Time Value of Money vs. Variable Percentage Withdrawal (VPW)

Non-investing personal finance issues including insurance, credit, real estate, taxes, employment and legal issues such as trusts and wills
User avatar
Topic Author
willthrill81
Posts: 10324
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Time Value of Money vs. Variable Percentage Withdrawal (VPW)

Post by willthrill81 » Mon Nov 05, 2018 10:33 am

[Split into a new topic from: Variable Percentage Withdrawal (VPW) --admin LadyGeek]
azanon wrote:
Mon Nov 05, 2018 8:55 am
jmk wrote:
Mon Nov 05, 2018 12:00 am
Looking at longevityillustrator for instance, I see that there is a 50% chance at 55 that either my wife and I will live to 100. So if I were to retire today this would be a reasonable number to use in the vpw equations for my first years of withdrawals. (100-55=45).
I'd find another actuarial website. 50% chance either you or your wife hit 100?!? That's extremely optimistic. Though I realize longevity is improving, today only 0.0173% of Americans live to 100. And don't confuse that with 1%. That's 0.0173% chance (source: http://www.genealogyintime.com/Genealog ... page1.html ). Now granted, I realize there's a slight increase starting at age 55 instead of 0, but it's not going to bump up that high!

If you want to round off to a whole number, for all intents and purposes, people don't live to 100.
Maybe not. The problem with the data you're referring to is that they are for the general populace and not a specific person. Factors such as health, ethnicity, wealth, family history, certain behaviors, etc. all have a significant relationship with longevity. For instance, depending on which longevity calculator I use, my (median) life expectancy is predicted to be anywhere from 92 to 101. If we amalgamate those, I probably have at least a 1/3 chance of surviving to 100. As such, VPW doesn't appeal to me personally because I'm not a huge fan of SPIAs.

For those wanting a more 'hands on' approach, a simple time value of money calculator can be used instead. You simply input the dollar amount of your portfolio that you wish to amortize (it could be all of your portfolio or some portion of it) over your remaining lifetime, the assumed real rate of return, and the period of time you want to amortize payments over. For instance, someone with $1 million to amortize over 30 years at a 3% real rate of return would withdraw $51k in the current year. This would be recalculated each subsequent year as the remaining amount of capital, the assumed rate of return, and the remaining years of withdrawals change. It only takes a few minutes at most once per year to do this.
“It's a dangerous business, Frodo, going out your door. You step onto the road, and if you don't keep your feet, there's no knowing where you might be swept off to.” J.R.R. Tolkien,The Lord of the Rings

azanon
Posts: 2376
Joined: Mon Nov 07, 2011 10:34 am
Location: Little Rock, AR
Contact:

Re: Variable Percentage Withdrawal (VPW)

Post by azanon » Tue Nov 06, 2018 11:33 am

willthrill81 wrote:
Mon Nov 05, 2018 10:33 am
azanon wrote:
Mon Nov 05, 2018 8:55 am
jmk wrote:
Mon Nov 05, 2018 12:00 am
Looking at longevityillustrator for instance, I see that there is a 50% chance at 55 that either my wife and I will live to 100. So if I were to retire today this would be a reasonable number to use in the vpw equations for my first years of withdrawals. (100-55=45).
I'd find another actuarial website. 50% chance either you or your wife hit 100?!? That's extremely optimistic. Though I realize longevity is improving, today only 0.0173% of Americans live to 100. And don't confuse that with 1%. That's 0.0173% chance (source: http://www.genealogyintime.com/Genealog ... page1.html ). Now granted, I realize there's a slight increase starting at age 55 instead of 0, but it's not going to bump up that high!

If you want to round off to a whole number, for all intents and purposes, people don't live to 100.
Maybe not. The problem with the data you're referring to is that they are for the general populace and not a specific person. Factors such as health, ethnicity, wealth, family history, certain behaviors, etc. all have a significant relationship with longevity. For instance, depending on which longevity calculator I use, my (median) life expectancy is predicted to be anywhere from 92 to 101. If we amalgamate those, I probably have at least a 1/3 chance of surviving to 100. As such, VPW doesn't appeal to me personally because I'm not a huge fan of SPIAs.

For those wanting a more 'hands on' approach, a simple time value of money calculator can be used instead. You simply input the dollar amount of your portfolio that you wish to amortize (it could be all of your portfolio or some portion of it) over your remaining lifetime, the assumed real rate of return, and the period of time you want to amortize payments over. For instance, someone with $1 million to amortize over 30 years at a 3% real rate of return would withdraw $51k in the current year. This would be recalculated each subsequent year as the remaining amount of capital, the assumed rate of return, and the remaining years of withdrawals change. It only takes a few minutes at most once per year to do this.
If you're a similar age to me, SPIAs might be quite a bit more attractive by the time we retire, if the prime rate keeps rising. I was browsing the data here this morning: http://www.fedprimerate.com/wall_street ... istory.htm and thought to myself, if the prime rate reaches at least mean (~ 7.0%), and especially if it passes 10%, by the time i retire, I might buy an inflation-indexed annuity at retirement up to my state guaranty limit (AR is 300K) with part of my 401(k) money (because the annuity interest rate will presumably also be high(er) too), and use VPW for the rest with a more stock-heavy portfolio than I would use otherwise because my floor is so high.

I admit VPW is more appealing to me though because I expect a federal pension to supplement my SS, and I'm naturally a low-risk kind of person evidenced by the fact that I am a fed.

User avatar
Topic Author
willthrill81
Posts: 10324
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: Variable Percentage Withdrawal (VPW)

Post by willthrill81 » Tue Nov 06, 2018 11:39 am

azanon wrote:
Tue Nov 06, 2018 11:33 am
willthrill81 wrote:
Mon Nov 05, 2018 10:33 am
azanon wrote:
Mon Nov 05, 2018 8:55 am
jmk wrote:
Mon Nov 05, 2018 12:00 am
Looking at longevityillustrator for instance, I see that there is a 50% chance at 55 that either my wife and I will live to 100. So if I were to retire today this would be a reasonable number to use in the vpw equations for my first years of withdrawals. (100-55=45).
I'd find another actuarial website. 50% chance either you or your wife hit 100?!? That's extremely optimistic. Though I realize longevity is improving, today only 0.0173% of Americans live to 100. And don't confuse that with 1%. That's 0.0173% chance (source: http://www.genealogyintime.com/Genealog ... page1.html ). Now granted, I realize there's a slight increase starting at age 55 instead of 0, but it's not going to bump up that high!

If you want to round off to a whole number, for all intents and purposes, people don't live to 100.
Maybe not. The problem with the data you're referring to is that they are for the general populace and not a specific person. Factors such as health, ethnicity, wealth, family history, certain behaviors, etc. all have a significant relationship with longevity. For instance, depending on which longevity calculator I use, my (median) life expectancy is predicted to be anywhere from 92 to 101. If we amalgamate those, I probably have at least a 1/3 chance of surviving to 100. As such, VPW doesn't appeal to me personally because I'm not a huge fan of SPIAs.

For those wanting a more 'hands on' approach, a simple time value of money calculator can be used instead. You simply input the dollar amount of your portfolio that you wish to amortize (it could be all of your portfolio or some portion of it) over your remaining lifetime, the assumed real rate of return, and the period of time you want to amortize payments over. For instance, someone with $1 million to amortize over 30 years at a 3% real rate of return would withdraw $51k in the current year. This would be recalculated each subsequent year as the remaining amount of capital, the assumed rate of return, and the remaining years of withdrawals change. It only takes a few minutes at most once per year to do this.
If you're a similar age to me, SPIAs might be quite a bit more attractive by the time we retire, if the prime rate keeps rising. I was browsing the data here this morning: http://www.fedprimerate.com/wall_street ... istory.htm and thought to myself, if the prime rate reaches at least mean (~ 7.0%), and especially if it passes 10%, by the time i retire, I might buy an inflation-indexed annuity at retirement up to my state guaranty limit (AR is 300K) with part of my 401(k) money (because the annuity interest rate will presumably also be high(er) too), and use VPW for the rest with a more stock-heavy portfolio than I would use otherwise because my floor is so high.

I admit VPW is more appealing to me though because I expect a federal pension to supplement my SS, and I'm naturally a low-risk kind of person evidenced by the fact that I am a fed.
Yes, higher payout rates in the future would be appealing, but I believe that only one company now offers inflation-indexed SPIAs. Considering that 70% of our SS benefits at age 70 would cover our anticipated essential spending, I'm not too concerned now about other non-portfolio sources of income.
“It's a dangerous business, Frodo, going out your door. You step onto the road, and if you don't keep your feet, there's no knowing where you might be swept off to.” J.R.R. Tolkien,The Lord of the Rings

azanon
Posts: 2376
Joined: Mon Nov 07, 2011 10:34 am
Location: Little Rock, AR
Contact:

Re: Variable Percentage Withdrawal (VPW)

Post by azanon » Tue Nov 06, 2018 11:55 am

willthrill81 wrote:
Tue Nov 06, 2018 11:39 am
azanon wrote:
Tue Nov 06, 2018 11:33 am
willthrill81 wrote:
Mon Nov 05, 2018 10:33 am
azanon wrote:
Mon Nov 05, 2018 8:55 am
jmk wrote:
Mon Nov 05, 2018 12:00 am
Looking at longevityillustrator for instance, I see that there is a 50% chance at 55 that either my wife and I will live to 100. So if I were to retire today this would be a reasonable number to use in the vpw equations for my first years of withdrawals. (100-55=45).
I'd find another actuarial website. 50% chance either you or your wife hit 100?!? That's extremely optimistic. Though I realize longevity is improving, today only 0.0173% of Americans live to 100. And don't confuse that with 1%. That's 0.0173% chance (source: http://www.genealogyintime.com/Genealog ... page1.html ). Now granted, I realize there's a slight increase starting at age 55 instead of 0, but it's not going to bump up that high!

If you want to round off to a whole number, for all intents and purposes, people don't live to 100.
Maybe not. The problem with the data you're referring to is that they are for the general populace and not a specific person. Factors such as health, ethnicity, wealth, family history, certain behaviors, etc. all have a significant relationship with longevity. For instance, depending on which longevity calculator I use, my (median) life expectancy is predicted to be anywhere from 92 to 101. If we amalgamate those, I probably have at least a 1/3 chance of surviving to 100. As such, VPW doesn't appeal to me personally because I'm not a huge fan of SPIAs.

For those wanting a more 'hands on' approach, a simple time value of money calculator can be used instead. You simply input the dollar amount of your portfolio that you wish to amortize (it could be all of your portfolio or some portion of it) over your remaining lifetime, the assumed real rate of return, and the period of time you want to amortize payments over. For instance, someone with $1 million to amortize over 30 years at a 3% real rate of return would withdraw $51k in the current year. This would be recalculated each subsequent year as the remaining amount of capital, the assumed rate of return, and the remaining years of withdrawals change. It only takes a few minutes at most once per year to do this.
If you're a similar age to me, SPIAs might be quite a bit more attractive by the time we retire, if the prime rate keeps rising. I was browsing the data here this morning: http://www.fedprimerate.com/wall_street ... istory.htm and thought to myself, if the prime rate reaches at least mean (~ 7.0%), and especially if it passes 10%, by the time i retire, I might buy an inflation-indexed annuity at retirement up to my state guaranty limit (AR is 300K) with part of my 401(k) money (because the annuity interest rate will presumably also be high(er) too), and use VPW for the rest with a more stock-heavy portfolio than I would use otherwise because my floor is so high.

I admit VPW is more appealing to me though because I expect a federal pension to supplement my SS, and I'm naturally a low-risk kind of person evidenced by the fact that I am a fed.
Yes, higher payout rates in the future would be appealing, but I believe that only one company now offers inflation-indexed SPIAs. Considering that 70% of our SS benefits at age 70 would cover our anticipated essential spending, I'm not too concerned now about other non-portfolio sources of income.
Oh I didn't know that was not commonly available. I know I can get one through my own retirement (federal FERS buys immediate annuities through Met-Life, and one option is an inflation-indexed one, though its capped at 3% annually I believe).

For the age 70 SS estimate, did you subtract 21% from the payment to be conservative (reference: cut in 2034 without reform)? Don't bite, just kidding a bit. For me, though, I'm not. I see how almost every year we don't have an appropriation from Congress to pay us and have to function under a CR for sometimes months, and threats of government shutdown, so I sure as heck don't trust Congress enough to defer SS. But that's just me.....

User avatar
Topic Author
willthrill81
Posts: 10324
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: Variable Percentage Withdrawal (VPW)

Post by willthrill81 » Tue Nov 06, 2018 12:00 pm

azanon wrote:
Tue Nov 06, 2018 11:55 am
willthrill81 wrote:
Tue Nov 06, 2018 11:39 am
azanon wrote:
Tue Nov 06, 2018 11:33 am
willthrill81 wrote:
Mon Nov 05, 2018 10:33 am
azanon wrote:
Mon Nov 05, 2018 8:55 am


I'd find another actuarial website. 50% chance either you or your wife hit 100?!? That's extremely optimistic. Though I realize longevity is improving, today only 0.0173% of Americans live to 100. And don't confuse that with 1%. That's 0.0173% chance (source: http://www.genealogyintime.com/Genealog ... page1.html ). Now granted, I realize there's a slight increase starting at age 55 instead of 0, but it's not going to bump up that high!

