Withdrawal rate for an early retirement

Discuss all general (i.e. non-personal) investing questions and issues, investing news, and theory.
Post Reply
YRT70
Posts: 1289
Joined: Sat Apr 27, 2019 8:51 am

Re: Withdrawal rate for an early retirement

Post by YRT70 »

willthrill81 wrote: Thu Nov 26, 2020 10:28 am
YRT70 wrote: Thu Nov 26, 2020 10:15 am
willthrill81 wrote: Thu Nov 26, 2020 9:55 am That brings up another point I've made already: I'm not sure that the rising equity glidepath's seeming improvement would have been statistically significant. If it isn't, then the results may be due to random chance.
Yes you've said that 3 times now :) And I'll keep saying this: while the differences in Kitces' model were relatively small, ERN's analysis showed more potent benefits for long retirements, which is what the OP is asking about.

And I'm not sure if you read what Kitces said in that podcast, but nothing he said there gives me the impression he has changed his mind about it.
If we agree that the improvement in 30 year SWRs by the rising equity glidepath is, at best, not significant, then let's agree to table that one.
In that specific model with those specific assumptions you're absolutely right that there was little difference between the 40/60 static and the 30>70 glide path. For other static AAs the improvement tended to be a bit better. And yes I would also like to see more work done with different models and assumptions.

But besides all the models, I also like the idea intuitively: taking a little less risk early on when SoRR is highest, at the cost of some upside potential. As Karsten phrased it in the conclusion I quoted above.
User avatar
CalPoppy
Posts: 313
Joined: Thu May 04, 2017 2:57 pm

Re: Withdrawal rate for an early retirement

Post by CalPoppy »

Rob1 wrote: Thu Nov 26, 2020 11:13 am For some perspective:

When the CFPs at Vanguard Personal Advisory Services conduct a portfolio analysis, they like to see an estimated success rate of 80% or better.

And Bernstein states: "But history teaches us that depriving ourselves to boost our 40-year success probability much beyond 80% is a fool’s errand, since all you are doing is increasing the probability of failure for political, economic, and military reasons relative to the failure of banal financial planning." More context here: http://www.efficientfrontier.com/ef/901/hell3.htm
And I’ll add: Targeting a 100% SWR based on all sorts of negative outlier assumptions (whether it be expenses, returns, life expectancy...) is too conservative for me. Particularly when other good solutions exist - such as adding expense flexibility and/or bit of human capital income, or using ABW.
Marseille07
Posts: 16054
Joined: Fri Nov 06, 2020 12:41 pm

Re: Withdrawal rate for an early retirement

Post by Marseille07 »

YRT70 wrote: Thu Nov 26, 2020 11:34 am
willthrill81 wrote: Thu Nov 26, 2020 10:28 am
YRT70 wrote: Thu Nov 26, 2020 10:15 am
willthrill81 wrote: Thu Nov 26, 2020 9:55 am That brings up another point I've made already: I'm not sure that the rising equity glidepath's seeming improvement would have been statistically significant. If it isn't, then the results may be due to random chance.
Yes you've said that 3 times now :) And I'll keep saying this: while the differences in Kitces' model were relatively small, ERN's analysis showed more potent benefits for long retirements, which is what the OP is asking about.

And I'm not sure if you read what Kitces said in that podcast, but nothing he said there gives me the impression he has changed his mind about it.
If we agree that the improvement in 30 year SWRs by the rising equity glidepath is, at best, not significant, then let's agree to table that one.
In that specific model with those specific assumptions you're absolutely right that there was little difference between the 40/60 static and the 30>70 glide path. For other static AAs the improvement tended to be a bit better. And yes I would also like to see more work done with different models and assumptions.

But besides all the models, I also like the idea intuitively: taking a little less risk early on when SoRR is highest, at the cost of some upside potential. As Karsten phrased it in the conclusion I quoted above.
I don't know, I don't see many fans of rising glidepath. It feels backwards to me, because the whole point of SoRR is you run into it early so you have more time to recover later.

And fundamentally, if you're comfortable going 90/10 at age 80, why couldn't you do so at 60?
BogleFan510
Posts: 1039
Joined: Tue Aug 04, 2020 2:13 pm

Re: Withdrawal rate for an early retirement

Post by BogleFan510 »

removed
Last edited by BogleFan510 on Fri Jul 09, 2021 7:40 pm, edited 1 time in total.
YRT70
Posts: 1289
Joined: Sat Apr 27, 2019 8:51 am

Re: Withdrawal rate for an early retirement

Post by YRT70 »

Marseille07 wrote: Thu Nov 26, 2020 11:42 am
YRT70 wrote: Thu Nov 26, 2020 11:34 am
willthrill81 wrote: Thu Nov 26, 2020 10:28 am
YRT70 wrote: Thu Nov 26, 2020 10:15 am
willthrill81 wrote: Thu Nov 26, 2020 9:55 am That brings up another point I've made already: I'm not sure that the rising equity glidepath's seeming improvement would have been statistically significant. If it isn't, then the results may be due to random chance.
Yes you've said that 3 times now :) And I'll keep saying this: while the differences in Kitces' model were relatively small, ERN's analysis showed more potent benefits for long retirements, which is what the OP is asking about.

And I'm not sure if you read what Kitces said in that podcast, but nothing he said there gives me the impression he has changed his mind about it.
If we agree that the improvement in 30 year SWRs by the rising equity glidepath is, at best, not significant, then let's agree to table that one.
In that specific model with those specific assumptions you're absolutely right that there was little difference between the 40/60 static and the 30>70 glide path. For other static AAs the improvement tended to be a bit better. And yes I would also like to see more work done with different models and assumptions.

But besides all the models, I also like the idea intuitively: taking a little less risk early on when SoRR is highest, at the cost of some upside potential. As Karsten phrased it in the conclusion I quoted above.
I don't know, I don't see many fans of rising glidepath.
Among Bogleheads it doesn't seem very popular I agree. That doesn't really matter to me though. Personally I think Michael Kitces, Karsten Jeske and Wade Pfau make good points for it.
It feels backwards to me, because the whole point of SoRR is you run into it early so you have more time to recover later.
If you're hit by a market crash early on your equity portion may have been hit too hard to recover and your nest egg may not support your intended withdrawal rate. That's what happened during the worst times to retire. The rising equity glide would have worked in those scenario's: it meaningfully increased the safe withdrawal rate, see the article on ERN: https://earlyretirementnow.com/2017/09/ ... glidepaths
And fundamentally, if you're comfortable going 90/10 at age 80, why couldn't you do so at 60?
I think the rising equity glide path has more to offer for early retirees like me. I would not be comfortable at 90/10 at the moment. I like the size of my safe assets as it is. I'm ok with slightly rising equity exposure. I'll see how far I have to take it. Chances are I will never have to take it to 90/10.
User avatar
geerhardusvos
Posts: 2046
Joined: Wed Oct 23, 2019 10:20 pm
Location: heavenlies

Re: Withdrawal rate for an early retirement

Post by geerhardusvos »

smitcat wrote: Thu Nov 26, 2020 7:52 am
willthrill81 wrote: Wed Nov 25, 2020 10:55 pm
Given the other information that geerhardusvos has provided, I suspect that even with early retirement that he'll still get a not insignificant amount of SS benefits down the road, which may reduce his needed withdrawals significantly. Just 10 years of maxing out the SS taxable maximum results in about $13.5k of annual SS benefits at age 62 and $25k at age 70, not counting the spousal benefit. Therefore, he may only be withdrawing 3-4% for 30-35 years and then withdrawing even less after that (if his portfolio hasn't done well). If so, that sounds really secure to me.
"Given the other information that geerhardusvos has provided, I suspect that even with early retirement that he'll still get a not insignificant amount of SS benefits down the road, which may reduce his needed withdrawals significantly"
Yes - a one or more of these that he has posted about makes all the difference in how you might view safe withdrawal rates..
- his wife may return to full time work
- his side gig makes money now and will continue
- SS will become a factor at some point
- he plans to expand the side gig and income
- the odds of an inheritance are good
Any of these would affect your views of both SWR and AA.
Not sure how this thread became about my specific situation, so sorry about that, OP. I appreciate Will’s thoughts and agree.

Our base expenses/budget will be supported by a 3 to 3.5% WR. None of the above needs to happen for us to survive and have a successful and happy cycle. All of those items are gravy (as we’ve discussed before), and even if some of them have a likelihood of happening like some SS or inheritance, I won’t need those neither do I expect them. If things get really really bad, I will cancel ESPN+ or go on one less vacation.

