Alternative to the 4% rule

[FinancialDave]

I have written a number of articles on this subject.

This is one that compares VPW, RMD, & the 4% over a specific couple of "good sequence" and "bad sequence" periods:

https://seekingalpha.com/article/2747405-safe-withdrawal-rates-part-2-variable-withdrawals

The summary tells the story over 35 years starting in 1973 & 1975.

It convinced me enough to use the RMD method in my own retirement (this is the 3rd year I have been using it.)

By the way nothing makes you use it based on your "real" age as long as you make sure you withhold enough to satisfy the IRS. I have written articles on using a flat starting rate of say 3.5% from your starting retirement up to age 70.5 if you start early, or if you want to be more aggressive and work your IRA down faster, I personally am doing and blogging about starting at an age that was 10 years older for those wanting to start at age 60.

Dave

 

[FinancialDave]

By the way for those above testing these methods using a spreadsheet, I suggest you only do it using real market data, as opposed to just picking some constant inflation and market return data.

Dave

 

[audreyh1]

I actually created my own VPW spreadsheet and I did it for several reasons.

Choices for the stock portion of your portfolio
1. 1/CAPE (aka 1/PE10). Not perfect but it is decent. You can always find this at: Shiller PE Ratio or on Professor Shiller's website. For example, today CAPE sits at 32.45 So 1/CAPE=3.08% Note, CAPE is derived by professor Shiller for the S&P 500 index. Probably pretty close if you use a total stock market fund. But if you deviate much from those two, you may need to adjust this number upwards or downwards. Or you can just use it and if your choice of stock fund has higher returns, than that, then the dollar amount withdrawn will automatically reflect it, eventually. This is probably the easiest metric to use as it's always kept up to date.

What about bonds?
1. The same websites above sometimes also give future bond returns.
2. You can also get a good idea by going to Morningstar and pull up your bond fund. The SEC yield is a decent predictor of nominal future returns for your bond fund. For example, the 30 day SEC yield for VBTLX is currently 3.12% according to Morningstar.

But you still need to take this number and convert it to a real return. So, you need a future inflation predictor to convert the nominal predicted return to a real predicted return. Again, some of the websites above provide an answer. Some alternatives to that:
1. University of Michigan has a survey with consumer expected inflation rates. It's available here: https://fred.stlouisfed.org/series/MICH In May of 2018, the data shows 2.8%
2. You can also look at the difference in yields between nominal bonds and TIPs of the same duration. In the ideal world that difference would be the expected inflation rate. Fortunately FRED has already done the calculation for you and its available here for 10 year expected inflation https://fred.stlouisfed.org/series/T10YIE As of July 6th, it's 2.12% Or you can look at the 5 year numbers which aren't much different and can be found here: https://fred.stlouisfed.org/series/T5YIFR

Now just weight the stock and bond expected returns as your AA. For example, if you're 60/40 then 0.6* stock expected + 0.4 * bond expected and plug that into the PMT equation each year to calculate your %withdrawal.

OK - am I correct in understanding that the stock performance prediction from CAPE10, using say Jan 1 value of 33.31, thus 1/CAPE10 being 3.00% is a real prediction?

I used the 30 day yield SEC value for AGG which was a 2.75% on 3/31/18, and 2.7% for inflation from the University of Michigan graph at about the same time. Came up with 0.05% real for bonds.

For a quick and dirty look I plugged these numbers into the VPW table sheet under the VPW Parameters Long Term Real Growth Trends - setting Override to YES.

For a 50/50 portfolio (48 depletion years) the initial recommended withdrawal rate dropped dramatically - from 4.1% using the Credit Suisse numbers, to 2.9%, and did not reach 4% until year 18.

Very interesting.

Maybe I'm not under withdrawing after all?

 

[indiajust]

I like one modification suggested for use during recessions and depressions. Take a fixed percentage, but if the portfolio declines dramatically, don't go below 95% of the previous year amount.


Dramatic improvement in quality of life by smoothing cash flow, but only a very slight increase in risk of running out of money.

