By Dr. James M. Dahle, WCI Founder
I am often asked to write more articles aimed at those who are retired or approaching retirement. In many ways, spending down your assets efficiently is a lot more complicated than accumulating them in the first place. In recent years, there are more and more articles being published on this topic, with Wade Pfau probably being the most prominent name in the field, at least among Boglehead types. In his 2016 book on Reverse Mortgages, he spent the first chapter on an excellent overview of the challenges of providing retirement income.
In this post, I will borrow heavily from his ideas, so I thought it only fair to plug his book a bit. So uh . . . buy his book. It's really good.
Your Balance Sheet
Your household has a balance sheet. As you'll recall from the mini-MBA they didn't give you in medical school, a balance sheet has the assets listed on the left and the liabilities listed on the right. That balance sheet for someone approaching or already in retirement might look like this:
Everyone's balance sheet will have a different dollar amount next to each of these assets and liabilities, but there are two keys to solid retirement income planning. The first is to make sure the assets cover the liabilities. If they do not, some adjustments need to be made to one side of the ledger or the other. Secondarily, you want to match the liabilities with the assets in the most “efficient” way possible. By efficient, I mean with the lowest likelihood of failure and the lowest tax bill without leaving money on the table. Even if you are well beyond “enough,” the more efficiently you manage this process, the more you can put toward your legacy.
Tapping Your Assets
You can use your assets to provide funds to spend in retirement in many different ways. Money is fungible, and while there are transaction costs and taxes to mind, you can often change it from one form to another relatively easily. For example, if you prefer to accumulate your assets using mutual funds and spend them as real estate rents, you can simply sell your mutual funds and buy income properties upon retiring.
But even for a given asset, such as home equity, there are a number of different ways you can use it to match your liabilities. By keeping your paid-off home and living in it, you can reduce your basic living expenses and then use the value of the home when you die as part of your legacy. You can sell the house and move into something smaller, investing the difference to help provide retirement income. You can rent out part of it with Airbnb and earn money from it as a retiree. You can take out a mortgage on your home prior to retirement and invest that money, hopefully earning a higher return than the home equity would provide you. You can also take out a home equity line of credit (HELOC) and spend it as you go throughout retirement. A reverse mortgage is another option.
One asset, many different uses. Which is best? Well, it depends on the entire picture of your household balance sheet.
Two Schools of Thought
There are really two schools of thought when it comes to retirement income planning. Pfau describes them well in his book:
“Two fundamentally different philosophies for retirement income planning—which I call “probability-based” and “safety-first”—diverge on the critical issue of where a retirement plan is best served: in the risk/reward tradeoffs of an equity portfolio, or the contractual guarantee of insurance products . . . those favoring investments (the probability-based) rely on the notion that the market will eventually provide favorable returns for most retirees. Though stock markets are volatile, stocks can be expected to outperform bonds over a reasonable amount of time. The investments crowd considers upside potential from a portfolio to be so significant that insurance solutions can only play a minimal role . . . Meanwhile, those favoring insurance (safety-first) believe contractual guarantees to be reliable and that staking your retirement income on the assumption that favorable market returns will eventually arrive is emotionally overwhelming and dangerous. The insurance side is clearly more concerned with the implications of market risk than those favoring investments . . . [they view] investment-only solutions as undesirable because the retiree retains all the longevity and market risks, which an insurance company is in a better position to manage.”
Obviously, this isn't an either/or proposition, despite the fact that many “advisors” basically only look at one side or the other of this continuum. Each of us is likely to find ourselves somewhere different on the continuum between investments and insurance, and that's fine. I get the impression that Pfau leans far more to the insurance side than I do, but that's primarily due to his pessimistic views about future market returns and his distaste for risk [for what it's worth: from 2017-2021, the S&P 500 rose at least 16.26% in four of those five years].
But it doesn't really matter how Pfau feels or how I feel about this issue. What matters is how you feel about it. The more you worry about leaving something on the table by being cautious, the more you'll lean toward the probability side. The more you worry about running out of money, the more you'll lean toward the insurance side.
The advantages of the probability approach include liquidity and the upside potential for increasing your standard of living and your legacy. Less money is also “probably” needed to reach your lifestyle and legacy goals. The advantages of the safety-first approach include less longevity risk, less sequence of returns risk, and less worry about declining cognitive abilities.
Pfau continues:
“For probability-based thinkers, a 90% chance of successfully meeting their goals is a more than reasonable starting point and the retiree can proceed with the plan. It has a high likelihood of success and that's enough for them. If future updates determine that the plan might be on course toward failure, a few changes, such as a small reduction in spending, should be sufficient to get the plan back on track.
Those identifying with the safety-first school, however, will not be comfortable with this level of risk, focusing instead on the 10% chance of failure. They make a distinction between essential expenses and discretionary expenses and seek a solution that practically eliminates the possibility of failure for meeting essential expenses. Jeopardizing success, they say, is only reasonable for discretionary expenses.”
As Bill Bernstein has said, “When you've won the game, stop playing.” It is important that you understand where both you and your advisor sit on this continuum.
At the far left side, you may carry a 100% equity portfolio throughout retirement and even carry a mortgage into retirement to keep more money in the market. If you have income properties, you would prefer to have two with a 50% mortgage on each of them rather than one paid-off mortgage.
At the far right side, you've annuitized almost all of your assets, reverse mortgaged your home, bought a permanent life insurance policy precisely equal to your legacy desires, and own a long-term care policy.
The wise course is likely somewhere in the middle. Pfau concludes his chapter, a masterpiece by itself wedged into a book on a rather specialized piece of it, with this graphic:
I love his 4 Ls—the goals of longevity, lifestyle, legacy, and liquidity.
