Downsides of VPW?

Discuss all general (i.e. non-personal) investing questions and issues, investing news, and theory.
AlohaJoe
Posts: 4494
Joined: Mon Nov 26, 2007 2:00 pm
Location: Saigon, Vietnam

Re: Downsides of VPW?

Post by AlohaJoe » Tue Jun 04, 2019 10:58 pm

bhsince87 wrote:
Tue Jun 04, 2019 10:20 pm
Yes, I'd much prefer spending 28k at 75 and 80k at 95 than having nothing left to spend at, say, age 85.

And zero to spend at age 86, 87, 88, 89, ....
You don't have zero to spend. You have your Social Security + SPIA.

The actual comparison most real retirees face isn't "28k vs 40k @ 75 in order to have 80k vs 0k @ 85". It is "68k vs 80k @ 75 in order to have 120k vs 40k @ 85".

randomguy
Posts: 7677
Joined: Wed Sep 17, 2014 9:00 am

Re: Downsides of VPW?

Post by randomguy » Tue Jun 04, 2019 11:08 pm

bhsince87 wrote:
Tue Jun 04, 2019 10:20 pm
randomguy wrote:
Tue Jun 04, 2019 9:47 pm
bhsince87 wrote:
Tue Jun 04, 2019 9:11 pm
AlohaJoe wrote:
Tue Jun 04, 2019 8:55 pm
bhsince87 wrote:
Tue Jun 04, 2019 8:31 pm
I don't understand this argument. Can you explain it in more detai? Maybe provide an example?
Open up the VPW spreadsheet, go to the "Backtesting" tab, and change C19 "Start Year" to 1966.

Then look at I32, the inflation-adjusted withdrawal for the 10th year of retirement when you are 74 years old. It is $28,000.

Then look at I48, the inflation-adjusted withdrawal for the 27th year of retirement when you are 92 years old. It is $59,000.
Would you have prefered the the result of 0$ left with the 4% withdrawal method?

And I'm not being facecitious. Some people might prefer that approach. We do have a fairly strong safety net here, at least. Different strokes for different folks.
Lets say you prefer VPWs results. Do you still think spending 28k at 75 and 80k at 95 is a good result? Or is that a downside of using VPW? I consider it a major downside. You have to decide if the upsides outweigh it, Just like every other choice you make in life.
Yes, I'd much prefer spending 28k at 75 and 80k at 95 than having nothing left to spend at, say, age 85.

And zero to spend at age 86, 87, 88, 89, ....
You are answering a different question. The question is simply is reduced spending in your prime years an upside or a downside? I am guessing everyone considers it a downside. You might prefer that downside to the downsides of other schemes.

Personally I would take the steady 38k/year (actually something like 42k since Bengen eventually discovered small value;)) and never running out of money to scrimping in my prime years and eating caviar when I am old. You might feel differently:)

User avatar
willthrill81
Posts: 10157
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: Downsides of VPW?

Post by willthrill81 » Wed Jun 05, 2019 12:11 am

AlohaJoe wrote:
Tue Jun 04, 2019 10:58 pm
bhsince87 wrote:
Tue Jun 04, 2019 10:20 pm
Yes, I'd much prefer spending 28k at 75 and 80k at 95 than having nothing left to spend at, say, age 85.

And zero to spend at age 86, 87, 88, 89, ....
You don't have zero to spend. You have your Social Security + SPIA.
It's not even that bad. First, the idea that retirees will blindly spend their portfolio down to zero is a myth. Second, many retirees will never survive to age 85, although many here seem to be acting as though they will live to 120, despite those over the age of 70 only having gotten about one extra year of life expectancy per decade since the 1950s.
“It's a dangerous business, Frodo, going out your door. You step onto the road, and if you don't keep your feet, there's no knowing where you might be swept off to.” J.R.R. Tolkien,The Lord of the Rings

User avatar
4nursebee
Posts: 1252
Joined: Sun Apr 01, 2012 7:56 am
Location: US

Re: Downsides of VPW?

Post by 4nursebee » Wed Jun 05, 2019 3:16 am

gasdoc wrote:
Tue Jun 04, 2019 4:48 am
4nursebee wrote:
Tue Jun 04, 2019 4:23 am
I/we intend to use this method. I do not know the research into downsides, really have not paid attention to that. I view the downside as having to have enough cash or cash equivalents available to tolerate (psychologically and financially) small withdrawals if needed. Because of this we are growing this portion of our assets above all others. We are aiming for 2 years out of pocket expenses, not including healthcare or taxes.
I am glad you brought this up, as I've heard it before and have not really understood the principle. It seems like the concept of keeping more cash so that one can use cash rather than sell other investments during downswings is a type of market timing. How do you know when to use your cash, and know that by using your cash you are not delaying the selling of your equities to a time when the equities prices are even lower? Aren't you trying to time the market?

gasdoc
OK, some questions for you.
Start portfolio 100k
VPW says take 5% of port per year.
Take 5% out first year. ($5,000)
Market goes down 20%. (95K x .8 = 76K)
Take out 5% of port year two. (76K x .05 = 3.8 K)

1. Is it market timing to do the above?
2. Would you feel better having a large cash reserve?
For me, the answers are no and yes.

Or at least that is my read of the wiki for this method: https://www.bogleheads.org/wiki/Variabl ... withdrawal
4nursebee

randomguy
Posts: 7677
Joined: Wed Sep 17, 2014 9:00 am

Re: Downsides of VPW?

Post by randomguy » Wed Jun 05, 2019 6:22 am

willthrill81 wrote:
Wed Jun 05, 2019 12:11 am
AlohaJoe wrote:
Tue Jun 04, 2019 10:58 pm
bhsince87 wrote:
Tue Jun 04, 2019 10:20 pm
Yes, I'd much prefer spending 28k at 75 and 80k at 95 than having nothing left to spend at, say, age 85.

And zero to spend at age 86, 87, 88, 89, ....
You don't have zero to spend. You have your Social Security + SPIA.
It's not even that bad. First, the idea that retirees will blindly spend their portfolio down to zero is a myth. Second, many retirees will never survive to age 85, although many here seem to be acting as though they will live to 120, despite those over the age of 70 only having gotten about one extra year of life expectancy per decade since the 1950s.
It is actually 92-94 :)

dcabler
Posts: 950
Joined: Wed Feb 19, 2014 11:30 am

Re: Downsides of VPW?

Post by dcabler » Wed Jun 05, 2019 6:43 am

4nursebee wrote:
Wed Jun 05, 2019 3:16 am
gasdoc wrote:
Tue Jun 04, 2019 4:48 am
4nursebee wrote:
Tue Jun 04, 2019 4:23 am
I/we intend to use this method. I do not know the research into downsides, really have not paid attention to that. I view the downside as having to have enough cash or cash equivalents available to tolerate (psychologically and financially) small withdrawals if needed. Because of this we are growing this portion of our assets above all others. We are aiming for 2 years out of pocket expenses, not including healthcare or taxes.
I am glad you brought this up, as I've heard it before and have not really understood the principle. It seems like the concept of keeping more cash so that one can use cash rather than sell other investments during downswings is a type of market timing. How do you know when to use your cash, and know that by using your cash you are not delaying the selling of your equities to a time when the equities prices are even lower? Aren't you trying to time the market?

gasdoc
OK, some questions for you.
Start portfolio 100k
VPW says take 5% of port per year.
Take 5% out first year. ($5,000)
Market goes down 20%. (95K x .8 = 76K)
Take out 5% of port year two. (76K x .05 = 3.8 K)

1. Is it market timing to do the above?
2. Would you feel better having a large cash reserve?
For me, the answers are no and yes.

Or at least that is my read of the wiki for this method: https://www.bogleheads.org/wiki/Variabl ... withdrawal
VPW never has you taking the same percentage out in consecutive years. The percentage to be used increases monotonically each year. See table on the wiki cited above. There is no market timing with this method. You're not changing your allocation based on anything. You're simply taking out an ever increasing percentage of whatever dollar amount is in your portfolio at the time.

azanon
Posts: 2360
Joined: Mon Nov 07, 2011 10:34 am
Location: Little Rock, AR
Contact:

Re: Downsides of VPW?

Post by azanon » Wed Jun 05, 2019 7:41 am

randomguy wrote:
Tue Jun 04, 2019 8:26 pm
longinvest wrote:
Tue Jun 04, 2019 4:45 pm

Randomguy,

As far as I know, nobody knows whether future markets will recover or not (think of Japan), and no portfolio withdrawal method can create returns that the market doesn't provide.

Our wiki's VPW page suggests that the retiree combines stable income with VPW withdrawals.

Do you have a better suggestion?
Adding stable income doesn't change anything. You can do that with any of these schemes. It doesn't change the fact that the portfolio fails to produce income when you might want it.

Better all comes down to your definition better. I might prefer to live on 40k/year for 28 years and then live on my stable income. You might prefer to live on 30k for 20 years and then 50k for the next 20. And someone else might prefer a steady 35k/year for 40 years. Or any of a zillion mixtures in between. You end up having to balance a bunch of different goals and risks and pick which one you can live with.
And so to clarify our discussion earlier, I approach it completely from a portfolio point of view, and put income variations secondary. In theory, the most efficient way a portfolio can be treated would be to spend less from it when it's down and more from it when it's high. Buy low, sell high. If you do that in exact proportions (what VPW does) then you're treating the portfolio most efficiently.

So I don't think of it as what income stability do I prefer. I think what is best for my portfolio. Now if I could have a smoother withdrawals with NO harm to the portfolio, now I'm interested. I guess its a similar reason to why I'm drawn to risk parity and/or low equity portfolios - because they're most efficient when designed that way (higher Sharpes and/or more return per unit of risk).

User avatar
willthrill81
Posts: 10157
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: Downsides of VPW?

