Alternative 4% rule

Callitaday2022

Recycles dryer sheets
Joined
Nov 19, 2015
Messages
128
Location
Colorado
If you were to take 4% of assets each year, and not 4% first year and then adjust for inflation in future years, does that help or hurt chances of not running out of money eventually?
Are there studies on this?
It seems like it would help longevity of retirement dollars as you take out less in down markets, and take out more if markets are up, but your portfolio would also be significantly up.
We want to enjoy life, but need to leave legacy for kids uncertain future medical issues.
 
If you only took 4% of whatever your assets were each year you would never run out of money.

Even if you were down to your last dollar, you would only be taking out .04 cents that year.
 
You can model this concept in Firecalc under the "Spending Model" section under the category of "Percentage of Remaining portfolio". You still put in your expenses on the first page, but you effectively can create the 4% spending.
Example - input 1m portfolio and 40k on the first page and use the spending model above.
If you wish for actual dollar spending to never decrease year to year, you input 100% for the fill in box under the % of Remaining Portfolio.
 
Are there studies on this?

Not exactly what you are asking for - but it reminds me of Variable Percentage Withdrawal from Bogleheads:

https://www.bogleheads.org/wiki/Variable_percentage_withdrawal

Variable percentage withdrawal (VPW) is a withdrawal method that adapts to the retiree's retirement horizon, asset allocation, and portfolio returns during retirement. It combines the best ideas of the constant-dollar, constant-percentage, and 1/N withdrawal methods to allow the retiree to spend most of his portfolio using return-adjusted withdrawals. By adapting withdrawals to market returns, VPW will never prematurely deplete the portfolio.
 
If you were to take 4% of assets each year, and not 4% first year and then adjust for inflation in future years, does that help or hurt chances of not running out of money eventually?
Are there studies on this?
It seems like it would help longevity of retirement dollars as you take out less in down markets, and take out more if markets are up, but your portfolio would also be significantly up.
We want to enjoy life, but need to leave legacy for kids uncertain future medical issues.
You would not run out of money.

Yes there are studies - Firecalc can model this method using historical data.

On a 50% 5 year treasury/50% total stock market portfolio, under the worst case historical scenario in Firecalc (1899) you would see an real income drop of over half - like down 56%, and it would take 22 years of gradual decline to get there. It would not look nearly so bad in nominal terms but your buying power would definitely be reduced.

The 1966 case is not quite as bad despite the high inflation although it happens more quickly. It takes 16 years for real income to drop 52% before recovering.

Both those results are way better IMO than the 4% + inflation adjusted income that means constant real income through year 25, but a smaller income in year 26 because you ran out of money, and then a big fat 0 from year 27 on in the 1966 case.

So you can see that you might experience large swings in income in real terms. You need to decide how you would handle this variable income.

And if you start with a good sequence, your income will rise more quickly perhaps than inflation, but you might want to put some aside for a rainy day.

I modeled all these scenarios in Firecalc. You can choose the % remaining portfolio method under the “Spending Models” tab and then model to your hearts content.

I use the % remaining portfolio withdrawal method, so I have modeled it pretty carefully for a simple asset allocation close enough to my case, and looked at a wide range of historical scenarios. Firecalc has data all the way back to 1890 or thereabouts for my cases, so it includes several decades of far more volatile markets and booms and busts in the US compared to what we have seen since WWII.
 
Last edited:
Fixed expenses covered by pension, so this is more "blow through the dough " ...but i would like a lot of that money!
Travel, charity, upgrade RV, classic cars, etc.
Any great books I can read on retirement spending and leaving a good inheritance?
Calling attorneys today to inquire about special needs trust.
 
There are many withdrawal models that you can try. Bogelheads has endless threads on this topic.

The one I use is to take a constant percentage (for me that is 3.3%) but with a floor that is a fixed dollar amount that I determined when we retired, and is inflation adjusted. That number is based on a bare-bones budget. When we retired the floor was about 2.8%.
 
There are many withdrawal models that you can try. Bogelheads has endless threads on this topic.

The one I use is to take a constant percentage (for me that is 3.3%) but with a floor that is a fixed dollar amount that I determined when we retired, and is inflation adjusted. That number is based on a bare-bones budget. When we retired the floor was about 2.8%.

"Living Off Your Monies" by Michael McClung details about 10 different withdrawal strategies, plus what was mentioned above with Bogleheads.
 
If you were to take 4% of assets each year, and not 4% first year and then adjust for inflation in future years, does that help or hurt chances of not running out of money eventually?
Are there studies on this?
It seems like it would help longevity of retirement dollars as you take out less in down markets, and take out more if markets are up, but your portfolio would also be significantly up.
We want to enjoy life, but need to leave legacy for kids uncertain future medical issues.

