Withdrawal rate for an early retirement
Re: Withdrawal rate for an early retirement
I'm planning to early retire at 40.
My method is to take the expected real return, reduce it by a third, and that's the withdrawal rate.
So if expected real return is 6%, then withdrawal rate is 4%. Right now withdrawal rate is 3.7%.
Expected (nominal) return of bonds is simply their yield. Expected inflation is 2%.
Expected real return of stocks is 100 divided by the rolling average of their PE ratio over 10 years or so. Right now that's 5.6% real for a global portfolio.
My method is to take the expected real return, reduce it by a third, and that's the withdrawal rate.
So if expected real return is 6%, then withdrawal rate is 4%. Right now withdrawal rate is 3.7%.
Expected (nominal) return of bonds is simply their yield. Expected inflation is 2%.
Expected real return of stocks is 100 divided by the rolling average of their PE ratio over 10 years or so. Right now that's 5.6% real for a global portfolio.
-
- Posts: 3509
- Joined: Sun Jan 07, 2018 11:52 am
Re: Withdrawal rate for an early retirement
That does not even makes sense since the dividend rate itself of my portfolio is around 2% so how on earth did he arrive at 0.5% LOL! Surely he did not use math.Marseille07 wrote: ↑Tue Jun 01, 2021 11:52 pm I wouldn't use 4% SWR in retirement, especially for an early retirement. Financial Samurai recommends 0.5% WR as our bond yields are really low.
I don't buy the low bond yield warnings, the reasoning for me is that capital has to flow somewhere to chase returns. If bond yields are low then capital will go to bid equities and other assets higher which is precisely what is happening now. It's a zero sum game. People don't put their money under their mattress.
Re: Withdrawal rate for an early retirement
It isn't "happening", it's already happened. TINA inflates the price of stocks, but it does so at the expense of long term returns - TINA doesn't make underlying earnings any higher. The chickens are laying the same size eggs, but they're more expensive to buy. This bodes ill for medium-term returns. This is NOT a zero sum game.. the time value of money can legitimately go down.stocknoob4111 wrote: ↑Tue Jul 06, 2021 8:54 am I don't buy the low bond yield warnings, the reasoning for me is that capital has to flow somewhere to chase returns. If bond yields are low then capital will go to bid equities and other assets higher which is precisely what is happening now.
-
- Posts: 16054
- Joined: Fri Nov 06, 2020 12:41 pm
Re: Withdrawal rate for an early retirement
I think those are good approaches. I'll see if cFireSim supports spending floor-type deals in conjunction with constant-% configuration.marcopolo wrote: ↑Tue Jul 06, 2021 12:40 am One approach that I think would be reasonable is to set a floor on withdrawals. That is a minimum amount you are able/willing to live on in a given year. Then, test to see if the constant percentage withdrawal amount ever falls below that threshold. An alternate approach is to use a constant percentage, with a constant dollar floor on withdrawals in any given year, and then see if the portfolio survived.
The two sound similar, but are a bit different. In the first case, if in any given year the constant percentage withdrawal would result in breaching your threshold, you declare a failure, even if the portfolio goes on to great gains in future years.
In the second, in such years, you go ahead and withdraw the floor amount even if it exceeds your target percentage. Then, if the market recovers, you are back to using constant percentage. Failure only occurs if portfolio runs out of money.
-
- Posts: 16054
- Joined: Fri Nov 06, 2020 12:41 pm
Re: Withdrawal rate for an early retirement
Yeah that's the thing, as far as I've run sims, even as high as 5.5% saw some portfolio growth (on average case), and of course constant-% never runs out of money. So then, how do we decide what's too high vs too low is really what I'm trying to measure objectively. Not easy, I guessAlohaJoe wrote: ↑Tue Jul 06, 2021 1:11 amUtility functions, certainty equivalent withdrawals, semideviations, the ulcer index. None of them have quite the same simple intuitive appeal of "failure rates", though.Marseille07 wrote: ↑Tue Jul 06, 2021 12:08 am My question is, what are some good ways to evaluate various constant-percentage withdrawal rates? Obviously the lower the WR, the more initial cost to generate the same amount of expense but safer (for example, if you need 100K/year then 4% needs 2.5M, 2% needs 5M, 1% needs 10M etc etc). But unlike constant-dollar SWR, there's no equivalent of "success rate" in constant-percentage. Any metrics we can use to objectively compare various withdrawal rates?
-
- Posts: 16054
- Joined: Fri Nov 06, 2020 12:41 pm
Re: Withdrawal rate for an early retirement
Yeah, and I fully accept that possibility. Discretionary spending budget is a must, and this will be cut during the downturns if you walk right into it.canadianbacon wrote: ↑Tue Jul 06, 2021 8:22 amIt is essentially VPW but more conservative because with VPW your WR increases (slowly) over your lifetime.Marseille07 wrote: ↑Tue Jul 06, 2021 12:08 am Has anyone studied "constant-percentage" withdrawal method for early retirees?
As far as I've run sims, we can actually go as high as 5.5% WR for a 30-year retirement horizon. While the portfolio doesn't grow much (obviously), it allows one to withdraw quite a bit of cash throughout the horizon. This sounds high but actually in-line with the VPW table.
As with VPW, it's a good idea to do some backtesting. While it's true that you can't easily deplete the portfolio, you may end up living on (for example) 60% of your starting budget for a decade or more.
-
- Posts: 16054
- Joined: Fri Nov 06, 2020 12:41 pm
Re: Withdrawal rate for an early retirement
I'll see what Merriman said. 6% is *fine* in that you can withdraw a lot of cash on a low initial cost, at the expense of your portfolio not growing much. I *feel* that's too high, personally, but trying to quantify why. So far my conclusion is basically "if you can afford to accumulate a lot of money, the lower WR, the better."David Jay wrote: ↑Tue Jul 06, 2021 8:34 amConstant percentage withdrawals work fine mathematically. Plus, you can never fully deplete your portfolio (math joke).Marseille07 wrote: ↑Tue Jul 06, 2021 12:08 amHas anyone studied "constant-percentage" withdrawal method for early retirees?
As far as I've run sims, we can actually go as high as 5.5% WR for a 30-year retirement horizon. While the portfolio doesn't grow much (obviously), it allows one to withdraw quite a bit of cash throughout the horizon. This sounds high but actually in-line with the VPW table.
Merriman is in the same ballpark with you - his data looks closer to 6% of remaining portfolio. The problem is the wild swings in withdrawal amount. For someone using those withdrawals directly for living expenses, seeing a 30% drop in cash flow from one year to the next is ugly.
