Variable Percentage Withdrawal (VPW)

Discuss all general (i.e. non-personal) investing questions and issues, investing news, and theory.
User avatar
Leif
Posts: 2653
Joined: Wed Sep 19, 2007 4:15 pm

Re: Variable Percentage Withdrawal (VPW)

Post by Leif » Sun Jun 30, 2019 10:17 am

There is no stop date on the pension input in the retirement worksheet, so how do you recommend a SS Spousal benefit at FRA, then own benefit at 70, be handled?

What I do now is add the spousal benefit at 66, then add my own benefit at 70 minus the spousal benefit.

I would like to see a stop date and additional (above 4) inputs for pension. Thanks.

Topic Author
longinvest
Posts: 3809
Joined: Sat Aug 11, 2012 8:44 am

Re: Variable Percentage Withdrawal (VPW)

Post by longinvest » Sun Jun 30, 2019 12:51 pm

Leif wrote:
Sun Jun 30, 2019 10:17 am
There is no stop date on the pension input in the retirement worksheet, so how do you recommend a SS Spousal benefit at FRA, then own benefit at 70, be handled?
Leif, I would handle them as two distinct pensions, one that starts at FRA (spousal benefit), and one that starts at 70 (SS in excess of spousal benefit).
Leif wrote:
Sun Jun 30, 2019 10:17 am
What I do now is add the spousal benefit at 66, then add my own benefit at 70 minus the spousal benefit.
Yes, exactly!
Leif wrote:
Sun Jun 30, 2019 10:17 am
I would like to see a stop date and additional (above 4) inputs for pension. Thanks.
It would be quite difficult to handle a time-limited pension in the VPW Worksheet.

I think that your current approach works.

I think that 4 pensions is sufficient for one person. Two spouses can use two separate worksheets, one per spouse.
Bogleheads investment philosophy | single-ETF balanced portfolio | VBAL

jaj2276
Posts: 463
Joined: Sat Apr 16, 2011 5:13 pm

Re: Variable Percentage Withdrawal (VPW)

Post by jaj2276 » Mon Jul 01, 2019 7:56 am

At the request of OP in another thread, I've been using VPW for 3 years (this year being the 4th) to make some of our yearly charitable giving from a DAF I started in 2016. At that time I was 40 and wanted this lump sum to cover 20 years of charitable contributions.

The portfolio started with $107,213. It's invested in the Moderate Growth portfolio which is 45% Balanced Index, 45% Wellington, and 10% Total Intl Stock. From that I put in 56% Domestic Stocks, 10% Intl Stocks, and 34% Domestic Bonds to the VPW spreadsheet. I haven't changed this allocation so if Wellington has shifted any, I haven't picked it up (I assume Balanced will remain 60/40 and Intl will be 100% intl stock).

Here are the portfolio values along with distributions:

Year, %WD, Port $Bal, WD $Amt
2016,7.0%,$107213,$7504
2017,7.3%,$110874,$8094
2018,7.6%,$106469,$8092


This year will be a 7.9% distribution and portfolio balance is currently at $105,165. The next withdrawal amount will take place in October so still plenty of time to make/lose money.
Last edited by jaj2276 on Mon Jul 01, 2019 9:24 am, edited 1 time in total.

Topic Author
longinvest
Posts: 3809
Joined: Sat Aug 11, 2012 8:44 am

Re: Variable Percentage Withdrawal (VPW)

Post by longinvest » Mon Jul 01, 2019 8:23 am

jaj2276 wrote:
Mon Jul 01, 2019 7:56 am
At the request of OP in another thread, I've been using VPW for 3 years (this year being the 4th) to make some of our yearly charitable giving from a DAF I started in 2016. At that time I was 40 and wanted this lump sum to cover 20 years of charitable contributions.

The portfolio started with $107,213. It's invested in the Moderate Growth portfolio which is 45% Balanced Index, 45% Wellington, and 10% Total Intl Stock. From that I put in 56% Domestic Stocks, 10% Intl Stocks, and 34% Domestic Bonds to the VPW spreadsheet. I haven't changed this allocation so Wellington has shifted any, I haven't picked it up (I assume Balanced will remain 60/40 and Intl will be 100% intl stock).

Here are the portfolio values along with distributions:

Year, %WD, Port $Bal, WD $Amt
2016,7.0%,$107213,$7504
2017,7.3%,$110874,$8094
2018,7.6%,$106469,$8092


This year will be a 7.9% distribution and portfolio balance is currently at $105,165. The next withdrawal amount will take place in October so still plenty of time to make/lose money.
Thanks, Jaj2276, for your report.
Bogleheads investment philosophy | single-ETF balanced portfolio | VBAL

jaj2276
Posts: 463
Joined: Sat Apr 16, 2011 5:13 pm

Re: Variable Percentage Withdrawal (VPW)

Post by jaj2276 » Mon Jul 01, 2019 9:30 am

longinvest wrote:
Mon Jul 01, 2019 8:23 am
...

Thanks, Jaj2276, for your report.
Of course, and THANK YOU for the amount of work you've put in to both creating the idea/spreadsheet and the ongoing enhancements/improvements/discussions about it.

I'm curious whether this specific tool is ever used by independent financial planners to come up with retirement strategies for their clients. I know if I were one this would be the most used tool in the toolbox.

Topic Author
longinvest
Posts: 3809
Joined: Sat Aug 11, 2012 8:44 am

Re: Variable Percentage Withdrawal (VPW)

Post by longinvest » Mon Jul 01, 2019 9:38 am

jaj2276 wrote:
Mon Jul 01, 2019 9:30 am
I'm curious whether this specific tool is ever used by independent financial planners to come up with retirement strategies for their clients. I know if I were one this would be the most used tool in the toolbox.
I really have no idea. I'm just a do-it-yourself index investor.

