50x minimum expenses to use VPW?

Discuss all general (i.e. non-personal) investing questions and issues, investing news, and theory.
Post Reply
Topic Author
mortal
Posts: 452
Joined: Mon Sep 15, 2008 11:15 pm

50x minimum expenses to use VPW?

Post by mortal » Wed Apr 24, 2019 8:39 pm

Howdy folks,

One thing I've noticed with VPW, when one looks at the 'Withdrawal Stats for Every Year' chart, is that the 'lower bound' for the inflation adjusted minimum withdrawal sometimes hovers close to 2% SWR.

Does this mean one should have about 50x their minimum expenses before they consider using VPW?

User avatar
Peter Foley
Posts: 4869
Joined: Fri Nov 23, 2007 10:34 am
Location: Lake Wobegon

Re: 50x minimum expenses to use VPW?

Post by Peter Foley » Wed Apr 24, 2019 9:52 pm

And VPW means?

stlutz
Posts: 5042
Joined: Fri Jan 02, 2009 1:08 am

Re: 50x minimum expenses to use VPW?

Post by stlutz » Wed Apr 24, 2019 10:03 pm

No. VPW is focused on consumption that is variable. You should have a base of other income as an underlying foundation. This includes things like Social Security, a pension, an annuity or a TIPS ladder. The allows you to just get by if you don't have other income.

Anything above that level would be covered by the variable portfolio.

VPW guarantees that you won't run out of money early; it doesn't guarantee any particular level of spending. And that is the reality of relying assets that can significantly change in price to fund consumption. For me, that's what I like about the VPW approach--it is relentlessly reality-based. What you can consume now is based on how much money you have now, and not on how much you had last year or hope to have next year.

quantAndHold
Posts: 3034
Joined: Thu Sep 17, 2015 10:39 pm

Re: 50x minimum expenses to use VPW?

Post by quantAndHold » Wed Apr 24, 2019 10:05 pm

I assume VPW means Variable Percentage Withdrawal, discussed on this wiki page.

I’m not understanding where you’re getting 2%. Can you explain?

Topic Author
mortal
Posts: 452
Joined: Mon Sep 15, 2008 11:15 pm

Re: 50x minimum expenses to use VPW?

Post by mortal » Thu Apr 25, 2019 6:20 pm

My apologies, I didn't do a very good job at explaining myself. Yes, VPW is variable percentage withdrawal, and I got the 2% SWR from graphs like the one below.
Image

This is 1M with a 60 / 40 stock / bond mix. Real withdrawals for someone retiring in the 1914s (ok, I cherry picked a pessimistic start date ) start at ~ 42k but quickly fall down to 22k before gradually recovering. Social security wouldn't be an option for me until about 20 years in because I'm looking at early retirement in my 40s. I might have to look further into a tips ladder or annuities but honestly, I was under the impression that's what my 40% in (g fund) bonds were for.
Last edited by mortal on Thu Apr 25, 2019 6:34 pm, edited 2 times in total.

User avatar
Phineas J. Whoopee
Posts: 8255
Joined: Sun Dec 18, 2011 6:18 pm

Re: 50x minimum expenses to use VPW?

Post by Phineas J. Whoopee » Thu Apr 25, 2019 6:26 pm

As written upthread, the Variable Percentage Withdrawal plan assumes your basic expenses are covered by something other than the part of your portfolio you're managing with VPW. They might be covered by some other part of your portfolio, or by another thing or things.

Nice work if you can get it.

PJW

User avatar
billthecat
Posts: 359
Joined: Tue Jan 24, 2017 2:50 pm

Re: 50x minimum expenses to use VPW?

Post by billthecat » Thu Apr 25, 2019 6:35 pm

Phineas J. Whoopee wrote:
Thu Apr 25, 2019 6:26 pm
As written upthread, the Variable Percentage Withdrawal plan assumes your basic expenses are covered by something other than the part of your portfolio you're managing with VPW. They might be covered by some other part of your portfolio, or by another thing or things.

Nice work if you can get it.

PJW
Does the VPW backtesting allow you to cap the spending at something less than what VPW allows? Yes, I know you don't have to spend what it allows - I'm interested in knowing the backtesting results if I cap the spending.
We cannot direct the winds but we can adjust our sails.

