Withdrawal rate for an early retirement

Discuss all general (i.e. non-personal) investing questions and issues, investing news, and theory.
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 11:30 am
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.
Sure, but the whole point is to eventually follow it (which means you spend at 4.5% or less), right? I don't think we're really disagreeing on anything here to be honest.
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 12:14 pm
willthrill81 wrote: Fri Nov 27, 2020 11:30 am
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

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.
Sure, but the whole point is to eventually follow it (which means you spend at 4.5% or less), right? I don't think we're really disagreeing on anything here to be honest.
I still just don't see the value in pulling money out of your investments to hold it in cash for no real benefit. But if it helps you feel better about funding your retirement, then go for it.
The Sensible Steward
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 11:26 am
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

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.
I think it will work. Now, for day-to-day (not even EF) I see myself having around 20K of cash in checking, not for emergencies but just to have some liquidity. But the 80K EF business can be tracked on spreadsheet while fully invested.
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 12:15 pm
Marseille07 wrote: Fri Nov 27, 2020 12:14 pm
willthrill81 wrote: Fri Nov 27, 2020 11:30 am
Marseille07 wrote: Fri Nov 27, 2020 11:10 am
willthrill81 wrote: Fri Nov 27, 2020 10:55 am

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.
Sure, but the whole point is to eventually follow it (which means you spend at 4.5% or less), right? I don't think we're really disagreeing on anything here to be honest.
I still just don't see the value in pulling money out of your investments to hold it in cash for no real benefit. But if it helps you feel better about funding your retirement, then go for it.
I'll just track it in excel or something. Both you and EnjoyIt are making good points, thank you.
EnjoyIt
Posts: 8244
Joined: Sun Dec 29, 2013 7:06 pm

Re: Withdrawal rate for an early retirement

Post by EnjoyIt »

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.
^ I put in bold above ^

You are so right. "Flexibility is king" and far more important than what strategy you use. VPW and ABW are just mathematical method to provide flexibility to the standard 4% withdrawal strategy.
If things are bad, one spends less. If things are good, one can spend more. Having $10s of thousands in flexibility is a pretty big deal. If I can make a reasonably comfortable life on $40k a year and then have another $20k-$40k of discretionary expenses. Not so easy for the Mr. Money Mustache crowd who retire on $24k a year, but here at bogleheads I think many really have that capability.
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 »

OP,

I just re-ran my numbers last night and my current conclusion is that 3% "constant-percentage" is fine to go with. I was part of the 2.0%~2.5% camp, which isn't terrible still, but I think I can bump it up a notch.

Even with 3% WR, I'll only withdraw 20K in checking then after that I'll track EF in excel; this way, the AA stays in the market thus my effective WR would be lower than 3%.
marcopolo
Posts: 8411
Joined: Sat Dec 03, 2016 9:22 am

Re: Withdrawal rate for an early retirement

Post by marcopolo »

willthrill81 wrote: Fri Nov 27, 2020 10:39 am
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

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).
I completely agree with what you are saying here.

But, weren't you making the exact opposite case earlier in the thread and suggesting having extra money stashed away in cash and using it during a market downturn provided significant benefits?
Once in a while you get shown the light, in the strangest of places if you look at it right.
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 12:43 pm OP,

I just re-ran my numbers last night and my current conclusion is that 3% "constant-percentage" is fine to go with. I was part of the 2.0%~2.5% camp, which isn't terrible still, but I think I can bump it up a notch.

Even with 3% WR, I'll only withdraw 20K in checking then after that I'll track EF in excel; this way, the AA stays in the market thus my effective WR would be lower than 3%.
Very nice. As most of us will agree, 3% is on the conservative side and will protect you from unprecedented economic turmoil. If it doesn't come to be in the next 5-10 years, odds are your portfolio will grow, grow, grow which isn't a bad thing. Who knows, you can use the extra growth and splurge on one off large expenses without putting your retirement at risk. It could be taking your family out for a nice vacation, buying a fancy retirement car, or whatever. My parents recently splurged on a personal guide when they went to Japan last year. It was great for them. Basically it was a personal translator, chauffeur, and guide all in one package.
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 »

EnjoyIt wrote: Fri Nov 27, 2020 1:04 pm
Marseille07 wrote: Fri Nov 27, 2020 12:43 pm OP,

I just re-ran my numbers last night and my current conclusion is that 3% "constant-percentage" is fine to go with. I was part of the 2.0%~2.5% camp, which isn't terrible still, but I think I can bump it up a notch.

Even with 3% WR, I'll only withdraw 20K in checking then after that I'll track EF in excel; this way, the AA stays in the market thus my effective WR would be lower than 3%.
Very nice. As most of us will agree, 3% is on the conservative side and will protect you from unprecedented economic turmoil. If it doesn't come to be in the next 5-10 years, odds are your portfolio will grow, grow, grow which isn't a bad thing. Who knows, you can use the extra growth and splurge on one off large expenses without putting your retirement at risk. It could be taking your family out for a nice vacation, buying a fancy retirement car, or whatever. My parents recently splurged on a personal guide when they went to Japan last year. It was great for them. Basically it was a personal translator, chauffeur, and guide all in one package.
Indeed, this is part of the reason why I want some cash on hand. A nice vacation wouldn't come out of EF but can be covered by the 20K in checking. I've been to Japan as well, a very nice country to visit.
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 »

marcopolo wrote: Fri Nov 27, 2020 12:49 pm
willthrill81 wrote: Fri Nov 27, 2020 10:39 am
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

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).
I completely agree with what you are saying here.

But, weren't you making the exact opposite case earlier in the thread and suggesting having extra money stashed away in cash and using it during a market downturn provided significant benefits?
That was part of an overall strategy to attempt to reduce SORR and was not merely mental accounting. However, even then, digging further into the results indicated that the method results in the same SWR as not using it, so there was no real benefit even then.
The Sensible Steward
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 11:26 am
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

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.
One thing that struck me is that, by accounting for EF virtually, we can remove the arbitrary 80K boundary and make it infinite. Since the money remains invested in the market, it's not going to matter if the EF size becomes 200K, 300K etc etc - if I later need this kind of money (a down payment, for example) then I know exactly how much I can withdraw. This is a brilliant idea.

