Bengen Floor & Ceiling Withdrawal Method

Discuss all general (i.e. non-personal) investing questions and issues, investing news, and theory.
Post Reply
Topic Author
sixtyforty
Posts: 330
Joined: Tue Nov 25, 2014 12:22 pm
Location: USA

Bengen Floor & Ceiling Withdrawal Method

Post by sixtyforty » Sun Apr 07, 2019 5:31 pm

I've been playing around with Tyler's excellent tool (see link below). From my findings, it appears that the Bengen Floor & Ceiling allows higher SWR with a 50/50 portfolio than withdrawal methods from Kitces Ratchet, Guyton-Klinger or the Clyatt's 95% rule. I really like the simplicity and the fact that Bengen was the developer of the original 4% rule. While the method enforces a 15% max reduction, it enables you to determine (1) If the reduction would support your current living expenses and (2) if not, what you would need to cut. It looks like a pretty good strategy and not sure why we don't hear more about it on this forum. Thoughts ?

BTW... Anyone know of supporting documentation that shows back-test using this method prior to 1970 ?

https://portfoliocharts.com/portfolio/r ... -spending/
"Simplicity is the ultimate sophistication" - Leonardo Da Vinci

User avatar
Tyler9000
Posts: 423
Joined: Fri Aug 21, 2015 11:57 am

Re: Bengen Floor & Ceiling Withdrawal Method

Post by Tyler9000 » Sun Apr 07, 2019 9:24 pm

sixtyforty wrote:
Sun Apr 07, 2019 5:31 pm
BTW... Anyone know of supporting documentation that shows back-test using this method prior to 1970 ?

Here's Bengen's paper on the Floor & Ceiling withdrawal method and Wade Pfau's summary. Both use supporting data for reasonably balanced stock/bond portfolios since 1926.

https://finalytiq.co.uk/wp-content/uplo ... art-IV.pdf

https://retirementresearcher.com/floor- ... -strategy/

The Pfau charts should look familiar, as they use the same format as the Portfolio Charts Retirement Spending tool. And one thing I've always respected about William Bengen is his continued research beyond the original assumptions used in his famous paper on safe withdrawal rates. I completely agree that many of his other ideas like the Floor & Ceiling withdrawal method deserve more attention, so I'm happy to see you start this discussion.

AlohaJoe
Posts: 4556
Joined: Mon Nov 26, 2007 2:00 pm
Location: Saigon, Vietnam

Re: Bengen Floor & Ceiling Withdrawal Method

Post by AlohaJoe » Sun Apr 07, 2019 9:36 pm

sixtyforty wrote:
Sun Apr 07, 2019 5:31 pm
It looks like a pretty good strategy and not sure why we don't hear more about it on this forum.
You don't hear about it for the same reason you don't hear more about Gummi's "Sensible Withdrawals" or Ingliss's "Feel Free Withdrawals" or Stout & Mitchell's "Model 3" or Pye's "Retrenchment Rule" or Blanchett's "Simple Formula" or Siegel & Waring's "ARVA".

Be honest: have you heard of any of those?

I think most people who have looked hard at withdrawal strategies have a natural aversion to anything that starts with "adjust by inflation every year" as a starting point. The hard-floor also makes people nervous. Ceilings are also not useful, so having one is usually pointless.

In the metrics I've seen Bengen's Floor-and-Ceiling is middle of the road. Other strategies do better, so it gets lost in the crowd of other also-ran strategies.

Topic Author
sixtyforty
Posts: 330
Joined: Tue Nov 25, 2014 12:22 pm
Location: USA

Re: Bengen Floor & Ceiling Withdrawal Method

Post by sixtyforty » Mon Apr 08, 2019 4:16 am

AlohaJoe wrote:
Sun Apr 07, 2019 9:36 pm
sixtyforty wrote:
Sun Apr 07, 2019 5:31 pm
It looks like a pretty good strategy and not sure why we don't hear more about it on this forum.
You don't hear about it for the same reason you don't hear more about Gummi's "Sensible Withdrawals" or Ingliss's "Feel Free Withdrawals" or Stout & Mitchell's "Model 3" or Pye's "Retrenchment Rule" or Blanchett's "Simple Formula" or Siegel & Waring's "ARVA".

Be honest: have you heard of any of those?

I think most people who have looked hard at withdrawal strategies have a natural aversion to anything that starts with "adjust by inflation every year" as a starting point. The hard-floor also makes people nervous. Ceilings are also not useful, so having one is usually pointless.