User avatar
billthecat
Posts: 359
Joined: Tue Jan 24, 2017 2:50 pm

Re: 50x minimum expenses to use VPW?

Post by billthecat » Thu Apr 25, 2019 6:37 pm

stlutz wrote:
Wed Apr 24, 2019 10:03 pm
No. VPW is focused on consumption that is variable. You should have a base of other income as an underlying foundation. This includes things like Social Security, a pension, an annuity or a TIPS ladder. The allows you to just get by if you don't have other income.

Anything above that level would be covered by the variable portfolio.

VPW guarantees that you won't run out of money early; it doesn't guarantee any particular level of spending. And that is the reality of relying assets that can significantly change in price to fund consumption. For me, that's what I like about the VPW approach--it is relentlessly reality-based. What you can consume now is based on how much money you have now, and not on how much you had last year or hope to have next year.
If you're completely reliant on your portfolio - at least until SS kicks in - how would you isolate VPW to your variable consumption?
We cannot direct the winds but we can adjust our sails.

User avatar
Phineas J. Whoopee
Posts: 8255
Joined: Sun Dec 18, 2011 6:18 pm

Re: 50x minimum expenses to use VPW?

Post by Phineas J. Whoopee » Thu Apr 25, 2019 6:38 pm

billthecat wrote:
Thu Apr 25, 2019 6:35 pm
Phineas J. Whoopee wrote:
Thu Apr 25, 2019 6:26 pm
As written upthread, the Variable Percentage Withdrawal plan assumes your basic expenses are covered by something other than the part of your portfolio you're managing with VPW. They might be covered by some other part of your portfolio, or by another thing or things.

Nice work if you can get it.

PJW
Does the VPW backtesting allow you to cap the spending at something less than what VPW allows? Yes, I know you don't have to spend what it allows - I'm interested in knowing the backtesting results if I cap the spending.
A bigger VPW expert than I will have to answer the detailed question, but backtesting is a dangerous ally. Foretesting would be a good one, but sadly is impossible.

PJW

User avatar
Phineas J. Whoopee
Posts: 8255
Joined: Sun Dec 18, 2011 6:18 pm

Re: 50x minimum expenses to use VPW?

Post by Phineas J. Whoopee » Thu Apr 25, 2019 6:43 pm

billthecat wrote:
Thu Apr 25, 2019 6:37 pm
stlutz wrote:
Wed Apr 24, 2019 10:03 pm
No. VPW is focused on consumption that is variable. You should have a base of other income as an underlying foundation. This includes things like Social Security, a pension, an annuity or a TIPS ladder. The allows you to just get by if you don't have other income.

Anything above that level would be covered by the variable portfolio.

VPW guarantees that you won't run out of money early; it doesn't guarantee any particular level of spending. And that is the reality of relying assets that can significantly change in price to fund consumption. For me, that's what I like about the VPW approach--it is relentlessly reality-based. What you can consume now is based on how much money you have now, and not on how much you had last year or hope to have next year.
If you're completely reliant on your portfolio - at least until SS kicks in - how would you isolate VPW to your variable consumption?
I'm not stlutz, but one would divide their portfolio into two pieces, mentally or otherwise. The first would have characteristics of a liability-matching portfolio. The other would be managed using VPW.

Regardless of how one thinks about one's portfolio or designs one's spreadsheet, one owns the totality of what one owns. There's no escaping that basic truth.

PJW

stlutz
Posts: 5042
Joined: Fri Jan 02, 2009 1:08 am

Re: 50x minimum expenses to use VPW?

Post by stlutz » Thu Apr 25, 2019 8:48 pm

Phineas J. Whoopee wrote:
Thu Apr 25, 2019 6:43 pm

I'm not stlutz, but one would divide their portfolio into two pieces, mentally or otherwise. The first would have characteristics of a liability-matching portfolio. The other would be managed using VPW.

Regardless of how one thinks about one's portfolio or designs one's spreadsheet, one owns the totality of what one owns. There's no escaping that basic truth.

PJW
I am not PWJ but I completely agree. :D What I proposed upthread is ultimately a way to rationally arrive at a sensible asset allocation.

Stock market declines of 50% happen with more regularity than we like to think. High inflation happens now and then as well.

If you need to live off your money and your portfolio declining by half means you have big problems, then starting with a good base of inflation-adjusted fixed income (or equivalent) is a good idea.

