withdrawal rates on portfolio charts

[mathjak107]

tyler creator of the website portfolio charts did an article on safe withdrawal rates . tyler is also creator of the golden butterfly portfolio.

one of the problems i see in the article is the term safe withdrawal rate is based on retiring in the worst of times which were 1907, 1939, 1937 , 1965 and 1966

all the references in the article start at 1970 which actually misses the worst of times which was 1965/1966 .

really the article in my opinion is about a withdrawal rate but not necessarily what we consider a safe withdrawal rate

https://portfoliocharts.com/2024/03/15/how-to-harness-the-flowing-nature-of-withdrawal-rate-math/

https://www.kitces.com/blog/what-returns-are-safe-withdrawal-rates-really-based-upon/

 

[DrRoy]

The math is valid, if you accept the more recent bias.

 

[mathjak107]

i wish portfolio visualizer went back earlier .

every failure has happened in the first 15 years of a 30 year retirement that we have had .

even the best bull markets couldn’t bring them back after that point and 4% failed .

it makes a big difference starting in 1966 vs 1970 as far as outcome goes.

in fact if we eliminate those failure dates , a safe withdrawal rate is actually 6-1/2% .

so a safe withdrawal rate is based on specific starting dates representing the worst years to start a retirement , and none are 1970.

last one was 1966 . if you started one year later in 1967 you were fine

 

[Exchme]

It's the nature of the beast, the world has changed.

Many of the things we take for granted today - 10year Treasuries, factor funds (large, small, value, growth, SCV, etc), tradable gold, etc. just didn't exist in the far past. The deeper we go into the past, the more those become estimated reconstructions, but maybe not something you could easily invest in.

 

[mathjak107]

rule a thumb is to have that safe withdrawal rate hold takes about a 2% real return average the first 15 years no matter what you use

Overall looking at the 30 year time frame things look actually pretty normal …this is where those who try to use other time frames get fouled up because when spending down it was the poor first 15 years that did them in


30 year outcomes , are fairly decent

1907 stocks returned 7.77% -- bonds 4.250-- rebalanced portfolio 7.02- - inflation 1.64--

1929 stocks 8.19% - - bonds 1.74%-- rebalanced portfolio 6.28-- inflation 1.69--

1937 stocks 10.12 - - bonds 2.13 - rebalanced portfolio -- 7.24 inflation-- 2.82

1966 stocks 10.23 - -bonds 7.85 -- rebalanced portfolio 9.56- - inflation 5.38

for comparison the 140 year average's were: stocks 8.39--bonds 2.85%--rebalanced portfolio 6.17% inflation 2.23%
..


so what made those time frames the worst ? what made them the worst is the fact in every single retirement time frame the outcome of that 30 year period was determined not by what happened over the 30 years but the entire outcome was decided in the first 15 years.
..


so lets look at the first 15 years in those time frames determined to be the worst we ever had.
..

1907--- stocks minus 1.47%---- bonds minus .39%-- rebalanced minus .70% ---inflation 1.64%
.

1929---stocks 1.07%---bonds 1.79%---rebalanced 2.29%--inflation 1.69%
.

1937---stocks -- 3.45%---bonds minus 3.07%-- rebalanced 1.23%--inflation 2.82%
.

1966-stocks minus .13%--bonds 1.08%--rebalanced .64%-- inflation 5.38%

it is those 15 year horrible time frames that the 4% safe withdrawal rate was born out of since you had to reduce from what could have been 6.50% as a swr down to just 4% to get through those worst of times.

 

[kevink]

tyler creator of the website portfolio charts did an article on safe withdrawal rates . tyler is also creator of the golden butterfly portfolio.

one of the problems i see in the article is the term safe withdrawal rate is based on retiring in the worst of times which were 1907, 1939, 1937 , 1965 and 1966

all the references in the article start at 1970 which actually misses the worst of times which was 1965/1966 .

really the article in my opinion is about a withdrawal rate but not necessarily what we consider a safe withdrawal rate

https://portfoliocharts.com/2024/03/15/how-to-harness-the-flowing-nature-of-withdrawal-rate-math/

https://www.kitces.com/blog/what-returns-are-safe-withdrawal-rates-really-based-upon/

Tyler answers the "why no data before 1970) in the FAQ on withdrawal rates on his site:

Why does your data only start in 1970? Isn’t that cherry picking?

Like other retirement studies, the timeframe covered is simply a matter of data availability.

The Trinity and Bengen studies used data going back to 1926 not because that was a magical date for safe withdrawal rates but because that’s what was readily available from Ibbotson Associates for the indices they studied. Wade Pfau and Michael Kitces used data since the early 1870’s because that’s what was freely available from Robert Shiller. And Pfau later did the same analysis for developed international markets since 1900 because that’s what was available from the DMS dataset. All three sources have limited information on diverse asset classes beyond large cap stocks, bonds, and T-bills.