-
- Posts: 16054
- Joined: Fri Nov 06, 2020 12:41 pm
Re: Withdrawal rate for an early retirement
Yeah 0.5% is way too conservative. I followed up and corrected myself upthread.stocknoob4111 wrote: ↑Tue Jul 06, 2021 8:54 amThat does not even makes sense since the dividend rate itself of my portfolio is around 2% so how on earth did he arrive at 0.5% LOL! Surely he did not use math.Marseille07 wrote: ↑Tue Jun 01, 2021 11:52 pm I wouldn't use 4% SWR in retirement, especially for an early retirement. Financial Samurai recommends 0.5% WR as our bond yields are really low.
I don't buy the low bond yield warnings, the reasoning for me is that capital has to flow somewhere to chase returns. If bond yields are low then capital will go to bid equities and other assets higher which is precisely what is happening now. It's a zero sum game. People don't put their money under their mattress.
Re: Withdrawal rate for an early retirement
I think it is pretty simple, two items explain almost all of the difference between 4% SWR and 6% of portfolio balance:Marseille07 wrote: ↑Tue Jul 06, 2021 9:42 amI *feel* that's too high, personally, but trying to quantify why.
1. By definition (because the withdrawal amount is fixed and the future is unknown), the 4% SWR must take into account the worst of future outcomes. The vast majority of those who withdraw along the general lines of a 4% SWR will have very happy heirs - they will leave behind huge portfolio balances. Look at any of the "spagetti" traces in any of the retirement calculators. Looking at historical data, pretty much only the 1966 - 1968 cadre challenge the sustainability of the 4% SWR. By contrast, the percentage withdrawal has a dynamic adjustment for poor future returns - the retiree withdraws less in down years. And presumably has less to live on...
2. 4% SWR has the built-in inflator which will result in withdrawing greater amounts from the portfolio over the decades. 6% of remaining portfolio does not.
[edit] BTW - speaking of worst cases, what happens to the 6% of portfolio rule if the market has a 50% downturn in the first year with a 60/40 portfolio? Interesting, no?
[second edit] To be more explicit for the broader audience: in the worst case (50% drop the first year, 2% inflation) the 4% SWR withdraws 4.02% of initial portfolio. The 6% of remaining portfolio, with a 60/40 sees a 30% drop, so the 6% rule withdraws 4.12% of initial portfolio. Very similar result from a worst case initial event, which flows back into item 1 above.
[third edit] Using Marseille' 5.5% of remaining portfolio instead of 6%, the withdrawal percentage would be 3.85% of initial portfolio with a 60/40 portfolio - less than the 4% SWR withdrawal of 4.02%.
Last edited by David Jay on Tue Jul 06, 2021 11:07 am, edited 3 times in total.
It's not an engineering problem - Hersh Shefrin | To get the "risk premium", you really do have to take the risk - nisiprius
-
- Posts: 16054
- Joined: Fri Nov 06, 2020 12:41 pm
Re: Withdrawal rate for an early retirement
Right. I mean 6% is way too high, but 4% of portfolio balance? 3%? Based on my study, lower-ish constant-percentage actually leaves a lot more money than SWR because SORR is mostly mitigated. You don't have a situation where the market drops by 50% and you withdraw 4% (effective WR of 7~8%) anyway.David Jay wrote: ↑Tue Jul 06, 2021 10:14 amI think it is pretty simple, two items explain almost all of the difference between 4% SWR and 6% of portfolio balance:Marseille07 wrote: ↑Tue Jul 06, 2021 9:42 amI *feel* that's too high, personally, but trying to quantify why.
1. By definition (because the withdrawal amount is fixed and the future is unknown), the 4% SWR must take into account the worst of future outcomes. The vast majority of those who withdraw along the general lines of a 4% SWR will have very happy heirs - they will leave behind huge portfolio balances. Look at any of the "spagetti" traces in any of the retirement calculators. Looking at historical data, pretty much only the 1966 - 1968 cadre challenge the sustainability of the 4% SWR. By contrast, the percentage withdrawal has a dynamic adjustment for poor future returns - the retiree withdraws less in down years. And presumably has less to live on...
2. 4% SWR has the built-in inflator which will result in withdrawing greater amounts from the portfolio over the decades. 6% of remaining portfolio does not.
Re: Withdrawal rate for an early retirement
Did you see my edit?Marseille07 wrote: ↑Tue Jul 06, 2021 10:25 amRight. I mean 6% is way too high, but 4% of portfolio balance? 3%? Based on my study, lower-ish constant-percentage actually leaves a lot more money than SWR because SORR is mostly mitigated. You don't have a situation where the market drops by 50% and you withdraw 4% (effective WR of 7~8%) anyway.
It's not an engineering problem - Hersh Shefrin | To get the "risk premium", you really do have to take the risk - nisiprius
-
- Posts: 16054
- Joined: Fri Nov 06, 2020 12:41 pm
Re: Withdrawal rate for an early retirement
I didn't, I must have been responding during. Constant-percentage can certainly run into years of poor withdrawals, and it is likely 6% is so high that your portfolio balance might not recover at all. Maybe this is something to look at in my sims.David Jay wrote: ↑Tue Jul 06, 2021 10:28 amDid you see my edit?Marseille07 wrote: ↑Tue Jul 06, 2021 10:25 amRight. I mean 6% is way too high, but 4% of portfolio balance? 3%? Based on my study, lower-ish constant-percentage actually leaves a lot more money than SWR because SORR is mostly mitigated. You don't have a situation where the market drops by 50% and you withdraw 4% (effective WR of 7~8%) anyway.
Re: Withdrawal rate for an early retirement
Not necessarily.
Depending on one's budget, during good years a retiree using automatic VPW withdrawals might end up saving money that isn't needed for expenses, thus building a cushion for down years.
Also, one can separate portfolio withdrawals from income distributions -- e.g. by having a savings account with X number of portfolio withdrawals but only distributing income based on a rolling average.
But in general I agree -- a VPW method requires some flexibility in spending compared to SWR.
-
- Posts: 16054
- Joined: Fri Nov 06, 2020 12:41 pm
Re: Withdrawal rate for an early retirement
I ran more sims and I'm liking 3.5% WR now based on p5 (the bottom 5 percentile) on 1M initial is 1.14M after 30 years. In other words, unless one's really unlucky, the portfolio balance is preserved and the average case is much much higher.