- Longevity: Make sure your assets can provide your essential expenses for your entire life
- Lifestyle: Allow you to live “the good life” in retirement
- Legacy: Leave something behind for those people and causes that you care about
- Liquidity: Make sure you're covered in case something bad happens
However, I also appreciate the way in which he matches assets to liabilities. The more reliable the income, the more it can be used for essential expenses. Meanwhile, those assets with a higher expected long-term return fund your lifestyle and legacy goals. Then you cover contingencies with a hodgepodge of assets ranging from insurance to the income of your children.
What do you think? Where do you fall on the continuum between probability-based and safety-first? How do you plan to match your assets with your liabilities in retirement? Comment below!
[This updated post was originally published in 2017.]
My general framework is to divide my portfolio into two buckets; the retirement bucket I’ll use for retirement spending, and the estate bucket that will be left to my heirs. The size of the retirement bucket would be in accordance with the 4% rule. The estate bucket would be the rest of my assets.
I would withdraw money from the retirement bucket using the 4% rule, which has been shown to be successful with a high probability. The estate bucket would be invested in 100% stocks, as that money is not for me, but for my children/grandchildren, who have a very long time horizon. Clearly, if there are some unforeseen circumstances where there is any risk that the retirement bucket may run dry, I could draw from the estate bucket.
I’ll admit this is very simplistic compared to the excellent work by Wade Pfau, but it works for me.
I’m assuming retirement bucket is 50% stocks and 50% bonds (that is where 4% rule came from i’ve read).
The 4% rule is not a fundamental law of physics and I prefer Wade Pfaus’s pragmatic approach.
Are you not assuming some risk with insurance products r.e. The insurer’s viability? I am looking toward retirement with dividend stocks and bonds providing enough dividend income to more that meet all reasonable needs, without the need to planned-sell assets to fund retirement.
Though others can provide alot more detail, insurance policies (annuity, whole life) are supervised by the State in which the insurance company is located. Each State is going to differ, but in the State of Illinois for instance there is a fund all insurance companies pay into. The fund can be tapped for the benefit of policy holders in case of failure of the insurance company. There is a significant amount of supervision by the State also, with lots of reporting on assets/liabilities etc. The big stick for the State though is the ability to shut down any insurance company from underwriting anything at a moment’s notice, so companies tend not to piss off the State insurance regulators too much.
Have you looked at the State of Illinois’ finances recently? Do you honestly believe they’d make you whole and continue your annuity income stream without disruption in the event their was a financial cataclysm such that several large insurers in the state went bust?
Sure, it’s a low prob event, but I wouldn’t want to bet my own well being on that ‘guarantee’.
Yes, I know about the State of Illinois and its finances (rather lack thereof). The distinction in this case is that a fund like this would be segregated from State access/obligations. Withdraws from the fund cannot occur except for the failure of the insurance company and an inability to pay contracted policy holders. Both Allstate and State Farm are located in and supervised by the State of Illinois from an insurance perspective, and neither had any issues during the recession of the late 2000’s. What I described is not a guarantee, it is an State mandated industry-fund financial backstop. My original post was to describe the insurance backstop mechanism, not to dissuade Clair from her approach that if structured and funded properly is better than some type of insurance contract.
Yes, there is some risk there, although if you keep the amounts below the state insurance guaranty corp limits it is significantly mitigated.
Your approach- to only spend the income, is the equivalent of using a 2-3% withdrawal rate. Sure, it works, but it works because you’re likely to leave tons of money on the table.
The state guaranty associations actually don’t have much money and neither the states nor the federal government is obligated to step in. The association tries to make people whole by taxing the intact insurers and facilitating them taking over policies of failed companies. In the past when insurers have gone under, the association has made people whole 94% of the time for annuities and 96% for insurance.
Boglehead types would not agree that “a 90% chance of successfully meeting their goals is a more than reasonable starting point.” Most using a probability-based approach to retirement (myself included) would want that number as close to 100% as possible.
I may not use the safety-first upon initial retirement, but it makes logical sense to move in that direction as one ages and cognitive abilities decline.
You can get as close to 100% as possible, but it’s going to cost you. No free lunch.
I definitely fall in the probability-based approach camp at this point in my life. I’m moving aggressively to get that probability as close to 100% as possible by building up my asset base using a combination of a well-diversified equity portfolio and rental properties. Can’t predict the future, but I feel if I achieve my goal early, and I continue to work as much as I enjoy, I should have plenty to live on. However, depending on how it goes over the next 10-15 years, I would consider adding in some safety-first components.
My family has a weird risk some others of us might share, and I guess the answer is self or outside insurance or reassuring calculation that it is unnecessary: Military/VA pension adequate for most of our current needs, which drops to 35% or so if husband dies. In fact if I die he gets a bumpup as he quits paying survivor benefit insurance on his pension. It’s why he can let his medical skills go but I probably shouldn’t (though maybe if I want to fully retire I should just get term life on him through my 60s or whenever I’m certain I’ll want to retire). Of course I’m less expensive than he is…
Jenn – we are similar, but I have stayed in the military long enough and at high enough rank that our pension will actually be enough to cover a pretty lavish lifestyle by itself, and will drop to 55% if I die. We got around that by “over-saving” such that the 4% rule (or 3.5 or whatever) will also cover that same happy lifestyle. The “over-saving” will wind up going into the “legacy” category most likely. We leave it in 100% equities since the military pension is a fairly “safe” asset. So I guess my advice is self-insure with your legacy assets.
So like us if the army retiree spouse dies the survivor lives on less plus draws more from the intended legacy than otherwise. Does your spouse think (s)he’ll need more than 55% if you’re gone? Hadn’t considered it until now but this risk is another good reason for us to have paid off our mortgage- just in case I’m a young widow my monthly fixed costs are a lot less already and having to move/sell will not be so acute an issue as if it weren’t to be paid off until we were 60+
We think now we don’t want to leave much legacy- but are in a cautious ‘we can loosen our spending after the college is all paid for’ phase, still sorting out how much we actually need to spend and what our other wants are and how much they would cost.