I keep reading in our forum that some financial advisors set up sophisticated Monte-Carlo simulations in search of a perfectly "safe withdrawal rate" to extract constant inflation-adjusted amounts out of a complex factor-based portfolio that includes derivatives and alternative assets while keeping the chances of bankrupting the client relatively low, something like 5%. Does this mean that 1 out of 20 clients ends up bankrupt? I think not, because I'm sure that a client is disqualified from service (e.g. stops being a client) as soon as his assets fall below a high-enough threshold.
Bogleheads investment philosophy | single-ETF balanced portfolio | VBAL

User avatar
Leif
Posts: 2653
Joined: Wed Sep 19, 2007 4:15 pm

Re: Variable Percentage Withdrawal (VPW)

Post by Leif » Mon Jul 01, 2019 10:35 am

longinvest wrote:
Sun Jun 30, 2019 12:51 pm

It would be quite difficult to handle a time-limited pension in the VPW Worksheet.

I think that your current approach works.

I think that 4 pensions is sufficient for one person. Two spouses can use two separate worksheets, one per spouse.
OK. So the retirement worksheet was designed for a single person. Now I see why only a single age is entered. So if married how do I split the assets between the two worksheets and then merge the results for a couple?

Thanks.

Topic Author
longinvest
Posts: 3809
Joined: Sat Aug 11, 2012 8:44 am

Re: Variable Percentage Withdrawal (VPW)

Post by longinvest » Mon Jul 01, 2019 10:46 am

Leif wrote:
Mon Jul 01, 2019 10:35 am
longinvest wrote:
Sun Jun 30, 2019 12:51 pm

It would be quite difficult to handle a time-limited pension in the VPW Worksheet.

I think that your current approach works.

I think that 4 pensions is sufficient for one person. Two spouses can use two separate worksheets, one per spouse.
OK. So the retirement worksheet was designed for a single person. So if married how do I split the assets then merge the results for a couple?
The Retirement sheet was designed to be simple and fit on a single screen. Adding many options and entries to it would make it complex and, more importantly, it wouldn't fit on a single screen anymore.

The VPW Worksheet can often work for a couple, when similar pensions are combined into a single entry. For example if both spouses are receiving a fixed work pension, both pensions can be combined into a single entry.

For the few cases where this doesn't work, it's easy enough to split the portfolio balance in half and use two separate worksheets, but be careful to always use the youngest spouse's age in both worksheets. This might cause the delayed Social Security pension of the older spouse to start at age 72, for example.
Bogleheads investment philosophy | single-ETF balanced portfolio | VBAL

Topic Author
longinvest
Posts: 3809
Joined: Sat Aug 11, 2012 8:44 am

Re: Variable Percentage Withdrawal (VPW)

Post by longinvest » Tue Jul 02, 2019 6:41 am

In a previous post, I've explained how to combine monthly VPW withdrawals with monthly silos (12 savings accounts named January to December) to significantly dampen short-term income fluctuations and increase short-term income predictability.

But, money is fungible. I intend to build an easy-to-use monthly silos spreadsheet that automatically calculates individual silo balances and distributes interest across them, so that the money is actually kept into a single high-interest savings account. Furthermore, the spreadsheet would suggest a single money transfer per month; either a withdrawal from the savings account or a deposit into the savings account to adjust the income provided by the last monthly VPW withdrawal.

The spreadsheet would require its user to enter the last monthly VPW withdrawal amount and the amount of interest received, would make its calculations, and would inform the user of:
  1. The amount to withdraw from the savings account or the amount to deposit into the savings account, depending on whether the current-month silo balance is higher or lower than the last monthly VPW withdrawal.
  2. The balance of the 12 monthly silos.
I'm planning to force the user to confirm the balance of the savings account (pre-withdrawal or pre-deposit), once amount of interest received is entered, before displaying the amount to withdraw or deposit, to detect erroneous data entry.

What do you think?
Bogleheads investment philosophy | single-ETF balanced portfolio | VBAL

Bnjneer
Posts: 25
Joined: Fri Feb 23, 2018 10:33 am

Re: Variable Percentage Withdrawal (VPW)

Post by Bnjneer » Tue Jul 02, 2019 7:58 am

I would like to see your new spreadsheet with the Silos. Thank you.

User avatar
2pedals
Posts: 869
Joined: Wed Dec 31, 2014 12:31 pm

Re: Variable Percentage Withdrawal (VPW)

Post by 2pedals » Tue Jul 02, 2019 10:09 am

I have a basic question about the social security entries. I am 59 and planning my retirement. We plan on taking SS for DW at 62 and DH at 70. Do I enter the quoted social security monthly estimated benefits from our social social security estimates at ages 62 and 70 respectively (i.e. do not use FRA amounts)?

Thank you, longinvest I really appreciate the VPW spreadsheet, grateful for the simple things.

User avatar
Leif
Posts: 2653
Joined: Wed Sep 19, 2007 4:15 pm

Re: Variable Percentage Withdrawal (VPW)

Post by Leif » Tue Jul 02, 2019 11:10 am

longinvest wrote:
Mon Jul 01, 2019 10:46 am
For the few cases where this doesn't work, it's easy enough to split the portfolio balance in half and use two separate worksheets, but be careful to always use the youngest spouse's age in both worksheets. This might cause the delayed Social Security pension of the older spouse to start at age 72, for example.
Thanks longinvest.

I'm trying to use a single spreadsheet. I just need to ignore a very small pension (I'm using all four slots).

With only one age I was using my age (2 years older), then for the pension info I adjusted my spouse's age to +2 to match my age at the time of the event.

