Is VPW a safer retirement strategy than SWR?

Discuss all general (i.e. non-personal) investing questions and issues, investing news, and theory.
zuma
Posts: 641
Joined: Thu Dec 29, 2016 11:15 am

Re: Variable Percentage Withdrawal (VPW)

Post by zuma »

nigel_ht wrote: Sat Jul 24, 2021 8:48 am Well...my guess would be SWR would work just as well for you and if you aren't spending up to the VPW limit there really isn't that much point to choosing that over SWR right?
Your arguments are not very convincing.

I use VPW as a structured way to provide regular monthly income. With this system my income will be variable, as it depends on the market's performance. However, my budget (i.e. expected spending) is calculated separately and takes into account the possibility of a severe market crash. My spending requirements are flexible (I'm single, no kids, no debt) and VPW provides an automated way to earn more when the market is doing well.

Earning more doesn't mean that I must spend everything that I earn. I still save for larger unplanned expenses, extra travel, etc. This is the same thing I did while working a normal job. There seems to be assumption in your analysis that retirees must spend all of their income immediately. During a bear market, I can cut discretionary expenses and/or dip into savings to meet my needs.

I have nothing against SWR if that works for you. It's a more conservative, predictable approach to retirement income and is probably a better choice if your budget is less flexible. Personally, I prefer having the built-in adjustments of VPW.
Topic Author
nigel_ht
Posts: 4742
Joined: Tue Jan 01, 2019 9:14 am

Re: Variable Percentage Withdrawal (VPW)

Post by nigel_ht »

zuma wrote: Sat Jul 24, 2021 9:50 am
nigel_ht wrote: Sat Jul 24, 2021 8:48 am Well...my guess would be SWR would work just as well for you and if you aren't spending up to the VPW limit there really isn't that much point to choosing that over SWR right?
Your arguments are not very convincing.

I use VPW as a structured way to provide regular monthly income. With this system my income will be variable, as it depends on the market's performance. However, my budget (i.e. expected spending) is calculated separately and takes into account the possibility of a severe market crash. My spending requirements are flexible (I'm single, no kids, no debt) and VPW provides an automated way to earn more when the market is doing well.
VPW doesn’t help you earn more (time in market) but helps you spend more. If you don’t need or want to spend more than nearly any withdrawal strategy works including winging it.

It means the needed WR is far below the SWR amount. A lot of us are in this category where our WR is less than 3%.
Earning more doesn't mean that I must spend everything that I earn.
The only money you are earning in retirement comes from your portfolio and any other actual income. Everything else is reducing your net worth and not “earning” but “spending”
I still save for larger unplanned expenses, extra travel…
Ah…why? It’s not “saving”. You already did that in the accumulation phase. Now you’re just spending slower.

You can do this in SWR too…”saving” unused budget one year for something else later.

This terminology is why I think VPW like strategies are less frugal. If folks are thinking of the withdrawals as “income” as opposed to spending that gives a different sort of mental model to retirement spending.
zuma
Posts: 641
Joined: Thu Dec 29, 2016 11:15 am

Re: Variable Percentage Withdrawal (VPW)

Post by zuma »

nigel_ht wrote: Sat Jul 24, 2021 9:58 am VPW doesn’t help you earn more (time in market) but helps you spend more. If you don’t need or want to spend more than nearly any withdrawal strategy works including winging it.
It's a metaphor.

I "earn" money in retirement just like I did before retirement: money appears in my checking account.
Topic Author
nigel_ht
Posts: 4742
Joined: Tue Jan 01, 2019 9:14 am

Re: Variable Percentage Withdrawal (VPW)

Post by nigel_ht »

zuma wrote: Sat Jul 24, 2021 9:50 am During a bear market, I can cut discretionary expenses and/or dip into savings to meet my needs.
Wait. Where are you keeping these “savings”? In like an actual HYS account or something?

If it just sits in your portfolio unused doesn’t VPW take care of it automatically even if it’s smeared across the remaining time? How would you know how much is available to “dip into” if you have to go above the VPW computed VPW amount?
zuma
Posts: 641
Joined: Thu Dec 29, 2016 11:15 am

Re: Variable Percentage Withdrawal (VPW)

Post by zuma »

nigel_ht wrote: Sat Jul 24, 2021 10:24 am
zuma wrote: Sat Jul 24, 2021 9:50 am During a bear market, I can cut discretionary expenses and/or dip into savings to meet my needs.
Wait. Where are you keeping these “savings”? In like an actual HYS account or something?

If it just sits in your portfolio unused doesn’t VPW take care of it automatically even if it’s smeared across the remaining time? How would you know how much is available to “dip into” if you have to go above the VPW computed VPW amount?
In my checking account. This is money that's available to spend (or not).
Topic Author
nigel_ht
Posts: 4742
Joined: Tue Jan 01, 2019 9:14 am

Re: Variable Percentage Withdrawal (VPW)

Post by nigel_ht »

zuma wrote: Sat Jul 24, 2021 10:21 am
nigel_ht wrote: Sat Jul 24, 2021 9:58 am VPW doesn’t help you earn more (time in market) but helps you spend more. If you don’t need or want to spend more than nearly any withdrawal strategy works including winging it.
It's a metaphor.

I "earn" money in retirement just like I did before retirement: money appears in my checking account.
Yes, but it’s an incorrect mental model. Before retirement the money that shows up comes from external source that increases your financial health. It really is earned and is income.

After retirement the money comes from what you saved during accumulation. If this exceeds the gains from the portfolio then your net worth drops. In a SORR scenario this means your reserves are lower…so in the 1965 example above you could deplete your portfolio if your minimum spend stays above the VPW computed amount.
Topic Author
nigel_ht
Posts: 4742
Joined: Tue Jan 01, 2019 9:14 am

Re: Variable Percentage Withdrawal (VPW)

Post by nigel_ht »

zuma wrote: Sat Jul 24, 2021 10:28 am
nigel_ht wrote: Sat Jul 24, 2021 10:24 am
zuma wrote: Sat Jul 24, 2021 9:50 am During a bear market, I can cut discretionary expenses and/or dip into savings to meet my needs.
Wait. Where are you keeping these “savings”? In like an actual HYS account or something?

If it just sits in your portfolio unused doesn’t VPW take care of it automatically even if it’s smeared across the remaining time? How would you know how much is available to “dip into” if you have to go above the VPW computed VPW amount?
In my checking account. This is money that's available to spend (or not).
Well if it’s a small amount and you don’t keep it for long then you are mostly spending everything VPW allows you to.

If not, and the balance is large and held for a long time…that seems…suboptimal.
zuma
Posts: 641
Joined: Thu Dec 29, 2016 11:15 am

Re: Variable Percentage Withdrawal (VPW)

Post by zuma »

nigel_ht wrote: Sat Jul 24, 2021 10:32 am
zuma wrote: Sat Jul 24, 2021 10:28 am
nigel_ht wrote: Sat Jul 24, 2021 10:24 am
zuma wrote: Sat Jul 24, 2021 9:50 am During a bear market, I can cut discretionary expenses and/or dip into savings to meet my needs.
Wait. Where are you keeping these “savings”? In like an actual HYS account or something?

If it just sits in your portfolio unused doesn’t VPW take care of it automatically even if it’s smeared across the remaining time? How would you know how much is available to “dip into” if you have to go above the VPW computed VPW amount?
In my checking account. This is money that's available to spend (or not).
Well if it’s a small amount and you don’t keep it for long then you are mostly spending everything VPW allows you to.