You can not run out of money with this method, Mathematically speaking. Though that doesn't mean that if you have a very long string of low returns that the withdrawals wouldn't be very low. But there is no protection for that sort of thing anyway.

Take a look at VPW over on bogleheads. In its purest form, it has a monotonically increasing % withdrawal of existing assets over a specified number of years. So it also won't automatically adjust for inflation, either. But between the percentage increasing and (hopefully) a little bit of tailwind on your returns, historically this method does fine and is relatively easy to implement.
 
i-orp is also good for an alternative withdrawal strategy. It's one of the few that will let you model dying with no money.
 
You can not run out of money with this method, Mathematically speaking. Though that doesn't mean that if you have a very long string of low returns that the withdrawals wouldn't be very low. But there is no protection for that sort of thing anyway.

Take a look at VPW over on bogleheads. In its purest form, it has a monotonically increasing % withdrawal of existing assets over a specified number of years. So it also won't automatically adjust for inflation, either. But between the percentage increasing and (hopefully) a little bit of tailwind on your returns, historically this method does fine and is relatively easy to implement.
What is "very low" - depends on the withdrawal rate and asset allocation. But historically, you can expect drops of a bit over 50% in real income terms. Is cut in half very low? Maybe - but it's not like a 90% drop in income.

Classic VPW doesn't do that much better.

Even with the monotonically increasing % withdrawal rate of VPW, in the worst scenarios you will still see a real drop in income from start of ballpark 50%. I modeled max drawdowns for the 50/50 case for 1906 and 1966 starting at the 4.2% withdrawal rate as recommended by the classic VPW spreadsheet for 50/50 and 46 depletion years. I saw drawdowns (real drops in portfolio) as large as 61% and 57% from starting years 1906 and 1966. Real income didn't drop as much - it dropped 53% and 47% from the start.

Conclusion: Classic VPW could still see drop in real income in the 50% ballpark under the worst case scenarios in spite of monotonically increasing withdrawal rates.
 
Last edited:
What is "very low" - depends on the withdrawal rate and asset allocation. But historically, you can expect drops of a bit over 50% in real income terms. Is cut in half very low? Maybe - but it's not like a 90% drop in income.

Classic VPW doesn't do any better.

Even with the monotonically increasing % withdrawal rate of VPW, in the worst scenarios you will still see a hit to real income worse than 50%. I modeled max drawdowns for the 50/50 case for 1906 and 1966 starting at the 4.2% withdrawal rate as recommended by the classic VPW spreadsheet for 50/50. I saw drawdowns (real drops in portfolio) as large as 61% and 57% from starting years 1906 and 1966.

Conclusion: Classic VPW might see slightly worse drops in real income during worst case scenarios compared to a constant 4%, both due to starting at a slightly higher initial withdrawal rate and the gradually increasing withdrawal % every year.

Right, a long/bad series of real returns is going to do bad things to both your portfolio and your withdrawals regardless of your withdrawal method. Hence, as discussed elsewhere on the forum, I prefer a version of VPW that takes into account valuations. Not perfect either by any means as no withdrawal method is capable of creating money where there is none. And as always, the future will most likely not look like the past. Folks on Bogleheads argue all the time whether an early 1900's scenario matters anymore in 2018 given the regulatory differences today. Same for the depression. Like many debates there, they'll never be resolved. :LOL:

One might also wish to consider the underlying assumptions for classic VPW for both stock and bonds - they're from a Credit Suisse report for worldwide stock and bond real returns. So, for US-centric investors, many historic withdrawal trajectories end up starting off pretty low and ending up extremely high at the end of retirement because of the US's history of strong market returns over the longer term. If I were to use classic VPW, I would probably consider increasing the historic real returns in the calculation to somewhere between the long term US real returns and worldwide returns. It won't do much to the two periods you mentioned since we had stagflation combined with a bear market in 1966 for example, but based on my own backtesting, for US-centric investors more trajectories start off with higher withdrawals and more of the late-cycle large withdrawal increases are reduced by doing this. Something everybody might not want, but it suits me.

Finally, even the VPW author believes that something like VPW works best when considered as part of a holistic income plan that includes SS, a Pension (if you're lucky), a TIPs and/or Ibond ladder, SPIA, etc. For myself, I long ago stopped looking at withdrawal methods without the rest of my plan in the spreadsheet. Definitely helps with the sleep-at-night factor when I do that. :D

Cheers...
 
Right, a long/bad series of real returns is going to do bad things to both your portfolio and your withdrawals regardless of your withdrawal method. Hence, as discussed elsewhere on the forum, I prefer a version of VPW that takes into account valuations. Not perfect either by any means as no withdrawal method is capable of creating money where there is none. And as always, the future will most likely not look like the past. Folks on Bogleheads argue all the time whether an early 1900's scenario matters anymore in 2018 given the regulatory differences today. Same for the depression. Like many debates there, they'll never be resolved. :LOL:

Cheers...
FYI - I updated my prior post as the real income drops weren't quite as bad as the worst case portfolio drawdowns in the VPW case due to the increasing withdrawal rate.

