Christine and John discuss the key findings from ‘The State of Retirement Income’ study that they recently published.
On this week’s episode, we’ll be chatting about “The State of Retirement Income” study that Christine and I recently co-authored with our colleague John Rekenthaler, who joins us for this conversation. This is the second year we’ve conducted this study, which examines how much retirees can safely withdraw in retirement. Much has changed since last year as the stock and bond markets have sold off and inflation has risen sharply. This can create uncertainty about whether retirement assets will sustain spending over a multidecade horizon, an issue we explore in this study. In this episode, I’ll be asking Christine and John about “The State of Retirement Income” study and key takeaways. For reference, you can find a link to the study in the show notes to this episode.
Safe Withdrawal Rates
“Determining Withdrawal Rates Using Historical Data,” by William Bengen, Journal of Financial Planning, October 1994.
“The State of Retirement Income,” by Christine Benz, Jeff Ptak, and John Rekenthaler, Morningstar.com, 2022.
“What’s a Safe Withdrawal Rate Today?” by Christine Benz and John Rekenthaler, Morningstar.com, Dec. 13, 2022.
“How Much Do You Know When It Comes to Preparing for Retirement? Fidelity’s Retirement IQ Survey Uncovers Significant Knowledge Gaps,” Fidelity.com, March 6, 2017.
“Bill Bengen: Revisiting Safe Withdrawal Rates,” The Long View podcast, Morningstar.com, Dec. 14, 2021.
“Jonathan Guyton: What the Crisis Means for Retirement Planning,” The Long View podcast, Morningstar.com, June 16, 2020.
“What’s a Safe Retirement Spending Rate for the Decades Ahead?” by Christine Benz and John Rekenthaler, Morningstar.com, Nov. 11, 2021.
Inflation
“How to Build a TIPS Ladder,” by John Rekenthaler, Morningstar.com, Dec. 8, 2022.
“Is It Too Late to Inflation-Proof Your Portfolio?” Christine Benz and Susan Dziubinski, Morningstar.com, Nov. 14, 2022.
The Retirement Spending Smile
“What Is the ‘Retirement Spending Smile’?” by Wade Pfau, retirementresearcher.com.
“Estimating the True Cost of Retirement,” by David Blanchett, Morningstar.com, Nov. 5, 2013.
The Required Minimum Distribution Approach
“Decision Rules and Maximum Initial Withdrawal Rates,” by Jonathan Guyton and William Klinger, Journal of Financial Planning, March 1, 2006.
“Yes, RMDs Can Improve Your Portfolio,” by Christine Benz, Morningstar.com, May 3, 2022.
“ABCs of RMDs,” by Christine Benz, Morningstar.com, June 15, 2022.
Sequence of Return
“Sequence Risk During Retirement,” by John Rekenthaler, Morningstar.com, Aug. 25, 2022.
“Don’t Panic on Sequence of Returns,” by Amy Arnott, Morningstar.com, May 16, 2022.
“How Worried Should New Retirees Be About Market Losses and High Inflation?” by Jeff Ptak and Christine Benz, Morningstar.com, Dec. 15, 2022.
Jeff Ptak: Hi, and welcome to The Long View. I’m Jeff Ptak, chief ratings officer for Morningstar Research Services.
On this week’s episode, we’ll be chatting about “The State of Retirement Income” study that Christine and I recently co-authored with our colleague John Rekenthaler, who joins us for this conversation. This is the second year we’ve conducted this study, which examines how much retirees can safely withdraw in retirement. Much has changed since last year as the stock and bond markets have sold off and inflation has risen sharply. This can create uncertainty about whether retirement assets will sustain spending over a multidecade horizon, an issue we explore in this study. In this episode, I’ll be asking Christine and John about “The State of Retirement Income” study and key takeaways. For reference, you can find a link to the study in the show notes to this episode.
John, welcome back to The Long View, and Christine, thanks for letting me turn the tables and pose questions to you and John.
Christine, to get us started, can you describe “The State of Retirement Income” study in broad terms? For example, what are we trying to assess in conducting this study?
Christine Benz: Figuring out what is a safe withdrawal rate for retirement is one of the biggest problems in financial planning, and we see that people really need help with this. In fact, several years ago, I remember being struck by a Fidelity study where people were surveyed on what they thought was a safe withdrawal rate. And some folks were saying that they thought 10% was a safe starting withdrawal rate. So, people clearly need help with benchmarking. But we wanted to achieve a couple of key things.
