Variable Percentage Withdrawal (VPW)

[Dodge]

Just came across an incredibly interesting thread on Bogleheads on Variable Percentage Withdrawal (VPW):

http://www.bogleheads.org/forum/viewtopic.php?f=10&t=120430&p=1761580#p1761563

I'll try my best to summarize the idea.  Let's assume you have a yearly spending rate of $40,000, here are some possible withdraw strategies:

1.  The standard 4% rule, also referred to as "Constant-Dollar".  So you amass a $1,000,000 portfolio, and withdraw the $40,000 plus inflation each year.  You are now Financially Independent, and can retire.  The risk of this advice, is that you're subject to a bad sequence of returns, in which case your portfolio can drop to 0.  If the bad sequence of returns doesn't show up, it is likely you will leave a large amount of money unspent.  And by the time you realize you have too big a pile of money, you might be too old to spend it.

2.  Constant-Percentage, where you withdraw 4% of your portfolio each year, even if that number ends up being less than your $40,000 yearly spending.  If you retire on a $1,000,000 portfolio, you'd have to be prepared to take up a side job if the portfolio doesn't provide the required $40,000 plus inflation.  This one does not have as much sequence of return risk.  The problem is that I am, again, quite likely to underspend and leave a lot of money unspent.  If I choose a higher percentage, it might be too high, leading to fast depletion, or too low, leading to lots of unspent money.

3.  Spend only the dividends.  At today's rates you'd need to save up significantly more money, or invest in significantly riskier assets, in order to live on just dividends.  To quote Senior Mustachian Sol, when people recommend this, "It's almost like you're a force of evil in the world, trying your hardest to keep people from finding happiness by suffering extra decades in soul-crushing jobs.  Stop it."

4.  The Variable Percentage Withdraw (VPW).  While the methods above try to ensure your portfolio lasts forever, VPW aims to deplete the portfolio during a specified time period (by default, when you turn 100 years old).  The OP of the Bogleheads thread (Longinvest) says this about other withdraw methods, "by withdrawing less in the earlier part of retirement, you are deferring your consumption from when you have the health and ability to enjoy it with your loved ones, only to save it for later days when you will be very old and have little ability to make good use of it."

VPW will always withdraw a higher (absolute) amount money than Constant-Percentage, with the important difference that VPW will deplete your portfolio while Constant-percentage won't.  VPW adapts to the effective returns (gains or losses) of your portfolio. It will not fail if a bad sequence of returns happens at the start of your withdrawal phase.  It will also increase your withdrawals if your portfolio gives you returns above what was expected.  Similar to Constant-Percentage, I need to be prepared to get a side-job if the portfolio doesn't provide the required $40,000 plus inflation.

I think the VPW fits perfectly for Mustacians who want to retire early, and have hobbies/activities which will likely be bringing in a bit of money anyway.  Especially since Mustacians are much less susceptible to inflation than the average person, so nominal returns go much further with us.  The Longinvest discusses this here.

To highlight this, I downloaded the calculator, and put in the numbers for a 35 year old who has $40,000 yearly spending, retiring from the 9-5 with $600,000 saved up:



Here's what the withdrawals look like using data starting in 1945:




You will definitely be relying on the side job for the first few years, but after a few years it probably won't be necessary.  Over the 65 years in the example above, you will have withdrawn a total of $21.7 million, all on the $600,000 initial portfolio size.  This also assumes you won't bring in any more money than required from the side jobs, which seems unlikely.

Longinvest gives a fantastic argument for VPW here, and the FAQ is available here for more details, but really that's the main idea.

I'm seriously considering using VPW with my portfolio, before the standard 4% rule would indicate I'm ready.  Is this riskier than I'm making it out to be?  What do you think?

[Ynari]

http://www.cfiresim.com/input.php has a VPW calculator.

According to a few runs of cfiresim, VPW is very similar to constant percentage, though could be better if your horizon is predictably less than 65 years.

For a 65 year retirement, comparing VPW to constant-percentage of 4.7%, CP results in you spending 2% less per year, on average, and ending with usually around 1/3 of the real value of your starting portfolio left over as opposed to 0%. So for a long retirement, CP hedges in case you can't predict the exact year of your death... For much shorter horizons, VPW tended to be better.