Re: Withdrawal rate for an early retirement
I know I have run the second approach I describe above in cFireSim in the past.Marseille07 wrote: ↑Tue Jul 06, 2021 9:33 amI think those are good approaches. I'll see if cFireSim supports spending floor-type deals in conjunction with constant-% configuration.marcopolo wrote: ↑Tue Jul 06, 2021 12:40 am One approach that I think would be reasonable is to set a floor on withdrawals. That is a minimum amount you are able/willing to live on in a given year. Then, test to see if the constant percentage withdrawal amount ever falls below that threshold. An alternate approach is to use a constant percentage, with a constant dollar floor on withdrawals in any given year, and then see if the portfolio survived.
The two sound similar, but are a bit different. In the first case, if in any given year the constant percentage withdrawal would result in breaching your threshold, you declare a failure, even if the portfolio goes on to great gains in future years.
In the second, in such years, you go ahead and withdraw the floor amount even if it exceeds your target percentage. Then, if the market recovers, you are back to using constant percentage. Failure only occurs if portfolio runs out of money.
Once in a while you get shown the light, in the strangest of places if you look at it right.
-
- Posts: 16054
- Joined: Fri Nov 06, 2020 12:41 pm
Re: Withdrawal rate for an early retirement
Yeah cFireSim is so good that we can simulate both scenariosmarcopolo wrote: ↑Tue Jul 06, 2021 11:49 amI know I have run the second approach I describe above in cFireSim in the past.Marseille07 wrote: ↑Tue Jul 06, 2021 9:33 amI think those are good approaches. I'll see if cFireSim supports spending floor-type deals in conjunction with constant-% configuration.marcopolo wrote: ↑Tue Jul 06, 2021 12:40 am One approach that I think would be reasonable is to set a floor on withdrawals. That is a minimum amount you are able/willing to live on in a given year. Then, test to see if the constant percentage withdrawal amount ever falls below that threshold. An alternate approach is to use a constant percentage, with a constant dollar floor on withdrawals in any given year, and then see if the portfolio survived.
The two sound similar, but are a bit different. In the first case, if in any given year the constant percentage withdrawal would result in breaching your threshold, you declare a failure, even if the portfolio goes on to great gains in future years.
In the second, in such years, you go ahead and withdraw the floor amount even if it exceeds your target percentage. Then, if the market recovers, you are back to using constant percentage. Failure only occurs if portfolio runs out of money.
Re: Withdrawal rate for an early retirement
How much of your expenses are discretionary? What happened to your spending in the past recessions. Most humans will find themselves inadvertently spending less during bad times. They will maybe travel less, delay expensive expected purchases, eat out less, just generally be a bit more thrifty. How much is that innate decreased spending worth. For example our budget is about 50%-60% discretionary. That is because we are debt free. Outside of utilities, insurance and food, everything else are wants and not needs. For us, it isn't that hard to cut spending by 25%. All of a sudden a 4% withdrawal rate is now 3%. Run some sims making a few cuts in spending during bad times and see how things fare. Cfiresim allows you to do that.Marseille07 wrote: ↑Tue Jul 06, 2021 11:40 am I ran more sims and I'm liking 3.5% WR now based on p5 (the bottom 5 percentile) on 1M initial is 1.14M after 30 years. In other words, unless one's really unlucky, the portfolio balance is preserved and the average case is much much higher.
A time to EVALUATE your jitters: |
viewtopic.php?p=1139732#p1139732
-
- Posts: 16054
- Joined: Fri Nov 06, 2020 12:41 pm
Re: Withdrawal rate for an early retirement
My discretionary spending was around 35%, I trimmed a bit and now around 25%.EnjoyIt wrote: ↑Tue Jul 06, 2021 10:16 pm How much of your expenses are discretionary? What happened to your spending in the past recessions. Most humans will find themselves inadvertently spending less during bad times. They will maybe travel less, delay expensive expected purchases, eat out less, just generally be a bit more thrifty. How much is that innate decreased spending worth. For example our budget is about 50%-60% discretionary. That is because we are debt free. Outside of utilities, insurance and food, everything else are wants and not needs. For us, it isn't that hard to cut spending by 25%. All of a sudden a 4% withdrawal rate is now 3%. Run some sims making a few cuts in spending during bad times and see how things fare. Cfiresim allows you to do that.
"All of a sudden a 4% withdrawal rate is now 3%" By this, I'm guessing you're talking about constant-dollar SWR of 4% becoming 3%. The beauty of constant-percentage approach is, your 4% stays 4% but becomes 25% less after a market crash.
Re: Withdrawal rate for an early retirement
From the reading I've done, there's no reason to plan for anything less than 3% for 40-60 year retirements. Flexibility and options for continued earning play a big role. Each situation is pretty different.
Re: Withdrawal rate for an early retirement
I just described human nature choosing to spend less during bad times and cutting back on discretionary expenses. In your example you are cutting to 3% effective spending because your 50/50 portfolio is down 50%. Reality is that when the market goes down, you would end up spending less anyways. It is just human nature. If the market goes down 15% you may very well cut your discretionary expense in half, maybe more. As a saver, your subconscious will do it for you. It is what allowed you to build all that wealth to begin with.Marseille07 wrote: ↑Tue Jul 06, 2021 10:52 pmMy discretionary spending was around 35%, I trimmed a bit and now around 25%.EnjoyIt wrote: ↑Tue Jul 06, 2021 10:16 pm How much of your expenses are discretionary? What happened to your spending in the past recessions. Most humans will find themselves inadvertently spending less during bad times. They will maybe travel less, delay expensive expected purchases, eat out less, just generally be a bit more thrifty. How much is that innate decreased spending worth. For example our budget is about 50%-60% discretionary. That is because we are debt free. Outside of utilities, insurance and food, everything else are wants and not needs. For us, it isn't that hard to cut spending by 25%. All of a sudden a 4% withdrawal rate is now 3%. Run some sims making a few cuts in spending during bad times and see how things fare. Cfiresim allows you to do that.
"All of a sudden a 4% withdrawal rate is now 3%" By this, I'm guessing you're talking about constant-dollar SWR of 4% becoming 3%. The beauty of constant-percentage approach is, your 4% stays 4% but becomes 25% less after a market crash.
I have a buddy who sold smart home equipment back in the day before there were do it yourself options. This was equipment that was put in very expensive homes. Ozzy Osbourne had such equipment as an example. He noted how when the market is down, sales decrease and when market goes up they increase. It is just human nature to spend less when the market is down.
A time to EVALUATE your jitters: |
viewtopic.php?p=1139732#p1139732
Re: Withdrawal rate for an early retirement
Im so interested in this topic, will return later today to attempt to catch up from the beginning.
-
- Posts: 16054
- Joined: Fri Nov 06, 2020 12:41 pm
Re: Withdrawal rate for an early retirement
I don't know how reliable the expected real return is. If it is reliable then it is probably a reasonable approach.Tamalak wrote: ↑Tue Jul 06, 2021 8:39 am I'm planning to early retire at 40.