So, you are saying I should use the younger age (my spouse) and adjust my age (-2) to match my younger spouse's age at the time of the event? Is that to make the results a bit more conservative? I'm fine with that if that is your logic.

Is age as of January 1 of the year?

Request - Could you open up some cells on the retirement spreadsheet that could be used for comments?

Topic Author
longinvest
Posts: 3809
Joined: Sat Aug 11, 2012 8:44 am

Re: Variable Percentage Withdrawal (VPW)

Post by longinvest » Tue Jul 02, 2019 3:33 pm

2pedals wrote:
Tue Jul 02, 2019 10:09 am
I have a basic question about the social security entries. I am 59 and planning my retirement. We plan on taking SS for DW at 62 and DH at 70. Do I enter the quoted social security monthly estimated benefits from our social social security estimates at ages 62 and 70 respectively (i.e. do not use FRA amounts)?

Thank you, longinvest I really appreciate the VPW spreadsheet, grateful for the simple things.
2pedals, I suggest that you download Neurosphere's Social Security Estimator and enter your historical and projected future income to properly estimate Social Security (SS) payments at ages 62 and 70.

I think that the SS site assumes that one will work until the claim age, giving an overstated estimate of future payments for someone delaying SS payments to age 70 while being retired.
Bogleheads investment philosophy | single-ETF balanced portfolio | VBAL

Topic Author
longinvest
Posts: 3809
Joined: Sat Aug 11, 2012 8:44 am

Re: Variable Percentage Withdrawal (VPW)

Post by longinvest » Tue Jul 02, 2019 3:39 pm

Leif wrote:
Tue Jul 02, 2019 11:10 am
longinvest wrote:
Mon Jul 01, 2019 10:46 am
For the few cases where this doesn't work, it's easy enough to split the portfolio balance in half and use two separate worksheets, but be careful to always use the youngest spouse's age in both worksheets. This might cause the delayed Social Security pension of the older spouse to start at age 72, for example.
Thanks longinvest.

I'm trying to use a single spreadsheet. I just need to ignore a very small pension (I'm using all four slots).

With only one age I was using my age (2 years older), then for the pension info I adjusted my spouse's age to +2 to match my age at the time of the event.

So, you are saying I should use the younger age (my spouse) and adjust my age (-2) to match my younger spouse's age at the time of the event? Is that to make the results a bit more conservative? I'm fine with that if that is your logic.

Is age as of January 1 of the year?

Request - Could you open up some cells on the retirement spreadsheet that could be used for comments?
I suggest to express everything, in the spreadsheet, based on your younger spouse's age (2 years younger). As you're delaying Social Security (SS) to age 70, your spouse will be 68 when you start collecting it. That's the "Leif SS" Start Age I would enter into the spreadsheet. (The 72 was wrong in my previous post).

Note that, in the context of a couple, it's always important to consider what will happen after the passing of one spouse. In particular, one of the two SS payments will disappear. It's important to make sure that the projected residual retirement income will be sufficient for the survivor.
Bogleheads investment philosophy | single-ETF balanced portfolio | VBAL

Metx
Posts: 19
Joined: Fri Apr 05, 2019 8:08 pm
Location: WA

Re: Variable Percentage Withdrawal (VPW)

Post by Metx » Wed Jul 03, 2019 10:01 pm

longinvest wrote:
Tue Jul 02, 2019 6:41 am
... I intend to build an easy-to-use monthly silos spreadsheet that automatically calculates individual silo balances and distributes interest across them, so that the money is actually kept into a single high-interest savings account. Furthermore, the spreadsheet would suggest a single money transfer per month; either a withdrawal from the savings account or a deposit into the savings account to adjust the income provided by the last monthly VPW withdrawal.
...
What do you think?
That sounds great, it would make the monthly silo approach feel more practical.
Thanks in general for your contributions, for me they upgrade VPW from an interesting concept to something actionable.

Topic Author
longinvest
Posts: 3809
Joined: Sat Aug 11, 2012 8:44 am

Re: Variable Percentage Withdrawal (VPW)

Post by longinvest » Thu Jul 04, 2019 9:04 am

OK. Silos will be replaced with a simpler process.

As reference, I'm copying a few posts discussing the difference between silos and using a simple ratio from another thread.
D-Dog wrote:
Tue Jul 02, 2019 3:06 pm
It seems like there is a simpler way to accomplish what you are trying to accomplish with the silos. Why not just prefund the bank account with N (you choose N) months of withdrawals at the beginning. Then each month add the normal VPW monthly withdrawal to the bank account and divide the resulting balance by (N+1). That is then your monthly income. Do the same thing every month. The larger you choose N the more the VPW withdrawals are smoothed. It’s not exactly the same as the silos but seems a lot easier.
AlohaJoe wrote:
Thu Jul 04, 2019 3:52 am
Here's a chart showing the difference in the two approaches over a randomly chosen 4-year period in the 1950s.

Image

I'll leave it to the reader to guess which line is which approach.
longinvest wrote:
Thu Jul 04, 2019 8:19 am
It's mathematically obvious that silos and a simple ratio will have relatively similar results. But, silos don't overweight the most recent withdrawal relative to previous withdrawals; silos only attribute 1/12th of the last withdrawal to income.

When I backest silos versus a simple 1/6.5 ratio on a 60/40 stocks/bonds portfolio with a 60/40 weighting of US/International stocks (using Portfolio Visualizer CASHX returns for cash) over the drawdown period from November 2007 to March 2009, I get:

Image

Note that I've started the left axis at $2,750 instead of $0. What we see is that silos were less reactive, in the short term, to the drawdown. The difference isn't huge; it's 50$ in March 2009, at the bottom of the drawdown.

