Amortization Based Withdrawal (ABW)

Non-investing personal finance issues including insurance, credit, real estate, taxes, employment and legal issues such as trusts and wills
AlohaJoe
Posts: 6022
Joined: Mon Nov 26, 2007 2:00 pm
Location: Saigon, Vietnam

Re: Using the Time Value of Money Formula to Determine Withdrawals: Year 2000 Retiree Example

Post by AlohaJoe »

michaeljc70 wrote: Sun Feb 07, 2021 9:15 am
Another thing is why would I spend money just because a formula tells me I can? Should I go to the casino and blow it?
I am always a bit saddened by how many Bogleheads are completely incapable of imagining using their portfolio for anything but spending on themselves.

ABW, and other similar approaches, give you guidelines for how much giving -- to family, to charities, to church, wherever your priorities are -- today instead of decades from now.

If you really can't think of anything better to do with your legacy than spend it at a casino I feel a bit sorry for the choices you've made over the many decades of your life while you were building your portfolio.
dcabler
Posts: 1931
Joined: Wed Feb 19, 2014 11:30 am

Re: Using the Time Value of Money Formula to Determine Withdrawals: Year 2000 Retiree Example

Post by dcabler »

michaeljc70 wrote: Sun Feb 07, 2021 9:15 am
jmk wrote: Sat Feb 06, 2021 6:57 pm
dknightd wrote: Thu Oct 29, 2020 12:11 pm And to me this is one of the biggest weakness of the PMT/TVM approach.
You have to predict when you will die.
You have to predict inflation.
You have to predict returns on investments.
You have to predict what your future expenses will be.
You have to predict when and how your cognitive ability will decline.
I suspect I have a 0% chance of predicting any one of those correctly.
These issues arise with any withdrawal approach one does oneself, other than hiring someone.
The PMT approach is one of the easiest to implement: it is as simple as one equation per year.
The IRS in fact uses a simple form of it in the RMD amounts, and if one wants to be really simple you can assume 0% return and the average life expectancy and have it as simple as 1/y, where y is the number of years one wants to live with rounding toward the low end.
True. Which is why I don't use any formal methodology to increase/decrease my spending in retirement.

A couple of other things I would add to the list above is you have to predict future tax rates/laws (and their impact on you) and you have to predict a potential inheritance (if applicable to you). We always say don't count on an inheritance, which is good advice, but it could potentially leave you with a lot more money than you planned for at an age where you are less likely to be spending more. So, I wouldn't count on any inheritance but I also wouldn't totally ignore a potential inheritance.

Another thing is why would I spend money just because a formula tells me I can? Should I go to the casino and blow it? Some years I want to take more/more expensive vacations. Some years I have big non-recurring expenses (new roof, new car, unexpected medical bill, etc.) Some years I might be in a pandemic and spend less :shock: I think I have a pretty good grasp if I can spend more or if I should spend less. Since I am in the first 5 years of retirement I am being conservative to avoid sequence of return problems.
Nothing in any of the formal withdrawal methods I know of compels one to either withdraw the full calculated amount, nor to spend it if one does. Not sure I know anybody who is retired who actually fully slavishly follows such a calculation, but rather uses it as a guideline since this is real life.

One form of smoothing for any of the more formal methods is to just leave any excess in the portfolio for the future if the calculated withdrawal is more than you actually need. Hopefully it continues to grow. Another is to go ahead and withdraw and bank it for a rainy day in case a year comes along where the withdrawal is less than needed expenses. And of course there are always unexpected expenses - real life is lumpy.

Also most formal withdrawal methods I'm aware of only calculate how much you can withdraw. It's up to the user to know whether that covers all of the expenses or not. If it does, great. If it doesn't, then there are the usual suggestions of reducing expenses, working longer & saving more (when possible). I'm not retired yet, but plan on using my own ABW spreadsheet that I created several years ago. And at least a few times a year, I pretend that I'm about to retire and I calculate what my withdrawal would be, along with estimated expenses & taxes as yet another way to answer the question: "Am I there yet?".

