 87 years. So you're talking about a 20-year difference based on race and gender. But what we find is that when we control for these other variables, like smoking and health and education, that those racial differences go away. So, the good news for advisors and planners is that they don't need to heavily take race into consideration when they project the lifespan of a client.

Adam: Oh, that's interesting. Yeah, I was wondering about the interaction about some of these data points. So, it sounds like, though, in isolation, things like the subjective self-described health, smoking history, those, by themselves, contribute to better estimates, and when put together, create a more accurate estimate. Is that correct?

Jeremy: Yeah, absolutely. So, the best thing that you can do is to incorporate all those simultaneously. If you incorporate them in isolation, you're getting part of the picture, but you're not getting the full picture, because it could conceivably be…I think we all know anecdotes of people who smoke, but they do great. And we know the opposite situation as well. We know that people who die of lung cancer, but they never smoked a cigarette in their life. So, it's important to put these variables together, in a single model. Now, that doesn't mean that you need to have 100 variables there to be able to capture lifespan accurately. In fact, that's one of the core questions that we're attempted to answer in the research that we're doing right now, is, if we were to choose a parsimonious model of lifespan prediction, based on, let's say, five to ten variables, what would those variables be? So, that's the research that we're currently doing. But it's important to have enough in there.

Adam: Yeah, I think that's a good point that you make, that using multiple variables could be helpful in these client conversations. I'm sure a lot of advisors out there listening have had the experience of a client who says, "Well, my mother or father either smoked every day, or engaged in other habits and lived to age 95 or 100, so I expect that I'm going to live to 95 or 100," as it were, too. Or even vice versa, for some reason, they might expect to have a shorter lifespan, but when you put multiple factors together, actually, their lifespan estimate could be more optimistic. So, I think not leaning on one particular data point, but combining together sounds like a pretty powerful tool for the advisors to use in these client conversations.

Jeremy: Yeah, I totally agree. And when we're including just one variable or two variables, again, we're sort of taking a small step away from a non-personalized approach. But the thing, again, that I would endorse is taking a larger step away. And I don't think this necessitates, again, taking in 100 intake variables to predict lifespan, but just a small enough number to be able to predict it accurately, I think is sufficient to formulating a really robust financial plan for somebody.

Free Online Tools That Allow Advisors To Better Estimate Client Lifespans [11:59]

Adam: So, if I'm an advisor, I say, okay, well... I might be able to ask a couple of these questions, if they've ever smoked their subjective health, but now, what do I do with that? Can you maybe talk about some of the resources that are available for advisors to help them sort of apply these and other factors, that might be useful in making personalized lifespan assumptions.

Jeremy: Yeah, so, there's plenty of calculators out there, that will help advisors use this intake information to project future lifespan. So, we've got a blog post that I think is out, or coming out very soon, which discusses these in good detail. And there's three that I highlight, although there's a panoply of others as well. And I just want to highlight these because they're particularly good, they've got particularly good data sources, and they provide different kinds of looks and feels in terms of client intake. So, first and foremost, the Society of Actuaries, of course, one of the lead authorities on actuarial science, they've got a calculator up that's called the Longevity Illustrator, which is pretty parsimonious. It's got a pretty minimal set of highly predictive inputs. The ones that I just outlined, smoking, health status, gender, birth year. And one of the nice things about their calculator is that they don't just project expected lifespan, but different percentiles of lifespan. So, in other words, if you want to be optimistic for a particular individual, and take their 90th percentile lifespan, that's the lifespan at which they're going to die, at or before, with 90% chance. If you want to be conservative and really optimistic about their future lifespan, it will report things like that. And then, if you want more intake variables, you've got a calculator from the University of Connecticut's Goldenson Actuarial Science Center. And there, they've got a bunch of other input variables, things on chronic conditions, exercise, diet, sleep, alcohol use, education, income. And this calculator not only projects lifespan, but it also projects healthy years as well. And health span is a very important input to lots of things, lots of aspects of financial planning, like, for example, the potential need for long-term care in the future.