Here's a similar comparison for the drawdown period of February 2018 to January 2019:

Image

But, to be fair, when I look at the overall November 2007 to January 2019 period, properly starting the left axis at $0, I get:

Image
Bogleheads investment philosophy | single-ETF balanced portfolio | VBAL

Topic Author
longinvest
Posts: 3809
Joined: Sat Aug 11, 2012 8:44 am

Re: Variable Percentage Withdrawal (VPW)

Post by longinvest » Thu Jul 04, 2019 9:06 am

Here's the proposed simpler monthly process (based on the proposition of forum member D-Dog):
longinvest wrote:
Thu Jul 04, 2019 8:19 am
I have to admit that the 1/6.5 ratio is simpler than silos and good enough.

OK, unless there's a big objection, I pronounce monthly silos dead.

To remove awkward ratios, I would simplify the monthly process further to:
  1. An initial withdrawal equal to half an annual VPW withdrawal is taken from the portfolio.
  2. A 1/6 ratio of the initial withdrawal is taken as income (for taxes and expenses).
  3. The remaining money from the initial withdrawal is put into a savings account.
  4. Each subsequent month:
    1. A withdrawal equal to 1/12 an annual VPW withdrawal is taken from the portfolio.
    2. An amount equal to 1/6 of the sum of the withdrawal and savings account is taken as income (for taxes and expenses):
      1. If income is greater than the withdrawal, missing money is withdrawn from the savings account.
      2. If income is smaller than the withdrawal, excess withdrawal money is deposited into the savings account.
This simpler process makes it possible to do monthly VPW withdrawals, with dampening savings account, using a simple sheet of paper, a calculator, and the VPW Table.

For example, if the retiree's age is 67, the 60/40 stocks/bonds portfolio balance is $943,738, and the savings account balance is $20,722, I get:
  1. The percentage in the VPW Table is multiplied by the portfolio balance and divided by 12. The current month portfolio withdrawal is ((5.1% X $943,738) / 12) = $4,011.
  2. Monthly income, for taxes and expenses, is equal to (($4,011 + $20,722) / 6) = $4,122. As a consequence:
    • The missing ($4,122 - $4,011) = $111 is withdrawn from the savings account.
What do you think?
Effectively, what do you think?
Last edited by longinvest on Thu Jul 04, 2019 12:00 pm, edited 2 times in total.
Bogleheads investment philosophy | single-ETF balanced portfolio | VBAL

User avatar
2pedals
Posts: 869
Joined: Wed Dec 31, 2014 12:31 pm

Re: Variable Percentage Withdrawal (VPW)

Post by 2pedals » Thu Jul 04, 2019 9:57 am

longinvest wrote:
Thu Jul 04, 2019 9:06 am
Here's the proposed simpler monthly process (based on the proposition of forum member D-Dog):
longinvest wrote:
Thu Jul 04, 2019 8:19 am
I have to admit that the 1/6.5 ratio is simpler than silos and good enough.

OK, unless there's a big objection, I pronounce monthly silos dead.

To remove awkward ratios, I would simplify the monthly process further to:
  1. An initial withdrawal equal to half an annual VPW withdrawal is taken from the portfolio.
  2. A 1/6 ratio of the initial withdrawal is taken as income (for taxes and expenses).
  3. The remaining money from the initial withdrawal is put into a savings account.
  4. Each subsequent month:
    1. A withdrawal equal to 1/12 an annual VPW withdrawal is taken from the portfolio.
    2. An amount equal to 1/6 of the sum of the withdrawal and savings account is taken as income (for taxes and expenses).
    3. If income is greater than the withdrawal, missing money is withdrawn from the savings account.
    4. If income is smaller than the withdrawal, excess withdrawal money is deposited into the savings account.
This simpler process makes it possible to do monthly VPW withdrawals, with dampening savings account, using a simple sheet of paper, a calculator, and the VPW Table.

For example, if the retiree's age is 67, the 60/40 stocks/bonds portfolio balance is $943,738, and the savings account balance is $20,722, I get:
  1. The percentage in the VPW Table is multiplied by the portfolio balance and divided by 12. The current month portfolio withdrawal is ((5.1% X $943,738) / 12) = $4,011.
  2. Monthly income, for taxes and expenses, is equal to (($4,011 + $20,722) / 6) = $4,122.
  3. The missing ($4,122 - $4,011) = $111 is withdrawn from the savings account.
What do you think?
Effectively, what do you think?
Perhaps you don't need a savings account depending on your needs or wants. Just a withdrawal calculation and then a withdrawal based on VPW and a virtual dampening account you mention.

Topic Author
longinvest
Posts: 3809
Joined: Sat Aug 11, 2012 8:44 am

Re: Variable Percentage Withdrawal (VPW)

Post by longinvest » Thu Jul 04, 2019 10:24 am

2pedals wrote:
Thu Jul 04, 2019 9:57 am
Perhaps you don't need a savings account depending on your needs or wants. Just a withdrawal calculation and then a withdrawal based on VPW and a virtual dampening account you mention.
Once money has been destroyed by markets, it's gone! It makes no sense to withdraw money from a portfolio based (in part) on a historical value one year ago. Of course, as long as both the loss and the withdrawal percentage are small enough, the harm to the portfolio will be small and won't make a big difference on backtesting charts.