Post by willthrill81 » Wed Jun 05, 2019 8:53 am

4nursebee wrote:
Wed Jun 05, 2019 3:16 am
gasdoc wrote:
Tue Jun 04, 2019 4:48 am
4nursebee wrote:
Tue Jun 04, 2019 4:23 am
I/we intend to use this method. I do not know the research into downsides, really have not paid attention to that. I view the downside as having to have enough cash or cash equivalents available to tolerate (psychologically and financially) small withdrawals if needed. Because of this we are growing this portion of our assets above all others. We are aiming for 2 years out of pocket expenses, not including healthcare or taxes.
I am glad you brought this up, as I've heard it before and have not really understood the principle. It seems like the concept of keeping more cash so that one can use cash rather than sell other investments during downswings is a type of market timing. How do you know when to use your cash, and know that by using your cash you are not delaying the selling of your equities to a time when the equities prices are even lower? Aren't you trying to time the market?

gasdoc
OK, some questions for you.
Start portfolio 100k
VPW says take 5% of port per year.
Take 5% out first year. ($5,000)
Market goes down 20%. (95K x .8 = 76K)
Take out 5% of port year two. (76K x .05 = 3.8 K)

1. Is it market timing to do the above?
2. Would you feel better having a large cash reserve?
For me, the answers are no and yes.

Or at least that is my read of the wiki for this method: https://www.bogleheads.org/wiki/Variabl ... withdrawal
If your idea is to use cash to fund your withdrawals during a stock downturn, then you are intending on using a dynamic asset allocation, which is a form of market timing.
“It's a dangerous business, Frodo, going out your door. You step onto the road, and if you don't keep your feet, there's no knowing where you might be swept off to.” J.R.R. Tolkien,The Lord of the Rings

randomguy
Posts: 7677
Joined: Wed Sep 17, 2014 9:00 am

Re: Downsides of VPW?

Post by randomguy » Wed Jun 05, 2019 10:27 am

azanon wrote:
Wed Jun 05, 2019 7:41 am
randomguy wrote:
Tue Jun 04, 2019 8:26 pm
longinvest wrote:
Tue Jun 04, 2019 4:45 pm

Randomguy,

As far as I know, nobody knows whether future markets will recover or not (think of Japan), and no portfolio withdrawal method can create returns that the market doesn't provide.

Our wiki's VPW page suggests that the retiree combines stable income with VPW withdrawals.

Do you have a better suggestion?
Adding stable income doesn't change anything. You can do that with any of these schemes. It doesn't change the fact that the portfolio fails to produce income when you might want it.

Better all comes down to your definition better. I might prefer to live on 40k/year for 28 years and then live on my stable income. You might prefer to live on 30k for 20 years and then 50k for the next 20. And someone else might prefer a steady 35k/year for 40 years. Or any of a zillion mixtures in between. You end up having to balance a bunch of different goals and risks and pick which one you can live with.
And so to clarify our discussion earlier, I approach it completely from a portfolio point of view, and put income variations secondary. In theory, the most efficient way a portfolio can be treated would be to spend less from it when it's down and more from it when it's high. Buy low, sell high. If you do that in exact proportions (what VPW does) then you're treating the portfolio most efficiently.

So I don't think of it as what income stability do I prefer. I think what is best for my portfolio. Now if I could have a smoother withdrawals with NO harm to the portfolio, now I'm interested. I guess its a similar reason to why I'm drawn to risk parity and/or low equity portfolios - because they're most efficient when designed that way (higher Sharpes and/or more return per unit of risk).
What is best for the portfolio isn't spending anything.:) But that isn't most peoples goal. As soon as you start taking money out you have to decide how you value a ton of factors from income stability, portfolio value, and portfolio depletion risk. VPW does harm to the portfolio so it is all about how much harm you are willing to accept and how little income you will accept. If you define the most harm as what VPW does and the least income as being what VPW provides, it is hard to beat VPW:) Relax either of those constraints and other systems might be better.

There are a zillion schemes out there for retirement spending. Any variable scheme has the same drawback of not guaranting income. Most are less aggressive than VPW at reacting at market conditions. That gives you more stable income but less income later and potentially the dips can be lower and longer. But the details all depend on the scheme and assumptions.

azanon
Posts: 2360
Joined: Mon Nov 07, 2011 10:34 am
Location: Little Rock, AR
Contact:

Re: Downsides of VPW?

Post by azanon » Wed Jun 05, 2019 10:44 am

randomguy wrote:
Wed Jun 05, 2019 10:27 am
azanon wrote:
Wed Jun 05, 2019 7:41 am
randomguy wrote:
Tue Jun 04, 2019 8:26 pm
longinvest wrote:
Tue Jun 04, 2019 4:45 pm

Randomguy,

As far as I know, nobody knows whether future markets will recover or not (think of Japan), and no portfolio withdrawal method can create returns that the market doesn't provide.

Our wiki's VPW page suggests that the retiree combines stable income with VPW withdrawals.

Do you have a better suggestion?
Adding stable income doesn't change anything. You can do that with any of these schemes. It doesn't change the fact that the portfolio fails to produce income when you might want it.

Better all comes down to your definition better. I might prefer to live on 40k/year for 28 years and then live on my stable income. You might prefer to live on 30k for 20 years and then 50k for the next 20. And someone else might prefer a steady 35k/year for 40 years. Or any of a zillion mixtures in between. You end up having to balance a bunch of different goals and risks and pick which one you can live with.
And so to clarify our discussion earlier, I approach it completely from a portfolio point of view, and put income variations secondary. In theory, the most efficient way a portfolio can be treated would be to spend less from it when it's down and more from it when it's high. Buy low, sell high. If you do that in exact proportions (what VPW does) then you're treating the portfolio most efficiently.

So I don't think of it as what income stability do I prefer. I think what is best for my portfolio. Now if I could have a smoother withdrawals with NO harm to the portfolio, now I'm interested. I guess its a similar reason to why I'm drawn to risk parity and/or low equity portfolios - because they're most efficient when designed that way (higher Sharpes and/or more return per unit of risk).
What is best for the portfolio isn't spending anything.:) But that isn't most peoples goal. As soon as you start taking money out you have to decide how you value a ton of factors from income stability, portfolio value, and portfolio depletion risk. VPW does harm to the portfolio so it is all about how much harm you are willing to accept and how little income you will accept. If you define the most harm as what VPW does and the least income as being what VPW provides, it is hard to beat VPW:) Relax either of those constraints and other systems might be better.

There are a zillion schemes out there for retirement spending. Any variable scheme has the same drawback of not guaranting income. Most are less aggressive than VPW at reacting at market conditions. That gives you more stable income but less income later and potentially the dips can be lower and longer. But the details all depend on the scheme and assumptions.
In fairness, I probably should give a full disclosure here. I'll likely be using a 30% equity portfolio or thereabouts or, more likely, something risk parity-ish. Massive fluctuation quite simply won't be my concern no matter what system I use. I have this massively unpopular belief that heavy equity portfolios are only appropriate early in the accumulation phase, and low risk portfolios are for retirement and income.

I did say it already though, in so many words. I don't think VPW is very appropriate for equity heavy portfolios. I'll just leave that statement in general terms and let everyone else decide what equity heavy actually is. Rephrased, if your withdrawals are jumping all over the place from year to year, your problem isn't VPW, it's your portfolio that has too much risk.

Admiral
Posts: 2047
Joined: Mon Oct 27, 2014 12:35 pm

Re: Downsides of VPW?

Post by Admiral » Wed Jun 05, 2019 10:51 am

longinvest wrote:
Tue Jun 04, 2019 3:32 pm
msa6 wrote:
Tue Jun 04, 2019 2:52 pm
As always, I really appreciate the wisdom of BH posters.

I'm going to do a little deeper dig on our expenses and give some more thought to how big a hit our portfolio could handle before we'd be forced to make really serious changes using a VPW strategy. We're a few years from SS, and there are no pensions, etc. in the mix. Regardless of what this reveals, the general concept of spending more freely when the portfolio looks good, and more carefully when it's struggling, very much feels to me like the right way to go.
Msa6,

I am considering to build an easy-to-use spreadsheet similar to the new variable savings rate (VSR) spreadsheet to help estimate the flexibility that the spending budget should accommodate, when combining VPW withdrawals from a balanced portfolio with current and future pensions.

In a few words, VSR is an extension of VPW to accumulation time. It estimates how much one should save in the upcoming year, given one's age, salary, portfolio balance and allocation, planned retirement age, and future pensions (with and without cost of living adjustments). Like VPW, it makes no attempt to predict future returns; it adjusts savings, annually, to the new portfolio balance, new salary, new shorter retirement horizon, etc.

The VSR spreadsheet provides the accumulating investor with two amounts:
  1. The amount that should be saved (monthly, semimonthly, or biweekly) during the upcoming year.
  2. An amount of additional savings that one should be able to make if markets misbehaved (e.g. stocks lost 50%). In other words, that's the amount that should be easy to cut in the spending budget. It's the flexibility that using VSR requires.
The spreadsheet also informs the accumulating investor about the projected retirement income, before and after loss.

Here's a screenshot:

Image

A similar VPW spreadsheet would ask the retiree about his age, portfolio balance and allocation, current and future pensions (with and without cost of living adjustments), and provide two amounts:
  1. A suggested amount to withdraw from the portfolio (annually, quarterly, or monthly) during the upcoming year.
  2. A withdrawal reduction amount if markets misbehaved (e.g. stocks lost 50%). In other words, that's the amount that should be easy to cut in the spending budget. It's the flexibility that VPW requires.
The spreadsheet would also inform the retiree about the resulting total retirement income (including pensions and withdrawal), before and after loss. Starting at age 80, it would inform the retiree about the residual income after age 100 and suggest to buy an inflation-indexed SPIA* with part (not all) of the portfolio, if residual income seems insufficient.

* Single-Premium Immediate Annuity.

Would such a spreadsheet (online and downloadable) be useful to you?
If you build it, we will come! Yes.