My method is to take the expected real return, reduce it by a third, and that's the withdrawal rate.
So if expected real return is 6%, then withdrawal rate is 4%. Right now withdrawal rate is 3.7%.
Expected (nominal) return of bonds is simply their yield. Expected inflation is 2%.
Expected real return of stocks is 100 divided by the rolling average of their PE ratio over 10 years or so. Right now that's 5.6% real for a global portfolio.
It's kind of interesting everyone talks about 4% SWR constant-dollar, yet in reality people actually use a percentage-based method, and yet there aren't many studies done on this front compared to the 4% SWR studies.
Re: Withdrawal rate for an early retirement
It's not reliable at all. It is, I hope, the approximate center of a low and broad bell curve. That's why I reduce it by a third as a safety measure.Marseille07 wrote: ↑Wed Jul 07, 2021 9:33 amI don't know how reliable the expected real return is. If it is reliable then it is probably a reasonable approach.Tamalak wrote: ↑Tue Jul 06, 2021 8:39 am I'm planning to early retire at 40.
My method is to take the expected real return, reduce it by a third, and that's the withdrawal rate.
So if expected real return is 6%, then withdrawal rate is 4%. Right now withdrawal rate is 3.7%.
Expected (nominal) return of bonds is simply their yield. Expected inflation is 2%.
Expected real return of stocks is 100 divided by the rolling average of their PE ratio over 10 years or so. Right now that's 5.6% real for a global portfolio.
It's kind of interesting everyone talks about 4% SWR constant-dollar, yet in reality people actually use a percentage-based method, and yet there aren't many studies done on this front compared to the 4% SWR studies.
-
- Posts: 16054
- Joined: Fri Nov 06, 2020 12:41 pm
Re: Withdrawal rate for an early retirement
I see. I still think the beauty of constant-percentage is that it's quite safe under 4~4.5%. Basically the reliability of the expected return isn't going to materially impact your outcome so long as you stay below 4.5%.
We're actually likely going to run into the opposite problem where our portfolio allows us to spend 150K/year when we don't need to spend that much.
Re: Withdrawal rate for an early retirement
It just so happens that Vanguard put out a new paper yesterday on this very topic. They seem to recommend the use of 4% SWR with dynamic spending adjustments for long retirement durations. They are saying this gets you to 90% success rate for a 50 year retirement.
Of course, to do that one would (I think) need to be flexible enough to reduce spending during down periods if needed. Having a significant fraction of expenses be discretionary would help.
FWIW - I plan to use a 3.5% SWR for a 45-50 year retirement horizon. When combined with social security (with an appropriate "haircut" if needed to account for funding issues) I think we will be OK.
Of course, to do that one would (I think) need to be flexible enough to reduce spending during down periods if needed. Having a significant fraction of expenses be discretionary would help.
FWIW - I plan to use a 3.5% SWR for a 45-50 year retirement horizon. When combined with social security (with an appropriate "haircut" if needed to account for funding issues) I think we will be OK.
Re: Withdrawal rate for an early retirement
Glancing at the article referenced on dynamic spending, they talk about setting a floor and ceiling of -1.5% and 5%, respectively.
Assuming you start with 1million portfolio and pull 4% the first year (40k), how do you apply the floor and ceiling?
Are they saying, at start of year 2, if your portfolio drops to 700k you can take a maximum of 5% from your portfolio value, that being 35,000? And thereby avoid taking 40k in year 2, which would technically be 5.7% of your year 2 portfolio value?
Assuming you start with 1million portfolio and pull 4% the first year (40k), how do you apply the floor and ceiling?
Are they saying, at start of year 2, if your portfolio drops to 700k you can take a maximum of 5% from your portfolio value, that being 35,000? And thereby avoid taking 40k in year 2, which would technically be 5.7% of your year 2 portfolio value?
Re: Withdrawal rate for an early retirement
Using your example, my understanding is that you would take $39,400 in year two ($40,000 x .985). In other words, you reduce the withdrawal amount by 1.5%, not 30%.Brianjp18 wrote: ↑Wed Sep 22, 2021 7:47 am Glancing at the article referenced on dynamic spending, they talk about setting a floor and ceiling of -1.5% and 5%, respectively.
Assuming you start with 1million portfolio and pull 4% the first year (40k), how do you apply the floor and ceiling?
Are they saying, at start of year 2, if your portfolio drops to 700k you can take a maximum of 5% from your portfolio value, that being 35,000? And thereby avoid taking 40k in year 2, which would technically be 5.7% of your year 2 portfolio value?
If your portfolio grows 20% the following year, you would then take a maximum of $41,370 ($39,400 x 1.05).
The floor/ceiling adjustment is always applied to the previous withdrawal amount.
-
- Posts: 16054
- Joined: Fri Nov 06, 2020 12:41 pm
Re: Withdrawal rate for an early retirement
Percentage-based methods would be the way to go. SWR (constant-dollar) is good for academic studies but not really practical.
Re: Withdrawal rate for an early retirement
In your example on year 2 you reduced the 40K pulled from year 1 by 1.5% (39,400). But, say your portfolio was cut in half in year one, down to 500,000 and you use the same rule, max reduction of 1.5% you would still pull 39,400. But now, compared to your portfolio you are pulling 8% (39,400/500,000), that does not seem like a safe rate to pull.zuma wrote: ↑Wed Sep 22, 2021 8:09 amUsing your example, my understanding is that you would take $39,400 in year two ($40,000 x .985). In other words, you reduce the withdrawal amount by 1.5%, not 30%.Brianjp18 wrote: ↑Wed Sep 22, 2021 7:47 am Glancing at the article referenced on dynamic spending, they talk about setting a floor and ceiling of -1.5% and 5%, respectively.
Assuming you start with 1million portfolio and pull 4% the first year (40k), how do you apply the floor and ceiling?
Are they saying, at start of year 2, if your portfolio drops to 700k you can take a maximum of 5% from your portfolio value, that being 35,000? And thereby avoid taking 40k in year 2, which would technically be 5.7% of your year 2 portfolio value?
If your portfolio grows 20% the following year, you would then take a maximum of $41,370 ($39,400 x 1.05).
The floor/ceiling adjustment is always applied to the previous withdrawal amount.
That is why I feel like the floor/ceiling is based off your portfolio value, not last year's withdrawal amount. For example, if year one you pulled 40K of 1mill (4%), then the following year when you plan to pull 40K, you ensure you are not pulling greater than 5% from your portfolio? So the 5% ceiling would ensure you don't pull greater than 5% out of your portfolio value, therefore avoiding depleting your portfolio too fast.