Much of the research on withdrawal rates, certainly Bill Bengen’s seminal research on what’s now called the 4% guideline anchors on historical market returns. We aim to look forward to incorporate current valuations, current bond yields. So, that’s, I would say, a key differentiator for our research. And then, another thing is that much of the research about safe withdrawal rates has pointed to the value of being willing to be flexible, being willing to be dynamic in terms of how much you take out and somewhat tethering your withdrawals to what’s going on in the market and your portfolio. A big part of our research, and frankly, a piece that I wish had gotten more attention from the media, and other people have covered this, is the pros and cons of different dynamic strategies. We explored several of them in the course of the paper. And finally, in the 2022 research, we wanted to examine this specific year that in a lot of ways has been just an awful convergence for retirees where they’ve had high inflation, they have had bad returns from bonds, at the same time, stocks are down. And Jeff, I know that it was a big contribution of yours to the paper—examining what retirement researchers call sequence risk, this idea of encountering a really bad market environment right out of the box. So, that was a dimension that we explored in particular in this year’s research.
Ptak: That’s really helpful. And we’re going to tackle a number of those topics—flexible withdrawal strategies, sequence risk—as we make our way through the conversation. Before we get to that though, John, maybe a little bit of context setting. When we looked at the history of starting safe withdrawal rates, going back all the way to the Depression, I think we found they could vary quite a bit it. And so, in your research, what factors have tended to push safe withdrawal rates higher or lower through time as we’ve observed?
John Rekenthaler: Well, there are three factors, Jeff. One is pretty obvious—market returns, the nominal returns on stocks and bonds. I should say that our base case for when we’re calculating withdrawal rates is a portfolio that consists 50% of stocks and 50% of bonds. It’s a half-and-half portfolio over a 30-year time period with the annual withdrawal amounts adjusted fully for inflation. So, they’re fixed in real terms. Market returns are one item. Inflation is another item. And the final one, probably the one that people think about the least, is volatility. A highly volatile market increases the number of steep down years, like the one we’ve had this year. And steep down years, steep losses are the real problem for withdrawal strategies, because then people need to withdraw their assets or spend from a pool, an investment pool that’s severely declined. It takes a bigger chunk out of that pool. And therefore, it’s harder to catch up from that point. So, it’s returns, inflation, and volatility.
And when you look at the history of what have been the highest withdrawal rates and what have been the lowest withdrawal rates, all three of those have been issues. Some of the lowest withdrawal rates surprisingly come from the first time period, 30-year time period that we looked at, from 1927 to 1956, which includes the Great Depression. And actually, the returns aren’t that bad for that, because in the second half of the period, both stocks and bonds performed—stocks in particular performed quite well—but volatility was so high early in the period and so, therefore, overall. And then, inflation in the 1970s. You had the 1930s because of volatility as well as low stock returns and inflation in the 1970s. And finally, looking forward, we have had relatively low projection compared with what people have been able to achieve over history. We’re on the low end of that, not because we think volatility is going to be high and not because our inflation estimate is that great either, but because we have relatively low projection for investment returns, because investments continue to be priced reasonably aggressively compared with historic standards.
Ptak: Christine, this is the second time we’ve run the study, the first one being in 2021. What are the most notable changes from the findings in last year’s study?
Benz: The best way to think about this and discuss this is in the context of that base case that John just referenced. So, we assumed a 50% stock, 50% fixed-income portfolio. We assumed a 30-year time horizon. And we assumed that the retiree was aiming for a 90% success rate—in other words, a 90% chance that he or she wouldn’t outlive his or her assets over that 30-year time horizon. And importantly, we assumed what we have short-handedly called a fixed real withdrawal system. So, we’re assuming that the retiree withdraws X percentage in year one of retirement and then inflation adjusts that dollar amount thereafter, which is the convention for thinking about safe withdrawal rates. And I know we’re going to discuss whether that’s the right way to approach it.
When we compare this year’s result versus last year’s, last year, it was 3.3%, a very low number, certainly relative to that 4% guideline. This year, it was 3.8% for that base case. And the key reason is that the return inputs are higher this year, in part because the market has been really difficult. So, we embedded higher bond returns. That’s a big factor in the mix. The bond returns we used this year—this is a 30-year forecast, which we received from our colleagues in Morningstar Investment Management—the bond return assumption is in the neighborhood of 5% for high-quality fixed income. And then, for equities, it’s in the neighborhood of—depending on the subasset class—ranging between 9% and 12%. And those got a big lift from the numbers we used last year. So, last year, we assumed 3% 30-year fixed-income returns and equity returns as low as 6% for core categories, like U.S. large growth. The higher return assumptions for both stocks and bonds gave that baseline, base case return assumption, and in turn, withdrawal rate assumption, a nice lift from last year.