(Just FYI Flexible Retirement Planner has this same functionality.)

jmk
Posts: 452
Joined: Tue Nov 01, 2011 7:48 pm

Re: Downsides of VPW?

Post by jmk » Wed Jun 05, 2019 11:06 am

While being very efficient, the vpw method tends to take out the most $ exponentially toward end of life, when you value the money least.

azanon
Posts: 2360
Joined: Mon Nov 07, 2011 10:34 am
Location: Little Rock, AR
Contact:

Re: Downsides of VPW?

Post by azanon » Wed Jun 05, 2019 11:07 am

jmk wrote:
Wed Jun 05, 2019 11:06 am
While being very efficient, the vpw method tends to take out the most $ exponentially toward end of life, when you value the money least.
My understanding is that this is incorrect. It's designed to take out approximately the same amount of inflation-adjusted dollars from start to finish. Longinvest, correct me if I'm wrong here.....

Maybe you're thinking of the "1/n withdrawal" method? (which does have that issue)

longinvest
Posts: 3485
Joined: Sat Aug 11, 2012 8:44 am

Re: Downsides of VPW?

Post by longinvest » Wed Jun 05, 2019 11:16 am

azanon wrote:
Wed Jun 05, 2019 11:07 am
jmk wrote:
Wed Jun 05, 2019 11:06 am
While being very efficient, the vpw method tends to take out the most $ exponentially toward end of life, when you value the money least.
My understanding is that this is incorrect. It's designed to take out approximately the same amount of inflation-adjusted dollars from start to finish. Longinvest, correct me if I'm wrong here.....
VPW withdrawals fluctuate with the markets the portfolio is invested into.

A retiree who retired in 1966 at age 65 and hit age 81 in 1982 at the start of a secular bull market has seen his withdrawals increase in the later part of his life (starting at age 81), when using a globally diversified 50/50 stocks/bonds allocation.

A retiree who retired in 1982 at age 65 and hit age 83 in 2000 at the start of a challenging start of century has seen his withdrawal decrease in the later part of his life (starting at age 83), when using a globally diversified 50/50 stocks/bonds allocation.

VPW doesn't get to choose whether markets will have high or low returns, and in which part of the retiree's life. This would require predicting future market returns. VPW simply adapts withdrawals the retiree's age, portfolio balance, and asset allocation. Markets determine the shape of withdrawals.

As Bogleheads know, nominal bonds dampen nominal portfolio fluctuations, and inflation-indexed bonds dampen inflation-indexed portfolio fluctuations. It is thus possible for a retiree using VPW to dampen withdrawal fluctuations by including bonds in his portfolio. It is suggested to use an allocation within the range 30/70 to 70/30 stocks/bonds.

Knowing that stocks can easily lose 50% in a short amount of time, someone whose income provides 50% from pensions (including Social Security) and 50% from VPW withdrawals can limit total income fluctuations to 15% by investing into a portfolio allocated (15% / 50%) / (1 - 50%) = 60% into stocks, (1 - 60%) = 40% into bonds, a 60/40 stocks/bonds portfolio.
Last edited by longinvest on Wed Jun 05, 2019 11:31 am, edited 2 times in total.
Bogleheads investment philosophy | single-ETF balanced portfolio | VBAL

azanon
Posts: 2360
Joined: Mon Nov 07, 2011 10:34 am
Location: Little Rock, AR
Contact:

Re: Downsides of VPW?

Post by azanon » Wed Jun 05, 2019 11:26 am

longinvest wrote:
Wed Jun 05, 2019 11:16 am
azanon wrote:
Wed Jun 05, 2019 11:07 am
jmk wrote:
Wed Jun 05, 2019 11:06 am
While being very efficient, the vpw method tends to take out the most $ exponentially toward end of life, when you value the money least.
My understanding is that this is incorrect. It's designed to take out approximately the same amount of inflation-adjusted dollars from start to finish. Longinvest, correct me if I'm wrong here.....
VPW withdrawals fluctuate with the markets the portfolio is invested into.

A retiree who retired in 1966 at age 65 and hit age 81 in 1982 at the start of a secular bull market has seen his withdrawals increase in the later part of his life (starting at age 81), when using a globally diversified 50/50 stocks/bonds allocation.

A retiree who retired in 1982 at age 65 and hit age 83 in 2000 at the start of a challenging start of century has seen his withdrawal decrease in the later part of his life (starting at age 83), when using a globally diversified 50/50 stocks/bonds allocation.

VPW doesn't get to choose whether markets will have high or low returns, and in which part of the retiree's life. This would require predicting future market returns. VPW simply adapts withdrawals the retiree's age, portfolio balance, and asset allocation. Markets determine the shape of withdrawals.
Right, but the angle I was coming at it was this; Assume that a portfolio returned it's exact expected return, with 0% volatility. Aren't the withdrawals (the increasing percentages by year) designed as a best estimate of exactly what they should be to give a perfectly smooth inflation-adjusted income? Contrast that with 1/n method which uses remaining lifespan as a primary variable for the withdrawal amount (along with portfolio fluctuations), which will naturally result in inflation-adjusted withdrawals rising by age.

Now if that's not true, and it is intentionally raising inflation-adjusted withdrawals along with the portfolio fluctuations, then I do need a new method.

longinvest
Posts: 3485
Joined: Sat Aug 11, 2012 8:44 am

Re: Downsides of VPW?

Post by longinvest » Wed Jun 05, 2019 11:33 am

azanon wrote:
Wed Jun 05, 2019 11:26 am
Assume that a portfolio returned it's exact expected return, with 0% volatility.
Expected return? You're probably talking about the actuarial method which uses future return predictions to calibrate withdrawals. VPW doesn't do that.

VPW has internal fixed growth trends. They are based on long-term (longer than a century) global stock and bond returns, including countries where they went to zero. The goal of growth trends is to loosely distinguish between high and low returns. They aren't meant as a precise indicator, nor as a future return prediction. It would be a mistake to replace the portfolio's growth trend with a future return prediction and call the result "VPW" as in "Bogleheads wiki VPW".

I don't think that anyone would be shocked if I said that a 5% real return is a high return for bonds, and that a -10% real return is a low return for stocks. That's what growth trends serve to loosely determine.
Bogleheads investment philosophy | single-ETF balanced portfolio | VBAL

azanon
Posts: 2360
Joined: Mon Nov 07, 2011 10:34 am
Location: Little Rock, AR
Contact:

Re: Downsides of VPW?

Post by azanon » Wed Jun 05, 2019 11:44 am

longinvest wrote:
Wed Jun 05, 2019 11:33 am
azanon wrote:
Wed Jun 05, 2019 11:26 am
Assume that a portfolio returned it's exact expected return, with 0% volatility.
Expected return? You're probably talking about the actuarial method who uses future return predictions to calibrate withdrawals. VPW doesn't do that.

VPW has an internal fixed growth trend. It is based on long-term (longer than a century) global stock and bond returns, including countries where they went to zero. The goal of the growth trend is to loosely distinguish between high and low returns. It isn't meant as a precise indicator, nor as a future return prediction. It would be a mistake to replace the growth trend with a return prediction and call the result "VPW" as in "Bogleheads wiki VPW".
So what jmk said actually is true? With a hypothetical portfolio that gave a fixed inflation-adjusted return with 0% volatility using VPW, the inflation-adjusted withdrawals would actually rise year by year? Hmm i didn't realize that. I thought those withdrawal values for the various % equity portfolios were designed to make it more linear.

Then why not use the 1/n method? That's it's only weakness. I thought VPW addresssed that issue... < confused >

longinvest
Posts: 3485
Joined: Sat Aug 11, 2012 8:44 am

Re: Downsides of VPW?

Post by longinvest » Wed Jun 05, 2019 11:54 am

azanon wrote:
Wed Jun 05, 2019 11:44 am
longinvest wrote:
Wed Jun 05, 2019 11:33 am
azanon wrote:
Wed Jun 05, 2019 11:26 am
Assume that a portfolio returned it's exact expected return, with 0% volatility.
Expected return? You're probably talking about the actuarial method who uses future return predictions to calibrate withdrawals. VPW doesn't do that.

VPW has an internal fixed growth trend. It is based on long-term (longer than a century) global stock and bond returns, including countries where they went to zero. The goal of the growth trend is to loosely distinguish between high and low returns. It isn't meant as a precise indicator, nor as a future return prediction. It would be a mistake to replace the growth trend with a return prediction and call the result "VPW" as in "Bogleheads wiki VPW".
So what jmk said actually is true? With a hypothetical portfolio that gave a fixed inflation-adjusted return with 0% volatility using VPW, the inflation-adjusted withdrawals would actually rise year by year? Hmm i didn't realize that. I thought those withdrawal values for the various % equity portfolios were designed to make it more linear.

Then why not use the 1/n method? That's it's only weakness. I thought VPW addresssed that issue... < confused >
1/n assumes that a portfolio tends to gravitate around a 0% real return.

Historically, the average real return of bonds and of stocks have been higher than 0%. Not always. Sometimes the returns of stocks or bonds were negative for at least a decade. VPW is meant to adapt withdrawals to market returns. Every year, when market returns are high, withdrawals increase and when market returns are low, withdrawals decrease. These annual fluctuations are the fluctuations that the retiree will mostly notice.

Now, over the entire retirement (with unknown-in-advance-death year), the long-term trend of withdrawals might slightly increase or decrease depending on whether returns are higher or lower than historical returns. But, this effect won't be really perceptible to the retiree, as he'll be mostly affected by the more apparent annual fluctuations and his own long-term changes in personal consumption patterns.
Bogleheads investment philosophy | single-ETF balanced portfolio | VBAL

azanon
Posts: 2360
Joined: Mon Nov 07, 2011 10:34 am
Location: Little Rock, AR
Contact:

Re: Downsides of VPW?