 

[Leo1277]

Actually one great spending strategy is the following: Look up in a table what your chance of dying in the next year is for your age; and that is the percentage of your current balance that you can withdraw. Adjust this rate every year; and if you have a medical condition, or consider yourself super healthy, you can adjust accordingly. Similarly, it is of course higher for men, so men can draw more quickly since they need to cover fewer years. If you retire early, it automatically means you can draw less. And if the market tanks, you draw the same percentage, but off a lesser total amount.

This obviously sounds morbid, but there are two very good and logical reasons for it: One, say your chance of dying is 5%, this means that you have a 95% chance of making it another year, so you need 95% of your money to cover this "eventuality". Two, it is actually very close to what the mandatory distribution percentages for tax deferred retirement accounts are. These also keep rising every year so as to hopefully nearly deplete the account during one's lifetime, but of course they only kick in in your seventies.

 

[Dtail]

I like one modification suggested for use during recessions and depressions. Take a fixed percentage, but if the portfolio declines dramatically, don't go below 95% of the previous year amount.


Dramatic improvement in quality of life by smoothing cash flow, but only a very slight increase in risk of running out of money.

+1
This is the Clyatt methodology and it is attached to the "% Remaining of Portfolio" methodology. This is modeled in Firecalc.
I plan on starting to use it next year.

 

[big-papa]

By the way for those above testing these methods using a spreadsheet, I suggest you only do it using real market data, as opposed to just picking some constant inflation and market return data.

Dave

FYI - the RMD method is essentially identical to VPW/PMT if you assume that the future expected return is 0%. This was discussed on bogleheads on the VPW thread.

 

[Dtail]

No problem! Yep, there's nothing that says you actually have to spend what you withdraw, so the sort of buffering you describe will work. Plus it also helps with those "unplanned expenses".
Cheers!

I started using this concept this year thanks to Audreyh1. Some posters here choose to put back into the portfolio any unused funds, which are then subject to the market performance again.
Money is fungible, but still rather maintain the separate bucket, that along with the 95% rule, I am hoping to never have to drastically reduce spending.

 

[big-papa]

OK - am I correct in understanding that the stock performance prediction from CAPE10, using say Jan 1 value of 33.31, thus 1/CAPE10 being 3.00% is a real prediction?

I used the 30 day yield SEC value for AGG which was a 2.75% on 3/31/18, and 2.7% for inflation from the University of Michigan graph at about the same time. Came up with 0.05% real for bonds.

For a quick and dirty look I plugged these numbers into the VPW table sheet under the VPW Parameters Long Term Real Growth Trends - setting Override to YES.

For a 50/50 portfolio (48 depletion years) the initial recommended withdrawal rate dropped dramatically - from 4.1% using the Credit Suisse numbers, to 2.9%, and did not reach 4% until year 18.

Very interesting.

Maybe I'm not under withdrawing after all?

You're close. Just remember, you have to take the geometric difference when calculating the real expected return, not the arithmetic difference. So the formula is (1+2.75%)*(1-2.70%)-1 which is -0.02%


And you're correct for 1/CAPE, essentially. It's 3.02%

And in my own spreadsheet, I get 2.89% for the first year, so close enough to what VPW is telling you.
By the way, the formula to get the % is PMT(rate, # years remaining, -1, 0, 1) where rate is the predicted real return of your portfolio. Makes it pretty easy to create your own spreadsheet.


What you're seeing lines up with most analysts expectations which are that the next decade +/- will have lower returns than the long term average. Just remember
1. The error bars of those predictions can be pretty big. If you're in accumulation mode, don't bet your retirement date on those numbers.
2. The low returns over a decade or so can take any trajectory, including basically flat for the whole time, or a huge crash followed by stellar returns. Or even 9 years of stellar returns followed by a big crash. And everything in between. That's why it's important to update the prediction annually.

 

[audreyh1]

Thanks again!

Right, I would definitely update the spreadsheet every year, as market valuations continually change and sometimes change drastically.

