100% stocks with 3% SWR with 60 years retirement period

Discuss all general (i.e. non-personal) investing questions and issues, investing news, and theory.
Post Reply
Topic Author
alibaba123
Posts: 6
Joined: Mon Dec 31, 2018 12:51 pm

100% stocks with 3% SWR with 60 years retirement period

Post by alibaba123 » Mon May 13, 2019 7:53 am

Hi all,

I've been doing some simulations on CFIRESIM and got the following results with a 1,000,000 portfolio and 3% SWR with 60 years retirement period:

100% Stocks:

Success Rate Avg. Portfolio at Retirement
100% $1,000,000
Ending Portfolio Yearly Withdrawals Total Withdrawals
Average: $20,212,739 $30,000 $1,830,000
Median: $18,819,291 $30,000 $1,830,000
St. Dev.: $10,439,963 $0 $0
Highest: $46,539,389 $30,000 $1,830,000
Lowest: $2,548,686 $30,000 $1,830,000


75% Stocks 25% Bonds:

Success Rate Avg. Portfolio at Retirement
100% $1,000,000
Ending Portfolio Yearly Withdrawals Total Withdrawals
Average: $10,209,683 $30,000 $1,830,000
Median: $9,145,287 $30,000 $1,830,000
St. Dev.: $5,012,639 $0 $0
Highest: $23,205,198 $30,000 $1,830,000
Lowest: $2,486,326 $30,000 $1,830,000


50% Stocks 50% Bonds:

Success Rate Avg. Portfolio at Retirement
100% $1,000,000
Ending Portfolio Yearly Withdrawals Total Withdrawals
Average: $6,107,843 $30,000 $1,830,000
Median: $4,735,748 $30,000 $1,830,000
St. Dev.: $3,810,146 $0 $0
Highest: $17,331,059 $30,000 $1,830,000
Lowest: $931,759 $30,000 $1,830,000


It seems to me that the safest allocation, safest meaning the lowest, average and median portfolio balance at the end of 60 years, is 100% stocks. However i often see that 100% stocks is strongly discouraged on this forum and considering the data i have retrieved I would like to know why? Is there something i'm missing or wrong about?

22twain
Posts: 1821
Joined: Thu May 10, 2012 5:42 pm

Re: 100% stocks with 3% SWR with 60 years retirement period

Post by 22twain » Mon May 13, 2019 8:22 am

The FIRE group notwithstanding, most people don't have the luxury of a 60 year retirement period.
My investing princiPLEs do not include absolutely preserving princiPAL.

Topic Author
alibaba123
Posts: 6
Joined: Mon Dec 31, 2018 12:51 pm

Re: 100% stocks with 3% SWR with 60 years retirement period

Post by alibaba123 » Mon May 13, 2019 8:36 am

22twain wrote:
Mon May 13, 2019 8:22 am
The FIRE group notwithstanding, most people don't have the luxury of a 60 year retirement period.
Hi 22twain,

Fair enough, I have revised the simulation to be 30 year periods:

100% stocks:

Success Rate Avg. Portfolio at Retirement
100% $1,000,000
Ending Portfolio Yearly Withdrawals Total Withdrawals
Average: $4,004,046 $30,000 $930,000
Median: $3,276,112 $30,000 $930,000
St. Dev.: $2,546,634 $0 $0
Highest: $11,375,936 $30,000 $930,000
Lowest: $706,676 $30,000 $930,000


75% stocks 25% bonds:

Success Rate Avg. Portfolio at Retirement
100% $1,000,000
Ending Portfolio Yearly Withdrawals Total Withdrawals
Average: $2,873,817 $30,000 $930,000
Median: $2,390,766 $30,000 $930,000
St. Dev.: $1,738,028 $0 $0
Highest: $7,170,438 $30,000 $930,000
Lowest: $553,632 $30,000 $930,000


50% stocks 25% bonds:

Success Rate Avg. Portfolio at Retirement
100% $1,000,000
Ending Portfolio Yearly Withdrawals Total Withdrawals
Average: $1,927,978 $30,000 $930,000
Median: $1,272,592 $30,000 $930,000
St. Dev.: $1,343,271 $0 $0
Highest: $5,883,614 $30,000 $930,000
Lowest: $370,310 $30,000 $930,000