3 to 5% WR is ridiculously conservative. Stop worrying about it!
VTSAX and chill
Marseille07
Posts: 16054
Joined: Fri Nov 06, 2020 12:41 pm

Re: Withdrawal rate for an early retirement

Post by Marseille07 »

YRT70 wrote: Thu Nov 26, 2020 12:02 pm If you're hit by a market crash early on your equity portion may have been hit too hard to recover and your nest egg may not support your intended withdrawal rate. That's what happened during the worst times to retire. The rising equity glide would have worked in those scenario's: it meaningfully increased the safe withdrawal rate, see the article on ERN: https://earlyretirementnow.com/2017/09/ ... glidepaths
Here's the thing though. Looking at those unfortunate retirees is a curve-fit example because:
a) they ran into hardship right after retiring
b) they did not run into a market crash after they raised glidpath

The reverse can easily be our case, where:
a) we have great markets right after retiring but we didn't benefit much because we were doing 50/50
b) as we raised glidepath to 80/20 at age 80, markets crashed hard

Of course, I'm not here to tell you what to do, and it's actually insightful to learn about rising glidepath although I am not a big fan of the idea.
User avatar
willthrill81
Posts: 32250
Joined: Thu Jan 26, 2017 2:17 pm
Location: USA
Contact:

Re: Withdrawal rate for an early retirement

Post by willthrill81 »

Marseille07 wrote: Thu Nov 26, 2020 12:35 pm
YRT70 wrote: Thu Nov 26, 2020 12:02 pm If you're hit by a market crash early on your equity portion may have been hit too hard to recover and your nest egg may not support your intended withdrawal rate. That's what happened during the worst times to retire. The rising equity glide would have worked in those scenario's: it meaningfully increased the safe withdrawal rate, see the article on ERN: https://earlyretirementnow.com/2017/09/ ... glidepaths
Here's the thing though. Looking at those unfortunate retirees is a curve-fit example because:
a) they ran into hardship right after retiring
b) they did not run into a market crash after they raised glidpath

The reverse can easily be our case, where:
a) we have great markets right after retiring but we didn't benefit much because we were doing 50/50
b) as we raised glidepath to 80/20 at age 80, markets crashed hard
You've identified what I also see to be a potentially big problem with the rising equity glidepath approach: if stocks do very well when you have a low allocation to them and subsequently crash when your equity allocation is higher, it can lead to a situation significantly worse than having a static AA. And with real bond yields almost certainly being negative over the next decade, having 50% or more of one's allocation in bonds right now seems even more problematic for early retirees.

I'm open to the idea but still skeptical.
The Sensible Steward
YRT70
Posts: 1289
Joined: Sat Apr 27, 2019 8:51 am

Re: Withdrawal rate for an early retirement

Post by YRT70 »

Marseille07 wrote: Thu Nov 26, 2020 12:35 pm The reverse can easily be our case, where:
a) we have great markets right after retiring but we didn't benefit much because we were doing 50/50
Sure. There's a cost to using a glide path too. And that's what you're describing. But if markets are good early on you've made it through the 'retirement danger zone', it's early on when SoRR can hit. So you got your protection against SoRR, you still made enough money for your retirement just less, and you end up with less money when you die. Have you read the articles from Kitces and ERN well? This is all described pretty well imo.

I'll quote the conclusion from Karsten again:
Early retirees need the power of equity expected returns to make the nest egg last for many decades. Even more so than the traditional retiree at age 65! But that exposes us to Sequence of Return Risk. An equity glidepath can alleviate some of the negative effects of Sequence of Return Risk. But it shouldn’t come as a surprise that you will never completely eliminate the risk. For a given withdrawal rate, say 3.5%, we can only reduce the failure rate while leaving some residual risk. And likewise, the 4% rule would still not be safe for today’s early retirees even with an equity glidepath.

Moreover, an equity glidepath is like an insurance policy. A hedge against a tail event! On average it will cost you money, but if and when you need it the most it will likely pay off. Exactly when the static stock/bond allocation paths had their worst sustainable safe withdrawal rates you get slightly better results but you also give up some of the upside if the equity market “decides” to rally some more right after your retirement. But that’s a good problem to have!
https://earlyretirementnow.com/2017/09/ ... lidepaths/
b) as we raised glidepath to 80/20 at age 80, markets crashed hard
Right, but then you likely have had good returns early on and your nest egg is that large that even a 50% crash of your 90% equity portion would still leave you with enough money to live the last years of your life.
Of course, I'm not here to tell you what to do, and it's actually insightful to learn about rising glidepath although I am not a big fan of the idea.
I think you're in a very different situation than me. I have no idea if a glide path makes any sense in your situation. I do believe it makes sense in my situation.
User avatar
vineviz
Posts: 14921
Joined: Tue May 15, 2018 1:55 pm
Location: Baltimore, MD

Re: Withdrawal rate for an early retirement

Post by vineviz »

marcopolo wrote: Thu Nov 26, 2020 11:19 am There is a lot of assumptions packed into the highlighted paragraph. They may or may not be true. The underlying distribution is likely quite complex, and the parameters may change over time.
There’s nothing particularly complex about the distribution of expected return. Certainly nothing that would invalidate the assumption that the distribution is adequately described by the first two moments.
Your three input formula does NOT produce THE expected value, it produces AN expected value, based on your assumptions, which may or may not be accurate. The inputs themselves are also subject to the same problem.
This is not a substantial critique: every expected value estimate is conditional on the process used to create it.

In any case, you can use whichever values for the inputs you’d like. There’s no way to get a 90% “safe” withdrawal rate going forward without using preposterous inputs.
"Far more money has been lost by investors preparing for corrections than has been lost in corrections themselves." ~~ Peter Lynch
User avatar
vineviz
Posts: 14921
Joined: Tue May 15, 2018 1:55 pm
Location: Baltimore, MD

Re: Withdrawal rate for an early retirement

Post by vineviz »

willthrill81 wrote: Thu Nov 26, 2020 10:33 am I agree that for more typical retirement periods (e.g. 30 years), the rising equity glidepath doesn't seem to be beneficial. But Karsten's analysis seemed to indicate pretty robustly that a rising equity glidepath worked well at improving the SWR for 60 year retirements when the starting stock allocation was at least 60%. Also, this improvement seemed to increase when starting valuations were high, as they are now.
What Karsten found was a fluke: a historical accident. It’s not a robust finding, and certainly not a universal one. Basically, it’s overfitting the model to a limited set of data and not a strategy that is likely to be helpful going forward.
"Far more money has been lost by investors preparing for corrections than has been lost in corrections themselves." ~~ Peter Lynch
YRT70
Posts: 1289
Joined: Sat Apr 27, 2019 8:51 am

Re: Withdrawal rate for an early retirement

Post by YRT70 »

vineviz wrote: Thu Nov 26, 2020 1:18 pm
willthrill81 wrote: Thu Nov 26, 2020 10:33 am I agree that for more typical retirement periods (e.g. 30 years), the rising equity glidepath doesn't seem to be beneficial. But Karsten's analysis seemed to indicate pretty robustly that a rising equity glidepath worked well at improving the SWR for 60 year retirements when the starting stock allocation was at least 60%. Also, this improvement seemed to increase when starting valuations were high, as they are now.
What Karsten found was a fluke: a historical accident. It’s not a robust finding, and certainly not a universal one. Basically, it’s overfitting the model to a limited set of data and not a strategy that is likely to be helpful going forward.
This does not look like a fluke to me. Every glide path starting at 60% provided a higher SWR than every static allocation. Even all the 40 to 100 glide paths beat the SWR of every static allocation.

Image

Source: https://earlyretirementnow.com/2017/09/ ... lidepaths/
truenorth418
Posts: 637
Joined: Wed Dec 19, 2012 6:38 am

Re: Withdrawal rate for an early retirement

Post by truenorth418 »

I retired at 47. I use a fixed % rate.

For the first 8 years I used 2.5% of previous year's ending balance.

55-65 I'll use 3%. But I have a small pension awaiting me at age 65 so there's some buffer there.

Another way is use the IRS's RMD %. At age 45 it will be a little less than 2%, but the rate will inch up a little bit each year. The factors were recently updated to reflect longer lifespans.

I'll use the RMD % from age 65, which is about 3%, and increase accordingly until death.
User avatar
vineviz
Posts: 14921
Joined: Tue May 15, 2018 1:55 pm
Location: Baltimore, MD

Re: Withdrawal rate for an early retirement

Post by vineviz »

YRT70 wrote: Thu Nov 26, 2020 1:32 pm
vineviz wrote: Thu Nov 26, 2020 1:18 pm
willthrill81 wrote: Thu Nov 26, 2020 10:33 am I agree that for more typical retirement periods (e.g. 30 years), the rising equity glidepath doesn't seem to be beneficial. But Karsten's analysis seemed to indicate pretty robustly that a rising equity glidepath worked well at improving the SWR for 60 year retirements when the starting stock allocation was at least 60%. Also, this improvement seemed to increase when starting valuations were high, as they are now.
What Karsten found was a fluke: a historical accident. It’s not a robust finding, and certainly not a universal one. Basically, it’s overfitting the model to a limited set of data and not a strategy that is likely to be helpful going forward.
This does not look like a fluke to me. Every glide path starting at 60% provided a higher SWR than every static allocation. Even all the 40 to 100 glide paths beat the SWR of every static allocation.

Image

Source: https://earlyretirementnow.com/2017/09/ ... lidepaths/
That’s exactly the kind of result you get from over-fitting: erratic and inconsistent results with “improvement” concentrated in most severe extremes.