EDIT: One minor downside is the reverse - because the money remains invested instead of taken out, a percentage of portfolio would also be larger, and the amount "virtually added to EF" will be bigger. This won't be corrected until the EF is actually consumed.
klaus14
Posts: 953
Joined: Sun Nov 25, 2018 6:43 pm

Re: Withdrawal rate for an early retirement

Post by klaus14 »

willthrill81 wrote: Fri Nov 27, 2020 10:19 am
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%.
Using a negative growth parameter in that spreadsheet is probably cleaner?
My investment algorithm: https://www.bogleheads.org/forum/viewtopic.php?f=10&t=351899&p=6112869#p6112869
klaus14
Posts: 953
Joined: Sun Nov 25, 2018 6:43 pm

Re: Withdrawal rate for an early retirement

Post by klaus14 »

vineviz wrote: Fri Nov 27, 2020 10:22 am
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.
Do you like a static allocation during retirement or increasing bond allocation like Vanguard does?
(i think everyone agrees bond alloc should go up till retirement and disagreement is about what to do afterwards)
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: Fri Nov 27, 2020 4:43 pm
willthrill81 wrote: Fri Nov 27, 2020 10:19 am
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%.
Using a negative growth parameter in that spreadsheet is probably cleaner?
It's not 'messy' to add 1-2% to your expected returns to achieve front-loading.
The Sensible Steward
klaus14
Posts: 953
Joined: Sun Nov 25, 2018 6:43 pm

Re: Withdrawal rate for an early retirement

Post by klaus14 »

willthrill81 wrote: Fri Nov 27, 2020 4:58 pm
klaus14 wrote: Fri Nov 27, 2020 4:43 pm
willthrill81 wrote: Fri Nov 27, 2020 10:19 am
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%.
Using a negative growth parameter in that spreadsheet is probably cleaner?
It's not 'messy' to add 1-2% to your expected returns to achieve front-loading.
it's messy in the sense that: how do you know real impact of using "fake" return numbers? what if you deplete your savings earlier than desired? using 'g' param doesn't have these problems.
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: Fri Nov 27, 2020 5:05 pm
willthrill81 wrote: Fri Nov 27, 2020 4:58 pm
klaus14 wrote: Fri Nov 27, 2020 4:43 pm
willthrill81 wrote: Fri Nov 27, 2020 10:19 am
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%.
Using a negative growth parameter in that spreadsheet is probably cleaner?
It's not 'messy' to add 1-2% to your expected returns to achieve front-loading.
it's messy in the sense that: how do you know real impact of using "fake" return numbers? what if you deplete your savings earlier than desired? using 'g' param doesn't have these problems.
You will not deplete your savings earlier than desired by using a larger rate of return in the TVM formula than the expected return. Your future withdrawals will just naturally shrink in real dollars, all else held equal.
The Sensible Steward
klaus14
Posts: 953
Joined: Sun Nov 25, 2018 6:43 pm

Re: Withdrawal rate for an early retirement

Post by klaus14 »

willthrill81 wrote: Fri Nov 27, 2020 5:10 pm
klaus14 wrote: Fri Nov 27, 2020 5:05 pm
willthrill81 wrote: Fri Nov 27, 2020 4:58 pm
klaus14 wrote: Fri Nov 27, 2020 4:43 pm
willthrill81 wrote: Fri Nov 27, 2020 10:19 am
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%.
Using a negative growth parameter in that spreadsheet is probably cleaner?
It's not 'messy' to add 1-2% to your expected returns to achieve front-loading.
it's messy in the sense that: how do you know real impact of using "fake" return numbers? what if you deplete your savings earlier than desired? using 'g' param doesn't have these problems.
You will not deplete your savings earlier than desired by using a larger rate of return in the TVM formula than the expected return. Your future withdrawals will just naturally shrink in real dollars, all else held equal.
I was trying to say: what if you are reducing your withdrawals in old age more than desired? Your method is very imprecise compared to g param.
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: Fri Nov 27, 2020 5:13 pm
willthrill81 wrote: Fri Nov 27, 2020 5:10 pm
klaus14 wrote: Fri Nov 27, 2020 5:05 pm
willthrill81 wrote: Fri Nov 27, 2020 4:58 pm
klaus14 wrote: Fri Nov 27, 2020 4:43 pm

Using a negative growth parameter in that spreadsheet is probably cleaner?
It's not 'messy' to add 1-2% to your expected returns to achieve front-loading.
it's messy in the sense that: how do you know real impact of using "fake" return numbers? what if you deplete your savings earlier than desired? using 'g' param doesn't have these problems.
You will not deplete your savings earlier than desired by using a larger rate of return in the TVM formula than the expected return. Your future withdrawals will just naturally shrink in real dollars, all else held equal.
I was trying to say: what if you are reducing your withdrawals in old age more than desired? Your method is very imprecise compared to g param.
We have no real risk of reducing withdrawals too much in this way because SS benefits will cover all of our essential spending from age 70 forward, leaving whatever remains in our portfolio for discretionary spending and possible LTC expenses.

I'd like to see a concrete example of how what you're recommending is objectively better. Right now, it looks like six of one and half a dozen of another.
The Sensible Steward
klaus14
Posts: 953
Joined: Sun Nov 25, 2018 6:43 pm

Re: Withdrawal rate for an early retirement

Post by klaus14 »

willthrill81 wrote: Fri Nov 27, 2020 5:23 pm We have no real risk of reducing withdrawals too much in this way because SS benefits will cover all of our essential spending from age 70 forward, leaving whatever remains in our portfolio for discretionary spending and possible LTC expenses.

I'd like to see a concrete example of how what you're recommending is objectively better. Right now, it looks like six of one and half a dozen of another.
Intuitively there is an equivalence between two methods, I was just saying g is more explicit. For example, you don't know how much you are reducing your consumption at year 20 with your method. You can calculate it obviously but it's not just there like g method.
My investment algorithm: https://www.bogleheads.org/forum/viewtopic.php?f=10&t=351899&p=6112869#p6112869
nigel_ht
Posts: 4742
Joined: Tue Jan 01, 2019 9:14 am

Re: Withdrawal rate for an early retirement

Post by nigel_ht »

geerhardusvos wrote: Thu Nov 26, 2020 9:37 am
deanmoriarty wrote: Thu Nov 26, 2020 9:06 am For my early retirement, at around 36 in a year or two (hopefully somewhere around $4M NW), I plan for 1.5% withdrawal rate.

Why? A lot of people on the forum feel that 3-4% should be safe. However, expenses are not always constant, especially for early retirement.

The natural consequence, in my mind, is to make your expected baseline of expenses a very low proportion of your net worth (1.5%). Alternatively, aggressively pad your current expenses to account for the much unexpected, and then use that number with a 3% rate. It’s the same thing.
Congrats on your upcoming retirement! I think you’re missing the point and also contradicting yourself in the statement. We don’t feel that 3 to 4% withdrawal rate is safe. It is and has been safe for extremely long periods of time. It’s just a fact. Having your baseline expenses at 3% withdrawal rate is historically completely (100%) safe, and anything else that could come up will very likely be covered because you are already taking an ultra safe withdrawal rate of 3%. You can also budget for the potential things that are unknown to ensure your nest egg supports an extra $10,000 per year if you wanted to. All of this can be done at or above a 3% withdrawal rate. I don’t think you realize how drastic of a difference there is in outcome between a 3.9% WR and a 3.8% WR. It dramatically changes the long-term outcome. It is beyond reasonable and way outside of what is necessary to go from a ridiculously safe 3.25% withdrawal rate down to 1.5% withdrawal rate. If you are able to do it, that is totally fine, more power to you. It’s just overwhelmingly unnecessary, especially if you have mostly equities. There is so much padding built into a 3% withdrawal rate, and people forget that. Over 30 year periods, 4% withdrawal rate has a ton of padding as well.

