Variable Percentage Withdrawal (VPW)

Discuss all general (i.e. non-personal) investing questions and issues, investing news, and theory.

Variable Percentage Withdrawal (VPW)

Postby longinvest » Fri Jul 26, 2013 12:12 am

PLEASE read the FREQUENTLY ASKED QUESTIONS (F.A.Q.) section of the following post before posting your question: viewtopic.php?f=10&t=120430&p=1761580#p1761563 .


DOWNLOAD LOCATION
You can find the link to download the latest version on the Bogleheads Wiki Variable Percentage Withdrawal page.

SCREENSHOT

Image


The Variable Percentage Withdrawal method was improved and simplified (relative to this initial post) to only require the retirement age and asset allocation. If you do not wish to download the above spreadsheet, see the following post for the improved version along with generic withdrawal tables: viewtopic.php?f=10&t=120430&p=1761580#p1761563 .


Hi,

I'm seeking comments on a withdrawal method I've thought up. I've been reading the forums for a while and I've read many Safe Withdrawal Rate (SWR) threads. I've also read the Withdrawal Methods Wiki page. Yet, I'm not satisfied with the methods I've seen :
  • Constant-Dollar : I would need to make a huge leap of faith to start withdrawing 4% (or whatever %) of the my portfolio, and index it to inflation every year. (1) This is scary! I would be subject to bad sequence of return risk, and (2) there is a very high probability that I will leave much money unspent (if the bad sequence doesn't show up). I might be too old to spend it, once I realize that I have too big a pile of money.
  • Constant-Percentage : This one does not have as much sequence of return risk (I can still end up with a much lower amount to spend than I expected, if I chose too high a percentage). The problem is that I am, again, quite likely to underspend and leave a lot of money unspent.
  • Spend Only the Dividends : I would likely overspend my bonds and underspend my equities, unless, of course, I had a 50/50 portfolio and rebalanced it regularly. Yet, I would still be very likely to leave a lot of money unspent.

Here's my idea. It is a Variable Percentage Withdrawal (VPW) method. It is a mix, somehow, of the Constant-Dollar and Constant-Percentage methods. Here's an example:

Let say I have a portfolio with an expected 3% real return. I want this portfolio to survive 30 years. I have no bequest motive, so I am happy to drawdown 100% of it.

At the start of year 1 of retirement, I withdraw 4.95% of my portfolio (and take the opportunity to rebalance it).

At the start of year 2 of retirement, I withdraw 5.06% of my remaining portfolio (and rebalance it).

At the start of year 3 of retirement, I withdraw 5.17% of my remaining portfolio (and rebalance it).

And so on, until year 30, where I withdraw 100% of my remaining portfolio. (No need to rebalance!)

So, even if the famous bad sequence of return was to happen, I wouldn't run out of money before the end of 30 years. As a bonus, I am sure to have spent it all; there's no risk of having a single dime left for year 31 (unless I died before year 30, of course).

The withdrawn amount varies as an increasing percentage of portfolio balance.

The percentages, above, are computed from a constant-payment 30-year annuity at 3%:
Payment: 0.049533
Balance
Year 1: 1.000000
Year 2: 0.978981 (= (1.000000 – 0.049533) * 1.03)
Year 3: 0.957331 (= ( 0.978981 – 0.049533) * 1.03)

Year 30: 0.049533
So, we get:
Year 1: 4.95% (= 0.049533 / 1.000000)
Year 2: 5.06% (= 0.049533 / 0.978981 )

Year 30: 100%

Now, I don't know about you, but personally, if I'm not dead at the end of 30 years, I'd like to have some minimal money left. So, as a first improvement, I would cap the increasing percentage at 20% to smooth out the final portfolio decline.

Another improvement is to cap the drawdown. Let say I want to drawdown 75% of my porfolio: I would apply the VPW method on this 75%, and I would only spend the expected return (3%) of the remaining (more precisely: (3/1.03) %, so that the capital can rebuild up). This would add up to:
Year 1: 4.44%
Year 2: 4.51%
Year 3: 4.59%

Year 25: 9.72%
Year 26: 9.95%
Year 27: 9.17%
...
Year 30: 7.09 %

Year 40: 3.67%
...
Year 50: 3.03%

With the two improvements, the rate increases until year 26. Then the rate starts declining. The “expected remaining balance” slides towards 25% of the original portfolio.

I am sure somebody has probably thought about something similar before. (I was partly inspired by the Canadian Registered Retirement Income Fund (RRIF) minimum withdrawal rules, probably similar to US RMD.)

I have an excel spread sheet to compute the percentages, given 3 inputs:
  • expected real return
  • number of years
  • Drawdown percentage
I would be more than happy to share it, but I don't know how to attach it to my post.

So, does VPW make sense?
Last edited by longinvest on Sun Feb 02, 2014 10:59 am, edited 16 times in total.
longinvest
 
Posts: 642
Joined: Sat Aug 11, 2012 9:44 am

Re: Variable Percentage Withdrawal

Postby bertilak » Fri Jul 26, 2013 7:25 am

longinvest wrote:So, does VPW make sense?

Under two related conditions:
  1. You have a variable NEED for income, and
  2. That need varies IN SYNC with the output from your VPW plan.
No-one really listens to anyone else, and if you try it sometime you will see why.
-- Mignon McLaughlin
User avatar
bertilak
 
Posts: 3094
Joined: Tue Aug 02, 2011 6:23 pm
Location: East of the Pecos, West of the Mississippi

Re: Variable Percentage Withdrawal

Postby JoMoney » Fri Jul 26, 2013 7:46 am

I'm a fan of the x/y method, where x=balance y=time you want the money to last.
For retirement purposes, this would be the same as the IRS Required Minimum Distribution (RMD) on an IRA, so "y" = the number of years you have remaining. The number would be based on an actuarial table that changes (i.e. the older you get, the longer you're expected to live).

There's no guarantees on how much your assets will be worth in the future, this method is very safe in making sure it never runs out and gives you a proportional withdrawal rate based on what you know today.
User avatar
JoMoney
 
Posts: 1272
Joined: Tue Jul 23, 2013 6:31 am

Re: Variable Percentage Withdrawal

Postby Cut-Throat » Fri Jul 26, 2013 8:00 am

I am using a VPW, but similar to one that was outlined in Bob's Financial Pages (Not sure if you can get at them anymore).

I take 3% of remaining portfolio balance until Age 70, Increase to 4% at age 75, Increase to 5% at age 80, increase to 6% at age 85.....re-evaluate at each 5 year interval.
According to Jim Otar's retirement book, a 3% withdrawal rate will never deplete the portfolio. Not sure I believe this, but if it is true, then a 3% of remaining portfolio should give an increasing amount to spend each year. (On average). We'll see how this works out. :sharebeer

If is far simpler than the one you outlined. More conservative also. It also avoids the damage that an early sequence of very bad market returns does to a portfolio. It makes you cut back spending if your portfolio shrinks.
User avatar
Cut-Throat
 
Posts: 1890
Joined: Sun Oct 17, 2010 10:46 am

Re: Variable Percentage Withdrawal

Postby siamond » Fri Jul 26, 2013 10:29 am

I guess you have to think about your personal objectives. Are you so sure that you'd like to have less money to begin with, and more money as you get older? Including when you get real old and rather frail?

Personally, I know I'd like to have a tad more money in my early days of retirement, as I guess I will be more active by then, but then my level of activity might decrease. That being said, my medical expenses will probably increase, so this is a bit of a tough call.

I have looking at the Guyton/Klinger variable method, and although this sounds a tad complicated (albeit easy to automate with a spreadsheet) and may require a bit of customization (e.g. define floor/ceiling on my spend to avoid silly situations), this makes a lot of sense to me.
http://schulmerichandassoc.homestead.com/Using_Decision_Rules_to_Create_Retirement_Withdrawal_Profiles.pdf
siamond
 
Posts: 522
Joined: Mon May 28, 2012 6:50 am

Re: Variable Percentage Withdrawal

Postby Cut-Throat » Fri Jul 26, 2013 10:39 am

siamond wrote: Are you so sure that you'd like to have less money to begin with, and more money as you get older? Including when you get real old and rather frail?

Personally, I know I'd like to have a tad more money in my early days of retirement, as I guess I will be more active by then, but then my level of activity might decrease. That being said, my medical expenses will probably increase, so this is a bit of a tough call.