If not, and the balance is large and held for a long time…that seems…suboptimal.
The actual amount varies depending on the circumstances. But optimisation is not the goal. The goal is to withdraw from the portfolio in a systematic way that is consistent with market returns.
Topic Author
nigel_ht
Posts: 4742
Joined: Tue Jan 01, 2019 9:14 am

Re: Variable Percentage Withdrawal (VPW)

Post by nigel_ht »

zuma wrote: Sat Jul 24, 2021 11:15 am
nigel_ht wrote: Sat Jul 24, 2021 10:32 am
zuma wrote: Sat Jul 24, 2021 10:28 am
nigel_ht wrote: Sat Jul 24, 2021 10:24 am
zuma wrote: Sat Jul 24, 2021 9:50 am During a bear market, I can cut discretionary expenses and/or dip into savings to meet my needs.
Wait. Where are you keeping these “savings”? In like an actual HYS account or something?

If it just sits in your portfolio unused doesn’t VPW take care of it automatically even if it’s smeared across the remaining time? How would you know how much is available to “dip into” if you have to go above the VPW computed VPW amount?
In my checking account. This is money that's available to spend (or not).
Well if it’s a small amount and you don’t keep it for long then you are mostly spending everything VPW allows you to.

If not, and the balance is large and held for a long time…that seems…suboptimal.
The actual amount varies depending on the circumstances. But optimisation is not the goal. The goal is to withdraw from the portfolio in a systematic way that is consistent with market returns.
I generally have no issues with mental accounting but this seems like a fairly extreme version where you pull money out of your savings portfolio to “save” it as cash.

Now if there is a RMD that makes sense but why wouldn’t you toss it into a taxable account using your preferred AA?

Optimization may not be a goal but we do believe in low ER as part of the BH strategy…by taking money out you aren’t going to actually spend is odd.

In any case it implies you have plenty of retirement savings that doing anything reasonable works…while not complex, I dunno what the VPW flexibility is really providing you.

But like you said, I’m happy it works for you. I have no issues with VPW, I just want to understand the failure modes and weaknesses of the strategy.

I do still have issue with the mental model where portfolio withdrawals is thought of as “income”…
zuma
Posts: 641
Joined: Thu Dec 29, 2016 11:15 am

Re: Variable Percentage Withdrawal (VPW)

Post by zuma »

nigel_ht wrote: Sat Jul 24, 2021 11:28 am I generally have no issues with mental accounting but this seems like a fairly extreme version where you pull money out of your savings portfolio to “save” it as cash.

Now if there is a RMD that makes sense but why wouldn’t you toss it into a taxable account using your preferred AA?

Optimization may not be a goal but we do believe in low ER as part of the BH strategy…by taking money out you aren’t going to actually spend is odd.
I will spend the money. Again, the point is that I'm not required to spend it all every month. I can use budgeting techniques like I did before retirement. In some cases, maybe I want to have a little extra cash available.

The benefit of VPW for me is that my withdrawal amounts are mechanical and based on actual market returns. My "income" (or whatever word you prefer) will fluctuate and I will adjust my spending accordingly. This is not a magic bullet approach, but it works for me.
longinvest
Posts: 5682
Joined: Sat Aug 11, 2012 8:44 am

Re: Variable Percentage Withdrawal (VPW) - Danger Will Robinson!

Post by longinvest »

zuma wrote: Sat Jul 24, 2021 12:08 pm My "income" (or whatever word you prefer) will fluctuate and I will adjust my spending accordingly. This is not a magic bullet approach, but it works for me.
Zuma, you're not alone adapting your spending to your income. That's what my wife and I have done, so far. When our household income increased, we increased both spending and savings. When it decreased, we decreased both spending and savings.

Budgeting techniques and, in the earlier years, low-cost debt to spread the cost of bigger items* over time when we had little liquidity took care of the mismatch between regular income and lumpy expenses. We're now able to manage without debt, and we prefer it that way, but it's a choice.

* We very carefully kept the cost of big items reasonable and debt payments as a small fraction of our take home pay, anticipating the possibility that either spouse could lose their job.

When we'll retire, we'll continue adapting our spending to our retirement income which will consist of non-portfolio income (in part delayed) and variable withdrawals from our balanced portfolio using VPW.
Variable Percentage Withdrawal (bogleheads.org/wiki/VPW) | One-Fund Portfolio (bogleheads.org/forum/viewtopic.php?t=287967)
Topic Author
nigel_ht
Posts: 4742
Joined: Tue Jan 01, 2019 9:14 am

Re: Variable Percentage Withdrawal (VPW)

Post by nigel_ht »

zuma wrote: Sat Jul 24, 2021 12:08 pm
nigel_ht wrote: Sat Jul 24, 2021 11:28 am I generally have no issues with mental accounting but this seems like a fairly extreme version where you pull money out of your savings portfolio to “save” it as cash.

Now if there is a RMD that makes sense but why wouldn’t you toss it into a taxable account using your preferred AA?

Optimization may not be a goal but we do believe in low ER as part of the BH strategy…by taking money out you aren’t going to actually spend is odd.
I will spend the money. Again, the point is that I'm not required to spend it all every month. I can use budgeting techniques like I did before retirement. In some cases, maybe I want to have a little extra cash available.

The benefit of VPW for me is that my withdrawal amounts are mechanical and based on actual market returns. My "income" (or whatever word you prefer) will fluctuate and I will adjust my spending accordingly. This is not a magic bullet approach, but it works for me.
Unless your job was commission or tips based or based on personal business profits you likely didn’t have very large fluctuations in income that requires you to adjust to.

Pre-retirement it probably looks more like SWR than VPW in bad years as your salary generally isn’t likely market driven. The biggest difference between SWR and most pre-retirement salary is you don’t get any raises in SWR except for COLA increases.

Like I said, if it works for you great. But VPW does have risks if you spend more in the first few years and run into SORR…you are forced to decide whether to increase your spend which risks portfolio depletion fast than it may have been with SWR or dropping below you minimum spend and “selling your house” to achieve the flexibility required.

The key issue is that folks where 4% spend is marginal to meet their spending needs will see VPW as a “safe” way to spend more than 4% and the risk of failure is obscured because VPW claims it never depletes before the end date.

If the user really needs 4.5% to meet their needs including SS this isn’t easily seen as it is with SWR and the computed historical failure calculations.
longinvest
Posts: 5682
Joined: Sat Aug 11, 2012 8:44 am

Re: Variable Percentage Withdrawal (VPW) - Danger Will Robinson!

Post by longinvest »

Guaranteed income can be bought, it's called an inflation-indexed SPIA* (joint life, for a couple).

* Single Premium Immediate Annuity.

It's expensive. The younger the retiree, the more expensive it gets. But it exists. An early retiree could use as small part of their portfolio to buy one* and gain partial certainty. The financial inability to do so might be indicative that the investor doesn't have the means to retire so early. The solution, in that case, is to continue working a few additional years and reconsider the situation.

* For a period of at most 30-year, a properly constructed non-rolling TIPS ladder could be constructed as a home-made term-certain annuity to provide bridge payments (coupons + maturing principal) until the start of a delayed pension.
Last edited by longinvest on Sat Jul 24, 2021 1:12 pm, edited 1 time in total.
Variable Percentage Withdrawal (bogleheads.org/wiki/VPW) | One-Fund Portfolio (bogleheads.org/forum/viewtopic.php?t=287967)
Topic Author
nigel_ht
Posts: 4742
Joined: Tue Jan 01, 2019 9:14 am

Re: Variable Percentage Withdrawal (VPW) - Danger Will Robinson!