 

[big-papa]

Thanks again!

Right, I would definitely update the spreadsheet every year, as market valuations continually change and sometimes change drastically.

No problem. Glad to help.

By the way, you can improve your expected return on the stock side by adding some international if you choose. How much has been a debate going on at BH literally for years and will never be resolved. Very roughly, the world market is about 50% US and 50% everybody and some choose to go that way. Others go lower on international. To each his/her own.

Anyway, if you go to the research associate website I noted earlier, and on the right hand side click on "charts" under Equities you can select CAPE. Using that, you get the following for current CAPE.

US Large is 31.6 so 1/31.6 is 3.2%
EAFE is 17.3 so 1/17.3 is 5.7%

Whether those expected returns or even the difference between those expected returns results in realized returns is anybody's guess, of course. But adding some international would increase the first year's VPW withdrawal using the 1/CAPE guesstimate method for future real returns.

Cheers.

 

[audreyh1]

No problem. Glad to help.

By the way, you can improve your expected return on the stock side by adding some international if you choose. How much has been a debate going on at BH literally for years and will never be resolved. Very roughly, the world market is about 50% US and 50% everybody and some choose to go that way. Others go lower on international. To each his/her own.

Anyway, if you go to the research associate website I noted earlier, and on the right hand side click on "charts" under Equities you can select CAPE. Using that, you get the following for current CAPE.

US Large is 31.6 so 1/31.6 is 3.2%
EAFE is 17.3 so 1/17.3 is 5.7%

Whether those expected returns or even the difference between those expected returns results in realized returns is anybody's guess, of course. But adding some international would increase the first year's VPW withdrawal using the 1/CAPE guesstimate method for future real returns.

Cheers.

I do have a chunk of international in my equity allocation (maybe about 20%). I was not going to bother to model it for the purposes of looking at withdrawal rates. But maybe it's worth a look.

 

[big-papa]

I do have a chunk of international in my equity allocation (maybe about 20%). I was not going to bother to model it for the purposes of looking at withdrawal rates. But maybe it's worth a look.

Yeah, I'm about 30% myself but mainly in EM and Int'l Small cap. One thing that's worth doing in your experiments is a quick sensitivity analysis is simply to look at the change in first year's withdrawal per unit change in the expected return.

Cheers...

 

[audreyh1]

Anyway, if you go to the research associate website I noted earlier, and on the right hand side click on "charts" under Equities you can select CAPE. Using that, you get the following for current CAPE.

US Large is 31.6 so 1/31.6 is 3.2%
EAFE is 17.3 so 1/17.3 is 5.7%

Whether those expected returns or even the difference between those expected returns results in realized returns is anybody's guess, of course. But adding some international would increase the first year's VPW withdrawal using the 1/CAPE guesstimate method for future real returns.

Cheers.

I'm familiar with the EAFE Index, but it isn't a 10 year averaged price to earnings valuation metric like CAPE10 is it? I thought it was

Where does the suggestion to use 1/EAFE come from? and how does that related to expected returns?

I see that the https://interactive.researchaffilia...currency=USD&model=ER&scale=LINEAR&terms=REAL link gives 4.8% expected real returns for the EAFE as of 6/30/18.

 

[97guns]

I never understood drawdown rules, it’s never been in my fire equation, if there’s an income stream to cover expense there’s no need to touch capital

 

[Dtail]

I never understood drawdown rules, it’s never been in my fire equation, if there’s an income stream to cover expense there’s no need to touch capital

I would think more than half the folks on this site are touching their capital.

 

[big-papa]

I'm familiar with the EAFE Index, but it isn't a 10 year averaged price to earnings valuation metric like CAPE10 is it? I thought it was

Where does the suggestion to use 1/EAFE come from? and how does that related to expected returns?

I see that the https://interactive.researchaffilia...currency=USD&model=ER&scale=LINEAR&terms=REAL link gives 4.8% expected real returns for the EAFE as of 6/30/18.