If you want to round off to a whole number, for all intents and purposes, people don't live to 100.
Maybe not. The problem with the data you're referring to is that they are for the general populace and not a specific person. Factors such as health, ethnicity, wealth, family history, certain behaviors, etc. all have a significant relationship with longevity. For instance, depending on which longevity calculator I use, my (median) life expectancy is predicted to be anywhere from 92 to 101. If we amalgamate those, I probably have at least a 1/3 chance of surviving to 100. As such, VPW doesn't appeal to me personally because I'm not a huge fan of SPIAs.

For those wanting a more 'hands on' approach, a simple time value of money calculator can be used instead. You simply input the dollar amount of your portfolio that you wish to amortize (it could be all of your portfolio or some portion of it) over your remaining lifetime, the assumed real rate of return, and the period of time you want to amortize payments over. For instance, someone with $1 million to amortize over 30 years at a 3% real rate of return would withdraw $51k in the current year. This would be recalculated each subsequent year as the remaining amount of capital, the assumed rate of return, and the remaining years of withdrawals change. It only takes a few minutes at most once per year to do this.
If you're a similar age to me, SPIAs might be quite a bit more attractive by the time we retire, if the prime rate keeps rising. I was browsing the data here this morning: http://www.fedprimerate.com/wall_street ... istory.htm and thought to myself, if the prime rate reaches at least mean (~ 7.0%), and especially if it passes 10%, by the time i retire, I might buy an inflation-indexed annuity at retirement up to my state guaranty limit (AR is 300K) with part of my 401(k) money (because the annuity interest rate will presumably also be high(er) too), and use VPW for the rest with a more stock-heavy portfolio than I would use otherwise because my floor is so high.

I admit VPW is more appealing to me though because I expect a federal pension to supplement my SS, and I'm naturally a low-risk kind of person evidenced by the fact that I am a fed.
Yes, higher payout rates in the future would be appealing, but I believe that only one company now offers inflation-indexed SPIAs. Considering that 70% of our SS benefits at age 70 would cover our anticipated essential spending, I'm not too concerned now about other non-portfolio sources of income.
Oh I didn't know that was not commonly available. I know I can get one through my own retirement (federal FERS buys immediate annuities through Met-Life, and one option is an inflation-indexed one, though its capped at 3% annually I believe).

For the age 70 SS estimate, did you subtract 21% from the payment to be conservative (reference: cut in 2034 without reform)? Don't bite, just kidding a bit. For me, though, I'm not. I see how almost every year we don't have an appropriation from Congress to pay us and have to function under a CR for sometimes months, and threats of government shutdown, so I sure as heck don't trust Congress enough to defer SS. But that's just me.....
It's easy to get a SPIA with a fixed rate of payout increases like 2-3%, but getting one that is inflation-indexed is nigh impossible these days.

The 70% number I referred to is based on currently scheduled SS benefits, so a 20-25% reduction would still enable us to cover our anticipated essential spending with SS benefits. But to be honest, I don't take SS into account at all for two reasons: (1) there is too much uncertainty about benefits, especially decades out, and (2) I want to be incentivized as much as possible to be solely responsible for all of our retirement income. I want whatever we get from SS to be the cherry on our already delicious frosted cake.
“It's a dangerous business, Frodo, going out your door. You step onto the road, and if you don't keep your feet, there's no knowing where you might be swept off to.” J.R.R. Tolkien,The Lord of the Rings

azanon
Posts: 2376
Joined: Mon Nov 07, 2011 10:34 am
Location: Little Rock, AR
Contact:

Re: Variable Percentage Withdrawal (VPW)

Post by azanon » Tue Nov 06, 2018 12:11 pm

willthrill81 wrote:
Tue Nov 06, 2018 12:00 pm
azanon wrote:
Tue Nov 06, 2018 11:55 am
willthrill81 wrote:
Tue Nov 06, 2018 11:39 am
azanon wrote:
Tue Nov 06, 2018 11:33 am
willthrill81 wrote:
Mon Nov 05, 2018 10:33 am


Maybe not. The problem with the data you're referring to is that they are for the general populace and not a specific person. Factors such as health, ethnicity, wealth, family history, certain behaviors, etc. all have a significant relationship with longevity. For instance, depending on which longevity calculator I use, my (median) life expectancy is predicted to be anywhere from 92 to 101. If we amalgamate those, I probably have at least a 1/3 chance of surviving to 100. As such, VPW doesn't appeal to me personally because I'm not a huge fan of SPIAs.

For those wanting a more 'hands on' approach, a simple time value of money calculator can be used instead. You simply input the dollar amount of your portfolio that you wish to amortize (it could be all of your portfolio or some portion of it) over your remaining lifetime, the assumed real rate of return, and the period of time you want to amortize payments over. For instance, someone with $1 million to amortize over 30 years at a 3% real rate of return would withdraw $51k in the current year. This would be recalculated each subsequent year as the remaining amount of capital, the assumed rate of return, and the remaining years of withdrawals change. It only takes a few minutes at most once per year to do this.
If you're a similar age to me, SPIAs might be quite a bit more attractive by the time we retire, if the prime rate keeps rising. I was browsing the data here this morning: http://www.fedprimerate.com/wall_street ... istory.htm and thought to myself, if the prime rate reaches at least mean (~ 7.0%), and especially if it passes 10%, by the time i retire, I might buy an inflation-indexed annuity at retirement up to my state guaranty limit (AR is 300K) with part of my 401(k) money (because the annuity interest rate will presumably also be high(er) too), and use VPW for the rest with a more stock-heavy portfolio than I would use otherwise because my floor is so high.

I admit VPW is more appealing to me though because I expect a federal pension to supplement my SS, and I'm naturally a low-risk kind of person evidenced by the fact that I am a fed.
Yes, higher payout rates in the future would be appealing, but I believe that only one company now offers inflation-indexed SPIAs. Considering that 70% of our SS benefits at age 70 would cover our anticipated essential spending, I'm not too concerned now about other non-portfolio sources of income.
Oh I didn't know that was not commonly available. I know I can get one through my own retirement (federal FERS buys immediate annuities through Met-Life, and one option is an inflation-indexed one, though its capped at 3% annually I believe).

For the age 70 SS estimate, did you subtract 21% from the payment to be conservative (reference: cut in 2034 without reform)? Don't bite, just kidding a bit. For me, though, I'm not. I see how almost every year we don't have an appropriation from Congress to pay us and have to function under a CR for sometimes months, and threats of government shutdown, so I sure as heck don't trust Congress enough to defer SS. But that's just me.....
It's easy to get a SPIA with a fixed rate of payout increases like 2-3%, but getting one that is inflation-indexed is nigh impossible these days.

The 70% number I referred to is based on currently scheduled SS benefits, so a 20-25% reduction would still enable us to cover our anticipated essential spending with SS benefits. But to be honest, I don't take SS into account at all for two reasons: (1) there is too much uncertainty about benefits, especially decades out, and (2) I want to be incentivized as much as possible to be solely responsible for all of our retirement income. I want whatever we get from SS to be the cherry on our already delicious frosted cake.
Hmm, ok I"m not sure if what I can get through FERS fits the former or latter category you describe, or is a hybrid of the two. It has an "increasing payment" option (vs. level payment), but the increase that you get is based on change in the CPI, and can't exceed 3% if the CPI is greater than 3%.

By taking SS early (and not "buying" an age 70 SS by spending down the portfolio), I'm also not dependent on SS either. But it would hurt pretty bad if the Fed government defaulted on my FERS pension. But gosh, if that happens, I imagine stocks and bonds aren't doing so hot either, because it would take something pretty disastrous for it to come to that!

User avatar
Lieutenant.Columbo
Posts: 1178
Joined: Sat Sep 05, 2015 9:20 pm
Location: Cicely AK

Re: Variable Percentage Withdrawal (VPW)

Post by Lieutenant.Columbo » Mon Nov 12, 2018 11:02 am

willthrill81 wrote:
Mon Nov 05, 2018 10:33 am
I probably have at least a 1/3 chance of surviving to 100
willthrill81,
what does this really mean? One out of 3 what?
willthrill81 wrote:
Mon Nov 05, 2018 10:33 am
For those wanting a more 'hands on' approach, a simple time value of money calculator can be used instead. You simply input the dollar amount of your portfolio that you wish to amortize (it could be all of your portfolio or some portion of it) over your remaining lifetime, the assumed real rate of return, and the period of time you want to amortize payments over. For instance, someone with $1 million to amortize over 30 years at a 3% real rate of return would withdraw $51k in the current year. This would be recalculated each subsequent year as the remaining amount of capital, the assumed rate of return, and the remaining years of withdrawals change. It only takes a few minutes at most once per year to do this.
Is there more on this that you'd recommend reading? This approach doesn't sound more 'hands-on' than VPW, is it?
I'd be interested to hear what longinvest thinks about your "time value of money calculator" suggestion.
Lt. Columbo: Well, what do you know. Here I am talking with some of the smartest people in the world, and I didn't even notice!

User avatar
Topic Author
willthrill81
Posts: 10324
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: Variable Percentage Withdrawal (VPW)

Post by willthrill81 » Mon Nov 12, 2018 11:24 am

Lieutenant.Columbo wrote:
Mon Nov 12, 2018 11:02 am
willthrill81 wrote:
Mon Nov 05, 2018 10:33 am
I probably have at least a 1/3 chance of surviving to 100
willthrill81,
what does this really mean? One out of 3 what?
I probably have a 33% probability of surviving to age 100.
Lieutenant.Columbo wrote:
Mon Nov 12, 2018 11:02 am
willthrill81 wrote:
Mon Nov 05, 2018 10:33 am
For those wanting a more 'hands on' approach, a simple time value of money calculator can be used instead. You simply input the dollar amount of your portfolio that you wish to amortize (it could be all of your portfolio or some portion of it) over your remaining lifetime, the assumed real rate of return, and the period of time you want to amortize payments over. For instance, someone with $1 million to amortize over 30 years at a 3% real rate of return would withdraw $51k in the current year. This would be recalculated each subsequent year as the remaining amount of capital, the assumed rate of return, and the remaining years of withdrawals change. It only takes a few minutes at most once per year to do this.
Is there more on this that you'd recommend reading? This approach doesn't sound more 'hands-on' than VPW, is it?
I'd be interested to hear what longinvest thinks about your "time value of money calculator" suggestion.
There isn't any reading on this that I'm aware of. It is more hands-on than VPW because it requires you to make return assumptions that are already embedded in the VPW tables. But I'd rather be in the driver's seat with all of those variables. My life expectancy could change quickly, for instance, with a medical diagnosis, and I might want to amortize the remainder of my portfolio over 10 years instead of 20. Or valuations might fall, resulting in an increase in the expected returns of stocks, resulting in an increase in the withdrawals. This is a very flexible yet relatively simple approach.

I can tell you right now that he's going to prefer VPW over it because VPW is his thing.
“It's a dangerous business, Frodo, going out your door. You step onto the road, and if you don't keep your feet, there's no knowing where you might be swept off to.” J.R.R. Tolkien,The Lord of the Rings

User avatar
Lieutenant.Columbo
Posts: 1178
Joined: Sat Sep 05, 2015 9:20 pm
Location: Cicely AK

Re: Variable Percentage Withdrawal (VPW)

Post by Lieutenant.Columbo » Mon Nov 12, 2018 11:53 am

willthrill81 wrote:
Mon Nov 12, 2018 11:24 am
I probably have a 33% probability of surviving to age 100.
I admit I still do not know what that means. Not trying to be silly or difficult. Does this mean that 33% of people "like you" (relatives, ethnicity, age, gender, geographic location, background, etc) have lived to 100?
Lieutenant.Columbo wrote:
Mon Nov 12, 2018 11:02 am
willthrill81 wrote:
Mon Nov 05, 2018 10:33 am
For those wanting a more 'hands on' approach, a simple time value of money calculator can be used instead. You simply input the dollar amount of your portfolio that you wish to amortize (it could be all of your portfolio or some portion of it) over your remaining lifetime, the assumed real rate of return, and the period of time you want to amortize payments over. For instance, someone with $1 million to amortize over 30 years at a 3% real rate of return would withdraw $51k in the current year. This would be recalculated each subsequent year as the remaining amount of capital, the assumed rate of return, and the remaining years of withdrawals change. It only takes a few minutes at most once per year to do this.
Is there more on this that you'd recommend reading? This approach doesn't sound more 'hands-on' than VPW, is it?
I'd be interested to hear what longinvest thinks about your "time value of money calculator" suggestion.

There isn't any reading on this that I'm aware of. It is more hands-on than VPW because it requires you to make return assumptions that are already embedded in the VPW tables. But I'd rather be in the driver's seat with all of those variables. My life expectancy could change quickly, for instance, with a medical diagnosis, and I might want to amortize the remainder of my portfolio over 10 years instead of 20. Or valuations might fall, resulting in an increase in the expected returns of stocks, resulting in an increase in the withdrawals. This is a very flexible yet relatively simple approach.

I can tell you right now that he's going to prefer VPW over it because VPW is his thing.
I understand longinvest will prefer VPW, but I bet he'll have a reasonable reason. Meaning VPW is "his thing" because VPW makes sense to him. And it is not that VPW makes sense to him because it is his thing*.
*I bet you Longinvest never thought Lieutenant.Columbo would stand up for him. :wink:
Lt. Columbo: Well, what do you know. Here I am talking with some of the smartest people in the world, and I didn't even notice!