Not sure who you are referring to, but I don't see why the methods explained here would have less money early in retirement and more as you get older.
User avatar
Cut-Throat
 
Posts: 1890
Joined: Sun Oct 17, 2010 10:46 am

Re: Variable Percentage Withdrawal

Postby longinvest » Fri Jul 26, 2013 11:13 am

bertilak wrote:
longinvest wrote:So, does VPW make sense?

Under two related conditions:
  1. You have a variable NEED for income, and
  2. That need varies IN SYNC with the output from your VPW plan.


bertilac, what is your solution?

I think that the two conditions should rather be:
  • I have a flexible spending plan, and
  • I am able to vary my spending according to the output from my VPW plan.
This implies that, in normal years, the provided income covers more than my strict minimum NEEDS, otherwise I should delay retirement until I have a cushion. This applies to other withdrawal strategies as well.

Also, I already live with variable/flexible spending during my working years. In years my income is lower, I lower my spending. It's not that difficult (as long as income is not too low). In retirement, government plans do provide a minimal cushion, in the worst case. In Canada, we have free health-care, Old Age Security (approx. $6,500/year), and CPP (similar to SS but much less generous, yet better funded).

So, let me compare how the different withdrawal strategies cope with my NEED for income relative to VPW :

Constant-Dollar:
If a bad sequence of return happens, I end up with nothing. So I better have a highly variable need indeed. :oops:

So, for me VPW requires a much less variable need for income (in the worst case, of course). Also, VPW is likely to provide a much higher income in most cases.

Constant-Percentage:
A big problem, here, is to choose an appropriate percentage. If I choose the expected real return as percentage, I will underspend. My income will be much lower in the earlier years than with VPW, and will vary with the bond/stock markets. If I choose a higher percentage, it might be too high, leading to fast depletion, or too low, leading to lots of unspent money.

To make a fair comparison, I will choose 4.44%, the same percentage as VPW's first year (in my 30 year @3% example). In the first year, Constant-Percentage and VPW will give me the same amount. But, in the second year, VPW will give me more, regardless of how the markets behave. Actually, during the selected drawdown period, VPW will always give a higher income than Constant-Percentage, regardless of market behavior (except when the 20% cap starts to limit spending, in the last years). Here's an example:

Year 1:
VPW: portfolio $1,000.00, income $44.43
C-P: portfolio $1,000.00, income $44.43

Year 2:
last year real return: -5.00%
VPW: portfolio $907.79, income $40.98
C-P: portfolio $907.79, income $40.33

Year 3:
last year real return: -10.00%
VPW: portfolio $780.13, income $35.81
C-P: portfolio $780.71, income $34.69

Year 3:
last year real return: 7.00%
VPW: portfolio $796.42, income $37.20
C-P: portfolio $798.24, income $35.47

Year 4:
last year real return: -3.00%
VPW: portfolio $736.45, income $35.03
C-P: portfolio $739.89, income $32.87

Year 4:
last year real return: 15.00%
VPW: portfolio $806.63, income $39.11
C-P: portfolio $813.07, income $36.13

In down years, VPW's income doesn't get down as much as Constant-Percentage. In positive years, VPW's income increases more than Constant-Percentage. The key idea is that the capital is being attacked by VPW, giving a much smoother ride.

In summary, VPW will deliver a more stable income than Constant-Percentage.

Spend Only the Dividends:

It is difficult to make a fair comparison of this strategy with VPW. Intuitively, I would say that SOtD will give a much lower income in the earlier years and it will leave much money unspent in the end.

As for stability of income, I can't say, but I prefer a higher variable income to a much lower income.

1/N Withdrawal Amounts

Actually, I forgot to include this one in my first post. It happens to correspond to VPW applied with the following parameters:
  • Expected real return: 0%
  • Number of years: N
  • Drawdown percentage: 100%

1/NWA underestimates the real return (except, maybe, a short-term fixed-income only portfolio), and ends up delivering an increasing income. In other words, it leads me to underspend during the earlier (and healthier) years of my life. It also doesn't cope with cases where I don't want to draw-down 100% (yet, this could be done similarly to VPW).

So, VPM is likely to provide a more stable income (instead of increasing) than 1/NWA.
longinvest
 
Posts: 642
Joined: Sat Aug 11, 2012 9:44 am

Re: Variable Percentage Withdrawal

Postby longinvest » Fri Jul 26, 2013 11:20 am

I'd like to make the Variable Percetage Withdrawal excel sheet available to bogleheads. I just don't know how to attach it. I am looking for help! :happy
longinvest
 
Posts: 642
Joined: Sat Aug 11, 2012 9:44 am

Re: Variable Percentage Withdrawal

Postby longinvest » Fri Jul 26, 2013 11:30 am

JoMoney wrote:I'm a fan of the x/y method, where x=balance y=time you want the money to last.
For retirement purposes, this would be the same as the IRS Required Minimum Distribution (RMD) on an IRA, so "y" = the number of years you have remaining. The number would be based on an actuarial table that changes (i.e. the older you get, the longer you're expected to live).

There's no guarantees on how much your assets will be worth in the future, this method is very safe in making sure it never runs out and gives you a proportional withdrawal rate based on what you know today.


JoMoney, I looked at the RMD tables. Wow! They are much less agressive than Canada's RRIF rules. RRIF rules force you to withdraw 7.38% of your portfolio at age 71 (up to 20% per year at age 94+).

I looked at the RMD rules. They are very interesting, indeed. Yet, it would seem that VPW gives me better control to increase (or not) my spending during the earlier and healthier part of my retirement, and it lets me control the maximum drawdown.

VPW shares the same properties as RMD: it never runs out (as long as you didn't select 100% drawdown) and gives you a proportional withdrawal based on your current portfolio balance. Maybe adding a mortality table component would improve it, but it would complexify it.
longinvest
 
Posts: 642
Joined: Sat Aug 11, 2012 9:44 am

Re: Variable Percentage Withdrawal

Postby longinvest » Fri Jul 26, 2013 11:55 am

Cut-Throat wrote:I am using a VPW, but similar to one that was outlined in Bob's Financial Pages (Not sure if you can get at them anymore).

I take 3% of remaining portfolio balance until Age 70, Increase to 4% at age 75, Increase to 5% at age 80, increase to 6% at age 85.....re-evaluate at each 5 year interval.
According to Jim Otar's retirement book, a 3% withdrawal rate will never deplete the portfolio. Not sure I believe this, but if it is true, then a 3% of remaining portfolio should give an increasing amount to spend each year. (On average). We'll see how this works out. :sharebeer

If is far simpler than the one you outlined. More conservative also. It also avoids the damage that an early sequence of very bad market returns does to a portfolio. It makes you cut back spending if your portfolio shrinks.


3% in your 70s seems low to me, if the expected return of your portfolio is 3%. This is during a hopefully active and healthy part of your retirement. I would prefer to spend more money while I am healthier.

As for VPW, if a bad sequence of return happens, it will lower your spending. The spending is not constant (it does not work like a SWR). The income is computed each year as a percentage of your remaining portfolio, much like the RMD approach. The difference, with RMD, is that the percentage is determined on the basis of 3 parameters that you specify:
  • The expected real return of your portfolio (based on your chosen asset allocation)
  • The percentage of your portfolio that you'd like to drawdown
  • The number of years to do the drawdown
At the end of the drawdown, your remaining portfolio will provide a (small) residual income based on the expected real return of your portfolio.

If you want to be ultra safe, simply select to drawdown 100% over a period that would lead you to extremely old age (115 years-old?). This will lower the rates during earlier years. :D

Example
Expected Real Return: 3.00%
Number of Years: 50
Drawdown: 100.00%
Year 1: 3.77%
...
Year 11: 4.20%
...
Year 21: 4.95%
...
Year 31: 6.53%
...
Year 41: 11.38%
...
Year 50: 20.00%
longinvest
 
Posts: 642
Joined: Sat Aug 11, 2012 9:44 am

Re: Variable Percentage Withdrawal

Postby bertilak » Fri Jul 26, 2013 11:58 am

longinvest wrote:
bertilak wrote:
longinvest wrote:So, does VPW make sense?

Under two related conditions:
  1. You have a variable NEED for income, and
  2. That need varies IN SYNC with the output from your VPW plan.


bertilac, what is your solution?

I think that the two conditions should rather be:
  • I have a flexible spending plan, and
  • I am able to vary my spending according to the output from my VPW plan.

My plan is to establish a safe floor on my income that meets my needs. The amount above that can be flexible.