But it doesn't change the basic point that these simpler % of portfolio withdrawal methods can see ballpark 50% drops in real income due to the brutal combination of market drops and inflation over long periods in the worst case scenarios. So one best go into them with eyes open.
 
Last edited:
If you were to take 4% of assets each year, and not 4% first year and then adjust for inflation in future years, does that help or hurt chances of not running out of money eventually?
Are there studies on this?
It seems like it would help longevity of retirement dollars as you take out less in down markets, and take out more if markets are up, but your portfolio would also be significantly up...

As others have pointed out, your stash will last forever, even if it gets down to the last dollar. :) You will withdraw 4c, and next year 3c, and so forth.

Well, it will not get that bad, but be prepared that your withdrawal may get down to only 60% of the the initial amount very quickly. And that's because your stash could shrink to 60% of the start value in just a few years when the market turned south.

It's very easy to see the above. Run FIRECalc with $1M, and a WR of 0%. Use a duration of 4, then 5, then 6 years. It is easily seen that even with no withdrawal at all, the $1M in the past could shrink to a bit above $600K.

The above happens if you are unlucky and have a bad sequence of returns right of the bat. But what if you are lucky, and your stash grows much more than the 4% that you withdraw? Yes, that's what happens to people who retired after the Great Recession, like yours truly.

So, let's say that despite your 4% WR of the current value, your $1M stash has grown to $2M, and you are now enjoying $80K WR instead of the $40K you started your retirement with.

The market can still shrink to 60% or down to $1.2M from the new $2M value in just a few years. If you are so accustomed to the $80K, can you scale back to $48K? If you have expanded your lifestyle to the $80K, you are going to be in trouble.
 
Last edited:
FYI - I updated my prior post as the real income drops weren't quite as bad as the worst case portfolio drawdowns in the VPW case due to the increasing withdrawal rate.

But it doesn't change the basic point of these simpler % of portfolio withdrawal methods can see ballpark 50% drops in real income due to the brutal combination of market drops and inflation over long periods. So one best go into them with eyes open.

Yup - that's the big tradeoff vs something like the 4% rule: Instead of a fixed (real) income stream that might deplete early, you have a method that can't deplete early - but you have to be able to tolerate the year-on-year income changes instead.
 
OP-Here are some links on this subject to explore. I found all of these very informative when I first read them, and I occasionally reread them just for the fun of it.

Pfau Paper comparing VWRs:

https://poseidon01.ssrn.com/deliver...5103081104008124105026091104088105116&EXT=pdf

Pfau (with others) comparing the whole spectrum of WR approaches (IMO, one of the best WR papers out there):

eDocument Viewer


And now, Kitces telling us that even “Safety First” is not always “safe”

https://www.kitces.com/blog/even-sa...stinction-is-risk-transfer-vs-risk-retention/
 
Last edited:
Here are some links on this subject to explore. I found all of these very informative when I first read them, and I occasionally reread them just for the fun of it.

Pfau Paper comparing VWRs:

https://poseidon01.ssrn.com/deliver...5103081104008124105026091104088105116&EXT=pdf

Pfau (with others) comparing the whole spectrum of WR approaches (IMO, one of the best WR papers out there):

eDocument Viewer


And now, Kitces telling us that even “Safety First” is not always “safe”

https://www.kitces.com/blog/even-sa...stinction-is-risk-transfer-vs-risk-retention/

Yep - that's a great paper. Pfau takes VPW (PMT/Actuarial) and modifies it a bit.
1. Instead of planning for a fixed number of years, he uses life expectancy and updates the remaining number of years each year.
2. Unlike classic VPW which uses long term worldwide stock and bond real returns, Kitces uses 10 year nominal treasury returns to seed the PMT calculation, which is different than what I've seen elsewhere. Makes it somewhat conservative with more upside if stocks perform, which they should. Reminds me of the ARVA paper which uses TIPs returns. Neither use either expected or historical stock returns as part of the calculation.

My favorite quote for actuarial methods. "This rule can be designed many different ways". Indeed!
 
Last edited:
I have to regularly remind DW that if we withdraw 4% our 30 year success rate is 95%. If we withdraw 5% our success rate drops to 74%.

It's a fine line.
 
I have to regularly remind DW that if we withdraw 4% our 30 year success rate is 95%. If we withdraw 5% our success rate drops to 74%.

It's a fine line.
Not that fine. A 5% withdrawal is 25% larger than a 4% withdrawal.
 
As others have pointed out, your stash will last forever, even if it gets down to the last dollar. :) You will withdraw 4c, and next year 3c, and so forth.