I had decided whole life was more expensive than SBP as expected. But now as I feel full retirement from medicine for me is unwise in my 50s, maybe term life on him until I expect I’d definitely retire would ease that concern. Will look at THOSE rates. We had let our term life lapse once we’d retired and the kids’ college funds were adequate (and the rates started rising).
Jenn, we are a military family too. My plan is not to take the SBP. I plan on having enough by age 54ish to retire mainly on the military pension, with enough saved in retirement accounts based on the 4% rule to be “self insured” for my wife if I die early. Also plan to have the house paid off by then or soon after. Don’t forget to throw in social security for surviving spouse too. I want to avoid the SBP tax by saving enough in other investments, if I live a long life, the legacy money will be unused investments we did not need.
For those not following, SBP in this case is not spontaneous bacterial peritonitis, but the survivor benefit plan, basically structuring the pension so your spouse still gets it if you die first.
Like most things in life, the answer likely lies in the middle, or with moderation. By funding (most of) the essential expenses with safety-first guaranteed income sources, like a single premium insurance annuity, the remaining discretionary expenses can be served with probability-based portfolio solutions. The 4% rule, or any SWR plan, basically gives a probability not of “success”, but of probability of NOT changing your plans. Modest spending changes if the plan probability dips due to portfolio/life events can reset the odds.
Most people will likely prefer greater certainty in their spending/lifestyle in retirement, especially as they get older, so narrowing the range of wealth outcomes, through transferring more assets to guaranteed income, is likely the optimal choice. Bifurcating your wealth into 2 pools – one to meet your lifestyle goals, the other, if your assets are sufficient, to legacy goals makes a lot of sense, akin to what Wall Street Physician ascribes.
Still, Pfau was done such excellent work in this field that the hope is the advisory industry quickly adopts it and more have this blended approach presented and implemented for societal benefit.
Michael Kitces has also said that idea of “likelihood of failure” in the probability models of retirement spending should instead be called “likelihood of changing one’s spending.” I’ve not even heard by hearsay of anyone who follows a 4% (or any other percentage) withdrawal strategy with no regard to the performance of their portfolio. So in reality, strategies like the “4% rule” are just a starting point for most people, but virtually everyone will reduce their spending at least temporarily if/when their portfolio suffers due to poor performance.
Count me among the more Probability based retirees. With an anticipated withdrawal rate down around 3%, the success rate is close enough to 100% for me. I’ll hold about 5 years of expenses in bonds and remain invested in equities with the rest.
Best,
-PoF
I am a Boglehead who is retired with a 100% equity portfolio
Bogleheads call me “crazy, naive, accepting an irrational level of risk, etc”
Plau and Dahle call me “Probability Based”……….I like that better…..Gordon
I call the bogleheads saving for a 2.5% withdrawal rate crazy. Maybe I should call them safety first instead.
Nope. That’s still crazy. I mean, if that’s all you WANT to spend, because you’d rather leave a huge legacy than spend it yourself, no big deal. But to only spend 2.5% because you’re afraid you’ll run out of money–those folks need to buy some SPIAs for required income and then bump up the withdrawal rate of what’s left for discretionary expenses.
At times in the recent past, Wade Pfau himself has actually called for a withdrawal rate below 3%. His recommendations are believed by most to be on the very conservative side of the spectrum.
Most who write on retirement are not themselves retired. Accordingly there are volumes of theories on spending in retirement which often are not practical. Wade Pfau’s probability vs safety first analysis is a good summary of these theories.
Once retired, when the piggy bank is set, the process of planning in retirement should, and can be, much simpler; this is particularly true for followers of this blog, most of whom will be HNWI. Retired individuals are in a position to see their actual spending patterns and match these up with retirement income. Most followers of this blog are not concerned with becoming destitute, but rather with knowing how much they should be spending each year, and the consequence of various possible outcomes for themselves and their heirs. The 90% probability thing becomes irrelevant. Rather, the concern is how will my retirement spending and bequests be affected if annual retirement income falls at different points within a range of possibilities.
What is needed is a good knowledge of actual annual spending patterns (no need to make the often meaningless distinction between discretionary and essential). Then the individual can use a comprehensive spreadsheet that will determine the amount available to spend each year under various scenarios, including reasonable worst case scenarios. These analyses are continually updated, and retirement planning becomes dynamic rather than static. Using this approach the retiree can decide for himself how he would deal with these scenarios in the context of his own situation, and revise investing and spending as deemed appropriate. There are no formulas that work.
I disagree that the piggy bank is set at retirement. It can easily grow (or shrink) depending on market performance.
I really like Dr. Berstein’s approach. Make sure you have a safe income stream for necessities and then comfortably be able to take some risk with the rest. Well, it turns out Social Security of $30k/yr per person is plenty to cover our basic necessities and then some. Meaning property tax, home maintenance, utilities, food, health insurance premiums and out of pocket expenses, and car vs taxi expenses. That is all in our oversized doctor house which may be downgraded as we need less room. If we find our fixed expenses to be higher, we can always purchase a small annuity to guarantee a more comfortable living. Everything else can be easily invested as per the 4% withdrawal rate. If markets go sour for a few years we should be more than capable of only doing 3 trips those few years as opposed to 4 or 5. Maybe staying in less fancy hotels and eating out a little less. Maybe delaying purchasing a new vehicle for another year or two.
I understand the hesitation or fear of the low 90 percent success rate. If you love your job and don’t want or need to retire then there is no issue with making a little more money to purchase your piece of mind. If you are getting tired, the documentation and regulatory burden is wearing on you and you are on the brinks of burnout, a few more years of working unnecessary wasting your life away, suffering, is just not worth it when all you need to do is cut expenses when times are rough. After all, historically the 4% rule left most people with way more money then they started with.