Post by nigel_ht »

longinvest wrote: Sat Jul 24, 2021 12:19 pm
zuma wrote: Sat Jul 24, 2021 12:08 pm My "income" (or whatever word you prefer) will fluctuate and I will adjust my spending accordingly. This is not a magic bullet approach, but it works for me.
Zuma, you're not alone adapting your spending to your income. That's what my wife and I have done, so far. When our household income increased, we increased both spending and savings. When it decreased, we decreased both spending and savings.
Has income varied by 50%? For some folks their income does vary by quite a lot year to year. For what I would guess to be the majority in the forums it does not.

Which means their pre-retirement spending/budgeting techniques aren’t battle tested with highly variable income amounts.

Your ability to adapt is fairly academic until you have to perform it under real world circumstances unless you’ve suffered a long stretch of unplanned unemployment.

Now I believe that Klangfool IS fully able to adapt given his past experience but most folks don’t actually have personal experience with long term un- or under- employment.

So I’m not banking on my theoretical ability to adapt to large unplanned withdrawal fluctuations to meet my planned expenses when I already know that the error bars for my estimated retirement expenses are very large if we run into unexpected health issues.

There is a reason we’re not shooting for 25X expenses but 33X expenses and it’s not for worse than historical market outcomes. Those are rare. Strokes, cancer, heart failure, car accidents, etc are not uncommon even for folks that take care of themselves.

These are sufficient unknown unknowns for expenses that SWR isn’t really that conservative.
longinvest
Posts: 5682
Joined: Sat Aug 11, 2012 8:44 am

Re: Variable Percentage Withdrawal (VPW) - Danger Will Robinson!

Post by longinvest »

nigel_ht wrote: Sat Jul 24, 2021 1:12 pm Has income varied by 50%?
Dear Nigel_ht,

Yes, it has dropped by even more. We adapted.

We all know people who lost their job, or had increased expenses while income dropped (e.g. having children), or divorced, or had many other major changes in their life situation putting them into a new financial situation.

Best regards,

longinvest
Last edited by longinvest on Sat Jul 24, 2021 1:19 pm, edited 1 time in total.
Variable Percentage Withdrawal (bogleheads.org/wiki/VPW) | One-Fund Portfolio (bogleheads.org/forum/viewtopic.php?t=287967)
Topic Author
nigel_ht
Posts: 4742
Joined: Tue Jan 01, 2019 9:14 am

Re: Variable Percentage Withdrawal (VPW) - Danger Will Robinson!

Post by nigel_ht »

longinvest wrote: Sat Jul 24, 2021 1:12 pm
nigel_ht wrote: Sat Jul 24, 2021 1:12 pm Has income varied by 50%?
Yes, it has dropped by even more. We adapted.
For a decade?
longinvest
Posts: 5682
Joined: Sat Aug 11, 2012 8:44 am

Re: Variable Percentage Withdrawal (VPW) - Danger Will Robinson!

Post by longinvest »

nigel_ht wrote: Sat Jul 24, 2021 1:18 pm
longinvest wrote: Sat Jul 24, 2021 1:12 pm
nigel_ht wrote: Sat Jul 24, 2021 1:12 pm Has income varied by 50%?
Yes, it has dropped by even more. We adapted.
For a decade?
Dear Nigel_ht,

Are you talking about the hypothetical Japanese 4% SWR retiree who died exactly 30 years after retirement and only got 3 withdrawals, then nothing for 27 years?*

* Source.

Best regards,

longinvest
Variable Percentage Withdrawal (bogleheads.org/wiki/VPW) | One-Fund Portfolio (bogleheads.org/forum/viewtopic.php?t=287967)
zuma
Posts: 641
Joined: Thu Dec 29, 2016 11:15 am

Re: Variable Percentage Withdrawal (VPW) - Danger Will Robinson!

Post by zuma »

nigel_ht wrote: Sat Jul 24, 2021 1:12 pm So I’m not banking on my theoretical ability to adapt to large unplanned withdrawal fluctuations to meet my planned expenses when I already know that the error bars for my estimated retirement expenses are very large if we run into unexpected health issues.
Then maybe VPW isn't for you?
reln
Posts: 718
Joined: Fri Apr 19, 2019 4:01 pm

Re: Variable Percentage Withdrawal (VPW) - Danger Will Robinson!

Post by reln »

Researchers generally agree that a VPW is better than constant (real or nominal) withdrawals. A further improvement also takes into account life expectancy.

All withdrawal methods can fail but constant spending methods are more likely to fail than VPWs.
User avatar
Zardoz
Posts: 140
Joined: Thu Sep 24, 2020 12:25 am

Re: Variable Percentage Withdrawal (VPW)

Post by Zardoz »

zuma wrote: Sat Jul 24, 2021 9:50 am I use VPW as a structured way to provide regular monthly income. With this system my income will be variable, as it depends on the market's performance. However, my budget (i.e. expected spending) is calculated separately and takes into account the possibility of a severe market crash. My spending requirements are flexible (I'm single, no kids, no debt) and VPW provides an automated way to earn more when the market is doing well.

Earning more doesn't mean that I must spend everything that I earn. I still save for larger unplanned expenses, extra travel, etc. This is the same thing I did while working a normal job. There seems to be assumption in your analysis that retirees must spend all of their income immediately. During a bear market, I can cut discretionary expenses and/or dip into savings to meet my needs.

I have nothing against SWR if that works for you. It's a more conservative, predictable approach to retirement income and is probably a better choice if your budget is less flexible. Personally, I prefer having the built-in adjustments of VPW.
This is the way I view VPW as well. Also, the VPW worksheet has been incredibly helpful because it allows me to see a reasonable range of spending for me, taking into account my portfolio, my age, and my expected Social Security benefits. As long as my basic spending needs could still be met under the "Stocks drop 50%" scenario shown in the worksheet, I feel comfortable with my finances in retirement.

For a few years I actually saw some of the discussions of VPW on this site and I largely ignored them, because I assumed that the VPW method was overly complex. But in fact it's not - it's simple and elegant. The SWR calculation as a rough rule of thumb combined with a low withdrawal rate may be effective enough for some people. However, I think VPW can allow those who are willing to embrace a somewhat flexible spending pattern to retire a little earlier and spend a little more during the early years of retirement.

I encourage everyone who is close to retirement age to do your own reading, starting with https://www.bogleheads.org/wiki/Variabl ... withdrawal and other threads on the forum until you have a firm understanding of the design behind the worksheet. Then try out the worksheet and the backtest spreadsheet, and draw your own conclusion.
Withdrawal Phase Plan: Equities <= 50% | TIPS, I Bonds | VPW Worksheet | TPAW | Social Security @70
SnowBog
Posts: 4700
Joined: Fri Dec 21, 2018 10:21 pm

Re: Variable Percentage Withdrawal (VPW)

Post by SnowBog »

nigel_ht wrote: Sat Jul 24, 2021 12:55 pm Unless your job was commission or tips based or based on personal business profits you likely didn’t have very large fluctuations in income that requires you to adjust to.
You forgot those with bonuses, stock options, RSUs, ESPP, those part of start-ups, those who have an equity/partner stake in the business, and a whole host of other forms of variable income such as side hustles. While this might still be a small segment of the population, it's not an insignificant segment...

In the past 5 years, I've had massive "variability" in my income due to some of these factors. One year had income that was roughly double the year prior, and which has not been as high since...