He’s pulling data from less than 3 independent 60-year historical period: no matter how many times you resample the same data, the statistics are not favorable.
"Far more money has been lost by investors preparing for corrections than has been lost in corrections themselves." ~~ Peter Lynch
EnjoyIt
Posts: 8244
Joined: Sun Dec 29, 2013 7:06 pm

Re: Withdrawal rate for an early retirement

Post by EnjoyIt »

Marseille07 wrote: Thu Nov 26, 2020 12:35 pm
YRT70 wrote: Thu Nov 26, 2020 12:02 pm If you're hit by a market crash early on your equity portion may have been hit too hard to recover and your nest egg may not support your intended withdrawal rate. That's what happened during the worst times to retire. The rising equity glide would have worked in those scenario's: it meaningfully increased the safe withdrawal rate, see the article on ERN: https://earlyretirementnow.com/2017/09/ ... glidepaths
Here's the thing though. Looking at those unfortunate retirees is a curve-fit example because:
a) they ran into hardship right after retiring
b) they did not run into a market crash after they raised glidpath

The reverse can easily be our case, where:
a) we have great markets right after retiring but we didn't benefit much because we were doing 50/50
b) as we raised glidepath to 80/20 at age 80, markets crashed hard

Of course, I'm not here to tell you what to do, and it's actually insightful to learn about rising glidepath although I am not a big fan of the idea.
If we are talking about early retirees, that means people who are in their early 50s and before. I can tell you that if I retire tomorrow and then a year later the market plummets bad, I would very likely pick up some income. I couldn’t help myself. If on the other hand the bad sequence of returns occurs 5-10 years later, by then my portfolio should be large enough to withstand those poor sequence of returns.
A time to EVALUATE your jitters: | viewtopic.php?p=1139732#p1139732
Topic Author
nptit
Posts: 102
Joined: Sun Mar 29, 2015 5:51 pm

Re: Withdrawal rate for an early retirement

Post by nptit »

Marseille07 wrote: Wed Nov 25, 2020 7:21 pm
EnjoyIt wrote: Wed Nov 25, 2020 7:19 pm
Marseille07 wrote: Wed Nov 25, 2020 6:55 pm
EnjoyIt wrote: Wed Nov 25, 2020 6:48 pm I believe the sub 3% people are going way overboard. It’s one thing to retire at retirement age and then figuring out you can comfortably live on 2.5% withdrawals.
Retirement age doesn't matter. If OP can hit 4M by 45 and figure out they can comfortably live on 2% (80K/yr) then nothing wrong with 2% or retiring early. Advocating such an idea isn't going way overboard.
This is the crux. Is the 2% because of fear or is it 2% because they have no desire to spend more. Very big difference here.
We need to ask the OP, but my understanding is that they are content with spending 80K/year. Of course, we can all increase spending if we want, like flying business / first class.
To make it clear, 80K/year should be plenty if we don’t have a mortgage. We are spending this much now with a mortgage. My wife and I are pretty frugal, we live below our means. 2% is more of no desire to go beyond. Likely we won’t get a luxury sport car even we have 10MM. Regarding the 4MM by 45 I believe it is very achievable by us, we are at 30 with a decent income and NW. It is projected to achieve in a few years. My wife also enjoys her job, so she will likely work more even we achieve the goal. I think she is confused by reading the post and we might end up just go this route as we won’t be suffered to get there.
Marseille07
Posts: 16054
Joined: Fri Nov 06, 2020 12:41 pm

Re: Withdrawal rate for an early retirement

Post by Marseille07 »

nptit wrote: Thu Nov 26, 2020 3:27 pm
Marseille07 wrote: Wed Nov 25, 2020 7:21 pm
EnjoyIt wrote: Wed Nov 25, 2020 7:19 pm
Marseille07 wrote: Wed Nov 25, 2020 6:55 pm
EnjoyIt wrote: Wed Nov 25, 2020 6:48 pm I believe the sub 3% people are going way overboard. It’s one thing to retire at retirement age and then figuring out you can comfortably live on 2.5% withdrawals.
Retirement age doesn't matter. If OP can hit 4M by 45 and figure out they can comfortably live on 2% (80K/yr) then nothing wrong with 2% or retiring early. Advocating such an idea isn't going way overboard.
This is the crux. Is the 2% because of fear or is it 2% because they have no desire to spend more. Very big difference here.
We need to ask the OP, but my understanding is that they are content with spending 80K/year. Of course, we can all increase spending if we want, like flying business / first class.
To make it clear, 80K/year should be plenty if we don’t have a mortgage. We are spending this much now with a mortgage. My wife and I are pretty frugal, we live below our means. 2% is more of no desire to go beyond. Likely we won’t get a luxury sport car even we have 10MM. Regarding the 4MM by 45 I believe it is very achievable by us, we are at 30 with a decent income and NW. It is projected to achieve in a few years. My wife also enjoys her job, so she will likely work more even we achieve the goal. I think she is confused by reading the post and we might end up just go this route as we won’t be suffered to get there.
Sounds like a good & safe plan to me, I like it. No reason to withdraw some extra money you don't end up spending.
Last edited by Marseille07 on Thu Nov 26, 2020 6:51 pm, edited 1 time in total.
klaus14
Posts: 953
Joined: Sun Nov 25, 2018 6:43 pm

Re: Withdrawal rate for an early retirement

Post by klaus14 »

ERN charts are quoted here a lot. But ERN himself uses return expectations in his favorite method (CAPE based SWR rules here). I plan using CAPE rule but i also validate it against return expectations for my portfolio as i explained here:
viewtopic.php?f=10&t=331020
My investment algorithm: https://www.bogleheads.org/forum/viewtopic.php?f=10&t=351899&p=6112869#p6112869
User avatar
willthrill81
Posts: 32250
Joined: Thu Jan 26, 2017 2:17 pm
Location: USA
Contact:

Re: Withdrawal rate for an early retirement

Post by willthrill81 »

klaus14 wrote: Thu Nov 26, 2020 4:51 pm ERN charts are quoted here a lot. But ERN himself uses return expectations in his favorite method (CAPE based SWR rules here).
Yep. The method he uses is basically the same as the amortization based withdrawal method that many of us have concluded is the 'best at accomplishing the most'. No simulations needed.
The Sensible Steward
Marseille07
Posts: 16054
Joined: Fri Nov 06, 2020 12:41 pm

Re: Withdrawal rate for an early retirement

Post by Marseille07 »

willthrill81 wrote: Thu Nov 26, 2020 7:50 pm
klaus14 wrote: Thu Nov 26, 2020 4:51 pm ERN charts are quoted here a lot. But ERN himself uses return expectations in his favorite method (CAPE based SWR rules here).
Yep. The method he uses is basically the same as the amortization based withdrawal method that many of us have concluded is the 'best at accomplishing the most'. No simulations needed.
How are you measuring 'best' and 'accomplishing the most'?
User avatar
willthrill81
Posts: 32250
Joined: Thu Jan 26, 2017 2:17 pm
Location: USA
Contact:

Re: Withdrawal rate for an early retirement

Post by willthrill81 »

Marseille07 wrote: Thu Nov 26, 2020 9:29 pm
willthrill81 wrote: Thu Nov 26, 2020 7:50 pm
klaus14 wrote: Thu Nov 26, 2020 4:51 pm ERN charts are quoted here a lot. But ERN himself uses return expectations in his favorite method (CAPE based SWR rules here).
Yep. The method he uses is basically the same as the amortization based withdrawal method that many of us have concluded is the 'best at accomplishing the most'. No simulations needed.
How are you measuring 'best' and 'accomplishing the most'?
The reason I bracketed that is because it's obviously a subjective assessment. However, the advantages of the ABW method are numerous. It is a mathematically driven approach that will not prematurely deplete a portfolio, can be easily adapted to current market conditions (i.e. by incorporating the current portfolio value and forward return estimates), and can incorporate multiple income streams and/or expenses at various points in time. I know of no other approach that can do all of that, and I've seen a lot of withdrawal methods. About the only real disadvantage of the approach is that it can get a bit complex, especially if your financial situation is complicated, but other users have developed easy to use spreadsheets that greatly help. Another potential disadvantage is that it may lead to what turns out to be unnecessarily volatile withdrawals, but this is true of any flexible withdrawal method; you don't know that the flexibility was unnecessary until after the fact.
The Sensible Steward
Marseille07
Posts: 16054
Joined: Fri Nov 06, 2020 12:41 pm

Re: Withdrawal rate for an early retirement

Post by Marseille07 »

willthrill81 wrote: Thu Nov 26, 2020 9:39 pm
Marseille07 wrote: Thu Nov 26, 2020 9:29 pm
willthrill81 wrote: Thu Nov 26, 2020 7:50 pm
klaus14 wrote: Thu Nov 26, 2020 4:51 pm ERN charts are quoted here a lot. But ERN himself uses return expectations in his favorite method (CAPE based SWR rules here).
Yep. The method he uses is basically the same as the amortization based withdrawal method that many of us have concluded is the 'best at accomplishing the most'. No simulations needed.
How are you measuring 'best' and 'accomplishing the most'?
The reason I bracketed that is because it's obviously a subjective assessment. However, the advantages of the ABW method are numerous. It is a mathematically driven approach that will not prematurely deplete a portfolio, can be easily adapted to current market conditions (i.e. by incorporating the current portfolio value and forward return estimates), and can incorporate multiple income streams and/or expenses at various points in time. I know of no other approach that can do all of that, and I've seen a lot of withdrawal methods. About the only real disadvantage of the approach is that it can get a bit complex, especially if your financial situation is complicated, but other users have developed easy to use spreadsheets that greatly help. Another potential disadvantage is that it may lead to what turns out to be unnecessarily volatile withdrawals, but this is true of any flexible withdrawal method; you don't know that the flexibility was unnecessary until after the fact.
I see, I think you described fair points on ABW. I mentioned elsewhere but volatile withdrawals are a non-issue with proper EF, at least for me.
YRT70
Posts: 1289
Joined: Sat Apr 27, 2019 8:51 am

Re: Withdrawal rate for an early retirement

Post by YRT70 »

vineviz wrote: Thu Nov 26, 2020 2:46 pm That’s exactly the kind of result you get from over-fitting: erratic and inconsistent results with “improvement” concentrated in most severe extremes.
That's exactly the goal: protection in the most extreme scenarios.