But I see no reason to do this when a simpler approach that involves no remembering of past withdrawal amounts or portfolio values can dampen short-term fluctuations safely (without harming the portfolio) using a small savings account, while collecting some interest to fight inflation.
Bogleheads investment philosophy | single-ETF balanced portfolio | VBAL

KarenC
Posts: 105
Joined: Mon Apr 27, 2015 7:25 am

Re: Variable Percentage Withdrawal (VPW)

Post by KarenC » Thu Jul 04, 2019 10:33 am

longinvest wrote:
Thu Jul 04, 2019 9:06 am
[…] Effectively, what do you think?
I think I would actually use this approach. (In contrast, my gut reaction to the silo approach was a hard “No.”)
"How much you know is less important than how clearly you understand where the borders of your ignorance begin." — Jason Zweig

Topic Author
longinvest
Posts: 3809
Joined: Sat Aug 11, 2012 8:44 am

Re: Variable Percentage Withdrawal (VPW)

Post by longinvest » Thu Jul 04, 2019 10:44 am

I've redone the simulation I had made for silos, but with a dampening savings account instead. Here are the previous average-inflation-adjusted silo income chart and new average-inflation-adjusted savings buffer income chart:
longinvest wrote:
Sat Jun 29, 2019 11:58 pm
Inflation-Adjusted Silo Income

So, here it is, finally, the chart of inflation-adjusted silo income derived from monthly VPW withdrawals at age 65 from a 50/50 stocks/bonds $1,000,000 portfolio (without taking into account any non-portfolio income), using 12-months-average inflation:

Image
Image

Silos do a somewhat better job at dampening fluctuations, but, as I wrote earlier, I'll concede that the savings buffer is good enough and much simpler.
Bogleheads investment philosophy | single-ETF balanced portfolio | VBAL

AlohaJoe
Posts: 4711
Joined: Mon Nov 26, 2007 2:00 pm
Location: Saigon, Vietnam

Re: Variable Percentage Withdrawal (VPW)

Post by AlohaJoe » Thu Jul 04, 2019 11:14 am

longinvest wrote:
Thu Jul 04, 2019 10:44 am
Silos do a somewhat better job at dampening fluctuations, but, as I wrote earlier, I'll concede that the savings buffer is good enough and simpler.
While I have a soft spot for spreadsheet-driven mathematical purity, I think this approach is actually better for one reason. One of the benefits of VPW over some alternatives (say, Guyton-Klinger's guardrails) is that it is plausible a spouse who isn't excited about finances could run it after you pass away. Having a simple monthly withdrawal methodology fits in with that as well. I can envision telling my spouse to withdraw VPW/12 every month and then spend bankaccount/6 every month.

A conceptually simpler approach also makes it easier for tinkerers to tinker. Maybe someone thinks 6 months is too much cash drag and 3 months is enough. Or maybe someone can't imagine not having 12 months of withdrawals already sitting in cash. Or maybe someone just got the equivalent 3-months of income from small inheritance and didn't see a reason to reinvest it just to withdraw in 60-90 days later. So they divide by 9 instead of 6 for a few months. Or some mutual fund they hold in taxable has a large distribution for some reason. Etc, etc.

User avatar
2pedals
Posts: 869
Joined: Wed Dec 31, 2014 12:31 pm

Re: Variable Percentage Withdrawal (VPW)

Post by 2pedals » Thu Jul 04, 2019 11:46 am

longinvest wrote:
Thu Jul 04, 2019 9:06 am
Here's the proposed simpler monthly process (based on the proposition of forum member D-Dog):
longinvest wrote:
Thu Jul 04, 2019 8:19 am
I have to admit that the 1/6.5 ratio is simpler than silos and good enough.

OK, unless there's a big objection, I pronounce monthly silos dead.

To remove awkward ratios, I would simplify the monthly process further to:
  1. An initial withdrawal equal to half an annual VPW withdrawal is taken from the portfolio.
  2. A 1/6 ratio of the initial withdrawal is taken as income (for taxes and expenses).
  3. The remaining money from the initial withdrawal is put into a savings account.
  4. Each subsequent month:
    1. A withdrawal equal to 1/12 an annual VPW withdrawal is taken from the portfolio.
    2. An amount equal to 1/6 of the sum of the withdrawal and savings account is taken as income (for taxes and expenses).
    3. If income is greater than the withdrawal, missing money is withdrawn from the savings account.
    4. If income is smaller than the withdrawal, excess withdrawal money is deposited into the savings account.
This simpler process makes it possible to do monthly VPW withdrawals, with dampening savings account, using a simple sheet of paper, a calculator, and the VPW Table.

For example, if the retiree's age is 67, the 60/40 stocks/bonds portfolio balance is $943,738, and the savings account balance is $20,722, I get:
  1. The percentage in the VPW Table is multiplied by the portfolio balance and divided by 12. The current month portfolio withdrawal is ((5.1% X $943,738) / 12) = $4,011.
  2. Monthly income, for taxes and expenses, is equal to (($4,011 + $20,722) / 6) = $4,122.
  3. The missing ($4,122 - $4,011) = $111 is withdrawn from the savings account.
What do you think?
Effectively, what do you think?
suggestion -- let the savings account do the math and eliminate steps 3 and 4:
  1. An initial withdrawal equal to 7 months an annual VPW withdrawal is taken from the portfolio.
  2. 1/7 (1 month annual basis) of initial withdrawal is taken as income (for taxes and expenses).
  3. The remaining money from the initial withdrawal is put into a savings account.
  4. Each subsequent month:
    1. An amount equal to 1/6 of the sum of the withdrawal and savings account is taken as income (for taxes and expenses). (Note 1-2 sequence reversed)
    2. A withdrawal equal to 1/12 an annual VPW withdrawal is taken from the portfolio and deposited into the savings account.
    3. If income is greater than the withdrawal, missing money is withdrawn from the savings account.
    4. If income is smaller than the withdrawal, excess withdrawal money is deposited into the savings account.
Last edited by 2pedals on Thu Jul 04, 2019 4:06 pm, edited 1 time in total.