Re: Withdrawal rate for an early retirement
Zuma- I’m sure you are right, just seems like that strategy wouldn’t protect you much from withdrawing a large percentage of your portfolio if the market was way down.
-
- Posts: 6234
- Joined: Mon Apr 11, 2011 12:28 am
Re: Withdrawal rate for an early retirement
Taking a rolling average of PE would overestimate real returns and assumes that if there’s been upward trending valuation the past 10 years that it will persist to some degree.Tamalak wrote: ↑Wed Jul 07, 2021 10:07 amIt's not reliable at all. It is, I hope, the approximate center of a low and broad bell curve. That's why I reduce it by a third as a safety measure.Marseille07 wrote: ↑Wed Jul 07, 2021 9:33 amI don't know how reliable the expected real return is. If it is reliable then it is probably a reasonable approach.Tamalak wrote: ↑Tue Jul 06, 2021 8:39 am I'm planning to early retire at 40.
My method is to take the expected real return, reduce it by a third, and that's the withdrawal rate.
So if expected real return is 6%, then withdrawal rate is 4%. Right now withdrawal rate is 3.7%.
Expected (nominal) return of bonds is simply their yield. Expected inflation is 2%.
Expected real return of stocks is 100 divided by the rolling average of their PE ratio over 10 years or so. Right now that's 5.6% real for a global portfolio.
It's kind of interesting everyone talks about 4% SWR constant-dollar, yet in reality people actually use a percentage-based method, and yet there aren't many studies done on this front compared to the 4% SWR studies.
Why not take 100/PE10 ? This is more conservative
20% VOO | 20% VXUS | 20% AVUV | 20% AVDV | 20% AVES
Re: Withdrawal rate for an early retirement
But this is how a constant-dollar withdrawal method like the classic '4% rule' is intended to work. The idea is that even if your portfolio value drops significantly, you can continue to pull the same amount every year for 30 years. In the article linked above, Vanguard is describing a method that's a compromise between this (i.e. the '4% rule') and a percentage-of-portfolio method (i.e. take 5% of the portfolio balance every year).Brianjp18 wrote: ↑Wed Sep 22, 2021 10:15 am In your example on year 2 you reduced the 40K pulled from year 1 by 1.5% (39,400). But, say your portfolio was cut in half in year one, down to 500,000 and you use the same rule, max reduction of 1.5% you would still pull 39,400. But now, compared to your portfolio you are pulling 8% (39,400/500,000), that does not seem like a safe rate to pull.
Re: Withdrawal rate for an early retirement
The 4% withdrawal strategy takes into account having a portfolio drop significantly early in retirement. The 4% number comes from our history that in the past, if the worst case in history repeats itself, one can withdraw 4% and adjust for inflation without running out of money for 30 years. 4% takes into account this 50% drop as you describe. So although you are withdrawing 8% that year, history, the market would recover and one would be fine. By adding a bit more variability to your withdrawal strategy such as decrease withdrawals by 1.5% in a bad year, one will increase the chance of that portfolio lasting over 30 years.Brianjp18 wrote: ↑Wed Sep 22, 2021 10:15 amIn your example on year 2 you reduced the 40K pulled from year 1 by 1.5% (39,400). But, say your portfolio was cut in half in year one, down to 500,000 and you use the same rule, max reduction of 1.5% you would still pull 39,400. But now, compared to your portfolio you are pulling 8% (39,400/500,000), that does not seem like a safe rate to pull.zuma wrote: ↑Wed Sep 22, 2021 8:09 amUsing your example, my understanding is that you would take $39,400 in year two ($40,000 x .985). In other words, you reduce the withdrawal amount by 1.5%, not 30%.Brianjp18 wrote: ↑Wed Sep 22, 2021 7:47 am Glancing at the article referenced on dynamic spending, they talk about setting a floor and ceiling of -1.5% and 5%, respectively.
Assuming you start with 1million portfolio and pull 4% the first year (40k), how do you apply the floor and ceiling?
Are they saying, at start of year 2, if your portfolio drops to 700k you can take a maximum of 5% from your portfolio value, that being 35,000? And thereby avoid taking 40k in year 2, which would technically be 5.7% of your year 2 portfolio value?
If your portfolio grows 20% the following year, you would then take a maximum of $41,370 ($39,400 x 1.05).
The floor/ceiling adjustment is always applied to the previous withdrawal amount.
That is why I feel like the floor/ceiling is based off your portfolio value, not last year's withdrawal amount. For example, if year one you pulled 40K of 1mill (4%), then the following year when you plan to pull 40K, you ensure you are not pulling greater than 5% from your portfolio? So the 5% ceiling would ensure you don't pull greater than 5% out of your portfolio value, therefore avoiding depleting your portfolio too fast.
The only caveat is that the future has to look better than the past. Add in a sprinkle of variability in your yearly withdrawals and you should be golden.
A time to EVALUATE your jitters: |
viewtopic.php?p=1139732#p1139732
Re: Withdrawal rate for an early retirement
Keep in mind that if the market drops by 50%, expected future return will likely be better so pulling 8% is not such a bad idea. Also, pulling only 5% of the portfolio, in this case $25k, may not be sufficient for the retiree, depending how much slack there are in the retiree's expense.Brianjp18 wrote: ↑Wed Sep 22, 2021 10:15 am
In your example on year 2 you reduced the 40K pulled from year 1 by 1.5% (39,400). But, say your portfolio was cut in half in year one, down to 500,000 and you use the same rule, max reduction of 1.5% you would still pull 39,400. But now, compared to your portfolio you are pulling 8% (39,400/500,000), that does not seem like a safe rate to pull.
-
- Posts: 16054
- Joined: Fri Nov 06, 2020 12:41 pm
Re: Withdrawal rate for an early retirement
I'm not sure if I agree. What kills your portfolio is selling lots of shares at the bottom, and pulling 8% after -50% is exactly that.jarjarM wrote: ↑Wed Sep 22, 2021 12:25 pm Keep in mind that if the market drops by 50%, expected future return will likely be better so pulling 8% is not such a bad idea. Also, pulling only 5% of the portfolio, in this case $25k, may not be sufficient for the retiree, depending how much slack there are in the retiree's expense.