But the reality is that most people start retirement with a good base of inflation-adjusted income, namely Social Security. You add that into the VPW charts shown and the numbers look less threatening.

Where those VPW charts do look scary is if you're planning to retire at 50 or something like that and you have to go a good long while with no other income beyond what your portfolio produces. If you're planning to live another 40 or 50 years, then you should expect to have one sustained period of high inflation during that span and 2 market declines of 50%. That's the typical case, not the worst case scenario.

The VPW charts reflect that reality. And that's perhaps the best feature of the VPW approach in my view--it's very much reality-based. You withdraw based on how much money you actually have right now and you don't pretend that stocks and annuities are the same thing.

AlohaJoe
Posts: 4425
Joined: Mon Nov 26, 2007 2:00 pm
Location: Saigon, Vietnam

Re: 50x minimum expenses to use VPW?

Post by AlohaJoe » Thu Apr 25, 2019 9:45 pm

billthecat wrote:
Thu Apr 25, 2019 6:35 pm
Does the VPW backtesting allow you to cap the spending at something less than what VPW allows? Yes, I know you don't have to spend what it allows - I'm interested in knowing the backtesting results if I cap the spending.
Before I provide the answer, first let's ask ourselves why we think this could possibly make a difference. Some kind of belief that low spending is primarily caused by "spending too much" (whatever that might mean) early on in our retirement, deplenishing our portfolio, and forcing us to make deeper cuts later on that we otherwise might have.

Image

Here we cap VPW at $50,000 (real). Keep in mind that you start out at $48,000, so this is essentially saying you never withdraw more than you start out withdrawing.

We can see that it makes essentially no difference, outside of a few years in the 1880s, which suggests that our mental model of how retirement & portfolios work isn't quite right and needs to be updated.

Let's pick a random year that had no real difference -- 1900 -- and zoom in on it.

Image

We can see the cap taking effect but we can also see later on that it doesn't actually help -- the minimum withdrawal is just as low as before.

If we look at the portfolio value over time, the reason why capping doesn't make a difference becomes clear

Image

A cap is just almost never going to save enough money to offset a market decline. A cap might save $10,000 here and there...but when your portfolio loses $200,000 .... you're still down a lot and need to cut your withdrawals.

In this case our hypothetical 1900 retiree saw market declines + high inflation make her portfolio go from $790,000 to $695,000 to $478,000 to $417,000 to $362,000. That's a four year period where the portfolio dropped by $428,000. Even if the "cap" saves you $20,000 a year (which it almost never does) it would take 20 years of savings to make up for that portfolio decline.

Topic Author
mortal
Posts: 452
Joined: Mon Sep 15, 2008 11:15 pm

Re: 50x minimum expenses to use VPW?

Post by mortal » Thu Apr 25, 2019 9:50 pm

I could understand such volatility while being 100% stocks, but this example has 40% of the portfolio in bonds. Isn't this roughly equivalent to a bond ladder? I don't really understand the mechanics of having a split portfolio. Should I have a logical partition where I'm withdrawing a 3% SWR on a portion, and VPW the remainder? That might just be mental accounting, but who knows maybe it's useful?

Edit: I just ran the numbers in the VPW spreadsheet, and was shocked at how well this worked. Maybe this was the entire point of the VPW portfolio and I was too dense to understand it. Have a look at the total income line on this chart! (Sorry for the yellow, I don't control the color of the total income line)

Image

Edit portfolio details:
This is a 1M portfolio that's 60% at 3%SWR, and 40% withdrawn with VPW. I need to think through the overall stocks / bonds allocation a bit more for it to be apples to apples
Last edited by mortal on Thu Apr 25, 2019 10:16 pm, edited 3 times in total.

AlohaJoe
Posts: 4425
Joined: Mon Nov 26, 2007 2:00 pm
Location: Saigon, Vietnam

Re: 50x minimum expenses to use VPW?