In the metrics I've seen Bengen's Floor-and-Ceiling is middle of the road. Other strategies do better, so it gets lost in the crowd of other also-ran strategies.
The Bengen Floor & Ceiling doesn't adjust for inflation, it's simply a constant percentage with guard rails. You mention 'Other strategies do better'.. which ones would those be ?
"Simplicity is the ultimate sophistication" - Leonardo Da Vinci

AlohaJoe
Posts: 4556
Joined: Mon Nov 26, 2007 2:00 pm
Location: Saigon, Vietnam

Re: Bengen Floor & Ceiling Withdrawal Method

Post by AlohaJoe » Mon Apr 08, 2019 6:20 am

sixtyforty wrote:
Mon Apr 08, 2019 4:16 am
The Bengen Floor & Ceiling doesn't adjust for inflation, it's simply a constant percentage with guard rails.
Sure it does. The floor & ceiling are adjusted for inflation every year. Bengen and others have mentioned various floors & ceilings but it appears the number Bengen used the most was a floor of 90% and a ceiling of 125%. So if you start out withdrawing 5% of the portfolio then the floor & ceiling become 4.5% inflation-adjusted and 6.25% inflation-adjusted. So with a tight floor & ceiling, as Bengen defined them, you are strongly bound by inflation-adjustments.

Now, you could argue that 5%, 90%, and 125% are all the wrong numbers. The wider you make the guardrails, the less bound you are to inflation-adjustments. But even with the numbers on PortfolioCharts we can see that it failed 4 times. And at other times it left the portfolio with way way way too much money. (In general having a ceiling has little benefit; you can see that by changing the ceiling to, say, 500% and noticing there's no impact on the 5 or so bad sequences.)

It's not a bad strategy; it often performs well and rarely performs poorly. If someone used it, or recommended it to others, I wouldn't argue with them about it.
You mention 'Other strategies do better'.. which ones would those be ?
Keep in mind that "works better" depends on a lot of personal preferences & judgment calls. Many of the differences around strategies are tradeoffs between the annual withdrawal stability and portfolio stability. That said, some of the strategies that often perform better than Bengen's Floor-to-ceiling are McClung's EM, longinvest's VPW, and just using the IRS RMD tables. You'll notice that all three of the "stronger" strategies take advantage of "how much time you've got left". The basic idea is that withdrawing 10% of your portfolio when you are 65 is reckless but withdrawing 10% of your portfolio when you are 93 is not; so take advantage of that reality when designing a strategy.

MikeG62
Posts: 1813
Joined: Tue Nov 15, 2016 3:20 pm
Location: New Jersey

Re: Bengen Floor & Ceiling Withdrawal Method

Post by MikeG62 » Mon Apr 08, 2019 6:51 am

Personally I have a bias toward WD strategies that contemplate adjustments to one's WD rate since the future is uncertain. I am familiar with (have read about) many of them, including Kitces' Ratcheting, Bengen's Floor & Ceiling and Guyton and Klinger's guardrails. I have chosen a modified version of G&K's guardrails (although I am starting with a lower initial WD rate, and therefore, different/customized lower and upper guardrails). Based upon the way I have the guardrails set, equities would have to suffer a massive decline to trigger a reduction to our WD rate. Yes I am conservative, but I also sleep pretty well at night with the plan I have developed having gone through the math (running it through MC simulations).

FWIW, we are in our forth year of retirement, so this is not something that I am thinking about down the road.
Real Knowledge Comes Only From Experience

Topic Author
sixtyforty
Posts: 330
Joined: Tue Nov 25, 2014 12:22 pm
Location: USA

Re: Bengen Floor & Ceiling Withdrawal Method

Post by sixtyforty » Mon Apr 08, 2019 12:17 pm

Tyler9000 wrote:
Sun Apr 07, 2019 9:24 pm
sixtyforty wrote:
Sun Apr 07, 2019 5:31 pm
BTW... Anyone know of supporting documentation that shows back-test using this method prior to 1970 ?

Here's Bengen's paper on the Floor & Ceiling withdrawal method and Wade Pfau's summary. Both use supporting data for reasonably balanced stock/bond portfolios since 1926.

https://finalytiq.co.uk/wp-content/uplo ... art-IV.pdf

https://retirementresearcher.com/floor- ... -strategy/

The Pfau charts should look familiar, as they use the same format as the Portfolio Charts Retirement Spending tool. And one thing I've always respected about William Bengen is his continued research beyond the original assumptions used in his famous paper on safe withdrawal rates. I completely agree that many of his other ideas like the Floor & Ceiling withdrawal method deserve more attention, so I'm happy to see you start this discussion.
Thanks Tyler. Could you confirm how your performing the calculation in the example you show for Bengen Floor to Celing ? I'm assuming it's applying the fixed % each year on the portfolio amount and then just determining if it's within the floor & ceiling (calculated with the floor and celing values on the first year). I don't see where the spending would be adjusted for inflation each year.
"Simplicity is the ultimate sophistication" - Leonardo Da Vinci

Topic Author
sixtyforty
Posts: 330
Joined: Tue Nov 25, 2014 12:22 pm
Location: USA