Re: Withdrawal rate for an early retirement
That's why 100% in stock during retirement is not a good idea. I'm an aggressive investor but I would never be 100% stock when I jumpMarseille07 wrote: ↑Wed Sep 22, 2021 12:32 pmI'm not sure if I agree. What kills your portfolio is selling lots of shares at the bottom, and pulling 8% after -50% is exactly that.jarjarM wrote: ↑Wed Sep 22, 2021 12:25 pm Keep in mind that if the market drops by 50%, expected future return will likely be better so pulling 8% is not such a bad idea. Also, pulling only 5% of the portfolio, in this case $25k, may not be sufficient for the retiree, depending how much slack there are in the retiree's expense.
-
- Posts: 16054
- Joined: Fri Nov 06, 2020 12:41 pm
Re: Withdrawal rate for an early retirement
My preference will be having 8-10x expense in cash/short term bond. If I get to 100x expense, then 90% is very possible for me In reality, I think my risk tolerance is probably in the 80+% range depending on economic news noises.Marseille07 wrote: ↑Wed Sep 22, 2021 1:08 pmWhich AA are you planning to use? I won't do 100% either, but 90%+ is possible imo.
-
- Posts: 16054
- Joined: Fri Nov 06, 2020 12:41 pm
Re: Withdrawal rate for an early retirement
Marseille:Marseille07 wrote: ↑Wed Sep 22, 2021 12:32 pmI'm not sure if I agree. What kills your portfolio is selling lots of shares at the bottom, and pulling 8% after -50% is exactly that.jarjarM wrote: ↑Wed Sep 22, 2021 12:25 pm Keep in mind that if the market drops by 50%, expected future return will likely be better so pulling 8% is not such a bad idea. Also, pulling only 5% of the portfolio, in this case $25k, may not be sufficient for the retiree, depending how much slack there are in the retiree's expense.
Your instincts are usually "right on", but in this case I believe that they may be misleading you. I did a bit of a deep dive into SORR a couple of years ago and it turns out that sequence of return risk comes primarily from the 8% withdrawal rate, not from the "selling stocks", which is certainly a contributing factor. Many on this forum are mistaken about this, saying things like: "I have 10 years in bonds, so SORR will not affect me". But they are still pulling 8% from their portfolio if they spend only bonds during the market downturn.
Additionally, in decumulation most folks are using withdrawals for rebalancing, so few will be selling "lots of stock" because they will have an excess percentage of bonds after the market drop. From a practical standpoint most will not be selling "lots of stock" unless they have a very high stock AA, but they still face SORR. They are withdrawing from bonds but they are still withdrawing 8% from their portfolio.
Run some numbers and I think you will find that this is the case...
It's not an engineering problem - Hersh Shefrin | To get the "risk premium", you really do have to take the risk - nisiprius
-
- Posts: 16054
- Joined: Fri Nov 06, 2020 12:41 pm
Re: Withdrawal rate for an early retirement
I believe you're correct. I was primarily talking about selling stocks since I don't own bonds, but you're correct that SORR shows up when you sell bonds by the same amount.David Jay wrote: ↑Wed Sep 22, 2021 1:32 pm Marseille:
Your instincts are usually "right on", but in this case I believe that they may be misleading you. I did a bit of a deep dive into SORR a couple of years ago and it turns out that sequence of return risk comes primarily from the 8% withdrawal rate, not from the "selling stocks", which is certainly a contributing factor. Many on this forum are mistaken about this, saying things like: "I have 10 years in bonds, so SORR will not affect me". But they are still pulling 8% from their portfolio if they spend only bonds during the market downturn.
Additionally, in decumulation most folks are using withdrawals for rebalancing, so few will be selling "lots of stock" because they will have an excess percentage of bonds after the market drop. From a practical standpoint most will not be selling "lots of stock" unless they have a very high stock AA, but they still face SORR. They are withdrawing from bonds but they are still withdrawing 8% from their portfolio.
Run some numbers and I think you will find that this is the case...
Re: Withdrawal rate for an early retirement
I don't like to use PE10 because it doesn't account for optimism or pessimism for future returns. If the market goes up 20% one year that isn't 'stealing' 20% from future returns - going up that much means that expectations were exceeded and so total returns including that 20% are expected to be higher now.Nathan Drake wrote: ↑Wed Sep 22, 2021 11:47 amTaking a rolling average of PE would overestimate real returns and assumes that if there’s been upward trending valuation the past 10 years that it will persist to some degree.Tamalak wrote: ↑Wed Jul 07, 2021 10:07 amIt's not reliable at all. It is, I hope, the approximate center of a low and broad bell curve. That's why I reduce it by a third as a safety measure.Marseille07 wrote: ↑Wed Jul 07, 2021 9:33 amI don't know how reliable the expected real return is. If it is reliable then it is probably a reasonable approach.Tamalak wrote: ↑Tue Jul 06, 2021 8:39 am I'm planning to early retire at 40.
My method is to take the expected real return, reduce it by a third, and that's the withdrawal rate.
So if expected real return is 6%, then withdrawal rate is 4%. Right now withdrawal rate is 3.7%.
Expected (nominal) return of bonds is simply their yield. Expected inflation is 2%.
Expected real return of stocks is 100 divided by the rolling average of their PE ratio over 10 years or so. Right now that's 5.6% real for a global portfolio.
It's kind of interesting everyone talks about 4% SWR constant-dollar, yet in reality people actually use a percentage-based method, and yet there aren't many studies done on this front compared to the 4% SWR studies.
Why not take 100/PE10 ? This is more conservative
I don't want to be conservative, I want to aim for the middle in my estimate of future returns. I want to be equally likely to be wrong high or wrong low. The 'conservatism' in my plan is keeping my WW to 2/3s of my expected returns. Heaping one safety margin on top of another is a way to never end up using your wealth at all.
I don't see how using a rolling PE overestimates returns by much. Historically PE has averaged 13-15 which would imply 7% real. Historical earnings have been 6.7% real which is pretty close.
Re: Withdrawal rate for an early retirement
Tamalak wrote: ↑Wed Sep 22, 2021 1:47 pmI don't like to use PE10 because it doesn't account for optimism or pessimism for future returns. If the market goes up 20% one year that isn't 'stealing' 20% from future returns - going up that much means that expectations were exceeded and so total returns including that 20% are expected to be higher now.Nathan Drake wrote: ↑Wed Sep 22, 2021 11:47 amTaking a rolling average of PE would overestimate real returns and assumes that if there’s been upward trending valuation the past 10 years that it will persist to some degree.Tamalak wrote: ↑Wed Jul 07, 2021 10:07 amIt's not reliable at all. It is, I hope, the approximate center of a low and broad bell curve. That's why I reduce it by a third as a safety measure.Marseille07 wrote: ↑Wed Jul 07, 2021 9:33 amI don't know how reliable the expected real return is. If it is reliable then it is probably a reasonable approach.Tamalak wrote: ↑Tue Jul 06, 2021 8:39 am I'm planning to early retire at 40.