One potential need is covered by health insurance, another by LTCI (long term care insurance).
No-one really listens to anyone else, and if you try it sometime you will see why.
-- Mignon McLaughlin
User avatar
bertilak
 
Posts: 3094
Joined: Tue Aug 02, 2011 6:23 pm
Location: East of the Pecos, West of the Mississippi

Re: Variable Percentage Withdrawal

Postby longinvest » Fri Jul 26, 2013 12:26 pm

bertilak wrote:
longinvest wrote:
bertilak wrote:
longinvest wrote:So, does VPW make sense?

Under two related conditions:
  1. You have a variable NEED for income, and
  2. That need varies IN SYNC with the output from your VPW plan.


bertilac, what is your solution?

I think that the two conditions should rather be:
  • I have a flexible spending plan, and
  • I am able to vary my spending according to the output from my VPW plan.

My plan is to establish a safe floor on my income that meets my needs. The amount above that can be flexible.

One potential need is covered by health insurance, another by LTCI (long term care insurance).


I agree with establishing a floor. But what about the flexible part? I am trying to come up with a flexible solution to the problems found in other withdrawal strategies proposed in the Boglehead Wiki.
longinvest
 
Posts: 642
Joined: Sat Aug 11, 2012 9:44 am

Re: Variable Percentage Withdrawal

Postby Cut-Throat » Fri Jul 26, 2013 12:40 pm

longinvest wrote: But what about the flexible part? I am trying to come up with a flexible solution to the problems found in other withdrawal strategies proposed in the Boglehead Wiki.


A Constant Percentage of Portfolio Balance does this.
User avatar
Cut-Throat
 
Posts: 1890
Joined: Sun Oct 17, 2010 10:46 am

Re: Variable Percentage Withdrawal

Postby longinvest » Fri Jul 26, 2013 12:57 pm

Cut-Throat wrote:
longinvest wrote: But what about the flexible part? I am trying to come up with a flexible solution to the problems found in other withdrawal strategies proposed in the Boglehead Wiki.


A Constant Percentage of Portfolio Balance does this.


A Constant Percentage is likely to underspend, in the earlier, younger, healthier part of my life, and let me die with a huge pile of money, or, alternately, it can drain my savings too quickly, if I pick too aggressive a percentage. It is quite tricky to pick an appropriate constant percentage. That's what VPW tries to solve.

VPW computes a sliding percentage according to my time horizon, the expected real return of my portfolio, and the percentage of my portfolio that I'd like to drawdown.
longinvest
 
Posts: 642
Joined: Sat Aug 11, 2012 9:44 am

Re: Variable Percentage Withdrawal

Postby Cut-Throat » Fri Jul 26, 2013 1:28 pm

longinvest wrote:A Constant Percentage is likely to underspend, it can drain my savings too quickly, if I pick too aggressive a percentage.


Actually your method would drain savings faster than a Constant percentage.....Percentages being equal in a down market.
User avatar
Cut-Throat
 
Posts: 1890
Joined: Sun Oct 17, 2010 10:46 am

Re: Variable Percentage Withdrawal

Postby uncaD » Fri Jul 26, 2013 1:33 pm

I think this is a terrific insight, thanks for posting. I had been thinking about this exact problem recently, recognizing that 1/NPR effectively assumes a real return of zero and would therefore (likely) result in increasing real dollar withdrawals over time.

I think one interesting implication is that for a 30 year period and a 100% drawdown percentage, you need to assume a real return of only 1.3% to get an initial (safe?) withdrawal rate of 4%.
uncaD
 
Posts: 14
Joined: Mon Jun 24, 2013 1:21 pm

Re: Variable Percentage Withdrawal

Postby longinvest » Fri Jul 26, 2013 1:50 pm

Cut-Throat wrote:
longinvest wrote:A Constant Percentage is likely to underspend, it can drain my savings too quickly, if I pick too aggressive a percentage.


Actually your method would drain savings faster than a Constant percentage.....Percentages being equal in a down market.


It all depends on the percentage. If you star C-P with the same rate as VPW, of course, VPW will drain the portfolio faster. But, C-P will leave a huge pile of unspent money.

If, instead, you start C-P with a higher percentage than VPW, you'll be spending more than than VPW in early down markets. Yet, at the end of the target drawdown period, you're still likely to leave a lot more unspent money.

The only way to leave little unspent money, with C-P, is to pick an overly aggressive spending rate that is likely to drain your portfolio in very few early years.

VPW provides a much smoother experience and a higher margin of error in selecting the rate. In othrer words, if you picked a 3% expected return, and the actual real return averaged 1%, you would only be over-draining your portfolio at an excess rate of 2%. (Your income will reduce by approx 2% per year, instead of staying relatively stable during drawdown years).

It would be much simpler if I could attach an excel sheet. What do other boglehead members do?
longinvest
 
Posts: 642
Joined: Sat Aug 11, 2012 9:44 am

Re: Variable Percentage Withdrawal

Postby longinvest » Fri Jul 26, 2013 1:57 pm

uncaD wrote:I think this is a terrific insight, thanks for posting. I had been thinking about this exact problem recently, recognizing that 1/NPR effectively assumes a real return of zero and would therefore (likely) result in increasing real dollar withdrawals over time.

I think one interesting implication is that for a 30 year period and a 100% drawdown percentage, you need to assume a real return of only 1.3% to get an initial (safe?) withdrawal rate of 4%.


Yes, you are right!

Code: Select all
Expected Real Return :         1.31%
Number of Years :         30
Drawdown :         100.00%

Year    Rate
1   4.00%
2   4.11%
3   4.23%
4   4.36%
5   4.50%
6   4.66%
7   4.82%
8   5.00%
9   5.19%
10   5.41%
11   5.64%
12   5.90%
13   6.19%
14   6.51%
15   6.88%
16   7.29%
17   7.76%
18   8.31%
19   8.94%
20   9.69%
21   10.60%
22   11.70%
23   13.08%
24   14.85%
25   17.21%
26   20.00%
27   20.00%
28   20.00%
29   20.00%
30   20.00%
longinvest
 
Posts: 642
Joined: Sat Aug 11, 2012 9:44 am

Re: Variable Percentage Withdrawal

Postby Cut-Throat » Fri Jul 26, 2013 2:05 pm

longinvest wrote:It all depends on the percentage. If you star C-P with the same rate as VPW, of course, VPW will drain the portfolio faster. But, C-P will leave a huge pile of unspent money.


Not if you annuitize at age 80.
User avatar
Cut-Throat
 
Posts: 1890
Joined: Sun Oct 17, 2010 10:46 am

Re: Variable Percentage Withdrawal

Postby Rodc » Fri Jul 26, 2013 2:23 pm

There is no way from a highly variable asset to set a fixed plan (variable payout or fixed payout) that you can let ride for decades on auto-pilot and spend the last penny on the day you die.

It is a fool's errand.

In some fashion you simply have to keep reexamining how things are going, what your needs are, the risks, etc. and making adjustments as you go. (Can you imagine setting a fixed spending plan up at age 30 and thinking it would never change over 30 yeas?)

To the extent you can be flexible in your spending it helps a lot (try to keep fixed costs fairly low so you can be flexible).

You might live longer than 30 years, or you might die early.

It is also worth noting that in real life people tend not to have constant spending needs that are known and fixed for 30 years so even if you could do the impossible it would likely be the wrong thing to do.

If you really want to spend all your money and not have any left over your best bet is to annuitize all or most of your money at some medium late age, or better over some span of years. That of course has its own risks.
We live a world with knowledge of the future markets has less than one significant figure. And people will still and always demand answers to three significant digits.
Rodc
 
Posts: 10059
Joined: Tue Jun 26, 2007 10:46 am

Re: Variable Percentage Withdrawal

Postby uncaD » Fri Jul 26, 2013 2:50 pm

Rodc wrote:In some fashion you simply have to keep reexamining how things are going, what your needs are, the risks, etc. and making adjustments as you go.


Nobody's saying you wouldn't do that. It's still helpful to have a model of your future withdrawals when you are planning your retirement today.
uncaD
 
Posts: 14
Joined: Mon Jun 24, 2013 1:21 pm

Re: Variable Percentage Withdrawal

Postby longinvest » Fri Jul 26, 2013 2:54 pm

Rodc wrote:There is no way from a highly variable asset to set a fixed plan (variable payout or fixed payout) that you can let ride for decades on auto-pilot and spend the last penny on the day you die.

It is a fool's errand.