Post by azanon » Wed Jun 05, 2019 12:00 pm

longinvest wrote:
Wed Jun 05, 2019 11:54 am
azanon wrote:
Wed Jun 05, 2019 11:44 am
longinvest wrote:
Wed Jun 05, 2019 11:33 am
azanon wrote:
Wed Jun 05, 2019 11:26 am
Assume that a portfolio returned it's exact expected return, with 0% volatility.
Expected return? You're probably talking about the actuarial method who uses future return predictions to calibrate withdrawals. VPW doesn't do that.

VPW has an internal fixed growth trend. It is based on long-term (longer than a century) global stock and bond returns, including countries where they went to zero. The goal of the growth trend is to loosely distinguish between high and low returns. It isn't meant as a precise indicator, nor as a future return prediction. It would be a mistake to replace the growth trend with a return prediction and call the result "VPW" as in "Bogleheads wiki VPW".
So what jmk said actually is true? With a hypothetical portfolio that gave a fixed inflation-adjusted return with 0% volatility using VPW, the inflation-adjusted withdrawals would actually rise year by year? Hmm i didn't realize that. I thought those withdrawal values for the various % equity portfolios were designed to make it more linear.

Then why not use the 1/n method? That's it's only weakness. I thought VPW addresssed that issue... < confused >
1/n assumes that a portfolio tends to gravitate around a 0% real return.

Historically, the average real return of bonds and of stocks have been higher than 0%. Not always. Sometimes returns were negative for at least a decade. VPW is meant to adapt withdrawals to market returns. When market returns are high, withdrawals increase. When market returns are low, withdrawals decrease. These are the fluctuations that the retiree will mostly notice.

Now, over the entire retirement (with unknown-in-advance-death year), the long-term trend of withdrawals might slightly increase or decrease depending on whether returns are higher or lower than historical lower returns. But, this effect won't be really perceptible to the retiree, as he'll be mostly affected by the more apparent annual fluctuations and his own long-term changes in consumption patterns.
Ok I think that matches my original understanding then, but I just didn't articulate it correctly.

jmk
Posts: 452
Joined: Tue Nov 01, 2011 7:48 pm

Re: Downsides of VPW?

Post by jmk » Sun Jun 09, 2019 3:33 pm

azanon wrote:
Wed Jun 05, 2019 11:07 am
jmk wrote:
Wed Jun 05, 2019 11:06 am
While being very efficient, the vpw method tends to take out the most $ exponentially toward end of life, when you value the money least.
My understanding is that this is incorrect. It's designed to take out approximately the same amount of inflation-adjusted dollars from start to finish. Longinvest, correct me if I'm wrong here.....

Maybe you're thinking of the "1/n withdrawal" method? (which does have that issue)
You're right, I was confusing the two. If returns were equal to the discount rate used in the the PMT formula, you'd get an even amount each year. The only way things would be off is if you assumed the wrong return (high or low) in retrospect, and of course as actual returns go up and down.

In fact, you could think of the actuarial pmt method (of which VPW is a variant) mathematically as: the 1/n discounted such that the amounts would be equal if returns matched the discount rate.

Now with VPW specifically, since it uses as its discount rate the Credit Suisse Global Investment Returns Yearbook long-term average, withdrawal rates will rise or fall over time depending on how the future unfolds relative to that average. The easiest way to see this is to compare to the Safe Withdrawal Rate, which by definition is the highest constant rate calculated in retrospect. VPW will be higher over time if the Credit Suisse average ends up being lower than true returns, and vice versa.

I personally prefer using a CAPE based discount rate and a simpler actuarial model-- not because I think CAPE magically predicts returns, but to move the yearly PMT amounts toward even. VPW has too many additional bells and whistles on the basic actuarial idea.

GuyInFL
Posts: 182
Joined: Thu Aug 04, 2016 7:17 pm

Re: Downsides of VPW?

Post by GuyInFL » Sun Jun 09, 2019 4:07 pm

corn18 wrote:
Tue Jun 04, 2019 4:18 pm
This thread has been a great read. I think this is my plan:

1. My COLA pension with SS @ 70 will cover 100% of expenses (including taxes).
2. I think I am going to set aside enough to bridge SS from 55 to 70. By set aside, I just mean mental accounting. All money will reside in a 60/40 BH portfolio. No SPIA. This should require about $700k of my projected $1.7M of savings.
3. Then I can VPW the rest as discretionary spend. I can either spend this or save it each year and if the crap hits the fan, I still have my 100% floor.
In a similar situation and am thinking about a rising glidepath for the SS bridge. Looks like you are going witjh a fixed AA.

MnD
Posts: 4273
Joined: Mon Jan 14, 2008 12:41 pm

Re: Downsides of VPW?

Post by MnD » Sun Jun 09, 2019 4:12 pm

dcabler wrote:
Tue Jun 04, 2019 11:19 am
bhsince87 wrote:
Tue Jun 04, 2019 10:48 am
I think the biggest downside of VPW, for some people, is the potential for significant income changes from year to year.

This can be an issue for folks who have always used a strict biweekly/monthly/yearly budget, and match cash flows of income/outgo.

We've been on a highly variable, bonus based salary for 20+ years, so this will not be an issue for us.
Yep, that's the primary downside of VPW and other PMT-based techniques. When compared to SWR, it's a tradeoff of fluctuating income year on year with a guarantee that your portfolio will last exactly as long as you planned for vs. a steady, income adjusted income stream with the danger of either running out of money before you planned or leaving a huge pile of money when you croak.
That's why I went with a hybrid of 5% of annual portfolio balance (throughout retirement) with a floor of a) 3% inflation-adjusted SWR from portfolio, plus b) one nearly maxed out and one above average SS claim down the road and c) a federal pension on 35+ years service.

We're fine with getting what the market provides on the portfolio income side including fluctuation and much better income outcomes in anything better than terrible sequences. Our expectation is checking out with no real portfolio growth versus runaway balances to upside or formulaic depletion, although if bad sequences do come to pass, significant portfolio depletion will occur just as it will with a low 3.X% inflation-adjusted SWR. The difference being the poor conversion of wealth to retirement income only occurs if needed, versus being guaranteed with a low inflation-adjusted SWR.

The things I didn't care for with VPW was no floor and a formulaic depletion of portfolio in advancing age. We have the greatest utility for income in retirement right now in our 50's when presumably we are the most active/healthy/energetic. Don't see the sense in adopting a plan that includes an increasing % of portfolio withdrawal strategy as one gets into the middle and later years of retirement - just because one can get away with it.

User avatar
willthrill81
Posts: 10157
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: Downsides of VPW?

Post by willthrill81 » Sun Jun 09, 2019 6:35 pm

MnD wrote:
Sun Jun 09, 2019 4:12 pm
dcabler wrote:
Tue Jun 04, 2019 11:19 am
bhsince87 wrote:
Tue Jun 04, 2019 10:48 am
I think the biggest downside of VPW, for some people, is the potential for significant income changes from year to year.

This can be an issue for folks who have always used a strict biweekly/monthly/yearly budget, and match cash flows of income/outgo.

We've been on a highly variable, bonus based salary for 20+ years, so this will not be an issue for us.
Yep, that's the primary downside of VPW and other PMT-based techniques. When compared to SWR, it's a tradeoff of fluctuating income year on year with a guarantee that your portfolio will last exactly as long as you planned for vs. a steady, income adjusted income stream with the danger of either running out of money before you planned or leaving a huge pile of money when you croak.
That's why I went with a hybrid of 5% of annual portfolio balance (throughout retirement) with a floor of a) 3% inflation-adjusted SWR from portfolio, plus b) one nearly maxed out and one above average SS claim down the road and c) a federal pension on 35+ years service.
For those who might not be aware of it, Portfolio Charts has a new tool that enables you to backtest such a strategy easily with many different asset classes going back to 1970. For a 60/40 AA, this strategy would have hit against the 3% floor about six times since 1970, ranging from 3 to 16 years into retirement, before going back up. This strategy definitely allows for volatile withdrawals, but as long as your essential spending is covered by the 3% floor, it sounds fine to me, especially in MnD's case since SS benefits and a pension are added on top.

Image
Last edited by willthrill81 on Sun Jun 09, 2019 6:45 pm, edited 1 time in total.
“It's a dangerous business, Frodo, going out your door. You step onto the road, and if you don't keep your feet, there's no knowing where you might be swept off to.” J.R.R. Tolkien,The Lord of the Rings

User avatar
corn18
Posts: 1392
Joined: Fri May 22, 2015 6:24 am

Re: Downsides of VPW?

Post by corn18 » Sun Jun 09, 2019 6:41 pm

GuyInFL wrote:
Sun Jun 09, 2019 4:07 pm
corn18 wrote:
Tue Jun 04, 2019 4:18 pm
This thread has been a great read. I think this is my plan:

1. My COLA pension with SS @ 70 will cover 100% of expenses (including taxes).
2. I think I am going to set aside enough to bridge SS from 55 to 70. By set aside, I just mean mental accounting. All money will reside in a 60/40 BH portfolio. No SPIA. This should require about $700k of my projected $1.7M of savings.
3. Then I can VPW the rest as discretionary spend. I can either spend this or save it each year and if the crap hits the fan, I still have my 100% floor.
In a similar situation and am thinking about a rising glidepath for the SS bridge. Looks like you are going witjh a fixed AA.
That's an excellent point. I hadn't thought about a rising glide path as I withdraw the SS bridge. Interesting approach that I will definitely consider when the time comes. How high would you go on stocks if the end point at age 70 was just going to be used for blowing that dough or legacy?
Don't do something, just stand there!

GuyInFL
Posts: 182
Joined: Thu Aug 04, 2016 7:17 pm

Re: Downsides of VPW?