The discussion is very similar to lump sum vs. DCA. Lump sum investing will beat DCA most of the time. But if the markets drop right after you lump summed, there's a good chance DCA would have done better. A rising equity glide path is a lot like DCA. Kitces explains it well in his article. https://www.kitces.com/blog/should-equi ... ly-better/
EnjoyIt
Posts: 8244
Joined: Sun Dec 29, 2013 7:06 pm

Re: Withdrawal rate for an early retirement

Post by EnjoyIt »

Marseille07 wrote: Thu Nov 26, 2020 9:46 pm
willthrill81 wrote: Thu Nov 26, 2020 9:39 pm
Marseille07 wrote: Thu Nov 26, 2020 9:29 pm
willthrill81 wrote: Thu Nov 26, 2020 7:50 pm
klaus14 wrote: Thu Nov 26, 2020 4:51 pm ERN charts are quoted here a lot. But ERN himself uses return expectations in his favorite method (CAPE based SWR rules here).
Yep. The method he uses is basically the same as the amortization based withdrawal method that many of us have concluded is the 'best at accomplishing the most'. No simulations needed.
How are you measuring 'best' and 'accomplishing the most'?
The reason I bracketed that is because it's obviously a subjective assessment. However, the advantages of the ABW method are numerous. It is a mathematically driven approach that will not prematurely deplete a portfolio, can be easily adapted to current market conditions (i.e. by incorporating the current portfolio value and forward return estimates), and can incorporate multiple income streams and/or expenses at various points in time. I know of no other approach that can do all of that, and I've seen a lot of withdrawal methods. About the only real disadvantage of the approach is that it can get a bit complex, especially if your financial situation is complicated, but other users have developed easy to use spreadsheets that greatly help. Another potential disadvantage is that it may lead to what turns out to be unnecessarily volatile withdrawals, but this is true of any flexible withdrawal method; you don't know that the flexibility was unnecessary until after the fact.
I see, I think you described fair points on ABW. I mentioned elsewhere but volatile withdrawals are a non-issue with proper EF, at least for me.
I don’t understand how an emergency fund helps you with that. Can you please elaborate? When do you use the fund? How do you replenish the fund?
A time to EVALUATE your jitters: | viewtopic.php?p=1139732#p1139732
Derpalator
Posts: 215
Joined: Sun Nov 15, 2015 1:52 pm

Re: Withdrawal rate for an early retirement

Post by Derpalator »

Rob1 wrote: Thu Nov 26, 2020 11:39 am
Rob1 wrote: Thu Nov 26, 2020 11:13 am For some perspective:

When the CFPs at Vanguard Personal Advisory Services conduct a portfolio analysis, they like to see an estimated success rate of 80% or better.

And Bernstein states: "But history teaches us that depriving ourselves to boost our 40-year success probability much beyond 80% is a fool’s errand, since all you are doing is increasing the probability of failure for political, economic, and military reasons relative to the failure of banal financial planning." More context here: http://www.efficientfrontier.com/ef/901/hell3.htm
And I’ll add: Targeting a 100% SWR based on all sorts of negative outlier assumptions (whether it be expenses, returns, life expectancy...) is too conservative for me. Particularly when other good solutions exist - such as adding expense flexibility and/or bit of human capital income, or using ABW.
Rob 1, I agree with both the first and second parts of your post. I initially wrote specifics about my own calculations but you say it more simply and better. Some say cash is king, I say FLEXIBILITY is king.

As an aside, I never would have had the wherewithal to grok personal finance had I not discovered this forum. Thanks and many blessings to all.
User avatar
vineviz
Posts: 14921
Joined: Tue May 15, 2018 1:55 pm
Location: Baltimore, MD

Re: Withdrawal rate for an early retirement

Post by vineviz »

YRT70 wrote: Thu Nov 26, 2020 11:33 pm
vineviz wrote: Thu Nov 26, 2020 2:46 pm That’s exactly the kind of result you get from over-fitting: erratic and inconsistent results with “improvement” concentrated in most severe extremes.
That's exactly the goal: protection in the most extreme scenarios.

The discussion is very similar to lump sum vs. DCA. Lump sum investing will beat DCA most of the time. But if the markets drop right after you lump summed, there's a good chance DCA would have done better. A rising equity glide path is a lot like DCA. Kitces explains it well in his article. https://www.kitces.com/blog/should-equi ... ly-better/
You misunderstand me. The ERN results supporting a rising equity glide path appear to be noise, an example of data mining.

The appearance of “ protection in the most extreme scenarios” is an illusion, not real protection.
"Far more money has been lost by investors preparing for corrections than has been lost in corrections themselves." ~~ Peter Lynch
YRT70
Posts: 1289
Joined: Sat Apr 27, 2019 8:51 am

Re: Withdrawal rate for an early retirement

Post by YRT70 »

vineviz wrote: Fri Nov 27, 2020 7:04 am
YRT70 wrote: Thu Nov 26, 2020 11:33 pm
vineviz wrote: Thu Nov 26, 2020 2:46 pm That’s exactly the kind of result you get from over-fitting: erratic and inconsistent results with “improvement” concentrated in most severe extremes.
That's exactly the goal: protection in the most extreme scenarios.

The discussion is very similar to lump sum vs. DCA. Lump sum investing will beat DCA most of the time. But if the markets drop right after you lump summed, there's a good chance DCA would have done better. A rising equity glide path is a lot like DCA. Kitces explains it well in his article. https://www.kitces.com/blog/should-equi ... ly-better/
You misunderstand me. The ERN results supporting a rising equity glide path appear to be noise, an example of data mining.

The appearance of “ protection in the most extreme scenarios” is an illusion, not real protection.
I think I understand you fine. I just think it's highly unlikely that these improvements are due to noise or random chance. The improvements are too large for it and the intuitive explanation of why it works also makes sense to me. If you got data showing that this is just noise I'd like to take a look at it.

Image
https://earlyretirementnow.com/2017/09/ ... lidepaths/
Marseille07
Posts: 16054
Joined: Fri Nov 06, 2020 12:41 pm

Re: Withdrawal rate for an early retirement

Post by Marseille07 »

EnjoyIt wrote: Fri Nov 27, 2020 3:22 am
Marseille07 wrote: Thu Nov 26, 2020 9:46 pm
willthrill81 wrote: Thu Nov 26, 2020 9:39 pm
Marseille07 wrote: Thu Nov 26, 2020 9:29 pm
willthrill81 wrote: Thu Nov 26, 2020 7:50 pm

Yep. The method he uses is basically the same as the amortization based withdrawal method that many of us have concluded is the 'best at accomplishing the most'. No simulations needed.
How are you measuring 'best' and 'accomplishing the most'?
The reason I bracketed that is because it's obviously a subjective assessment. However, the advantages of the ABW method are numerous. It is a mathematically driven approach that will not prematurely deplete a portfolio, can be easily adapted to current market conditions (i.e. by incorporating the current portfolio value and forward return estimates), and can incorporate multiple income streams and/or expenses at various points in time. I know of no other approach that can do all of that, and I've seen a lot of withdrawal methods. About the only real disadvantage of the approach is that it can get a bit complex, especially if your financial situation is complicated, but other users have developed easy to use spreadsheets that greatly help. Another potential disadvantage is that it may lead to what turns out to be unnecessarily volatile withdrawals, but this is true of any flexible withdrawal method; you don't know that the flexibility was unnecessary until after the fact.
I see, I think you described fair points on ABW. I mentioned elsewhere but volatile withdrawals are a non-issue with proper EF, at least for me.
I don’t understand how an emergency fund helps you with that. Can you please elaborate? When do you use the fund? How do you replenish the fund?
The poster is concerned about unexpected 50K in expenditure, which they don't know when / if it will happen.

You build up EF simply by saving some of your monthly withdrawals. For example, let's say you have 2M and doing 4% = 6.7K/mo. You only spend 4K/mo and put the rest of cash in EF, which builds up over time.

I know you regard this EF as uninvested money down the toilet. However, this approach keeps your WR intact while providing liquidity in case of emergencies.
User avatar
willthrill81
Posts: 32250
Joined: Thu Jan 26, 2017 2:17 pm
Location: USA
Contact:

Re: Withdrawal rate for an early retirement

Post by willthrill81 »

Marseille07 wrote: Fri Nov 27, 2020 9:34 am
EnjoyIt wrote: Fri Nov 27, 2020 3:22 am
Marseille07 wrote: Thu Nov 26, 2020 9:46 pm
willthrill81 wrote: Thu Nov 26, 2020 9:39 pm
Marseille07 wrote: Thu Nov 26, 2020 9:29 pm

How are you measuring 'best' and 'accomplishing the most'?
The reason I bracketed that is because it's obviously a subjective assessment. However, the advantages of the ABW method are numerous. It is a mathematically driven approach that will not prematurely deplete a portfolio, can be easily adapted to current market conditions (i.e. by incorporating the current portfolio value and forward return estimates), and can incorporate multiple income streams and/or expenses at various points in time. I know of no other approach that can do all of that, and I've seen a lot of withdrawal methods. About the only real disadvantage of the approach is that it can get a bit complex, especially if your financial situation is complicated, but other users have developed easy to use spreadsheets that greatly help. Another potential disadvantage is that it may lead to what turns out to be unnecessarily volatile withdrawals, but this is true of any flexible withdrawal method; you don't know that the flexibility was unnecessary until after the fact.
I see, I think you described fair points on ABW. I mentioned elsewhere but volatile withdrawals are a non-issue with proper EF, at least for me.
I don’t understand how an emergency fund helps you with that. Can you please elaborate? When do you use the fund? How do you replenish the fund?
The poster is concerned about unexpected 50K in expenditure, which they don't know when / if it will happen.