And in each of those years, we adapted our spending and savings accordingly. We'll do the same during retirement...

And for what it's worth, we won't be retiring when VPW (or SWR or any other method) just barely shows enough for us to retire... We'll retire when the models show that we'll have "enough" to live the type of retirement we want, which includes lots of discretionary expenses. And guess what - those expenses can be reduced or eliminated if markets are really bad - just like we did while working...
nigel_ht wrote: Sat Jul 24, 2021 12:55 pm ...
Like I said, if it works for you great. But VPW does have risks if you spend more in the first few years and run into SORR…you are forced to decide whether to increase your spend which risks portfolio depletion fast than it may have been with SWR or dropping below you minimum spend and “selling your house” to achieve the flexibility required.
SORR is a risk - regardless of withdrawal methods. SWR is not "immune" to it... And VPW is not "more at risk" to it...

As others have pointed out, if you really want to avoid SORR and have guaranteed income, then you may want to consider something like a Liability Matching Portfolio (LMP) often constructed with inflation adjusted sources like I Bonds or TIPS.
nigel_ht wrote: Sat Jul 24, 2021 12:55 pm The key issue is that folks where 4% spend is marginal to meet their spending needs will see VPW as a “safe” way to spend more than 4% and the risk of failure is obscured because VPW claims it never depletes before the end date.

If the user really needs 4.5% to meet their needs including SS this isn’t easily seen as it is with SWR and the computed historical failure calculations.
Your comment makes no sense to me...

My working theory is your are implying that someone trying to figure out if they can retire would (made up numbers for example):
  • Figure out their annual expenses - let's say $50k
  • Look at SWR of 4% and see it could only support $49k/year...
  • So they look at VPW and see it could support $51k in the first year, and they fail to pay attention to the "required flexibility" which shows it might only support $26k in some years...
  • So they blindly go ahead and retire using VPW only to find out they never saved enough (as they clearly didn't understand VPW at all...)
That is illogical to me... Anyone using VPW is expected to understand that the "income" (aka withdrawals) from the portfolio are variable. Anyone who fails to understand that, is missing the point of VPW...

Unfortunately, I think this includes you... You've stated a few times how VPW is supposed to "guarantee minimumal income", that it's a "bug" that it does not... Again, to me that's illogical...

Other than an LMP and/or an income source like SPIA, pension, social security (and some of these are questionable) - there is no "guaranteed" income... SWR does not magically give you "guaranteed" withdrawals...

Sure you can find an SWR % (say < 3%) that has never failed a backtest in meeting your "minimum" expenses. It's highly likely - but not guaranteed - that it will continue to do so going forward. But as others, such as longinvest have pointed out - you are trading one (in your mind) "risk" of not having enough income in a given year for other risks such as a higher risk of either dying ultra rich or broke (SWR is rarely in the middle).

Those of us that understand VPW accept what it is, and isn't, and we use it (or plan to use it) accordingly.
Last edited by SnowBog on Sat Jul 24, 2021 5:52 pm, edited 1 time in total.
marcopolo
Posts: 8446
Joined: Sat Dec 03, 2016 9:22 am

Re: Variable Percentage Withdrawal (VPW)

Post by marcopolo »

nigel_ht wrote: Sat Jul 24, 2021 9:07 am
longinvest wrote: Sat Jul 24, 2021 8:56 am Dear Nigel_ht,

This long (and historical) thread is about the Variable Percentage Withdrawal (VPW) method and retirement approach, its objectives, its design, and for helping its users on a general level (personal issues being best discussed in user-specific threads in the Personal Investments forum).

There are already many other threads for discussing the choice of a withdrawal method for retirement and a new thread about it can also be started.

The fact that I commented on an unfair comparison between VPW and SWR doesn't justify derailing this thread.

Best regards,

longinvest
My points relate to failure modes of VPW and should be on topic.

I modified the backtest spreadsheet to add a floor calculation. Not entirely sure if I did it all correctly (did it during a swim meet) but if interested I can send you the modified excel.

For 1965 a 4% SWR didn't fail (35% US 25% INT 40% Bonds) but with a 3.8% floor (5% flexibility) VPW did fail because of a 12 year stretch of overspending.

Image

So this is a historical scenario where overspending in the early years resulted in increased risk of failure vs SWR. VPW would require more than 5% flexibility to not run out of money while meeting the estimated minimum retirement spending without other sources of income.

Now granted, if you put SS into the mix then VPW doesn't fail. But neither would SWR in the historical failure scenarios. For a "normal" 30 year retirement that lasts from 30 years to age 100 your retirement age is 70. If you retire earlier then 30 years won't cover until 100.

A retirement at "full retirement age of 65-67" has only 3-5 years before SS kicks in even if you delay until 70.

This kind of analysis is more critical for those retiring say 55 or younger…we’re more likely to need a longer than 30 year retirement than those retiring at 67.

—-

Edit: 5% not 20%

I am not a huge fan of VPW. If you want to use the general idea, an amortization approach like ABW or TPAW, seems like a better option.

Having said that, your example of 5% flexibility seems quite restrictive. Most people's spending probably varies more than that under normal circumstances. During lean years, ability to cut expenses is likely significantly more than that.

If you prefer such a strict spending profile, perhaps a withdrawal method with the word "variable" right in its name is not appropriate for you.
Once in a while you get shown the light, in the strangest of places if you look at it right.
User avatar
Tejfyy
Posts: 224
Joined: Mon Aug 26, 2019 9:18 pm

Re: Variable Percentage Withdrawal (VPW) - Danger Will Robinson!

Post by Tejfyy »

I have this topic bookmarked and check in every now and then, that is every month or so. I'm using the VPW in "early retirement" after long hours spent understanding it initially and then via the threads that ensue here. I'm consistently reassured of its validity and reliability as a tool to guide me through the 30+ remaining years of life. Much of that reassurance comes from @longinvest's patient discussions with others, not so much those about the tool itself but about the bigger picture, life as it were. Once you've got your ship in order, there's something to be said for floating more and steering less, or as someone who'd been through hell and back once said to me: life isn't an enema, stop.
User avatar
NearlyRetired
Posts: 350
Joined: Fri Aug 09, 2019 12:56 pm
Location: United Kingdom

Re: Variable Percentage Withdrawal (VPW) - Danger Will Robinson!

Post by NearlyRetired »

For me it is quite simple.

The 4% "rule" is based on historical analysis (which BTW only considers investment performance and must be adjusted for fees and inflation). This is a backward looking calculation and assumes that the past describes a likely future. This may, or may not, be accurate.

VPW and other Amortization mechanisms (not sure if VPW is an Amortization technique) at least rely on current valuations and mathematically calculate an amount based on "today". It may be variable, but one could argue it's a better sign post to spending ability.

At least VPW gives you a feel for what you may have to reduce you income to if the market drops (and if you can live with that then all is well - if not then you need to rethink).

I would rather have an adaptive formula rather than something that suggests based on past performance, a figure will work.

For myself, I have an amount I need to spend (split between must spend and would like to spend). I will always go the the "would like to spend" figure , but also compare that against other withdrawal calculators for a sanity check
To err is to be human, to really mess up, use a computer
Topic Author
nigel_ht
Posts: 4742
Joined: Tue Jan 01, 2019 9:14 am

Re: Variable Percentage Withdrawal (VPW) - Danger Will Robinson!