User avatar
siamond
Posts: 4771
Joined: Mon May 28, 2012 5:50 am

Re: Variable Percentage Withdrawal (VPW)

Post by siamond » Mon Nov 12, 2018 12:11 pm

willthrill81 wrote:
Mon Nov 05, 2018 10:33 am
For those wanting a more 'hands on' approach, a simple time value of money calculator can be used instead. You simply input the dollar amount of your portfolio that you wish to amortize (it could be all of your portfolio or some portion of it) over your remaining lifetime, the assumed real rate of return, and the period of time you want to amortize payments over. For instance, someone with $1 million to amortize over 30 years at a 3% real rate of return would withdraw $51k in the current year. This would be recalculated each subsequent year as the remaining amount of capital, the assumed rate of return, and the remaining years of withdrawals change. It only takes a few minutes at most once per year to do this.
This is quite close to what I've been doing for a few years now (as an early retiree with a fixed AA), and yeah, I much prefer such 'hands on' recalculated approach - although admittedly, this is for people who are a bit more numbers-inclined than the average retiree.

The primary difference is that I include in the time value of money math the other expected income flows (e.g. SSA, Pension, possible SPIA, inheritance, possible house downsizing, etc), assembling a small table of future retirement years to do so. This can also model some one-time extraordinary expenses if needs be. Quite trivial to do for someone familiar with spreadsheets and very flexible to accommodate individual situations and also run scenario analysis (ah, those SPIAs!).

User avatar
Topic Author
willthrill81
Posts: 10324
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: Variable Percentage Withdrawal (VPW)

Post by willthrill81 » Mon Nov 12, 2018 12:25 pm

Lieutenant.Columbo wrote:
Mon Nov 12, 2018 11:53 am
willthrill81 wrote:
Mon Nov 12, 2018 11:24 am
I probably have a 33% probability of surviving to age 100.
I admit I still do not know what that means. Not trying to be silly or difficult. Does this mean that 33% of people "like you" (relatives, ethnicity, age, gender, geographic location, background, etc) have lived to 100?
Basically yes. In the same sense that a fair die has a 1/3 chance of coming up '1' or '2', I have a 33% chance of surviving to age 100. In a large group of people with similar characteristics and familial history as myself, 1/3 are expected to survive to age 100.
“It's a dangerous business, Frodo, going out your door. You step onto the road, and if you don't keep your feet, there's no knowing where you might be swept off to.” J.R.R. Tolkien,The Lord of the Rings

User avatar
Lieutenant.Columbo
Posts: 1178
Joined: Sat Sep 05, 2015 9:20 pm
Location: Cicely AK

Re: Variable Percentage Withdrawal (VPW)

Post by Lieutenant.Columbo » Mon Nov 12, 2018 12:44 pm

willthrill81 wrote:
Mon Nov 12, 2018 12:25 pm
Lieutenant.Columbo wrote:
Mon Nov 12, 2018 11:53 am
willthrill81 wrote:
Mon Nov 12, 2018 11:24 am
I probably have a 33% probability of surviving to age 100.
I admit I still do not know what that means. Not trying to be silly or difficult. Does this mean that 33% of people "like you" (relatives, ethnicity, age, gender, geographic location, background, etc) have lived to 100?
Basically yes. In the same sense that a fair die has a 1/3 chance of coming up '1' or '2', I have a 33% chance of surviving to age 100. In a large group of people with similar characteristics and familial history as myself, 1/3 are expected to survive to age 100.
thank you
Lt. Columbo: Well, what do you know. Here I am talking with some of the smartest people in the world, and I didn't even notice!

AlohaJoe
Posts: 4511
Joined: Mon Nov 26, 2007 2:00 pm
Location: Saigon, Vietnam

Re: Variable Percentage Withdrawal (VPW)

Post by AlohaJoe » Mon Nov 12, 2018 7:59 pm

willthrill81 wrote:
Mon Nov 12, 2018 11:24 am
There isn't any reading on this that I'm aware of.
http://howmuchcaniaffordtospendinretire ... gspot.com/

Here's the article published in the Journal of Personal Finance: http://howmuchcaniaffordtospendinretire ... 062014.pdf

Here's a few published in Advisor Perspectives (an online magazine for financial advisors):
http://www.advisorperspectives.com/arti ... er-advisor
http://www.advisorperspectives.com/arti ... -ever-need
https://www.advisorperspectives.com/art ... s-spending

Here's the spreadsheet: http://howmuchcaniaffordtospendinretire ... heets.html

The earliest publication I know of using this kind of approach is Gordon Pye in his 18-year old article "Sustainable Investment Withdrawals" in the Journal of Financial Planning. He ended up writing a number of articles -- and a book -- about using PMT for retirement withdrawals around 1999-2001. For whatever reason it never really caught on. Maybe it was too much math for the retirement planners at the time? The second wave came in 2006 when Stout & Mitchell published a series of papers, starting with "Dynamic Retirement Withdrawal Planning" in the Financial Services Review.

Ken Steiner (the website at the beginning) began his website in March 2010 which, along with longinvest's work since 2013, seems to be the third wave and (slowly) making progress against the constant drumbeat of fixed dollar withdrawals.

User avatar
Topic Author
willthrill81
Posts: 10324
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: Variable Percentage Withdrawal (VPW)

Post by willthrill81 » Mon Nov 12, 2018 8:16 pm

AlohaJoe wrote:
Mon Nov 12, 2018 7:59 pm
willthrill81 wrote:
Mon Nov 12, 2018 11:24 am
There isn't any reading on this that I'm aware of.
http://howmuchcaniaffordtospendinretire ... gspot.com/

Here's the article published in the Journal of Personal Finance: http://howmuchcaniaffordtospendinretire ... 062014.pdf

Here's a few published in Advisor Perspectives (an online magazine for financial advisors):
http://www.advisorperspectives.com/arti ... er-advisor
http://www.advisorperspectives.com/arti ... -ever-need
https://www.advisorperspectives.com/art ... s-spending

Here's the spreadsheet: http://howmuchcaniaffordtospendinretire ... heets.html

The earliest publication I know of using this kind of approach is Gordon Pye in his 18-year old article "Sustainable Investment Withdrawals" in the Journal of Financial Planning. He ended up writing a number of articles -- and a book -- about using PMT for retirement withdrawals around 1999-2001. For whatever reason it never really caught on. Maybe it was too much math for the retirement planners at the time? The second wave came in 2006 when Stout & Mitchell published a series of papers, starting with "Dynamic Retirement Withdrawal Planning" in the Financial Services Review.

Ken Steiner (the website at the beginning) began his website in March 2010 which, along with longinvest's work since 2013, seems to be the third wave and (slowly) making progress against the constant drumbeat of fixed dollar withdrawals.
Great! Thank you.

I really like their smoothing algorithm (i.e. no more than a +/- 10% change in annual withdrawals). I would expect some amount of smoothing would occur as well if you were using something like 1/CAPE for expected stock returns because, everything else held equal, a reduction in price (and your stock holdings) would be accompanied by an increase in the expected returns going forward.
“It's a dangerous business, Frodo, going out your door. You step onto the road, and if you don't keep your feet, there's no knowing where you might be swept off to.” J.R.R. Tolkien,The Lord of the Rings

User avatar
LadyGeek
Site Admin
Posts: 54285
Joined: Sat Dec 20, 2008 5:34 pm
Location: Philadelphia
Contact:

Re: Time Value of Money vs. Variable Percentage Withdrawal (VPW)

Post by LadyGeek » Mon Nov 12, 2018 9:14 pm

willthrill81 - I split your discussion into a new thread, as it was focused on the Time Value of Money as applied to your situation. The main discussion was getting derailed.
Wiki To some, the glass is half full. To others, the glass is half empty. To an engineer, it's twice the size it needs to be.

User avatar
siamond
Posts: 4771
Joined: Mon May 28, 2012 5:50 am

Re: Variable Percentage Withdrawal (VPW)

Post by siamond » Mon Nov 12, 2018 9:48 pm

willthrill81 wrote:
Mon Nov 12, 2018 11:24 am
There isn't any reading on this that I'm aware of.
When I was researching the whole topic of retirement withdrawal methods around 2012/13, I remember stumbling upon various references to financial advisors having used actuarial methods (i.e. PMT-based approaches) for quite a few years. I mean, this is a pretty natural idea for anybody having basic knowledge of financing concepts. Solid retirement research is still a pretty new field, and a decade or two ago, very few people were bothering to publish on the topic, besides the endless parroting of the Trinity study. I suspect this concept is an age-old idea, it's just that it wasn't publicized. Unfortunately, I am not as well organized as AlohaJoe and I didn't keep pointers about said references.

AlohaJoe
Posts: 4511
Joined: Mon Nov 26, 2007 2:00 pm
Location: Saigon, Vietnam

Re: Variable Percentage Withdrawal (VPW)

Post by AlohaJoe » Mon Nov 12, 2018 10:04 pm

siamond wrote:
Mon Nov 12, 2018 9:48 pm
Unfortunately, I am not as well organized as AlohaJoe and I didn't keep pointers about said references.
You just need to install https://www.mendeley.com :D

User avatar
siamond
Posts: 4771
Joined: Mon May 28, 2012 5:50 am

Re: Variable Percentage Withdrawal (VPW)

Post by siamond » Mon Nov 12, 2018 10:06 pm

willthrill81 wrote:
Mon Nov 12, 2018 8:16 pm
I would expect some amount of smoothing would occur as well if you were using something like 1/CAPE for expected stock returns because, everything else held equal, a reduction in price (and your stock holdings) would be accompanied by an increase in the expected returns going forward.
Yes, indeed, and I do exactly that, among a few other folks I know of. You can find some of my analysis in this long and somewhat contentious thread. This more recent thread has also some valuable discussion on the matter.

Personally, I ended up simplifying and settling on only using the dynamic expected returns model (ala 1/CAPE), as you're right, this does provide decent smoothing on its own and it's downright simple if you're willing to be 'hands on' as you put it. This might not be enough for some situations though (e.g. oil crisis) which are not valuation crises. If this occurs, I'll probably come back to one of the short-term smoothing algorithms. Let's see what happens during the next crisis...

Instead of re-hashing a smoothing discussion, it might be more productive to focus this thread on its title though, and possible use of time value of money concepts to model a full retirement plan. This was mentioned in a few random posts in the past, but never in a focused enough manner.

User avatar
Topic Author
willthrill81
Posts: 10324
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: Time Value of Money vs. Variable Percentage Withdrawal (VPW)

Post by willthrill81 » Mon Nov 12, 2018 10:16 pm

So far, the only drawback of this annuitization approach to VPW that I can see is that it requires the user to input a couple of unknowns, the number of years to amortize over and the expected rate of return, to compute the current year's withdrawals. For most Bogleheads, this seems like a very minimal drawback. Can anyone identify any other disadvantages compared to VPW?

In truth, VPW is just a specific application of this annuitization approach (i.e. number of years to amortize is equal to 100 minus your current age, and return assumptions are historic).
“It's a dangerous business, Frodo, going out your door. You step onto the road, and if you don't keep your feet, there's no knowing where you might be swept off to.” J.R.R. Tolkien,The Lord of the Rings

User avatar
BlueEars
Posts: 3700
Joined: Sat Mar 10, 2007 12:15 am
Location: West Coast

Re: Time Value of Money vs. Variable Percentage Withdrawal (VPW)

Post by BlueEars » Mon Nov 12, 2018 10:26 pm

willthrill81 wrote:
Mon Nov 12, 2018 10:16 pm
So far, the only drawback of this annuitization approach to VPW that I can see is that it requires the user to input a couple of unknowns, the number of years to amortize over and the expected rate of return, to compute the current year's withdrawals. For most Bogleheads, this seems like a very minimal drawback. Can anyone identify any other disadvantages compared to VPW?

In truth, VPW is just a specific application of this annuitization approach (i.e. number of years to amortize is equal to 100 minus your current age, and return assumptions are historic).
I'm somewhat confused by what I've read in some of the responses above. I don't think VPW requires any use of an SPIA. Maybe Longinvest mentioned an SPIA as a possible path?

VPW uses "Start Age" and "Last Withdrawal Age" so if one assumes he will get to 100 he can use a "Last Withdrawal Age" (LWA) of maybe 110 should he want to leave an estate or want to keep the portfolio from going below a certain amount during a worst case simulation. For instance, Joe has $2M and does not want to go below $1M for the retirement start date of 1966. He sets LWA = 110 and sees that his 60/40 AA stays above an inflation adjusted $1M for all years after 1966. Joe hopes that the 1966 scenario is about as bad as it could get for him in the future.

User avatar
Topic Author
willthrill81
Posts: 10324
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: Time Value of Money vs. Variable Percentage Withdrawal (VPW)

Post by willthrill81 » Mon Nov 12, 2018 10:30 pm

BlueEars wrote:
Mon Nov 12, 2018 10:26 pm
willthrill81 wrote:
Mon Nov 12, 2018 10:16 pm
So far, the only drawback of this annuitization approach to VPW that I can see is that it requires the user to input a couple of unknowns, the number of years to amortize over and the expected rate of return, to compute the current year's withdrawals. For most Bogleheads, this seems like a very minimal drawback. Can anyone identify any other disadvantages compared to VPW?