Ptak: Sticking with you, Christine, one interesting finding was that the starting safe withdrawal rate for the fixed real approach that you’ve been describing, it didn’t increase even if you tilted the portfolio more toward stocks. Why do you think that is, and what do you think the implications are for retirees who are thinking about how to calibrate that stocks/bonds mix in their retirement funds?
Benz: I remember when the three of us initially embarked on this research, and we came out with that 3.3% number last year. And I remember my gut response was like, well, let’s dial up the equity exposure and see what happens. And the reason that the lower equity weightings actually support reasonably high withdrawal rates is just that bond returns are much less variable than equity returns. John referenced that we’re thinking about and embedding standard deviation assumptions, and the standard deviation for bonds is just so much lower than is the case for equities. So, that’s the major reason. If we’re looking at that 3.8% starting withdrawal rate for this year, one thing we found is that you could dial equity exposure all the way down to 30% of assets and still be able to use a 3.8% number as a starting safe withdrawal rate.
Rekenthaler: To add to what Christine said, I think one thing that will come through this discussion is, we’re talking about a single number—in this case, this 3.8% number that’s moved up from 3.3%. And that’s useful. It shows directionally up 0.5%, and I think that’s very useful and accurate information. But there’s a lot of complexity that’s built into this estimate that underlies this. And one thing to be said, or an additional matter to discuss with this is, we’re running trials—in this case, 1,000 different simulations go into this estimate of a 90% success rate. And when you push up, you add more equities and you have more equity portfolio, you’re going to get more trials, circumstances that the coin flips in the wrong direction and you have a sequence of bad returns and that is going to affect when you’re trying to have a 90% success rate, which is a pretty high success rate. You’re trying to succeed 9 times out of 10. And that’s where the volatility, the extra volatility that stocks have really damages chances for a high percentage of a safe withdrawal rate.
But one thing that should be noted is, for the median case, if investment returns don’t go particularly badly or even particularly well, but average, you do have a much higher ending balance with equities than you do with more conservative portfolios. There’s rarely, or very often, not a single best answer when we’re doing this work. It depends on what you’re looking for. And we have defined a 90% success rate with a fixed system, where the investor must achieve 90% success and there are no exceptions in the spending plan allowed. And in that case, it calls for a conservative portfolio given those constraints. But somebody is willing to be more flexible and is, in particular, interested in potentially having a higher balance at the end, and leaving money to heirs or to charity would certainly wish to consider a higher stock allocation.
Ptak: Exactly. Our definition of success itself can have some bearing on what the “safe withdrawal rate” is if you set it …
Rekenthaler: That’s true for anybody, anywhere that is specifying what a retirement plan can be. There are always assumptions in there often not discussed. I think we’re getting our assumptions out front, at least, that really affects the final council that’s given.
Ptak: I want to return to inflation. Christine, in last year’s study, we assumed inflation would average around 2% per year over the following three decades, but it’s run, obviously, quite a bit above the forecast over the past year. Given that, how did we factor inflation into this year’s study?
Benz: As with the return assumptions, we received the inflation forecast from our colleagues in Morningstar Investment Management as well. And so, you’re right, Jeff, that when we got the 30-year forecast from them this year, it was substantially higher, 2.8% versus 2.2%, which is what we used last time around. But I think people might be surprised. Inflation, as we talked today, is what, 7% or so. But I think that 2.8% number assumes that inflation will level out to a more normal level in the years and certainly in the decades ahead.
Rekenthaler: Quietly, core inflation is averaged 3.0% the last two months. So, we’ll see if that continues. But actually, it has subsided to that level over the last two months or so. But clearly, with the inflation estimate—it’s a 30-year estimate—that’s a very long time period. And the belief is that the Federal Reserve will be quite vigilant and not permit 7% inflation rate for sustained period of time.
Ptak: John, how would this starting safe withdrawal rate change if we held everything else constant—so, market returns constant, but assumed that inflation would be, say, twice as high as what we’ve projected? That’s not real world, obviously, because inflation and market returns, there’s interplay there. But under those simplistic assumptions, do we know how the starting safe withdrawal rate would change?
Rekenthaler: We do. It’d be a disaster. You’d drop from 3.8% to 2.6%. You’d be at 2.6%. I should point out, we have not experienced in the United States a 30-year period like that. Because as you state, when inflation rises, bond yields adjust, they go higher. The nominal returns on bond, if inflation were 5.6%—which is what would happen if you double our rate—bond yields would rise and the nominal return on bonds would surely be higher than the 5% that we have projected for that 30-year. And stock returns would rise too, because inflation seeps into stock returns as well. We saw that in the 1970s. So, I don’t think that would occur. But if it did occur, yeah, spend a whole lot less.