Post by nigel_ht »

longinvest wrote: Sat Jul 24, 2021 1:21 pm
nigel_ht wrote: Sat Jul 24, 2021 1:18 pm
longinvest wrote: Sat Jul 24, 2021 1:12 pm
nigel_ht wrote: Sat Jul 24, 2021 1:12 pm Has income varied by 50%?
Yes, it has dropped by even more. We adapted.
For a decade?
Dear Nigel_ht,

Are you talking about the hypothetical Japanese 4% SWR retiree who died exactly 30 years after retirement and only got 3 withdrawals, then nothing for 27 years?*

* Source.

Best regards,

longinvest
Sure, we could using the Nikkei instead of S&P 500 at a 60/40 with 35% Nikkei and 25% US or ex-Japan if those numbers are handy.

1937 Japan is an outlier for Japan…unless you think Washington DC and other major cities in the US will be firebombed and Austin (11th largest city like Hiroshima) and Atlanta (38th) will get nuked in the next few years.

So 1990 Japanese bubble would be a good case study.

Comparing apples to apples is the best policy don’t you think?

The other thing is that 4% SWR isn’t 4% but 3.8% for 30 years. In the case of 1965, VPW fails if the there is a floor applied whereas SWR does not.
Topic Author
nigel_ht
Posts: 4742
Joined: Tue Jan 01, 2019 9:14 am

Re: Variable Percentage Withdrawal (VPW) - Danger Will Robinson!

Post by nigel_ht »

reln wrote: Sat Jul 24, 2021 1:42 pm Researchers generally agree that a VPW is better than constant (real or nominal) withdrawals. A further improvement also takes into account life expectancy.

All withdrawal methods can fail but constant spending methods are more likely to fail than VPWs.
Only if you assume there is no minimum spending.

Which is an incorrect assertion.
Topic Author
nigel_ht
Posts: 4742
Joined: Tue Jan 01, 2019 9:14 am

Re: Variable Percentage Withdrawal (VPW) - Danger Will Robinson!

Post by nigel_ht »

longinvest wrote: Sat Jul 24, 2021 1:12 pm
nigel_ht wrote: Sat Jul 24, 2021 1:12 pm Has income varied by 50%?
Dear Nigel_ht,

Yes, it has dropped by even more. We adapted.

We all know people who lost their job, or had increased expenses while income dropped (e.g. having children), or divorced, or had many other major changes in their life situation putting them into a new financial situation.

Best regards,

longinvest
Ah, missed your edit. A lot of times these have resulted in “selling your house” which you used as an example of flexibility…
Topic Author
nigel_ht
Posts: 4742
Joined: Tue Jan 01, 2019 9:14 am

Re: Variable Percentage Withdrawal (VPW) - Danger Will Robinson!

Post by nigel_ht »

zuma wrote: Sat Jul 24, 2021 1:29 pm
nigel_ht wrote: Sat Jul 24, 2021 1:12 pm So I’m not banking on my theoretical ability to adapt to large unplanned withdrawal fluctuations to meet my planned expenses when I already know that the error bars for my estimated retirement expenses are very large if we run into unexpected health issues.
Then maybe VPW isn't for you?
For me it doesn’t matter. We have saved enough that almost anything reasonable works based on our past spending patterns. Our minimum SWR amount, barring left tail health expenses, is less than 3%.
Topic Author
nigel_ht
Posts: 4742
Joined: Tue Jan 01, 2019 9:14 am

Re: Variable Percentage Withdrawal (VPW) - Danger Will Robinson!

Post by nigel_ht »

longinvest wrote: Sat Jul 24, 2021 1:10 pm Guaranteed income can be bought, it's called an inflation-indexed SPIA* (joint life, for a couple).

* Single Premium Immediate Annuity.
If you can find one the costs will be extremely high in comparison to nominal annuities or 2% annuities…

Given there is little competition for inflation indexed SPIAs there’s no real incentive to offer better.

SPIAs have a place in retirement planning. I’m not sure that one should depend on them existing in a form or coat you like when retirement comes unless retirement is in the near future.

That you may not be able to afford (or want to pay for) a hugely overpriced product doesn’t necessarily mean you didn’t save enough. Just that it’s a terrible product.
Topic Author
nigel_ht
Posts: 4742
Joined: Tue Jan 01, 2019 9:14 am

Re: Variable Percentage Withdrawal (VPW) - Danger Will Robinson!

Post by nigel_ht »

NearlyRetired wrote: Sat Jul 24, 2021 2:18 pm For me it is quite simple.

The 4% "rule" is based on historical analysis (which BTW only considers investment performance and must be adjusted for fees and inflation). This is a backward looking calculation and assumes that the past describes a likely future. This may, or may not, be accurate.

VPW and other Amortization mechanisms (not sure if VPW is an Amortization technique) at least rely on current valuations and mathematically calculate an amount based on "today". It may be variable, but one could argue it's a better sign post to spending ability.

At least VPW gives you a feel for what you may have to reduce you income to if the market drops (and if you can live with that then all is well - if not then you need to rethink).
How does it do so without using historical analysis or Monte Carlo?
I would rather have an adaptive formula rather than something that suggests based on past performance, a figure will work.

For myself, I have an amount I need to spend (split between must spend and would like to spend). I will always go the the "would like to spend" figure , but also compare that against other withdrawal calculators for a sanity check
How do you know how often the VPW has historically dropped you below your must spend figure?

A table should be easy to generate but the current backtest spreadsheet doesn’t appear to take the starting values and run it through all the years in the dataset to tell you that historically the minimum withdrawal was $X.

Perhaps that information is available some place.
Topic Author
nigel_ht
Posts: 4742
Joined: Tue Jan 01, 2019 9:14 am

Re: Variable Percentage Withdrawal (VPW)

Post by nigel_ht »

marcopolo wrote: Sat Jul 24, 2021 1:59 pm
I am not a huge fan of VPW. If you want to use the general idea, an amortization approach like ABW or TPAW, seems like a better option.

Having said that, your example of 5% flexibility seems quite restrictive. Most people's spending probably varies more than that under normal circumstances. During lean years, ability to cut expenses is likely significantly more than that.

If you prefer such a strict spending profile, perhaps a withdrawal method with the word "variable" right in its name is not appropriate for you.
Yes 5% is rather restrictive. Of course someone with 25% flexibility could use a 3% SWR as a baseline…or carve out a portfolio amount the meets the minimum estimated spending rate with a SWR and flexibility spend the rest.

The example I used before was someone who only needed $30K to meet expected expenses with a $1M portfolio could do a 3.8% SWR on $800K and then have $200K to flexibly spend as they like over their first decade. After which SS likely isn’t that far away.

If you really have a lot of flexibility between need and want then nearly any reasonable withdrawal strategy works.

I haven’t run the numbers for ABW or TPAW yet so I won’t comment on them.
Seasonal
Posts: 3045
Joined: Sun May 21, 2017 1:49 pm

Re: Variable Percentage Withdrawal (VPW)

Post by Seasonal »

nigel_ht wrote: Sat Jul 24, 2021 9:06 amIf the method does not guarantee your minimum expenses in a historical worst case scenario then you have a known failure percentage based on historical outcomes.
This is tautologically true. The real issue for retirees is the future and history may not be a great guide to the future. History provides some guidance, but is far from a guarantee. This is especially true at a time of very low interest rates and very high valuations.