You build up EF simply by saving some of your monthly withdrawals. For example, let's say you have 2M and doing 4% = 6.7K/mo. You only spend 4K/mo and put the rest of cash in EF, which builds up over time.

I know you regard this EF as uninvested money down the toilet. However, this approach keeps your WR intact while providing liquidity in case of emergencies.
It's not 'down the toilet', but it's bona fide mental accounting. Whether the money used to pay for an unexpected expenses comes from your EF or your portfolio doesn't matter; the point is that you no longer have it. The WR coming from your assets is the same either way.
The Sensible Steward
User avatar
willthrill81
Posts: 32250
Joined: Thu Jan 26, 2017 2:17 pm
Location: USA
Contact:

Re: Withdrawal rate for an early retirement

Post by willthrill81 »

YRT70 wrote: Fri Nov 27, 2020 8:49 am
vineviz wrote: Fri Nov 27, 2020 7:04 am
YRT70 wrote: Thu Nov 26, 2020 11:33 pm
vineviz wrote: Thu Nov 26, 2020 2:46 pm That’s exactly the kind of result you get from over-fitting: erratic and inconsistent results with “improvement” concentrated in most severe extremes.
That's exactly the goal: protection in the most extreme scenarios.

The discussion is very similar to lump sum vs. DCA. Lump sum investing will beat DCA most of the time. But if the markets drop right after you lump summed, there's a good chance DCA would have done better. A rising equity glide path is a lot like DCA. Kitces explains it well in his article. https://www.kitces.com/blog/should-equi ... ly-better/
You misunderstand me. The ERN results supporting a rising equity glide path appear to be noise, an example of data mining.

The appearance of “ protection in the most extreme scenarios” is an illusion, not real protection.
I think I understand you fine. I just think it's highly unlikely that these improvements are due to noise or random chance. The improvements are too large for it and the intuitive explanation of why it works also makes sense to me. If you got data showing that this is just noise I'd like to take a look at it.

Image
https://earlyretirementnow.com/2017/09/ ... lidepaths/
I understand both yours and vineviz's position. Yes, the improvement in the SWRs was seemingly substantial, but the sample size is too small for the improvement to be statistically significant. That's why I said that I would like to see more testing of this method. For instance, how well did it work in other nations in the past?
The Sensible Steward
Marseille07
Posts: 16054
Joined: Fri Nov 06, 2020 12:41 pm

Re: Withdrawal rate for an early retirement

Post by Marseille07 »

willthrill81 wrote: Fri Nov 27, 2020 9:48 am It's not 'down the toilet', but it's bona fide mental accounting. Whether the money used to pay for an unexpected expenses comes from your EF or your portfolio doesn't matter; the point is that you no longer have it. The WR coming from your assets is the same either way.
I see value in this mental accounting and budgeting. It's OK if you don't. I'm not forcing anything, just addressing the question by EnjoyIt.
User avatar
willthrill81
Posts: 32250
Joined: Thu Jan 26, 2017 2:17 pm
Location: USA
Contact:

Re: Withdrawal rate for an early retirement

Post by willthrill81 »

Marseille07 wrote: Fri Nov 27, 2020 10:00 am
willthrill81 wrote: Fri Nov 27, 2020 9:48 am It's not 'down the toilet', but it's bona fide mental accounting. Whether the money used to pay for an unexpected expenses comes from your EF or your portfolio doesn't matter; the point is that you no longer have it. The WR coming from your assets is the same either way.
I see value in this mental accounting and budgeting. It's OK if you don't. I'm not forcing anything, just addressing the question by EnjoyIt.
Yes, if you feel good about it, it's not going to hurt you in the instance you described.
The Sensible Steward
Marseille07
Posts: 16054
Joined: Fri Nov 06, 2020 12:41 pm

Re: Withdrawal rate for an early retirement

Post by Marseille07 »

willthrill81 wrote: Fri Nov 27, 2020 10:01 am
Marseille07 wrote: Fri Nov 27, 2020 10:00 am
willthrill81 wrote: Fri Nov 27, 2020 9:48 am It's not 'down the toilet', but it's bona fide mental accounting. Whether the money used to pay for an unexpected expenses comes from your EF or your portfolio doesn't matter; the point is that you no longer have it. The WR coming from your assets is the same either way.
I see value in this mental accounting and budgeting. It's OK if you don't. I'm not forcing anything, just addressing the question by EnjoyIt.
Yes, if you feel good about it, it's not going to hurt you in the instance you described.
I'm actually curious how you keep your WR in check without having some cash on hand. If you're selling AA as you go, how do you prevent overspending like buying 2 teslas every month?
User avatar
willthrill81
Posts: 32250
Joined: Thu Jan 26, 2017 2:17 pm
Location: USA
Contact:

Re: Withdrawal rate for an early retirement

Post by willthrill81 »

Marseille07 wrote: Fri Nov 27, 2020 10:11 am
willthrill81 wrote: Fri Nov 27, 2020 10:01 am
Marseille07 wrote: Fri Nov 27, 2020 10:00 am
willthrill81 wrote: Fri Nov 27, 2020 9:48 am It's not 'down the toilet', but it's bona fide mental accounting. Whether the money used to pay for an unexpected expenses comes from your EF or your portfolio doesn't matter; the point is that you no longer have it. The WR coming from your assets is the same either way.
I see value in this mental accounting and budgeting. It's OK if you don't. I'm not forcing anything, just addressing the question by EnjoyIt.
Yes, if you feel good about it, it's not going to hurt you in the instance you described.
I'm actually curious how you keep your WR in check without having some cash on hand. If you're selling AA as you go, how do you prevent overspending like buying 2 teslas every month?
Spending money is spending money, regardless which mental account you withdraw the money from. Calling one account 'EF' and another 'portfolio' doesn't change anything in that regard. For instance, if you have $1 million in your portfolio and another $50k in your EF, then you have $1.05 million. Whether you withdraw $50k from your EF or your portfolio doesn't matter; you'll still be left with $1 million either way.

We're planning on using the amortization based withdrawal method. We have no EF now and won't have one in retirement either. If we have to withdraw more than we otherwise would in the current year, the method automatically adjusts for that by reducing future withdrawals, all else being equal. So if the method indicates that based on our situation and assumptions that we could withdraw $50k but we actually withdraw $100k, then our future withdrawals will be something less than $50k, all else being equal.

As an aside, one of the really neat things about this approach is that if we want to 'front-load' our withdrawals, which we want to do since we plan on spending more in our 50s and 60s than later in life due to travel and such, this is very easily accomplished by using a larger than expected rate of return in the formula. So if we expected forward real returns of 3%, we could front-load withdrawals by computing our withdrawals with a rate of return greater than 3%, say 4% or 5%.
Last edited by willthrill81 on Fri Nov 27, 2020 10:22 am, edited 1 time in total.
The Sensible Steward
User avatar
vineviz
Posts: 14921
Joined: Tue May 15, 2018 1:55 pm
Location: Baltimore, MD

Re: Withdrawal rate for an early retirement

Post by vineviz »

YRT70 wrote: Fri Nov 27, 2020 8:49 am I think I understand you fine. I just think it's highly unlikely that these improvements are due to noise or random chance. The improvements are too large for it and the intuitive explanation of why it works also makes sense to me. If you got data showing that this is just noise I'd like to take a look at it.
Take the 1/31/1966 result as an example. The SWR for people who retired in precisely that month with a 60/40 portfolio was 3.84%. Simplistically matching that allocation requires a glide from 20% equity to 100% equity (to average 60% over the same period). In order to represent a statistically significant improvement in the SWR, you'd need an increase of about 0.34%. The actual improvement is only 0.14%.

In other words, noise.

Another way to illustrate the problem is simply to roll the retirement date forward or back by a few months. ERN reported a (statistically insignificant) improvement for retirements starting on precisely 1/31/1966. The SWR for retirements starting six months earlier or six months later was basically the same as 1/31/1966 using the static 60/40 allocation, and for those retirees the rising equity glide path produced either no improvement or modest (statistically insignificant) reduction in SWR.

In other words, overfitting.

Another way to even more simply conceptualize the problem with the ERN analysis is this: we knew BEFORE the analysis that these were the three WORST periods for a static allocation in the sample. The likelihood of finding a better strategy for THOSE PARTICULAR PERIODS by testing a bunch of random allocation glide paths is pretty high, simply due to the chance that a random strategy will outperform a strategy you already know does poorly.