And then finally, I wanted to highlight the calculator that's provided by Northwestern Mutual, which takes in insurance data as input data, that it estimates a model on. And this is a really, really long kind of intake survey, which includes all the items, pretty much, from the University of Connecticut calculator that I just mentioned, but then also includes certain types of high-risk behaviors, like seatbelt use, driving, drug use, and so on. This is insurance data, so they've got all sorts of invasive data that goes into their model. And then last but not least, I wanted to mention my own company's calculator, ShoreUp Retirement Solutions. We've got a lifespan calculator that's also a Social Security optimizer, which estimates lifespan, and gives higher percentile lifespans if the planner or the advisor wants to be conservative, and then concurrently optimizes Social Security for that particular person. And that's also related to some of the research that I did with my co-authors on the benefit of incorporating these personalized lifespan projections into retirement planning, in the context of Social Security.

Adam: Yeah, that's great. And are all of these calculators free, or is it a subscription?

Jeremy: Yeah, all the ones that I mentioned are free.

Adam: Oh, okay. Terrific. So, yeah, for advisors out there, if you'd like to try out these different calculators, when I saw them, of course, I had to immediately do it for myself. But if you'd like to check them out, you can go to kitces.com/FAT2, that will take you to Jeremy's terrific article on this topic, which has links to all of these calculators.

Incorporating Lifespan Assumptions Into Social Security Claiming Decisions [15:43]

Adam: So, now, I think that's a really good point, that it's not so difficult to sort of run these numbers, and, again, free tools to help you do that. Now, in terms of practical purposes, in terms of being able to leverage these sort of more accurate assumptions, perhaps one example I know that you discussed in your article is about Social Security. How might having this more accurate approach help with the Social Security claiming decision?

Jeremy: Yeah. So, Social Security is a very obvious place that one can plug these kinds of personalized lifespan estimates into. And I think that it's widely known among planners, advisors, about how basic Social Security retirement benefits work. So, first of all, people are generally eligible from age 62 to age 70, so there is some variation around that, for spouses with dependent children, and disabled spouses and so on. But generally speaking, people are eligible for Social Security retirement benefits between 62 and 70. And the later they claim in that age range, the higher their monthly benefit is in real terms, in other words, in terms of buying power, spending power. So, what this implies is that if you've got a shorter expected lifespan, you're better off claiming earlier, because you don't want to miss out on those early benefits from age 62 to 70. Whereas if you've got a very long expected lifespan, you're generally better off claiming later, because, when you've got a lot of monthly payments that are going to accrue to you as a result of your longer lifespan, then you want to make the most of those payments. You want to increase those payments as much as possible, by claiming later.

So, it's just a general sort of fact pattern that the shorter your lifespan, the earlier you want to claim, the longer your lifespan, the later you want to claim. So, personalized lifespan estimation is something that plugs in very naturally to the Social Security estimation, I'm sorry, the Social Security claiming decision. And so the research that I did with my co-authors, Qi Sun, of Pacific Life Insurance and Sean Zhang, of John Carroll University, studies the personalized lifespan estimation issue within this very quantifiable context, because the rules of Social Security are really well-defined. And what we found, somewhat to our surprise, actually, is that personalized lifespan estimation really does make a difference in the amount of around $10,000 on average, across individuals. And this finding's really robust, regardless of how you cut the sample. If you just look at high wealth people, that would be typical financial planning and advising clients, this result still holds, that people, on average, are going to do better, on the order of $10,000 or more, per person, if they incorporate personalized lifespan projections into their Social Security claiming decision.

And this finding kind of extrapolates beyond Social Security benefits. I mean, if it just had to do with that, then, I don't know, it'd be kind of limited importance, I would say. I mean, Social Security is an important benefit to the vast majority of Americans, but with higher-wealth people, it becomes less and less important. But it also matters for things like retirement savings thresholds and withdrawal rates. So, in the blog, we kind of did this analysis, to show that if you look at someone with a quite short expected lifespan at the age of 65, so, someone who smokes, who's in poor health, female, that that person is going to have a median expected lifespan of around 80 years. And someone who's got sort of the opposite attributes, are in excellent health, they don't smoke, they're going to have an expected lifespan of around 96 years, that there's really marked, striking, striking differences between the amount that the former person would need in order to retire comfortably, versus the latter person would need. So, the former person would need an amount that's on the order around $450,000 to $600,000, if you take some variation in the rate of return that they're going to get on the portfolio, whereas the latter person's going to need strictly more. They're going to need something on the order of $620,000 to $1 million in order to retire comfortably. So, this is a really important input variable that planners should take seriously when they construct a plan when it comes to retirement.