Post by GuyInFL » Mon Jun 10, 2019 7:11 am

corn18 wrote:
Sun Jun 09, 2019 6:41 pm
GuyInFL wrote:
Sun Jun 09, 2019 4:07 pm
corn18 wrote:
Tue Jun 04, 2019 4:18 pm
This thread has been a great read. I think this is my plan:

1. My COLA pension with SS @ 70 will cover 100% of expenses (including taxes).
2. I think I am going to set aside enough to bridge SS from 55 to 70. By set aside, I just mean mental accounting. All money will reside in a 60/40 BH portfolio. No SPIA. This should require about $700k of my projected $1.7M of savings.
3. Then I can VPW the rest as discretionary spend. I can either spend this or save it each year and if the crap hits the fan, I still have my 100% floor.
In a similar situation and am thinking about a rising glidepath for the SS bridge. Looks like you are going with a fixed AA.
That's an excellent point. I hadn't thought about a rising glide path as I withdraw the SS bridge. Interesting approach that I will definitely consider when the time comes. How high would you go on stocks if the end point at age 70 was just going to be used for blowing that dough or legacy?
That's the $64,000 question. I'm inclined to go with 75/25 or 80/20 (Stocks/Bonds) at 70 once SS kicks in since really all of the fixed and most of the discretionary expenses are covered.
Assuming $50K of your and spouse SS, you could reserve...
$20K per year (20K*15 = 300K) plus withdraw $30K per year ($1700K-$300K = $1400K, so $30K/$1400K = 2.1%) pretty sustainably or
$30K per year (30K*15 = 450K) plus withdraw $20K per year ($1700K-$450K = $1250K, so $20K/$1250K = 1.5%)

So an 80/20 ratio at 70 would start as about (.2*$1700K = $340K) + 300K or 450K which is 640K (62/38) or 790K (54/46)
Similarly 75/25 (.25*1700K = $425K) +300K or 450K which is $725K (57/43) or $875K (49/51)
Rounding these numbers generates a range of 60/40 => 50/50 which strikes me as pretty reasonable.

Either of these should be pretty safe with a higher expected ending portfolio with the $20K / year fixed withdrawal compared to the $30K / year withdrawal. Obviously the risk is higher (with $20K/year withdrawal) as well.
Also be sure to consider what happens when either you or your spouse passes on. You'll drop to 1 SS and you'll need to consider if the COLA pension decreases if you pass on.

Always looking for comments on the strategy above.

User avatar
corn18
Posts: 1392
Joined: Fri May 22, 2015 6:24 am

Re: Downsides of VPW?

Post by corn18 » Mon Jun 10, 2019 8:28 am

GuyInFL wrote:
Mon Jun 10, 2019 7:11 am
corn18 wrote:
Sun Jun 09, 2019 6:41 pm
GuyInFL wrote:
Sun Jun 09, 2019 4:07 pm
corn18 wrote:
Tue Jun 04, 2019 4:18 pm
This thread has been a great read. I think this is my plan:

1. My COLA pension with SS @ 70 will cover 100% of expenses (including taxes).
2. I think I am going to set aside enough to bridge SS from 55 to 70. By set aside, I just mean mental accounting. All money will reside in a 60/40 BH portfolio. No SPIA. This should require about $700k of my projected $1.7M of savings.
3. Then I can VPW the rest as discretionary spend. I can either spend this or save it each year and if the crap hits the fan, I still have my 100% floor.
In a similar situation and am thinking about a rising glidepath for the SS bridge. Looks like you are going with a fixed AA.
That's an excellent point. I hadn't thought about a rising glide path as I withdraw the SS bridge. Interesting approach that I will definitely consider when the time comes. How high would you go on stocks if the end point at age 70 was just going to be used for blowing that dough or legacy?
That's the $64,000 question. I'm inclined to go with 75/25 or 80/20 (Stocks/Bonds) at 70 once SS kicks in since really all of the fixed and most of the discretionary expenses are covered.
Assuming $50K of your and spouse SS, you could reserve...
$20K per year (20K*15 = 300K) plus withdraw $30K per year ($1700K-$300K = $1400K, so $30K/$1400K = 2.1%) pretty sustainably or
$30K per year (30K*15 = 450K) plus withdraw $20K per year ($1700K-$450K = $1250K, so $20K/$1250K = 1.5%)

So an 80/20 ratio at 70 would start as about (.2*$1700K = $340K) + 300K or 450K which is 640K (62/38) or 790K (54/46)
Similarly 75/25 (.25*1700K = $425K) +300K or 450K which is $725K (57/43) or $875K (49/51)
Rounding these numbers generates a range of 60/40 => 50/50 which strikes me as pretty reasonable.

Either of these should be pretty safe with a higher expected ending portfolio with the $20K / year fixed withdrawal compared to the $30K / year withdrawal. Obviously the risk is higher (with $20K/year withdrawal) as well.
Also be sure to consider what happens when either you or your spouse passes on. You'll drop to 1 SS and you'll need to consider if the COLA pension decreases if you pass on.

Always looking for comments on the strategy above.
I'll have to set up a model in excel. That will allow me to do some sensitivity analysis. And good point on survivor income. That will definitely play a role on whether I let the glide path rise or not after 70. Yeah, more numbers to crunch!
Don't do something, just stand there!

longinvest
Posts: 3485
Joined: Sat Aug 11, 2012 8:44 am

Re: Downsides of VPW?

Post by longinvest » Tue Jun 11, 2019 5:44 am

msa6 wrote:
Tue Jun 04, 2019 3:59 pm
longinvest wrote:
Tue Jun 04, 2019 3:32 pm
Would such a spreadsheet (online and downloadable) be useful to you?
Since you were nice enough to ask...yes! (Guessing it would be useful to others as well...thanks.)
One Ping wrote:
Tue Jun 04, 2019 7:13 pm
longinvest, will your new VSR spreadsheet be integrated with VPW so one would have a seamless planning tool covering from the start of accumulation phase through retirement to the end of withdrawal phase?
I've just made the announcement. It's a unified VPW accumulation and retirement worksheet.

Here are two screenshots.

First, the Retirement sheet:

Image

Second, the Accumulation sheet:

Image

Enjoy!
Bogleheads investment philosophy | single-ETF balanced portfolio | VBAL

User avatar
One Ping
Posts: 686
Joined: Thu Sep 24, 2015 4:53 pm

Re: Downsides of VPW?

Post by One Ping » Tue Jun 11, 2019 10:13 am

Cool. Thanks, longinvest. I look forward to checking it out.
"Re-verify our range to target ... one ping only."

bradshaw1965
Posts: 743
Joined: Tue Jul 31, 2007 9:36 pm

Re: Downsides of VPW?

Post by bradshaw1965 » Tue Jun 11, 2019 11:31 am

willthrill81 wrote:
Wed Jun 05, 2019 12:11 am
AlohaJoe wrote:
Tue Jun 04, 2019 10:58 pm
bhsince87 wrote:
Tue Jun 04, 2019 10:20 pm
Yes, I'd much prefer spending 28k at 75 and 80k at 95 than having nothing left to spend at, say, age 85.

And zero to spend at age 86, 87, 88, 89, ....
You don't have zero to spend. You have your Social Security + SPIA.
It's not even that bad. First, the idea that retirees will blindly spend their portfolio down to zero is a myth. Second, many retirees will never survive to age 85, although many here seem to be acting as though they will live to 120, despite those over the age of 70 only having gotten about one extra year of life expectancy per decade since the 1950s.
This is the chart that makes the comparison to the 1950's a little more real for me, especially for upper-middle class people.

Image


And having dealt with a little too much cognitive decline caregiving I'm much more concerned about quality of life then early death.

User avatar
willthrill81
Posts: 10157
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: Downsides of VPW?

Post by willthrill81 » Tue Jun 11, 2019 11:53 am

bradshaw1965 wrote:
Tue Jun 11, 2019 11:31 am
willthrill81 wrote:
Wed Jun 05, 2019 12:11 am
AlohaJoe wrote:
Tue Jun 04, 2019 10:58 pm
bhsince87 wrote:
Tue Jun 04, 2019 10:20 pm
Yes, I'd much prefer spending 28k at 75 and 80k at 95 than having nothing left to spend at, say, age 85.

And zero to spend at age 86, 87, 88, 89, ....
You don't have zero to spend. You have your Social Security + SPIA.
It's not even that bad. First, the idea that retirees will blindly spend their portfolio down to zero is a myth. Second, many retirees will never survive to age 85, although many here seem to be acting as though they will live to 120, despite those over the age of 70 only having gotten about one extra year of life expectancy per decade since the 1950s.
This is the chart that makes the comparison to the 1950's a little more real for me, especially for upper-middle class people.

Image


And having dealt with a little too much cognitive decline caregiving I'm much more concerned about quality of life then early death.
It's interesting to see how framing a message in one way impacts our interpretation of it. Most people would react to their statement differently if it were worded like this: "Almost two out of three women and four out of five men will not reach age 90."
Last edited by willthrill81 on Tue Jun 11, 2019 2:01 pm, edited 1 time in total.
“It's a dangerous business, Frodo, going out your door. You step onto the road, and if you don't keep your feet, there's no knowing where you might be swept off to.” J.R.R. Tolkien,The Lord of the Rings

bradshaw1965
Posts: 743
Joined: Tue Jul 31, 2007 9:36 pm

Re: Downsides of VPW?

Post by bradshaw1965 » Tue Jun 11, 2019 1:00 pm

willthrill81 wrote:
Tue Jun 11, 2019 11:53 am
bradshaw1965 wrote:
Tue Jun 11, 2019 11:31 am
willthrill81 wrote:
Wed Jun 05, 2019 12:11 am
AlohaJoe wrote:
Tue Jun 04, 2019 10:58 pm
bhsince87 wrote:
Tue Jun 04, 2019 10:20 pm
Yes, I'd much prefer spending 28k at 75 and 80k at 95 than having nothing left to spend at, say, age 85.

And zero to spend at age 86, 87, 88, 89, ....
You don't have zero to spend. You have your Social Security + SPIA.
It's not even that bad. First, the idea that retirees will blindly spend their portfolio down to zero is a myth. Second, many retirees will never survive to age 85, although many here seem to be acting as though they will live to 120, despite those over the age of 70 only having gotten about one extra year of life expectancy per decade since the 1950s.
This is the chart that makes the comparison to the 1950's a little more real for me, especially for upper-middle class people.