In some fashion you simply have to keep reexamining how things are going, what your needs are, the risks, etc. and making adjustments as you go. (Can you imagine setting a fixed spending plan up at age 30 and thinking it would never change over 30 yeas?)

To the extent you can be flexible in your spending it helps a lot (try to keep fixed costs fairly low so you can be flexible).

You might live longer than 30 years, or you might die early.

It is also worth noting that in real life people tend not to have constant spending needs that are known and fixed for 30 years so even if you could do the impossible it would likely be the wrong thing to do.

If you really want to spend all your money and not have any left over your best bet is to annuitize all or most of your money at some medium late age, or better over some span of years. That of course has its own risks.


I mostly agree with you about constantly re-evaluating one's plan and circumstances. Yet, one has to use some guiding principles to compute an approximate amount of income to extract from a portfolio, at a given point of time. Simply picking a magical 4% rate, independently from one's portfolio composition and time horizon, seems foolish.

For instance, VPW could be used to estimate the withdrawal rates for the first 5 years of retirement, let say at age 65, using a 30-year 50% drawdown with 2% expected real return. This would provide the following 5 percentages:
Code: Select all
Year    Rate
1   3.55%
2   3.59%
3   3.63%
4   3.68%
5   3.73%


At 70, you can reevaluate you health, your horizon, your portfolio's balance and expected return. (You can even chose to buy an annuity at that point, if you don't care about losing liquidity).

The withdrawal methods proposed in the Boglehead Wiki are probably meant as guiding principles to make such a computation. I meant VPW as an additional withdrawal method. I just think that VPW is a better estimate than 1/N, constant-dollar, or constant-percentage. Maybe I am wrong, but I haven't seen an argument proving otherwise, so far in this thread.

Saying that annuities are better is comparing apples and oranges. I do not wish to compare a withdrawal method with a SPIA; I do not think that this would be fair, as they are completely different things. (If annuities are the only sensible solution in town, why does the Bogleheads Wiki contain a Withdrawal Method page?)

I do not think that I have found an ultimate solution to all retirement planning problems; far from it. But, I do hope that VPW is a proposal worthy of debate (in comparison to other withdrawal methods, not annuities) in this forum.

Debate brings ideas forward. Already, JoMoney's reference to the RMD method is raising some ideas to better smooth out the tail of the drawdown (currently topped at 20%).

Thanks for the feedback!
Last edited by longinvest on Fri Jul 26, 2013 3:38 pm, edited 1 time in total.
longinvest
 
Posts: 642
Joined: Sat Aug 11, 2012 9:44 am

Re: Variable Percentage Withdrawal

Postby longinvest » Fri Jul 26, 2013 3:10 pm

uncaD wrote:
Rodc wrote:In some fashion you simply have to keep reexamining how things are going, what your needs are, the risks, etc. and making adjustments as you go.


Nobody's saying you wouldn't do that. It's still helpful to have a model of your future withdrawals when you are planning your retirement today.


uncaD, I have to learn to be as concise as you! Thanks.
longinvest
 
Posts: 642
Joined: Sat Aug 11, 2012 9:44 am

Re: Variable Percentage Withdrawal

Postby siamond » Fri Jul 26, 2013 4:08 pm

Cut-Throat wrote:
siamond wrote: Are you so sure that you'd like to have less money to begin with, and more money as you get older? Including when you get real old and rather frail?

Personally, I know I'd like to have a tad more money in my early days of retirement, as I guess I will be more active by then, but then my level of activity might decrease. That being said, my medical expenses will probably increase, so this is a bit of a tough call.


Not sure who you are referring to, but I don't see why the methods explained here would have less money early in retirement and more as you get older.


I was referring to the OP, but you're right, I didn't think past my nose... I was just looking at the increasing withdrawal rates, but if the portfolio shrinks, then the absolute $$ may indeed not increase!
siamond
 
Posts: 522
Joined: Mon May 28, 2012 6:50 am

Re: Variable Percentage Withdrawal

Postby Rodc » Fri Jul 26, 2013 4:23 pm

uncaD wrote:
Rodc wrote:In some fashion you simply have to keep reexamining how things are going, what your needs are, the risks, etc. and making adjustments as you go.


Nobody's saying you wouldn't do that. It's still helpful to have a model of your future withdrawals when you are planning your retirement today.


A plan is one thing. A silly plan is another. First, many posts read as if people actually believe their hyper-precise plans. Some believe the "4% rule" really is a plan (and many tinker like crazy to fine-tune the percentage), rather than a back of the envelope approximation. This occurs over many threads. And in many publications (see the thousand and one threads and published articles on the 4% "rule"). If you actually know better you are ahead of the curve. Go to the head of the class. :)

But, since one has to keep adjusting their plan it makes no sense to be overly complicated. Many of the plans in this thread, and the OP, are fine illustrations of the problem of people trying to get three significant digit answers out of 1 significant digit of knowledge.

Honestly, do you really think it is sensible to have a plan to that goes from 4.95% one year to 5.17% two years later? Talk about false precision. Other so-called plans are no better.

The OP and many others are making this far too complicated. This is wasted energy at best.
We live a world with knowledge of the future markets has less than one significant figure. And people will still and always demand answers to three significant digits.
Rodc
 
Posts: 10059
Joined: Tue Jun 26, 2007 10:46 am

Re: Variable Percentage Withdrawal

Postby uncaD » Fri Jul 26, 2013 5:21 pm

Rodc wrote:A plan is one thing. A silly plan is another.


This plan is no "sillier" than the "1/N Withdrawal Amount" method, which has its own entry in the Bogleheads Wiki.

Rodc wrote:This is wasted energy at best.


So is golf, IMHO. Others apparently disagree. To each his own.
uncaD
 
Posts: 14
Joined: Mon Jun 24, 2013 1:21 pm

Re: Variable Percentage Withdrawal

Postby Hexdump » Sat Jul 27, 2013 11:24 am

Rodc wrote:
A plan is one thing. A silly plan is another.

But, since one has to keep adjusting their plan it makes no sense to be overly complicated. Many of the plans in this thread, and the OP, are fine illustrations of the problem of people trying to get three significant digit answers out of 1 significant digit of knowledge.

Honestly, do you really think it is sensible to have a plan to that goes from 4.95% one year to 5.17% two years later? Talk about false precision. Other so-called plans are no better.

The OP and many others are making this far too complicated. This is wasted energy at best.


No offense meant Rodc but I don't think this is silly at all. And the %s of 4.95 one year to 5.17 two years later are not the plan. They are the answers that his plans spreadsheet spits out. His plan was to run out of money the day he dies, his spreadsheet tries to approximate his goal.

He did not fret and worry about each percent, the computer did all the work once he formulated his plan into some kind of equation.

As you say, if he elects to go along with the recommended %s, he will need to do periodic reviews to see if it is working and if his projections are on track. Doesn't seem silly at all to me.

I give you that worrying about .001 is fruitless and in my case the worry level is at .05.
3.5 vs 4.0 might make a big difference later on.
User avatar
Hexdump
 
Posts: 1436
Joined: Fri Oct 10, 2008 8:28 am
Location: Houston, Texas

Re: Variable Percentage Withdrawal

Postby umfundi » Sat Jul 27, 2013 12:04 pm

It's not that difficult, in my opinion. You don't know the future, so there's no point in calculating it in great detail.

Pick a remaining life expectancy, say, N=30. This year take 1/30 = 3.3%. Next year take 1/29 = 3.4% of the remaining balance. The year after that, 1/28 = 3.6%, and so on. Done.

If there is no market change, this is withdrawing a constant dollar amount each year. The idea is that the payment is not indexed for inflation, but that a market increase over time should provide some compensation for inflation.

If you want higher payments to start, use your actual life expectancy which you can find in the Social Security tables and elsewhere. Compute the withdrawal every year based on your then current life expectancy.

Yes, this does provide you with a potentially fluctuating withdrawal, but I believe it largely eliminates sequence of returns risk. If the market drops, you get less, but that change is spread over the remaining life of the plan.