Post by AlohaJoe » Thu Apr 25, 2019 10:07 pm

mortal wrote:
Thu Apr 25, 2019 9:50 pm
I could understand such volatility while being 100% stocks, but this example has 40% of the portfolio in bonds. Isn't this roughly equivalent to a bond ladder?
Sure, it is roughly equivalent to a bond ladder. But I'm not sure why that matters? A bond ladder does very little to help maintain your inflation-adjusted portfolio value. Over the time period shown above bonds had real returns like -10%, -7%, -10%, -16%, -11%, -15%. A bond ladder isn't going to do anything to save you from inflation. Following WW1 inflation in the US was 12%, 19%, 17%, 16%. A bond ladder got obliterated and wasn't safe in any way when measured in real terms.

User avatar
billthecat
Posts: 359
Joined: Tue Jan 24, 2017 2:50 pm

Re: 50x minimum expenses to use VPW?

Post by billthecat » Thu Apr 25, 2019 10:49 pm

AlohaJoe wrote:
Thu Apr 25, 2019 9:45 pm
billthecat wrote:
Thu Apr 25, 2019 6:35 pm
Does the VPW backtesting allow you to cap the spending at something less than what VPW allows? Yes, I know you don't have to spend what it allows - I'm interested in knowing the backtesting results if I cap the spending.
<snip>

We can see that it makes essentially no difference, outside of a few years in the 1880s, which suggests that our mental model of how retirement & portfolios work isn't quite right and needs to be updated.

<snip>
That's interesting but hard to believe. You're saying that even though VPW says I can withdraw $130,000, if I limit myself to never going above $84,000, it won't make a difference? That's really counter-intuitive.

When selecting VPW as the spending model under cfiresim, the portfolio lasts longer the lower you make the upper bound (and lower bound) for spending. So your post would seem to undermine the results at cfiresim.
We cannot direct the winds but we can adjust our sails.

User avatar
willthrill81
Posts: 9696
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: 50x minimum expenses to use VPW?

Post by willthrill81 » Thu Apr 25, 2019 11:17 pm

The VPW is just one particular application of the broader time value of money formula. VPW includes some simplistic assumptions that make it user friendly, but these might or might not be particularly useful or appealing to some. For those . An explanation of the concept is discussed in this thread, and a specific example as how this could have played out for year 2000 retirees using this approach was provided in this thread.

Another approach that sounds potentially right up the OP's alley is a new calculator found at Portfolio Charts titled "Retirement Spending." Using data going back to 1970, it allows users to backtest the results of a strategy where they start withdrawing X% of a portfolio each year and then potentially adjust that percentage up or down based on portfolio performance. Importantly for the OP, it also allows the user to specify a spending floor and ceiling in relation to the starting percentage (e.g. starting 5% of portfolio each year, reduced by a maximum of 5% each year that the portfolio is down to a minimum of 3% of the portfolio each year). Account triggers can also be modeled (e.g. don't make any reductions in the percentage withdrawn unless the portfolio drops by at least 25% of its starting inflation-adjusted value).
“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: 9696
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: 50x minimum expenses to use VPW?

Post by willthrill81 » Thu Apr 25, 2019 11:24 pm

billthecat wrote:
Thu Apr 25, 2019 10:49 pm
AlohaJoe wrote:
Thu Apr 25, 2019 9:45 pm
billthecat wrote:
Thu Apr 25, 2019 6:35 pm
Does the VPW backtesting allow you to cap the spending at something less than what VPW allows? Yes, I know you don't have to spend what it allows - I'm interested in knowing the backtesting results if I cap the spending.
<snip>

We can see that it makes essentially no difference, outside of a few years in the 1880s, which suggests that our mental model of how retirement & portfolios work isn't quite right and needs to be updated.

<snip>
That's interesting but hard to believe. You're saying that even though VPW says I can withdraw $130,000, if I limit myself to never going above $84,000, it won't make a difference? That's really counter-intuitive.

When selecting VPW as the spending model under cfiresim, the portfolio lasts longer the lower you make the upper bound (and lower bound) for spending. So your post would seem to undermine the results at cfiresim.
It's mathematically impossible for VPW to prematurely deplete the portfolio, so I don't know what you mean by the "portfolio lasting longer" by introducing a withdrawal ceiling.
“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

AlohaJoe
Posts: 4425
Joined: Mon Nov 26, 2007 2:00 pm
Location: Saigon, Vietnam

Re: 50x minimum expenses to use VPW?