It seems that 100% stocks still comes up ahead, even at the lowest ending balance

RJC
Posts: 167
Joined: Fri Dec 14, 2018 1:40 pm

Re: 100% stocks with 3% SWR with 60 years retirement period

Post by RJC » Mon May 13, 2019 8:42 am

I've wondered this myself. I don't think there has been a 15-year period where stocks were down. Maybe because most folks have enough by retirement that they do not need to take on the additional volatility?

dbr
Posts: 28881
Joined: Sun Mar 04, 2007 9:50 am

Re: 100% stocks with 3% SWR with 60 years retirement period

Post by dbr » Mon May 13, 2019 8:48 am

alibaba123 wrote:
Mon May 13, 2019 7:53 am


It seems to me that the safest allocation, safest meaning the lowest, average and median portfolio balance at the end of 60 years, is 100% stocks. However i often see that 100% stocks is strongly discouraged on this forum and considering the data i have retrieved I would like to know why? Is there something i'm missing or wrong about?
In post this result all the time. CFIRESIM is nice but even in FireCalc you can just look at the chart of portfolio trajectories at different asset allocations and this fact stands out like a sore thumb.

But, I think where the nit would be is that how much money you die with is hardly a measure of safety. The measure of safety is how bad are the worst outcomes and how many of them are there. The effect of asset allocation on that is very small compared to the effect of withdrawal rate. In any model using historical data any asset allocation is safe at a low enough withdrawal rate. You loaded the dice by lowering the withdrawal rate to 3%, which never fails at any length of time in data over the last hundred years or more (as CFIRESIM and FIRECALC do that calculation).

If you increase the withdrawal rate to start to get some failures you will find that the longer the timeline the more stock you want to have. It will still be true that the more stock you have the wealthier you heirs will be. At shorter timelines 100% stock is not the calculated optimum, though the difference in failure rates is not significant.

So the reason people don't invest 100% stocks in retirement probably includes the following:

1. People don't want to experience the wide swings in wealth in holding 100% stocks. The fear that this time is different and market really is going to disintegrate can be real. The one actual mistake is to panic and sell out.
2. People want the outcome to be more predicatable.
3. The ranges of outcomes for both 100% stocks and less in stocks overlap significantly. There are meaningful chances a low stock portfolio will actually end at greater wealth than a high stock portfolio. You can see that in the lowest, mean, and highest data in your list. Therefore it is not clear there is really a game here that one needs to play.
4. However one looks at it the principle of diversification should be honored. Somehow, someway one does not want to be caught in something really bad happening to the one asset one has. I personally think diversification should go in the direction of annuitizing rather than of holding bonds, but that is a discussion.
5. A lot of people just don't understand how investing works and somehow think that bonds are for "safety" and that stocks are "risky". Wait, maybe those people do understand something about how investing works.

Larry Swedroe in his advocacy of the need, ability, and willingness approach to risk in investing says the trump card is to take no more risk than necessary, to recognize that the marginal utility of wealth is a diminishing function, and that people tend to underestimate risk.

Vanguard Fan 1367
Posts: 839
Joined: Wed Feb 08, 2017 3:09 pm

Re: 100% stocks with 3% SWR with 60 years retirement period

Post by Vanguard Fan 1367 » Mon May 13, 2019 9:13 am

Since I read books like How To Prepare for the Coming Bad Times and The Coming Economic Earthquake I am interested in safe withdrawel rates.