Especially since you have such control over your living expenses, I think you are in need of some recalibration here.
4% isn’t safe retiring at age 36. Why are you talking about 30 year periods in a scenario where it doesn’t apply?

Oh right, because the light is better over here.
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 »

klaus14 wrote: Fri Nov 27, 2020 4:49 pm
Do you like a static allocation during retirement or increasing bond allocation like Vanguard does?
(i think everyone agrees bond alloc should go up till retirement and disagreement is about what to do afterwards)
I can’t think of a scenario in which a rising bond allocation makes sense.

And to be fair, some investors might prefer a rising equity allocation DESPITE the fact that it doesn’t increase the expected SWR. High degrees of loss aversion or regret aversion might cause such a preference.
"Far more money has been lost by investors preparing for corrections than has been lost in corrections themselves." ~~ Peter Lynch
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 »

nigel_ht wrote: Fri Nov 27, 2020 7:47 pm
geerhardusvos wrote: Thu Nov 26, 2020 9:37 am
deanmoriarty wrote: Thu Nov 26, 2020 9:06 am For my early retirement, at around 36 in a year or two (hopefully somewhere around $4M NW), I plan for 1.5% withdrawal rate.

Why? A lot of people on the forum feel that 3-4% should be safe. However, expenses are not always constant, especially for early retirement.

The natural consequence, in my mind, is to make your expected baseline of expenses a very low proportion of your net worth (1.5%). Alternatively, aggressively pad your current expenses to account for the much unexpected, and then use that number with a 3% rate. It’s the same thing.
Congrats on your upcoming retirement! I think you’re missing the point and also contradicting yourself in the statement. We don’t feel that 3 to 4% withdrawal rate is safe. It is and has been safe for extremely long periods of time. It’s just a fact. Having your baseline expenses at 3% withdrawal rate is historically completely (100%) safe, and anything else that could come up will very likely be covered because you are already taking an ultra safe withdrawal rate of 3%. You can also budget for the potential things that are unknown to ensure your nest egg supports an extra $10,000 per year if you wanted to. All of this can be done at or above a 3% withdrawal rate. I don’t think you realize how drastic of a difference there is in outcome between a 3.9% WR and a 3.8% WR. It dramatically changes the long-term outcome. It is beyond reasonable and way outside of what is necessary to go from a ridiculously safe 3.25% withdrawal rate down to 1.5% withdrawal rate. If you are able to do it, that is totally fine, more power to you. It’s just overwhelmingly unnecessary, especially if you have mostly equities. There is so much padding built into a 3% withdrawal rate, and people forget that. Over 30 year periods, 4% withdrawal rate has a ton of padding as well.

Especially since you have such control over your living expenses, I think you are in need of some recalibration here.
4% isn’t safe retiring at age 36. Why are you talking about 30 year periods in a scenario where it doesn’t apply?
Retiring at age 36, you have about a 50 to 55 year retirement in the best case of scenarios. Did you know the 4% WR has worked ~90% of the time with a 75/25 portfolio for those time horizons? And 70% of the time with a 5% WR... many will probably die way sooner than 90 years old...

If that’s not good enough (and it is for many people), success looks even better when considering Social Security, other possible income, and the ability to be a little bit flexible with the withdrawal rate, especially during down markets.
VTSAX and chill
palanzo
Posts: 2146
Joined: Thu Oct 10, 2019 4:28 pm

Re: Withdrawal rate for an early retirement

Post by palanzo »

geerhardusvos wrote: Fri Nov 27, 2020 9:58 pm
nigel_ht wrote: Fri Nov 27, 2020 7:47 pm
geerhardusvos wrote: Thu Nov 26, 2020 9:37 am
deanmoriarty wrote: Thu Nov 26, 2020 9:06 am For my early retirement, at around 36 in a year or two (hopefully somewhere around $4M NW), I plan for 1.5% withdrawal rate.

Why? A lot of people on the forum feel that 3-4% should be safe. However, expenses are not always constant, especially for early retirement.

The natural consequence, in my mind, is to make your expected baseline of expenses a very low proportion of your net worth (1.5%). Alternatively, aggressively pad your current expenses to account for the much unexpected, and then use that number with a 3% rate. It’s the same thing.
Congrats on your upcoming retirement! I think you’re missing the point and also contradicting yourself in the statement. We don’t feel that 3 to 4% withdrawal rate is safe. It is and has been safe for extremely long periods of time. It’s just a fact. Having your baseline expenses at 3% withdrawal rate is historically completely (100%) safe, and anything else that could come up will very likely be covered because you are already taking an ultra safe withdrawal rate of 3%. You can also budget for the potential things that are unknown to ensure your nest egg supports an extra $10,000 per year if you wanted to. All of this can be done at or above a 3% withdrawal rate. I don’t think you realize how drastic of a difference there is in outcome between a 3.9% WR and a 3.8% WR. It dramatically changes the long-term outcome. It is beyond reasonable and way outside of what is necessary to go from a ridiculously safe 3.25% withdrawal rate down to 1.5% withdrawal rate. If you are able to do it, that is totally fine, more power to you. It’s just overwhelmingly unnecessary, especially if you have mostly equities. There is so much padding built into a 3% withdrawal rate, and people forget that. Over 30 year periods, 4% withdrawal rate has a ton of padding as well.

Especially since you have such control over your living expenses, I think you are in need of some recalibration here.
4% isn’t safe retiring at age 36. Why are you talking about 30 year periods in a scenario where it doesn’t apply?
Retiring at age 36, you have about a 50 to 55 year retirement in the best case of scenarios. Did you know the 4% WR has worked ~90% of the time with a 75/25 portfolio for those time horizons? And 70% of the time with a 5% WR... many will probably die way sooner than 90 years old...

If that’s not good enough (and it is for many people), success looks even better when considering Social Security, other possible income, and the ability to be a little bit flexible with the withdrawal rate, especially during down markets.
Unless he started working at age 1 how much Social Security will he get? Will he be eligible for Medicare?
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 »

palanzo wrote: Fri Nov 27, 2020 10:03 pm
geerhardusvos wrote: Fri Nov 27, 2020 9:58 pm
nigel_ht wrote: Fri Nov 27, 2020 7:47 pm
geerhardusvos wrote: Thu Nov 26, 2020 9:37 am
deanmoriarty wrote: Thu Nov 26, 2020 9:06 am For my early retirement, at around 36 in a year or two (hopefully somewhere around $4M NW), I plan for 1.5% withdrawal rate.

Why? A lot of people on the forum feel that 3-4% should be safe. However, expenses are not always constant, especially for early retirement.