Post by AlohaJoe » Fri Apr 26, 2019 1:14 am

billthecat wrote:
Thu Apr 25, 2019 10:49 pm
AlohaJoe wrote:
Thu Apr 25, 2019 9:45 pm
billthecat wrote:
Thu Apr 25, 2019 6:35 pm
Does the VPW backtesting allow you to cap the spending at something less than what VPW allows? Yes, I know you don't have to spend what it allows - I'm interested in knowing the backtesting results if I cap the spending.
<snip>

We can see that it makes essentially no difference, outside of a few years in the 1880s, which suggests that our mental model of how retirement & portfolios work isn't quite right and needs to be updated.

<snip>
That's interesting but hard to believe.
It shouldn't be hard to believe, since I provided multiple charts showing how the theory actually worked in practice.
You're saying that even though VPW says I can withdraw $130,000, if I limit myself to never going above $84,000, it won't make a difference? That's really counter-intuitive.
I don't know what counts as "make a difference". We were solely talking about the minimum withdraw that occurs.
When selecting VPW as the spending model under cfiresim, the portfolio lasts longer the lower you make the upper bound (and lower bound) for spending. So your post would seem to undermine the results at cfiresim.
No, my post doesn't undermine the results at cfiresim. My post was about the minimum withdraw made. You are now talking about how long a portfolio lasts. They are different things.

I don't see a way to make cfiresim tell me how long a VPW portfolio will last. Can you provide a description of how you make it do that?

In any case, it isn't a mystery. VPW, by design, will draw a portfolio down to $0 by the end of 35 years. If you cap it, then of course VPW can't spend fast enough to reach that goal. Eventually you will reach a point where VPW will spend the portfolio down to $0 but Capped-VPW will still have money left, so it can run for another year (or two or ten).

For instance, take a look at a 1980 retiree (this time with a cap of $80,000 real dollar):

Image

You can see that we hit the cap pretty quickly -- around year 10 or so -- and remain capped for the rest of our retirement. Which means, just the simple mathematics of what VPW is -- that VPW will spend the portfolio down to $0 and the Capped-VPW portfolio will still have a value of $1.7 million.

Image

With a cap of $80,000 it could take another 20+ years to spend that down. So one "ran out" after 35 years while the other "ran out" after 55+ years.

But it isn't really a meaningful comparison because VPW did exactly what you told it to do. If you didn't want to run out after 35 years, you shouldn't have told it that.

In reality, the difference doesn't matter because:

1) The whole "VPW spends down to exactly $0" is artificial. In reality, someone at age 85 or 90 is going to start annuitising.
2) In reality, most people are going to find it very hard to practically spend anywhere near what VPW says they can. If you retire at age 65 spending $48,000 a year, you almost certainly aren't going to spend $130,000 a year when you are 90 even if that's what VPW says you can.

So it is clear that the only thing a cap does it extend the life of a portfolio beyond 35 years in simulations. But if you want to do that, then just tell VPW to use 40 or 45 years instead of 35. Or annuitise when you turn 90.

AlohaJoe
Posts: 4425
Joined: Mon Nov 26, 2007 2:00 pm
Location: Saigon, Vietnam

Re: 50x minimum expenses to use VPW?

Post by AlohaJoe » Fri Apr 26, 2019 1:17 am

willthrill81 wrote:
Thu Apr 25, 2019 11:24 pm
It's mathematically impossible for VPW to prematurely deplete the portfolio, so I don't know what you mean by the "portfolio lasting longer" by introducing a withdrawal ceiling.
See my previous post but the short version:

VPW is guaranteed to spend the portfolio after 35 years. That's mathematically what it is designed to do.

If you have a cap, then VPW can't spend fast enough and after 35 years the portfolio will usually be (much) greater than $0. That means it will last 36, 37, or more years before reaching $0.

I don't think it is a useful comparison, though.

User avatar
willthrill81
Posts: 9696
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: 50x minimum expenses to use VPW?

Post by willthrill81 » Fri Apr 26, 2019 9:45 am

AlohaJoe wrote:
Fri Apr 26, 2019 1:17 am
willthrill81 wrote:
Thu Apr 25, 2019 11:24 pm
It's mathematically impossible for VPW to prematurely deplete the portfolio, so I don't know what you mean by the "portfolio lasting longer" by introducing a withdrawal ceiling.
See my previous post but the short version:

VPW is guaranteed to spend the portfolio after 35 years. That's mathematically what it is designed to do.