Ptak: Staying with you, John, as I know, you’ve written recently on this topic. For those retirees who have been maybe rattled by the spike in inflation, can’t stomach the possibility of seeing their spending power erode in the future or even the thought of that. What would you suggest they do? What has your research revealed?
Rekenthaler: That’s a little bit of a different scenario than what our paper is addressing, Jeff. Again, look at our assumptions. Our assumptions are that inflation will rise but be manageable and that stock and bond returns will have positive real yields. They will be higher than the rate of inflation. And actually, if inflation does double, that won’t necessarily invalidate what we do because it’s quite likely that stock and bond yields will also move in sync with that. That’s what has historically happened over long periods of time. But let’s say, we’re in a situation where that doesn’t occur—inflation runs rampant, and equity and bond prices over a long period of time do not keep pace. In that case, what an investor would want is something that’s outside of the scope of this study, and that would be a ladder of TIPS—Treasury Inflation-Protected Securities. That you could absolutely lock in inflation-adjusted real income or a real return. I’ve been talking to several people who are quite well-informed who have been building TIPS portfolios and assembling TIPS portfolios just to guard against that possibility. They’re not doing it with their entire portfolio or even a majority of their portfolios. But they’re doing it with part of it, just to lock in and make sure if the worst case for inflation does arrive that they’re protected. And that’s really the tool that one would use would be Treasury Inflation-Protected Securities.
Benz: It seems like that could be a really elegant solution for especially someone’s fixed spending where maybe in lieu of an annuity, someone could consider employing that ladder TIPS portfolio that John just talked about. But I think it does make sense for wealthier folks.
Rekenthaler: Right. Both TIPS and annuities guard against… They have guaranteed aspects to them, but there’s a key difference between them in that annuities typically are not inflation-adjusted.
Benz: Absolutely. And you’re spending your principal.
Rekenthaler: With the TIPS ladder, effectively you’re spending your principal, too. So, that’s the similarity of them. They’re both spending strategies. They’re not spending strategies that’s going to end up potentially building a portfolio as is often the case with the portfolios that we model in our study. But they do operate very differently. If somebody is worried about high inflation, annuities are not the way to go, not the way they’re normally sold, whereas TIPS will provide that protection.
Benz: Right.
Ptak: And, John, you mentioned that your TIPS ladder idea. I think you wrote a piece on that recently, which we will include in the show notes. Christine, one method we examine for the first time in the study is a withdrawal approach where the retiree doesn’t increase spending by the full amount of inflation. The notion is that because spending doesn’t typically keep up with inflation over the full retirement horizon anyway, it makes sense to model in spending adjustments that are slightly lower than CPI. What was the starting safe withdrawal rate under that method?
Benz: That method, which we short-handed as the inflation haircut, arrived at a 4.3% starting safe withdrawal amount versus that 3.8% in our base case. So, that holds all of the other constants the same. So, we assume a 50-50 portfolio, that 30-year time horizon, 90% success rate. But it arrives at a higher starting safe withdrawal. And I think that’s valuable because one thing that we had wanted to include in this year’s research, which was something that looks at how retirees actually spend. So, in communicating with David Blanchett—our former colleague at Morningstar, who is now at PGIM—he talked to us about what has often been called the retirement spending smile, his research on how retirees actually spend over their lifecycle. And he has found this pattern that somewhat resembles a smile, where you’ve got people spending higher levels early on in retirement, which is quite intuitive, those pent-up demand years of retirement when people are feeling good and maybe doing heavy travel. Then he found that spending generally trails down later in retirement, certainly in the mid- to-later years of retirement and then it may increase a little bit later on. But in talking to him, his advice was, well, yeah, when I look at this on average, retirees spend less than the inflation rate. They spend about 1 percentage point less than the inflation rate. So, for this base case, we modeled in a 1.8% inflation rate, effectively giving that 2.8% inflation rate a 1 percentage point haircut. So, that’s a strategy that retirees could consider.
Ptak: John, that method seems to check a lot of boxes to the withdrawal amounts were about as stable as the traditional Bengen approach, and it boasted the second-best median-ending balance of the methods we tested, which would seem to be well-suited to those that have a stronger bequest motive. What do you see as the biggest drawbacks of this approach?
Rekenthaler: I don’t think there are any large drawbacks associated with the approach of ratcheting down spending slowly over time by taking not the full inflation adjustment. One thing to point out is, it’s not free money. You get a higher starting withdrawal rate. But because the inflation adjustment is cut by 1% per year, over time the real spending amount does decline, and you end up with a similar withdrawal rate over the entire 30-year history. So, we’re not transmitting lead into gold here. We’re just changing the spending pattern to one that, in most cases, it is more realistic. I’m not planning on retiring, but I’m at the age when coming up on 62 when people often do retire, and my wife and I are healthy and active and we’re doing lots of different things, and perhaps, we will not be doing as many of those things 20 years from now. We’re not the only ones that have young adult children that perhaps require assistance at times, and that would also change.