If 4% inflation adjusted fails, it fails spectacularly. The usual advice is to adjust spending if this method appears likely to fail, which just brings you to variable income that may not meet your spending goals. A planned variable income method, such as VPW, won't fail and has the virtue of providing guidance throughout, rather than the seat of the pants adjustments that may be necessary with 4% inflation adjusted.
Topic Author
nigel_ht
Posts: 4742
Joined: Tue Jan 01, 2019 9:14 am

Re: Variable Percentage Withdrawal (VPW)

Post by nigel_ht »

Seasonal wrote: Sat Jul 24, 2021 3:16 pm
nigel_ht wrote: Sat Jul 24, 2021 9:06 amIf the method does not guarantee your minimum expenses in a historical worst case scenario then you have a known failure percentage based on historical outcomes.
This is tautologically true. The real issue for retirees is the future and history may not be a great guide to the future. History provides some guidance, but is far from a guarantee. This is especially true at a time of very low interest rates and very high valuations.
I guess my point is that with SWR we know what that failure rate is. With VPW, unless there is a table of minimum withdrawals you don’t know how flexible you need to be in the worst case.

Nor do you know how often VPW fails if you have to withdraw at a rate above the computed rate to meet that minimum spend.
If 4% inflation adjusted fails, it fails spectacularly.
Well, given it’s the most conservative for 30 years (well, 3.8) you have a worse than historical outcome. So if you have used VPW and spent more than you would have under SWR then your portfolio is less likely to survive the scenario OR your lifestyle will need a significant adjustment OR both.

The other interesting aspect is that you have a fair probability of not making it to the end of your 30 year period in a normal retirement (age 65)…
A planned variable income method, such as VPW, won't fail and has the virtue of providing guidance throughout, rather than the seat of the pants adjustments that may be necessary with 4% inflation adjusted.
What guidance does VPW provide if you have to spend more than the computed amount?
4nursebee
Posts: 2664
Joined: Sun Apr 01, 2012 7:56 am
Location: US

Re: Variable Percentage Withdrawal (VPW) - Danger Will Robinson!

Post by 4nursebee »

The alarmist title is kinda click baity isn't it OP?

It is kind of a basis of all models that if you spend more than you have you will fail.

Is this the only scenario you ran or did you just choose the one that failed?

We are using VPW. To help with not spending too much early on if times were bad we built a two year expenses emergency fund. I believe this would assist in the example you provide.
Pale Blue Dot
User avatar
Nestegg_User
Posts: 2112
Joined: Wed Aug 05, 2009 1:26 pm

Re: Variable Percentage Withdrawal (VPW) - Danger Will Robinson!

Post by Nestegg_User »

sorry, but I think that you have set up a strawman

the genesis of VPW, if I recall, was to allow a potentially higher spend than a static wr (which was based upon initial retirement conditions)....and all that was an outgrowth of an earlier thread examining potential retirement withdrawal possibilities (SWR, Guyton-Klinger, VPW or other variable withdrawal methods (AMT, etc).

I remember that the method requires that the prospective user examine the potential outcome for a drop of X% (they might consider 50% or even higher) to determine application___ if that's below threshold then they need to push some of their portfolio into a more guaranteed income, whether it be a SPIA (often the most preferred route, and usually one with staggered purchases), a MYGA for short duration (which can be replicated over one's needed period, i.e. one can establish a new one every say four years, as wanted), or a I bond or other bond (to maturity so as to not be at risk of loss). Otherwise, your potential retiree is in the "grey zone" per an early paper and might be cautioned that they are at risk under most all withdrawal conditions.
That you are noting the most obvious just reinforces that; the prospective retiree needs to have a higher level of savings __no amount of shell games is gonna change that, whether it's SWR or not.
We've used a modified Guyton-Klinger to examine maximum safe spending, but haven't come close to that spend...and probably won't as I'll be able to start my SS (already well past eligibility) if we need more income (spouse has slightly higher PIA... I'm delaying so as to do some Roth conversions).
GK has less variation than VPW, but other variable methods have variations that can be limited by setting up limits to either "income" (spending) growth on the upside so as to mitigate spending drops when markets aren't good.
The classic "what happens when markets drop 50+% and stay down for two decades" has been asked countless times (how often has that occurred?? I believe that's a null set)... those with any savings at all will be best prepared...ain't going to the G&B mentality.
Normchad
Posts: 5648
Joined: Thu Mar 03, 2011 6:20 am

Re: Variable Percentage Withdrawal (VPW)

Post by Normchad »

Zardoz wrote: Sat Jul 24, 2021 1:43 pm
zuma wrote: Sat Jul 24, 2021 9:50 am I use VPW as a structured way to provide regular monthly income. With this system my income will be variable, as it depends on the market's performance. However, my budget (i.e. expected spending) is calculated separately and takes into account the possibility of a severe market crash. My spending requirements are flexible (I'm single, no kids, no debt) and VPW provides an automated way to earn more when the market is doing well.

Earning more doesn't mean that I must spend everything that I earn. I still save for larger unplanned expenses, extra travel, etc. This is the same thing I did while working a normal job. There seems to be assumption in your analysis that retirees must spend all of their income immediately. During a bear market, I can cut discretionary expenses and/or dip into savings to meet my needs.

I have nothing against SWR if that works for you. It's a more conservative, predictable approach to retirement income and is probably a better choice if your budget is less flexible. Personally, I prefer having the built-in adjustments of VPW.
This is the way I view VPW as well. Also, the VPW worksheet has been incredibly helpful because it allows me to see a reasonable range of spending for me, taking into account my portfolio, my age, and my expected Social Security benefits. As long as my basic spending needs could still be met under the "Stocks drop 50%" scenario shown in the worksheet, I feel comfortable with my finances in retirement.

For a few years I actually saw some of the discussions of VPW on this site and I largely ignored them, because I assumed that the VPW method was overly complex. But in fact it's not - it's simple and elegant. The SWR calculation as a rough rule of thumb combined with a low withdrawal rate may be effective enough for some people. However, I think VPW can allow those who are willing to embrace a somewhat flexible spending pattern to retire a little earlier and spend a little more during the early years of retirement.

I encourage everyone who is close to retirement age to do your own reading, starting with https://www.bogleheads.org/wiki/Variabl ... withdrawal and other threads on the forum until you have a firm understanding of the design behind the worksheet. Then try out the worksheet and the backtest spreadsheet, and draw your own conclusion.
I’m right there with you. I ignored the VPW posts for years.

Once I really looked into it, I found it incredibly helpful. I can’t thank longinvest enough for creating the spreadsheet and spending so much time helping answer peoples questions about it.

Eventually, each of us will have to decide what we are going do. How we are gonna to handle this. Different people will make different decisions. There is no guaranteed not up fail approach.

I really like VPW. I like that it lets me include my SS in the model. I won’t use it verbatim though, but I do anticipate updating it frequently and using it as a great tool to sanity check myself.

As with any approach though, people need to understand what they are doing, and what the pitfalls might be. If you use VPW, and don’t know what “variable” part is, that’s on you. (Especially since the sheet explicitly shows you what a steep decline means for withdrawals). But that’s true for ABW, SWR, annuities, etc. there is no silver bullet.
Topic Author
nigel_ht
Posts: 4742
Joined: Tue Jan 01, 2019 9:14 am

Re: Variable Percentage Withdrawal (VPW) - Danger Will Robinson!

Post by nigel_ht »

4nursebee wrote: Sat Jul 24, 2021 3:45 pm The alarmist title is kinda click baity isn't it OP?
Lol…maybe.
It is kind of a basis of all models that if you spend more than you have you will fail.

Is this the only scenario you ran or did you just choose the one that failed?
It would be better to run the whole set of data but I just added a few columns to the existing backtest spreadsheet.
SnowBog
Posts: 4700
Joined: Fri Dec 21, 2018 10:21 pm

Re: Variable Percentage Withdrawal (VPW) - Danger Will Robinson!