If you have a cap, then VPW can't spend fast enough and after 35 years the portfolio will usually be (much) greater than $0. That means it will last 36, 37, or more years before reaching $0.

I don't think it is a useful comparison, though.
Good point.
“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

Topic Author
mortal
Posts: 452
Joined: Mon Sep 15, 2008 11:15 pm

Re: 50x minimum expenses to use VPW?

Post by mortal » Fri Apr 26, 2019 6:51 pm

I am curious as to what you guys think about using a low (3%) normal SWR to provide for basic expenses and VPW for the remainder?

Here, I tested a 1M portfolio, 600K of which operates under a 3% SWR, yielding an inflation adjusted 18k / yr and 400k operates under VPW. You can see the combination of the two in the 'total income' line.

Image

One fundamental assumption around this, is that a 3% SWR is safe. As in, if this fails, odds are good you're living in a world where you have bigger concerns than the performance of your portfolio, think WWIII, skynet, who knows :shock:

On the plus side, your blind faith in a 3% SWR for the 'basic expenses' part of your portfolio means you're not holding a ton of bonds hoping to tame the volatility of VPW. Everything the VPW portion grants you is gravy. Do you guys think anything is fundamentally unsound here? This might become my new mental model for retirement planning.

pastabatman
Posts: 12
Joined: Sun Oct 22, 2017 1:15 pm

Re: 50x minimum expenses to use VPW?

Post by pastabatman » Sun Apr 28, 2019 2:40 pm

mortal wrote:
Fri Apr 26, 2019 6:51 pm
I am curious as to what you guys think about using a low (3%) normal SWR to provide for basic expenses and VPW for the remainder?
So, you’re trying to ‘self annuitize’ with the 3% SWR portfolio to substitute for the usual suggestion of pairing VPW with a real annuity or pension. Annuities & portfolios behave in fundamentally different ways, so you would not really be satisfying the annuity role. Also, VPW addresses SWR’s flaws in a simple and elegant way. A hybrid VPW/SWR plan increases complexity and re-introduces SWR’s flaws.

If one does not have a pension or SS and does not want to annuitize, then I think the best way to use VPW is to simply be prepared to adjust one’s total spending down to historical 2.x % lows as dictated by the formula. You’ll get a secure base similar to what you seek from your 3% SWR bucket while allowing more potential for higher spending in good years.

User avatar
NotYourAverageJones
Posts: 68
Joined: Thu Nov 01, 2018 5:12 am
Location: An Upscale Sandbox, in a Hot as Hell Playground!

Re: 50x minimum expenses to use VPW?

Post by NotYourAverageJones » Mon Apr 29, 2019 2:51 am

willthrill81 wrote:
Thu Apr 25, 2019 11:17 pm
Another approach that sounds potentially right up the OP's alley is a new calculator found at Portfolio Charts titled "Retirement Spending." Using data going back to 1970, it allows users to backtest the results of a strategy where they start withdrawing X% of a portfolio each year and then potentially adjust that percentage up or down based on portfolio performance. Importantly for the OP, it also allows the user to specify a spending floor and ceiling in relation to the starting percentage (e.g. starting 5% of portfolio each year, reduced by a maximum of 5% each year that the portfolio is down to a minimum of 3% of the portfolio each year). Account triggers can also be modeled (e.g. don't make any reductions in the percentage withdrawn unless the portfolio drops by at least 25% of its starting inflation-adjusted value).
Thank you for the above portfolio chart link. I am finding it very useful! :beer

AlohaJoe
Posts: 4425
Joined: Mon Nov 26, 2007 2:00 pm
Location: Saigon, Vietnam

Re: 50x minimum expenses to use VPW?

Post by AlohaJoe » Mon Apr 29, 2019 5:33 am

mortal wrote:
Fri Apr 26, 2019 6:51 pm
I am curious as to what you guys think about using a low (3%) normal SWR to provide for basic expenses and VPW for the remainder?
I think it is unnecessarily complicated and doesn't seem to offer much of a benefit

I say it is complicated because if you want a spending floor then just have a spending floor. Say "I don't care what VPW says, I'm never going to withdraw less than $29,500 a year because that's what I need for my basic expenses". Don't do some complicated thing where you have one spending rule for part of your portfolio and another spending rule for another part.