Image


And having dealt with a little too much cognitive decline caregiving I'm much more concerned about quality of life then early death.
It's interesting to see how framing a message in one way impacts our interpretation of it. Most people would react to their statement differently if it were worded like this: "Almost two out of three women and four out of five men will reach age 90."
Makes sense. I'm definitely not embracing the idea of significant life extension. I am however thinking the probability of reaching the upper edges of what we think of as old and probably frail or with significant cognitive decline are pretty high.

bhsince87
Posts: 2282
Joined: Thu Oct 03, 2013 1:08 pm

Re: Downsides of VPW?

Post by bhsince87 » Tue Jun 11, 2019 1:37 pm

willthrill81 wrote:
Tue Jun 11, 2019 11:53 am
bradshaw1965 wrote:
Tue Jun 11, 2019 11:31 am
willthrill81 wrote:
Wed Jun 05, 2019 12:11 am
AlohaJoe wrote:
Tue Jun 04, 2019 10:58 pm
bhsince87 wrote:
Tue Jun 04, 2019 10:20 pm
Yes, I'd much prefer spending 28k at 75 and 80k at 95 than having nothing left to spend at, say, age 85.

And zero to spend at age 86, 87, 88, 89, ....
You don't have zero to spend. You have your Social Security + SPIA.
It's not even that bad. First, the idea that retirees will blindly spend their portfolio down to zero is a myth. Second, many retirees will never survive to age 85, although many here seem to be acting as though they will live to 120, despite those over the age of 70 only having gotten about one extra year of life expectancy per decade since the 1950s.
This is the chart that makes the comparison to the 1950's a little more real for me, especially for upper-middle class people.

Image


And having dealt with a little too much cognitive decline caregiving I'm much more concerned about quality of life then early death.
It's interesting to see how framing a message in one way impacts our interpretation of it. Most people would react to their statement differently if it were worded like this: "Almost two out of three women and four out of five men will reach age 90."

Yes that would certainly impact the way I interpret it.

What you said is completely wrong, though, according to the chart. Did you accidentally leave out the word "not"?
Retirement: When you reach a point where you have enough. Or when you've had enough.

User avatar
willthrill81
Posts: 10157
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: Downsides of VPW?

Post by willthrill81 » Tue Jun 11, 2019 2:01 pm

bhsince87 wrote:
Tue Jun 11, 2019 1:37 pm
What you said is completely wrong, though, according to the chart. Did you accidentally leave out the word "not"?
Yes, and I've corrected my post.
“It's a dangerous business, Frodo, going out your door. You step onto the road, and if you don't keep your feet, there's no knowing where you might be swept off to.” J.R.R. Tolkien,The Lord of the Rings

User avatar
willthrill81
Posts: 10157
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: Downsides of VPW?

Post by willthrill81 » Tue Jun 11, 2019 2:03 pm

bradshaw1965 wrote:
Tue Jun 11, 2019 1:00 pm
Makes sense. I'm definitely not embracing the idea of significant life extension. I am however thinking the probability of reaching the upper edges of what we think of as old and probably frail or with significant cognitive decline are pretty high.
It depends on what you believe "pretty high" to be. Some believe a 20% probability (e.g. likelihood of a 65 year old American male to reach age 90) to be high, while others do not.
“It's a dangerous business, Frodo, going out your door. You step onto the road, and if you don't keep your feet, there's no knowing where you might be swept off to.” J.R.R. Tolkien,The Lord of the Rings

bradshaw1965
Posts: 743
Joined: Tue Jul 31, 2007 9:36 pm

Re: Downsides of VPW?

Post by bradshaw1965 » Tue Jun 11, 2019 3:48 pm

willthrill81 wrote:
Tue Jun 11, 2019 2:03 pm
bradshaw1965 wrote:
Tue Jun 11, 2019 1:00 pm
Makes sense. I'm definitely not embracing the idea of significant life extension. I am however thinking the probability of reaching the upper edges of what we think of as old and probably frail or with significant cognitive decline are pretty high.
It depends on what you believe "pretty high" to be. Some believe a 20% probability (e.g. likelihood of a 65 year old American male to reach age 90) to be high, while others do not.
factor in this:
The Rich Live Longer Everywhere. For the Poor, Geography Matters.
https://www.nytimes.com/interactive/201 ... death.html

and the average Boglehead stretches up to an 88 year old life expectancy for men. In 30 years it's within reason to think that's 90.

Again, not dramatic life extension but if it's a coin flip to hit 90 I'm much more worried that I'm not frail and have compressed morbidity at the end of life. I feel like I'm hijacking the thread but a Boglehead should take all of this into account.

longinvest
Posts: 3485
Joined: Sat Aug 11, 2012 8:44 am

Re: Downsides of VPW?

Post by longinvest » Tue Jun 11, 2019 6:58 pm

One Ping wrote:
Tue Jun 11, 2019 10:13 am
Cool. Thanks, longinvest. I look forward to checking it out.
Just make sure to get version 1.4 (or newer).
Bogleheads investment philosophy | single-ETF balanced portfolio | VBAL

User avatar
gordoni2
Posts: 196
Joined: Wed Aug 15, 2007 12:20 am
Contact:

Re: Downsides of VPW?

Post by gordoni2 » Tue Jun 11, 2019 10:44 pm

As I see it VPW and other PMT() based schemes have two main downsides:

First VPW is blind to risk aversion. Most people care more about a consumption decline of 50% than a consumption increase of 50%. To guard against a drastic consumption decline I suspect you should use a lower assumed rate of return than used by VPW. Note that Merton's portfolio problem provides an exact solution to how much to consume (and what asset allocation to use) in the absence of guaranteed income. I'm not volunteering, but it would be an interesting exercise to compare the recommendations of VPW to the exact results of Merton/Samuelson for different risk aversions.

Second VPW operates over a fixed number of years. In reality the probability of reaching various ages gradually declines. It would be nice to have a scheme that reflected this by placing less and less weight on the outcomes for older and older ages. Dynamic programming can be used to overcome both these downsides, but would be challenging to implement in a spreadsheet.

That said, VPW and PMT() schemes, even if only followed loosely, are far superior to SWR and any other fixed consumption scheme.

User avatar
willthrill81
Posts: 10157
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: Downsides of VPW?

Post by willthrill81 » Tue Jun 11, 2019 10:56 pm

gordoni2 wrote:
Tue Jun 11, 2019 10:44 pm
As I see it VPW and other PMT() based schemes have two main downsides:

First VPW is blind to risk aversion. Most people care more about a consumption decline of 50% than a consumption increase of 50%. To guard against a drastic consumption decline I suspect you should use a lower assumed rate of return than used by VPW. Note that Merton's portfolio problem provides an exact solution to how much to consume (and what asset allocation to use) in the absence of guaranteed income. I'm not volunteering, but it would be an interesting exercise to compare the recommendations of VPW to the exact results of Merton/Samuelson for different risk aversions.
Any PMT based approach, of which VPW is only one, can be as conservative as the investor wishes. This conservatism can be manifested in either the portfolio or in the number of years over which the payments are spread.
gordoni2 wrote:
Tue Jun 11, 2019 10:44 pm
Second VPW operates over a fixed number of years. In reality the probability of reaching various ages gradually declines. It would be nice to have a scheme that reflected this by placing less and less weight on the outcomes for older and older ages. Dynamic programming can be used to overcome both these downsides, but would be challenging to implement in a spreadsheet.
I prefer a PMT approach to the, in my mind, overly simplistic VPW approach for several reasons, one of which is because I wish to front-load my withdrawals in retirement somewhat so that we have more money when we're younger, healthier, and, most importantly, still alive. I plan to accomplish this front-loading by overestimating my portfolio's returns by 1-2% annually.
“It's a dangerous business, Frodo, going out your door. You step onto the road, and if you don't keep your feet, there's no knowing where you might be swept off to.” J.R.R. Tolkien,The Lord of the Rings

trueblueky
Posts: 1482
Joined: Tue May 27, 2014 3:50 pm

Re: Downsides of VPW?

Post by trueblueky » Wed Jun 12, 2019 7:36 am

gordoni2 wrote:
Tue Jun 11, 2019 10:44 pm
As I see it VPW and other PMT() based schemes have two main downsides:

First VPW is blind to risk aversion. Most people care more about a consumption decline of 50% than a consumption increase of 50%. To guard against a drastic consumption decline I suspect you should use a lower assumed rate of return than used by VPW. Note that Merton's portfolio problem provides an exact solution to how much to consume (and what asset allocation to use) in the absence of guaranteed income. I'm not volunteering, but it would be an interesting exercise to compare the recommendations of VPW to the exact results of Merton/Samuelson for different risk aversions.
One could maintain a floor through Social Security, pension, or SPIA and use VPW for the rest.

Second VPW operates over a fixed number of years. In reality the probability of reaching various ages gradually declines. It would be nice to have a scheme that reflected this by placing less and less weight on the outcomes for older and older ages. Dynamic programming can be used to overcome both these downsides, but would be challenging to implement in a spreadsheet.

That said, VPW and PMT() schemes, even if only followed loosely, are far superior to SWR and any other fixed consumption scheme.

jmk
Posts: 452
Joined: Tue Nov 01, 2011 7:48 pm

Re: Downsides of VPW?