I think it’s realistic. But again, this is a model. Individual circumstances can change because people can suddenly have, for whatever reasons, higher spending needs than they thought they had, and in particular, healthcare issues could crop up, which is why David Blanchett talked about a smile, about it rising at the end. So, that probably can and should be addressed by a separate pool of money going toward long-term care or something like that, insurance. But overall, I think this is a sensible way for people to go. They could pull out a little more money, withdraw money earlier. You get over the 4% mark that is the general rule of thumb, if you have that as a plan. And it pretty much matches what people do. When I’ve discussed this topic, I hear from retirees, and they say, “Yeah, exactly. I’m trying to maintain my purchasing power within reason, but I’m not a fanatic about it, and I’m spending a little bit less over time.” So, I think there’s a lot to be said for the approach.
Benz: I think so, too. And just to pick up on something John was talking about with the long-term-care expenses: I like the idea of actually segregating—if someone is self-funding long-term care, segregating those funds from the spendable assets. So, for your modeling purposes, as you think about your spendable portfolio, separate out any long-term-care fund assets. I sometimes think of it as maybe a fourth bucket in the portfolio where it’s possibly going to fund long-term-care needs, possibly would be your bequest fund, or your fund if you live to be a 105 or something like that. But I like the idea of people who do not have any sort of long-term-care insurance segregating those funds from their spendable funds.
Ptak: We’ve talked a lot about the starting safe withdrawal rate, but with flexible spending approaches, which we’re going to spend a little bit more time talking about now, that measure might not tell the whole story. John, can you talk about the notion of a lifetime withdrawal rate and how these different methods stack up on that basis?
Rekenthaler: Yes. Our paper discusses two withdrawal rates. One is the starting withdrawal rate, so that’s the amount of money, percentage of the portfolio that one would withdraw in the first year, such as the number we’ve said, 3.8%. So, you have $1 million portfolio. You can withdraw $38,000 from it. And in our base case method, which is fixed real, that number does not change, it’s $38,000 inflation-adjusted throughout the 30-year life span, which means that the lifetime withdrawal rate, which is the lifetime withdrawal rate we define as an average withdrawal rate over the 30 years. What’s the withdrawal rate in year one, what’s the withdrawal rate in year two, year three, and so forth? We add up all those withdrawal rates, divide it by 30. That’s the average withdrawal rate. It’s the same: It’s 3.8%, because it’s fixed. The spending is fixed in real terms.
But every other method that we evaluate, the spending is not fixed. And in fact, some of the flexible-spending methods, it varies quite a bit. It will go up. Again, I’m always speaking in inflation-adjusted returns. It will go down depending upon the market returns. So, what you have very often is a pattern where you can have, because you have the flexibility, a higher starting withdrawal rate. But you can’t necessarily sustain that withdrawal rate throughout the time period. It will come down often. Overall, the lifetime withdrawal rate does remain higher for these flexible methods. And we have several methods that we haven’t discussed, but the fixed rate of 3.8% under the base case goes from anywhere from 3.9% all the way up to 5.4% for these various flexible methods. That’s the lifetime rate. So, you not only get a higher starting rate, you do get a higher lifetime rate, but the lifetime rates are not as much higher as the starting rate is, if that makes sense. So, there’s not as much of a bump when you take the entire 30 years into consideration.
Earlier, I said, you can’t turn lead into gold. And now, I’m talking about higher lifetime rates. So, am I contradicting myself? Well, I try not to. Two things going on here. One, the ending balances are different for these strategies. The fixed-base case that we have has the very highest ending balance, or among the very highest ending balances, on average—or median, the median case at year 30. These other strategies are more aggressive about spending down the portfolio during the bull markets. So, what that means in practice is, if the markets go up a lot, their withdrawal rates go up a lot, and therefore, that contributes toward the higher lifetime rate. But that’s only under the bull market cases. And it’s important to recognize finally to say that with a fixed case, nobody is locked in, nobody is committed, nobody has signed a contract that says I’m going to spend 3.8% for the next 30 years inflation-adjusted, and I’m never going to change that. If there’s a strong bull market as we had, say, during the 2010s or during the 1990s, a portfolio that’s 50% stocks, 50% bonds is likely to double in real terms over that time, even as you’re taking money out of it. The person who has a fixed real plan probably will end up increasing their rates. So, they’ll be flexible, in practice, they will do some of what these flexible schemes do. These are model results. Again, they are instructive for helping us understand how different decisions, the effect, the size of the effect, and the direction of the effect the different decisions can make. But we need to recognize that they are models, and in practice, people will mix and match some of these strategies, and they are not locked into what decision they make.