I also don't think it offers a large benefit. Let's take a look at one of the scenarios you are so worried about: 1906

Image

Just by eye-balling, it doesn't look like a particularly impressive improvement but let's check the details.

For the early part of your retirement you withdraw a lot less than VPW would suggest. Over the first 12 years (from when you are 65 to 77) you withdraw a total of $85,000 less. Then VPW drops quite low and your strategy:
  • Withdraw $30,000 instead of $27,000
  • $30,000 instead of $26,000
  • $29,000 instead of $24,000
  • $28,000 instead of $21,000
  • $30,000 instead of $26,000
So there's a 5 year sequence where your strategy ends up withdrawing $3,000 to $7,000 a year more than plain VPW. But let's take a step back and look at that in the context of your entire portfolio.

Withdrawing $28,000 at that point means withdrawing 8.4% of your entire portfolio. In the past 5 years your portfolio has gone from: $773,000 to $332,000. That's why the VPW withdrawal is so low.

Do you think it is realistic that someone who is 77 years old and has seen their portfolio get crushed by 58% in the past 5 years is going to be totally okay withdrawing over 8% in a single go?

How's your blind faith in 3% feeling right about then? Strong enough that you'd ignore VPW and withdraw that extra $7,000? Constant dollar spending strategies are totally implausible in the real world when things go bad. No one is going to stick to them because, in the real world, they don't know that things will bounce back soon enough or fast enough to save their portfolio.

I think the alleged benefit is entirely theoretical and a real retiree would abandon your 3% rule when things really go south.

User avatar
willthrill81
Posts: 9696
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: 50x minimum expenses to use VPW?

Post by willthrill81 » Mon Apr 29, 2019 9:21 am

NotYourAverageJones wrote:
Mon Apr 29, 2019 2:51 am
willthrill81 wrote:
Thu Apr 25, 2019 11:17 pm
Another approach that sounds potentially right up the OP's alley is a new calculator found at Portfolio Charts titled "Retirement Spending." Using data going back to 1970, it allows users to backtest the results of a strategy where they start withdrawing X% of a portfolio each year and then potentially adjust that percentage up or down based on portfolio performance. Importantly for the OP, it also allows the user to specify a spending floor and ceiling in relation to the starting percentage (e.g. starting 5% of portfolio each year, reduced by a maximum of 5% each year that the portfolio is down to a minimum of 3% of the portfolio each year). Account triggers can also be modeled (e.g. don't make any reductions in the percentage withdrawn unless the portfolio drops by at least 25% of its starting inflation-adjusted value).
Thank you for the above portfolio chart link. I am finding it very useful! :beer
No problem, but thank Tyler9000 for putting it together. He's done a lot of yeoman's work on that site.
“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
billthecat
Posts: 359
Joined: Tue Jan 24, 2017 2:50 pm

Re: 50x minimum expenses to use VPW?

Post by billthecat » Sat May 04, 2019 12:17 pm

Phineas J. Whoopee wrote:
Thu Apr 25, 2019 6:43 pm
billthecat wrote:
Thu Apr 25, 2019 6:37 pm
stlutz wrote:
Wed Apr 24, 2019 10:03 pm
No. VPW is focused on consumption that is variable. You should have a base of other income as an underlying foundation. This includes things like Social Security, a pension, an annuity or a TIPS ladder. The allows you to just get by if you don't have other income.

Anything above that level would be covered by the variable portfolio.

VPW guarantees that you won't run out of money early; it doesn't guarantee any particular level of spending. And that is the reality of relying assets that can significantly change in price to fund consumption. For me, that's what I like about the VPW approach--it is relentlessly reality-based. What you can consume now is based on how much money you have now, and not on how much you had last year or hope to have next year.
If you're completely reliant on your portfolio - at least until SS kicks in - how would you isolate VPW to your variable consumption?
I'm not stlutz, but one would divide their portfolio into two pieces, mentally or otherwise. The first would have characteristics of a liability-matching portfolio. The other would be managed using VPW.

Regardless of how one thinks about one's portfolio or designs one's spreadsheet, one owns the totality of what one owns. There's no escaping that basic truth.

PJW
Suppose your portfolio is 60/40 stock/fixed income. Now you split off a chunk of cash, put it in CDs to fund a base of other income as an underlying foundation. Do you rebalance the remaining portion back to 60/40 or would you include the CD base in your asset allocation?
We cannot direct the winds but we can adjust our sails.