Keith
Déjà Vu is not a prediction
umfundi
 
Posts: 3361
Joined: Tue Jun 07, 2011 6:26 pm

Re: Variable Percentage Withdrawal

Postby Sheepdog » Sat Jul 27, 2013 12:31 pm

Needs are not constant in retirement. You must plan for such things as required large home maintenance items (major remodeling, new roof, driveway replacement, etc.)...automobile replacements in certain years......major unplanned medical expenses including large hearing or dental expenses which will popup unexpectedly. This is to name just a few above the average and normal expenses that will pop up. I am not even including that round the world cruise you would like to take or that home theater you have yearned for because they are not required . You should plan for those nevertheless. Will whatever percentage withdrawal method you use take care of those? When they occur, will they derail your withdrawal plan resulting in financial difficulties in your late retirement? No plan is perfect, but you should realize that you will have to take them in your considerations. If you have excess income, then you can relax somewhat. If you must maintain a tight retirement income and spending plan to maintain your lifestyle for a long life span (100?) like I do, then you should have a plan for the unplanned.

This plan has worked for me in my 15 years. It requires an annual review of the annual spending and future needs to control withdrawals. In short, I planned for an average annual withdrawal percentage, not a specific set annual withdrawal percentage. The withdrawals will be variable year to year, low in some years in order to pay for the large expenses in others, yet keep within the average of multiple years. It requires me to limit spending in down market years, but I have never wanted.

My goal was a withdrawal of 4.5% a year. That has worked for my lifestyle....good living, new autos for my wife and I on regular basis, nice vacations when income is good, enjoy sports and entertainment, etc. every year. Actual annual withdrawals have been variable to meet spending needs and wants. They have ranged from 3.14% to as high as 6.57%, but the average is almost exactly the planned 4.5% in 14 years.

Things have worked so well that as I approached the beginning of my 15th year this year and age 80, I decided that I can increase my future average withdrawal percentage to 5.5%. At 80, what the heck, my investment balance has increased at 4.5% withdrawal. This budget percentage withdrawal increase should allow us to do even more things such as gifting and more expensive and elective spending if we wish.
Jim
People should not say everything they think. They should think about everything they say.
User avatar
Sheepdog
 
Posts: 3590
Joined: Tue Feb 27, 2007 4:05 pm
Location: Indiana, retired 1998 at age 65

Re: Variable Percentage Withdrawal

Postby longinvest » Sat Jul 27, 2013 7:28 pm

umfundi wrote:It's not that difficult, in my opinion. You don't know the future, so there's no point in calculating it in great detail.

Pick a remaining life expectancy, say, N=30. This year take 1/30 = 3.3%. Next year take 1/29 = 3.4% of the remaining balance. The year after that, 1/28 = 3.6%, and so on. Done.

If there is no market change, this is withdrawing a constant dollar amount each year. The idea is that the payment is not indexed for inflation, but that a market increase over time should provide some compensation for inflation.

If you want higher payments to start, use your actual life expectancy which you can find in the Social Security tables and elsewhere. Compute the withdrawal every year based on your then current life expectancy.

Yes, this does provide you with a potentially fluctuating withdrawal, but I believe it largely eliminates sequence of returns risk. If the market drops, you get less, but that change is spread over the remaining life of the plan.

Keith


Keith,

VPW can be seen as a configurable 1/N. See the end of my earlier post: viewtopic.php?f=10&t=120430#p1760093
longinvest
 
Posts: 642
Joined: Sat Aug 11, 2012 9:44 am

Re: Variable Percentage Withdrawal

Postby longinvest » Sun Jul 28, 2013 9:01 am

[Edited to add:] Now available as a dynamic spreadsheet. Just enter your Asset Allocation and it generates a table for you:

This dynamic spreadsheet allows you to backtest VPW.


I've computed a set of Variable Percentage Withdrawal (VPW) Tables.

Here's how to use the VPW Tables:
  1. At the start of each year, find the percentage P for your current age in the VPW Table specific to the Asset Allocation of your portfolio.
  2. Multiply this percentage by the remaining balance of your portfolio. Withdrawal = P X Balance.


Here is an example for a 65-year old person with a 40% Stocks / 60% Bonds $1,000,000 portfolio:
  • At age 65, the percentage (age 65 in the 40/60 VPW Table) is 4.6%, so she withdraws $46,000.
  • At age 66, the percentage is 4.6%. Assuming there was a bad year, and the market dropped 5%, the portfolio balance is ($1,000,000 - $46,000) - 5% = $906,300. She withdraws $906,300 X 4.6% = $42,000.
  • At age 67, the percentage is 4.7%. Assuming market increased 10%,
    the portfolio balance is ($906,300 - $42,000) + 10% = $950,730. She withdraws $950,730 X 4.7% = $45,000.
  • And so on.


Here are the VPW Tables.

Code: Select all
VPW Table for
Asset Allocation: 20% Stocks / 80% Bonds
Depletion age: 100
Age   Percentage
40   3,1%
41   3,1%
42   3,2%
43   3,2%
44   3,2%
45   3,2%
46   3,3%
47   3,3%
48   3,3%
49   3,4%
50   3,4%
51   3,4%
52   3,5%
53   3,5%
54   3,5%
55   3,6%
56   3,6%
57   3,7%
58   3,7%
59   3,8%
60   3,8%
61   3,9%
62   4,0%
63   4,0%
64   4,1%
65   4,2%
66   4,2%
67   4,3%
68   4,4%
69   4,5%
70   4,6%
71   4,7%
72   4,8%
73   5,0%
74   5,1%
75   5,3%
76   5,4%
77   5,6%
78   5,8%
79   6,0%
80   6,2%
81   6,5%
82   6,8%
83   7,1%
84   7,4%
85   7,8%
86   8,3%
87   8,8%
88   9,5%
89   10,2%
90   11,1%
91   12,2%
92   13,6%
93   15,3%
94   17,7%
95   21,0%
96   25,9%
97   34,1%
98   50,6%
99   100,0%

Code: Select all
VPW Table for
Asset Allocation: 40% Stocks / 60% Bonds
Depletion age: 100
Age   Percentage
40   3,6%
41   3,6%
42   3,6%
43   3,6%
44   3,7%
45   3,7%
46   3,7%
47   3,7%
48   3,8%
49   3,8%
50   3,8%
51   3,9%
52   3,9%
53   3,9%
54   4,0%
55   4,0%
56   4,1%
57   4,1%
58   4,1%
59   4,2%
60   4,3%
61   4,3%
62   4,4%
63   4,4%
64   4,5%
65   4,6%
66   4,6%
67   4,7%
68   4,8%
69   4,9%
70   5,0%
71   5,1%
72   5,2%
73   5,3%
74   5,5%
75   5,6%
76   5,8%
77   5,9%
78   6,1%
79   6,3%
80   6,6%
81   6,8%
82   7,1%
83   7,4%
84   7,8%
85   8,2%
86   8,6%
87   9,2%
88   9,8%
89   10,5%
90   11,4%
91   12,5%
92   13,9%
93   15,6%
94   18,0%
95   21,2%
96   26,1%
97   34,3%
98   50,8%
99   100,0%

Code: Select all
VPW Table for
Asset Allocation: 60% Stocks / 40% Bonds
Depletion age: 100
Age   Percentage
40   4,0%
41   4,1%
42   4,1%
43   4,1%
44   4,1%
45   4,2%
46   4,2%
47   4,2%
48   4,2%
49   4,3%
50   4,3%
51   4,3%
52   4,3%
53   4,4%
54   4,4%
55   4,5%
56   4,5%
57   4,5%
58   4,6%
59   4,6%
60   4,7%
61   4,7%
62   4,8%
63   4,8%
64   4,9%
65   5,0%
66   5,1%
67   5,1%
68   5,2%
69   5,3%
70   5,4%
71   5,5%
72   5,6%
73   5,7%
74   5,9%
75   6,0%
76   6,2%
77   6,3%
78   6,5%
79   6,7%
80   6,9%
81   7,2%
82   7,5%
83   7,8%
84   8,1%
85   8,5%
86   9,0%
87   9,5%
88   10,1%
89   10,8%
90   11,7%
91   12,8%
92   14,2%
93   15,9%
94   18,2%
95   21,5%
96   26,4%
97   34,6%
98   50,9%
99   100,0%

Code: Select all
VPW Table for
Asset Allocation: 80% Stocks / 20% Bonds
Depletion age: 100
Age   Percentage
40   4,6%
41   4,6%
42   4,6%
43   4,6%
44   4,6%
45   4,6%
46   4,7%
47   4,7%
48   4,7%
49   4,7%
50   4,8%
51   4,8%
52   4,8%
53   4,9%
54   4,9%
55   4,9%
56   5,0%
57   5,0%
58   5,0%
59   5,1%
60   5,1%
61   5,2%
62   5,2%
63   5,3%
64   5,3%
65   5,4%
66   5,5%
67   5,6%
68   5,6%
69   5,7%
70   5,8%
71   5,9%
72   6,0%
73   6,1%
74   6,3%
75   6,4%
76   6,5%
77   6,7%
78   6,9%
79   7,1%
80   7,3%
81   7,5%
82   7,8%
83   8,1%
84   8,5%
85   8,9%
86   9,3%
87   9,8%
88   10,4%
89   11,2%
90   12,0%
91   13,1%
92   14,5%
93   16,2%
94   18,5%
95   21,8%
96   26,6%
97   34,8%
98   51,1%
99   100,0%


Frequently Asked Questions

Q: There is no VPW Table with my precise Asset Allocation. What do I do?
A: Pick the Asset Allocation closest to your Asset Allocation. If you have a 25% Stocks / 75% Bonds portfolio, pick the 20% Stocks / 80% Bonds table. Alternately, you can use the dynamic spreadsheet.