Post by jmk » Wed Jun 12, 2019 7:53 am

gordoni2 wrote:
Tue Jun 11, 2019 10:44 pm
As I see it VPW and other PMT() based schemes have two main downsides:

First VPW is blind to risk aversion. Most people care more about a consumption decline of 50% than a consumption increase of 50%. To guard against a drastic consumption decline I suspect you should use a lower assumed rate of return than used by VPW.
1) You can set any discount rate you want with PMT()and and I personally use the risk free rate for just this reason. (The consequence of this of course is withdrawals will likely go up with time, with more money toward end of life, which many may not want.). Other folks use mixed discount rates for different parts of the portfolio (like the risk free rate for necessities but ER for withdrawals of discretionary income)—in other words pmt buckets

2) You can also add an informal utility adjustment into any year to reflect changing risk aversion and/or changing utility of money. For instance if I value a dollar more in my 60s than 90s, I might add 10% per year to the PMT() calculation for those years.
Second VPW operates over a fixed number of years. In reality the probability of reaching various ages gradually declines. It would be nice to have a scheme that reflected this by placing less and less weight on the outcomes for older and older ages. Dynamic programming can be used to overcome both these downsides, but would be challenging to implement in a spreadsheet.
You don’t need dynamic programming: each year you use the 25th or 10th percentile for longevity from standard social security or insurance tables in the pmt() formula. So while n might be 93 when one is 55, it might be 98 when one is 70.
Last edited by jmk on Wed Jun 12, 2019 8:00 am, edited 2 times in total.

azanon
Posts: 2360
Joined: Mon Nov 07, 2011 10:34 am
Location: Little Rock, AR
Contact:

Re: Downsides of VPW?

Post by azanon » Wed Jun 12, 2019 8:00 am

willthrill81 wrote:
Tue Jun 11, 2019 10:56 pm
gordoni2 wrote:
Tue Jun 11, 2019 10:44 pm
As I see it VPW and other PMT() based schemes have two main downsides:

First VPW is blind to risk aversion. Most people care more about a consumption decline of 50% than a consumption increase of 50%. To guard against a drastic consumption decline I suspect you should use a lower assumed rate of return than used by VPW. Note that Merton's portfolio problem provides an exact solution to how much to consume (and what asset allocation to use) in the absence of guaranteed income. I'm not volunteering, but it would be an interesting exercise to compare the recommendations of VPW to the exact results of Merton/Samuelson for different risk aversions.
Any PMT based approach, of which VPW is only one, can be as conservative as the investor wishes. This conservatism can be manifested in either the portfolio or in the number of years over which the payments are spread.
gordoni2 wrote:
Tue Jun 11, 2019 10:44 pm
Second VPW operates over a fixed number of years. In reality the probability of reaching various ages gradually declines. It would be nice to have a scheme that reflected this by placing less and less weight on the outcomes for older and older ages. Dynamic programming can be used to overcome both these downsides, but would be challenging to implement in a spreadsheet.
I prefer a PMT approach to the, in my mind, overly simplistic VPW approach for several reasons, one of which is because I wish to front-load my withdrawals in retirement somewhat so that we have more money when we're younger, healthier, and, most importantly, still alive. I plan to accomplish this front-loading by overestimating my portfolio's returns by 1-2% annually.
Yeah I don't know why VPW gets criticized so much for its variability in year-to-year payments, and no one seems to want to blame an insistence of so many to hold a higher percentage of equities in retirement with no working capital left than even the global market portfolio has (GMP has ~ 43% equities, give or take a few percent).

Did he say the probably of reaching certain ages decline over time? I'm pretty sure it rises, actually.

User avatar
willthrill81
Posts: 10157
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: Downsides of VPW?

Post by willthrill81 » Wed Jun 12, 2019 9:28 am

azanon wrote:
Wed Jun 12, 2019 8:00 am
willthrill81 wrote:
Tue Jun 11, 2019 10:56 pm
gordoni2 wrote:
Tue Jun 11, 2019 10:44 pm
As I see it VPW and other PMT() based schemes have two main downsides:

First VPW is blind to risk aversion. Most people care more about a consumption decline of 50% than a consumption increase of 50%. To guard against a drastic consumption decline I suspect you should use a lower assumed rate of return than used by VPW. Note that Merton's portfolio problem provides an exact solution to how much to consume (and what asset allocation to use) in the absence of guaranteed income. I'm not volunteering, but it would be an interesting exercise to compare the recommendations of VPW to the exact results of Merton/Samuelson for different risk aversions.
Any PMT based approach, of which VPW is only one, can be as conservative as the investor wishes. This conservatism can be manifested in either the portfolio or in the number of years over which the payments are spread.
gordoni2 wrote:
Tue Jun 11, 2019 10:44 pm
Second VPW operates over a fixed number of years. In reality the probability of reaching various ages gradually declines. It would be nice to have a scheme that reflected this by placing less and less weight on the outcomes for older and older ages. Dynamic programming can be used to overcome both these downsides, but would be challenging to implement in a spreadsheet.
I prefer a PMT approach to the, in my mind, overly simplistic VPW approach for several reasons, one of which is because I wish to front-load my withdrawals in retirement somewhat so that we have more money when we're younger, healthier, and, most importantly, still alive. I plan to accomplish this front-loading by overestimating my portfolio's returns by 1-2% annually.
Yeah I don't know why VPW gets criticized so much for its variability in year-to-year payments, and no one seems to want to blame an insistence of so many to hold a higher percentage of equities in retirement with no working capital left than even the global market portfolio has (GMP has ~ 43% equities, give or take a few percent).
Post #10,000!! :D

Many don't find the GMP percentages to be compelling for any individual, and the fact that SWRs have historically been maximized with equity holdings being 50-75% of the portfolio is persuasive to many. But the bottom line is that one can change one's portfolio to be as conservative as desirable in order to smooth out withdrawals. If one is using a PMT approach, withdrawals can be smoothed out significantly by using 1/CAPE as a forward return estimate for stocks rather than the straight historic returns that VPW uses; current TIPS yield can be used as a forward return for bonds of the same duration.
“It's a dangerous business, Frodo, going out your door. You step onto the road, and if you don't keep your feet, there's no knowing where you might be swept off to.” J.R.R. Tolkien,The Lord of the Rings

User avatar
siamond
Posts: 4744
Joined: Mon May 28, 2012 5:50 am

Re: Downsides of VPW?

Post by siamond » Wed Jun 12, 2019 9:51 am

randomguy wrote:
Tue Jun 04, 2019 9:00 am
azanon wrote:
Tue Jun 04, 2019 8:24 am
If I'm understanding this right, the catch to that alternative method is that it introduces some Sequencing of Return Risk (SoRR), which is arguably the greatest risk a portfolio is exposed to in early retirement. VPW has no SoRR.
Yeah but it has a lot of sequence of income risk.
This is a terrific wording. Sequence of income risk vs. sequence of returns risk. This is very concisely put while capturing the crux of the issue very well. Well done, sir.

PS. McClung opposed 'loss of income' risk to 'loss of value' risk. Similar idea, although he was more focused on long-term issues than year-over-year issues.

azanon
Posts: 2360
Joined: Mon Nov 07, 2011 10:34 am
Location: Little Rock, AR
Contact:

Re: Downsides of VPW?

Post by azanon » Wed Jun 12, 2019 10:03 am

willthrill81 wrote:
Wed Jun 12, 2019 9:28 am
azanon wrote:
Wed Jun 12, 2019 8:00 am
willthrill81 wrote:
Tue Jun 11, 2019 10:56 pm
gordoni2 wrote:
Tue Jun 11, 2019 10:44 pm
As I see it VPW and other PMT() based schemes have two main downsides:

First VPW is blind to risk aversion. Most people care more about a consumption decline of 50% than a consumption increase of 50%. To guard against a drastic consumption decline I suspect you should use a lower assumed rate of return than used by VPW. Note that Merton's portfolio problem provides an exact solution to how much to consume (and what asset allocation to use) in the absence of guaranteed income. I'm not volunteering, but it would be an interesting exercise to compare the recommendations of VPW to the exact results of Merton/Samuelson for different risk aversions.
Any PMT based approach, of which VPW is only one, can be as conservative as the investor wishes. This conservatism can be manifested in either the portfolio or in the number of years over which the payments are spread.
gordoni2 wrote:
Tue Jun 11, 2019 10:44 pm
Second VPW operates over a fixed number of years. In reality the probability of reaching various ages gradually declines. It would be nice to have a scheme that reflected this by placing less and less weight on the outcomes for older and older ages. Dynamic programming can be used to overcome both these downsides, but would be challenging to implement in a spreadsheet.
I prefer a PMT approach to the, in my mind, overly simplistic VPW approach for several reasons, one of which is because I wish to front-load my withdrawals in retirement somewhat so that we have more money when we're younger, healthier, and, most importantly, still alive. I plan to accomplish this front-loading by overestimating my portfolio's returns by 1-2% annually.
Yeah I don't know why VPW gets criticized so much for its variability in year-to-year payments, and no one seems to want to blame an insistence of so many to hold a higher percentage of equities in retirement with no working capital left than even the global market portfolio has (GMP has ~ 43% equities, give or take a few percent).
Post #10,000!! :D

Many don't find the GMP percentages to be compelling for any individual, and the fact that SWRs have historically been maximized with equity holdings being 50-75% of the portfolio is persuasive to many. But the bottom line is that one can change one's portfolio to be as conservative as desirable in order to smooth out withdrawals. If one is using a PMT approach, withdrawals can be smoothed out significantly by using 1/CAPE as a forward return estimate for stocks rather than the straight historic returns that VPW uses; current TIPS yield can be used as a forward return for bonds of the same duration.
An alternative POV is that any tilt away from a GMP is an active bet. Also, my POV would be that, if anything, one single individual should be more conservative, if anything, vs. a GMP because a single individual bears more risk, per se, than the collective global markets do. Or perhaps a better way to say that, is the collective global markets is going to include institutional investors which should have more risk capacity than a single investor.

But that was actually all besides my point. My point was "your" VPW is probably volatile primarily because your #1 asset is stocks, not because it's VPW. So I"m just saying I think it's unfair to accuse the method of causing the volatility if one is insistent on having an equity percentage higher than, say, Vanguard Target Retirement Income fund.

randomguy
Posts: 7677
Joined: Wed Sep 17, 2014 9:00 am

Re: Downsides of VPW?