Ptak: And I think that one of the messages that was loud and clear is that these flexible-spending methods, like all of the methods we examined, involve trade-offs. Christine, I wondered if you could talk about that in the context of something like the RMD method, which isn’t a method we’ve talked quite as much about. What are the pros and cons of that RMD approach, and in which other method do you think it contrasts most strongly with?
Benz: I tend to group these flexible withdrawal strategies into two key categories. One would be more modest tweaks, mainly if the portfolio has had losses, and those tend to deliver, as you would expect, a more stable stream of income. And then, two others that we examined take haircuts in bad markets, but they also give raises, as John was just talking about, in good market environments. So, we tested two in this year’s study and in last year’s research as well. One is called the guardrails approach, which was pioneered by Jonathan Guyton and William Klinger. And the other is simply an RMD-style method where you are annually updating your withdrawal rate based on whatever your portfolio balance is, divided by your life expectancy. This approach does deliver the highest lifetime payout of any of the methods we tested, and in that respect, it’s very efficient. It gets the retiree taking less in a year like 2022 and taking more in the series of years that we had from 2019 through 2021. If people are taking RMDs, they know this, that the RMD is forcing them to spend from their IRA or other tax-deferred account on an ongoing basis.
But you just referenced the trade-offs, Jeff, and the big trade-off for the RMD method is volatility, that you get buffeted around by whatever your portfolio is doing at any given point in time, and the higher your equity exposure is in that portfolio, the more volatility you’ll have in that portfolio and in turn, the more volatile your cash flows will be. So, volatility is the big disadvantage with the RMD approach. I would say for a lot of people where their portfolio is a key income source, I would say that variability may make that RMD approach a nonstarter from the standpoint of a livable cash flow system. And the other issue is that because it does urge retirees to spend more on an ongoing basis, it gives them those raises in good markets. It’s one of the worst—I guess, it’s the worst, from the standpoint of bequests of having money left over at the end. So, it’s best suited for people who are wealthier, who have other income sources that they’re relying on, and also for people who have a consumption mindset, who want to spend as much as they can from their portfolio during their own lifetimes without really worrying about having leftovers.
Rekenthaler: There is also a greater risk of longevity risk as well, because we’ve built this model with a 30-year outlook and RMD aggressively spends to that year-30 point. And if you live year 31 or year 32, in a lot of those cases that we’ve defined as successful, there’s not a lot left. Whereas with more traditional fixed methods, there tends to be quite a bit left in the kitty, as you said. So, something that’s worth pointing out as well. We have a 5.4% lifetime withdrawal rate, on average, for the RMD method as opposed to 3.8%. You don’t get that 1.6% without a lot of trade-offs. That extra 1.6 percentage points—there are a lot of trade-offs. I would not just jump and say, wow, 5.4%, that’s the best number, let’s go for it.
Benz: Right.
Rekenthaler: Possibly, but understand how much variability there is in the cash flow and how low that account is likely to be at year 30.
Ptak: It seems like the method where there’s a greater payoff to tilting the portfolio to equity is something like the guardrail spending approach, which we’ve alluded to previously. It was popularized by advisor Jonathan Guyton. I think we found that a portfolio split 80-20 between stocks and bonds would support a 5.6% annual lifetime withdrawal, which I think is almost 2% higher than what the fixed real withdrawal method can support. So, what is it about that method that allows for such sizable withdrawals and why is it better suited, in your opinion, to equity-heavy portfolios than some of the other methods?
Benz: John and I and you, Jeff, we’ve all talked about our favorite of these methods, and I think we all are attracted to the guardrails strategy, and anecdotally, it seems like a lot of financial advisors I talk to use some version, maybe a modified version, of guardrails. But as I said, it does give people a lift in their paydays when the market is up, and often the market is up. And so, that’s the key reason why it does deliver such a nice lift to portfolios, especially equity-heavy portfolios. But it does come with more variability certainly than some of the very modest tweaks that one might make to a fixed real withdrawal system. People have to be comfortable with the trade-offs. It’s also, like the RMD method, very focused on getting people to consume and enjoy the fruits of their labor and so, therefore, maybe less appropriate for people who are very bequest-minded.
Rekenthaler: I think of guardrails as a sober version of the RMD method. RMD is tipsy.