The natural consequence, in my mind, is to make your expected baseline of expenses a very low proportion of your net worth (1.5%). Alternatively, aggressively pad your current expenses to account for the much unexpected, and then use that number with a 3% rate. It’s the same thing.
Congrats on your upcoming retirement! I think you’re missing the point and also contradicting yourself in the statement. We don’t feel that 3 to 4% withdrawal rate is safe. It is and has been safe for extremely long periods of time. It’s just a fact. Having your baseline expenses at 3% withdrawal rate is historically completely (100%) safe, and anything else that could come up will very likely be covered because you are already taking an ultra safe withdrawal rate of 3%. You can also budget for the potential things that are unknown to ensure your nest egg supports an extra $10,000 per year if you wanted to. All of this can be done at or above a 3% withdrawal rate. I don’t think you realize how drastic of a difference there is in outcome between a 3.9% WR and a 3.8% WR. It dramatically changes the long-term outcome. It is beyond reasonable and way outside of what is necessary to go from a ridiculously safe 3.25% withdrawal rate down to 1.5% withdrawal rate. If you are able to do it, that is totally fine, more power to you. It’s just overwhelmingly unnecessary, especially if you have mostly equities. There is so much padding built into a 3% withdrawal rate, and people forget that. Over 30 year periods, 4% withdrawal rate has a ton of padding as well.

Especially since you have such control over your living expenses, I think you are in need of some recalibration here.
4% isn’t safe retiring at age 36. Why are you talking about 30 year periods in a scenario where it doesn’t apply?
Retiring at age 36, you have about a 50 to 55 year retirement in the best case of scenarios. Did you know the 4% WR has worked ~90% of the time with a 75/25 portfolio for those time horizons? And 70% of the time with a 5% WR... many will probably die way sooner than 90 years old...

If that’s not good enough (and it is for many people), success looks even better when considering Social Security, other possible income, and the ability to be a little bit flexible with the withdrawal rate, especially during down markets.
Unless he started working at age 1 how much Social Security will he get? Will he be eligible for Medicare?
Yes, he will certainly be eligible for Medicare at age 65. I said the following below regarding SS benefits up thread:
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.
The Sensible Steward
palanzo
Posts: 2146
Joined: Thu Oct 10, 2019 4:28 pm

Re: Withdrawal rate for an early retirement

Post by palanzo »

willthrill81 wrote: Fri Nov 27, 2020 10:18 pm
palanzo wrote: Fri Nov 27, 2020 10:03 pm
geerhardusvos wrote: Fri Nov 27, 2020 9:58 pm
nigel_ht wrote: Fri Nov 27, 2020 7:47 pm
geerhardusvos wrote: Thu Nov 26, 2020 9:37 am

Congrats on your upcoming retirement! I think you’re missing the point and also contradicting yourself in the statement. We don’t feel that 3 to 4% withdrawal rate is safe. It is and has been safe for extremely long periods of time. It’s just a fact. Having your baseline expenses at 3% withdrawal rate is historically completely (100%) safe, and anything else that could come up will very likely be covered because you are already taking an ultra safe withdrawal rate of 3%. You can also budget for the potential things that are unknown to ensure your nest egg supports an extra $10,000 per year if you wanted to. All of this can be done at or above a 3% withdrawal rate. I don’t think you realize how drastic of a difference there is in outcome between a 3.9% WR and a 3.8% WR. It dramatically changes the long-term outcome. It is beyond reasonable and way outside of what is necessary to go from a ridiculously safe 3.25% withdrawal rate down to 1.5% withdrawal rate. If you are able to do it, that is totally fine, more power to you. It’s just overwhelmingly unnecessary, especially if you have mostly equities. There is so much padding built into a 3% withdrawal rate, and people forget that. Over 30 year periods, 4% withdrawal rate has a ton of padding as well.

Especially since you have such control over your living expenses, I think you are in need of some recalibration here.
4% isn’t safe retiring at age 36. Why are you talking about 30 year periods in a scenario where it doesn’t apply?
Retiring at age 36, you have about a 50 to 55 year retirement in the best case of scenarios. Did you know the 4% WR has worked ~90% of the time with a 75/25 portfolio for those time horizons? And 70% of the time with a 5% WR... many will probably die way sooner than 90 years old...

If that’s not good enough (and it is for many people), success looks even better when considering Social Security, other possible income, and the ability to be a little bit flexible with the withdrawal rate, especially during down markets.
Unless he started working at age 1 how much Social Security will he get? Will he be eligible for Medicare?
Yes, he will certainly be eligible for Medicare at age 65. I said the following below regarding SS benefits up thread:
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.
You're right. He will be eligible. We don't know if he has enough credits to get free Part A.

I am referring to deanmoriarty not geerhardusvos. High NW but not necessarily high taxable Social Security income.
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 10:22 am
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%.
We must be looking at other data. I'm talking about the 60 year retirements. The SWR for 60/40 static was 3.325%. The 60->100 glide path improved it to 3.561%, that's a 6.6% improvement. With a 2M portfolio that would translate to a $ 4720 per year difference. Even if you believe that's not statistically significant, it didn't perform worse than the static allocations (for SWR). In the 2 other worst times to retire it did substantially better than 60/40 static.

This is all with the dataset and methods that ERN worked with. He tends to use similar methods for his other articles. Historic back testing has serious issues, I'm aware.

In Kitces' calculations for shorter retirements the best equity glide path beats all static allocations, sometimes with a very small difference (likely statistically insignificant), sometimes with larger differences.

Dr. David Graham (FiPhysician) also tested glide paths. He's the fourth one to find that glide paths beat 60/40 static.
https://www.fiphysician.com/modeling-ri ... Glidepaths

Is it all just noise, flukes and data mining? I suspect you will say 'yes' but I will say 'probably not'.
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.
It could be chance. But it could also be a more effective AA during particularly bad times to retire.

Now we have 4 (I'd say) reasonably smart people who claim that glide paths can alleviate SoRR. And they also have some data to back it up.

Could they be wrong? Sure. But I'd have to see compelling data to believe that.

For anyone interested, Karsten Jeske was on the Morningstar podcast (Oct 2020) and he talks about glide paths:
https://the-long-view.simplecast.com/ep ... transcript
marcopolo
Posts: 8411
Joined: Sat Dec 03, 2016 9:22 am

Re: Withdrawal rate for an early retirement

Post by marcopolo »

YRT70 wrote: Sat Nov 28, 2020 1:52 am
vineviz wrote: Fri Nov 27, 2020 10:22 am
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%.
We must be looking at other data. I'm talking about the 60 year retirements. The SWR for 60/40 static was 3.325%. The 60->100 glide path improved it to 3.561%, that's a 6.6% improvement. With a 2M portfolio that would translate to a $ 4720 per year difference. Even if you believe that's not statistically significant, it didn't perform worse than the static allocations (for SWR). In the 2 other worst times to retire it did substantially better than 60/40 static.