Cheers.
michaeljc70
Posts: 7826
Joined: Thu Oct 15, 2015 3:53 pm

Re: Using the Time Value of Money Formula to Determine Withdrawals: Year 2000 Retiree Example

Post by michaeljc70 »

AlohaJoe wrote: Sun Feb 07, 2021 10:51 am
michaeljc70 wrote: Sun Feb 07, 2021 9:15 am
Another thing is why would I spend money just because a formula tells me I can? Should I go to the casino and blow it?
I am always a bit saddened by how many Bogleheads are completely incapable of imagining using their portfolio for anything but spending on themselves.

ABW, and other similar approaches, give you guidelines for how much giving -- to family, to charities, to church, wherever your priorities are -- today instead of decades from now.

If you really can't think of anything better to do with your legacy than spend it at a casino I feel a bit sorry for the choices you've made over the many decades of your life while you were building your portfolio.
I think that is a bit harsh. I retired based on having enough to live a life that makes me happy including satisfying my hobbies and interests, general expense, giving gifts, etc. Retiring in my 40s with 20+ years until FRA for SS and having relatives that have lived to 95 puts me in a bit of a different situation than someone retiring at 65 with investments, SS and a pension. As pointed out by someone else in the post you partially quoted, there are a lot of variables in retirement planning/withdrawing. You hope you get them right, but there is more risk when you have a potential 50 year retirement ahead of you. It is not that I cannot possibly think of something to spend excess money on but I am not going to start throwing money around because the stock market was good for a few years at this point in my retirement.
User avatar
Topic Author
willthrill81
Posts: 25278
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: Using the Time Value of Money Formula to Determine Withdrawals: Year 2000 Retiree Example

Post by willthrill81 »

michaeljc70 wrote: Sun Feb 07, 2021 11:14 am
AlohaJoe wrote: Sun Feb 07, 2021 10:51 am
michaeljc70 wrote: Sun Feb 07, 2021 9:15 am
Another thing is why would I spend money just because a formula tells me I can? Should I go to the casino and blow it?
I am always a bit saddened by how many Bogleheads are completely incapable of imagining using their portfolio for anything but spending on themselves.

ABW, and other similar approaches, give you guidelines for how much giving -- to family, to charities, to church, wherever your priorities are -- today instead of decades from now.

If you really can't think of anything better to do with your legacy than spend it at a casino I feel a bit sorry for the choices you've made over the many decades of your life while you were building your portfolio.
I think that is a bit harsh. I retired based on having enough to live a life that makes me happy including satisfying my hobbies and interests, general expense, giving gifts, etc. Retiring in my 40s with 20+ years until FRA for SS and having relatives that have lived to 95 puts me in a bit of a different situation than someone retiring at 65 with investments, SS and a pension. As pointed out by someone else in the post you partially quoted, there are a lot of variables in retirement planning/withdrawing. You hope you get them right, but there is more risk when you have a potential 50 year retirement ahead of you. It is not that I cannot possibly think of something to spend excess money on but I am not going to start throwing money around because the stock market was good for a few years at this point in my retirement.
If you don't need or want to spend as much as the ABW indicates that you can, then don't. Doing so will further pad your finances.

I think that AlohaJoe's point was simply that you could give the surplus to those in need.
“Good and ill have not changed since yesteryear; nor are they one thing among Elves and Dwarves and another among Men.” J.R.R. Tolkien, The Lord of the Rings
GAAP
Posts: 1324
Joined: Fri Apr 08, 2016 12:41 pm

Re: Using the Time Value of Money Formula to Determine Withdrawals: Year 2000 Retiree Example

Post by GAAP »

michaeljc70 wrote: Sun Feb 07, 2021 9:15 am Another thing is why would I spend money just because a formula tells me I can? Should I go to the casino and blow it? Some years I want to take more/more expensive vacations. Some years I have big non-recurring expenses (new roof, new car, unexpected medical bill, etc.) Some years I might be in a pandemic and spend less :shock: I think I have a pretty good grasp if I can spend more or if I should spend less. Since I am in the first 5 years of retirement I am being conservative to avoid sequence of return problems.
I use the output from a process like this as an input to create a budgetary number for setting a safe upper limit to the spending rate for the year. Other inputs include the planned non-recurring items. But the budgetary limit is not absolute, nor is it really a target to hit. Actual spending is an entirely separate thing -- just guided by the results of computation somewhat like ABW.
“Adapt what is useful, reject what is useless, and add what is specifically your own.” ― Bruce Lee
ryman554
Posts: 1380
Joined: Sun Jan 12, 2014 9:44 pm