My method is to take the expected real return, reduce it by a third, and that's the withdrawal rate.
So if expected real return is 6%, then withdrawal rate is 4%. Right now withdrawal rate is 3.7%.
Expected (nominal) return of bonds is simply their yield. Expected inflation is 2%.
Expected real return of stocks is 100 divided by the rolling average of their PE ratio over 10 years or so. Right now that's 5.6% real for a global portfolio.
It's kind of interesting everyone talks about 4% SWR constant-dollar, yet in reality people actually use a percentage-based method, and yet there aren't many studies done on this front compared to the 4% SWR studies.
Why not take 100/PE10 ? This is more conservative
I don't want to be conservative, I want to aim for the middle in my estimate of future returns. I want to be equally likely to be wrong high or wrong low. The 'conservatism' in my plan is keeping my WW to 2/3s of my expected returns. Heaping one safety margin on top of another is a way to never end up using your wealth at all.
I don't see how using a rolling PE overestimates returns by much. Historically PE has averaged 13-15 which would imply 7% real. Historical earnings have been 6.7% real which is pretty close.
+1..This seems to be a balanced approach. So many posts on BH where one is withdrawing
2.5%
+extra 3 yrs of cash
+discounting portfolio due to PE, CAPE, ABC123
+adding 25% to spending for just in case
Actual portfolio prob at 70x...uh can I retire?
- willthrill81
- Posts: 32250
- Joined: Thu Jan 26, 2017 2:17 pm
- Location: USA
- Contact:
Re: Withdrawal rate for an early retirement
If this were accurate, then we shouldn't have seen anything like the so-called 'Siegel constant' (i.e., approximately 7% real returns over very long-term periods), yet we have. Higher returns now have tended to result in lower expected returns going forward. It isn't random that U.S. stocks followed the 25% returns of the late 1990s by losing out to inflation in the 2000s.
The Sensible Steward
-
- Posts: 6234
- Joined: Mon Apr 11, 2011 12:28 am
Re: Withdrawal rate for an early retirement
I do not think this is a safe assumption. P/E values can go up greatly without any underlying significant improvement to the fundamentals, that’s exactly what has happened this decade for US TSMTamalak wrote: ↑Wed Sep 22, 2021 1:47 pmI don't like to use PE10 because it doesn't account for optimism or pessimism for future returns. If the market goes up 20% one year that isn't 'stealing' 20% from future returns - going up that much means that expectations were exceeded and so total returns including that 20% are expected to be higher now.Nathan Drake wrote: ↑Wed Sep 22, 2021 11:47 amTaking a rolling average of PE would overestimate real returns and assumes that if there’s been upward trending valuation the past 10 years that it will persist to some degree.Tamalak wrote: ↑Wed Jul 07, 2021 10:07 amIt's not reliable at all. It is, I hope, the approximate center of a low and broad bell curve. That's why I reduce it by a third as a safety measure.Marseille07 wrote: ↑Wed Jul 07, 2021 9:33 amI don't know how reliable the expected real return is. If it is reliable then it is probably a reasonable approach.Tamalak wrote: ↑Tue Jul 06, 2021 8:39 am I'm planning to early retire at 40.
My method is to take the expected real return, reduce it by a third, and that's the withdrawal rate.
So if expected real return is 6%, then withdrawal rate is 4%. Right now withdrawal rate is 3.7%.
Expected (nominal) return of bonds is simply their yield. Expected inflation is 2%.
Expected real return of stocks is 100 divided by the rolling average of their PE ratio over 10 years or so. Right now that's 5.6% real for a global portfolio.
It's kind of interesting everyone talks about 4% SWR constant-dollar, yet in reality people actually use a percentage-based method, and yet there aren't many studies done on this front compared to the 4% SWR studies.
Why not take 100/PE10 ? This is more conservative
I don't want to be conservative, I want to aim for the middle in my estimate of future returns. I want to be equally likely to be wrong high or wrong low. The 'conservatism' in my plan is keeping my WW to 2/3s of my expected returns. Heaping one safety margin on top of another is a way to never end up using your wealth at all.
I don't see how using a rolling PE overestimates returns by much. Historically PE has averaged 13-15 which would imply 7% real. Historical earnings have been 6.7% real which is pretty close.
20% VOO | 20% VXUS | 20% AVUV | 20% AVDV | 20% AVES
Re: Withdrawal rate for an early retirement
Regardless of the rate you choose to start, keep in mind that a straight inflation adjustment will not be sufficient to maintain relative purchasing power over an extended period. You actually need to keep up with at least real wage growth at your starting level, and ideally real GDP growth, Otherwise those luxuries of today will eventually become necessities that you can't afford. 30 years ago, cell phones were nearly nonexistent, the internet wasn't here, and most people didn't actually own computers at home.
“Adapt what is useful, reject what is useless, and add what is specifically your own.” ― Bruce Lee
Re: Withdrawal rate for an early retirement
That's true, and in particular the Fed's ability to mess with the time value of money at will is a wrench in my approach.Nathan Drake wrote: ↑Wed Sep 22, 2021 3:03 pmI do not think this is a safe assumption. P/E values can go up greatly without any underlying significant improvement to the fundamentals, that’s exactly what has happened this decade for US TSMTamalak wrote: ↑Wed Sep 22, 2021 1:47 pmI don't like to use PE10 because it doesn't account for optimism or pessimism for future returns. If the market goes up 20% one year that isn't 'stealing' 20% from future returns - going up that much means that expectations were exceeded and so total returns including that 20% are expected to be higher now.Nathan Drake wrote: ↑Wed Sep 22, 2021 11:47 amTaking a rolling average of PE would overestimate real returns and assumes that if there’s been upward trending valuation the past 10 years that it will persist to some degree.Tamalak wrote: ↑Wed Jul 07, 2021 10:07 amIt's not reliable at all. It is, I hope, the approximate center of a low and broad bell curve. That's why I reduce it by a third as a safety measure.Marseille07 wrote: ↑Wed Jul 07, 2021 9:33 am
I don't know how reliable the expected real return is. If it is reliable then it is probably a reasonable approach.
It's kind of interesting everyone talks about 4% SWR constant-dollar, yet in reality people actually use a percentage-based method, and yet there aren't many studies done on this front compared to the 4% SWR studies.
Why not take 100/PE10 ? This is more conservative
I don't want to be conservative, I want to aim for the middle in my estimate of future returns. I want to be equally likely to be wrong high or wrong low. The 'conservatism' in my plan is keeping my WW to 2/3s of my expected returns. Heaping one safety margin on top of another is a way to never end up using your wealth at all.