Benz: Jeff, I want to jump in and turn the tables here, because you worked on the section of our paper that focused on sequence risk, and you examine sequence risk through the lens of 2022′s terrible market environment. Can you talk about sequence risk? What sequence risk is? Is it just a starting point?
Ptak: It’s the risk of running out of money in retirement caused by losses early on in retirement. And so, the question is, why do early losses matter? Well, they can forestall the stock and bond gains a retiree needs to grow retirement funds over time and they can also force retirees to sell assets to support spending at inopportune times when stocks and bonds boast more attractive expected returns. So, when you take the two together, that can increase the chance of a retiree outliving his or her assets by the end of retirement. That’s really sequence-of-returns risk in a nutshell.
Benz: Does 2022 epitomize sequence risk? And, I’m wondering if you can talk about what are the parallels for this period in modern market history?
Ptak: I think it does epitomize sequence risk in that we’ve seen both stocks and bonds fall, and we’ve seen inflation rise rapidly, which is an unholy trinity, I suppose, you’d say. Have we seen parallels to this? We have, but not many where we’ve seen all three presenting challenges to retirees at the same time and the way they have recently. In fact, I took a look, and a portfolio split 50-50 between stocks and bonds, it lost around 20%, give or take, after inflation over the year ended Sept. 30, 2022, and that’s the 20-poorest showing over a 12-month period since the Depression, and this is according to Ibbotson data. And so, that puts it in the second-worst percentile of all of these rolling 12-month periods going back to the Depression. So, while we’ve seen conditions like these from time to time before, they’ve definitely been the exception, not the rule.
Benz: John, you’ve written about inflation as a form of sequence risk, and we’ve put the question of inflation as a form of sequence risk to several of our guests on The Long View. Can you talk about what are the implications for a retiree if high inflation shows up early on in someone’s retirement versus if it shows up when you’ve been retired already for 20 years or something like that?
Rekenthaler: Yes, it’s not good. You’d rather have the bad news come later. It’s a similar story as with market returns, because there’s a ripple effect. And if it shows up early, there’s a ripple effect through time. If there are price increases in year one, that inflation rate effectively, those price increases, prices are higher for the next 29 years, all things being equal, than they would be. And that cuts into the investment pool because the spending goes up by the amount of inflation. So, spending goes up quite a bit. And if the investment pool is down, you’re spending more out of a declining pool at the wrong time potentially, as Jeff said. So, it’s really the same kind of story as with bad market returns. All bad news for a retiree portfolio is best pushed off later in the portfolio, in the later years. In fact, if the good years, or let’s say, the first 10 years have strong returns, low inflation, the investor is really set. It’s really pretty hard for bad news to take down a portfolio over the next 20 years if the first 10 are so strong. But if the first 10 are weak, it’s a lot more vulnerable.
Benz: Jeff, let’s look at the implications of that for people who embarked on retirement this year in 2022. If they had taken our 2021 guidance to heart and started with a 3.3% withdrawal this year, how is this year’s turn of events affected the odds that they will have funds left over in 30 years from now?
Ptak: The odds have worsened, whereas under our prior assumptions, we thought retirees taking that initial 3.3% withdrawal had a 90% likelihood of reaching the end of retirement without exhausting their assets. Those odds of success have fallen to about 78%. And that’s even after incorporating the higher expected stock and bond returns that we referenced earlier in the conversation. That maybe doesn’t seem like a huge deterioration, but it does mean that the odds of failure went from 1 in 10 under our previous assumptions to almost 1 in 4, which is not nothing.
Benz: And how about people who took out an initial withdrawal that was closer to 4% this year, using the standard guidance and then inflation-adjusted that amount thereafter. How has this year’s bad convergence affected their odds of having money left over after 30 years?
Ptak: If someone took an initial 4% withdrawal, which, as you know, exceeded the 3.3% we estimated as the initial safe withdrawal in last year’s study, they’re looking at a more sobering picture for sure. The odds of reaching the end of retirement successfully, where they haven’t exhausted their assets are a little less than 50-50. I think we found it was around 46% or 48%, forgive me. So, for them, they need to think seriously about some steps that they can take to perhaps recalibrate spending so that they’re not courting as much risk of exhausting their assets by the end of an assumed 30-year retirement horizon.
Benz: What steps should retirees take if they’re in that situation and they fear that perhaps they have overspent this year given what’s been going on? What steps should they take to course-correct if they’re worried that they overdid it this year?