Topic Author
longinvest
Posts: 3809
Joined: Sat Aug 11, 2012 8:44 am

Re: Variable Percentage Withdrawal (VPW)

Post by longinvest » Thu Jul 04, 2019 3:06 pm

2pedals wrote:
Thu Jul 04, 2019 11:46 am
suggestion ...
OK, I've eliminated (e.g. rephrased) steps 3 and 4:
longinvest wrote:
Thu Jul 04, 2019 8:19 am
  1. An initial withdrawal equal to half an annual VPW withdrawal is taken from the portfolio.
  2. A 1/6 ratio of the initial withdrawal is taken as income (for taxes and expenses).
  3. The remaining money from the initial withdrawal is put into a savings account.
  4. Each subsequent month:
    1. A withdrawal equal to 1/12 an annual VPW withdrawal is taken from the portfolio.
    2. An amount equal to 1/6 of the sum of the withdrawal and savings account is taken as income (for taxes and expenses):
      1. If income is greater than the withdrawal, missing money is withdrawn from the savings account.
      2. If income is smaller than the withdrawal, excess withdrawal money is deposited into the savings account.
I think that this is good enough for now. It's what I'll do in the forward test.
Bogleheads investment philosophy | single-ETF balanced portfolio | VBAL

Bnjneer
Posts: 25
Joined: Fri Feb 23, 2018 10:33 am

Re: Variable Percentage Withdrawal (VPW)

Post by Bnjneer » Sat Jul 06, 2019 4:32 pm

Can you please point me to where I can see how the pensions are handled (e.g. Mathematically, how do the pensions affect the withdrawal rate?)

Thank you.

Topic Author
longinvest
Posts: 3809
Joined: Sat Aug 11, 2012 8:44 am

Re: Variable Percentage Withdrawal (VPW)

Post by longinvest » Sat Jul 06, 2019 6:07 pm

Bnjneer wrote:
Sat Jul 06, 2019 4:32 pm
Can you please point me to where I can see how the pensions are handled (e.g. Mathematically, how do the pensions affect the withdrawal rate?)
Bnjneer,

I'll use the example of the forward test described in this post, which involves two pensions, a delayed one and a fixed one.

If you download the new VPW Accumulation And Retirement Worksheet, you'll be able to enter the data of the forward test and look at the "Calculations" section below the main screen of the Retirement sheet.

Mainly, the VPW worksheet is doing three calculations: a calculation for the delayed $2,000 Social Security pension, a calculation for the fixed $1,000 pension, and a VPW withdrawal calculation. Let's do that:

Social Security

The retiree will get $2,000/month in 5 years. That's $24,000/year. The VPW Table tells us that the percentage for a 5-year withdrawal with a 60/40 stocks/bonds scenario (e.g. age 95) is 21.5%.

So, we need to put aside (on paper) ($24,000 / 21.5%) = $111,628 for Social Security bridge withdrawals.

Fixed Pension

The $1,000/month work pension isn't indexed to inflation. To dampen the erosion of inflation, only 65.7% of the pension is spent and the rest (34.3%) is invested into the portfolio. That's ($1,000 X 12 X 34.3%) = $4,116/year.

If you want to understand how the 65.7% was determined, I've explained the calculation process in this post. The simpler formula used in the spreadsheet was derived by forum member #Cruncher and is stated in this post.

VPW Withdrawal

The retiree has a $1,000,000 portfolio, but we've put aside (on paper) $111,628 as bridge for Social Security. So, we're left with $888,372 for VPW withdrawals. At age 65 with a 60/40 stocks/bonds portfolio, the percentage in the VPW Table is 5.0%. This gives us ($888,372 X 5.0%) = $44,419.

Putting It All Together

On an annual basis, the retiree plans to withdraw $24,000 in replacement of future Social Security payments, to invest $4,116 to dampen the ravages of inflation on his fixed work pension, and to take a $44,419 VPW withdrawal. This sums up to ($24,000 - $4,116 + $44,419) = $64,303. On a monthly basis, this is ($64,303 / 12) = $5,358.

The small $2 difference with the VPW Worksheet's suggested $5,356 is due to rounding.
Bogleheads investment philosophy | single-ETF balanced portfolio | VBAL

Bnjneer
Posts: 25
Joined: Fri Feb 23, 2018 10:33 am

Re: Variable Percentage Withdrawal (VPW)

Post by Bnjneer » Sat Jul 06, 2019 7:43 pm

Awesome explanation by example. Thank you so much. I imagine this will be asked by others in the future. Do you think it would be worthwhile adding this on a tab in the spreadsheet for documentation?

You did a very nice job on this spreadsheet and I look forward to following the post with regards to this spreadsheet viewtopic.php?p=4627254#p4627254

Thank you

Topic Author
longinvest
Posts: 3809
Joined: Sat Aug 11, 2012 8:44 am

Re: Variable Percentage Withdrawal (VPW)

Post by longinvest » Fri Jul 26, 2019 6:59 am

🎂 🎉 🎈 🎉 🎈 🎉 🎂

This thread was started exactly six years ago on Friday July 26, 2013. The development of VPW that resulted, among other things, into the creation of the VPW Table, the VPW Backtesting Spreadsheet that includes historical US data (1871-2018) and Canadian data (1970-2018), and the easy-to-use VPW Accumulation And Retirement Worksheet that covers both accumulation and retirement phases while taking into account future and current pensions (with and without cost-of-living adjustments) wouldn't have been possible without the generous help, suggestions, comments, and feedback of many members of the Bogleheads forum (and of its Canadian sister FWF). Thank you!