In other words, data mining.
"Far more money has been lost by investors preparing for corrections than has been lost in corrections themselves." ~~ Peter Lynch
User avatar
willthrill81
Posts: 32250
Joined: Thu Jan 26, 2017 2:17 pm
Location: USA
Contact:

Re: Withdrawal rate for an early retirement

Post by willthrill81 »

vineviz wrote: Fri Nov 27, 2020 10:22 am Another way to even more simply conceptualize the problem with the ERN analysis is this: we knew BEFORE the analysis that these were the three WORST periods for a static allocation in the sample. The likelihood of finding a better strategy for THOSE PARTICULAR PERIODS by testing a bunch of random allocation glide paths is pretty high, simply due to the chance that a random strategy will outperform a strategy you already know does poorly.
And that's an inherent problem with the SWR methodology in general; by making any improvement to the one worst period on record, the SWR will, by definition, go up. But you have no idea that what worked in the single worst historic period will work in an equally poor situation going forward; in truth, the likelihood that it will work as well is low.

That's why I would like to see this method tested with historic data from other countries.
Last edited by willthrill81 on Fri Nov 27, 2020 10:25 am, edited 1 time in total.
The Sensible Steward
Marseille07
Posts: 16054
Joined: Fri Nov 06, 2020 12:41 pm

Re: Withdrawal rate for an early retirement

Post by Marseille07 »

willthrill81 wrote: Fri Nov 27, 2020 10:19 am
Marseille07 wrote: Fri Nov 27, 2020 10:11 am
willthrill81 wrote: Fri Nov 27, 2020 10:01 am
Marseille07 wrote: Fri Nov 27, 2020 10:00 am
willthrill81 wrote: Fri Nov 27, 2020 9:48 am It's not 'down the toilet', but it's bona fide mental accounting. Whether the money used to pay for an unexpected expenses comes from your EF or your portfolio doesn't matter; the point is that you no longer have it. The WR coming from your assets is the same either way.
I see value in this mental accounting and budgeting. It's OK if you don't. I'm not forcing anything, just addressing the question by EnjoyIt.
Yes, if you feel good about it, it's not going to hurt you in the instance you described.
I'm actually curious how you keep your WR in check without having some cash on hand. If you're selling AA as you go, how do you prevent overspending like buying 2 teslas every month?
Spending money is spending money, regardless which mental account you withdraw the money from. Calling one account 'EF' and another 'portfolio' doesn't change anything in that regard. For instance, if you have $1 million in your portfolio and another $50k in your EF, then you have $1.05 million. Whether you withdraw $50k from your EF or your portfolio doesn't matter; you'll still be left with $1 million either way.

We're planning on using the amortization based withdrawal method. We have no EF now and won't have one in retirement either. If we have to withdraw more than we otherwise would in the current year, the method automatically adjusts for that by reducing future withdrawals, all else being equal. So if the method indicates that based on our situation and assumptions that we could withdraw $50k but we actually withdraw $100k, then our future withdrawals will be something less than $50k, all else being equal.
I see, that makes sense. I think the difference between you and I is that you're treating WR as something long-term to eventually balance out; I'm looking at it as the upper limit. To each their own though.
User avatar
willthrill81
Posts: 32250
Joined: Thu Jan 26, 2017 2:17 pm
Location: USA
Contact:

Re: Withdrawal rate for an early retirement

Post by willthrill81 »

Marseille07 wrote: Fri Nov 27, 2020 10:25 am
willthrill81 wrote: Fri Nov 27, 2020 10:19 am
Marseille07 wrote: Fri Nov 27, 2020 10:11 am
willthrill81 wrote: Fri Nov 27, 2020 10:01 am
Marseille07 wrote: Fri Nov 27, 2020 10:00 am

I see value in this mental accounting and budgeting. It's OK if you don't. I'm not forcing anything, just addressing the question by EnjoyIt.
Yes, if you feel good about it, it's not going to hurt you in the instance you described.
I'm actually curious how you keep your WR in check without having some cash on hand. If you're selling AA as you go, how do you prevent overspending like buying 2 teslas every month?
Spending money is spending money, regardless which mental account you withdraw the money from. Calling one account 'EF' and another 'portfolio' doesn't change anything in that regard. For instance, if you have $1 million in your portfolio and another $50k in your EF, then you have $1.05 million. Whether you withdraw $50k from your EF or your portfolio doesn't matter; you'll still be left with $1 million either way.

We're planning on using the amortization based withdrawal method. We have no EF now and won't have one in retirement either. If we have to withdraw more than we otherwise would in the current year, the method automatically adjusts for that by reducing future withdrawals, all else being equal. So if the method indicates that based on our situation and assumptions that we could withdraw $50k but we actually withdraw $100k, then our future withdrawals will be something less than $50k, all else being equal.
I see, that makes sense. I think the difference between you and I is that you're treating WR as something long-term to eventually balance out; I'm looking at it as the upper limit. To each their own though.
Going back to your example, what will you do if/when your EF is exhausted in retirement? This is the identical problem that others who also use the 'bucket' strategy face: whether/how you will refill buckets in a consistent, logical, historically sound manner.
The Sensible Steward
Marseille07
Posts: 16054
Joined: Fri Nov 06, 2020 12:41 pm

Re: Withdrawal rate for an early retirement

Post by Marseille07 »

willthrill81 wrote: Fri Nov 27, 2020 10:27 am
Marseille07 wrote: Fri Nov 27, 2020 10:25 am
willthrill81 wrote: Fri Nov 27, 2020 10:19 am
Marseille07 wrote: Fri Nov 27, 2020 10:11 am
willthrill81 wrote: Fri Nov 27, 2020 10:01 am

Yes, if you feel good about it, it's not going to hurt you in the instance you described.
I'm actually curious how you keep your WR in check without having some cash on hand. If you're selling AA as you go, how do you prevent overspending like buying 2 teslas every month?
Spending money is spending money, regardless which mental account you withdraw the money from. Calling one account 'EF' and another 'portfolio' doesn't change anything in that regard. For instance, if you have $1 million in your portfolio and another $50k in your EF, then you have $1.05 million. Whether you withdraw $50k from your EF or your portfolio doesn't matter; you'll still be left with $1 million either way.

We're planning on using the amortization based withdrawal method. We have no EF now and won't have one in retirement either. If we have to withdraw more than we otherwise would in the current year, the method automatically adjusts for that by reducing future withdrawals, all else being equal. So if the method indicates that based on our situation and assumptions that we could withdraw $50k but we actually withdraw $100k, then our future withdrawals will be something less than $50k, all else being equal.
I see, that makes sense. I think the difference between you and I is that you're treating WR as something long-term to eventually balance out; I'm looking at it as the upper limit. To each their own though.
Going back to your example, what will you do if/when your EF is exhausted in retirement? This is the identical problem that others who also use the 'bucket' strategy face: whether/how you will refill buckets in a consistent, logical, historically sound manner.
Not sure if I'm following. EF of 100K (let's say) is a bucket of money at which point you stop withdrawing money. It's very unlikely to exhaust at all, because once it dips then you're going to withdraw 6.7K/mo or whatever from AA to fill it up.

Even if the EF somehow exhausts, the situation is still better than withdrawing 4% the whole time because you weren't withdrawing while EF was full, which has lowered your effective WR below 4% thus your portfolio grew more.
User avatar
willthrill81
Posts: 32250
Joined: Thu Jan 26, 2017 2:17 pm
Location: USA
Contact:

Re: Withdrawal rate for an early retirement

Post by willthrill81 »

Marseille07 wrote: Fri Nov 27, 2020 10:33 am
willthrill81 wrote: Fri Nov 27, 2020 10:27 am
Marseille07 wrote: Fri Nov 27, 2020 10:25 am
willthrill81 wrote: Fri Nov 27, 2020 10:19 am
Marseille07 wrote: Fri Nov 27, 2020 10:11 am

I'm actually curious how you keep your WR in check without having some cash on hand. If you're selling AA as you go, how do you prevent overspending like buying 2 teslas every month?
Spending money is spending money, regardless which mental account you withdraw the money from. Calling one account 'EF' and another 'portfolio' doesn't change anything in that regard. For instance, if you have $1 million in your portfolio and another $50k in your EF, then you have $1.05 million. Whether you withdraw $50k from your EF or your portfolio doesn't matter; you'll still be left with $1 million either way.

We're planning on using the amortization based withdrawal method. We have no EF now and won't have one in retirement either. If we have to withdraw more than we otherwise would in the current year, the method automatically adjusts for that by reducing future withdrawals, all else being equal. So if the method indicates that based on our situation and assumptions that we could withdraw $50k but we actually withdraw $100k, then our future withdrawals will be something less than $50k, all else being equal.
I see, that makes sense. I think the difference between you and I is that you're treating WR as something long-term to eventually balance out; I'm looking at it as the upper limit. To each their own though.
Going back to your example, what will you do if/when your EF is exhausted in retirement? This is the identical problem that others who also use the 'bucket' strategy face: whether/how you will refill buckets in a consistent, logical, historically sound manner.
Not sure if I'm following. EF of 100K (let's say) is a bucket of money at which point you stop withdrawing money. It's very unlikely to exhaust at all, because once it dips then you're going to withdraw 6.7K/mo or whatever from AA to fill it up.

Even if the EF somehow exhausts, the situation is still better than withdrawing 4% the whole time because you weren't withdrawing while EF was full, which has lowered your effective WR below 4% thus your portfolio grew more.
It seems that you're trying to keep a 'separate' (but not really in effect) EF so you can keep your 'portfolio' withdrawals from exceeding a certain threshold. As I said, that's unlikely to hurt you much, but I don't see how you believe that it would benefit you either. By not investing the $100k, you're missing out on the potential gains on it.