This is all with the dataset and methods that ERN worked with. He tends to use similar methods for his other articles. Historic back testing has serious issues, I'm aware.

In Kitces' calculations for shorter retirements the best equity glide path beats all static allocations, sometimes with a very small difference (likely statistically insignificant), sometimes with larger differences.

Dr. David Graham (FiPhysician) also tested glide paths. He's the fourth one to find that glide paths beat 60/40 static.
https://www.fiphysician.com/modeling-ri ... Glidepaths

Is it all just noise, flukes and data mining? I suspect you will say 'yes' but I will say 'probably not'.
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.
It could be chance. But it could also be a more effective AA during particularly bad times to retire.

Now we have 4 (I'd say) reasonably smart people who claim that glide paths can alleviate SoRR. And they also have some data to back it up.

Could they be wrong? Sure. But I'd have to see compelling data to believe that.

For anyone interested, Karsten Jeske was on the Morningstar podcast (Oct 2020) and he talks about glide paths:
https://the-long-view.simplecast.com/ep ... transcript
Not sure how much the assertion is strengthened by having four people agree that it help a little, aren't they all mining the same data?

From what i have read, the improvement, if any, seems to be kind of marginal, and probably has to be weighed against behavioral risk you might introduce with high equity allocations (and higher volatility) at advanced age.
Once in a while you get shown the light, in the strangest of places if you look at it right.
YRT70
Posts: 1289
Joined: Sat Apr 27, 2019 8:51 am

Re: Withdrawal rate for an early retirement

Post by YRT70 »

marcopolo wrote: Sat Nov 28, 2020 3:00 am
YRT70 wrote: Sat Nov 28, 2020 1:52 am
vineviz wrote: Fri Nov 27, 2020 10:22 am
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%.
We must be looking at other data. I'm talking about the 60 year retirements. The SWR for 60/40 static was 3.325%. The 60->100 glide path improved it to 3.561%, that's a 6.6% improvement. With a 2M portfolio that would translate to a $ 4720 per year difference. Even if you believe that's not statistically significant, it didn't perform worse than the static allocations (for SWR). In the 2 other worst times to retire it did substantially better than 60/40 static.

This is all with the dataset and methods that ERN worked with. He tends to use similar methods for his other articles. Historic back testing has serious issues, I'm aware.

In Kitces' calculations for shorter retirements the best equity glide path beats all static allocations, sometimes with a very small difference (likely statistically insignificant), sometimes with larger differences.

Dr. David Graham (FiPhysician) also tested glide paths. He's the fourth one to find that glide paths beat 60/40 static.
https://www.fiphysician.com/modeling-ri ... Glidepaths

Is it all just noise, flukes and data mining? I suspect you will say 'yes' but I will say 'probably not'.
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.
It could be chance. But it could also be a more effective AA during particularly bad times to retire.

Now we have 4 (I'd say) reasonably smart people who claim that glide paths can alleviate SoRR. And they also have some data to back it up.

Could they be wrong? Sure. But I'd have to see compelling data to believe that.

For anyone interested, Karsten Jeske was on the Morningstar podcast (Oct 2020) and he talks about glide paths:
https://the-long-view.simplecast.com/ep ... transcript
Not sure how much the assertion is strengthened by having four people agree that it help a little, aren't they all mining the same data?
I don't think so. ERN uses the historic data set. Kitces, Pfau and FiPhysician use simulation tools with various assumptions.

From what i have read, the improvement, if any, seems to be kind of marginal, and probably has to be weighed against behavioral risk you might introduce with high equity allocations (and higher volatility) at advanced age.
I agree on the behavioural risks. For longer retirements the improvement in SWR was sometimes more than marginal.

Image
Source: https://earlyretirementnow.com/2017/09/ ... lidepaths/
KarenC
Posts: 263
Joined: Mon Apr 27, 2015 7:25 am

Re: Withdrawal rate for an early retirement

Post by KarenC »

klaus14 wrote: Fri Nov 27, 2020 6:24 pm
willthrill81 wrote: Fri Nov 27, 2020 5:23 pm We have no real risk of reducing withdrawals too much in this way because SS benefits will cover all of our essential spending from age 70 forward, leaving whatever remains in our portfolio for discretionary spending and possible LTC expenses.

I'd like to see a concrete example of how what you're recommending is objectively better. Right now, it looks like six of one and half a dozen of another.
Intuitively there is an equivalence between two methods, I was just saying g is more explicit. For example, you don't know how much you are reducing your consumption at year 20 with your method. You can calculate it obviously but it's not just there like g method.
I've incorporated the g parameter into my custom versions of the VPW, siamond's ASB, and ABW spreadsheets. (Happily, the "year 0" W/D numbers are pretty close to each other.) One nice thing about the ABW spreadsheet is that it'll show how a particular g value will affect withdrawals in future (modulo actual returns, of course); that makes it easy to tweak the g value to one's personal comfort. (I'm using a negative g value because I think it makes sense to front-load withdrawals.)
"The first principle is that you must not fool yourself—and you are the easiest person to fool." — Richard P. Feynman
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: Sat Nov 28, 2020 3:00 am Not sure how much the assertion is strengthened by having four people agree that it help a little, aren't they all mining the same data?

From what i have read, the improvement, if any, seems to be kind of marginal, and probably has to be weighed against behavioral risk you might introduce with high equity allocations (and higher volatility) at advanced age.
Indeed, having multiple sources that all repeat variations of the same mistakes isn’t really helpful.

A rising equity glide path doesn’t reliably change the SWR, but an unreliable analysis might make it appear to do so.
"Far more money has been lost by investors preparing for corrections than has been lost in corrections themselves." ~~ Peter Lynch
Marseille07
Posts: 16054
Joined: Fri Nov 06, 2020 12:41 pm

Re: Withdrawal rate for an early retirement

Post by Marseille07 »

YRT70 wrote: Sat Nov 28, 2020 3:31 am I agree on the behavioural risks. For longer retirements the improvement in SWR was sometimes more than marginal.

Image
Source: https://earlyretirementnow.com/2017/09/ ... lidepaths/
I don't like this chart because it is not testing rising glidepath. "Worst time to retire" (i.e. right before a market crash) is not the same as "worst time for rising glidpath."

In order to really test what rising glidepath is made of, we need to look at the years such that a retirement horizon is *ending* when the markets crashed, such as 2009-01-01.
klaus14
Posts: 953
Joined: Sun Nov 25, 2018 6:43 pm

Re: Withdrawal rate for an early retirement

Post by klaus14 »

vineviz wrote: Fri Nov 27, 2020 8:46 pm
klaus14 wrote: Fri Nov 27, 2020 4:49 pm
Do you like a static allocation during retirement or increasing bond allocation like Vanguard does?
(i think everyone agrees bond alloc should go up till retirement and disagreement is about what to do afterwards)
I can’t think of a scenario in which a rising bond allocation makes sense.