Ptak: We explore a number of methods in the paper, and we’ve talked about a number of those whereby the retiree throttles back on spending a bit and that can help as it gives the portfolio a chance to recover through subsequent market gains. Whereas a continuation of inflation-adjusted spending, it courts the risk of short-circuiting that compounding process. That’s sequence-of-returns risk defined. One simple, even if it isn’t painless approach, is to cut spending by 10% following a losing year. That’s one of the flexible withdrawal approaches that we’ve evaluated. When we examined that in the case of a retiree who took an initial 3.3% withdrawal, we found that it boosted the odds of successfully completing retirement from 78% to 84%, which is a material improvement. You’re not quite to 90% again. But you’re about halfway there and it’s a relatively simple-to-implement step in the event that you’re concerned about longevity risk.
Benz: One question I have for you is just sequence risk. I guess by definition, it’s mainly relevant to people who have just begun retirement, and a lot of the numbers that we’ve been talking about are mainly relevant to people who have just started retirement. How about those people who have been retired for 10 or 15 years or more? Sounds like they should probably worry less about sequence risk. What should they use to benchmark how much they’re taking out currently is on track with where it should be?
Ptak: I would like to think that they would use methods similar to what we’ve used in this paper, where they’re making certain assumptions about inflation and attainable market returns over their investment horizon and see what that might portend for their assets given various spending rates. And so, they’ll be going through the same sorts of progressions that we’ve gone through in conducting this study albeit over a shorter time horizon. I would welcome your and John’s perspective on this, because I think that you’ve done a lot of excellent work and thinking on it. I tend to think of the risk as being less acute for someone who is further along in retirement just given the fact that they by definition should have less horizon. And if they also presumably have enjoyed the benefits of compounding through the first, say, 10 or 15 years of retirement, the market has worked for them notwithstanding the inflation-adjusted spending that they might have done along the way, and so, they’re not quite as vulnerable to sequence-of-returns risk as somebody that is just embarking on retirement. But Christine, John, what do you think? Do you think that maybe that somewhat understates the risk that some of these folks could be courting?
Rekenthaler: No, I think that’s correct. There’s two ways I look at it. One is, in our models, we’re modeling what effectively is the worst 10% of outcomes for a 90% success rate. So, the typical median portfolio people are going to have—unless they have the RMD strategy, which is very aggressive spend down and maybe to a little lesser extent the guardrail strategy—they’re going to have a nice cushion built up and they can afford to take some hits. And the other item is, sequence of risk by definition is sequence over a long time period. It’s about the ripple effect of what happens now playing out over many years. If you don’t have that many years left to live, there’s less sequence risk.
Benz: We have an exhibit in the paper that looks at shorter time periods, not just the 30-year time period, but also, shorter and longer time periods. So, you can look at different asset allocations, like a 50-50 asset allocation would lend itself to a 6.6% withdrawal rate for someone who has a 15-year time horizon. I like that as a benchmark. RMDs, I think, can be a check that retirees can use. If they’re spending in line with their RMDs, that’s probably a reasonable approach. And those are two benchmarks. Because I do think that people often have that need to check up on whatever they’re spending on an ongoing basis, not just in year one.
Ptak: Well, Christine and John, this has been a really useful recap of The Retirement Income study. Thanks so much for sharing your time and insights today.
Rekenthaler: Sure thing, Jeff.
Benz: Well, thanks so much, Jeff. It’s been fun working with you and John on this.
Rekenthaler: Absolutely.
Ptak: Thanks for joining us on The Long View. If you could, please take a minute to subscribe to and rate the podcast on Apple, Spotify, or wherever you get your podcasts.
You can follow us on Twitter @Syouth1, which is, S-Y-O-U-T-H and the number 1.
Benz: And @Christine_Benz.
Ptak: George Castady is our engineer for the podcast and Kari Greczek produces the show notes each week.
Finally, we’d love to get your feedback. If you have a comment or a guest idea, please email us at TheLongView@Morningstar.com. Until next time, thanks for joining us.
(Disclaimer: This recording is for informational purposes only and should not be considered investment advice. Opinions expressed are as of the date of recording. Such opinions are subject to change. The views and opinions of guests on this program are not necessarily those of Morningstar, Inc. and its affiliates. Morningstar and its affiliates are not affiliated with this guest or his or her business affiliates unless otherwise stated. Morningstar does not guarantee the accuracy, or the completeness of the data presented herein. Jeff Ptak is an employee of Morningstar Research Services LLC. Morningstar Research Services is a subsidiary of Morningstar, Inc. and is registered with and governed by the U.S. Securities and Exchange Commission. Morningstar Research Services shall not be responsible for any trading decisions, damages or other losses resulting from or related to the information, data analysis, or opinions, or their use. Past performance is not a guarantee of future results. All investments are subject to investment risk, including possible loss of principal. Individuals should seriously consider if an investment is suitable for them by referencing their own financial position, investment objectives and risk profile before making any investment decision.)