As noted further up thread, using a 'cash bucket' to withdraw from rather than one's portfolio would not have been of significant historic benefit to retirees (i.e. the outcome would have been virtually identical whether the cash bucket was used or not).
The Sensible Steward
EnjoyIt
Posts: 8244
Joined: Sun Dec 29, 2013 7:06 pm

Re: Withdrawal rate for an early retirement

Post by EnjoyIt »

Marseille07 wrote: Fri Nov 27, 2020 9:34 am
EnjoyIt wrote: Fri Nov 27, 2020 3:22 am
Marseille07 wrote: Thu Nov 26, 2020 9:46 pm
willthrill81 wrote: Thu Nov 26, 2020 9:39 pm
Marseille07 wrote: Thu Nov 26, 2020 9:29 pm

How are you measuring 'best' and 'accomplishing the most'?
The reason I bracketed that is because it's obviously a subjective assessment. However, the advantages of the ABW method are numerous. It is a mathematically driven approach that will not prematurely deplete a portfolio, can be easily adapted to current market conditions (i.e. by incorporating the current portfolio value and forward return estimates), and can incorporate multiple income streams and/or expenses at various points in time. I know of no other approach that can do all of that, and I've seen a lot of withdrawal methods. About the only real disadvantage of the approach is that it can get a bit complex, especially if your financial situation is complicated, but other users have developed easy to use spreadsheets that greatly help. Another potential disadvantage is that it may lead to what turns out to be unnecessarily volatile withdrawals, but this is true of any flexible withdrawal method; you don't know that the flexibility was unnecessary until after the fact.
I see, I think you described fair points on ABW. I mentioned elsewhere but volatile withdrawals are a non-issue with proper EF, at least for me.
I don’t understand how an emergency fund helps you with that. Can you please elaborate? When do you use the fund? How do you replenish the fund?
The poster is concerned about unexpected 50K in expenditure, which they don't know when / if it will happen.

You build up EF simply by saving some of your monthly withdrawals. For example, let's say you have 2M and doing 4% = 6.7K/mo. You only spend 4K/mo and put the rest of cash in EF, which builds up over time.

I know you regard this EF as uninvested money down the toilet. However, this approach keeps your WR intact while providing liquidity in case of emergencies.
If all you need to spend is $4k/month then why pull extra you don’t need? Why not just pull what you need from your portfolio when you actually need it and not months or years in advance? I just don’t understand the advantage of your strategy to just keep removing money from your portfolio that you do not intend to use.

$50k outside of your 60/40 portfolio could be earning a historic average of 8.7%-3% inflation. If returns are muted we can estimate maybe 5% just to make math simple. Having that $50k emergency fund costs $2.5k a year plus any future growth on that $2.5k. To me that is very expensive mental accounting with no real benefit. Even at 4% return this mental accounting costs $2k a year. I don’t know about you, but if I’m living on an average $80k a year, I would hate to waste $2k on mind games.
A time to EVALUATE your jitters: | viewtopic.php?p=1139732#p1139732
Marseille07
Posts: 16054
Joined: Fri Nov 06, 2020 12:41 pm

Re: Withdrawal rate for an early retirement

Post by Marseille07 »

willthrill81 wrote: Fri Nov 27, 2020 10:39 am It seems that you're trying to keep a 'separate' (but not really in effect) EF so you can keep your 'portfolio' withdrawals from exceeding a certain threshold. As I said, that's unlikely to hurt you much, but I don't see how you believe that it would benefit you either. By not investing the $100k, you're missing out on the potential gains on it.

As noted further up thread, using a 'cash bucket' to withdraw from rather than one's portfolio would not have been of significant historic benefit to retirees (i.e. the outcome would have been virtually identical whether the cash bucket was used or not).
Kind of. As I said, I see WR as the upper limit, not something to adjust annually. So I withdraw X% upfront then set aside leftover cash as EF until the bucket is full. Sure I'm missing on the potential gains, but if I'm safe to withdraw X% anyway then that's fine from a withdrawal perspective.
Last edited by Marseille07 on Fri Nov 27, 2020 10:55 am, edited 1 time in total.
Marseille07
Posts: 16054
Joined: Fri Nov 06, 2020 12:41 pm

Re: Withdrawal rate for an early retirement

Post by Marseille07 »

EnjoyIt wrote: Fri Nov 27, 2020 10:46 am
Marseille07 wrote: Fri Nov 27, 2020 9:34 am
EnjoyIt wrote: Fri Nov 27, 2020 3:22 am
Marseille07 wrote: Thu Nov 26, 2020 9:46 pm
willthrill81 wrote: Thu Nov 26, 2020 9:39 pm

The reason I bracketed that is because it's obviously a subjective assessment. However, the advantages of the ABW method are numerous. It is a mathematically driven approach that will not prematurely deplete a portfolio, can be easily adapted to current market conditions (i.e. by incorporating the current portfolio value and forward return estimates), and can incorporate multiple income streams and/or expenses at various points in time. I know of no other approach that can do all of that, and I've seen a lot of withdrawal methods. About the only real disadvantage of the approach is that it can get a bit complex, especially if your financial situation is complicated, but other users have developed easy to use spreadsheets that greatly help. Another potential disadvantage is that it may lead to what turns out to be unnecessarily volatile withdrawals, but this is true of any flexible withdrawal method; you don't know that the flexibility was unnecessary until after the fact.
I see, I think you described fair points on ABW. I mentioned elsewhere but volatile withdrawals are a non-issue with proper EF, at least for me.
I don’t understand how an emergency fund helps you with that. Can you please elaborate? When do you use the fund? How do you replenish the fund?
The poster is concerned about unexpected 50K in expenditure, which they don't know when / if it will happen.

You build up EF simply by saving some of your monthly withdrawals. For example, let's say you have 2M and doing 4% = 6.7K/mo. You only spend 4K/mo and put the rest of cash in EF, which builds up over time.

I know you regard this EF as uninvested money down the toilet. However, this approach keeps your WR intact while providing liquidity in case of emergencies.
If all you need to spend is $4k/month then why pull extra you don’t need? Why not just pull what you need from your portfolio when you actually need it and not months or years in advance? I just don’t understand the advantage of your strategy to just keep removing money from your portfolio that you do not intend to use.

$50k outside of your 60/40 portfolio could be earning a historic average of 8.7%-3% inflation. If returns are muted we can estimate maybe 5% just to make math simple. Having that $50k emergency fund costs $2.5k a year plus any future growth on that $2.5k. To me that is very expensive mental accounting with no real benefit. Even at 4% return this mental accounting costs $2k a year. I don’t know about you, but if I’m living on an average $80k a year, I would hate to waste $2k on mind games.
I think that's fair, I mean at the end of the day it's a trade-off between liquidity and potential gains.
User avatar
willthrill81
Posts: 32250
Joined: Thu Jan 26, 2017 2:17 pm
Location: USA
Contact:

Re: Withdrawal rate for an early retirement

Post by willthrill81 »

Marseille07 wrote: Fri Nov 27, 2020 10:52 am
willthrill81 wrote: Fri Nov 27, 2020 10:39 am It seems that you're trying to keep a 'separate' (but not really in effect) EF so you can keep your 'portfolio' withdrawals from exceeding a certain threshold. As I said, that's unlikely to hurt you much, but I don't see how you believe that it would benefit you either. By not investing the $100k, you're missing out on the potential gains on it.

As noted further up thread, using a 'cash bucket' to withdraw from rather than one's portfolio would not have been of significant historic benefit to retirees (i.e. the outcome would have been virtually identical whether the cash bucket was used or not).
Kind of. As I said, I see WR as the upper limit, not something to adjust annually.
Why?
The Sensible Steward
EnjoyIt
Posts: 8244
Joined: Sun Dec 29, 2013 7:06 pm

Re: Withdrawal rate for an early retirement

Post by EnjoyIt »

Marseille07 wrote: Fri Nov 27, 2020 10:54 am
EnjoyIt wrote: Fri Nov 27, 2020 10:46 am
Marseille07 wrote: Fri Nov 27, 2020 9:34 am
EnjoyIt wrote: Fri Nov 27, 2020 3:22 am
Marseille07 wrote: Thu Nov 26, 2020 9:46 pm

I see, I think you described fair points on ABW. I mentioned elsewhere but volatile withdrawals are a non-issue with proper EF, at least for me.
I don’t understand how an emergency fund helps you with that. Can you please elaborate? When do you use the fund? How do you replenish the fund?
The poster is concerned about unexpected 50K in expenditure, which they don't know when / if it will happen.

You build up EF simply by saving some of your monthly withdrawals. For example, let's say you have 2M and doing 4% = 6.7K/mo. You only spend 4K/mo and put the rest of cash in EF, which builds up over time.

I know you regard this EF as uninvested money down the toilet. However, this approach keeps your WR intact while providing liquidity in case of emergencies.
If all you need to spend is $4k/month then why pull extra you don’t need? Why not just pull what you need from your portfolio when you actually need it and not months or years in advance? I just don’t understand the advantage of your strategy to just keep removing money from your portfolio that you do not intend to use.

$50k outside of your 60/40 portfolio could be earning a historic average of 8.7%-3% inflation. If returns are muted we can estimate maybe 5% just to make math simple. Having that $50k emergency fund costs $2.5k a year plus any future growth on that $2.5k. To me that is very expensive mental accounting with no real benefit. Even at 4% return this mental accounting costs $2k a year. I don’t know about you, but if I’m living on an average $80k a year, I would hate to waste $2k on mind games.
I think that's fair, I mean at the end of the day it's a trade-off between liquidity and potential gains.
Just thinking about a solution for those who like buckets.