In truth, VPW is just a specific application of this annuitization approach (i.e. number of years to amortize is equal to 100 minus your current age, and return assumptions are historic).
I'm somewhat confused by what I've read in some of the responses above. I don't think VPW requires any use of an SPIA. Maybe Longinvest mentioned an SPIA as a possible path?
From the Wiki on VPW:
Around age 80, if you're still alive, it is important to consider using part (but not all) of your remaining portfolio to buy an inflation-indexed Single Premium Immediate Annuity (SPIA), so that total non-portfolio income (including Social Security, pension, and other lifelong income) is sufficient to live comfortably, independently of future portfolio withdrawals. This aims to reduce the financial risks associated with living past age 100.
BlueEars wrote:
Mon Nov 12, 2018 10:26 pm
VPW uses "Start Age" and "Last Withdrawal Age" so if one assumes he will get to 100 he can use a "Last Withdrawal Age" (LWA) of maybe 110 should he want to leave an estate or want to keep the portfolio from going below a certain amount during a worst case simulation. For instance, Joe has $2M and does not want to go below $1M for the retirement start date of 1966. He sets LWA = 110 and sees that his 60/40 AA stays above an inflation adjusted $1M for all years after 1966. Joe hopes that the 1966 scenario is about as bad as it could get for him in the future.
Something like that is a reasonable alternative to a SPIA for those interested in protecting against living beyond age 100. By definition, however, it means that you're withdrawing less today to protect against what may be a remote possibility for some people (but not remote at all for me and many others). It is estimated that there are fewer than 600 people alive today who are 110 or older. Only 1 in 1,000 of those who survive to age 100 go on to reach 110. That being said, I had a great-great-great grandfather live to 111, and nearly all of my father's aunts and uncles who didn't die in accidents survived to 85 or beyond. So longevity runs in our family.
“It's a dangerous business, Frodo, going out your door. You step onto the road, and if you don't keep your feet, there's no knowing where you might be swept off to.” J.R.R. Tolkien,The Lord of the Rings

2015
Posts: 2906
Joined: Mon Feb 10, 2014 2:32 pm

Re: Time Value of Money vs. Variable Percentage Withdrawal (VPW)

Post by 2015 » Mon Nov 12, 2018 10:42 pm

willthrill81 wrote:
Mon Nov 12, 2018 10:16 pm
So far, the only drawback of this annuitization approach to VPW that I can see is that it requires the user to input a couple of unknowns, the number of years to amortize over and the expected rate of return, to compute the current year's withdrawals. For most Bogleheads, this seems like a very minimal drawback. Can anyone identify any other disadvantages compared to VPW?

In truth, VPW is just a specific application of this annuitization approach (i.e. number of years to amortize is equal to 100 minus your current age, and return assumptions are historic).
I don't get it. How can inputting an expected rate of return in a complex adaptive system be anything but a major drawback? Get that wrong and the whole proposition flies out the window. To me, it's another example of the overconfident flying blind in a dynamic system (investing), a most perilous combination.

I also agree that annuitization is not a required aspect of VPW. It is an option, but not a requirement when using VPW. I will almost definitely not be using annuitization along with VPW. In fact, as my discretionary and non-discretionary liabilities are currently matched, I may not use VPW at all. OTOH, if my lap dance addiction gets worse and I feel the need to increase my lap dances from 10 to 20 times per day I may dip into the risk portfolio by means of VPW.

User avatar
Topic Author
willthrill81
Posts: 10324
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: Time Value of Money vs. Variable Percentage Withdrawal (VPW)

Post by willthrill81 » Mon Nov 12, 2018 10:50 pm

2015 wrote:
Mon Nov 12, 2018 10:42 pm
willthrill81 wrote:
Mon Nov 12, 2018 10:16 pm
So far, the only drawback of this annuitization approach to VPW that I can see is that it requires the user to input a couple of unknowns, the number of years to amortize over and the expected rate of return, to compute the current year's withdrawals. For most Bogleheads, this seems like a very minimal drawback. Can anyone identify any other disadvantages compared to VPW?

In truth, VPW is just a specific application of this annuitization approach (i.e. number of years to amortize is equal to 100 minus your current age, and return assumptions are historic).
I don't get it. How can inputting an expected rate of return in a complex adaptive system be anything but a major drawback? Get that wrong and the whole proposition flies out the window. To me, it's another example of the overconfident flying blind in a dynamic system (investing), a most perilous combination.
Yes and no. It's true that if your actual returns are different from your expected returns, your portfolio and withdrawals will be different as well. But this method is somewhat self-correcting in this regard. If my expected returns don't materialize, then my portfolio (PV in a time-value of money formula) will go down, and my withdrawals (PMT) will go down as well, even if I don't change the expected returns. If my actual returns are lower than my expected returns, it just means that I've front-loaded my withdrawals to some extent, which is probably a good move for many retirees anyway and something that I'm thinking about purposefully doing for two reasons: (1) I want to enjoy as much of my portfolio as I prudently can when I'm younger, healthier, and alive and (2) retirees' inflation-adjusted spending tends to decline throughout their retirement naturally.
2015 wrote:
Mon Nov 12, 2018 10:42 pm
I also agree that annuitization is not a required aspect of VPW. It is an option, but not a requirement when using VPW. I will almost definitely not be using annuitization along with VPW. In fact, as my discretionary and non-discretionary liabilities are currently matched, I may not use VPW at all. OTOH, if my lap dance addiction gets worse and I feel the need to increase my lap dances from 10 to 20 times per day I may dip into the risk portfolio by means of VPW.
The 'official' VPW approach does indeed use annuitization. By following it, your portfolio would be exhausted by age 100.
“It's a dangerous business, Frodo, going out your door. You step onto the road, and if you don't keep your feet, there's no knowing where you might be swept off to.” J.R.R. Tolkien,The Lord of the Rings

2015
Posts: 2906
Joined: Mon Feb 10, 2014 2:32 pm

Re: Time Value of Money vs. Variable Percentage Withdrawal (VPW)

Post by 2015 » Mon Nov 12, 2018 11:07 pm

willthrill81 wrote:
Mon Nov 12, 2018 10:50 pm
2015 wrote:
Mon Nov 12, 2018 10:42 pm
willthrill81 wrote:
Mon Nov 12, 2018 10:16 pm
I don't get it. How can inputting an expected rate of return in a complex adaptive system be anything but a major drawback? Get that wrong and the whole proposition flies out the window. To me, it's another example of the overconfident flying blind in a dynamic system (investing), a most perilous combination.
Yes and no. It's true that if your actual returns are different from your expected returns, your portfolio and withdrawals will be different as well. But this method is somewhat self-correcting in this regard. If my expected returns don't materialize, then my portfolio (PV in a time-value of money formula) will go down, and my withdrawals (PMT) will go down as well, even if I don't change the expected returns. If my actual returns are lower than my expected returns, it just means that I've front-loaded my withdrawals to some extent, which is probably a good move for many retirees anyway and something that I'm thinking about purposefully doing for two reasons: (1) I want to enjoy as much of my portfolio as I prudently can when I'm younger, healthier, and alive and (2) retirees' inflation-adjusted spending tends to decline throughout their retirement naturally.

There is much debate about "front-loading" one's withdrawals during the early years of retirement, something I've personally never been comfortable with. I am familiar with the "smile" pattern of spending in retirement, but it's not anything I want to bet on. Perhaps this can be viewed as a personal preference based on one's risk tolerance (my risk tolerance is quite low!). I can say this, from experience, not from theory: having my liabilities matched has brought me great peace of mind while allowing me to focus on things beyond my portfolio.

2015 wrote:
Mon Nov 12, 2018 10:42 pm
I also agree that annuitization is not a required aspect of VPW. It is an option, but not a requirement when using VPW. I will almost definitely not be using annuitization along with VPW. In fact, as my discretionary and non-discretionary liabilities are currently matched, I may not use VPW at all. OTOH, if my lap dance addiction gets worse and I feel the need to increase my lap dances from 10 to 20 times per day I may dip into the risk portfolio by means of VPW.
The 'official' VPW approach does indeed use annuitization. By following it, your portfolio would be exhausted by age 100.
Please link or quote where longinvest has explicitly stated annuitization is a required aspect of VPW, because I've not seen that in any of his explanations. AFAICT, whether one opts for annuitization when using VPW would probably depend on the following factors: (a) the degree to which one's liabilities are matched outside the risk portfolio; (b) the size of one's RP (on which VPW calculations are based); and (c) how religiously one intends to withdraw from the RP.

User avatar
TimeRunner
Posts: 1519
Joined: Sat Dec 29, 2012 9:23 pm

Re: Time Value of Money vs. Variable Percentage Withdrawal (VPW)

Post by TimeRunner » Mon Nov 12, 2018 11:12 pm

The mother of the couple who lives next door, who lives with them, turned 99 last month. She's very spry, gets up and does 1/2 hour of calisthenics every morning, takes a five or six mile walk every day, and lives in an upstairs bedroom so she climbs a flight of stairs multiple times a day. It does seem that if one takes care of oneself, is genetically blessed, has a quality support system, and has access to quality healthcare, that one can live a lot longer than previous generations. It's wise to plan for that if you fit those specs.
Man assumes more intelligence than dolphins because he achieves so much, while dolphins just have a good time. Dolphins believe they're more intelligent than man - for precisely the same reason. (HGG)

User avatar
Topic Author
willthrill81
Posts: 10324
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: Time Value of Money vs. Variable Percentage Withdrawal (VPW)

Post by willthrill81 » Mon Nov 12, 2018 11:28 pm

2015 wrote:
Mon Nov 12, 2018 11:07 pm
Please link or quote where longinvest has explicitly stated annuitization is a required aspect of VPW, because I've not seen that in any of his explanations.
It's right from the Wiki. VPW completely exhausts the portfolio by age 100.
“It's a dangerous business, Frodo, going out your door. You step onto the road, and if you don't keep your feet, there's no knowing where you might be swept off to.” J.R.R. Tolkien,The Lord of the Rings

longinvest
Posts: 3505
Joined: Sat Aug 11, 2012 8:44 am

Re: Time Value of Money vs. Variable Percentage Withdrawal (VPW)

Post by longinvest » Mon Nov 12, 2018 11:40 pm

willthrill81 wrote:
Mon Nov 05, 2018 10:33 am
For those wanting a more 'hands on' approach, a simple time value of money calculator can be used instead. You simply input the dollar amount of your portfolio that you wish to amortize (it could be all of your portfolio or some portion of it) over your remaining lifetime, the assumed real rate of return, and the period of time you want to amortize payments over. For instance, someone with $1 million to amortize over 30 years at a 3% real rate of return would withdraw $51k in the current year. This would be recalculated each subsequent year as the remaining amount of capital, the assumed rate of return, and the remaining years of withdrawals change. It only takes a few minutes at most once per year to do this.
I've developed the VPW-adv spreadsheet specifically for advanced users who enjoy spreadsheets (and possibly time value of money calculations). It combines various calculations to calculate the current year's withdrawal. It takes into account current and future pensions, with and without cost of living adjustments (COLA), and the desire for a residual portfolio primarily for lifelong liquidity and possibly for bequest.

I've kept that spreadsheet separate from the main VPW topic, because it fails one the main objectives I have for VPW: simplicity. When I die, I don't know how long my spreadsheet will continue to work in the hands of my wife. On the other hand, I can easily transcribe the appropriate column of the VPW table on a paper document where I explain how to use it. This considerably lowers the dependency on technology. (Writing this post reminds me that I'll have to actually do it!)

There are some fundamental decisions that have gone into VPW's design that distinguishes it from a general PMT formula.

A) For one thing, VPW's internal rate isn't based on a future return prediction. It's an asset-allocation-specific fixed rate that intends to be lower than high returns, and higher than low returns. It doesn't require precision but it has to be fixed; it's part of what gives VPW robustness and simplicity. The fixed rate will allow for higher spending in good years, and lower spending in bad years.

Many users of the general PMT method take a different approach. They try to estimate future returns and use their prediction as internal rate to calculate a withdrawal. Each year, they update their prediction to calculate the next withdrawal. A series of perfect prediction would obviously lead to perfect constant-dollar withdrawals. In other words, their objective is to recreate SWR, but with more safety (no premature depletion).

There are various reasons (other than simplicity) why I didn't adopt this prediction approach for VPW. Here are some important ones:
  • Regardless of all the ink about the accuracy (or not) of various medium or long-term future return metrics, I do not believe that any single metric can predict the future returns of a single stock investment. I see no way, as an example, for a ratio of current price related to past few years of earnings to have told me anything trustworthy about the future returns of Enron (which went bankrupt) or Apple (which became the largest worldwide capitalization). We simply don't know the future.
  • I don't believe that anyone who really knew how to predict future returns with sufficient accuracy to make a difference would tell me about it; he would be insane to do so instead of keeping the secret to himself and becoming a billionaire. By telling so many others about it that a random person like me would end up hearing about it, he would kill the advantage of the prediction. Think about it, if everybody knew years in advance the exact date at which Enron was going bankrupt and that Apple would soar, nobody would have put any money into Enron, or they would have sold their shares, dropping its price to the point where the remaining upcoming dividends, until bankruptcy, would return as much as Apple ahead. Of course, at the same time, people would fight to buy Apple shares, increasing their price and reducing their future returns appropriately so that they're more or less equal to Enron's. Markets are adaptive, no static.
  • Lastly, even if we were able to perfectly predict future returns, using this perfect prediction would lead to bad portfolio management with PMT-based withdrawals: "sell more shares when low", and "sell less shares when high". In other words, it would result into a reverse dollar cost averaging, a damaging thing for a portfolio.
A good time to withdraw more money from a portfolio is when shares are expensive. A good time to withdraw less is when their share values are down. VPW does just that, naturally, without needing to predict future returns as it's a purely reactive system.