It's not 1/EAFE. It's 1/CAPE for the EAFE index. Likewise it's 1/CAPE for the SP500 index. CAPE can be calculated for any index, if you have the data. Professor Shiller keeps up with CAPE on his website only for the SP500. Research associates apparently also keeps up with CAPE for EAFE as well as other indexes.

Also I originally gave the list of companies who have their own predictions. It doesn't mean that they're using 1/CAPE to make them. Most have their own proprietary models that may or may not include using CAPE in the calculation. 1/CAPE (at least for US Large blend like Sp500) is easily available at any time and is reasonably good at large time frame smoothing and keeping your withdrawals (in real terms) from shooting to the moon later in your retirement by spreading, on average, more even spending during your retirement - unlike "classic" VPW.

 

[big-papa]

I would think more than half the folks on this site are touching their capital.

I would believe that as well. But, sure, if you're fortunate enough to have significant amounts of other streams of income such as rental property or a pension, no need to touch your investments. And if dividends from your investments with or without other streams give you what you need to live on, then why not? I know over on the bogleheads forum, the idea of dividend streams comes up all the time with some pretty interesting insights. Might want to go check out some of the threads over there.


That said there are many of us who either don't have those streams available or who wish to maximize the amount available to be spent and/or have little interest in leaving an inheritance. For us, spending down our capital gives us what we need. And there are plenty of methods that put the odds in our favor of not running out of money before we run out of life.

 

[audreyh1]

It's not 1/EAFE. It's 1/CAPE for the EAFE index. Likewise it's 1/CAPE for the SP500 index. CAPE can be calculated for any index, if you have the data. Professor Shiller keeps up with CAPE on his website only for the SP500. Research associates apparently also keeps up with CAPE for EAFE as well as other indexes.

Also I originally gave the list of companies who have their own predictions. It doesn't mean that they're using 1/CAPE to make them. Most have their own proprietary models that may or may not include using CAPE in the calculation. 1/CAPE (at least for US Large blend like Sp500) is easily available at any time and is reasonably good at large time frame smoothing and keeping your withdrawals (in real terms) from shooting to the moon later in your retirement by spreading, on average, more even spending during your retirement - unlike "classic" VPW.

OK - that was my confusion. I didn't realize there was a CAPE10 calculated for the EAFE index or where to find it.

 

[audreyh1]

I never understood drawdown rules, it’s never been in my fire equation, if there’s an income stream to cover expense there’s no need to touch capital

I have no problems with touching capital. That's what we have saved and invested for all our lives and I'm not invested to generate income streams. I'm focused on total return, and I only care that we don't run out of money before we die. That's where the drawdown rules come into play.

 

[Hermit]

We don't have any particular withdrawal approach.

When I first retired 6 1/2 years ago, we were living solely off investments. I set up an automatic monthly transfer from our retirement portfolio to our local bank account that, at the time, was 3% annual WR. However, we have some lumpy expenses in November (insurances and property taxes) so I would do some occasional special transfers so our overall WR was probably more like 4%. That monthly transfer is the same today as it was in 2012.

5 years later I started my pension, which increased our monthly cash inflows by 23%. Between adding the pension and our investments doing very well we loosened the purse strings a bit and spend more freely now that we did when we first retired.

Our portfolio today is 125% of what it was when I retired, and that is after buyng a winter condo for cash, building a 2-car garage for cash, etc.

We have sufficient redundancy and our WR is now is a little higher then when we first retired because we are spending more but it is low enough that I don't fret about withdrawals at all... I just monitor our WR.

IMHO, that is a good withdrawal strategy. I would call it "spend what you want, within reason". The monitoring keeps it within reason.

 

[big-papa]

OK - that was my confusion. I didn't realize there was a CAPE10 calculated for the EAFE index or where to find it.

Where to find it was in the first part of my post post. Right hand side under charts on the research affilliates link. It'll show current and historical CAPE for several indexes.

 

[audreyh1]

Where to find it was in the first part of my post post. Right hand side under charts on the research affilliates link. It'll show current and historical CAPE for several indexes.