Unless some sort of disaster happens 3 percent seems pretty safe on the withdrawal rate. Interesting study on the stock mix.

longinvest
Posts: 3342
Joined: Sat Aug 11, 2012 8:44 am

Re: 100% stocks with 3% SWR with 60 years retirement period

Post by longinvest » Mon May 13, 2019 9:20 am

alibaba123 wrote:
Mon May 13, 2019 7:53 am
It seems to me that the safest allocation, safest meaning the lowest, average and median portfolio balance at the end of 60 years, is 100% stocks. However i often see that 100% stocks is strongly discouraged on this forum and considering the data i have retrieved I would like to know why? Is there something i'm missing or wrong about?
There was just another thread discussing the same subject. Maybe the answer is to question the idea of extracting a fixed inflation-indexed amount of money, every year, from a portfolio of fluctuating assets and consider using a variable withdrawal method instead: Firecalc is telling me to be [aggressive]: 85/15. Should I trust the advice?
Bogleheads investment philosophy | single-ETF balanced portfolio | VBAL

dbr
Posts: 28881
Joined: Sun Mar 04, 2007 9:50 am

Re: 100% stocks with 3% SWR with 60 years retirement period

Post by dbr » Mon May 13, 2019 9:25 am

longinvest wrote:
Mon May 13, 2019 9:20 am
alibaba123 wrote:
Mon May 13, 2019 7:53 am
It seems to me that the safest allocation, safest meaning the lowest, average and median portfolio balance at the end of 60 years, is 100% stocks. However i often see that 100% stocks is strongly discouraged on this forum and considering the data i have retrieved I would like to know why? Is there something i'm missing or wrong about?
There was just another thread discussing the same subject. Maybe the answer is to question the idea of extracting a fixed inflation-indexed amount of money, every year, from a portfolio of fluctuating assets and consider using a variable withdrawal method instead: Firecalc is telling me to be [aggressive]: 85/15. Should I trust the advice?
It would certainly seem to be true that withdrawal profile is a more significant lever on preventing retirement failure while still trying to spend money than is asset allocation. On the other hand if the objective is to die as rich as possible, then the tactic would undoubtedly be to spend nothing and put everything in stocks. I think most people here don't even think about how much money they will die with when contemplating retirement. Ironically, one of the most "secure" retirement tools, annuities, and also devices such as TIPS ladders start by taking away the assets entirely at the get go. Yes, the TIPS ladder also does this if followed to the letter.

YRT70
Posts: 76
Joined: Sat Apr 27, 2019 8:51 am

Re: 100% stocks with 3% SWR with 60 years retirement period

Post by YRT70 » Mon May 13, 2019 9:57 am

Interesting question. I posted something similar yesterday. Longinvest just linked to it.

I found this an interesting article by Wade Pfau https://retirementresearcher.com/willia ... s-safemax/

badger42
Posts: 470
Joined: Thu Apr 09, 2015 9:01 am

Re: 100% stocks with 3% SWR with 60 years retirement period

Post by badger42 » Mon May 13, 2019 10:43 am

The thing about blindly extending a simulation is that we have to remember 60 years is a REALLY long time.

60 years ago was 1959. The world was VERY different in 1959 (and not just because my parents were little kids). The global economy was still being rebuilt after World War II, computers were not yet really commercially viable, long distance telephone service was a fairly recent miracle, and stocks were traded by humans.

The conditions in 60 years are probably even less predictable than today would have been 60 years ago.

I would argue that's a really good argument for diversification. If I were building a 60 year portfolio, I would look at various flavors of tail risk and outperformance / underperformance - at a minimum, something like 10% in TIPS, 10% in real estate, maybe 5% in commodities would likely get you something closer to an "All weather" portfolio for 60 years than 100% in equities.

60 years of "the world changed" is really hard to represent in a simulation; this is why there's still some art to allocation, not 100% science.

dayzero
Posts: 1014
Joined: Fri May 16, 2008 11:04 pm

Re: 100% stocks with 3% SWR with 60 years retirement period

Post by dayzero » Mon May 13, 2019 10:50 am

Good analysis of 60 year retirement periods:

https://earlyretirementnow.com/2016/12/ ... t-1-intro/

aristotelian
Posts: 5593
Joined: Wed Jan 11, 2017 8:05 pm

Re: 100% stocks with 3% SWR with 60 years retirement period

Post by aristotelian » Mon May 13, 2019 11:02 am

Yes, if you look at ERN's charts, increasing % of stocks generally has higher success rates over long time periods. It is only when you get under 30 years that you really need to worry about market risk. I think this really puts in perspective that "risk" is not the same as volatility, and investors should be equally concerned with inflation risk and market risk.