And to be fair, some investors might prefer a rising equity allocation DESPITE the fact that it doesn’t increase the expected SWR. High degrees of loss aversion or regret aversion might cause such a preference.
I see. Then you think static allocation through retirement is reasonable.
My investment algorithm: https://www.bogleheads.org/forum/viewtopic.php?f=10&t=351899&p=6112869#p6112869
Aelron
Posts: 71
Joined: Wed Oct 07, 2020 6:49 pm

Re: Withdrawal rate for an early retirement

Post by Aelron »

Rising equity allocation during retirement doesn't seem to make sense to me, and I think I've put my finger on the reason. The studies advocating for that tend to define success or failure depending upon whether withdrawals could be maintained for the entire retirement. While that seems reasonable on first look, I don't think it captures people's actual preferences. I know I'd much prefer to have my portfolio be exhausted 1 year before the end of retirement rather than 5 years before the end of retirement, although both cases are equally considered a "failure" in many studies. Given the simple success/failure metric, someone late in retirement after a market downturn may realize that no reasonable return from a 50/50 portfolio could have any chance of success, but an 80/20 portfolio did have some chance of success. They would then increase the equity allocation to 80/20. However, the thing that this overlooks is that while the chance of the portfolio lasting until the end of retirement increases, so does the chance of having a more spectacular failure (failing earlier).
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 »

klaus14 wrote: Sat Nov 28, 2020 3:50 pm I see. Then you think static allocation through retirement is reasonable.
I do.
"Far more money has been lost by investors preparing for corrections than has been lost in corrections themselves." ~~ Peter Lynch
klaus14
Posts: 953
Joined: Sun Nov 25, 2018 6:43 pm

Re: Withdrawal rate for an early retirement

Post by klaus14 »

vineviz wrote: Sat Nov 28, 2020 4:19 pm
klaus14 wrote: Sat Nov 28, 2020 3:50 pm I see. Then you think static allocation through retirement is reasonable.
I do.
I think this makes sense. Especially for early retirees. Because retirement year 1 is not that much different than year 10. You still have a very long retirement horizon.
My investment algorithm: https://www.bogleheads.org/forum/viewtopic.php?f=10&t=351899&p=6112869#p6112869
Marseille07
Posts: 16054
Joined: Fri Nov 06, 2020 12:41 pm

Re: Withdrawal rate for an early retirement

Post by Marseille07 »

klaus14 wrote: Sat Nov 28, 2020 4:22 pm
vineviz wrote: Sat Nov 28, 2020 4:19 pm
klaus14 wrote: Sat Nov 28, 2020 3:50 pm I see. Then you think static allocation through retirement is reasonable.
I do.
I think this makes sense. Especially for early retirees. Because retirement year 1 is not that much different than year 10. You still have a very long retirement horizon.
I do too. As long as my expenses are covered, I'm fine being conservative and possibly leaving too much on the table. I'd much rather be in that situation than running out of money.
Marseille07
Posts: 16054
Joined: Fri Nov 06, 2020 12:41 pm

Re: Withdrawal rate for an early retirement

Post by Marseille07 »

Aelron wrote: Sat Nov 28, 2020 4:12 pm Rising equity allocation during retirement doesn't seem to make sense to me, and I think I've put my finger on the reason. The studies advocating for that tend to define success or failure depending upon whether withdrawals could be maintained for the entire retirement. While that seems reasonable on first look, I don't think it captures people's actual preferences. I know I'd much prefer to have my portfolio be exhausted 1 year before the end of retirement rather than 5 years before the end of retirement, although both cases are equally considered a "failure" in many studies. Given the simple success/failure metric, someone late in retirement after a market downturn may realize that no reasonable return from a 50/50 portfolio could have any chance of success, but an 80/20 portfolio did have some chance of success. They would then increase the equity allocation to 80/20. However, the thing that this overlooks is that while the chance of the portfolio lasting until the end of retirement increases, so does the chance of having a more spectacular failure (failing earlier).
It doesn't make much sense. The issue is, studies presented focus on "bad start years" of retirement, not "bad end years" of retirement when rising glidepath suffers bigly. Obviously, if rising glidpath starts off bad years with 60/40 then ends with 100/0 when the markets are good, it would perform well. What we want to know is the opposite of that.
EnjoyIt
Posts: 8244
Joined: Sun Dec 29, 2013 7:06 pm

Re: Withdrawal rate for an early retirement

Post by EnjoyIt »

klaus14 wrote: Sat Nov 28, 2020 4:22 pm
vineviz wrote: Sat Nov 28, 2020 4:19 pm
klaus14 wrote: Sat Nov 28, 2020 3:50 pm I see. Then you think static allocation through retirement is reasonable.
I do.
I think this makes sense. Especially for early retirees. Because retirement year 1 is not that much different than year 10. You still have a very long retirement horizon.
It is and it isn’t. It is because technically every year you can restart your 4%. It isn’t because historically the market returns well over 4% and in 10 years if the worst hasn’t occurred, very likely the portfolio is much larger and can sustain larger/longer drawdowns.
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 »

EnjoyIt wrote: Sat Nov 28, 2020 6:26 pm
klaus14 wrote: Sat Nov 28, 2020 4:22 pm
vineviz wrote: Sat Nov 28, 2020 4:19 pm
klaus14 wrote: Sat Nov 28, 2020 3:50 pm I see. Then you think static allocation through retirement is reasonable.
I do.
I think this makes sense. Especially for early retirees. Because retirement year 1 is not that much different than year 10. You still have a very long retirement horizon.
It is and it isn’t. It is because technically every year you can restart your 4%. It isn’t because historically the market returns well over 4% and in 10 years if the worst hasn’t occurred, very likely the portfolio is much larger and can sustain larger/longer drawdowns.
That's the beauty of constant-percentage. Your withdrawals automatically adjust.
Last edited by Marseille07 on Sat Nov 28, 2020 6:38 pm, edited 2 times in total.
EnjoyIt
Posts: 8244
Joined: Sun Dec 29, 2013 7:06 pm

Re: Withdrawal rate for an early retirement

Post by EnjoyIt »