How about a spreadsheet that takes this EF fund into account. Instead of physically taking the extra cash out every month, you allocate the unused portion in a spread sheet. This can be calculated quarterly and will provide the best of both worlds. Money stays invested, but still earmarked for said emergency. I think taking 20 minutes a quarter to run some tallies is well worth $2k or more a year.
A time to EVALUATE your jitters: | viewtopic.php?p=1139732#p1139732
Marseille07
Posts: 16054
Joined: Fri Nov 06, 2020 12:41 pm

Re: Withdrawal rate for an early retirement

Post by Marseille07 »

willthrill81 wrote: Fri Nov 27, 2020 10:55 am
Marseille07 wrote: Fri Nov 27, 2020 10:52 am
willthrill81 wrote: Fri Nov 27, 2020 10:39 am It seems that you're trying to keep a 'separate' (but not really in effect) EF so you can keep your 'portfolio' withdrawals from exceeding a certain threshold. As I said, that's unlikely to hurt you much, but I don't see how you believe that it would benefit you either. By not investing the $100k, you're missing out on the potential gains on it.

As noted further up thread, using a 'cash bucket' to withdraw from rather than one's portfolio would not have been of significant historic benefit to retirees (i.e. the outcome would have been virtually identical whether the cash bucket was used or not).
Kind of. As I said, I see WR as the upper limit, not something to adjust annually.
Why?
Because it is the upper limit...literally. Let me give you a contrived example.

Say you have 2M and ABW says 4.5%. You spent 1M, way overspending; you adjust ABW, now it says 0.5%. You spent another 1M, now out of money. Well, what is this person doing exactly?

At the end of the day, you have to treat the percentage for what it is - the upper limit.
User avatar
CalPoppy
Posts: 313
Joined: Thu May 04, 2017 2:57 pm

Re: Withdrawal rate for an early retirement

Post by CalPoppy »

Derpalator wrote: Fri Nov 27, 2020 4:18 am
Rob1 wrote: Thu Nov 26, 2020 11:39 am
Rob1 wrote: Thu Nov 26, 2020 11:13 am For some perspective:

When the CFPs at Vanguard Personal Advisory Services conduct a portfolio analysis, they like to see an estimated success rate of 80% or better.

And Bernstein states: "But history teaches us that depriving ourselves to boost our 40-year success probability much beyond 80% is a fool’s errand, since all you are doing is increasing the probability of failure for political, economic, and military reasons relative to the failure of banal financial planning." More context here: http://www.efficientfrontier.com/ef/901/hell3.htm
And I’ll add: Targeting a 100% SWR based on all sorts of negative outlier assumptions (whether it be expenses, returns, life expectancy...) is too conservative for me. Particularly when other good solutions exist - such as adding expense flexibility and/or bit of human capital income, or using ABW.
Rob 1, I agree with both the first and second parts of your post. I initially wrote specifics about my own calculations but you say it more simply and better. Some say cash is king, I say FLEXIBILITY is king.

As an aside, I never would have had the wherewithal to grok personal finance had I not discovered this forum. Thanks and many blessings to all.
:sharebeer
Marseille07
Posts: 16054
Joined: Fri Nov 06, 2020 12:41 pm

Re: Withdrawal rate for an early retirement

Post by Marseille07 »

EnjoyIt wrote: Fri Nov 27, 2020 11:06 am
Marseille07 wrote: Fri Nov 27, 2020 10:54 am
EnjoyIt wrote: Fri Nov 27, 2020 10:46 am
Marseille07 wrote: Fri Nov 27, 2020 9:34 am
EnjoyIt wrote: Fri Nov 27, 2020 3:22 am

I don’t understand how an emergency fund helps you with that. Can you please elaborate? When do you use the fund? How do you replenish the fund?
The poster is concerned about unexpected 50K in expenditure, which they don't know when / if it will happen.

You build up EF simply by saving some of your monthly withdrawals. For example, let's say you have 2M and doing 4% = 6.7K/mo. You only spend 4K/mo and put the rest of cash in EF, which builds up over time.

I know you regard this EF as uninvested money down the toilet. However, this approach keeps your WR intact while providing liquidity in case of emergencies.
If all you need to spend is $4k/month then why pull extra you don’t need? Why not just pull what you need from your portfolio when you actually need it and not months or years in advance? I just don’t understand the advantage of your strategy to just keep removing money from your portfolio that you do not intend to use.

$50k outside of your 60/40 portfolio could be earning a historic average of 8.7%-3% inflation. If returns are muted we can estimate maybe 5% just to make math simple. Having that $50k emergency fund costs $2.5k a year plus any future growth on that $2.5k. To me that is very expensive mental accounting with no real benefit. Even at 4% return this mental accounting costs $2k a year. I don’t know about you, but if I’m living on an average $80k a year, I would hate to waste $2k on mind games.
I think that's fair, I mean at the end of the day it's a trade-off between liquidity and potential gains.
Just thinking about a solution for those who like buckets.

How about a spreadsheet that takes this EF fund into account. Instead of physically taking the extra cash out every month, you allocate the unused portion in a spread sheet. This can be calculated quarterly and will provide the best of both worlds. Money stays invested, but still earmarked for said emergency. I think taking 20 minutes a quarter to run some tallies is well worth $2k or more a year.
Yes, this is a very good solution, I like it. Let me think about how it might look like.
EnjoyIt
Posts: 8244
Joined: Sun Dec 29, 2013 7:06 pm

Re: Withdrawal rate for an early retirement

Post by EnjoyIt »

Marseille07 wrote: Fri Nov 27, 2020 11:15 am
EnjoyIt wrote: Fri Nov 27, 2020 11:06 am
Marseille07 wrote: Fri Nov 27, 2020 10:54 am
EnjoyIt wrote: Fri Nov 27, 2020 10:46 am
Marseille07 wrote: Fri Nov 27, 2020 9:34 am

The poster is concerned about unexpected 50K in expenditure, which they don't know when / if it will happen.

You build up EF simply by saving some of your monthly withdrawals. For example, let's say you have 2M and doing 4% = 6.7K/mo. You only spend 4K/mo and put the rest of cash in EF, which builds up over time.

I know you regard this EF as uninvested money down the toilet. However, this approach keeps your WR intact while providing liquidity in case of emergencies.
If all you need to spend is $4k/month then why pull extra you don’t need? Why not just pull what you need from your portfolio when you actually need it and not months or years in advance? I just don’t understand the advantage of your strategy to just keep removing money from your portfolio that you do not intend to use.

$50k outside of your 60/40 portfolio could be earning a historic average of 8.7%-3% inflation. If returns are muted we can estimate maybe 5% just to make math simple. Having that $50k emergency fund costs $2.5k a year plus any future growth on that $2.5k. To me that is very expensive mental accounting with no real benefit. Even at 4% return this mental accounting costs $2k a year. I don’t know about you, but if I’m living on an average $80k a year, I would hate to waste $2k on mind games.
I think that's fair, I mean at the end of the day it's a trade-off between liquidity and potential gains.
Just thinking about a solution for those who like buckets.

How about a spreadsheet that takes this EF fund into account. Instead of physically taking the extra cash out every month, you allocate the unused portion in a spread sheet. This can be calculated quarterly and will provide the best of both worlds. Money stays invested, but still earmarked for said emergency. I think taking 20 minutes a quarter to run some tallies is well worth $2k or more a year.
Yes, this is a very good solution, I like it. Let me think about how it might look like.
That’s awesome. Please share your thoughts on it in the future.
A time to EVALUATE your jitters: | viewtopic.php?p=1139732#p1139732
User avatar
willthrill81
Posts: 32250
Joined: Thu Jan 26, 2017 2:17 pm
Location: USA
Contact:

Re: Withdrawal rate for an early retirement

Post by willthrill81 »

Marseille07 wrote: Fri Nov 27, 2020 11:10 am
willthrill81 wrote: Fri Nov 27, 2020 10:55 am
Marseille07 wrote: Fri Nov 27, 2020 10:52 am
willthrill81 wrote: Fri Nov 27, 2020 10:39 am It seems that you're trying to keep a 'separate' (but not really in effect) EF so you can keep your 'portfolio' withdrawals from exceeding a certain threshold. As I said, that's unlikely to hurt you much, but I don't see how you believe that it would benefit you either. By not investing the $100k, you're missing out on the potential gains on it.

As noted further up thread, using a 'cash bucket' to withdraw from rather than one's portfolio would not have been of significant historic benefit to retirees (i.e. the outcome would have been virtually identical whether the cash bucket was used or not).
Kind of. As I said, I see WR as the upper limit, not something to adjust annually.
Why?
Because it is the upper limit...literally. Let me give you a contrived example.

Say you have 2M and ABW says 4.5%. You spent 1M, way overspending; you adjust ABW, now it says 0.5%. You spent another 1M, now out of money. Well, what is this person doing exactly?

At the end of the day, you have to treat the percentage for what it is - the upper limit.
That's such an extreme example that I don't see the value in it. No withdrawal method will save you from reckless spending.

If the ABW says you can withdraw 4.5% and you withdraw more than that, it just means that your future withdrawals according to what ABW indicates you can withdraw will be lower than 4.5%, all else being equal.
The Sensible Steward
Post Reply