Q: Why not simply use a Constant-Dollar 4% withdrawal, indexed to inflation, instead?
A: Because while aiming at keeping relatively stable withdrawals, VPW adapts to the effective returns (gains or losses) of your portfolio. It will not fail if a bad sequence of returns happens at the start of your withdrawal phase. It will also increase your withdrawals if your portfolio gives you returns above what was expected. It also tunes the withdrawals to your age, if you are an early retiree, while also taking into account your Asset Allocation. A more aggressive portfolio will allow higher withdrawals (subject to higher volatility, of course).

Q: Why not simply use a Constant-Percentage withdrawal instead?
A: Because VPW takes into account your current age, and allows you to eat more and more into your capital, as you age, while trying to smooth out the withdrawals. If you start Constant-Percentage and VPW using the same percentage and the same portfolio return sequence, VPW will always withdraw a higher (absolute) amount money than Constant-Percentage, with the important difference that VPW will deplete your portfolio while Constant-percentage won't. (The above VPW tables deplete your portfolio at age 100, the dynamic spreadsheet allows you to select a different depletion period).

Q: Why not use 1/N, instead?
A: Actually, VPW and 1/N are identical when the expected real return of your portfolio is 0% and N = 100 - your age. Unlike 1/N, VPW takes into account your Asset Allocation to smooth out the withdrawals. With most portfolios, 1/N tends to produce increasing withdrawals.

Q: Why use these tables instead of RMD tables?
A: Because VPW takes into account your Asset Allocation, in addition to your current age.

Q: Why not use a lifetime inflation-indexed single-premium immediate annuity (SPIA), instead?
A: Who told you that you couldn't also use an annuity with part of your money? Some people like to keep a certain amount of liquidity. VPW can be used for the liquid part of your portfolio, if you wish. You could also keep a money reserve on the side on which you don't apply VPW. The VPW table is simply a tool, not an overall retirement solution.

Q: Must I withdraw the full amount implied by the year-specific VPW percentage?
A: You don't have to. The percentage you obtain from your asset allocation and your current age, each year, is a maximum; it would be unsafe to withdraw more, but you can withdraw less.

Q: What if life happens and circumstances change?
A: Just re-evaluate your situation and portfolio every now and then. The VPW table is simply a tool, it cannot predict the future and anticipate all life events.

Enjoy!

[Edited to remove false precision]
[Edited to base the withdrawals on asset allocation and portfolio depletion at age 100]
[Edited to improve the FAQ]
[7/31/2013: Edited to show the tables generated by the latest spreadsheet. Removed the cash allocation. Adapted the example and the FAQ accordingly]
[8/3/2013: Edited to update the links and fix typos]
[8/18/2013: Edited to update the links]
Last edited by longinvest on Sun Feb 02, 2014 11:19 am, edited 26 times in total.
longinvest
 
Posts: 642
Joined: Sat Aug 11, 2012 9:44 am

Re: Variable Percentage Withdrawal

Postby Rodc » Sun Jul 28, 2013 9:20 am

So, what do you think? is that simpler?


It is said that Generals say, "No battle was ever won without a plan, and no battle was ever won according to a plan".

Retirement is like that too.

My personal preference is that plans, while good to have, should have precision that reflects the inherent uncertainty in whatever one is planning.

I don't think this plan is so much bad, and just way overly precise. To start with any plan that starts with guessing the real return of your portfolio is possibly in error in the first digit, and is certainly in error in subsequent digits. You have in your example numbers to the nearest dollar, why not to the nearest penny? Planning to either is not realistic.

As others have noted, in addition to my issue of false precision, there is the fact that spending is not nearly that smooth in real life. A plan would be better if that were explicitly planned for. This is really just another issue of the plan being overly precise.

But to each his own. If this makes you happy you should go for it.

Added: After long thought, and years of reading posts here, and just thinking about the well known fact from psychology that humans are very poorly wired to intuitively understand probability and make rational decisions in the face of uncertainty, I think people, otherwise very smart people, have an extremely hard time really at a gut level dealing with the level of uncertainty in investing and retirement planning that exists despite all our hard work to make it otherwise. Thus we agonize over hyper-precise rebalancing schemes, worry about precisely how to set an asset allocation, despite huge amounts of evidence that no one can estimate future returns with any accuracy we read and many (including our resident experts both those that post regularly and not so regularly) insist that to invest well we much make such predictions. And why we have such fascination with various safe withdrawal rate schemes.

I'm sympathetic. It took me a long time, and much modeling and calculating, not just in things like the above withdrawal rates, but more importantly in studying the sensitivity of results to changes in assumptions to get to where I am today. I would suggest that if anyone likes precise calculations, much can be learned by looking at the changes in results due to sensitivity to errors in assumptions. As a hint, just as compound growth in your portfolio over decades is a great thing, the compound growth in errors over decades is just as powerful.

One could certainly at age 30 calculate to the nearest dollar what they plan to spend at age 60. And people could get together to discuss different methods of how to do this calculation to the nearest dollar. But would it be very useful compared to a back of the envelop estimate? Would it even be useful at all?
Last edited by Rodc on Sun Jul 28, 2013 10:00 am, edited 1 time in total.
We live a world with knowledge of the future markets has less than one significant figure. And people will still and always demand answers to three significant digits.
Rodc
 
Posts: 10059
Joined: Tue Jun 26, 2007 10:46 am

Re: Variable Percentage Withdrawal

Postby longinvest » Sun Jul 28, 2013 9:46 am

Rodc wrote:
So, what do you think? is that simpler?


It is said that Generals say, "No battle was ever won without a plan, and no battle was ever won according to a plan".

Retirement is like that too.

My personal preference is that plans, while good to have, should have precision that reflects the inherent uncertainty in whatever one is planning.

I don't think this plan is so much bad, and just way overly precise. To start with any plan that starts with guessing the real return of your portfolio is possibly in error in the first digit, and is certainly in error in subsequent digits. You have in your example numbers to the nearest dollar, why not to the nearest penny? Planning to either is not realistic.


You're are absolutely right! I'll edit my message. I'll reduce the percentages to a single decimal, and round the withdrawals.

Thanks!
longinvest
 
Posts: 642
Joined: Sat Aug 11, 2012 9:44 am

Re: Variable Percentage Withdrawal

Postby Cut-Throat » Sun Jul 28, 2013 9:56 am

longinvest wrote:So, what do you think? is that simpler?


The problem that I have with it, it that it asks you to 'guess' the return of your portfolio; which is like guessing your WR.
A withdrawal amount for a 65 year old could range from 1.82% to 5.11%. This does not help a retiree, but only begs more questions.
Withdrawal Rates are based on historical worse case conditions.

How about 4% of portfolio balance?

If you're pessimistic (Which most of us are here) a 1.82% of portfolio balance is kind of paltry compared to 4%.
This looks like you scrimping in your early years to me, unless you are wildly optimistic.
User avatar
Cut-Throat
 
Posts: 1890
Joined: Sun Oct 17, 2010 10:46 am

Re: Variable Percentage Withdrawal

Postby Rodc » Sun Jul 28, 2013 10:03 am

longinvest wrote:
Rodc wrote:
So, what do you think? is that simpler?


It is said that Generals say, "No battle was ever won without a plan, and no battle was ever won according to a plan".

Retirement is like that too.

My personal preference is that plans, while good to have, should have precision that reflects the inherent uncertainty in whatever one is planning.

I don't think this plan is so much bad, and just way overly precise. To start with any plan that starts with guessing the real return of your portfolio is possibly in error in the first digit, and is certainly in error in subsequent digits. You have in your example numbers to the nearest dollar, why not to the nearest penny? Planning to either is not realistic.