Marseille07 wrote: Sat Nov 28, 2020 6:36 pm
EnjoyIt wrote: Sat Nov 28, 2020 6:26 pm
klaus14 wrote: Sat Nov 28, 2020 4:22 pm
vineviz wrote: Sat Nov 28, 2020 4:19 pm
klaus14 wrote: Sat Nov 28, 2020 3:50 pm I see. Then you think static allocation through retirement is reasonable.
I do.
I think this makes sense. Especially for early retirees. Because retirement year 1 is not that much different than year 10. You still have a very long retirement horizon.
It is and it isn’t. It is because technically every year you can restart your 4%. It isn’t because historically the market returns well over 4% and in 10 years if the worst hasn’t occurred, very likely the portfolio is much larger and can sustain larger/longer drawdowns.
That's the beauty of constant-percentage. Your withdrawals automatically adjust.
And, you can never run out.
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: Sat Nov 28, 2020 6:36 pm
EnjoyIt wrote: Sat Nov 28, 2020 6:26 pm
klaus14 wrote: Sat Nov 28, 2020 4:22 pm
vineviz wrote: Sat Nov 28, 2020 4:19 pm
klaus14 wrote: Sat Nov 28, 2020 3:50 pm I see. Then you think static allocation through retirement is reasonable.
I do.
I think this makes sense. Especially for early retirees. Because retirement year 1 is not that much different than year 10. You still have a very long retirement horizon.
It is and it isn’t. It is because technically every year you can restart your 4%. It isn’t because historically the market returns well over 4% and in 10 years if the worst hasn’t occurred, very likely the portfolio is much larger and can sustain larger/longer drawdowns.
That's the beauty of constant-percentage. Your withdrawals automatically adjust.
The big problem with constant-percentage is that it makes your withdrawals as volatile as your portfolio.
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: Sat Nov 28, 2020 7:50 pm
Marseille07 wrote: Sat Nov 28, 2020 6:36 pm
EnjoyIt wrote: Sat Nov 28, 2020 6:26 pm
klaus14 wrote: Sat Nov 28, 2020 4:22 pm
vineviz wrote: Sat Nov 28, 2020 4:19 pm

I do.
I think this makes sense. Especially for early retirees. Because retirement year 1 is not that much different than year 10. You still have a very long retirement horizon.
It is and it isn’t. It is because technically every year you can restart your 4%. It isn’t because historically the market returns well over 4% and in 10 years if the worst hasn’t occurred, very likely the portfolio is much larger and can sustain larger/longer drawdowns.
That's the beauty of constant-percentage. Your withdrawals automatically adjust.
The big problem with constant-percentage is that it makes your withdrawals as volatile as your portfolio.
Well, we already went over it though. My plan is to simply have 20K in cash then 80K more in EF, virtually tracked in excel. At that point, it's a moot point if my withdrawal was 6K in one month and 4.5K in another.
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: Sat Nov 28, 2020 7:52 pm
willthrill81 wrote: Sat Nov 28, 2020 7:50 pm
Marseille07 wrote: Sat Nov 28, 2020 6:36 pm
EnjoyIt wrote: Sat Nov 28, 2020 6:26 pm
klaus14 wrote: Sat Nov 28, 2020 4:22 pm

I think this makes sense. Especially for early retirees. Because retirement year 1 is not that much different than year 10. You still have a very long retirement horizon.
It is and it isn’t. It is because technically every year you can restart your 4%. It isn’t because historically the market returns well over 4% and in 10 years if the worst hasn’t occurred, very likely the portfolio is much larger and can sustain larger/longer drawdowns.
That's the beauty of constant-percentage. Your withdrawals automatically adjust.
The big problem with constant-percentage is that it makes your withdrawals as volatile as your portfolio.
Well, we already went over it though. My plan is to simply have 20K in cash then 80K more in EF, virtually tracked in excel. At that point, it's a moot point if my withdrawal was 6K in one month and 4.5K in another.
If the market doesn't drop too much for too long, that will work. But a 60/40 portfolio with 4% withdrawn as a constant percentage was down 46% in inflation-adjusted dollars from Jan., 2000, to Feb., 2009, and didn't fully recover until August of this year. I doubt that your EF would cover that gap for that long.
Last edited by willthrill81 on Sat Nov 28, 2020 7:55 pm, edited 1 time in total.
The Sensible Steward
smitcat
Posts: 13227
Joined: Mon Nov 07, 2016 9:51 am

Re: Withdrawal rate for an early retirement

Post by smitcat »

Marseille07 wrote: Sat Nov 28, 2020 7:52 pm
willthrill81 wrote: Sat Nov 28, 2020 7:50 pm
Marseille07 wrote: Sat Nov 28, 2020 6:36 pm
EnjoyIt wrote: Sat Nov 28, 2020 6:26 pm
klaus14 wrote: Sat Nov 28, 2020 4:22 pm

I think this makes sense. Especially for early retirees. Because retirement year 1 is not that much different than year 10. You still have a very long retirement horizon.
It is and it isn’t. It is because technically every year you can restart your 4%. It isn’t because historically the market returns well over 4% and in 10 years if the worst hasn’t occurred, very likely the portfolio is much larger and can sustain larger/longer drawdowns.
That's the beauty of constant-percentage. Your withdrawals automatically adjust.
The big problem with constant-percentage is that it makes your withdrawals as volatile as your portfolio.
Well, we already went over it though. My plan is to simply have 20K in cash then 80K more in EF, virtually tracked in excel. At that point, it's a moot point if my withdrawal was 6K in one month and 4.5K in another.
Curious what would happen if you were to burn through 20% of the $100K buffer?
What about if you burned through 50% of that $100K buffer?
Marseille07
Posts: 16054
Joined: Fri Nov 06, 2020 12:41 pm

Re: Withdrawal rate for an early retirement

Post by Marseille07 »

smitcat wrote: Sat Nov 28, 2020 7:55 pm
Marseille07 wrote: Sat Nov 28, 2020 7:52 pm
willthrill81 wrote: Sat Nov 28, 2020 7:50 pm
Marseille07 wrote: Sat Nov 28, 2020 6:36 pm
EnjoyIt wrote: Sat Nov 28, 2020 6:26 pm

It is and it isn’t. It is because technically every year you can restart your 4%. It isn’t because historically the market returns well over 4% and in 10 years if the worst hasn’t occurred, very likely the portfolio is much larger and can sustain larger/longer drawdowns.
That's the beauty of constant-percentage. Your withdrawals automatically adjust.
The big problem with constant-percentage is that it makes your withdrawals as volatile as your portfolio.
Well, we already went over it though. My plan is to simply have 20K in cash then 80K more in EF, virtually tracked in excel. At that point, it's a moot point if my withdrawal was 6K in one month and 4.5K in another.
Curious what would happen if you were to burn through 20% of the $100K buffer?
What about if you burned through 50% of that $100K buffer?
Since the EF is tracked virtually, it's still part of AA. I'd have to rebalance enough to generate the cash then monthly withdrawals would start refilling the gap back up.