Still, most people cannot handle the volatility of 100% stocks. There could also be a Black Swan/Japan scenario that changes the numbers. Some people may not be comfortable with the risk regardless of what has happened in backtesting. Accepting the lower expected return of bonds may be a good price to pay for reduced volatility.

How do the numbers look at 3.5%?

RAchip
Posts: 292
Joined: Sat May 07, 2016 7:31 pm

Re: 100% stocks with 3% SWR with 60 years retirement period

Post by RAchip » Mon May 13, 2019 11:05 am

The data is clear. Bonds are for suckers if you are investing for 30+ years.


Topic Author
alibaba123
Posts: 6
Joined: Mon Dec 31, 2018 12:51 pm

Re: 100% stocks with 3% SWR with 60 years retirement period

Post by alibaba123 » Mon May 13, 2019 11:15 am

aristotelian wrote:
Mon May 13, 2019 11:02 am
Yes, if you look at ERN's charts, increasing % of stocks generally has higher success rates over long time periods. It is only when you get under 30 years that you really need to worry about market risk. I think this really puts in perspective that "risk" is not the same as volatility, and investors should be equally concerned with inflation risk and market risk.

Still, most people cannot handle the volatility of 100% stocks. There could also be a Black Swan/Japan scenario that changes the numbers. Some people may not be comfortable with the risk regardless of what has happened in backtesting. Accepting the lower expected return of bonds may be a good price to pay for reduced volatility.

How do the numbers look at 3.5%?
At 3.5 % withdrawal rate with a 30 hear horizon:

100% stocks:

Success Rate Avg. Portfolio at Retirement
100% $1,000,000
Ending Portfolio Yearly Withdrawals Total Withdrawals
Average: $3,476,293 $35,000 $1,085,000
Median: $2,971,258 $35,000 $1,085,000
St. Dev.: $2,467,201 $0 $0
Highest: $10,431,724 $35,000 $1,085,000
Lowest: $72,469 $35,000 $1,085,000


75% stocks 25% bonds:

Success Rate Avg. Portfolio at Retirement
100% $1,000,000
Ending Portfolio Yearly Withdrawals Total Withdrawals
Average: $2,440,064 $35,000 $1,085,000
Median: $1,877,666 $35,000 $1,085,000
St. Dev.: $1,699,201 $0 $0
Highest: $6,535,564 $35,000 $1,085,000
Lowest: $23,144 $35,000 $1,085,000


50% stocks 50% bonds:

Success Rate Avg. Portfolio at Retirement
98.29% $1,000,000
Ending Portfolio Yearly Withdrawals Total Withdrawals
Average: $1,576,003 $35,000 $1,085,000
Median: $996,005 $35,000 $1,085,000
St. Dev.: $1,311,995 $0 $0
Highest: $5,403,704 $35,000 $1,085,000
Lowest: $-67,350 $35,000 $1,085,000


100% stocks still comes out ahead. I've tried changing the withdrawal method to constant percentage, which I assume longinvest's suggestion of VPW is a subset of, and still 100% stocks comes out ahead.

User avatar
Tyler9000
Posts: 410
Joined: Fri Aug 21, 2015 11:57 am

Re: 100% stocks with 3% SWR with 60 years retirement period

Post by Tyler9000 » Mon May 13, 2019 11:18 am

alibaba123 wrote:
Mon May 13, 2019 7:53 am
It seems to me that the safest allocation, safest meaning the lowest, average and median portfolio balance at the end of 60 years, is 100% stocks. However i often see that 100% stocks is strongly discouraged on this forum and considering the data i have retrieved I would like to know why? Is there something i'm missing or wrong about?
For me, the biggest issue is that not all stocks and bonds are created equal. Mix things up beyond the default US S&P500 and intermediate term bonds used in the calculators you're studying, and the results can change quite a bit where much lower percentages of stocks did significantly better. Basically, most old-school retirement calculations are ignorant to even the simplest ideas of portfolio theory such as international diversification. Just because there's lots of data does not mean the data applies to you!