Post by randomguy » Wed Jun 12, 2019 10:11 am

bradshaw1965 wrote:
Tue Jun 11, 2019 1:00 pm

Makes sense. I'm definitely not embracing the idea of significant life exension. I am however thinking the probability of reaching the upper edges of what we think of as old and probably frail or with significant cognitive decline are pretty high.
Longevity averages are very deceptive. Most people make it to 65 no matter how much they abuse their bodies. After that it starts catching up to you.

As a healthy male, nonsmoker who isn't obese, at age 65 my life expectancy is ~90-92 depending what calculator. As a smoker I would expect to be dead at 80. Have some health issues and that number can drop to 75. Obviously these are all probabilities. The smoker can't count on dying at 80 or the nonsmoker living to 92. The advantages of the higher numbers is there is less upside. Some smoker is going to live to 95 and beat their estimate by 15 years. Basically nobody is going to make it to 107:)

azanon
Posts: 2360
Joined: Mon Nov 07, 2011 10:34 am
Location: Little Rock, AR
Contact:

Re: Downsides of VPW?

Post by azanon » Wed Jun 12, 2019 11:08 am

siamond wrote:
Wed Jun 12, 2019 9:51 am
randomguy wrote:
Tue Jun 04, 2019 9:00 am
azanon wrote:
Tue Jun 04, 2019 8:24 am
If I'm understanding this right, the catch to that alternative method is that it introduces some Sequencing of Return Risk (SoRR), which is arguably the greatest risk a portfolio is exposed to in early retirement. VPW has no SoRR.
Yeah but it has a lot of sequence of income risk.
This is a terrific wording. Sequence of income risk vs. sequence of returns risk. This is very concisely put while capturing the crux of the issue very well. Well done, sir.

PS. McClung opposed 'loss of income' risk to 'loss of value' risk. Similar idea, although he was more focused on long-term issues than year-over-year issues.
I like the wording too, but I'd add that if someone accomplished the initial steps of VPW correctly (put into place a floor level of income to cover basic expenses), then the sequence of income risk with VPW should be zero or near zero. So VPW does have a "you calculated your basic living expenses wrong, risk" though. Or, if someone chose to roll the dice and not build that initial layer high enough, and put needed money in the VPW portfolio instead, then there using some different method of retirement income that sort of resembles VPW but it isn't VPW.

User avatar
siamond
Posts: 4744
Joined: Mon May 28, 2012 5:50 am

Re: Downsides of VPW?

Post by siamond » Wed Jun 12, 2019 11:09 am

azanon wrote:
Wed Jun 12, 2019 10:03 am
My point was "your" VPW is probably volatile primarily because your #1 asset is stocks, not because it's VPW. So I"m just saying I think it's unfair to accuse the method of causing the volatility if one is insistent on having an equity percentage higher than, say, Vanguard Target Retirement Income fund.
Well, no, it's both. For sure, the portfolio's volatility (more precisely its current value) is an input to the PMT formula underlying VPW and the outputs are directly dependent on the inputs, hence the point you're making, the more volatile the AA, the bigger the sequence of income risk. BUT. One also needs to acknowledge that the PMT formula has the clear mathematical characteristic of NOT providing any dampening whatsoever. Other withdrawal methods (G-K and others) have ingrained dampening mechanisms. So... it's both (the portfolio's AA *and* the specifics of the withdrawal method).

Please note that my previous paragraph is NOT judgmental, I am well aware of the fact that some people like this lack of dampening while other people have a strong problem with it. This isn't my point. My point is just to be factual, the sequence of income is a direct consequence of BOTH the portfolio's volatility AND the mathematical nature of VPW.

randomguy
Posts: 7677
Joined: Wed Sep 17, 2014 9:00 am

Re: Downsides of VPW?

Post by randomguy » Wed Jun 12, 2019 12:22 pm

azanon wrote:
Wed Jun 12, 2019 11:08 am
siamond wrote:
Wed Jun 12, 2019 9:51 am
randomguy wrote:
Tue Jun 04, 2019 9:00 am
azanon wrote:
Tue Jun 04, 2019 8:24 am
If I'm understanding this right, the catch to that alternative method is that it introduces some Sequencing of Return Risk (SoRR), which is arguably the greatest risk a portfolio is exposed to in early retirement. VPW has no SoRR.
Yeah but it has a lot of sequence of income risk.
This is a terrific wording. Sequence of income risk vs. sequence of returns risk. This is very concisely put while capturing the crux of the issue very well. Well done, sir.

PS. McClung opposed 'loss of income' risk to 'loss of value' risk. Similar idea, although he was more focused on long-term issues than year-over-year issues.
I like the wording too, but I'd add that if someone accomplished the initial steps of VPW correctly (put into place a floor level of income to cover basic expenses), then the sequence of income risk with VPW should be zero or near zero. So VPW does have a "you calculated your basic living expenses wrong, risk" though. Or, if someone chose to roll the dice and not build that initial layer high enough, and put needed money in the VPW portfolio instead, then there using some different method of retirement income that sort of resembles VPW but it isn't VPW.
Any scheme works if you don't need it to do anything. Who cares if I am using the 7% rule and run out of money since I have income streams that cover all my expenses?


And it should be pointed out that bonds don't really help with the income volatility issue.
70/30 52k->26k
30/70 44k->23k
Your income decline is a bit smoother with bonds but you still have to deal with about a 50% cut in spending over your 1st 15 years of retirement. The issue isn't stock volatility. It is poor returns from both stocks and bonds over the rate limiting period.

azanon
Posts: 2360
Joined: Mon Nov 07, 2011 10:34 am
Location: Little Rock, AR
Contact:

Re: Downsides of VPW?

Post by azanon » Wed Jun 12, 2019 12:50 pm

randomguy wrote:
Wed Jun 12, 2019 12:22 pm
azanon wrote:
Wed Jun 12, 2019 11:08 am
siamond wrote:
Wed Jun 12, 2019 9:51 am
randomguy wrote:
Tue Jun 04, 2019 9:00 am
azanon wrote:
Tue Jun 04, 2019 8:24 am
If I'm understanding this right, the catch to that alternative method is that it introduces some Sequencing of Return Risk (SoRR), which is arguably the greatest risk a portfolio is exposed to in early retirement. VPW has no SoRR.
Yeah but it has a lot of sequence of income risk.
This is a terrific wording. Sequence of income risk vs. sequence of returns risk. This is very concisely put while capturing the crux of the issue very well. Well done, sir.

PS. McClung opposed 'loss of income' risk to 'loss of value' risk. Similar idea, although he was more focused on long-term issues than year-over-year issues.
I like the wording too, but I'd add that if someone accomplished the initial steps of VPW correctly (put into place a floor level of income to cover basic expenses), then the sequence of income risk with VPW should be zero or near zero. So VPW does have a "you calculated your basic living expenses wrong, risk" though. Or, if someone chose to roll the dice and not build that initial layer high enough, and put needed money in the VPW portfolio instead, then there using some different method of retirement income that sort of resembles VPW but it isn't VPW.
Any scheme works if you don't need it to do anything. Who cares if I am using the 7% rule and run out of money since I have income streams that cover all my expenses?


And it should be pointed out that bonds don't really help with the income volatility issue.
70/30 52k->26k
30/70 44k->23k
Your income decline is a bit smoother with bonds but you still have to deal with about a 50% cut in spending over your 1st 15 years of retirement. The issue isn't stock volatility. It is poor returns from both stocks and bonds over the rate limiting period.
Is this where i say, you caught me? That any money that you absolutely must have to live on shouldn't be invested in the stock market? Oh well, my run was good while it lasted. Yes, I think you need a retirement income method that works 100% of the time. So do a lot of retirement income experts (I'll spare you the list).

The actual VPW method, at the very least, makes this implication upfront so I don't know what to say about this critique. Yes, because of the instructions, it has no way of failing. That's right.

It's easier to adjust to more gradual spending declines. I can say that as matter-of-fact, right, or consider it an opinion if you must.

azanon
Posts: 2360
Joined: Mon Nov 07, 2011 10:34 am
Location: Little Rock, AR
Contact:

Re: Downsides of VPW?

Post by azanon » Wed Jun 12, 2019 12:59 pm

siamond wrote:
Wed Jun 12, 2019 11:09 am
azanon wrote:
Wed Jun 12, 2019 10:03 am
My point was "your" VPW is probably volatile primarily because your #1 asset is stocks, not because it's VPW. So I"m just saying I think it's unfair to accuse the method of causing the volatility if one is insistent on having an equity percentage higher than, say, Vanguard Target Retirement Income fund.
Well, no, it's both. For sure, the portfolio's volatility (more precisely its current value) is an input to the PMT formula underlying VPW and the outputs are directly dependent on the inputs, hence the point you're making, the more volatile the AA, the bigger the sequence of income risk. BUT. One also needs to acknowledge that the PMT formula has the clear mathematical characteristic of NOT providing any dampening whatsoever. Other withdrawal methods (G-K and others) have ingrained dampening mechanisms. So... it's both (the portfolio's AA *and* the specifics of the withdrawal method).

Please note that my previous paragraph is NOT judgmental, I am well aware of the fact that some people like this lack of dampening while other people have a strong problem with it. This isn't my point. My point is just to be factual, the sequence of income is a direct consequence of BOTH the portfolio's volatility AND the mathematical nature of VPW.
I knew I should have bolded "primarily". And of course there's going to be some volatility at any equity weighting, it's just when does that become a hardship. In theory, it should never become a hardship since this is for discretionary expenses.

I've said it already - I probably wouldn't use VPW for higher than 50%. For those that didn't notice, that actually includes longinvest, since I think his current balanced fund has 60%. But that's my opinion. For 60%, I'd probably prefer something like what Vanguard Managed payout uses (a 3-yr-smoothed, constant % method). I mean, why not. If you're going to have an endowment level of equities more appropriate for growing a portfolio, then at least use a perpetual method of withdrawal that can also sustain the inflation-adjusted principle.

Post Reply