Last time I said I would have 100K of EF in cash, I got jumped on by people saying I'm leaving too much money on the table uninvested. They can't argue now I have too little on the table :-D
Marseille07
Posts: 16054
Joined: Fri Nov 06, 2020 12:41 pm

Re: Withdrawal rate for an early retirement

Post by Marseille07 »

willthrill81 wrote: Sat Nov 28, 2020 7:55 pm
Marseille07 wrote: Sat Nov 28, 2020 7:52 pm
willthrill81 wrote: Sat Nov 28, 2020 7:50 pm
Marseille07 wrote: Sat Nov 28, 2020 6:36 pm
EnjoyIt wrote: Sat Nov 28, 2020 6:26 pm

It is and it isn’t. It is because technically every year you can restart your 4%. It isn’t because historically the market returns well over 4% and in 10 years if the worst hasn’t occurred, very likely the portfolio is much larger and can sustain larger/longer drawdowns.
That's the beauty of constant-percentage. Your withdrawals automatically adjust.
The big problem with constant-percentage is that it makes your withdrawals as volatile as your portfolio.
Well, we already went over it though. My plan is to simply have 20K in cash then 80K more in EF, virtually tracked in excel. At that point, it's a moot point if my withdrawal was 6K in one month and 4.5K in another.
If the market doesn't drop too much for too long, that will work. But a 60/40 portfolio with 4% withdrawn as a constant percentage was down 46% in inflation-adjusted dollars from Jan., 2000, to Feb., 2009, and didn't fully recover until August of this year. I doubt that your EF would cover that gap for that long.
Not sure if I follow the point you're making. ABW presenting you with a fix percentage number for the whole year would run into the same fate. If you disagree, please explain how.

EDIT: I think I know where you're coming from. ABW would make adjustments annually to stabilize withdrawals. I think that's a fair point, though there's a cost of doing that (no free lunch).
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: Sat Nov 28, 2020 10:01 pm
willthrill81 wrote: Sat Nov 28, 2020 7:55 pm
Marseille07 wrote: Sat Nov 28, 2020 7:52 pm
willthrill81 wrote: Sat Nov 28, 2020 7:50 pm
Marseille07 wrote: Sat Nov 28, 2020 6:36 pm

That's the beauty of constant-percentage. Your withdrawals automatically adjust.
The big problem with constant-percentage is that it makes your withdrawals as volatile as your portfolio.
Well, we already went over it though. My plan is to simply have 20K in cash then 80K more in EF, virtually tracked in excel. At that point, it's a moot point if my withdrawal was 6K in one month and 4.5K in another.
If the market doesn't drop too much for too long, that will work. But a 60/40 portfolio with 4% withdrawn as a constant percentage was down 46% in inflation-adjusted dollars from Jan., 2000, to Feb., 2009, and didn't fully recover until August of this year. I doubt that your EF would cover that gap for that long.
Not sure if I follow the point you're making. ABW presenting you with a fix percentage number for the whole year would run into the same fate. If you disagree, please explain how.
Not if you're using dynamic returns in your withdrawal model, as noted in the OP of this thread. Withdrawals barely skipped a beat throughout the GFC.
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: Sat Nov 28, 2020 10:06 pm
Marseille07 wrote: Sat Nov 28, 2020 10:01 pm
willthrill81 wrote: Sat Nov 28, 2020 7:55 pm
Marseille07 wrote: Sat Nov 28, 2020 7:52 pm
willthrill81 wrote: Sat Nov 28, 2020 7:50 pm

The big problem with constant-percentage is that it makes your withdrawals as volatile as your portfolio.
Well, we already went over it though. My plan is to simply have 20K in cash then 80K more in EF, virtually tracked in excel. At that point, it's a moot point if my withdrawal was 6K in one month and 4.5K in another.
If the market doesn't drop too much for too long, that will work. But a 60/40 portfolio with 4% withdrawn as a constant percentage was down 46% in inflation-adjusted dollars from Jan., 2000, to Feb., 2009, and didn't fully recover until August of this year. I doubt that your EF would cover that gap for that long.
Not sure if I follow the point you're making. ABW presenting you with a fix percentage number for the whole year would run into the same fate. If you disagree, please explain how.
Not if you're using dynamic returns in your withdrawal model, as noted in the OP of this thread. Withdrawals barely skipped a beat throughout the GFC.
Right. While I don't deny the benefit of that approach, I think enough EF can absorb the shock. It's not like monthly withdrawals would stop completely - 6K/mo might become 4K/mo but that's still some money to refill your EF buffer. And obviously there's no reason I can't cut spending seeing Lehman Brothers fall.

Just a thought experiment, but we all have some EF buffer that's sufficient. I would say for most people's monthly day-to-day needs are met with 20K just fine. But if for whatever reason that's not sufficient, prepare 50K, 100K or whatever. The point is, there's some amount somewhere above which you stop worrying about portfolio volatility because you have enough cash stocked up.
marcopolo
Posts: 8411
Joined: Sat Dec 03, 2016 9:22 am

Re: Withdrawal rate for an early retirement

Post by marcopolo »

Marseille07 wrote: Sat Nov 28, 2020 7:52 pm
willthrill81 wrote: Sat Nov 28, 2020 7:50 pm
Marseille07 wrote: Sat Nov 28, 2020 6:36 pm
EnjoyIt wrote: Sat Nov 28, 2020 6:26 pm
klaus14 wrote: Sat Nov 28, 2020 4:22 pm

I think this makes sense. Especially for early retirees. Because retirement year 1 is not that much different than year 10. You still have a very long retirement horizon.
It is and it isn’t. It is because technically every year you can restart your 4%. It isn’t because historically the market returns well over 4% and in 10 years if the worst hasn’t occurred, very likely the portfolio is much larger and can sustain larger/longer drawdowns.
That's the beauty of constant-percentage. Your withdrawals automatically adjust.
The big problem with constant-percentage is that it makes your withdrawals as volatile as your portfolio.
Well, we already went over it though. My plan is to simply have 20K in cash then 80K more in EF, virtually tracked in excel. At that point, it's a moot point if my withdrawal was 6K in one month and 4.5K in another.

What does it mean to virtually track 80k in an EF?
What is it actually invested in?
What advantage is there to this mental, or in this case spreadsheet accounting of dollars in different virtual buckets?

During working years I can see keeping an EF as a way to get through possible periods of unemployment without touching your portfolio.

When you are in retire, it seem your entire portfolio is essentially your EF. Pull from it what you need.
Once in a while you get shown the light, in the strangest of places if you look at it right.
Marseille07
Posts: 16054
Joined: Fri Nov 06, 2020 12:41 pm

Re: Withdrawal rate for an early retirement

Post by Marseille07 »

marcopolo wrote: Sat Nov 28, 2020 11:03 pm What does it mean to virtually track 80k in an EF?
What is it actually invested in?
What advantage is there to this mental, or in this case spreadsheet accounting of dollars in different virtual buckets?

During working years I can see keeping an EF as a way to get through possible periods of unemployment without touching your portfolio.

When you are in retire, it seem your entire portfolio is essentially your EF. Pull from it what you need.
The idea isn't that complicated, instead of actually selling AA and withdrawing cash, you just keep tabs on a spreadsheet so the money stays invested in the AA.

"Pull from it what you need" is right, but you need some way to track your spending to avoid overspending. That's what I use virtual buckets for.
Post Reply