Adam: Yeah, and I think throughout the discussion, another theme is that the advisor doesn't necessarily have to take the output of maybe one of the calculators that we discussed earlier, and that's sort of the end-all, be-all. There's room for adjustment, based on a client's risk preferences, for example. So, if, you mentioned sort of the 90% threshold earlier, or perhaps taking the median output and adding a few years on to that, but as I understand, it seems as though, even if you did that, you use the output and then add years on, that's still likely to be more accurate than using sort of a generic number.

Jeremy: Yeah, and then that's kind of the ethic that we're trying to espouse in this research and in the blog piece, is that I think it's very sensible, this is a very, very sensible practice that planners and advisors pursue, and that is to add some kind of cushion, to be conservative about lifespan, to hedge against longevity risk. That makes total sense. But what makes a little bit less sense is doing that in a monolithic way, to assume that even, in spite of the fact that I smoke, that I've got poor parental longevity, I don't take care of myself, I've got multiple chronic conditions, that my optimistic lifespan is going to be the same as yours, Adam, if you don't smoke, you exercise, you take good care of yourself, your parents and your grandparents lived till 100, that my optimistic lifespan and your optimistic lifespan would be the same, it's not really well-rationalized. So it'd be better to take one of these calculators, and add a cushion, as you pointed out, depending on how conservative you want to be, and maybe take the 90th, the 70th percent of lifestyle, or the 90th percentile lifespan, in order to hedge against longevity risk, to the degree that the planner and the client want to.

Adam: Yeah, and I should note as well, if anyone wants to get nerdy, as we often do here at Kitces, within the article, we have a lot of great graphics showing Jeremy's research, some of the outputs on Social Security and these other findings. So, again, if you go to kitces.com/FAT2, you can check that out.

Strategies To Tactfully Raise Lifespan Discussions With Clients [21:58]

Adam: So, let's say I'm one of our advisor listeners. And so, they say, "Great. I really want to...I'm interested in doing these more personalized lifespan assumptions," but, as we alluded to earlier, one of the potential hang ups is, that some of these questions could be a little uncomfortable, especially if you're meeting with a relatively newer client. What are some approaches that advisors might take to sort of get this conversation started in a tactful manner?

Jeremy: Yeah. So, first and foremost, this is a fantastic question, and a really big barrier to having these kinds of conversations is just the level of discomfort that can potentially be on both sides. So, it's an issue, it's an elephant in the room that we absolutely need to address. First and foremost, I think that the discomfort in broaching these topics oftentimes comes primarily on the advisor side, more so than on the client side. So, an advisor will find, if they broach this in a delicate way, that clients can be surprisingly open to discussing these issues, because an advisor is a trusted person, who's supposed to take this information in confidence. And I think the majority of clients see that person that way, that this is a relationship of trust.

Now, in order to make that conversation more comfortable on both sides, there's standard approaches that people can use. And I'm going to borrow some material from a well-known financial planning researcher in mind out there. Her name's Meghaan Lurtz. And she's done some content for Kitces. She's got her own Substack. I think it's called "(Less) Lonely Money." And she endorses an approach where the planner takes this therapy-based, conversational approach, and they don't start with hard questions first. So, they do a lot of rapport-building. They do a lot of listening. They do a lot of empathizing and asking about vision, asking about getting to know this person in a general way, before the advisor starts getting into hard numbers and questions on, that could be uncomfortable. And then, another practice which she endorses, which makes utter sense to me is asking permission to broach these topics, to say, "Look, I wanted to ask you about these sort of health-related issues. Is it okay that I do so?" And I think that advisors would find, if they were to try to adopt this practice, that clients are surprisingly open to discussing these things.