Psychologically speaking, I imagine the reason that very high percentages of stocks are often discouraged here is that the board is generally a little older and experienced enough to remember just how painful a true market drawdown in the 50% range is to an investor. It's the difference between knowledge and wisdom -- even the theoretically best plan is useless if you can't stick with it through the worst times. Everyone has a different tolerance for that type of thing, but in my experience people are never as strong as they think they will be when it seems like the financial world is crumbling around them. It's important to know thyself.

And finally, I personally believe very long 60-year retirements require a different mindset. Rather than looking at safe withdrawal rates, the more practical measure is the perpetual withdrawal rate that preserved inflation-adjusted principal. And rather than thinking about maximizing end portfolio balances, it's really about building a comfortable, sustainable life. You're doing a great job studying the data, but make sure you're also thinking beyond the numbers right in front of you to the larger design challenge. There's so much more to life than simply maximizing returns.
Last edited by Tyler9000 on Mon May 13, 2019 12:05 pm, edited 2 times in total.

Random Poster
Posts: 1775
Joined: Wed Feb 03, 2010 10:17 am

Re: 100% stocks with 3% SWR with 60 years retirement period

Post by Random Poster » Mon May 13, 2019 11:32 am

alibaba123 wrote:
Mon May 13, 2019 11:15 am

At 3.5 % withdrawal rate with a 30 hear horizon:

50% stocks 50% bonds:

Success Rate Avg. Portfolio at Retirement
98.29% $1,000,000
Curious that the 98.29% rate you show doesn't match the 100% rate shown by ERN in the post immediately prior to yours.

Topic Author
alibaba123
Posts: 6
Joined: Mon Dec 31, 2018 12:51 pm

Re: 100% stocks with 3% SWR with 60 years retirement period

Post by alibaba123 » Mon May 13, 2019 11:41 am

Random Poster wrote:
Mon May 13, 2019 11:32 am
alibaba123 wrote:
Mon May 13, 2019 11:15 am

At 3.5 % withdrawal rate with a 30 hear horizon:

50% stocks 50% bonds:

Success Rate Avg. Portfolio at Retirement
98.29% $1,000,000
Curious that the 98.29% rate you show doesn't match the 100% rate shown by ERN in the post immediately prior to yours.
It could be due to the fees. I had input it as 0.25%. I just saw on ERN'S blog that he used a 0.05% fee rate, which when i input it in i get 100% success as well.

YRT70
Posts: 76
Joined: Sat Apr 27, 2019 8:51 am

Re: 100% stocks with 3% SWR with 60 years retirement period

Post by YRT70 » Mon May 13, 2019 11:45 am

Curious what you get with this Vanguard tool. https://retirementplans.vanguard.com/VG ... ggCalc.jsf

In my calculations it's showing a benefit of higher bond allocations where other tools didn't.

J295
Posts: 1998
Joined: Sun Jan 01, 2012 11:40 pm

Re: 100% stocks with 3% SWR with 60 years retirement period

Post by J295 » Mon May 13, 2019 11:50 am

Living on $30,000 per year not an option for us.

User avatar
HomerJ
Posts: 12769
Joined: Fri Jun 06, 2008 12:50 pm

Re: 100% stocks with 3% SWR with 60 years retirement period

Post by HomerJ » Mon May 13, 2019 11:55 am

alibaba123 wrote:
Mon May 13, 2019 7:53 am
It seems to me that the safest allocation, safest meaning the lowest, average and median portfolio balance at the end of 60 years, is 100% stocks. However i often see that 100% stocks is strongly discouraged on this forum and considering the data i have retrieved I would like to know why? Is there something i'm missing or wrong about?
Try 4%. 100% stocks definitely failed during the Great Depression if you were pulling 4% a year.

I'm surprised 3% didn't fail also, but maybe that was a low enough withdrawal to survive. Most people would have had a heart-attack on that ride though.

Retire with a $1 million, be down to $620,000 or so after the 1st year. Then down into the low $300,000s after the second year.

And into the $200,000s by year 3.

Year 4, it would have doubled back into the $400,000s.