Re: Using the Time Value of Money Formula to Determine Withdrawals: Year 2000 Retiree Example

Post by ryman554 »

michaeljc70 wrote: Sun Feb 07, 2021 9:15 am
A couple of other things I would add to the list above is you have to predict future tax rates/laws (and their impact on you) and you have to predict a potential inheritance (if applicable to you). We always say don't count on an inheritance, which is good advice, but it could potentially leave you with a lot more money than you planned for at an age where you are less likely to be spending more. So, I wouldn't count on any inheritance but I also wouldn't totally ignore a potential inheritance.
Tax rates are tricksy, given many different sources of income and taxation of which (roth vs ss vs taxable vs tira vs pension vs lottery winnings....) A spitball approach is what I'm using -- my effective tax rate over the past N years and estimates based on what's left in the portfolio and how it will be taxed (marginally)... gets me close/overstimates it going forward. Easy to put into a PMT-based system, as it just lops off the top from your budget so you can see the after tax estimate, but otherwise does not change the actual math.

Inheritance is also trivial, if you are using time-value-of-money. It's just a one-year income stream of unknown amount at an unknown time, but once you know that, the math works just the same as any other income stream (like, say, paying for kids/grandkids college, except in reverse!)

RIght now I have it in my worksheet as getting an inheritance in 20 years of $0. I've gamed it out by moving the amount and timing of it, and, well, it changes things a lot if the inheritance is in the 7 figure range. So I don't count on it until it's more sure, but I do run scenarios with it on to see how much of a budget it can add.. $10-$20k per annum maximally, or so.
User avatar
siamond
Posts: 5819
Joined: Mon May 28, 2012 5:50 am

Re: Using the Time Value of Money Formula to Determine Withdrawals: Year 2000 Retiree Example

Post by siamond »

ryman554 wrote: Tue Feb 09, 2021 3:12 pmRIght now I have it in my worksheet as getting an inheritance in 20 years of $0. I've gamed it out by moving the amount and timing of it, and, well, it changes things a lot if the inheritance is in the 7 figure range. So I don't count on it until it's more sure, but I do run scenarios with it on to see how much of a budget it can add.. $10-$20k per annum maximally, or so.
Perfect! This is what a flexible ABW spreadsheet can do for you, don't be afraid of uncertainties, just run scenarios and see what goes...
jmk
Posts: 588
Joined: Tue Nov 01, 2011 7:48 pm

Re: Using the Time Value of Money Formula to Determine Withdrawals: Year 2000 Retiree Example

Post by jmk »

Wrench wrote: Sat Feb 06, 2021 7:41 pm
jmk wrote: Sat Feb 06, 2021 6:57 pm
dknightd wrote: Thu Oct 29, 2020 12:11 pm And to me this is one of the biggest weakness of the PMT/TVM approach.
You have to predict when you will die.
You have to predict inflation.
You have to predict returns on investments.
You have to predict what your future expenses will be.
You have to predict when and how your cognitive ability will decline.
I suspect I have a 0% chance of predicting any one of those correctly.
These issues arise with any withdrawal approach one does oneself, other than hiring someone.
The PMT approach is one of the easiest to implement: it is as simple as one equation per year.
The IRS in fact uses a simple form of it in the RMD amounts, and if one wants to be really simple you can assume 0% return and the average life expectancy and have it as simple as 1/y, where y is the number of years one wants to live with rounding toward the low end.
There are lots of ways to create stable retirement income that do NOT require a PMT calculation. Here's one: work/save until you can take social security and then purchase an annuity that will cover the balance of your expenses. Invest remaining monies in an asset allocation for which you feel comfortable, and use those funds for one time/special expenses. You need not know ANY of the issues raised by dknightd, and you require NO calculations beyond RMDs. The approach in this thread is fine and can work well for the technically literate, and cognitively able. BUT, there are other ways to achieve an acceptable withdrawal approach that work fine too. As in all things financial, there is no one size fits all approach.
Getting an annuity to cover balance of expenses is certainly a credible alternative to PMT. But notice, again, that you still have to answer when you will die, expenses, inflation etc from the list above. You need to know these to know when to get an annuity and what kind. Either way there are a lot of unknowns. PMT or no PMT. I'm not at all saying everyone has to use PMT, just that the list of questions above is not the deal breaker it is being presented.
jmk
Posts: 588
Joined: Tue Nov 01, 2011 7:48 pm