I think this post is a good start to get people in the frame of mind about retirement spending, but I found the “balance sheet” to be somewhat odd as it mostly contains income statement items. Likely this oversimplification will lead to confusion and errors in planning.
Unfortunately what is necessary to do this correctly will require a bit of work, i.e. a three statement financial model with (i) actual assets and actual liabilities on the balance sheet, (ii) an income statement with actual income, expenses, and appreciation/ depreciation of investments, and (iii) a cash flow statement backing out non-cash items like appreciation on investments and taking into account things like the cash generated (used) from sale (purchase) of investments, issuance or pay down of debt, etc.
Simple sensitivities around spending and market returns can be run to see “safety” or “cushion” in retirement . There really is no way to know what a person’s financial condition will look like if there is a 50% market downturn when the person is 80 unless it is modeled out.
Awesome post. I thoroughly enjoy your blog!
Probability-based here. Although I’m also schizophrenic, planning a very low withdrawal rate of ~2% (at least to start), a few years of cash for short-term spending needs, and a very aggressive equity portfolio (the majority of my portfolio is in emerging market index funds right now). So a very conservative withdrawal rate and some cash buffer to mitigate risk of a very aggressive portfolio in order to maximize investment gains. Mostly I do this because my spending desires are low compared to my investments and I enjoy investing as a hobby.
Counter intuitively, using a very conservative withdrawal rate allows you to manage the risk of an aggressive investment portfolio, maximizing your potential investment returns but still protecting you from a personal financial disaster. Mostly likely a decade from now I’ll be trying to figure out what to do with all the extra money I have, while still enjoying life nicely on the spending level I have. I’ll be okay with this “problem” if it occurs.
I think that problem is very likely with a 2% withdrawal rate.
What’s your reasoning for having a majority of your portfolio in emerging market index funds?
If you’re interested, I go into the details here:
https://saveinvestbecomefree.com/2016/11/02/extreme-rebalancing/#more-1502
Probability based here also. Safety firsters will want to buy SPIAs. I enjoy things like asset allocation and SWRs so I have dialed back my equity percentage and increased some short term holdings and am hoping for the best. I think SPIAs make sense for some people.
1. Mental acuity issues.
2. History of buying into get rich quick schemes.
3. history of panicking with corrections or bears.
4. Planning on traveling extensively to remote areas with no internet access.
No matter what since these decisions are related to the future, your assumptions are critical. In addition your risk tolerance is important. Now I am very intolerant of risk, don’t desire an expensive life style so I spend less than my income each year. I have reserves for some expected future occurrences and if unexpected events happen I can liquidate some capital. This is very conservative with a very low probability of running out of money.
Understand as best you can conditions, yourself, and your assumptions then decide.
May I add another “L” to the Pfau list of goals? . Lots of sleep.
Seriously, I like how he captured the four areas.
We’ll be starting off at 2.5% WR but know with costs of teenage kids, uncertainty of healthcare costs and perhaps some funding required to supplement existing nicely funded college funds, that WR can move up. Throw in some more EU travel splurging and we could quickly get to 3-3.25%. Our low WR also reflects another favorite topic of Mr. Pfau – sequence of returns.
We have another year before pulling the plug and no doubt our plan will get tweaked further, yet still stay on the end of conservative.
Great post, I really enjoyed reading this.
If you have enough assets that 2.5% even seems doable, you’re going to be just fine. I swear the comments section on this post is like the Superbowl of Physician Finance. It’s like a race to the largest portfolio, the maximum frugality, and the lowest possible withdrawal rate. What’s next, someone invested 100% in individual TIPS who only needs 1% of his portfolio a year? Moderation in all things folks.
At least you recognize your WR, even the 3.25% one that includes all that college and European travel, is STILL a low rate.
For 1900-2015, using a 50:50 portfolio:
“With a 50/50 asset allocation, the 4% rule did not survive in any country, though it came very close in the U.S. (3.94%) and Canada (3.96%). Even allowing for a 10% failure rate, 4% made the cut only in Canada, the U.S., New Zealand, and Denmark.
In eleven of twenty countries, the SAFEMAX fell below 3%. World War II-era Japan, in particular, faced the sort of crisis suggested by William Bernstein, as the SAFEMAX was only 0.27% for 1937 retirees. The 4% rule would have supported expenditures for only three years.Hyperinflation in Austria led a hypothetical retiree at the start of World War I to only sustain a 0.07% withdrawal rate from their portfolio. Shockingly, the 4% rule would have failed more than half of the time for countries including Spain, Germany, France, Italy, and Austria. Italians attempting to use the 4% rule in their domestic financial markets would have actually faced failure in 76% of the historical periods.”
This is from Pfau: https://retirementresearcher.com/4-rule-work-around-world/
The US may end up at the bottom of world return charts in the future, rather than near the top.
A withdrawal of 2.5% should be safe, but then again it apparently failed in 8/22 developed countries over the last 115 years (see exhibit in link above).
CM-
Here’s the problem with both your comment and the data you cite:
You are using a probability-based model while looking for the certainty available only from the safety-first model. If you need absolute certainty that you won’t run out of money, i.e. a “SAFEMAX”, then use some SPIAs and reverse mortgages and the like. If you can roll with the punches, then you’re likely to be able to spend more/leave more by sticking with stocks and real estate. Actually using nothing but a 4% rule or a 2.5% rule or whatever no matter what happens over your investing career as your retirement spending plan is pretty foolish given the alternatives.
After the Bengen study, many retirees and wannabe early retirees have assumed that a 4% withdrawal rate is essentially foolproof. Beyond that, you’ve implied on a number of occasions that it’s ridiculous to use a 2.5% SWR.