Post by SnowBog »

Looks like the threads have been merged...

For anyone interested, arguably this thread "started" with OP here: viewtopic.php?t=353481 asking "Is VPW/ABW a frugal retirement method?"

Candidly, that thread was hard to follow...

Anyway, there was a cross post into the long standing VPW thread, which shifted some of this discussion over there.

That was split out into its own thread here...
Last edited by SnowBog on Sat Jul 24, 2021 5:56 pm, edited 1 time in total.
marcopolo
Posts: 8446
Joined: Sat Dec 03, 2016 9:22 am

Re: Variable Percentage Withdrawal (VPW) - Danger Will Robinson!

Post by marcopolo »

NearlyRetired wrote: Sat Jul 24, 2021 2:18 pm For me it is quite simple.

The 4% "rule" is based on historical analysis (which BTW only considers investment performance and must be adjusted for fees and inflation). This is a backward looking calculation and assumes that the past describes a likely future. This may, or may not, be accurate.

VPW and other Amortization mechanisms (not sure if VPW is an Amortization technique) at least rely on current valuations and mathematically calculate an amount based on "today". It may be variable, but one could argue it's a better sign post to spending ability.

At least VPW gives you a feel for what you may have to reduce you income to if the market drops (and if you can live with that then all is well - if not then you need to rethink).

I would rather have an adaptive formula rather than something that suggests based on past performance, a figure will work.

For myself, I have an amount I need to spend (split between must spend and would like to spend). I will always go the the "would like to spend" figure , but also compare that against other withdrawal calculators for a sanity check

I think you are giving VPW too much credit for taking current conditions (valuations) into account.

If I recall correctly, it uses a fixed "portfolio growth" number that seems to have been pulled outbof think air, and does not take into account any such thing as current valuations. Can you point to something describing how it is doing anything more complicated than that based on valuations?

It does take into account your current portfolio value.
Once in a while you get shown the light, in the strangest of places if you look at it right.
User avatar
LadyGeek
Site Admin
Posts: 95696
Joined: Sat Dec 20, 2008 4:34 pm
Location: Philadelphia
Contact:

Re: Is VPW/ABW a frugal retirement strategy?

Post by LadyGeek »

Upon further review (and requested by a member), I merged nigel_ht's thread into an earlier similar discussion.
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.
Seasonal
Posts: 3045
Joined: Sun May 21, 2017 1:49 pm

Re: Variable Percentage Withdrawal (VPW)

Post by Seasonal »

nigel_ht wrote: Sat Jul 24, 2021 3:29 pm
Seasonal wrote: Sat Jul 24, 2021 3:16 pm
nigel_ht wrote: Sat Jul 24, 2021 9:06 amIf the method does not guarantee your minimum expenses in a historical worst case scenario then you have a known failure percentage based on historical outcomes.
This is tautologically true. The real issue for retirees is the future and history may not be a great guide to the future. History provides some guidance, but is far from a guarantee. This is especially true at a time of very low interest rates and very high valuations.
I guess my point is that with SWR we know what that failure rate is. With VPW, unless there is a table of minimum withdrawals you don’t know how flexible you need to be in the worst case.
We don't know that the failure rate will be. We know what it has been.
nigel_ht wrote: Sat Jul 24, 2021 3:29 pmNor do you know how often VPW fails if you have to withdraw at a rate above the computed rate to meet that minimum spend.
That's true of any withdrawal method, including 4% SWR and VPW.
reln
Posts: 718
Joined: Fri Apr 19, 2019 4:01 pm

Re: Variable Percentage Withdrawal (VPW) - Danger Will Robinson!

Post by reln »

nigel_ht wrote: Sat Jul 24, 2021 2:25 pm
reln wrote: Sat Jul 24, 2021 1:42 pm Researchers generally agree that a VPW is better than constant (real or nominal) withdrawals. A further improvement also takes into account life expectancy.

All withdrawal methods can fail but constant spending methods are more likely to fail than VPWs.
Only if you assume there is no minimum spending.

Which is an incorrect assertion.
You're incorrect. Even if there is no minimum spending set in the withdrawal method (if breached, the scenario fails), the VPWs are superior to constant spending. In other words VPWs have better outcomes over more scenarios. This doesn't mean VPW has a better outcome in every scenario. You just happen to find 1 scenario in which it fails and constant spending succeeds; and you have extrapolated too far from that 1 example.
FactualFran
Posts: 2777
Joined: Sat Feb 21, 2015 1:29 pm

Re: Variable Percentage Withdrawal (VPW) - Danger Will Robinson!

Post by FactualFran »

NearlyRetired wrote: Sat Jul 24, 2021 2:18 pm The 4% "rule" is based on historical analysis (which BTW only considers investment performance and must be adjusted for fees and inflation). This is a backward looking calculation and assumes that the past describes a likely future. This may, or may not, be accurate
The "4% rule" is based on the analysis of historical returns and inflation. There is no need for an additional adjustment for inflation. The recent fees of index funds have been very small relative to the index returns that were used in the analysis. Such index funds were not available during the earlier years of the historical returns that were used in the analysis.

There is no overall assumption that the past describes a likely future. However, someone who uses an initial withdrawal rate equal to the maximum rate that did not deplete a portfolio using historical inflation-adjusted returns is depending on future sequences of inflation-adjusted returns being no worse than the worst sequence in the historical returns.
reln
Posts: 718
Joined: Fri Apr 19, 2019 4:01 pm

Re: Variable Percentage Withdrawal (VPW) - Danger Will Robinson!

Post by reln »

FactualFran wrote: Sat Jul 24, 2021 8:49 pm
NearlyRetired wrote: Sat Jul 24, 2021 2:18 pm The 4% "rule" is based on historical analysis (which BTW only considers investment performance and must be adjusted for fees and inflation). This is a backward looking calculation and assumes that the past describes a likely future. This may, or may not, be accurate
The "4% rule" is based on the analysis of historical returns and inflation. There is no need for an additional adjustment for inflation. The recent fees of index funds have been very small relative to the index returns that were used in the analysis. Such index funds were not available during the earlier years of the historical returns that were used in the analysis.

There is no overall assumption that the past describes a likely future. However, someone who uses an initial withdrawal rate equal to the maximum rate that did not deplete a portfolio using historical inflation-adjusted returns is depending on future sequences of inflation-adjusted returns being no worse than the worst sequence in the historical returns.
The 4% rule failed in most developed countries. It only worked in USA and Canada, historically. A more global "x% rule" would be closer to 2.8%, historically.

We just happen to refer only too "the worst sequence in the historical returns" of the *most successful* stock market in the historical returns. IMO, that violates the presumptive "conservative assumptions" the 4% rule is based on.

VPW based on assets and life expectancy makes more sense.
User avatar
Nestegg_User
Posts: 2112
Joined: Wed Aug 05, 2009 1:26 pm

Re: Variable Percentage Withdrawal (VPW) - Danger Will Robinson!

Post by Nestegg_User »

reln wrote: Sat Jul 24, 2021 9:12 pm
FactualFran wrote: Sat Jul 24, 2021 8:49 pm
NearlyRetired wrote: Sat Jul 24, 2021 2:18 pm The 4% "rule" is based on historical analysis (which BTW only considers investment performance and must be adjusted for fees and inflation). This is a backward looking calculation and assumes that the past describes a likely future. This may, or may not, be accurate
The "4% rule" is based on the analysis of historical returns and inflation. There is no need for an additional adjustment for inflation. The recent fees of index funds have been very small relative to the index returns that were used in the analysis. Such index funds were not available during the earlier years of the historical returns that were used in the analysis.