Re: Amortization Based Withdrawal (ABW)

Post by jmk »

.....
Last edited by jmk on Mon Mar 08, 2021 1:18 am, edited 1 time in total.
jmk
Posts: 588
Joined: Tue Nov 01, 2011 7:48 pm

Re: Amortization Based Withdrawal (ABW)

Post by jmk »

corn18 wrote: Sun Feb 07, 2021 10:01 am If I wanted to model my chosen 2% real return, how would I do that in my simple approach? I thought about doing a NPV on the expenses with the rate = 2%.
Basically, yes, that is what you'd do. It's the same thing, but instead of discount rate of 0% like you did in your original analysis (raw costs and assets, assuming 0% growth), you'd have 2% growth, which would account for the time value of money. Most would advice you run NPV of both income and expenses, not just expenses. There are controversies about what you should use as the discount rate: many say your expected rate of return across portfolio, others say the safe rate.
User avatar
Ben Mathew
Posts: 1685
Joined: Tue Mar 13, 2018 11:41 am
Location: Seattle

Re: Amortization Based Withdrawal (ABW)

Post by Ben Mathew »

I think it will be useful to add the following two "simulator" spreadsheets as companions to the basic ABW calculator in the wiki:

ABW simulator
ABW monte carlo simulator

These spreadsheets show what withdrawals would be for a given sequence of returns.

The first spreadsheet (ABW simulator) shows what withdrawals will be for one particular sequence of returns. The screenshot below shows how it works. You enter a sequence of returns in the orange column. The corresponding withdrawals are calculated in the green column. I entered the historical returns for a 35/65 portfolio for the last 36 years (1985-2020). The resulting withdrawals range from $41,341 at age 65 to 150,553 at age 100. The reason the withdrawals grew faster than the scheduled g=0.5% is because the portfolio returns were higher than the expected 3%.

Image

The second spreadsheet (ABW monte carlo simulator) simply reruns this simulation 500 times by randomly drawing returns from a given set of returns. The screenshot below shows the results of randomly drawing returns from the 150 year historical returns of a 35/65 portfolio. (There is also an option to reduce historical returns to bring it in line with future expectations, if the user wants.)

Image

The results of the monte carlo simulation are summarized using percentiles. It shows the tradeoff between risk and return in the ABW withdrawal plan. The user can adjust their plan by changing asset allocation and g till they arrive at their preferred plan.

If there are no objections, I will add these two spreadsheets to the "basic calculator" section of the wiki.
User avatar
CalPoppy
Posts: 278
Joined: Thu May 04, 2017 2:57 pm

Re: Amortization Based Withdrawal (ABW)

Post by CalPoppy »

Ben Mathew wrote: Tue Jul 20, 2021 11:40 am The first spreadsheet (ABW simulator) shows what withdrawals will be for one particular sequence of returns. The screenshot below shows how it works. You enter a sequence of returns in the orange column. The corresponding withdrawals are calculated in the green column. I entered the historical returns for a 35/65 portfolio for the last 36 years (1985-2020). The resulting withdrawals range from $41,341 at age 65 to 150,553 at age 100. The reason the withdrawals grew faster than the scheduled g=0.5% is because the portfolio returns were higher than the expected 3%.
Hi Ben,

The sheets look great! Can you please help me understand the interplay between expected return and the historic returns?
User avatar
Ben Mathew
Posts: 1685
Joined: Tue Mar 13, 2018 11:41 am
Location: Seattle

Re: Amortization Based Withdrawal (ABW)

Post by Ben Mathew »

CalPoppy wrote: Tue Jul 20, 2021 4:00 pm Hi Ben,

The sheets look great! Can you please help me understand the interplay between expected return and the historic returns?
Sure. The spreadsheet itself does not force any link between expected return and simulated returns. So, for example, you are free to put in a 3% expected return in the ABW plan, but simulate using a distribution that has a 5% expected return. While it seems reasonable that the expected return of the simulated returns should equal the expected return of the ABW plan, I leaves it to the user to make that happen. I think this allows the user more flexibility in considering all the what-ifs.