B) VPW also targets a specific last withdrawal age of 99. This age isn't random and it isn't the highest age a human has reached, nor the median age (e.g. life expectancy). I've explained in the VPW thread why I've kept this last withdrawal age for the VPW table. It has to do with the objective of balancing liquidity and getting a low-enough cost to insure against the financial risk of long life by buying an inflation-indexed Single-Premium Immediate Annuity (SPIA).

The problem, with inflation-indexed SPIAs, is that they're expensive, especially in young age. They have to be expensive as they have to be built (internally) by insurance companies using fixed income products. The insurance company can't simply dump the money into the stock market or a balanced portfolio and hope for the best. It has to make sure it will be able to deliver the payments! The longer one waits to buy one, the "cheaper" it becomes. Ideally, one would like to buy the SPIA very late, maybe at age 90 or 95 (if still alive, of course)... But, insurance companies stop selling them (at reasonable price) to people who survived too long; the pool of new annuitants becomes too small and the financial risk, for the insurance company, becomes too big.

VPW is set up so that the payout on a joint inflation-indexed annuity more or less competes with VPW table percentages at age 80 for a typical retirement portfolio with at least 50% in bonds. The payout for a single male will likely be much higher than the VPW percentage. In other words, VPW is setup to make partial annuitization as palatable as possible at age 80. The goal isn't to annuitize the entire residual portfolio, but to only annuitize part of the residual portfolio (if necessary!) to dampen the financial risk of living beyond age 100. Some people might not need this. The goal is for non-portfolio income to be sufficient, in itself, to live comfortably (but no more) until death. The residual portfolio, after annuitization, will provide liquidity (for unanticipated money needs) and extra spending and giving.


I'm all for advanced users to go satisfy their need to develop their own custom retirement model using PMT formulas. But, I suggest keeping VPW's fundamental principles as part of their design. The VPW-adv spreadsheet could be an interesting starting point or maybe even be sufficient in itself... :wink:
Bogleheads investment philosophy | single-ETF balanced portfolio | VBAL

2015
Posts: 2906
Joined: Mon Feb 10, 2014 2:32 pm

Re: Time Value of Money vs. Variable Percentage Withdrawal (VPW)

Post by 2015 » Mon Nov 12, 2018 11:54 pm

willthrill81 wrote:
Mon Nov 12, 2018 11:28 pm
2015 wrote:
Mon Nov 12, 2018 11:07 pm
Please link or quote where longinvest has explicitly stated annuitization is a required aspect of VPW, because I've not seen that in any of his explanations.
It's right from the Wiki. VPW completely exhausts the portfolio by age 100.
Ok, I see where you got that, but the wiki states:
VPW is best used in conjunction with guaranteed base income from Social Security, pensions, and, if necessary, inflation-indexed Single Premium Immediate Annuity (SPIA).
Emphasis added

In my case, annuitization isn't needed because SS delayed until 70 coupled with other guaranteed base income meets non-discretionary as well discretionary needs. VPW doesn't have to exhaust the PF by 100. For those outliers who believe they'll live beyond 100 or for those wishing to leave a legacy the spreadsheet can be used. Remember, VPW is based on the risk portfolio which functions outside the lifelong guaranteed base income so the portfolio end date is flexible.

AlohaJoe
Posts: 4511
Joined: Mon Nov 26, 2007 2:00 pm
Location: Saigon, Vietnam

Re: Time Value of Money vs. Variable Percentage Withdrawal (VPW)

Post by AlohaJoe » Mon Nov 12, 2018 11:55 pm

willthrill81 wrote:
Mon Nov 12, 2018 10:16 pm
Can anyone identify any other disadvantages compared to VPW?
A few years ago I wrote two blog posts exploring the two main variables in this approach (length of time & the rate):

https://medium.com/@justusjp/flavors-of ... e09aed5be1
https://medium.com/@justusjp/flavors-of ... 486445a4d1

I think there aren't straightforward answers to either of them but mortality is slightly more tractable: you can go to one of the many longevity estimators/calculators, figure out the joint mortality for you & your spouse, and then decide whether you want to target the 80th percentile, the 90th percentile, the 99th percentile, etc. Do you recalculate it every year?

Picking the rate is harder because so much judgment is involved. You could do 1/CAPE, which has some potential benefits. But your portfolio is probably not 100% S&P 500 -- and the things that aren't the S&P 500 in your portfolio are there because you expect them to have different return characteristics than the S&P 500. So using 1/CAPE for them seems...weird. You could do individual estimates (e.g. REITs are estimated this, EM is estimated that, bonds are estimated at this, and so on) and the weight them according to your portfolio and add it all up.

But then there's the question of "front-loading" expenses. The emerging research seems pretty clear that -- at least for non-early, normal retirees -- you want to do that to some extent. But how much? And what does that means in terms of the rate parameter?

Anyway, all of that stuff applies to VPW as well, it is just that someone else picked some plausible numbers for everyone.

The weaknesses of a withdrawal strategy like this aren't really unique to it -- they are common to virtually all systematic withdrawal strategies.

How do you handle largish lump sums in the future? (Let's say you know/think that in year 10 of retirement you'll need an extra $120,000 to pay for something.) How do you handle changing spending profiles? (Realistically, one spouse will die sooner than the other and expenses will go down slightly because of that.) How do you handle cognitive decline? (What if you die and your spouse is left with a spreadsheet full of PMT this and expected return that? Or what if you just turn 87 and can't handle it anymore?) And so on.

Most retirement plans provide solutions for those in a way that is orthogonal to the systematic withdrawals per se. "Buy a SPIA when you get old enough". "Put that $120,000 into TIPS". Etc.

User avatar
Topic Author
willthrill81
Posts: 10324
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: Time Value of Money vs. Variable Percentage Withdrawal (VPW)

Post by willthrill81 » Tue Nov 13, 2018 12:08 am

AlohaJoe wrote:
Mon Nov 12, 2018 11:55 pm
willthrill81 wrote:
Mon Nov 12, 2018 10:16 pm
Can anyone identify any other disadvantages compared to VPW?
A few years ago I wrote two blog posts exploring the two main variables in this approach (length of time & the rate):

https://medium.com/@justusjp/flavors-of ... e09aed5be1
https://medium.com/@justusjp/flavors-of ... 486445a4d1

I think there aren't straightforward answers to either of them but mortality is slightly more tractable: you can go to one of the many longevity estimators/calculators, figure out the joint mortality for you & your spouse, and then decide whether you want to target the 80th percentile, the 90th percentile, the 99th percentile, etc. Do you recalculate it every year?
Assuming that those calculators are accurate, you probably should recalcuate it every year. But that only takes a few minutes.
AlohaJoe wrote:
Mon Nov 12, 2018 11:55 pm
Picking the rate is harder because so much judgment is involved. You could do 1/CAPE, which has some potential benefits. But your portfolio is probably not 100% S&P 500 -- and the things that aren't the S&P 500 in your portfolio are there because you expect them to have different return characteristics than the S&P 500. So using 1/CAPE for them seems...weird. You could do individual estimates (e.g. REITs are estimated this, EM is estimated that, bonds are estimated at this, and so on) and the weight them according to your portfolio and add it all up.
I suspect that most Bogleheads are using broad market indices like TSM and ex-U.S., and CAPE and other valuation metrics are fairly easy to find for those. Presumably, tilting toward something like SCV or EM would be for the purpose of attempting to outperform these indices, so still using 1/CAPE seems adequately conservative to me.
AlohaJoe wrote:
Mon Nov 12, 2018 11:55 pm
But then there's the question of "front-loading" expenses. The emerging research seems pretty clear that -- at least for non-early, normal retirees -- you want to do that to some extent. But how much? And what does that means in terms of the rate parameter?
Based on the data indicating that real spending declines by 1% annually, I'd say that using a rate of return that's 1% higher than your expectation should front-load withdrawals adequately.
AlohaJoe wrote:
Mon Nov 12, 2018 11:55 pm
Anyway, all of that stuff applies to VPW as well, it is just that someone else picked some plausible numbers for everyone.
I think this is a very key point. VPW is indeed making return assumptions. Failing to meet those assumptions does not deplete the portfolio prematurely, but that's not the case with this general annuitization approach either. I may want to use a very different set of assumptions than someone else, and it's easy to input these into a time value formula, easier than a spreadsheet to me.
AlohaJoe wrote:
Mon Nov 12, 2018 11:55 pm
How do you handle largish lump sums in the future? (Let's say you know/think that in year 10 of retirement you'll need an extra $120,000 to pay for something.) How do you handle changing spending profiles? (Realistically, one spouse will die sooner than the other and expenses will go down slightly because of that.)
I probably wouldn't want to completely annuitize my portfolio, but this is easily managed by just changing the future value part of the PMT formula to some fixed dollar amount you want to hang on to for some reason (e.g. long-term care, 'guaranteed' legacy).
AlohaJoe wrote:
Mon Nov 12, 2018 11:55 pm
How do you handle cognitive decline? (What if you die and your spouse is left with a spreadsheet full of PMT this and expected return that? Or what if you just turn 87 and can't handle it anymore?) And so on.

Most retirement plans provide solutions for those in a way that is orthogonal to the systematic withdrawals per se. "Buy a SPIA when you get old enough". "Put that $120,000 into TIPS". Etc.
Cognitive decline is a big factor to keep in mind. There are no easy solutions to the problem either IMHO. A SPIA with a COLA would be more appealing in that situation to me than otherwise, but it's certainly not a one-size-fits-all approach either.
“It's a dangerous business, Frodo, going out your door. You step onto the road, and if you don't keep your feet, there's no knowing where you might be swept off to.” J.R.R. Tolkien,The Lord of the Rings

longinvest
Posts: 3505
Joined: Sat Aug 11, 2012 8:44 am

Re: Time Value of Money vs. Variable Percentage Withdrawal (VPW)

Post by longinvest » Tue Nov 13, 2018 12:13 am

AlohaJoe wrote:
Mon Nov 12, 2018 11:55 pm
How do you handle largish lump sums in the future? (Let's say you know/think that in year 10 of retirement you'll need an extra $120,000 to pay for something.) How do you handle changing spending profiles? (Realistically, one spouse will die sooner than the other and expenses will go down slightly because of that.) How do you handle cognitive decline? (What if you die and your spouse is left with a spreadsheet full of PMT this and expected return that? Or what if you just turn 87 and can't handle it anymore?) And so on.
As I wrote on the VPW thread, the VPW percentage is a guide to determine a reasonable withdrawal:
longinvest wrote:
Fri Jul 20, 2018 8:43 pm
I insist: the [VPW] table isn't meant to be used with ultimate precision. If the VPW table indicates 4.8%, taking 5% won't lead to premature portfolio depletion, and taking 4.5% won't lead to dying with a gigantic pile of unspent money. The percentages are the result of a calculation based on a wild-ass guess; they inherit the "wild-ass guess" property. So, they should be used as a guide to indicate approximately how much of the portfolio should be withdrawn today to pay taxes and fund the upcoming year's expenses given the current portfolio balance, the target asset allocation for the upcoming year, and the current age of the retiree.

Whether the specific percentage in the VPW table is 4.8% or 4.9% isn't important. These are rounded values resulting from a calculation based on a wild-ass guess. What's important is to understand that it wouldn't be sustainable to withdraw 7% or 8% instead of 4.8%, and keep over-withdrawing like that regularly. I've kept one decimal of precision (after being rightfully chastised by forum member Rodc, early in this thread, for the false precision of using the default two decimals of Microsoft Excel/Libreoffice Calc) so that annual percentage adjustments remain small. The difference between 4% and 5% of a portfolio is big; the difference between 4.4% and 4.5% is much smaller and leads to a more reasonable annual adjustment.
In other words, one could easily withdraw more from the portfolio for one year, as long as the additional amount isn't too big relative to the portfolio's balance. VPW is adaptive. The next year, the VPW withdrawal will be calculated on the residual portfolio at that point.

Life isn't as tidy as a spreadsheet. VPW will adapt to the retiree's unanticipated life events. Don't get me wrong: VPW isn't magical; it won't create money that the portfolio doesn't deliver. But, it will adapt future withdrawals to the state of the portfolio.