Re: Bengen Floor & Ceiling Withdrawal Method

Post by sixtyforty » Mon Apr 08, 2019 12:32 pm

AlohaJoe wrote:
Mon Apr 08, 2019 6:20 am
sixtyforty wrote:
Mon Apr 08, 2019 4:16 am
The Bengen Floor & Ceiling doesn't adjust for inflation, it's simply a constant percentage with guard rails.
Sure it does. The floor & ceiling are adjusted for inflation every year. Bengen and others have mentioned various floors & ceilings but it appears the number Bengen used the most was a floor of 90% and a ceiling of 125%. So if you start out withdrawing 5% of the portfolio then the floor & ceiling become 4.5% inflation-adjusted and 6.25% inflation-adjusted. So with a tight floor & ceiling, as Bengen defined them, you are strongly bound by inflation-adjustments.

Now, you could argue that 5%, 90%, and 125% are all the wrong numbers. The wider you make the guardrails, the less bound you are to inflation-adjustments. But even with the numbers on PortfolioCharts we can see that it failed 4 times. And at other times it left the portfolio with way way way too much money. (In general having a ceiling has little benefit; you can see that by changing the ceiling to, say, 500% and noticing there's no impact on the 5 or so bad sequences.)

It's not a bad strategy; it often performs well and rarely performs poorly. If someone used it, or recommended it to others, I wouldn't argue with them about it.
You mention 'Other strategies do better'.. which ones would those be ?
Keep in mind that "works better" depends on a lot of personal preferences & judgment calls. Many of the differences around strategies are tradeoffs between the annual withdrawal stability and portfolio stability. That said, some of the strategies that often perform better than Bengen's Floor-to-ceiling are McClung's EM, longinvest's VPW, and just using the IRS RMD tables. You'll notice that all three of the "stronger" strategies take advantage of "how much time you've got left". The basic idea is that withdrawing 10% of your portfolio when you are 65 is reckless but withdrawing 10% of your portfolio when you are 93 is not; so take advantage of that reality when designing a strategy.
Good feedback. Interestingly, I went back to McClungs book and to my surprise the Bengen Floor to Ceiling method was one of the strategies he tested. On page 107 he actually remarks that it is an overall strong performer. If you look at the matrix on that page, it achieves fairly high withdrawal rates, especially compared to the other strategies. It's not quite as good as his EM etc, but considering it's simplicity, not bad.
"Simplicity is the ultimate sophistication" - Leonardo Da Vinci

ThrustVectoring
Posts: 742
Joined: Wed Jul 12, 2017 2:51 pm

Re: Bengen Floor & Ceiling Withdrawal Method

Post by ThrustVectoring » Mon Apr 08, 2019 1:15 pm

AlohaJoe wrote:
Mon Apr 08, 2019 6:20 am
(In general having a ceiling has little benefit; you can see that by changing the ceiling to, say, 500% and noticing there's no impact on the 5 or so bad sequences.)
Yeah, the entire idea of a "ceiling" is flawed and doesn't correctly handle the two cases you're concerned about - a market rise followed by a drop, and strong market performance that leaves you with more money than you can spend.

What you want is a withdrawal strategy that saves the first chunk of returns from your portfolio, but then lets you spend a portion of any excessive returns. So something like "spend X% of your portfolio at retirement (inflation adjusted), and Y% of any excess over Z% of your portfolio". If you graph spending against portfolio value, it'll look flat until it's at the right-hand side of the graph, then a relatively sharply increasing line.

Alternatively and equivalently, you could have a constant spend and sell call options at the strike price where you'd increase spending.
Current portfolio: 60% VTI / 40% VXUS

User avatar
AtlasShrugged?
Posts: 689
Joined: Wed Jul 15, 2015 6:08 pm

Re: Bengen Floor & Ceiling Withdrawal Method

Post by AtlasShrugged? » Mon Apr 08, 2019 1:46 pm

That said, some of the strategies that often perform better than Bengen's Floor-to-ceiling are McClung's EM, longinvest's VPW, and just using the IRS RMD tables. You'll notice that all three of the "stronger" strategies take advantage of "how much time you've got left". The basic idea is that withdrawing 10% of your portfolio when you are 65 is reckless but withdrawing 10% of your portfolio when you are 93 is not; so take advantage of that reality when designing a strategy.
AlohaJoe....Great comment. I have learned much reading your 'take' on this topic. Are you familiar with Stanford's recent study on withdrawal strategies? I stumbled across this on 'The Retirement Manifesto' recently. In essence, my takeaway was the ideal seemed to be delaying SSA to 70, and use the IRS RMD tables. That plan was rated best overall because it is simple (not necessarily easy). But I digress.