Simulated returns could be anything--drawn from a historical distribution, drawn from a mathematical distribution, or just arbitrary made up numbers. Historical returns seem to be the most widely used in such simulations and so I used them in the examples I gave. But a concern that some people have is that because of high current valuations, expected future returns may be lower than historical returns. So I've made it easy in the spreadsheet to modify the historical returns to reduce it by some percentage points before using it in the simulation.
User avatar
CalPoppy
Posts: 278
Joined: Thu May 04, 2017 2:57 pm

Re: Amortization Based Withdrawal (ABW)

Post by CalPoppy »

Ben Mathew wrote: Tue Jul 20, 2021 5:07 pm
CalPoppy wrote: Tue Jul 20, 2021 4:00 pm Hi Ben,

The sheets look great! Can you please help me understand the interplay between expected return and the historic returns?
Sure. The spreadsheet itself does not force any link between expected return and simulated returns. So, for example, you are free to put in a 3% expected return in the ABW plan, but simulate using a distribution that has a 5% expected return. While it seems reasonable that the expected return of the simulated returns should equal the expected return of the ABW plan, I leaves it to the user to make that happen. I think this allows the user more flexibility in considering all the what-ifs.

Simulated returns could be anything--drawn from a historical distribution, drawn from a mathematical distribution, or just arbitrary made up numbers. Historical returns seem to be the most widely used in such simulations and so I used them in the examples I gave. But a concern that some people have is that because of high current valuations, expected future returns may be lower than historical returns. So I've made it easy in the spreadsheet to modify the historical returns to reduce it by some percentage points before using it in the simulation.
Thanks for the explanation!
klaus14
Posts: 733
Joined: Sun Nov 25, 2018 7:43 pm

Re: Amortization Based Withdrawal (ABW)

Post by klaus14 »

Dumb Question:

Is the return parameter in the spreadsheet arithmetic mean return or CAGR ?

I'd guess CAGR.
My investment algorithm: https://www.bogleheads.org/forum/viewtopic.php?f=10&t=351899&p=6112869#p6112869
sixtyforty
Posts: 571
Joined: Tue Nov 25, 2014 12:22 pm
Location: USA

Re: Amortization Based Withdrawal (ABW)

Post by sixtyforty »

I really like the idea of the simulators. Would it be possible to include future cash flows ? A $1M current portfolio with simulated returns will not yield the same results as a $1M portfolio that also receives a pension or SS. The differences could be significant.
"Simplicity is the ultimate sophistication" - Leonardo Da Vinci
User avatar
Ben Mathew
Posts: 1685
Joined: Tue Mar 13, 2018 11:41 am
Location: Seattle

Re: Amortization Based Withdrawal (ABW)

Post by Ben Mathew »

klaus14 wrote: Wed Jul 21, 2021 7:02 am Dumb Question:

Is the return parameter in the spreadsheet arithmetic mean return or CAGR ?

I'd guess CAGR.
Yes, it's CAGR.
User avatar
Ben Mathew
Posts: 1685
Joined: Tue Mar 13, 2018 11:41 am
Location: Seattle

Re: Amortization Based Withdrawal (ABW)

Post by Ben Mathew »

sixtyforty wrote: Wed Jul 21, 2021 8:47 am I really like the idea of the simulators. Would it be possible to include future cash flows ? A $1M current portfolio with simulated returns will not yield the same results as a $1M portfolio that also receives a pension or SS. The differences could be significant.
One way to handle a pension would be to set aside a part of your portfolio to build a bond bridge to cover the gap till the pension begins. Then you can enter the remaining funds in the simulator to model what will happen with your variable withdrawals.

For more options on incorporating future income and expenses, see total portfolio allocation and withdrawal (TPAW). That has the basic simulator (first spreadsheet) already. I'll add a monte carlo simulator (the second spreadsheet) in the next few weeks.
klaus14
Posts: 733
Joined: Sun Nov 25, 2018 7:43 pm

Re: Amortization Based Withdrawal (ABW)

Post by klaus14 »

Aiming to retire at age 40. I am thinking using ABW in the following fashion:

- Set duration to 30 years
- Set ending balance to 25%

At age 70, start receiving social security and invest a portion of the remaining balance to SPIA. If one expects these two to cover essential expenses, is this a reasonable approach?
My investment algorithm: https://www.bogleheads.org/forum/viewtopic.php?f=10&t=351899&p=6112869#p6112869
User avatar
CalPoppy
Posts: 278
Joined: Thu May 04, 2017 2:57 pm