For spending more in earlier retirement years, here's a simple trick (from another thread):
longinvest wrote:
Wed Oct 24, 2018 8:04 pm
As for spending more in one's 60s and 70s, I would do the following. To get, let's say, $15,000 more per year from 62 to 78, I would put ((78 - 61) X $15,000) = $255,000 into a high-interest savings account, CDs, or a short-term bond fund and withdraw 1/17 of the balance at 62, 1/16 of the balance at 63, 1/15 of the balance at 64, and so on.
Bogleheads investment philosophy | single-ETF balanced portfolio | VBAL

User avatar
Topic Author
willthrill81
Posts: 10324
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: Time Value of Money vs. Variable Percentage Withdrawal (VPW)

Post by willthrill81 » Tue Nov 13, 2018 12:16 am

Since this thread was created to keep the VPW thread from getting derailed, let's please not turn this thread into another one focused on VPW.
“It's a dangerous business, Frodo, going out your door. You step onto the road, and if you don't keep your feet, there's no knowing where you might be swept off to.” J.R.R. Tolkien,The Lord of the Rings

BigoteGato
Posts: 46
Joined: Fri Mar 11, 2016 3:05 pm

Re: Variable Percentage Withdrawal (VPW)

Post by BigoteGato » Tue Nov 13, 2018 12:50 am

willthrill81 wrote:
Mon Nov 12, 2018 12:25 pm
Lieutenant.Columbo wrote:
Mon Nov 12, 2018 11:53 am
willthrill81 wrote:
Mon Nov 12, 2018 11:24 am
I probably have a 33% probability of surviving to age 100.
I admit I still do not know what that means. Not trying to be silly or difficult. Does this mean that 33% of people "like you" (relatives, ethnicity, age, gender, geographic location, background, etc) have lived to 100?
Basically yes. In the same sense that a fair die has a 1/3 chance of coming up '1' or '2', I have a 33% chance of surviving to age 100. In a large group of people with similar characteristics and familial history as myself, 1/3 are expected to survive to age 100.
I won't claim expertise, but there are several ways to interpret probabilities. The "33% of people like you" viewpoint seems like a frequentist interpretation. An alternative is the subjectivist view - how much would you be willing to bet that the event happens. Take a look at https://xkcd.com/1132/ for a frequentist vs. Bayesian example.
BigoteGato

User avatar
Topic Author
willthrill81
Posts: 10324
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: Variable Percentage Withdrawal (VPW)

Post by willthrill81 » Tue Nov 13, 2018 2:02 pm

siamond wrote:
Mon Nov 12, 2018 10:06 pm
Instead of re-hashing a smoothing discussion, it might be more productive to focus this thread on its title though, and possible use of time value of money concepts to model a full retirement plan. This was mentioned in a few random posts in the past, but never in a focused enough manner.
Then let's continue the discussion in that vein. Why could a TVM (i.e. annuitization) approach work or not work to model a full retirement plan? Given that it's a generalized form of the VPW approach, I see no reason why not beyond the aforementioned 'hands on' aspect, which is frankly a feature more than a drawback for me at least.

As you note, this approach could be easily used in conjunction with known future liabilities and income streams. Excel would make this fairly simple.
“It's a dangerous business, Frodo, going out your door. You step onto the road, and if you don't keep your feet, there's no knowing where you might be swept off to.” J.R.R. Tolkien,The Lord of the Rings

User avatar
siamond
Posts: 4771
Joined: Mon May 28, 2012 5:50 am

Re: Variable Percentage Withdrawal (VPW)

Post by siamond » Tue Nov 13, 2018 3:52 pm

willthrill81 wrote:
Tue Nov 13, 2018 2:02 pm
Then let's continue the discussion in that vein. Why could a TVM (i.e. annuitization) approach work or not work to model a full retirement plan? Given that it's a generalized form of the VPW approach, I see no reason why not beyond the aforementioned 'hands on' aspect, which is frankly a feature more than a drawback for me at least.

As you note, this approach could be easily used in conjunction with known future liabilities and income streams. Excel would make this fairly simple.
Yes, indeed, if one is minimally spreadsheet-inclined AND has a bit of a clue about Annuitization and Present Value concepts (as embodied by PMT and NPV() spreadsheet functions), it is fairly trivial to do. And extremely useful.

First step is to realize that annuitizing a given portfolio of value PV over a given # of years is as simple as PMT(rate-of-return, #years, -PV, 0, 1). Be careful to the 'minus' sign in front of the portfolio value. This formula is the core of VPW, which uses specific input values for rate-of-return and #periods. Any actuarial method I've read about does basically the same thing. Personally, I prefer to compute those input values in my own way and refresh every year, but that's just me.

Second step is to make a realistic scenario for future cash flows until you're truly expected to push the daisies. Social Security, Pensions, part-time wages for a while, inheritance, house downsizing, etc. Note that this can include income as well as special expenses that go beyond a regular budget (e.g. buy a vacation home, really big wedding for daddy's daughter, LTC expenses in the old years, etc). It seems easier to do it in inflation-adjusted terms, but one has to be careful with pensions, as many of them are NOT inflation-adjusted (or only partly).

Third step: the handy NPV() function can transform such future cash flows in a present value PV2. And then a simple PMT function on PV2, with the *same* rate of return used in NPV() and the same duration as the cash flows scenario, will provide a second annuitized value.

Fourth step: add the outcome of both PMT computations and you have your spending budget (pre-tax) for the coming year. Rinse and repeat the following year (optionally adjusting one assumption or another). Optionally, run some alternate scenarios to figure out the effect (e.g. delay SSA or not being a good example).

I personally follow that process, and I also ran simulations for a few fellow Bogleheads from our local chapter, and this proved very handy. It also showed me that nobody has the same sequence of cash flows, that the exception IS the rule, so you'd better keep the math very open ('hands on').

User avatar
siamond
Posts: 4771
Joined: Mon May 28, 2012 5:50 am

Re: Time Value of Money vs. Variable Percentage Withdrawal (VPW)

Post by siamond » Tue Nov 13, 2018 4:28 pm

Spreadsheet-inclined people will find the process very simple and very flexible. Now what are the drawbacks?

- it is a little more complicated to model an SPIA scenario, although not that much (it's essentially a big 'expense' the year you transform part of your portfolio in an SPIA annual income flow). And then one has to make assumptions about the inflation rider (or lack thereof).

- to be truly convinced this is a valid approach, one needs to have (or develop) a solid understanding of net present value (aka time value of money) concepts. I found it a tad challenging to explain it to some folks - even people with basic spreadsheet knowledge.

- it isn't directly obvious what rate should be used in the NPV() / PMT () steps. Personally, I settled for the historical numbers used in VPW (while I do something different for the regular PMT math related to my regular portfolio). Interestingly enough, the outcome of such math is NOT terribly sensitive to the rate you pick (4% or 5% or 6% will not make such a big difference), which is reassuring.

- either your spouse/partner or somebody trustworthy will need to have the ability to do such math if you were to suddenly disappear. Or maybe to follow a simplified version of it. I made sure to explain (and document) such process to both my wife and one of my sons. Good news is that by the time we reach 70, the cash flow part will become much simpler (LTC $$ still in the equation though).

- now for the biggie (aka double-whammy risk). Before the bulk of the cash flow incomes start trickling in (e.g. before SSA/Pensions kick in), the result of this process is to spend more shares from the regular savings portfolio than one would otherwise. Which is fine, it totally makes sense to use more of it before the SSA years and less of it later. BUT. A bad sequence of returns can really hurt, making it a double-whammy of 'selling low'. In some extreme cases, one could bankrupt their savings before reaching the SSA years, which would be rather unfortunate!

I don't have a good answer for the last point. If a deep crisis occurs in my late 60s, I will definitely cancel my SSA postponement plans, and this will help (numerically AND behaviorally). If a deep crisis occurs before that, I will probably get itchy, hence regulate myself and spend less from the 'NPV/PMT' present value than allowed by the corresponding math, but I don't have a formulaic plan for that. One could also use a TIPS ladder on some of the savings to create *some* stability (the default recommendation of VPW, clearly suboptimal, but also clearly reassuring). Finally, if one's savings are limited enough to be easily depleted, well, you probably can't play that game even if a huge pension/inheritance is waiting for you, and you need to work longer, sorry.

Personally, I ran some backtesting to see how robust my plan would have worked against some of the bad decades, and I wouldn't have settled on such approach without being comfortable enough with the outcome. Now admittedly, this is a bit more complicated to do than to run a couple of NPV/PMT formulas. There is room for opportunity for a good cFIREsim-like online tool here! :wink:
Last edited by siamond on Mon Nov 19, 2018 7:48 pm, edited 1 time in total.

AlohaJoe
Posts: 4511
Joined: Mon Nov 26, 2007 2:00 pm
Location: Saigon, Vietnam

Re: Variable Percentage Withdrawal (VPW)

Post by AlohaJoe » Tue Nov 13, 2018 6:47 pm

siamond wrote:
Tue Nov 13, 2018 3:52 pm
Second step is to make a realistic scenario for future cash flows until you're truly expected to push the daisies. Social Security, Pensions, part-time wages for a while, inheritance, house downsizing, etc. Note that this can include income as well as special expenses that go beyond a regular budget (e.g. buy a vacation home, really big wedding for daddy's daughter, LTC expenses in the old years, etc).

[...]

It also showed me that nobody has the same sequence of cash flows, that the exception IS the rule, so you'd better keep the math very open ('hands on').
I think this is the single biggest benefit of this approach over all other approaches. It is naive to pretend that all future expenses from the day of retirement until the day we die are going to a nice, smooth, straightforward annual amount. This is especially true for people who retire even a few years earlier than normal (thereby extending their retirement planning window to 4+ decades). We can look back at the previous 4 decades of our spending and realise how variable and unpredictable it can be, we shouldn't pretend that is all going to change just because we are retiring.

Does that mean we will come up with a perfect plan? That we will never have to adapt or adjust? Of course not. As the saying goes, plans are useless but planning is indispensable. And this approach strongly encourages to do some of that hard thinking. If we're going to be retired for 40 years...are we really never going to want to spend an extra $80,000 one year to remodel the kitchen? Even simple things like car replacement every 5-10 years becomes a bit less handwavy and more planned.

And, as you said, all of that also applies on the income/asset side as well.

I also understand that not everyone is as hands on as I am. But I think there is some value (and peace of mind) in being able to figure out for one's self if you'll still be "okay" if something dramatic changes 10 or 20 years into retirement. (What if your pension gets cut when your old employer goes bankrupt? Or what if, despite your assuming otherwise, your parents actually do leave you $150,000 inheritance.)

AlohaJoe
Posts: 4511
Joined: Mon Nov 26, 2007 2:00 pm
Location: Saigon, Vietnam

Re: Time Value of Money vs. Variable Percentage Withdrawal (VPW)

Post by AlohaJoe » Tue Nov 13, 2018 7:27 pm

siamond wrote:
Tue Nov 13, 2018 4:28 pm
I don't have a good answer for the last point. If a deep crisis occurs in my late 60s, I will definitely cancel my SSA postponement plans, and this will help (numerically AND behaviorally). If a deep crisis occurs before that, I will probably get itchy, hence regulate myself and spend less from the 'NPV/PMT' present value than allowed by the corresponding math, but I don't have a formulaic plan for that.
Here are two things people can do to make it a bit more formulaic, if they are so inclined.

#1: Many of us have this vague notion that annuities are a decent idea but work best for only a small number of situations: if you have plenty of money, annuitising is throwing it away; if you don't have enough money, buying an annuity just locks in failure. Jim Otar was the first I saw to talk about this. But where exactly does that "annuity boundary" lie for us in our circumstances? If our portfolio drops enough does it ever make sense to annuitise some or all of it?

We know that figuring out complicated and emotion-laden questions under pressure isn't ideal. Who wants to lie in bed at 2am trying to figure out if they should annuitise before the markets open tomorrow in case the crash-of-a-lifetime continues?

I've been thinking about this a bit more since reading this recent blog post: https://rivershedge.blogspot.com/2018/1 ... ndary.html but I haven't actually formalised anything yet. But I think there's scope for building something like that into a spreadsheet, at least as a flashing warning light.

#2: You can build in some simple math to do the spending cuts for you. These aren't fool-proof or anything but it at least takes out some of the random human guessing. "Should I spend $5,000 less than PMT says or $8,000 less?" Keep in mind that I'm (almost) as concerned about under-spending as over-spending. By having some math, my hope is avoid my own knee-jerk reactions. "OMG!!! The market is down 10% I need to cut spending 50% immediately!"

John Walton published a series of articles in Advisor Perspectives in 2016, including "A New Framework for Comparing Withdrawal Strategies". It took me a few readings to warm to his concept (he explained it in what I thought was a quite roundabout way) -- and it isn't a silver bullet or anything. He actually talks about PMT, which he calls "mortgage withdrawals", but he applies a "tilt" to the result of the PMT calculation.

The interesting thing about his tilt mechanism is that the same math can be used to tilt towards income stability or capital stability. The math is simply

Code: Select all

 PMT * ((current_capital / target_capital) ^ tilt) 
Current capital is easy; that's just your current portfolio value.

Target capital is whatever you goal is. You can get that number is one of two ways: you can do PV calculations or you just pick a nice, round nominal number.

The tilt is a number between -1 and +1. Positive numbers favor capital stability. Negative numbers favor income stability.

Here's how it might work in practice.

Say that PMT says you can withdraw $50,000 this year. Your portfolio is $1.2 million now but before the epic market crash it was $1.5 million and you'd prefer if your portfolio were back at that mark. 1.2/1.5 = 0.8.