There is no overall assumption that the past describes a likely future. However, someone who uses an initial withdrawal rate equal to the maximum rate that did not deplete a portfolio using historical inflation-adjusted returns is depending on future sequences of inflation-adjusted returns being no worse than the worst sequence in the historical returns.
The 4% rule failed in most developed countries. It only worked in USA and Canada, historically. A more global "x% rule" would be closer to 2.8%, historically.

We just happen to refer only too "the worst sequence in the historical returns" of the *most successful* stock market in the historical returns. IMO, that violates the presumptive "conservative assumptions" the 4% rule is based on.

VPW based on assets and life expectancy makes more sense.
you might want to correct that__see Estrada (2018) J of Retirement, "Maximum Withdrawal Rates: An Empirical and Global Perspective" wherein most of those involved in the WW2 conflict (invaded or were invaded) had low P5 withdrawal rates (5% probably of failure, (i'm quoting only the same rate as the 4% rule) for a 60/40 portfolio) whereas others had equal or close to the 4% level [Australia 3.7%, Canada 4.1%, Denmark 4.0% (surprisingly), Netherlands 3.6% (also surprisingly), New Zealand 4.0%, South Africa 3.9%... but then there's Sweden 3.3%, Switzerland 3.0% (surprisingly), and with the UK down at 3.4%]. These certainly are well above your 2.8%... but you are probably looking at countries that weren't as well developed, like Portugal or Ireland which neither had a strong economy in earlier years. (Estrada had another paper that basically showed that the level you are thinking about is closer to 3.5%, so indeed lower than 4% but not by as large of a variance. That's why, in part, that we use a 3.5% value in our planning (now in our retirement, still use it)...as it places less importance on "US exceptionalism". )
Topic Author
nigel_ht
Posts: 4742
Joined: Tue Jan 01, 2019 9:14 am

Re: Variable Percentage Withdrawal (VPW)

Post by nigel_ht »

Seasonal wrote: Sat Jul 24, 2021 7:20 pm
nigel_ht wrote: Sat Jul 24, 2021 3:29 pm
Seasonal wrote: Sat Jul 24, 2021 3:16 pm
nigel_ht wrote: Sat Jul 24, 2021 9:06 amIf the method does not guarantee your minimum expenses in a historical worst case scenario then you have a known failure percentage based on historical outcomes.
This is tautologically true. The real issue for retirees is the future and history may not be a great guide to the future. History provides some guidance, but is far from a guarantee. This is especially true at a time of very low interest rates and very high valuations.
I guess my point is that with SWR we know what that failure rate is. With VPW, unless there is a table of minimum withdrawals you don’t know how flexible you need to be in the worst case.
We don't know that the failure rate will be. We know what it has been.
You don’t know what the historical failure rate is for VPW.

You do know the historical failure rate for 4% SWR.

If numbers exist for VPW that would be awesome.

What the lowest historical withdrawal was. How long VPW can stay below 10/20/30% of the original withdrawal amount, etc.

How flexible do you have to be for how long.

How often VPW fails if you have a minimum expense amount and for how long. Does it ever deplete the portfolio.
Topic Author
nigel_ht
Posts: 4742
Joined: Tue Jan 01, 2019 9:14 am

Re: Variable Percentage Withdrawal (VPW) - Danger Will Robinson!

Post by nigel_ht »

reln wrote: Sat Jul 24, 2021 8:30 pm
nigel_ht wrote: Sat Jul 24, 2021 2:25 pm
reln wrote: Sat Jul 24, 2021 1:42 pm Researchers generally agree that a VPW is better than constant (real or nominal) withdrawals. A further improvement also takes into account life expectancy.

All withdrawal methods can fail but constant spending methods are more likely to fail than VPWs.
Only if you assume there is no minimum spending.

Which is an incorrect assertion.
You're incorrect. Even if there is no minimum spending set in the withdrawal method (if breached, the scenario fails), the VPWs are superior to constant spending. In other words VPWs have better outcomes over more scenarios. This doesn't mean VPW has a better outcome in every scenario. You just happen to find 1 scenario in which it fails and constant spending succeeds; and you have extrapolated too far from that 1 example.
Since you know I’m incorrect it should be simple to show me how often minimum spending amounts are breached, for how long, etc.

How does VPW have a better outcome over more scenarios when SWR percentages are the maximum safe amount to withdraw historically? A SWR value isn’t really “safe” unless the historical failure rate is 0…for a 30 year that has been 3.8%.

I showed only one example because the backtest spreadsheet doesn’t appear to make it easy to evaluate every 30 cohort in one pass. Given you know I’m wrong presumably you can show me the results for every 30 year period from 1871 onward…
Topic Author
nigel_ht
Posts: 4742
Joined: Tue Jan 01, 2019 9:14 am

Re: Variable Percentage Withdrawal (VPW) - Danger Will Robinson!

Post by nigel_ht »

Nestegg_User wrote: Sat Jul 24, 2021 3:49 pm sorry, but I think that you have set up a strawman
Why? I would just like to see the outcomes across the historical data set…how is that a strawman?
the genesis of VPW, if I recall, was to allow a potentially higher spend than a static wr
Which is great…but it’s still useful to know failure modes and frequency for historical scenarios.
I remember that the method requires that the prospective user examine the potential outcome for a drop of X% (they might consider 50% or even higher) to determine application___
For how long? How often does this occur in the historical data set? There is a difference between a rapid drop and recovery like 2020 and one that last years.
SnowBog
Posts: 4700
Joined: Fri Dec 21, 2018 10:21 pm

Re: Is VPW/ABW a frugal retirement strategy?

Post by SnowBog »

https://calculator.ficalc.app/

I believe has data back to 1871, and you can use "constant dollar" (aka SWR) or VPW. That said, I'm not a fan of their VPW model - which uses the original table to completely exhaust portfolio by 100 (which is not what the longinvest model uses).

If I recall, one of your original scenarios was $1M 60/40, with $30k minimum income (adjusted for inflation) for 30 year retirement.

VPW shows 100% success https://calculator.ficalc.app?additiona ... gyName=vpw

SWR (4%) shows 95% success, with 8% "nearly failed", and > 16% "large ending portfolio) l: https://calculator.ficalc.app?additiona ... tantDollar

Personally, I'm not convinced the future is going to look exactly like the past... But to each their own...
SnowBog
Posts: 4700
Joined: Fri Dec 21, 2018 10:21 pm

Re: Variable Percentage Withdrawal (VPW) - Danger Will Robinson!

Post by SnowBog »

nigel_ht wrote: Sat Jul 24, 2021 11:49 pm For how long? How often does this occur in the historical data set? There is a difference between a rapid drop and recovery like 2020 and one that last years.
And since no one knows what the future holds... And it may not look like the past (at least not any specific sequence of returns)...

I'm not sure I understand the point - or the obsession with back testing... The point of VPW is it adapts regardless of what happens (but yes - might not meet your imposed minimum spending amount under adverse conditions). With SWR, you hope that you pick a safe SWR and the future does not deviate too much from the past... Either way, there's risk of running into a really bad sequence of returns... But the flip side is also possible, running into a great sequence of returns (as is everything in between).

But my prior post linked to a tool you might find useful to run whatever scenarios help you sleep at night.
Post Reply