In my opinion, that’s a bridge way too far, for two reasons:
1. Data extended from 1900 to 2015, and to other developed markets, shows that Bengen would have published a much lower number (much less than 2.5% in fact) for his SWR if he had used the broader data set, and
2. The failure rate is based on the historical data set which includes all of the periods of low and average stock and bond valuations, but the germane data set would include only periods with stock and bond valuations comparable to the present.
Whether it’s better to manage retirement with SPIAs and reverse mortgages (maybe so) than with an inflation-adjusted SWR from a stock/bond portfolio is a different topic.
My point is that a 2.5% “SWR” from a stock/bond portfolio isn’t safe on the basis of history, and it is even less safe given current stock and bond valuations.
CM-
The world isn’t safe. Stop looking for safety where it doesn’t exist. The question to ask isn’t “what withdrawal rate is safe?” The question is “What is the most reasonable way to fund my retirement?” Start from there and you’ll end in the right place. I really don’t care what safe withdrawal rate you, Pfau, and Bengen can put together because I don’t plan to use it as my retirement spending plan and if you’re smart, you won’t either. The Trinity study’s purpose was to show that 6, 8, 10% wasn’t safe, not to dictate whether your retirement plan should be a blind withdrawal 2.5%, 3.5%, or 4.5% method.
But hey, you want to spend 2% of your portfolio, knock yourself out. It really doesn’t bother me one bit. But bear this in mind. If you buy 30 year TIPS paying 0%, a 2% withdrawal rate gives you a portfolio guaranteed to keep up with inflation for 50 years. 50 years. At 0% real. Roll those numbers around in your brain for a bit and decide whether you think that’s a reasonable approach for you personally to take with your money. Then consider that 30 year TIPS are paying almost 1%. So that approach would get you even more than 50 years. Guaranteed.
I’m not saying 4% is foolproof. I’m saying it’s reasonable, especially for someone smart enough to adjust as they go along, especially if they have some guaranteed income, which is a far better strategy than a blind withdrawal every year no matter what is happening to the portfolio.
Right on target WCI. Those who are seriously considering a starting withdrawal rate under 3% would very likely be well served by amortizing enough of their assets (i.e. SPIAs, bond ladders, reverse mortgages) to cover their necessary expenses, a strategy that even Wade Pfau recommends for many.
Using data going back to 1970 (due to availability), a 60/40 portfolio comprised of strictly non-U.S. global equities and intermediate term treasuries had a safe withdrawal rate over 30 years of 4.8%. So the idea that someone who starts with 4% and then adjusts as the portfolio allows/requires seems well supported by the data.
Hahahaha!
“World War II-era Japan, in particular, faced the sort of crisis suggested by William Bernstein, as the SAFEMAX was only 0.27% for 1937 retirees.”
Yeah, if you retired in 1937 in Japan, you didn’t end up playing golf 5 days a week while jet setting to Europe a couple of times a year. Maybe we could specifically look at retirees in the area of Nagasaki circa 1944 and what their SAFEMAX was.
Yes, if global total war occurs in 2025, or Best Korea sparks a nuclear exchange with the US New Year 2020, anyone reading this site will probably be glad to have retired in 2018 and spent time with family and friends, as SWR and asset allocation will likely be relegated to ‘first world problems’ with issues like ‘access to clean water’ and ‘fuel for staying warm in winter’ taking precedence.
And as many readers are physicians, odds are likely that a ruptured aneurysm small collection of platelets in the wrong place, or random malignancy will ensure that our ‘SAFEMAX’ was far too conservative and we’ll leave money on the table at death that we could have enjoyed spending (or not working to earn) in life.
TIPS ladder and 1% WR. LOL.
Moderation in everything, including moderation.! A rule of living by the great scientist Arnold Beckman.
We have been told that we are too conservative although CM below would beg to differ. Early Retirement Now did a fantastic set of posts on the SWR, including equities / bond ratios and SS impact among other factors. SWR gold IMHO for FIRE planners.
Thanks for the shout-out, Mr. PIE! The series with currently 16 parts starts here:
https://earlyretirementnow.com/2016/12/07/the-ultimate-guide-to-safe-withdrawal-rates-part-1-intro/
I know very few instances of facing substantial (or even not that substantial) risks where I would accept a 10% failure rate:
Going to the airport, would I cut it so short that I might miss my flight with a 10% probability? Probably not!
As the gas station, would I accept a 10% risk of the pump setting my car on fire? Certainly not.
In retirement, would I accept a 10% risk of running out of money at age 70? Certainly not.
That said, a completely blind and fixed withdrawal rule, of course, is only the starting point. Only to be used to roughly gauge the starting SWR. I agree that a dynamic rule is the way to go. It will never run out of money but might suffer a temporary drop in purchasing power. But the “temporary” could mean decade-long droughts with very significant reductions in withdrawals (see part 11 in my series). Sometimes I hear in the FIRE blog world that all we need is a little flexibility, maybe delay the inflation adjustment for a few years, or get a side gig (at age 70?) to deal with adverse returns. That might not work as well as people think.
Why would you have a 10% failure rate? Because you’re not going to adjust your withdrawals a bit if you hit a period similar to the 1930s? Or stagflation? Or 2008? No adjustment at all? Really? If that’s so, you should lean more toward the safety first approach, buy a few SPIAs and TIPS ladders etc.
But when I think about it, a 90% chance of success, a likely outcome of dying with 2.7X what I started with, enough guaranteed income to live comfortably, and the ability to adjust as I go? Yea, I’ll take that, especially when the alternative is going on half as many trips during retirement and then dying with 3 times the money.
You are making exactly my point. Because I don’t like the idea of fixing one withdrawal amount today and sticking with it I’d prefer the dynamic rule. I’d be crazy to not adjust if returns are poor and I’d be crazy to not increase withdrawals if stocks keep up their rally.