Well, it will not get that bad, but be prepared that your withdrawal may get down to only 60% of the the initial amount very quickly. And that's because your stash could shrink to 60% of the start value in just a few years when the market turned south.

It's very easy to see the above. Run FIRECalc with $1M, and a WR of 0%. Use a duration of 4, then 5, then 6 years. It is easily seen that even with no withdrawal at all, the $1M in the past could shrink to a bit above $600K.

The above happens if you are unlucky and have a bad sequence of returns right of the bat. But what if you are lucky, and your stash grows much more than the 4% that you withdraw? Yes, that's what happens to people who retired after the Great Recession, like yours truly.

So, let's say that despite your 4% WR of the current value, your $1M stash has grown to $2M, and you are now enjoying $80K WR instead of the $40K you started your retirement with.

The market can still shrink to 60% or down to $1.2M from the new $2M value in just a few years. If you are so accustomed to the $80K, can you scale back to $48K? If you have expanded your lifestyle to the $80K, you are going to be in trouble.

If most of one's increase involves discretionary spending, one can make it work in a more flexible way on the way down.
 
I take “the market” to be equities. And though the market may drop 50% for example, I assume most retirees aren’t 100% equities, closer to 60:40 in which case the portfolio drop should be closer to 30%. Still quite a haircut but I assume no one here is 100% exposed to “the market.” And most have some other sources of income like SS, pensions, annuities, or others so total income would drop (significantly) less than 30%. So the belt tightening should be less than market fluctuations.
 
I take “the market” to be equities. And though the market may drop 50% for example, I assume most retirees aren’t 100% equities, closer to 60:40 in which case the portfolio drop should be closer to 30%. Still quite a haircut but I assume no one here is 100% exposed to “the market.” And most have some other sources of income like SS, pensions, annuities, or others so total income would drop (significantly) less than 30%. So the belt tightening should be less than market fluctuations.

That is true in any given year. But over multiple years is what matters.

In the worst sequence of returns series, real income from a portfolio can drop by 50% or more even for the 50/50 portfolio when one is using a ~4% withdrawal method (incl. classic VPW) based on annual value of portfolio. It will take 15-20 years to get there, but does gradually recover, and for the straight 4% of remaining portfolio almost recovers the initial portfolio in the worst case sequence. VPW is designed to deplete the portfolio more aggressively without running out, as long as you give yourself a long enough time period.

The traditional 4% of initial portfolio plus inflation adjustment won’t see this slow decline in real income, but will just suddenly run out of money at year 26/27.

The point is that for someone considering a %remaining portfolio method, whether it is the straight percent, Clyatt's 95% method, or classic VPW with a long enough horizon, the worst case scenarios represent real income drops in the 50% ballpark. So you have to think about how you might manage this unlikely case. If more than half of your spending is discretionary, and you also expect your spending to decline after a decade anyway, that might be good enough.
 
Last edited:
That is true in any given year. But over multiple years is what matters.

In the worst sequence of returns series, real income can drop by 50% or more even for the 50/50 portfolio when one is using a ~4% withdrawal method (incl. classic VPW) based on annual value of portfolio. It will take 15-20 years to get there, but does gradually recover, and for the straight 4% of remaining portfolio almost recovers the initial portfolio in the worst case sequence. VPW is designed to deplete the portfolio more aggressively without running out, as long as you give yourself a long enough time period.

The traditional 4% of initial portfolio plus inflation adjustment won’t see this slow decline in real income, but will just suddenly run out of money at year 26/27.

audreyh1 is correct. A particularly insidious time to have retired is the late 1960's. You have a bear market combined with stagflation in the early 1970's. So when you look at withdrawals in real (not nominal) terms relative to the first year's withdrawal, you see a pretty significant dip that takes a long time to recover when using a variable withdrawal method like fixed % or classic VPW.

Investing isn't a complete science and there's no denying an emotional aspect to it. I have a pretty strong memory of those days. It's why I put a good bit of my bond holdings in TIPs/Ibonds. It won't completely eliminate that drop, but I'm betting that it will reduce the depth and shorten the duration if a similar event happens. There is no way to know with absolute certainty as neither TIPs nor Ibonds existed during that time. Some researchers have made attempts to recreate what TIPs returns series might have looked like back then. That, combined with how these instruments are actually supposed to act give me some reassurance, however. And like many choices in a portfolio, "might help, won't hurt much if I'm wrong" applies. :LOL:
 
Fixed expenses covered by pension, so this is more "blow through the dough " ...but i would like a lot of that money!
Travel, charity, upgrade RV, classic cars, etc.


There is no way to know for sure, what strategy will work best.

If I had my basics covered with pensions like you do, I'd try to keep the withdrawal strategy as simple as possible (like a fixed percentage) so that you can stay the course easily as you get older.
 
Back
Top Bottom