User avatar
Phineas J. Whoopee
Posts: 8255
Joined: Sun Dec 18, 2011 6:18 pm

Re: 50x minimum expenses to use VPW?

Post by Phineas J. Whoopee » Sat May 04, 2019 2:06 pm

billthecat wrote:
Sat May 04, 2019 12:17 pm
Phineas J. Whoopee wrote:
Thu Apr 25, 2019 6:43 pm
billthecat wrote:
Thu Apr 25, 2019 6:37 pm
stlutz wrote:
Wed Apr 24, 2019 10:03 pm
No. VPW is focused on consumption that is variable. You should have a base of other income as an underlying foundation. This includes things like Social Security, a pension, an annuity or a TIPS ladder. The allows you to just get by if you don't have other income.

Anything above that level would be covered by the variable portfolio.

VPW guarantees that you won't run out of money early; it doesn't guarantee any particular level of spending. And that is the reality of relying assets that can significantly change in price to fund consumption. For me, that's what I like about the VPW approach--it is relentlessly reality-based. What you can consume now is based on how much money you have now, and not on how much you had last year or hope to have next year.
If you're completely reliant on your portfolio - at least until SS kicks in - how would you isolate VPW to your variable consumption?
I'm not stlutz, but one would divide their portfolio into two pieces, mentally or otherwise. The first would have characteristics of a liability-matching portfolio. The other would be managed using VPW.

Regardless of how one thinks about one's portfolio or designs one's spreadsheet, one owns the totality of what one owns. There's no escaping that basic truth.

PJW
Suppose your portfolio is 60/40 stock/fixed income. Now you split off a chunk of cash, put it in CDs to fund a base of other income as an underlying foundation. Do you rebalance the remaining portion back to 60/40 or would you include the CD base in your asset allocation?
If I may repeat myself I'll repeat myself: Regardless of how one thinks about one's portfolio or designs one's spreadsheet, one owns the totality of what one owns. There's no escaping that basic truth.

In terms of how you think about it, or even account for it, you would take a portion of the assets to create a liability-matching portfolio, then allocate the rest as makes sense under the changed circumstances, in the present context using VPW to exhaust it after 35 years.

There is no asset allocation formula, even setting aside mental accounting, so I can't provide one.

PJW

pastabatman
Posts: 12
Joined: Sun Oct 22, 2017 1:15 pm

Re: 50x minimum expenses to use VPW?

Post by pastabatman » Thu May 09, 2019 3:39 pm

billthecat wrote:
Sat May 04, 2019 12:17 pm
Suppose your portfolio is 60/40 stock/fixed income. Now you split off a chunk of cash, put it in CDs to fund a base of other income as an underlying foundation. Do you rebalance the remaining portion back to 60/40 or would you include the CD base in your asset allocation?
The 'CD/TIPS ladder/annuity/etc.' would be your LMP, and the remainder would be balanced according to your risk tolerance between equities and bonds. For example, I plan to start VPW at 60, and SS at 70, so will have a 10 year gap without SS. So I will set aside 10 years worth of expenses into some kind of bond ladder, which will be my LMP. The remainder will be my portfolio to last my entire retirement, and probably set to 60/40 stocks/bonds.

Now, if one needs LMP to last 30+ years, I'm not sure what LMP vehicle would make sense. In that case, as I mentioned previously, I would not do an LMP, but instead plan to be able to live off only 2.x % VPW draw (historical low) in some years.

User avatar
billthecat
Posts: 359
Joined: Tue Jan 24, 2017 2:50 pm

Re: 50x minimum expenses to use VPW?

Post by billthecat » Wed May 15, 2019 11:33 pm

AlohaJoe wrote:
Fri Apr 26, 2019 1:14 am
I don't see a way to make cfiresim tell me how long a VPW portfolio will last. Can you provide a description of how you make it do that?
Without bounds, cfiresim shows an ending balance of $22-40K (average of $32k) - i.e. zero effectively. My observation was that when I include an upper bound on spending I could very comfortably live within, cfiresim shows that I reach the target age (95) with $2-37M (average of $9M). That's what I meant by lasts longer.
We cannot direct the winds but we can adjust our sails.

Post Reply