But: my extensive experience with simulating both static and dynamic withdrawal rules is the following: A dumb static rule might have a 10% chance of completely running out of money. A dynamic rule like Guyton-Klinger never runs out of money but has a 10% probability of really nasty, decade-long reductions of withdrawals by 40%, even 50%+, see part 10:
https://earlyretirementnow.com/2017/02/15/the-ultimate-guide-to-safe-withdrawal-rates-part-10-guyton-klinger/
Also: don’t get carried away with the average and median final values computed by Trinity, Kitces etc. They average over all retirement starting points, including starting dates with much more attractive equity valuations. Have they simulated what’s the median portfolio value after 30+ years conditional on an initial CAPE of 30? I did!
Being pessimistic is seductive because it sounds so smart, but it is the optimists that have triumphed so far. Some of the future return projections that the pessimists come up with are ridiculous. If I really thought I was only going to get 1% real out of equities over my investment horizon, I would invest in something else like real estate or other small businesses. Consider that you are surrounded by real estate properties selling at cap rates of 4-6. So if you really don’t think you can get real returns in that range from equities over your time horizon, it might be time to change your investment strategy. “But bonds only yield 2% these days so my SWR is only 2.5%,” say the pessimists. Yet somehow I keep finding hard money loans at 8.5-11%. Every week. Are they riskier than treasuries? Sure. But they’re backed by real property that can be foreclosed on.
https://www.whitecoatinvestor.com/making-different-choices-due-to-low-expected-returns/
At any rate, I’m not talking about a dynamic rule of any kind. I’m talking about the Taylor Larimore philosophy of just keeping an eye on your portfolio balance. In 2008-2009, you tighten your belt and spend less. After markets recover, then you go on the European vacation. People are amazingly adaptable to changes in their financial circumstances. We do this with our earned income throughout our lives but those who are trying to come up with withdrawal rules seems to think we instantly lose the ability to do this on the eve of our retirement. We don’t. Particularly for a high earner/strong saver who is likely to require a very low withdrawal rate (or perhaps no withdrawal at all between Social Security, pensions, and a SPIA) to meet basic expenses.
I mean if you really expect 1% real out of your equities and -1% real out of your fixed income, all the more reason to annuitize a big chunk of your nest egg a la safety first. Let the insurance companies run those risks you’re so worried about.
WCI You are right that dynamic rules don’t work. You are wrong that the Taylor Larimore wing-it philosophy is the way to go. The right answer is to plan as you go, as Taylor does, but to have adequate continually updated information to make the right decisions on investment and spending. This stuff is really simple and all the dynamic rules and SWR rules are nonsense.
I think “winging it” carries a negative connotation that probably isn’t warranted. I prefer the term “being flexible using adequate continually updated information to make the right decisions on investment and spending.” Looks like you do too.
Yes, but retirees still get hung up on all the rules and percentages discussed in this thread. Simply knowing personal spending patterns and future cash flows under various scenarios, is all that is needed for any high net worth retiree.
Who said equity returns will be 1% real? I certainly didn’t. Jack Bogle used 2% (4% minus 2% inflation), which I found a bit too conservative. Personally, I am using ~3% real for equities for the next 10 years, while the CAPE normalizes back to the low 20s and then the long-term returns after that. But as I stated elsewhere, retirement plans don’t usually fail because average returns are low but because of sequence risk.
I like real estate, too. A cap rate of 4-6 (after all the costs, depreciation, management fees etc.) is possible. I invest in multi-family RE through a few private equity funds. They even get to the upper end of that range, thanks to leverage. I doubt that a 4-6 cap rate in this very illiquid, non-transparent, idiosyncratic and unregulated market would justify a similar real rate of return for equities, though. In fact, after applying 2-3% haircut to my real estate expected returns I land right back where I was with my equity expected returns.
Guyton-Klinger has very nasty and long-lasting spending reductions. 40-50%, sometimes 60%. For decades. CAPE-based rules are a little better, but can’t avoid long-lasting reductions in spending either. I’m doubtful that any kind of magic Taylor Larimore “just keeping an eye on my portfolio” approach can change that arithmetic. Quite the opposite, until anybody provides me with some hard evidence and simulations I’d surmise that “just keeping an eye on my portfolio” is exactly what Guyton-Klinger does. But I’m open for discussion. If someone can show me a simple, intuitive, quantitative and replicable dynamic spending rule a la Larimore that a) has lower spending reductions and b) shorter spending reductions than the other quantifiable and replicable rules, I’m all ears. Until then I’d maintain that a dynamic rule has a roughly10% failure rate too: Tightening the belt with a 50% spending reduction and delaying that European vacation for 20 years is something that actually happened in the past with that rule.
ERN, WADR, the problem is that there is a world of difference between those planning for retirement and those already retired. There is some minimal validity to SWR and various other “rules” to assist in conceptual planning prior to retirement. Once retired, with “enough”, all of those rules are meaningless. It is as important to spend the right amount, not too little or too much, and all of the rules, percentages and techniques being discussed in this thread are meaningless. Unfortunately, most of the advice on this thread is theoretical, and coming from those who are not yet retired.
That’s life. If you work at a job that pays $30K you don’t vacation in Europe. If what you’re comfortable taking out of your portfolio due to poor recent sequence of returns is only $30K, you don’t go to Europe either. If you want guarantees, seek them from guaranteed income sources, not ever more complicated withdrawal rules.
So no, I’m not going to come up with a rule because the issue I have with all this retirement planning is using a rule in the first place. What I’m saying is be flexible. You’re probably going to be able to do just fine spending something around 4% of your stash. But if sequence of returns risk shows up early in your retirement, you need a plan for how to deal with that. i.e. spend less for a few years or put more of your nest egg into guaranteed sources of income. I don’t need a rule to tell me to spend less. I just delay buying that new car a year, or I buy a cheaper car, or I vacation in Wyoming instead of France or I go out to eat once a month instead of once a week.