But the standard 4% rule and either of the variable-income options produce very different results that matter in a psychological sense. The 4% rule gives people a steady, mostly predictable income. VPW gives you the prospect of very high income in return for (unpredictable) years of low income. These are just two radically different approaches to retirement. I don't think there's anything wrong with either choice, but I think a percentage based strategy only *really* works if you have extremely flexible annual budget or easily used alternative revenue streams.  For instance, I would only be comfortable with the VPW scenario you gave if my minimum budget was under 20k, not 40k. Otherwise you are definitely going to be working part-time for about a decade, which to me just means this is a way to turn FU money into FI money after quitting a full time job.

Personally, VPW or constant percentage would feel a little too much like playing the lottery. "This year my income is *spins a wheel* ---!"  Both the 4% rule and VPW could leave me with tons of money to spend when I'm 90 (in your chart, at 90 you've got a 120k income!), and they both have a chance of "failure" if failure means not being able to cover my basic expenses alternative revenue streams. But the 4% rule is easier to plan my life around.

[Dodge]

http://www.cfiresim.com/input.php has a VPW calculator.

According to a few runs of cfiresim, VPW is very similar to constant percentage, though could be better if your horizon is predictably less than 65 years.

For a 65 year retirement, comparing VPW to constant-percentage of 4.7%, CP results in you spending 2% less per year, on average, and ending with usually around 1/3 of the real value of your starting portfolio left over as opposed to 0%. So for a long retirement, CP hedges in case you can't predict the exact year of your death... For much shorter horizons, VPW tended to be better.

But the standard 4% rule and either of the variable-income options produce very different results that matter in a psychological sense. The 4% rule gives people a steady, mostly predictable income. VPW gives you the prospect of very high income in return for (unpredictable) years of low income. These are just two radically different approaches to retirement. I don't think there's anything wrong with either choice, but I think a percentage based strategy only *really* works if you have extremely flexible annual budget or easily used alternative revenue streams.  For instance, I would only be comfortable with the VPW scenario you gave if my minimum budget was under 20k, not 40k. Otherwise you are definitely going to be working part-time for about a decade, which to me just means this is a way to turn FU money into FI money after quitting a full time job.

Personally, VPW or constant percentage would feel a little too much like playing the lottery. "This year my income is *spins a wheel* ---!"  Both the 4% rule and VPW could leave me with tons of money to spend when I'm 90 (in your chart, at 90 you've got a 120k income!), and they both have a chance of "failure" if failure means not being able to cover my basic expenses alternative revenue streams. But the 4% rule is easier to plan my life around.

Thanks for your post, it definitely got me thinking about this.  After playing with cfiresim a bit, I came to an interesting (to me anyway!) idea.  If I set it to the Original VPW method, set the spending floor to 3% of the portfolio, and the ceiling to 4% of the portfolio, it completely prevents portfolio failure.  I picked a starting year, 1937, which fails (you run out of money) with the normal 4% rule on a 60 year timeline, and tried it with these new settings:



Then I tried again with a 30 year timeline, on the worst failure I could find when using the standard 4%:



In both of these scenarios, you'll need a part-time job, or a side-gig, earning at least $10,000 a year ($833 a month) during many years of your retirement, maybe a decade or more.  But the standard 4% withdrawal method shows this was quite necessary.  Had you not done this, your portfolio would have failed.  This might be a good withdraw strategy to follow if you're worried about running out of money early.

[Dodge]

In both of these scenarios, you'll need a part-time job, or a side-gig, earning at least $10,000 a year ($833 a month) during many years of your retirement, maybe a decade or more.  But the standard 4% withdrawal method shows this was quite necessary.  Had you not done this, your portfolio would have failed.  This might be a good withdraw strategy to follow if you're worried about running out of money early.

Or continue accumulation before retirement to permit a 3% withdrawal rate to support your lifestyle. For me, that's hopefully two extra years of working before retirement.

But if you do it that way, you're guaranteeing extra years of working, for something that has about a 10% chance of occurring.  If you instead consider yourself FI when you're at the standard 4%, and have a good way of detecting if you're on the path of portfolio failure, isn't that preferred?  That way you're only obligated to work more if you have to.  $833 a month is like 1 or 2 photography jobs (a few hours commitment) for me, and chances are very high that I'd never have to worry about it.

Honestly, as MMM says, we will all very likely be bringing in money anyway, as our hobbies and skills grow.  MMM says he can't help but bring in money, because he enjoys the "work"!  He touches on this here:

It's all about the safety margin
MrMoneyMustache vs the internet retirement police

[BuildingFrugalHabits]

It's interesting comparing the absolute number with the inflation adjusted.  750,800 is equivalent to just $63,570 ??  On another note, at least having a paid off house makes for a nice inflation hedge. 