If you pick a tilt of 0.3 then 0.8^0.3 = 0.93. Instead of $50,000 take 93% of $50,000.
If you pick a tilt of 0.5 then 0.8^0.5 = 0.89. Instead of $50,000 take 89% of $50,000.
If you pick a tilt of 1.0 then 0.8^1 = 0.8. Instead of $50,000 take 80% of $50,000.

I do this and picked a largish tilt, since I have a relatively large amount of discretionary spending and have the flexibility to vary my spending by tens of thousands of dollars a year if I want to/need to.

User avatar
Topic Author
willthrill81
Posts: 10324
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: Variable Percentage Withdrawal (VPW)

Post by willthrill81 » Tue Nov 13, 2018 7:52 pm

siamond wrote:
Tue Nov 13, 2018 3:52 pm
willthrill81 wrote:
Tue Nov 13, 2018 2:02 pm
Then let's continue the discussion in that vein. Why could a TVM (i.e. annuitization) approach work or not work to model a full retirement plan? Given that it's a generalized form of the VPW approach, I see no reason why not beyond the aforementioned 'hands on' aspect, which is frankly a feature more than a drawback for me at least.

As you note, this approach could be easily used in conjunction with known future liabilities and income streams. Excel would make this fairly simple.
Yes, indeed, if one is minimally spreadsheet-inclined AND has a bit of a clue about Annuitization and Present Value concepts (as embodied by PMT and NPV() spreadsheet functions), it is fairly trivial to do. And extremely useful.

First step is to realize that annuitizing a given portfolio of value PV over a given # of years is as simple as PMT(rate-of-return, #years, -PV, 0, 1). Be careful to the 'minus' sign in front of the portfolio value. This formula is the core of VPW, which uses specific input values for rate-of-return and #periods. Any actuarial method I've read about does basically the same thing. Personally, I prefer to compute those input values in my own way and refresh every year, but that's just me.

Second step is to make a realistic scenario for future cash flows until you're truly expected to push the daisies. Social Security, Pensions, part-time wages for a while, inheritance, house downsizing, etc. Note that this can include income as well as special expenses that go beyond a regular budget (e.g. buy a vacation home, really big wedding for daddy's daughter, LTC expenses in the old years, etc). It seems easier to do it in inflation-adjusted terms, but one has to be careful with pensions, as many of them are NOT inflation-adjusted (or only partly).

Third step: the handy NPV() function can transform such future cash flows in a present value PV2. And then a simple PMT function on PV2, with the *same* rate of return used in NPV() and the same duration as the cash flows scenario, will provide a second annuitized value.

Fourth step: add the outcome of both PMT computations and you have your spending budget (pre-tax) for the coming year. Rinse and repeat the following year (optionally adjusting one assumption or another). Optionally, run some alternate scenarios to figure out the effect (e.g. delay SSA or not being a good example).

I personally follow that process, and I also ran simulations for a few fellow Bogleheads from our local chapter, and this proved very handy. It also showed me that nobody has the same sequence of cash flows, that the exception IS the rule, so you'd better keep the math very open ('hands on').
All of that sounds very reasonable to me. Yes, it involves some effort, but most worthwhile activities do. What I really like about this approach is that you can mold the assumptions around your particular situation rather than just take some generic assumptions that may not be applicable to you. The ability to model specific events, changing market conditions, changing life expectancy, bequest amounts, etc. makes this process very appealing. Also, it becomes very easy to back- or front-load withdrawals, depending on one's desires.
“It's a dangerous business, Frodo, going out your door. You step onto the road, and if you don't keep your feet, there's no knowing where you might be swept off to.” J.R.R. Tolkien,The Lord of the Rings

User avatar
Topic Author
willthrill81
Posts: 10324
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: Time Value of Money vs. Variable Percentage Withdrawal (VPW)

Post by willthrill81 » Tue Nov 13, 2018 7:55 pm

AlohaJoe wrote:
Tue Nov 13, 2018 7:27 pm
siamond wrote:
Tue Nov 13, 2018 4:28 pm
I don't have a good answer for the last point. If a deep crisis occurs in my late 60s, I will definitely cancel my SSA postponement plans, and this will help (numerically AND behaviorally). If a deep crisis occurs before that, I will probably get itchy, hence regulate myself and spend less from the 'NPV/PMT' present value than allowed by the corresponding math, but I don't have a formulaic plan for that.
Here are two things people can do to make it a bit more formulaic, if they are so inclined.

#1: Many of us have this vague notion that annuities are a decent idea but work best for only a small number of situations: if you have plenty of money, annuitising is throwing it away; if you don't have enough money, buying an annuity just locks in failure. Jim Otar was the first I saw to talk about this. But where exactly does that "annuity boundary" lie for us in our circumstances? If our portfolio drops enough does it ever make sense to annuitise some or all of it?

We know that figuring out complicated and emotion-laden questions under pressure isn't ideal. Who wants to lie in bed at 2am trying to figure out if they should annuitise before the markets open tomorrow in case the crash-of-a-lifetime continues?

I've been thinking about this a bit more since reading this recent blog post: https://rivershedge.blogspot.com/2018/1 ... ndary.html but I haven't actually formalised anything yet. But I think there's scope for building something like that into a spreadsheet, at least as a flashing warning light.

#2: You can build in some simple math to do the spending cuts for you. These aren't fool-proof or anything but it at least takes out some of the random human guessing. "Should I spend $5,000 less than PMT says or $8,000 less?" Keep in mind that I'm (almost) as concerned about under-spending as over-spending. By having some math, my hope is avoid my own knee-jerk reactions. "OMG!!! The market is down 10% I need to cut spending 50% immediately!"

John Walton published a series of articles in Advisor Perspectives in 2016, including "A New Framework for Comparing Withdrawal Strategies". It took me a few readings to warm to his concept (he explained it in what I thought was a quite roundabout way) -- and it isn't a silver bullet or anything. He actually talks about PMT, which he calls "mortgage withdrawals", but he applies a "tilt" to the result of the PMT calculation.

The interesting thing about his tilt mechanism is that the same math can be used to tilt towards income stability or capital stability. The math is simply

Code: Select all

 PMT * ((current_capital / target_capital) ^ tilt) 
Current capital is easy; that's just your current portfolio value.

Target capital is whatever you goal is. You can get that number is one of two ways: you can do PV calculations or you just pick a nice, round nominal number.

The tilt is a number between -1 and +1. Positive numbers favor capital stability. Negative numbers favor income stability.

Here's how it might work in practice.

Say that PMT says you can withdraw $50,000 this year. Your portfolio is $1.2 million now but before the epic market crash it was $1.5 million and you'd prefer if your portfolio were back at that mark. 1.2/1.5 = 0.8.

If you pick a tilt of 0.3 then 0.8^0.3 = 0.93. Instead of $50,000 take 93% of $50,000.
If you pick a tilt of 0.5 then 0.8^0.5 = 0.89. Instead of $50,000 take 89% of $50,000.
If you pick a tilt of 1.0 then 0.8^1 = 0.8. Instead of $50,000 take 80% of $50,000.

I do this and picked a largish tilt, since I have a relatively large amount of discretionary spending and have the flexibility to vary my spending by tens of thousands of dollars a year if I want to/need to.
I'll have to mull that one over. We too plan on discretionary spending being at least 50% of our withdrawals, making them very flexible, but I find myself drawn to the smoothing guideline you referred to earlier (i.e. no more than +/- 10% annual change in withdrawals).
“It's a dangerous business, Frodo, going out your door. You step onto the road, and if you don't keep your feet, there's no knowing where you might be swept off to.” J.R.R. Tolkien,The Lord of the Rings

User avatar
LadyGeek
Site Admin
Posts: 54285
Joined: Sat Dec 20, 2008 5:34 pm
Location: Philadelphia
Contact:

Re: Time Value of Money vs. Variable Percentage Withdrawal (VPW)

Post by LadyGeek » Tue Nov 13, 2018 8:07 pm

A quick diversion for investors who may not understand the calculations under discussion:

The Time Value of Money calculations are well worth learning. Please spend some time to understand this most basic, but essential, part of finance and investing.

Once the concepts are understood, you can solve any financial equation. Calculating loan payments, mortgages, and interest rates (paid and received) are just a small example of the things you can do.

- See: Time Value of Money Introduction | TVMCalcs.com
- Also see the wiki: Comparing investments

This thread is using the concepts to estimate how much you can spend while in retirement.

If anyone has a question about the Time Value of Money calculations, or you try one on your own and get stuck, please start a thread in the Investing - Theory, News & General forum and we'll help you work through it.
Wiki To some, the glass is half full. To others, the glass is half empty. To an engineer, it's twice the size it needs to be.

User avatar
siamond
Posts: 4771
Joined: Mon May 28, 2012 5:50 am

Re: Time Value of Money vs. Variable Percentage Withdrawal (VPW)

Post by siamond » Tue Nov 13, 2018 8:27 pm

AlohaJoe wrote:
Tue Nov 13, 2018 7:27 pm
siamond wrote:
Tue Nov 13, 2018 4:28 pm
I don't have a good answer for the last point. If a deep crisis occurs in my late 60s, I will definitely cancel my SSA postponement plans, and this will help (numerically AND behaviorally). If a deep crisis occurs before that, I will probably get itchy, hence regulate myself and spend less from the 'NPV/PMT' present value than allowed by the corresponding math, but I don't have a formulaic plan for that.
Here are two things people can do to make it a bit more formulaic, if they are so inclined. [...]
Thanks for the pointers, but I think such approaches are more related to the basic topic of smoothing and/or define a floor for the REGULAR part of the withdrawal (the one derived from a regular PMT calculation -or else- on the regular portfolio). Here I'm fine, I have my way to proceed...

My hesitation is more about the outcome of the NPV/PMT math derived from the future cash flows. Which is kind of funny money because it's a (hopefully fairly safe) bet on the future, while all you have as concrete $$ to spend that year are shares stuck in a deep drawdown... I guess I should run more simulations of troublesome situations to better appreciate the desirable dynamics here. I do agree with you that a lack of a plan isn't a good thing in times of crisis.

PS. it could be tempting to sell bonds by then to hold on for a few years (after all, we're speaking pre-SSA years here, hard to believe we'll see multiple deep crises in a row until we're eligible for SSA), and only rebalance once the stock market is in better shape and/or SSA can be obtained, but this is something I ruled out in my regular AA rebalancing, so I am not too hot about it...

User avatar
Topic Author
willthrill81
Posts: 10324
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: Time Value of Money vs. Variable Percentage Withdrawal (VPW)

Post by willthrill81 » Tue Nov 13, 2018 9:37 pm

siamond wrote:
Tue Nov 13, 2018 8:27 pm
AlohaJoe wrote:
Tue Nov 13, 2018 7:27 pm
siamond wrote:
Tue Nov 13, 2018 4:28 pm
I don't have a good answer for the last point. If a deep crisis occurs in my late 60s, I will definitely cancel my SSA postponement plans, and this will help (numerically AND behaviorally). If a deep crisis occurs before that, I will probably get itchy, hence regulate myself and spend less from the 'NPV/PMT' present value than allowed by the corresponding math, but I don't have a formulaic plan for that.
Here are two things people can do to make it a bit more formulaic, if they are so inclined. [...]
Thanks for the pointers, but I think such approaches are more related to the basic topic of smoothing and/or define a floor for the REGULAR part of the withdrawal (the one derived from a regular PMT calculation -or else- on the regular portfolio). Here I'm fine, I have my way to proceed...

My hesitation is more about the outcome of the NPV/PMT math derived from the future cash flows. Which is kind of funny money because it's a (hopefully fairly safe) bet on the future, while all you have as concrete $$ to spend that year are shares stuck in a deep drawdown... I guess I should run more simulations of troublesome situations to better appreciate the desirable dynamics here. I do agree with you that a lack of a plan isn't a good thing in times of crisis.

PS. it could be tempting to sell bonds by then to hold on for a few years (after all, we're speaking pre-SSA years here, hard to believe we'll see multiple deep crises in a row until we're eligible for SSA), and only rebalance once the stock market is in better shape and/or SSA can be obtained, but this is something I ruled out in my regular AA rebalancing, so I am not too hot about it...
So you're concerned that following the PMT approach might lead you to sell equities at a loss in a bear market? How is that different from any other withdrawal approach?
“It's a dangerous business, Frodo, going out your door. You step onto the road, and if you don't keep your feet, there's no knowing where you might be swept off to.” J.R.R. Tolkien,The Lord of the Rings

User avatar
siamond
Posts: 4771
Joined: Mon May 28, 2012 5:50 am

Re: Time Value of Money vs. Variable Percentage Withdrawal (VPW)

Post by siamond » Tue Nov 13, 2018 9:46 pm

willthrill81 wrote:
Tue Nov 13, 2018 9:37 pm
So you're concerned that following the PMT approach might lead you to sell equities at a loss in a bear market? How is that different from any other withdrawal approach?
I have no concern with regular withdrawal approaches, with a formula directly derived from the regular portfolio. Well, with the sound approaches, that is. It's the double whammy of hitting the low-price shares due to regular withdrawals PLUS the cash flow math which rattles me a bit. Notably knowing that the regular withdrawal should be adaptive and get lower during a crisis (even with some smoothing in action), while the outcome of the cash flow math will not. Maybe this is silly, I don't know, I am just sharing where I am on this topic, that's all.