I think people spend too much time examining rules and playing with FireCalc and the like trying to impose a precision on the process that you simply cannot have. Should people be familiar with the data and what rules worked in the past and in what circumstances they failed? Sure. Should they spend a little time with some calculators? Sure. And then they should live life, adjusting to their circumstances as they go.
WCI, we are sort of in agreement, but you still suggest a fixed SWR, say 4%, and then being flexible in a downturn. In the real world of retirement for HNW, it is also necessary to make sure you are not spending too little. Spending includes gifts. That is why a detailed analysis of annual available spend, under various scenarios, in the context of actual spending is necessary.
The only “fixed” I suggest is starting somewhere reasonable. If you start in the 4% range, that’s reasonable. If you start at 10% or 1%, that’s not reasonable.
The 4% rule is more useful to tell you how large your nest egg ought to be before retiring than it is as an actual withdrawal plan.
Last I checked, there is no disaster that occurs if you spend too little. What disaster are you thinking of? A missed opportunity? Paying too much estate tax? Interesting use of the term “necessary.”
It is necessary for HNW to determine if they are gifting enough to kids, grandkids’ 529s, charity, etc. It’s not about financial rules, it is about real life and real people once retired — who needs what and when. Incidentally, estate planning also comes into play.
I just find it odd to see the words “necessary” and “gift” in the same sentence.
It’s not necessary to make gifts. It is necessary to determine your financial capacity to do so, if you so choose.
Right, but that all comes out of the same pot as spending. If poor returns show up early in your retirement, you spend/gift less.
It’s like I’m talking to a bunch of engineers here. 🙂 Engineers spend their careers crafting solutions to overcome future uncertainty. They build bridges twice as strong as they need to, but not four times as strong for instance. But in some things, you simply cannot eliminate uncertainty. It is an inherent part of the system. So you deal with it as best you can. I have a lot of uncertainty in my day job. I discharge lots of people from the ER without a diagnosis for their concerning symptoms, but with a follow-up plan and some careful return precautions. We’ve lowered their risks, but not eliminated them completely. Investing and retirement planning are like that. You cannot eliminate the uncertainty completely. So you do something reasonable that is likely to work and make a reasonable contingency plan for that small percentage of the time when it doesn’t. You muddle through.
It is often difficult to understand the perspective of others. As a retiree, my perspective on financial planning for retirement is totally different than it was pre-retirement. The difference is that I know what my views are on actual spending. I know how much I can spend under different scenarios, including reasonable worst case, and plan accordingly. From my perspective, all of the rules, guidelines, percentages, etc. that are being bandied about here are meaningless. Welcome to the real world.
Why didn’t you know how much you spent and your views on it before you retired? I know how much I spend and what my views are on it. While we all grow and change and mature as we go through life, I think in this era of squishy retirement that you’re drawing a hard line where none exists. It’s like these guys worried about how they’re going to pay for health care in retirement. I don’t worry about that because I’m paying for health care now. I know exactly what it costs.
If you now know how you want to prioritize all of your spending when you retire, your world is very different than mine. My approach is anything but a hard line. My views on my actual spending are changing and they must be correlated with my current annual available spend analysis. The key is to have all the information available and to be flexible with investing and spending. Decisions are simple when all relevant information is available.
Unfortunate that there is such a change in how you spend your money upon retiring. If I retired tomorrow, I expect I would spend my money exactly the way I’m spending it today. So my best guideline for how I will spend in retirement is how I am spending right now.
But it’s hard to disagree with the notion that “decisions are simple when all relevant information is available.” Unfortunately, I think a lot of relevant information in determining how much one can spend in retirement is unknowable without a crystal ball.
Agree regarding earlyretirementnow. Kudos.
Of course you can guess on retirement spending based upon today’s spending, but in the real world things change considerably over the retirement years, e.g. six new grandkids, resulting in spending changes. Once retired, with known investments and other income sources, it is possible to estimate annual future cash flows under various scenarios, and plan spending depending on whatever scenario life throws at you. The analysis is continually updated. Easy peasy.
John,
I do not understand why having a 529 changes anything. A 529 is no different than antithetical expense. If you have enough money, you find it. If you don’t, guess what either eat rice and beans or don’t fund the 529s. You make that choice.
The 529 is not a mandatory expense you know.
The key is how you “find it”. With proper analysis of actual spending and cash flow scenarios, you may well determine that there is no problem with rice and beans, but best to back off the caviar and champagne in order to fund the 529s.
Engineer here, not a doctor but have been retired 3+ years now (age 59). Have to agree with WCI, spending habits remain largely the same in retirement. After a flurry of international travel the first five months we have only been on occasional family vacations, and still on the same budget as while working, including a “savings” category. Have to say all the financial planning and “what-if” scenarios have boiled down to a simplistic approach: Make sure you spend less than your guaranteed income on ordinary living expenses, the rest is gravy.
I assume by guaranteed income you mean pensions or annuities. You are fortunate to have all of your ordinary living expenses covered by that income. Actually, I was also in that position 3 years out of retirement, and my thinking pretty much aligned with yours. After that, however, came six grandkids and other changes in lifestyle. Among other things I needed to decide about six 529 funds and amounts to contribute to each. For many people that (plus other inevitable changes) can eat up all the gravy. That is when detailed “what if” scenarios, of the type I described above, come in very handy.
Funding 529 accounts for six grandkids IS gravy — an unnecessary extravagance that is fine for you to spend your money on if you can afford it and it brings you happiness. Yes, it means you can’t afford other things like more travel spending or a nicer car, but, hopefully, you’ve decided to deploy your excess money in this way because you derive more pleasure from knowing that your grandkids’ education is covered than you would from taking another trip to Europe or driving around in a Ferrari. I probably would too.
First, let me say I find your blog very well written and informative, even though I am not a physician. My daughter is a 3rd year medical student, and I have directed her to your blog and given her your book as a Christmas present with the advice to at least read the chapter headings so that, when an issue like repaying loans comes up, she can read an evaluation from someone who has her back and understands her situation.