My question(s) for you pertain to McClung's Living Off Your Money. Are you aware of any update the author might have made? It has been 3-4 years since publication (his data went to 2010 as I recall), and I am wondering if the additional data points (through 2018) would help in seeing whether he really nailed it or not? Does it make the argument stronger or weaker to adopt his proposed system. The next question is this: Has anyone really tried to make the maintenance phase of his framework simpler? There is considerable upfront work needed (plugging in a bunch of numbers from disparate data sources). I'd be afraid as an old guy I would mess it up (cognitive decline).

Honestly, in comparing the two methods (Stanford Longevity recommendation versus Prime Harvesting [edit: EM, ECM]), the keep working and delay SSA to 70, then just do RMDs sounds better and better.

Am I way off-base in my thinking?
“If you don't know, the thing to do is not to get scared, but to learn.”

User avatar
Tyler9000
Posts: 423
Joined: Fri Aug 21, 2015 11:57 am

Re: Bengen Floor & Ceiling Withdrawal Method

Post by Tyler9000 » Mon Apr 08, 2019 2:07 pm

sixtyforty wrote:
Mon Apr 08, 2019 12:17 pm
Thanks Tyler. Could you confirm how your performing the calculation in the example you show for Bengen Floor to Celing ? I'm assuming it's applying the fixed % each year on the portfolio amount and then just determining if it's within the floor & ceiling (calculated with the floor and celing values on the first year). I don't see where the spending would be adjusted for inflation each year.
Practically speaking, the Bengen Floor & Ceiling method multiplies the account value by the withdrawal rate every year and then checks that it’s within the desired limits. But the charts translate everything to real returns, which turns normal inflation into horizontal lines to help visualize constant purchasing power. So the one trick in applying the F&C method is to realize that the limits will also increase in nominal terms with inflation over the years.

AlohaJoe
Posts: 4556
Joined: Mon Nov 26, 2007 2:00 pm
Location: Saigon, Vietnam

Re: Bengen Floor & Ceiling Withdrawal Method

Post by AlohaJoe » Mon Apr 08, 2019 8:23 pm

AtlasShrugged? wrote:
Mon Apr 08, 2019 1:46 pm
AlohaJoe....Great comment. I have learned much reading your 'take' on this topic. Are you familiar with Stanford's recent study on withdrawal strategies? I stumbled across this on 'The Retirement Manifesto' recently. In essence, my takeaway was the ideal seemed to be delaying SSA to 70, and use the IRS RMD tables. That plan was rated best overall because it is simple (not necessarily easy). But I digress.
Yes, I'm familiar with the plan. One of the authors -- Steve Vernon -- has a recent book Retirement Game-Changers -- that covers similar ground in a slightly more readable format. I agree it is a good default one for many retirees. (Though many people seem to ignore the authors' other suggestions about reverse mortgages, annuities, the portfolio being 100% stocks, and using actuarial methods to handle "income disconuities".)

Keep in mind that it is for "middle-income", which they define as $100,000 to $1,000,000 in assets (including home equity). Most Bogleheads aren't middle-income by that definition, though it is still a good starting point for building a plan.
My question(s) for you pertain to McClung's Living Off Your Money. Are you aware of any update the author might have made? It has been 3-4 years since publication (his data went to 2010 as I recall), and I am wondering if the additional data points (through 2018) would help in seeing whether he really nailed it or not?
A few of us exchanged email with him for a few months after the book came out. He started a new job and said he was swamped with that. As you said, it has been 3-4 years since then, though. He did say that not including any PMT-based systems was his biggest oversight, so perhaps any eventual update would also include that.
The next question is this: Has anyone really tried to make the maintenance phase of his framework simpler? There is considerable upfront work needed (plugging in a bunch of numbers from disparate data sources). I'd be afraid as an old guy I would mess it up (cognitive decline).
I'm not sure there's too much to mess up once the plan is going. There's a spreadsheet which handles most of it. That said, I agree with you that cognitive decline is an issue that nearly all systems don't pay enough attention to. But I also think that, realistically, you could easily use a more complex/sophisticated/whatever plan in the early years of one's retirement and switch to something simpler later on. After all, most of us are primarily concerned with the early years: if things aren't going off the rail in the first ~10 years or whatever then any withdrawal strategy will work fine, after all.

Greg in Idaho
Posts: 113
Joined: Tue Dec 27, 2016 12:59 pm

Re: Bengen Floor & Ceiling Withdrawal Method

Post by Greg in Idaho » Tue Apr 09, 2019 10:44 am

As many have mentioned, this is one of the areas (withdrawal/spending methods in retirement) that is underdeveloped, at least compared to resources on investment. Would love to see more work in this area, as McClung is not the most accessible treatment, and his work appears to have spurred a bunch of thought since he published it (I don't find it that difficult, but it does take some work to take in, process, and use).

Post Reply