[kyith]

the very problem with variable percentage is the loss of purchasing power, and the very problem tends to be solved by financial independent folks because most of them can work when purchasing power is lost.


Zolt's TPA rule is rather interesting and more applicable i felt for international folks where the 4% rule is rather risky

http://www.targetpercentage.com/wp-content/uploads/2013/06/Zolt-JFP-Article-Jan-2013-Higher-Safe-WR1.pdf

[hodedofome]

In retirement I'll probably vary my withdrawal rate between 3-5%, depending on how the portfolio is doing. Good years take out 4%, great years take out 5%, bad years take out 3% and always consider or have side gigs going.

[kyith]

In retirement I'll probably vary my withdrawal rate between 3-5%, depending on how the portfolio is doing. Good years take out 4%, great years take out 5%, bad years take out 3% and always consider or have side gigs going.

the standard rate is as if doing execution based on a starting premise hoping things will be ok. its rather blind. the variation like what you did is like assessing what is left and the situation current and making sensible adjustment.

an even stronger is the dynamic spending talked about by financial planner Dirk cotton here > http://theretirementcafe.blogspot.sg/2015/02/dominated-strategies-and-dynamic.html

A Dynamic Spending strategy will recalculate a sustainable withdrawal rate annually by considering updated portfolio balance, an updated life expectancy, changes in risk tolerance over time, changes in expected future returns and changes in spending.

Whether the retiree's portfolio balance trends downward or upward, Dynamic Spending will provide a better payoff than either SWR strategy because it considers remaining life expectancy. As remaining life expectancy declines throughout retirement, risk of ruin is reduced and the sustainable withdrawal rate increases. (The sustainable withdrawal amount will decrease if portfolio losses exceed the benefit of the life expectancy decrease.)

Spending gains due to decreasing life expectancy increase exponentially. Even if portfolio value remained flat throughout retirement, decreasing life expectancy would more than double spending by the end of a long retirement (see chart). SWR-V and SWR-F ignore this increase.

[forummm]

Here's what the withdrawals look like using data starting in 1945:

I wouldn't start with that year. You're entering a historically great period of market increases. The cFIREsim scenarios will get you more years of testing.

Personally, I am planning to use variable withdrawal, but I think about it a little differently. I have essentially 3 sets of money.

1) Base amount I'd be OK living on every year x 25. This is the bare bones--where if it's not in here, I'm OK not getting it for 5 years, but it includes everything that I'm willing to work longer to save up and pay for

2) A specific dollar amount saved for certain expenses in the future, such as the kids college, supplementing social security, etc. These dollar amounts will continue to grow, but I don't count on them to live off of for a really long time. The college fund portion would get invested more conservatively as the kids got older. Same with the SS supplement as we got close to 67 (if we even need it by then).

3) A variable withdrawal fund. This chunk of money would allow us to do optional things (like big international travel) in times when the market has been great, but to completely stop spending from when the market is bad. I could not find any calculator available to help me plan for this fund. The cFIREsim guy said it would be too complicated to add to his. But I created an Excel sheet to model all the possible 30-year scenarios using historical S&P 500 returns. Then I used cFIREsim to play with some of the results I had. I don't have the file with me now, but I think it was something like if I withdrew only in years where the market went up (which is typically most years), I could average a withdrawal rate of over 6-7% per year (IIRC), with a pretty high rate of still having money left at the end (it's all optional spending, so I don't need a 95% success rate). So the spending might look like 0%, 12%, 12%, 0%, 0%, 20%, etc, but averaging out to 6-7. I haven't found an equation to use for this. But I'm highly numerical, so I don't mind having to make a decision each year about how much I feel comfortable spending based on recent market performance.

[johnny847]

I think this is all worth considering if you don't have any kids, or if you have kids you and don't have any plans on leaving them an inheritance.

However, the entire point of VPW is to completely deplete your portfolio by the time you die. If you want to leave an inheritance, VPW doesn't make sense unless you set the depletion date to be ahead of your actual best guess of the year of your death.
I guess VPW could make sense despite wanting to leave your kids an inheritance if you change up the calculations a little bit and want to leave your kids a fixed amount of money upon your death.

[hodedofome]

Personally, I'd rather give my kids chunks of money on a regular basis and tell them to use it to max out their 401k, than give them several hundred thousand dollars upon my death. That would teach them the value of regular savings and compound returns. When they've built up their own nest egg, they'll realize they don't need any inheritance.