Found it. Thanks.

 

[audreyh1]

No problem. Glad to help.

By the way, you can improve your expected return on the stock side by adding some international if you choose. How much has been a debate going on at BH literally for years and will never be resolved. Very roughly, the world market is about 50% US and 50% everybody and some choose to go that way. Others go lower on international. To each his/her own.

Anyway, if you go to the research associate website I noted earlier, and on the right hand side click on "charts" under Equities you can select CAPE. Using that, you get the following for current CAPE.

US Large is 31.6 so 1/31.6 is 3.2%
EAFE is 17.3 so 1/17.3 is 5.7%

Whether those expected returns or even the difference between those expected returns results in realized returns is anybody's guess, of course. But adding some international would increase the first year's VPW withdrawal using the 1/CAPE guesstimate method for future real returns.

Cheers.

OK - I updated to take into account my international allocation, and initial withdrawal rate increased to 3.2%.

These are definitely more conservative withdrawal rates than with any other method I have modeled. But it is the only one that takes into account starting market valuations.

 

[big-papa]

OK - I updated to take into account my international allocation, and initial withdrawal rate increased to 3.2%.

These are definitely more conservative withdrawal rates than with any other method I have modeled. But it is the only one that takes into account starting market valuations.

Yep you've got the idea!

And this sort of thinking has led me (and others I correspond with) on some different thought processes.

1. Should your AA during retirement look the same as your AA during the accumulation years? Funds that use glidepaths would have you think so. But there are enough articles out there, as well as my own backtesting, that really make me wonder whether being more conservative is really the right answer. It certainly helps the "sleep at night" factor, especially if another 2008 scenario raises its head. But it might not be the best for the long term, especially with a variable withdrawal method.
2. Are my bond holdings only there to help me sleep at night? They certainly reduce volatility but is there a cost. Researcher Kitces has advocated a "bond tent" where the % of bond holdings rise as you approach retirement then slowly decrease as retirement progresses. But his work was mainly with SWR in mind, not a valuation-based method like this. This is because it's been shown that a bad sequence of returns in the early years of retirement can have a detrimental effect on how long the portfolio lasts for an SWR method. That's not really the case for a variable method because a bad sequence of returns early on mainly just reduces the size of the early withdrawals instead but has no bearing on how long the portfolio will last.

Anyway, my portfolio is, like yours, 50/50. But it has a very heavy dose of MCV, SCV along with some EM and Int'l small cap. So on the equity side, I have somewhat higher expected long term returns (and a higher first year withdrawal rate) but no doubt higher expected volatility which I try and balance out with 50% Intermediate Treasury bonds and 1/cape type smoothing. But there are other ways to think of this.

Early on I mentioned that in bond heavy portfolios, there is less smoothing by using expected returns than with a stock heavy portfolio. I know of one person on BH who only has 20% bonds. His 80% stock is half US and half International divided up as 20% TotalUS, 10% SCV, 10%MCV and 20% Total Int'l, 10%EM, 10%Int'l small. He is using 1/CAPE for long term smoothing but also using a shorter term smoothing method as well. In other words, he's counting on a very high stock allocation to keep the % withdrawn high and smoothing methods, and not so much bonds, to keep the volatility of his withdrawals down.

I have no idea what the absolute right answer here is. Using valuations makes intuitive sense to me and solves several problems vs. basing the withdrawals on super long term returns. Short term smoothing distorts the withdrawals and may, if you have a number of down years, reduce the size of your portfolio quicker, perhaps resulting in permanently lower withdrawals for the duration of retirement vs. not having short term smoothing. On the other hand, it is nice not to have your year on year withdrawals fluctuate too terribly much.

For now, I plan to stay the course since I'm not yet retired. But a thought I've had is that once retirement starts, I would slowly increase the stock portion of my portfolio to somewhere between 60 and 70% as I move through retirement. Got plenty of time to consider and I don't tend/need/want to make any big changes at this point...

Cheers - Big-Papa