Second, I am in my third month of retirement. I will 67 in a few months with a small pension and will otherwise be living completely off my accumulated portfolio for three years until my wife and I both take Social Security. At that point, I am estimating that Social Security and the pension will provide my base income for our modest needs and my investment portfolio will provide for contingencies like long term care or wants like trips, etc.
Right now I have 20/45/35 cash/bonds/stocks portfolio. I will mostly spend down the cash over these three years to preserve investments. Most people would say this is a fairly conservative portfolio. At his point though, since I am spending such a large chunk of my portfolio, I want to protect from sequence of returns risk, especially in an environment when the market is overvalued and we have had a very long bull market.
So, as SS claiming looms, I will adjust my equity allocation based on current circumstances, hopefully upwards, as Pfau and Kitces have suggested. This just seems to me a reasonable way to manage risk.
All my adult life I have played strategy games that have an element of skill but also an element of luck or risk. Many people would say that the people who win these games are “lucky” but in fact the same players tend to win a game much more frequently than others (like me, for instance!). Even backgammon, which has a high number of dice rolls, has players who consistently win. The reason is the best players MINIMIZE the number of risks they have to take in order to win. They know when to take a chance and when to “play the odds”. Applying that logic to investing, most people cannot avoid taking some risk, they just need to minimize that in relation to their objectives. Hopefully that is what I am doing with my plan.
Now that I am in retirement, I am trying to make my focus not so much on all the bad things that can happen to the money I have saved (though it can), but rather all of the benefits I can reap from taking some risks in my investments at the appropriate amount and time. It is a good thing to have some fear, but I think your life will be more fulfilling if we can look ahead to what we can gain and enjoy the fruits of that. If that doesn’t come to pass, I am at ease with living on less.
Thanks for spreading the message!
When you retire and and ALL your income is derived from SS and investments, MOST like myself go very conservative a la marginal utility of wealth theme; look it up with swedroe’s name
I just finished Pfau’s book safety first. It made me buy his probability book. I love reading his writing. I think I fall more on the probability side until you reach retirement age and then when you begin to derisk maybe supplement some immediate annuities using the safety-first approach. I think I will probably read his reverse mortgage book too.
Thanks, WCI, for a good article with some stimulating discussion in the comments. Thanks also for all you do and the contributions you make to all of our financial educations. Much appreciated.
One comment I wanted to make is that Bengen’s study which resulted in the so-called “4% rule” suggested a very low probability of running out of money over a 30 year time horizon. For people who are retiring in their 50s or even younger, they probably have to think in terms of a 40-50 year horizon, with more conservative withdrawal rates. Those retiring very early are likely to have low guaranteed sources of income from SS and pensions, making conservative assumptions even more important.
I ran some numbers in PortfolioVisualizer using Monte Carlo Simulation with a 70/30 portfolio (70% SPY, 30% BND) and the worst 2 years up front (to somewhat account for sequence of returns risk):
– Using a 4% withdrawal rate (adjusted for inflation), it showed a 91% success rate over a 30-year horizon but only an 83.5% success rate for 40 years, a 78% success rate over 50 years, and a 74% success rate over 60 years.
– Using a 3.5% withdrawal rate, the success rates are: 97% for 30 years, 93% for 40 years, 91% for 50 years, and 89% for 60 years.
– Using a 3.0% withdrawal rate, the success rates are: 99% for 30 years, 98% for 40 years, 97% for 50 years, and 96% for 60 years.
Things look a lot worse if you assume a worse sequence of returns risk. With the worst 5 years up front, a 4% withdrawal rate had a 53% success rate over 30 years, 30% success rate over 40 years, 16% success rate over 50 years, and 9% over 60 years. The corresponding numbers for a 3.5% withdrawal rate are: 67% for 30 years, 42% for 40 years, 27% for 50 years, and 16% for 60 years. The corresponding numbers for a 3.0% withdrawal rate are: 80% for 30 years, 59% for 40 years, 41% for 50 years, and 28% for 60 years. I think that target withdrawal rates need to be adjusted for the planning horizon, as well as at least some moderate effect of sequence of returns risk.
And if the research into anti-aging and extending health spans is even moderately successful before we get too old to benefit from it, we may have to think about financial plans that will support us until the age of 100-110 (or possibly even older). Additionally, for those couples with more than a couple years age gap between the spouses, a longer planning horizon may be required. I want a financial plan that has a higher than 90% success rate over a 40-year planning horizon assuming the first 2-3 bad years up front, since I want a low probability that I am going to have to significantly reduce my spending. For me, a 4% target withdrawal rate is a bit too high because it produces unacceptable failure rates over 40+ years. From my point of view, it is worth being a bit conservative in the financial assumptions to ensure that we can still travel quite a bit internationally until we are 80+, be generous with our resources, and that we aren’t forced into a miserly existence if we do live to 100+. This is a rather skewed, one-sided risk distribution, with virtually no downside to having too much money and a very large downside to running out of money.
As I get older and later in my career, this topic interests me more and more.
Since I recently gave a talk called “Mini-MBA” at a recent awesome national conference (WCICON22) I feel compelled to clarify one issue.
In retirement planning, they talk about “liability matching” when they talk of the timing of expenses. That makes sense.
But do realize that an actual Balance Sheet is a financial statement and there are no expenses listed on that. Expenses show up on the Income Statement.
There are certainly links between the two statements though since assets can produce income and liabilities can produce expenses. Both of those flow to the Balance Sheet.
It’s funny. I was thinking of replying, but then realized it was a post from 2017 where I previously made several comments. I said everything then that I would say now. People actually retired do not need the Pfau rules and formulas. There is a much better wat to monitor retirement planning through an Annual Available Spend spreadsheet. I wont repeat my prior comments.