🎂 🎉 🎈 🎉 🎈 🎉 🎂
Bogleheads investment philosophy | single-ETF balanced portfolio | VBAL

User avatar
LadyGeek
Site Admin
Posts: 56321
Joined: Sat Dec 20, 2008 5:34 pm
Location: Philadelphia
Contact:

Re: Variable Percentage Withdrawal (VPW)

Post by LadyGeek » Sat Jul 27, 2019 11:23 am

This thread is now in the Investing - Theory, News & General forum (theory).

By developer request (longinvest), the user base is now large enough to support a dedicated discussion for "theory of operation" (how the spreadsheet works) and a dedicated discussion for support (I need help using the spreadsheet).

This thread will be used for discussion of the spreadsheet internals (how the spreadsheet works).

longinvest will start a new thread for support shortly.
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
willthrill81
Posts: 12605
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: Variable Percentage Withdrawal (VPW)

Post by willthrill81 » Sat Jul 27, 2019 11:51 pm

LadyGeek wrote:
Sat Jul 27, 2019 11:23 am
This thread is now in the Investing - Theory, News & General forum (theory).
In the past, threads explicitly focusing on withdrawal strategies were moved by the mods to the 'personal finance' forum. Is that no longer the place for discussion of that topic? I'd just like some clarification. Thanks.
“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

dbr
Posts: 30079
Joined: Sun Mar 04, 2007 9:50 am

Re: Variable Percentage Withdrawal (VPW)

Post by dbr » Sun Jul 28, 2019 12:36 pm

willthrill81 wrote:
Sat Jul 27, 2019 11:51 pm
LadyGeek wrote:
Sat Jul 27, 2019 11:23 am
This thread is now in the Investing - Theory, News & General forum (theory).
In the past, threads explicitly focusing on withdrawal strategies were moved by the mods to the 'personal finance' forum. Is that no longer the place for discussion of that topic? I'd just like some clarification. Thanks.
I am a little puzzled too. What this actually amounts to is a classic case of needing a (actually two) stickies. We have a couple of those and there has been great reluctance to add more. VPW is in the Wiki, but the Wiki is not designed for discussion, so there is a bit of a disconnect there as well. If there are going to be two ongoing threads on theory and on operation of a tool, should that Wiki entry also be maintained.

It could also be this forum is not the best site to host a dedicated tool. I don't think we host any other tools that are as specific as this one, but maybe we do.

User avatar
SevenBridgesRoad
Posts: 444
Joined: Sat Jul 07, 2018 12:14 am

Re: Variable Percentage Withdrawal (VPW)

Post by SevenBridgesRoad » Mon Jul 29, 2019 12:00 am

longinvest wrote:
Fri Jul 26, 2019 6:59 am
🎂 🎉 🎈 🎉 🎈 🎉 🎂

This thread was started exactly six years ago on Friday July 26, 2013.

🎂 🎉 🎈 🎉 🎈 🎉 🎂
Happy birthday, VPW thread! Thank you longinvest. My wife and I are using the VPW method and using the Retirement Worksheet. Know that your work is being used in real life by real life retirees. Appreciated.
Retired 2018 age 61 | 30:70 VTINX with two deferred single premium annuities and deferred SS | VPW Method | Sleeping very well at night

Alex Frakt
Founder
Posts: 10861
Joined: Fri Feb 23, 2007 1:06 pm
Location: Chicago
Contact:

Re: Variable Percentage Withdrawal (VPW)

Post by Alex Frakt » Fri Aug 02, 2019 8:16 pm

Returned to Investing - Theory, News and General forum. This is quite obviously an investing theory discussion.

User avatar
Harry Livermore
Posts: 63
Joined: Thu Apr 04, 2019 5:32 am

Re: Variable Percentage Withdrawal (VPW)

Post by Harry Livermore » Tue Sep 10, 2019 5:12 am

longinvest,
Thanks so much for creating this tool. I have been following a couple of threads about VPW over the last couple of days, and one user was struggling with a calculation error because they entered "YES" or "NO" incorrectly on the "already started?" cell. It seems like asking for a start date AND an "already started?" question is redundant (and in that particular user's case, counterproductive)
Why not lock the cells that say "YES" or "NO" to the "pension started?" question, and have it return a "YES" or "NO" automatically based on the start age and current age entries?
Unless there is some change to the resultant withdrawal amount? Does the formula assume higher benefits accrue in future years if one has not started the pension but is eligible? That would be the case for SS, and many pensions, but not all pensions.
Thanks again. What a great resource for Bogleheads.
Cheers

Topic Author
longinvest
Posts: 3809
Joined: Sat Aug 11, 2012 8:44 am

Re: Variable Percentage Withdrawal (VPW)

Post by longinvest » Tue Sep 10, 2019 7:15 am

Harry Livermore wrote:
Tue Sep 10, 2019 5:12 am
Why not lock the cells that say "YES" or "NO" to the "pension started?" question, and have it return a "YES" or "NO" automatically based on the start age and current age entries?
Thanks for the suggestion. I had initially considered automatically detecting whether the pension is delayed or not, but I decided otherwise. Here's why.

I've put this redundant check in place because an erroneous "Start Age" entry could have a significant impact on the suggested withdrawal amount. I think that it's important to request that the user confirms whether the pension is started or not. I considered the consequences of annoying the user with an ERROR message in case of incoherent input, or providing an inappropriate withdrawal amount suggestion when the user forgets to change or update the "Start Age". I concluded that the annoyance was milder than providing an inappropriate withdrawal amount.

Most users, especially new ones, don't pay much attention to an automatically calculated secondary cell. I might as well completely remove the "Already Started" entry, if I was to automatically calculate it.

It's about protecting the user. Take the example of a user who has just retired at age 66 and anticipates to start receiving a pension 6 months later at age 66 and a half. If there wasn't an "Already Started" cell and the "Start Age" was set to "66", the spreadsheet would automatically assume that the pension was already started, which could lead to a significantly lower withdrawal suggestion. By explicitly requesting an "Already Started" entry, it forces the user to realize that "Start Age" should be set to 67. This wouldn't prevent an incorrect setting of "67" for a pension delayed to age 67.5, but the consequences would be milder, in that case.