I don't see how using a rolling PE overestimates returns by much. Historically PE has averaged 13-15 which would imply 7% real. Historical earnings have been 6.7% real which is pretty close.
The fundamental issue I've been facing is:
-When P/E changes due to optimism/pessimism I don't want to change my expected future % returns, because the adjusted expectation has already been clawed into the present.
-When P/E changes due to some other reason (time value of money shift, supply and demand, competing forms of investments becoming more or less attractive), I DO want to change my expected future returns as a %, because the dollar value of my portfolio has changed but the dollar value of what I'll be getting hasn't.
-However, I don't know how to distinguish between P/E changes due to market sentiment and P/E changes due to other reasons. The approach I've settled on is assuming that P/E changes due to market sentiment are sudden and short term, while other reasons are long term (like the shift upwards in P/E we've been seeing since the 1990s). Therefore, expected future returns should be according to rolling P/E of the last 10 years or so, and not the P/E we've got right now (which should reflect market sentiment).
-But the Fed's ability to, for example, drop rates suddenly to 0 as they did in 2020, is a non-market-sentiment but also sudden adjustment to the equilibrium P/E the market is seeking.
-One approach I'm considering is using forward PE as a metric instead of historical PE. Forward PE very neatly separates out what the market actually expects to get. However, in order to figure out how to translate forward PE into expected actual return, I need a lot of historical data, and I can't find historical PE before 1990
Re: Withdrawal rate for an early retirement
Thanks Zuma. I read back through the article and the linked PDF and I understand what you are saying now. It’s basically the 4% rule with a little up or down wiggle based on the market, but ultimately protects you from having to cut expenses dramatically during a down market while also reigning in the excitement and temptation to pull excess amounts after a good year.zuma wrote: ↑Wed Sep 22, 2021 12:10 pmBut this is how a constant-dollar withdrawal method like the classic '4% rule' is intended to work. The idea is that even if your portfolio value drops significantly, you can continue to pull the same amount every year for 30 years. In the article linked above, Vanguard is describing a method that's a compromise between this (i.e. the '4% rule') and a percentage-of-portfolio method (i.e. take 5% of the portfolio balance every year).Brianjp18 wrote: ↑Wed Sep 22, 2021 10:15 am In your example on year 2 you reduced the 40K pulled from year 1 by 1.5% (39,400). But, say your portfolio was cut in half in year one, down to 500,000 and you use the same rule, max reduction of 1.5% you would still pull 39,400. But now, compared to your portfolio you are pulling 8% (39,400/500,000), that does not seem like a safe rate to pull.
Re: Withdrawal rate for an early retirement
So it’s basically like the historically proven safe withdrawal rate (4%) plus a little extra insurance, granting you a little freedom to pull more in an up market, but reigning back in a down market, up to a max of 1.5% less than the value you withdrew last year.EnjoyIt wrote: ↑Wed Sep 22, 2021 12:23 pmThe 4% withdrawal strategy takes into account having a portfolio drop significantly early in retirement. The 4% number comes from our history that in the past, if the worst case in history repeats itself, one can withdraw 4% and adjust for inflation without running out of money for 30 years. 4% takes into account this 50% drop as you describe. So although you are withdrawing 8% that year, history, the market would recover and one would be fine. By adding a bit more variability to your withdrawal strategy such as decrease withdrawals by 1.5% in a bad year, one will increase the chance of that portfolio lasting over 30 years.Brianjp18 wrote: ↑Wed Sep 22, 2021 10:15 amIn your example on year 2 you reduced the 40K pulled from year 1 by 1.5% (39,400). But, say your portfolio was cut in half in year one, down to 500,000 and you use the same rule, max reduction of 1.5% you would still pull 39,400. But now, compared to your portfolio you are pulling 8% (39,400/500,000), that does not seem like a safe rate to pull.zuma wrote: ↑Wed Sep 22, 2021 8:09 amUsing your example, my understanding is that you would take $39,400 in year two ($40,000 x .985). In other words, you reduce the withdrawal amount by 1.5%, not 30%.Brianjp18 wrote: ↑Wed Sep 22, 2021 7:47 am Glancing at the article referenced on dynamic spending, they talk about setting a floor and ceiling of -1.5% and 5%, respectively.
Assuming you start with 1million portfolio and pull 4% the first year (40k), how do you apply the floor and ceiling?
Are they saying, at start of year 2, if your portfolio drops to 700k you can take a maximum of 5% from your portfolio value, that being 35,000? And thereby avoid taking 40k in year 2, which would technically be 5.7% of your year 2 portfolio value?
If your portfolio grows 20% the following year, you would then take a maximum of $41,370 ($39,400 x 1.05).
The floor/ceiling adjustment is always applied to the previous withdrawal amount.
That is why I feel like the floor/ceiling is based off your portfolio value, not last year's withdrawal amount. For example, if year one you pulled 40K of 1mill (4%), then the following year when you plan to pull 40K, you ensure you are not pulling greater than 5% from your portfolio? So the 5% ceiling would ensure you don't pull greater than 5% out of your portfolio value, therefore avoiding depleting your portfolio too fast.
The only caveat is that the future has to look better than the past. Add in a sprinkle of variability in your yearly withdrawals and you should be golden.
Quick question. If you decide on pulling 4% (40k) in year one. To calculate year two you would add the inflation percentage from year one to that value (say 2%; so year 2 you plan to pull 40,800)?
Then how do you use that 48,800 relative to your portfolio value to decide how much more/less to pull for year 2?
Re: Withdrawal rate for an early retirement
That’s true. 40k down to 25k is an unrealistic adjustment. I guess the key is to find a comfortable middle ground. Luckily we do have Bengen’s research to give us some confidence in pulling money during bad times. I guess the topic of sequence of return risk is pretty well covered by his data as long as we don’t go above 4%.jarjarM wrote: ↑Wed Sep 22, 2021 12:25 pmKeep in mind that if the market drops by 50%, expected future return will likely be better so pulling 8% is not such a bad idea. Also, pulling only 5% of the portfolio, in this case $25k, may not be sufficient for the retiree, depending how much slack there are in the retiree's expense.Brianjp18 wrote: ↑Wed Sep 22, 2021 10:15 am
In your example on year 2 you reduced the 40K pulled from year 1 by 1.5% (39,400). But, say your portfolio was cut in half in year one, down to 500,000 and you use the same rule, max reduction of 1.5% you would still pull 39,400. But now, compared to your portfolio you are pulling 8% (39,400/500,000), that does not seem like a safe rate to pull.