[BuffaloStache]

Personally, I'd rather give my kids chunks of money on a regular basis and tell them to use it to max out their 401k, than give them several hundred thousand dollars upon my death. That would teach them the value of regular savings and compound returns. When they've built up their own nest egg, they'll realize they don't need any inheritance.

I have a friend whose parents matched her 100% of all contributions she made into her Roth IRA (up to the annual contribution limit) until she reached the age of 21. This effectively gave her an "employer match" even for her crappy high school menial jobs, and taught her the value of saving for retirement. Plus it set her up extremely well on the path towards FI. I fully plan on implementing this with my kids if I'm able.

[RecoveringCarClown]

Personally, I'd rather give my kids chunks of money on a regular basis and tell them to use it to max out their 401k, than give them several hundred thousand dollars upon my death. That would teach them the value of regular savings and compound returns. When they've built up their own nest egg, they'll realize they don't need any inheritance.

I have a friend whose parents matched her 100% of all contributions she made into her Roth IRA (up to the annual contribution limit) until she reached the age of 21. This effectively gave her an "employer match" even for her crappy high school menial jobs, and taught her the value of saving for retirement. Plus it set her up extremely well on the path towards FI. I fully plan on implementing this with my kids if I'm able.

Great idea.

[CanuckExpat]

I was reading more about the VPW method, and downloaded the spreadsheet. (Did read the Bogleheads wiki entry, but only two pages of that epic thread)
I thought I'd necromance this thread to see if anyone had any experience using this in practice.

It's an interesting idea, and I think everyone's withdrawals will be variable in practice. This provides nice structure and a nice spreadsheet based tool. I'm trying to think how much it actually gains you in practice vs fixed withdrawals, or just winging it.

One question, perhaps more general, is in years where you find you need to spend less than the model suggest you can, what do you do with the extra? Leave it unspent? Move it into cash reserves? Just spend or donate it?

[SnackDog]

I'm a huge fan of VPW and plan to use it.   We don't have kids so no issue there. Even so, we will plan to live to 100 so if we croak early our "mistake" in forecasting our longevity will mean there is money leftover (in addition to whatever property we have at the end). So, even if we had kids I don't think the plan to deplete all funds would be a big deal.

VPW works well for us because we have a lot of downside flexibility in our spending (could easily grind down to 2% withdrawal rate I reckon).  I like how VPW lets us spend more in the early years (4.7%), assuming the market cooperates, when we will be more active and interested in enjoying hedonistic pursuits.  I think our biggest problem will be overcoming frugal tendencies and loosening up a bit.

I find the MMM forum focuses a lot on figuring out WHEN to retire, but not enough to how to spend (withdrawal strategies) once retired.  Bogleheads are light years ahead in that respect and good at reminding us that 4% is a fine rule of thumb for forecasting when you will have enough to commence retirement, but not much use after that.

[DreamFIRE]

I've seen a lot posted on this forum about withdrawing, but I have my own personal spreadsheet best optimized for my situation, which includes throttling MAGI for ACA and optimizing to minimize for taxes.

You might want to consider lowering your withdrawals with today's valuations:

https://forum.mrmoneymustache.com/investor-alley/cape-and-safe-withdrawal-rates/

[SnackDog]

I've seen a lot posted on this forum about withdrawing, but I have my own personal spreadsheet best optimized for my situation, which includes throttling MAGI for ACA and optimizing to minimize for taxes.

You might want to consider lowering your withdrawals with today's valuations:

https://forum.mrmoneymustache.com/investor-alley/cape-and-safe-withdrawal-rates/

The cool thing about VPW is that it is 100% backtested and has worked for every valuation.  You must prepare to grind down your withdrawals in the worst scenarios, but as long as you are ok with that it works. The difference is huge - 4.7% in the first year vs something like 3% if you are scared of current CAPE.

I am aware of the tax issues but these will be completely different for each individual. Depends on age, minimum income needs, portfolio, real estate, irreducible income, etc. There very few useful rules of thumb when it comes to the US tax code! In my case it is more valuable to do Roth Conversions up to the top of the 24% bracket for ten years than it is to get reduced ACA costs.  I will maximize taxes in early retirement but minimize long term particularly as I will have high RMDs as it is.

[complainy_pants]

FWIW, Harry sit will be implementing the VPW in his retirement starting now.