You're are absolutely right! I'll edit my message. I'll reduce the percentages to a single decimal, and round the withdrawals.

Thanks!


:)

BTW we were both typing at the same time when I added by post script above.
We live a world with knowledge of the future markets has less than one significant figure. And people will still and always demand answers to three significant digits.
Rodc
 
Posts: 10059
Joined: Tue Jun 26, 2007 10:46 am

Re: Variable Percentage Withdrawal

Postby longinvest » Sun Jul 28, 2013 10:21 am

Cut-Throat wrote:
longinvest wrote:So, what do you think? is that simpler?


The problem that I have with it, it that it asks you to 'guess' the return of your portfolio; which is like guessing your WR.
A withdrawal amount for a 65 year old could range from 1.82% to 5.11%. This does not help a retiree, but only begs more questions.
Withdrawal Rates are based on historical worse case conditions.

How about 4% of portfolio balance?

If you're pessimistic (Which most of us are here) a 1.82% of portfolio balance is kind of paltry compared to 4%.
This looks like you scrimping in your early years to me, unless you are wildly optimistic.


I could make a couple of improvements:

1- Reduce the target deplition age to 110 or, even 100 (instead of 120). That will up the percentages significantly.
2- Replace the column headings with a portfolio asset allocation. Something like:
  • Cash (-2%) / Bonds (2%) / Stocks (5%)
  • 50/50/0 (0%)
  • 25/75/0 (1%)
  • 0/100/0 (2%)
  • 0/66/34 (3%)
  • 0/33/67 (4%)
  • 0/0/100 (5%)

Here's the result:

[Edited to add:] The final version is presented in viewtopic.php?f=10&t=120430&p=1761563#p1761563

Code: Select all
   Asset Allocation Cash/Bond/Stock(Expected Real Return)               
Age   50/50/0(0%)   25/75/0(1%)   0/100/0(2%)   0/66/34(3%)   0/33/67(%4)   0/0/100(5%)

30   1.4%   2.0%   2.6%   3.3%   4.1%   4.9%
31   1.4%   2.0%   2.6%   3.3%   4.1%   4.9%
32   1.5%   2.0%   2.7%   3.4%   4.1%   4.9%
33   1.5%   2.0%   2.7%   3.4%   4.1%   5.0%
34   1.5%   2.1%   2.7%   3.4%   4.2%   5.0%
35   1.5%   2.1%   2.7%   3.4%   4.2%   5.0%
36   1.6%   2.1%   2.7%   3.4%   4.2%   5.0%
37   1.6%   2.1%   2.8%   3.4%   4.2%   5.0%
38   1.6%   2.2%   2.8%   3.5%   4.2%   5.0%
39   1.6%   2.2%   2.8%   3.5%   4.2%   5.0%
40   1.7%   2.2%   2.8%   3.5%   4.3%   5.0%
41   1.7%   2.2%   2.8%   3.5%   4.3%   5.0%
42   1.7%   2.3%   2.9%   3.6%   4.3%   5.1%
43   1.8%   2.3%   2.9%   3.6%   4.3%   5.1%
44   1.8%   2.3%   2.9%   3.6%   4.3%   5.1%
45   1.8%   2.3%   3.0%   3.6%   4.3%   5.1%
46   1.9%   2.4%   3.0%   3.7%   4.4%   5.1%
47   1.9%   2.4%   3.0%   3.7%   4.4%   5.1%
48   1.9%   2.5%   3.0%   3.7%   4.4%   5.2%
49   2.0%   2.5%   3.1%   3.7%   4.4%   5.2%
50   2.0%   2.5%   3.1%   3.8%   4.5%   5.2%
51   2.0%   2.6%   3.2%   3.8%   4.5%   5.2%
52   2.1%   2.6%   3.2%   3.8%   4.5%   5.3%
53   2.1%   2.7%   3.2%   3.9%   4.6%   5.3%
54   2.2%   2.7%   3.3%   3.9%   4.6%   5.3%
55   2.2%   2.7%   3.3%   4.0%   4.6%   5.4%
56   2.3%   2.8%   3.4%   4.0%   4.7%   5.4%
57   2.3%   2.8%   3.4%   4.0%   4.7%   5.4%
58   2.4%   2.9%   3.5%   4.1%   4.8%   5.5%
59   2.4%   3.0%   3.5%   4.1%   4.8%   5.5%
60   2.5%   3.0%   3.6%   4.2%   4.9%   5.6%
61   2.6%   3.1%   3.6%   4.3%   4.9%   5.6%
62   2.6%   3.1%   3.7%   4.3%   5.0%   5.6%
63   2.7%   3.2%   3.8%   4.4%   5.0%   5.7%
64   2.8%   3.3%   3.8%   4.4%   5.1%   5.8%
65   2.9%   3.4%   3.9%   4.5%   5.2%   5.8%
66   2.9%   3.4%   4.0%   4.6%   5.2%   5.9%
67   3.0%   3.5%   4.1%   4.7%   5.3%   6.0%
68   3.1%   3.6%   4.2%   4.8%   5.4%   6.0%
69   3.2%   3.7%   4.3%   4.9%   5.5%   6.1%
70   3.3%   3.8%   4.4%   5.0%   5.6%   6.2%
71   3.4%   4.0%   4.5%   5.1%   5.7%   6.3%
72   3.6%   4.1%   4.6%   5.2%   5.8%   6.4%
73   3.7%   4.2%   4.7%   5.3%   5.9%   6.5%
74   3.8%   4.3%   4.9%   5.4%   6.0%   6.6%
75   4.0%   4.5%   5.0%   5.6%   6.2%   6.8%
76   4.2%   4.7%   5.2%   5.7%   6.3%   6.9%
77   4.3%   4.8%   5.4%   5.9%   6.5%   7.1%
78   4.5%   5.0%   5.6%   6.1%   6.7%   7.2%
79   4.8%   5.3%   5.8%   6.3%   6.9%   7.4%
80   5.0%   5.5%   6.0%   6.5%   7.1%   7.6%
81   5.3%   5.7%   6.3%   6.8%   7.3%   7.9%
82   5.6%   6.0%   6.5%   7.1%   7.6%   8.1%
83   5.9%   6.4%   6.9%   7.4%   7.9%   8.4%
84   6.2%   6.7%   7.2%   7.7%   8.3%   8.8%
85   6.7%   7.1%   7.6%   8.1%   8.6%   9.2%
86   7.1%   7.6%   8.1%   8.6%   9.1%   9.6%
87   7.7%   8.2%   8.6%   9.1%   9.6%   10.1%
88   8.3%   8.8%   9.3%   9.8%   10.2%   10.7%
89   9.1%   9.5%   10.0%   10.5%   11.0%   11.5%
90   10.0%   10.5%   10.9%   11.4%   11.9%   12.3%
91   11.1%   11.6%   12.0%   12.5%   12.9%   13.4%
92   12.5%   12.9%   13.4%   13.8%   14.3%   14.7%
93   14.3%   14.7%   15.1%   15.6%   16.0%   16.5%
94   16.7%   17.1%   17.5%   17.9%   18.3%   18.8%
95+   20.0%   20.0%   20.0%   20.0%   20.0%   20.0%

What do you think?
Last edited by longinvest on Sun Jul 28, 2013 11:20 am, edited 1 time in total.
longinvest
 
Posts: 642
Joined: Sat Aug 11, 2012 9:44 am

Re: Variable Percentage Withdrawal

Postby Cut-Throat » Sun Jul 28, 2013 10:30 am

longinvest wrote:What do you think?


I like it better. Age 100 is plenty long enough. And Asset Allocation takes the guessing out of Returns.

It would also look easier with Asset Allocations being more in line with other models. Instead of 34% stocks and 66% Bonds......make it 30% Stocks/70% Bonds etc......More Round numbers instead of Odd numbers.
User avatar
Cut-Throat
 
Posts: 1890
Joined: Sun Oct 17, 2010 10:46 am

Re: Variable Percentage Withdrawal

Postby longinvest » Sun Jul 28, 2013 11:03 am

Cut-Throat wrote:
longinvest wrote:What do you think?


I like it better. Age 100 is plenty long enough. And Asset Allocation takes the guessing out of Returns.

It would also look easier with Asset Allocations being more in line with other models. Instead of 34% stocks and 66% Bonds......make it 30% Stocks/70% Bonds etc......More Round numbers instead of Odd numbers.