Now, there's other kinds of approaches that advisors can use as well, which I think are really well-rationalized, so, for example, one is to use an intake that's sort of direct to the client, and the client sort of inputs these things in a private way, and they've got optionality with it. And it's also, of course, these days, really important to have security assurances with this data as well. And finally, my background is in behavioral economics, so I'm going to refer to stuff from behavioral science, which is near and dear to my heart. There's ways to frame shorter expected lifespan, that make it more palatable for clients. One is to benchmark or anchor your estimates to lower ranges of lifespan, like 80, and then let's say your estimate pops up that it's 85, 90, 95. Well, that looks a lot better to a client than if you start with, "Hey, the average lifespan at your age going forward is 90," and then they find that theirs is 80. That might not make them feel very good.

So, there's that. And then there's also ways, believe it or not, of framing short lifespan in a favorable way. So, although it's, all else equal, a good thing to live a long time, provided you're going to have healthy years going into the future, but the drawback of that, from a financial standpoint, is that you need more money to retire, that you can withdraw less from your retirement assets over time, and that you may need to have a longer working life. So, there are some distinct, at least financial, advantages to having shorter expected lifespan. You can retire earlier, you don't have to work as late, and you can potentially enjoy the early years of retirement more. So, that's another behavioral approach that one could use, in order to take some of the anxiety away from short expected lifespan.

Adam: Yeah, and getting back to the point earlier about the inputs that are needed to have these assumptions, the advisor's not asking for a blood test or anything like that. It's personal, but still just asking questions here. And also, to the point, it's interesting that you note that we're looking at sort of ranges here. It's not I put your information in a calculator and the crystal ball says "you're going to die when you turn 80," or 75, or whatever it says. It's ranges and percentiles, because this is sort of inherently unknowable.

Jeremy: Correct.

Adam: And I'd imagine part of this conversation, too, is just getting the client to think ahead into the future, in terms of what their life is going to be like at that age, where they think they're going to be. Perhaps that can serve as a bridge, as it were, to sort of the deeper conversation about, well, what does your financial life look like here? What does your health look like there, amongst the other factors?

Jeremy: Mm-hmm. Yeah. Yeah, totally agree.

One Key Takeaway For Advisors [27:12]

Adam: So, before we finish up, one question we always like to ask our guests here on "Financial Advisor Technician" is, for our advisor listeners, given that we've talked about so much today, from the different ways to estimate lifespan, why having a personal lifespan is important, and how to broach this conversation with clients, what's sort of the one key takeaway you'd like our advisor listeners to have from this conversation?

Jeremy: I think the one key takeaway is that advisors spend a lot of time with intake of lots of different financial variables and goals-related inputs, when it comes to client intake, in order to personalize a financial plan for a client, and put comparatively less time in a very critical input, which is the length of retirement, the length of the retirement plan, which is linked to lifespan. And that that particular input is really critical to a financial plan when it comes to savings thresholds, withdrawal rates, Social Security claiming decisions, if the person has a defined benefit pension, what they ought to do with that, annuitization decisions, insurance, etc., etc. And so, because it's got such wide-ranging implications for retirement planning, it makes sense for the planner to put thought and time and energy into that critical input, and personalizing and tailoring it for the particular client's situation, and that it's not that hard to do so, not just in a behavioral sense, to be able to broach the topic with the client, but to find free resources out there that help them pinpoint that number a lot better for a particular client.

Adam: Yeah. That makes a lot of sense. Well, thank you so much, Jeremy, for joining us today on the "Financial Advisor Technician" podcast.

Jeremy: Thanks so much, Adam. Really appreciate the time, and I enjoyed talking to you and your audience.

Adam: And for our listeners, if you'd like to dig deeper into personalized lifespan estimates, including links to all the calculators discussed during the episode, you can go to kitces.com/FAT2, to read the full-length article on this topic. We'll also include a link to that in the episode description in your podcast player.

And as a reminder, Kitces Premier members can earn CE credit for taking quizzes on our technical content, and also have access to our regular CE-eligible webinars, recordings of which can be found in the Members section. If you're interested in becoming a Premier member, we'll put a link in the episode description as well.

Also, if you're enjoying the "Financial Advisor Technician" podcast, please leave us a rating or review on your favorite podcast platform, to help others discover the show. Thanks again for listening, and we'll see you next week on the "Financial Advisor Technician" podcast.