Good data for different markets is very difficult to find. If you know where to find reliable older data, please contact me.

Also, note that when this calculator was first released it only had data to 1972. I have been able to extend it to 1970 based on new information, and will continue to add more years as they become available.

 

[mathjak107]

but the fact is the term safe withdrawal rate is based on those time frames pre 1966 …starting in 1970 is like starting in 2000 or 2008 .

those are times that are not the worst outcomes and were really not failure periods at all.

so while they may have been tougher times there is nothing unique about starting in 1970 ,2000 or 2008 as they were not failure points nor would one have had to reduce from 4% to get thru them and that is the whole point .

firecalc uses the shiller data set which goes back to 1871

 

[audreyh1]

i wish portfolio visualizer went back earlier .

every failure has happened in the first 15 years of a 30 year retirement that we have had .

even the best bull markets couldn’t bring them back after that point and 4% failed .

it makes a big difference starting in 1966 vs 1970 as far as outcome goes.

in fact if we eliminate those failure dates , a safe withdrawal rate is actually 6-1/2% .

so a safe withdrawal rate is based on specific starting dates representing the worst years to start a retirement , and none are 1970.

last one was 1966 . if you started one year later in 1967 you were fine

The worst cases I have seen in my FIREcalc models have been before 1926 and most predating The Federal Reserve. The 1966 one came in 12th worst for my particular scenario. FIREcalc data goes back to 1889 or something like that.

However it’s fine with me having data from some really rough periods. It’s a good test of the robustness of my plans as far as I’m concerned.

I’m not using the Bengen/Trinity withdrawal method so many of the details don’t apply.

 

[mathjak107]

according to kitces if we start after 1966 and eliminate those worst outcomes then a safe withdrawal rate would be 6-1/2% not 4% .

but then it wouldn’t be a safe withdrawal rate by the standards set up by safe max and also the trinity study , since it wouldn’t be based on the worst of times .

safe max didn’t include 1965/1966 yet since the 30 year period didn’t complete when bill did his safe max work .

the trinity included 1965/1966 .

firecalc goes back even farther to 1871 so it includes 1907 .

worst results on record according to kitces were 1965/1966

so starting in 1970 eliminates all the time frames the safe withdrawal rate is based on , regardless of tyler’s thinking

i personally have had this discussion with tyler myself

https://www.kitces.com/blog/what-returns-are-safe-withdrawal-rates-really-based-upon/

 

[big-papa]

The SIMBA spreadsheet available on Bogleheads has data that goes back significantly farther for many of the asset types people care about. And it will also calculate an SWR over rolling periods.

The latest version is here: https://www.bogleheads.org/forum/viewtopic.php?p=7726003#p7726003

I personally don't like anything that resembles a fixed withdrawal rate, adjusted for inflation like SWR is. It does make the math a lot easier if all you want to measure are the odds (in the past) of a complete catastrophe.

I'm using amortization for my withdrawal calculations. There are plenty of examples around including VPW, ABW and TPAW, all discussed and available over on bogleheads. Also known as the "actuarial method" by some and covered by Karsten on Early Retirement Now at https://earlyretirementnow.com/2019...-rate-without-simulations-swr-series-part-33/

Cheers
Big-Papa

 

[CarbonaNotGlue]

Thanks for sharing the article, but the comments concerning recency bias are definitely valid. As others have noted, Big ERN over at his site has done far more extensive analysis over a larger time frame and with differing portfolio compositions. Based on his work, I’m planning on keeping my withdrawal rate below 3.25%. Here is a nice chart summarizing his extensive work -

https://i0.wp.com/earlyretirementnow.com/wp-content/uploads/2016/12/swr-part2-table1.png

 

[mathjak107]

Thanks for sharing the article, but the comments concerning recency bias are definitely valid. As others have noted, Big ERN over at his site has done far more extensive analysis over a larger time frame and with differing portfolio compositions. Based on his work, I’m planning on keeping my withdrawal rate below 3.25%. Here is a nice chart summarizing his extensive work -

https://i0.wp.com/earlyretirementnow.com/wp-content/uploads/2016/12/swr-part2-table1.png

here is yet another version with a little different outcome .

but all are based on the same worst outcomes pre 1970

most notable is at 30 years and 4% , 100% equities did a bit worse then 50/50 as far as success rate compared to the one above ,but certainly close enough

 

[mathjak107]

The worst cases I have seen in my FIREcalc models have been before 1926 and most predating The Federal Reserve. The 1966 one came in 12th worst for my particular scenario. FIREcalc data goes back to 1889 or something like that.

However it’s fine with me having data from some really rough periods. It’s a good test of the robustness of my plans as far as I’m concerned.

I’m not using the Bengen/Trinity withdrawal method so many of the details don’t apply.

we use 95/5 soweuse a difference formula too .

we find that while our goalposts each year are 4% of the balance we actually draw a lot less then that.

so every 4-5 years we get a new high end car