Thanks! OK, I did this (but with 0/60/40 and 0/40/60, as they are popular allocations). I also fixed the example and FAQ accordingly: viewtopic.php?f=10&t=120430&p=1761563#p1761563

Any other improvement?
longinvest
 
Posts: 642
Joined: Sat Aug 11, 2012 9:44 am

Re: Variable Percentage Withdrawal

Postby Cut-Throat » Sun Jul 28, 2013 11:13 am

longinvest wrote:Thanks! OK, I did this (but with 0/60/40 and 0/40/60, as they are popular allocations). I also fixed the example and FAQ accordingly: viewtopic.php?f=10&t=120430&p=1761563#p1761563

Any other improvement?


It would really be handy to include the Stock/Bond Numbers as variable inputs. If you are really handy with Xcel (Which I'm Not), you could have a Slider Bar (For a Stock Allocation) that would change percentages 'on the fly'. Then you would not even need 5 Columns of Percentages, Just one that changes with manipulation of the Slider Bar. Then the 'user' could change the Allocation to whatever Odd number they wanted. 62% Stocks/38% Bonds as an example.

If you do this.....I'll take a copy (You can e-mail it to me) :happy
User avatar
Cut-Throat
 
Posts: 1890
Joined: Sun Oct 17, 2010 10:46 am

Re: Variable Percentage Withdrawal

Postby longinvest » Sun Jul 28, 2013 11:17 am

Cut-Throat wrote:
longinvest wrote:Thanks! OK, I did this (but with 0/60/40 and 0/40/60, as they are popular allocations). I also fixed the example and FAQ accordingly: viewtopic.php?f=10&t=120430&p=1761563#p1761563

Any other improvement?


It would really be handy to include the Stock/Bond Numbers as variable inputs. If you are really handy with Xcel (Which I'm Not), you could have a Slider Bar (For a Stock Allocation) that would change percentages 'on the fly'. Then you would not even need 5 Columns of Percentages, Just one that changes with manipulation of the Slider Bar. Then the 'user' could change the Allocation to whatever Odd number they wanted. 62% Stocks/38% Bonds as an example.

If you do this.....I'll take a copy (You can e-mail it to me) :happy


Hey! You're looking for false precision! The table above is precise enough. :D
longinvest
 
Posts: 642
Joined: Sat Aug 11, 2012 9:44 am

Re: Variable Percentage Withdrawal

Postby Cut-Throat » Sun Jul 28, 2013 11:21 am

longinvest wrote:Hey! You're looking for false precision! The table above is precise enough. :D


If you do this, I promise to only move the Slider Bar to Round Asset Allocation (No false Precision for me!) :mrgreen:
User avatar
Cut-Throat
 
Posts: 1890
Joined: Sun Oct 17, 2010 10:46 am

Re: Variable Percentage Withdrawal

Postby longinvest » Sun Jul 28, 2013 12:56 pm

Cut-Throat wrote:
longinvest wrote:Hey! You're looking for false precision! The table above is precise enough. :D


If you do this, I promise to only move the Slider Bar to Round Asset Allocation (No false Precision for me!) :mrgreen:


OK, I don't know how make a sliding bar, but I made an Excel spreadsheet (with OpenOffice) where you can enter your bond/stock allocation and it generates the table. I've reduced the expected return increment to 0.5% so that it can gracefully handle a 50/50 allocation.

How can I share it here?
longinvest
 
Posts: 642
Joined: Sat Aug 11, 2012 9:44 am

Re: Variable Percentage Withdrawal

Postby longinvest » Sun Jul 28, 2013 1:31 pm

I don't know how to back-test VPW. Doing something like a FireCalc on it to see how the withdrawals fluctuate would be cool. Does anybody know how to do that?
longinvest
 
Posts: 642
Joined: Sat Aug 11, 2012 9:44 am

Re: Variable Percentage Withdrawal

Postby Cut-Throat » Sun Jul 28, 2013 1:47 pm

Not sure if there is a better method to share your spreadsheet, but the instructions here should work.

http://personalweb.about.com/od/linkingideas/a/408xlsfiles.htm
User avatar
Cut-Throat
 
Posts: 1890
Joined: Sun Oct 17, 2010 10:46 am

Re: Variable Percentage Withdrawal

Postby umfundi » Sun Jul 28, 2013 2:04 pm

I think you can save it on Google Docs and allow anyone to read it, then post the link here.

Or, use something like DropBox. Pick one of these:

http://www.ebizmba.com/articles/file-sharing-websites

Keith
Déjà Vu is not a prediction
umfundi
 
Posts: 3361
Joined: Tue Jun 07, 2011 6:26 pm

Re: Variable Percentage Withdrawal

Postby longinvest » Sun Jul 28, 2013 2:16 pm

umfundi wrote:I think you can save it on Google Docs and allow anyone to read it, then post the link here.

Or, use something like DropBox. Pick one of these:

http://www.ebizmba.com/articles/file-sharing-websites

Keith


Thanks for the tip. Here it is: [Edited to point to the latest version:] https://www.dropbox.com/s/0dmzcswmn4qzj ... drawal.xls

Please let me know if the link works, and if the spreadsheet works too.
Last edited by longinvest on Wed Jul 31, 2013 1:25 am, edited 1 time in total.
longinvest
 
Posts: 642
Joined: Sat Aug 11, 2012 9:44 am

Re: Variable Percentage Withdrawal

Postby umfundi » Sun Jul 28, 2013 2:33 pm

I accessed the file and changed the inputs. It seems to work. (I used MS Excel 2010.)

I could not figure out how to see the hidden cells so I could look at the formulas.

Keith
Déjà Vu is not a prediction
umfundi
 
Posts: 3361
Joined: Tue Jun 07, 2011 6:26 pm

Re: Variable Percentage Withdrawal

Postby longinvest » Sun Jul 28, 2013 2:41 pm

umfundi wrote:I accessed the file and changed the inputs. It seems to work. (I used MS Excel 2010.)

I could not figure out how to see the hidden cells so I could look at the formulas.

Keith


Just go into Tools->Protect->Sheet and unprotect the sheet (there is no password). Then select columns B and E, and click on Format->Column->Optimal Width.
longinvest
 
Posts: 642
Joined: Sat Aug 11, 2012 9:44 am

Re: Variable Percentage Withdrawal

Postby longinvest » Mon Jul 29, 2013 3:03 am

I was able to build a back-testing spreadsheet using data extracted from Simba's backtesting spreadsheets. It uses a Three Fund Portfolio and adjusts returns to the CPI-U.

You provide an asset allocation (bond/stock), a stock allocation (US/international), a portfolio size, a retirement age, and a retirement year (1972 or later).

Here is the VPW Back Testing spreadsheet: [Edited to point to the latest version:] https://www.dropbox.com/s/0dmzcswmn4qzj ... drawal.xls

Feedback is welcome.
Last edited by longinvest on Wed Jul 31, 2013 1:26 am, edited 1 time in total.
longinvest
 
Posts: 642
Joined: Sat Aug 11, 2012 9:44 am

Re: Variable Percentage Withdrawal

Postby Cut-Throat » Mon Jul 29, 2013 4:20 am

longinvest wrote:I was able to build a back-testing spreadsheet using data extracted from Simba's backtesting spreadsheets. It uses a Three Fund Portfolio and adjusts returns to the CPI-U.

You provide an asset allocation (bond/stock), a stock allocation (US/international), a portfolio size, a retirement age, and a retirement year (1972 or later).

Here is the VPW Back Testing spreadsheet: https://www.dropbox.com/s/m29bpcctv78to ... ckTest.xls

Feedback is welcome.


Nice Work!

The program does 'blow up' if a year earlier than 1972 is entered. I entered 1955 and it spit out some numbers, but eventually croaked.

What would be nice is another column of numbers next to the ones you just added. comparing a Constant Percentage (That increases 1% every 5 years) of Withdrawal (variable starting percent Input), showing Portfolio Balance and Withdrawal amount. So, you could see the 'smoothing' effect of the VPW. Note -- the Constant Percent with increase every 5 years was proposed by Stein and Demuth.

Too bad we only have data after 1972.
User avatar
Cut-Throat
 
Posts: 1890
Joined: Sun Oct 17, 2010 10:46 am

Next

Return to Investing - Theory, News & General

Who is online

Users browsing this forum: BL, Cut-Throat, kttpn, MoneyMythsSolved, noktem, sscritic and 62 guests