Re: Amortization Based Withdrawal (ABW)

Post by CalPoppy »

klaus14 wrote: Thu Jul 22, 2021 8:30 am Aiming to retire at age 40. I am thinking using ABW in the following fashion:

- Set duration to 30 years
- Set ending balance to 25%

At age 70, start receiving social security and invest a portion of the remaining balance to SPIA. If one expects these two to cover essential expenses, is this a reasonable approach?
I would:
- Use your full time horizon (from age 40 until estimated end of life)
- Enter social security cash flows in the appropriate years
- Enter an amount for a SPIA as a large cash outflow at age 70

I use the "PV-based Annuitized Spending" spreadsheet found here http://bit.ly/2WuNmvf
User avatar
Ben Mathew
Posts: 1685
Joined: Tue Mar 13, 2018 11:41 am
Location: Seattle

Re: Amortization Based Withdrawal (ABW)

Post by Ben Mathew »

klaus14 wrote: Thu Jul 22, 2021 8:30 am Aiming to retire at age 40. I am thinking using ABW in the following fashion:

- Set duration to 30 years
- Set ending balance to 25%

At age 70, start receiving social security and invest a portion of the remaining balance to SPIA. If one expects these two to cover essential expenses, is this a reasonable approach?
Sounds like you want to use ABW for the gap years 40-70 and then rely primarily on SS and the SPIA. Your income during the gap years will be variable. Then when you turn 70 and SS kicks in, your income will become much more certain and likely quite different from what it was earlier. If that is in line with what you want, your strategy should work. But if you don't want that pre-70 post-70 dichotomy, you could make a plan where your pre SS income is more consistent with post SS income. One way is to build a bond bridge to age 70 and use ABW to calculate withdrawals on the remaining portfolio. My thoughts on when that would be appropriate, and an alternative for when it's not appropriate, are here:

Managing gap years without a bridge to Social Security
klaus14
Posts: 733
Joined: Sun Nov 25, 2018 7:43 pm

Re: Amortization Based Withdrawal (ABW)

Post by klaus14 »

Ben Mathew wrote: Thu Jul 22, 2021 10:38 am
klaus14 wrote: Thu Jul 22, 2021 8:30 am Aiming to retire at age 40. I am thinking using ABW in the following fashion:

- Set duration to 30 years
- Set ending balance to 25%

At age 70, start receiving social security and invest a portion of the remaining balance to SPIA. If one expects these two to cover essential expenses, is this a reasonable approach?
Sounds like you want to use ABW for the gap years 40-70 and then rely primarily on SS and the SPIA. Your income during the gap years will be variable. Then when you turn 70 and SS kicks in, your income will become much more certain and likely quite different from what it was earlier. If that is in line with what you want, your strategy should work. But if you don't want that pre-70 post-70 dichotomy, you could make a plan where your pre SS income is more consistent with post SS income. One way is to build a bond bridge to age 70 and use ABW to calculate withdrawals on the remaining portfolio. My thoughts on when that would be appropriate, and an alternative for when it's not appropriate, are here:

Managing gap years without a bridge to Social Security
Thank you Ben, I'll think about the method you suggested.

In my method, I have the ending balance knob available for tuning. 25% sounds arbitrary but I selected it to make pre-70 and post-70 close to each other (I hypothetically invest all ending balance to SPIA with today's SPIA payout rate). If I update numbers before every withdrawal accordingly, wouldn't my method avoid the problem you described?
My investment algorithm: https://www.bogleheads.org/forum/viewtopic.php?f=10&t=351899&p=6112869#p6112869
User avatar
Ben Mathew
Posts: 1685
Joined: Tue Mar 13, 2018 11:41 am
Location: Seattle

Re: Amortization Based Withdrawal (ABW)

Post by Ben Mathew »

klaus14 wrote: Thu Jul 22, 2021 11:49 am
Ben Mathew wrote: Thu Jul 22, 2021 10:38 am
klaus14 wrote: Thu Jul 22, 2021 8:30 am Aiming to retire at age 40. I am thinking using ABW in the following fashion:

- Set duration to 30 years
- Set ending balance to 25%

At age 70, start receiving social security and invest a portion of the remaining balance to SPIA. If one expects these two to cover essential expenses, is this a reasonable approach?
Sounds like you want to use ABW for the gap years 40-70 and then rely primarily on SS and the SPIA. Your income during the gap years will be variable. Then when you turn 70 and SS kicks in, your income will become much more certain and likely quite different from what it was earlier. If that is in line with what you want, your strategy should work. But if you don't want that pre-70 post-70 dichotomy, you could make a plan where your pre SS income is more consistent with post SS income. One way is to build a bond bridge to age 70 and use ABW to calculate withdrawals on the remaining portfolio. My thoughts on when that would be appropriate, and an alternative for when it's not appropriate, are here:

Managing gap years without a bridge to Social Security
Thank you Ben, I'll think about the method you suggested.

In my method, I have the ending balance knob available for tuning. 25% sounds arbitrary but I selected it to make pre-70 and post-70 close to each other (I hypothetically invest all ending balance to SPIA with today's SPIA payout rate). If I update numbers before every withdrawal accordingly, wouldn't my method avoid the problem you described?
Yes, if you keep updating the terminal balance to set SPIA income + SS equal to your current ABW withdrawal, the pre-70 and post-70 income levels will align. So that's good.

Still, note that with your method, you would be taking all of your portfolio risk before age 70 and then stop taking risk thereafter. There are some benefits to spreading risk across your whole life. You can get a higher return at lower risk because of better time diversification. The TPAW thread has more on this. But if you prefer to avoid risk after age 70, I think your method makes sense.
klaus14
Posts: 733
Joined: Sun Nov 25, 2018 7:43 pm

Re: Amortization Based Withdrawal (ABW)

Post by klaus14 »

Ben Mathew wrote: Thu Jul 22, 2021 12:25 pm
klaus14 wrote: Thu Jul 22, 2021 11:49 am
Ben Mathew wrote: Thu Jul 22, 2021 10:38 am
klaus14 wrote: Thu Jul 22, 2021 8:30 am Aiming to retire at age 40. I am thinking using ABW in the following fashion:

- Set duration to 30 years
- Set ending balance to 25%

At age 70, start receiving social security and invest a portion of the remaining balance to SPIA. If one expects these two to cover essential expenses, is this a reasonable approach?
Sounds like you want to use ABW for the gap years 40-70 and then rely primarily on SS and the SPIA. Your income during the gap years will be variable. Then when you turn 70 and SS kicks in, your income will become much more certain and likely quite different from what it was earlier. If that is in line with what you want, your strategy should work. But if you don't want that pre-70 post-70 dichotomy, you could make a plan where your pre SS income is more consistent with post SS income. One way is to build a bond bridge to age 70 and use ABW to calculate withdrawals on the remaining portfolio. My thoughts on when that would be appropriate, and an alternative for when it's not appropriate, are here:

Managing gap years without a bridge to Social Security
Thank you Ben, I'll think about the method you suggested.

In my method, I have the ending balance knob available for tuning. 25% sounds arbitrary but I selected it to make pre-70 and post-70 close to each other (I hypothetically invest all ending balance to SPIA with today's SPIA payout rate). If I update numbers before every withdrawal accordingly, wouldn't my method avoid the problem you described?
Yes, if you keep updating the terminal balance to set SPIA income + SS equal to your current ABW withdrawal, the pre-70 and post-70 income levels will align. So that's good.

Still, note that with your method, you would be taking all of your portfolio risk before age 70 and then stop taking risk thereafter. There are some benefits to spreading risk across your whole life. You can get a higher return at lower risk because of better time diversification. The TPAW thread has more on this. But if you prefer to avoid risk after age 70, I think your method makes sense.
This is also good analysis.
I have the "what percentage of ending balance should go to SPIA" knob too. I could use that if i wanted to take some risk at age 70. But yes, it will be hard to equalize that risk to pre-70. I think it is fine to err on the side of having less risk post-70.

Still, i understand this method is not as systematic as TPAW. But it is simpler and i understand the mechanics of it :) I'll study TPAW at some point though.
My investment algorithm: https://www.bogleheads.org/forum/viewtopic.php?f=10&t=351899&p=6112869#p6112869
Post Reply