User avatar
siamond
Posts: 4771
Joined: Mon May 28, 2012 5:50 am

Re: Time Value of Money vs. Variable Percentage Withdrawal (VPW)

Post by siamond » Wed Nov 14, 2018 9:59 am

It finally remembered that I made a fairly simple example of such NPV/PMT math that I shared with somebody a year ago... Here it is, for whatever it's worth (check the cell notes; I didn't make a lot of efforts to make it fully educational, but it should be fairly self-explanatory):

https://docs.google.com/spreadsheets/d/ ... sp=sharing

GAAP
Posts: 850
Joined: Fri Apr 08, 2016 12:41 pm

Re: Time Value of Money vs. Variable Percentage Withdrawal (VPW)

Post by GAAP » Wed Nov 14, 2018 11:54 am

willthrill81 wrote:
Mon Nov 12, 2018 10:16 pm
So far, the only drawback of this annuitization approach to VPW that I can see is that it requires the user to input a couple of unknowns, the number of years to amortize over and the expected rate of return, to compute the current year's withdrawals. For most Bogleheads, this seems like a very minimal drawback. Can anyone identify any other disadvantages compared to VPW?

In truth, VPW is just a specific application of this annuitization approach (i.e. number of years to amortize is equal to 100 minus your current age, and return assumptions are historic).
Treating the retirement portfolio as a perpetuity can eliminate one of the unknowns.

The essence of a perpetuity is that the net real result from withdrawals and returns is zero. Stated another way, you need to recover the portfolio value after a drop due to the withdrawal. The perpetuity withdrawal rate is the reciprocal of the recovery rate.

The PMT function can be used to achieve this -- assuming you estimate the rate of return correctly. In this idealized world, the formula for the withdrawal rate is PMT(Rate,1,-1,1,1). This particular form of the PMT equation is somewhat unique, you can actually use any positive value for NPER and get exactly the same result.

Assuming your internal rate estimate is reasonably close, the perpetuity function actually describes the upper limit of how much you can withdraw. If you really want to recover or grow prior value, you must withdraw less than this amount. Otherwise, over time you converge on a decreasing number. Taking less than the perpetuity amount leaves some space for growth and as an error margin for rough real-world estimates. In a world where the long-term trend is for growth, this effectively means you're generally in a slow accumulation mode.

I use that general form with a combination of 1/CAPE10 to estimate stock returns, a somewhat arbitrary (and low) estimate for bond returns, and a bit of "tweaking" to provide a little more smoothing and value protection.

A perpetuity method explicitly intends to not draw down the portfolio, yet provide reasonably reliable income. If you want to spend your last dime on the day you die, then this is not for you.

User avatar
Topic Author
willthrill81
Posts: 10324
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: Time Value of Money vs. Variable Percentage Withdrawal (VPW)

Post by willthrill81 » Wed Nov 14, 2018 2:07 pm

GAAP wrote:
Wed Nov 14, 2018 11:54 am
willthrill81 wrote:
Mon Nov 12, 2018 10:16 pm
So far, the only drawback of this annuitization approach to VPW that I can see is that it requires the user to input a couple of unknowns, the number of years to amortize over and the expected rate of return, to compute the current year's withdrawals. For most Bogleheads, this seems like a very minimal drawback. Can anyone identify any other disadvantages compared to VPW?

In truth, VPW is just a specific application of this annuitization approach (i.e. number of years to amortize is equal to 100 minus your current age, and return assumptions are historic).
Treating the retirement portfolio as a perpetuity can eliminate one of the unknowns.

The essence of a perpetuity is that the net real result from withdrawals and returns is zero. Stated another way, you need to recover the portfolio value after a drop due to the withdrawal. The perpetuity withdrawal rate is the reciprocal of the recovery rate.

The PMT function can be used to achieve this -- assuming you estimate the rate of return correctly. In this idealized world, the formula for the withdrawal rate is PMT(Rate,1,-1,1,1). This particular form of the PMT equation is somewhat unique, you can actually use any positive value for NPER and get exactly the same result.

Assuming your internal rate estimate is reasonably close, the perpetuity function actually describes the upper limit of how much you can withdraw. If you really want to recover or grow prior value, you must withdraw less than this amount. Otherwise, over time you converge on a decreasing number. Taking less than the perpetuity amount leaves some space for growth and as an error margin for rough real-world estimates. In a world where the long-term trend is for growth, this effectively means you're generally in a slow accumulation mode.

I use that general form with a combination of 1/CAPE10 to estimate stock returns, a somewhat arbitrary (and low) estimate for bond returns, and a bit of "tweaking" to provide a little more smoothing and value protection.

A perpetuity method explicitly intends to not draw down the portfolio, yet provide reasonably reliable income. If you want to spend your last dime on the day you die, then this is not for you.
Correct me if I'm wrong, but if the future value (FV) is equal to the present value (PV), then the withdrawals (PMT) is equal to the (real) rate multiplied by the PV. If so, that definitely is a conservative route, especially since you should be using a historically low real rate of return.
“It's a dangerous business, Frodo, going out your door. You step onto the road, and if you don't keep your feet, there's no knowing where you might be swept off to.” J.R.R. Tolkien,The Lord of the Rings

GAAP
Posts: 850
Joined: Fri Apr 08, 2016 12:41 pm

Re: Time Value of Money vs. Variable Percentage Withdrawal (VPW)

Post by GAAP » Wed Nov 14, 2018 5:17 pm

willthrill81 wrote:
Wed Nov 14, 2018 2:07 pm
GAAP wrote:
Wed Nov 14, 2018 11:54 am
willthrill81 wrote:
Mon Nov 12, 2018 10:16 pm
So far, the only drawback of this annuitization approach to VPW that I can see is that it requires the user to input a couple of unknowns, the number of years to amortize over and the expected rate of return, to compute the current year's withdrawals. For most Bogleheads, this seems like a very minimal drawback. Can anyone identify any other disadvantages compared to VPW?

In truth, VPW is just a specific application of this annuitization approach (i.e. number of years to amortize is equal to 100 minus your current age, and return assumptions are historic).
Treating the retirement portfolio as a perpetuity can eliminate one of the unknowns.

The essence of a perpetuity is that the net real result from withdrawals and returns is zero. Stated another way, you need to recover the portfolio value after a drop due to the withdrawal. The perpetuity withdrawal rate is the reciprocal of the recovery rate.

The PMT function can be used to achieve this -- assuming you estimate the rate of return correctly. In this idealized world, the formula for the withdrawal rate is PMT(Rate,1,-1,1,1). This particular form of the PMT equation is somewhat unique, you can actually use any positive value for NPER and get exactly the same result.

Assuming your internal rate estimate is reasonably close, the perpetuity function actually describes the upper limit of how much you can withdraw. If you really want to recover or grow prior value, you must withdraw less than this amount. Otherwise, over time you converge on a decreasing number. Taking less than the perpetuity amount leaves some space for growth and as an error margin for rough real-world estimates. In a world where the long-term trend is for growth, this effectively means you're generally in a slow accumulation mode.

I use that general form with a combination of 1/CAPE10 to estimate stock returns, a somewhat arbitrary (and low) estimate for bond returns, and a bit of "tweaking" to provide a little more smoothing and value protection.

A perpetuity method explicitly intends to not draw down the portfolio, yet provide reasonably reliable income. If you want to spend your last dime on the day you die, then this is not for you.
Correct me if I'm wrong, but if the future value (FV) is equal to the present value (PV), then the withdrawals (PMT) is equal to the (real) rate multiplied by the PV. If so, that definitely is a conservative route, especially since you should be using a historically low real rate of return.
The return has to be slightly more than the withdrawal rate since it is only generated by the net PV after withdrawals. For example, with an estimated 4% real rate of return, the perpetuity withdrawal rate calculates to 3.85%.

Under current conditions, that does mean pretty small withdrawal rates. My AA is fairly heavy in equities, and they are internationally distributed with a lower 1/CAPE10 value than a strictly domestic allocation -- which helps increase the rate somewhat. Certainly, the conservatism in the method makes it easier to ignore market volatility.

jmk
Posts: 454
Joined: Tue Nov 01, 2011 7:48 pm

Re: Variable Percentage Withdrawal (VPW)

Post by jmk » Wed Nov 14, 2018 6:26 pm

siamond wrote:
Tue Nov 13, 2018 3:52 pm
willthrill81 wrote:
Tue Nov 13, 2018 2:02 pm
Then let's continue the discussion in that vein. Why could a TVM (i.e. annuitization) approach work or not work to model a full retirement plan? Given that it's a generalized form of the VPW approach, I see no reason why not beyond the aforementioned 'hands on' aspect, which is frankly a feature more than a drawback for me at least.

As you note, this approach could be easily used in conjunction with known future liabilities and income streams. Excel would make this fairly simple.
Yes, indeed, if one is minimally spreadsheet-inclined AND has a bit of a clue about Annuitization and Present Value concepts (as embodied by PMT and NPV() spreadsheet functions), it is fairly trivial to do. And extremely useful.
Ken Steiner provides a spreadsheet to do a version of what you describe on his website. I personally prefer his method to VPW since I can put in the inputs, including longevity, discount rate, misc expenses and incomes etc. His version is:
Accumulated Savings + PV Income from Other Sources = PV Future Non-Recurring Expenses + PV Future Recurring Annual Spending Budgets
http://howmuchcaniaffordtospendinretire ... eets.html

User avatar
siamond
Posts: 4771
Joined: Mon May 28, 2012 5:50 am

Re: Time Value of Money vs. Variable Percentage Withdrawal (VPW)

Post by siamond » Fri Nov 16, 2018 3:04 pm

siamond wrote:
Tue Nov 13, 2018 4:28 pm
- now for the biggie. Before the bulk of the cash flow incomes start trickling in (e.g. before SSA/Pensions kick in), the result of this process is to spend more shares from the regular savings portfolio than one would otherwise. Which is fine, it totally makes sense to use more of it before the SSA years and less of it later. BUT. A bad sequence of returns can really hurt, making it a double-whammy of 'selling low'. In some extreme cases, one could bankrupt their savings before reaching the SSA years, which would be rather unfortunate!
I came back to this issue. I extended the Google Sheet I previously shared, adding separate tabs to simulate a full scenario of somebody early retiring 10 years before getting solid fixed income (e.g. SSA/Pension) and then enjoying 30 more years of retirement. Here is the link, please duplicate the Google sheet to play by yourself:
https://docs.google.com/spreadsheets/d/ ... sp=sharing

This is definitely more hands-on, so please bear with me and ponder the formulas (and cell notes) as you go... Check scenario 1 first:
- 10 years of pre-SS fixed income, with a lump sum coming mid-way (house downsizing), then SS kicks off and 30 more years of retirement ensue
- Solid portfolio of $1M to start with
- Columns A to P basically work like the simple example of the 1st tab, NPV+PMT math on cash flows, regular PMT math on regular portfolio, add both quantities to get a spending budget in column M (and P)
- Ignore columns N and O for now
- Just one small thing, I added a 'grace period' of 10 years to the regular PMT math to leave some money at the end (e.g. bequest and/or LTC and/or extra longevity insurance). Set cell K4 to zero to come back to a more VPW-like logic.
- Check columns R and S, you can select which AA to test with and the historical year to start from (1965/66/73 are really bad ones)
- Play around a bit and you'll probably convince yourself that this all looks pretty much ok

Then check scenario 2:
- much smaller savings portfolio to start with, no house downsizing
- some part-time work while transitioning in retirement provides a modicum of extra income for a few years
- the expected SS/Pension is quite high (hence a high outcome of the NPV/PMT math on cash flows), but...
- ... the regular portfolio is typically emptied before SS/Pension kicks in, which is rather troublesome!

Time to check column O:
- set cell O1 to TRUE (scenario 2), and you'll see that the spending budget (column P) is constrained to do the best with lemons, i.e. spend all the savings before SS/Pension kicks in, but NOT prematurely.
- more generally, one probably has some kind of minimum target in mind for their savings before SS/Pension kicks in. In this extreme case (scenario 2), $0 seems the way to go, but usually one would probably want to keep some amount of savings to beef up income in the SS/Pension years.
- if we come back to scenario 1, I picked $400,000 (i.e. $20,000 per year for a 5% SWR).

How does this work? In column O, for the pre-SS years, I computed a spending (upper) cap defined as:
- recurring income if any (e.g. some part-time work), plus...
- a PMT formula running over the years remaining till SS/Pension kicks in (i.e. 10 years, then 9, etc)...
- said PMT formula uses a very dire rate of (real) return, i.e. 0%, and...
- it aims at a final value (cell O4, which is $0 in scenario 2 and $400,000 in scenario 1)
- point being to NOT deplete the regular portfolio beyond such final value
- the formula in column P uses the spending gate results (and also the floor from column N) to regulate spending.

A couple of additional subtleties:
- this is also regulated by the regular PMT formula on the regular portfolio as a minimum, to not become overly pessimistic (maybe this is debatable?)
- note that I computed $400k in scenario 1 using a 5% SWR and not a 3.5% or 4% SWR because the gate math is ALREADY assuming a really bad time in the pre SS/Pension years, so piling up worst case upon worst case doesn't seem sensible.

Yeah, I know, we're drifting away from simplicity goals, but for somebody "hands on", I think this makes reasonable sense. Picking up the right Final-Value for the pre-SS years isn't fully straightforward though. Feedback welcome.

Post Reply