Personally, I lean towards keeping the current design. What do you and others think?
Bogleheads investment philosophy | single-ETF balanced portfolio | VBAL

User avatar
backofbeyond
Posts: 242
Joined: Tue Apr 01, 2008 1:07 pm

Re: Variable Percentage Withdrawal (VPW)

Post by backofbeyond » Tue Sep 10, 2019 11:50 am

I was watching a youtube video called Techniques and Strategies for Creating a Retirement Income You Can Live With. It featured Prof Wade Pfau along with several other experts. The intended audience is other financial planners. But one of the discussion points that Prof Pfau mentioned was VPW (he didn't call it that, but it's the same idea). And he said something to the effect that it was the most sound economic withdrawal strategy.

However, he also said it was a hard sell to clients as they couldn't necessarily expect to get all the income they needed each year due to market volatility.

I know this string has battled that critic around with regards to floors etc, so I'm not trying to bring that up again. I only mention it because he seemed to think it was the most logical strategy out there.

And yes, I plan on VPWing when I retire.
The question isn't at what age I want to retire, it is at what income. - George Foreman

User avatar
Harry Livermore
Posts: 63
Joined: Thu Apr 04, 2019 5:32 am

Re: Variable Percentage Withdrawal (VPW)

Post by Harry Livermore » Tue Sep 10, 2019 12:05 pm

Makes perfect sense to me, thanks for the clarification!
I actually think complex spreadsheets benefit from having "checks" like this anyway. The more the user thinks about the mechanics of what's being computed, the more likely they'll spot something amiss, rather than just relying on the output.
Again, what a great tool. Many thanks for working so hard on it!
Cheers

MrDrinkingWater
Posts: 119
Joined: Tue Dec 23, 2014 11:30 am

Re: Variable Percentage Withdrawal (VPW)

Post by MrDrinkingWater » Wed Sep 11, 2019 9:38 am

@longinvest: If you are taking a poll, I like your current design. I see the point of making the change, too, but I lean toward your position of protecting the user from the mistake you've cited.

azanon
Posts: 2544
Joined: Mon Nov 07, 2011 10:34 am

Re: Variable Percentage Withdrawal (VPW)

Post by azanon » Wed Sep 11, 2019 10:33 am

backofbeyond wrote:
Tue Sep 10, 2019 11:50 am
I was watching a youtube video called Techniques and Strategies for Creating a Retirement Income You Can Live With. It featured Prof Wade Pfau along with several other experts. The intended audience is other financial planners. But one of the discussion points that Prof Pfau mentioned was VPW (he didn't call it that, but it's the same idea). And he said something to the effect that it was the most sound economic withdrawal strategy.

However, he also said it was a hard sell to clients as they couldn't necessarily expect to get all the income they needed each year due to market volatility.
I would say the only reason that VPW is a hard sell to those clients is because they're not using/planning with VPW properly. VPW calls for (or is recommended to) being used with a combination of social security, pensions, and in necessary, annuities. So if they follow that step correctly, there'll be a floor of income with no volatility.

Second, they're probably trying to combine VPW with a growth portfolio (I'd define as > or = to 50% equities), and then blaming the VPW method for the variability instead of the growth portfolio. I agree with Rick, that the Center of Gravity for retirement portfolios starts at around 30% equities, and is only modified from that if a particular situation calls for it (e.g. it will be combined with large fixed income streams such as dual social securities in a marriage plus maybe a pension on top of that).

With a proper floor and 30-40% equity portfolio, the only reason that problem would still remain is because they haven't saved enough money yet to retire. In other words, at least from my POV, the problem never actually gets to the VPW method itself.

User avatar
SevenBridgesRoad
Posts: 444
Joined: Sat Jul 07, 2018 12:14 am

Re: Variable Percentage Withdrawal (VPW)

Post by SevenBridgesRoad » Wed Sep 11, 2019 11:30 am

azanon wrote:
Wed Sep 11, 2019 10:33 am
backofbeyond wrote:
Tue Sep 10, 2019 11:50 am
...However, he also said it was a hard sell to clients as they couldn't necessarily expect to get all the income they needed each year due to market volatility...
I would say the only reason that VPW is a hard sell to those clients is because they're not using/planning with VPW properly. VPW calls for (or is recommended to) being used with a combination of social security, pensions, and in necessary, annuities. So if they follow that step correctly, there'll be a floor of income with no volatility.

Second, they're probably trying to combine VPW with a growth portfolio (I'd define as > or = to 50% equities), and then blaming the VPW method for the variability instead of the growth portfolio. I agree with Rick, that the Center of Gravity for retirement portfolios starts at around 30% equities, and is only modified from that if a particular situation calls for it (e.g. it will be combined with large fixed income streams such as dual social securities in a marriage plus maybe a pension on top of that).

With a proper floor and 30-40% equity portfolio, the only reason that problem would still remain is because they haven't saved enough money yet to retire. In other words, at least from my POV, the problem never actually gets to the VPW method itself.
Good points azanon. And the Retirement Worksheet really makes this clear. There are four slots called Pension #1-4 for a reason. For a couple these could be filled with two Social Security checks and an annuity or two and/or a pension. The Annual Income After Loss section shows a worse case scenario Portfolio Withdrawal. If that number won't meet one's expenses, don't blame the method. Change the inputs.
Retired 2018 age 61 | 30:70 VTINX with two deferred single premium annuities and deferred SS | VPW Method | Sleeping very well at night

Post Reply