Not going to check further. I did go further once before and discovered 4% withdrawals were not sustainable. You were completely broke about 20 years in.

Maybe 3% was low enough to let you bounce back. But having at least some bonds (which did well) would have made those first years a LOT easier on your stomach.
Last edited by HomerJ on Mon May 13, 2019 12:10 pm, edited 1 time in total.
The J stands for Jay

User avatar
HomerJ
Posts: 12769
Joined: Fri Jun 06, 2008 12:50 pm

Re: 100% stocks with 3% SWR with 60 years retirement period

Post by HomerJ » Mon May 13, 2019 12:00 pm

alibaba123 wrote:
Mon May 13, 2019 7:53 am
Hi all,

I've been doing some simulations on CFIRESIM and got the following results with a 1,000,000 portfolio and 3% SWR with 60 years retirement period:

100% Stocks:

Success Rate Avg. Portfolio at Retirement
100% $1,000,000
Ending Portfolio Yearly Withdrawals Total Withdrawals
Average: $20,212,739 $30,000 $1,830,000
Median: $18,819,291 $30,000 $1,830,000
St. Dev.: $10,439,963 $0 $0
Highest: $46,539,389 $30,000 $1,830,000
Lowest: $2,548,686 $30,000 $1,830,000
Oh I see part of the problem. You're not adjusting your withdrawals for inflation at all.

$30,000 from $1 million is a pretty low withdrawal rate. Not adjusting for inflation makes it a lot worse. By year 30, you'd have to live on today's equivalent of $15,000 a year, and by year 50, you'd be living on the equivalent of $8000 a year in today's money.

And that's assuming low inflation.

No thanks.
The J stands for Jay

Topic Author
alibaba123
Posts: 6
Joined: Mon Dec 31, 2018 12:51 pm

Re: 100% stocks with 3% SWR with 60 years retirement period

Post by alibaba123 » Mon May 13, 2019 12:09 pm

HomerJ wrote:
Mon May 13, 2019 12:00 pm
alibaba123 wrote:
Mon May 13, 2019 7:53 am
Hi all,

I've been doing some simulations on CFIRESIM and got the following results with a 1,000,000 portfolio and 3% SWR with 60 years retirement period:

100% Stocks:

Success Rate Avg. Portfolio at Retirement
100% $1,000,000
Ending Portfolio Yearly Withdrawals Total Withdrawals
Average: $20,212,739 $30,000 $1,830,000
Median: $18,819,291 $30,000 $1,830,000
St. Dev.: $10,439,963 $0 $0
Highest: $46,539,389 $30,000 $1,830,000
Lowest: $2,548,686 $30,000 $1,830,000
Oh I see part of the problem. You're not adjusting your withdrawals for inflation at all.

$30,000 from $1 million is a pretty low withdrawal rate. Not adjusting for inflation makes it a lot worse. By year 30, you'd have to live on today's equivalent of $15,000 a year, and by year 50, you'd be living on the equivalent of $8000 a year in today's money.

And that's assuming low inflation.

No thanks.
Hi HomerJ, I did select adjust for inflation. I believe the numbers you see under withdrawal and total withdrawals are in inflation adjusted terms. If i select "not inflation-adjusted" the ending balances are higher.

mariezzz
Posts: 620
Joined: Mon Oct 02, 2017 11:02 pm

Re: 100% stocks with 3% SWR with 60 years retirement period

Post by mariezzz » Mon May 13, 2019 12:12 pm

There have been threads on this forum and elsewhere which discuss the fact that for periods longer than 30 years retirement, the effect of one's allocation doesn't meaningfully change, compared to about a 30 year retirement.

User avatar
HomerJ
Posts: 12769
Joined: Fri Jun 06, 2008 12:50 pm

Re: 100% stocks with 3% SWR with 60 years retirement period

Post by HomerJ » Mon May 13, 2019 12:13 pm

alibaba123 wrote:
Mon May 13, 2019 12:09 pm
HomerJ wrote:
Mon May 13, 2019 12:00 pm
alibaba123 wrote:
Mon May 13, 2019 7:53 am
Hi all,

I've been doing some simulations on CFIRESIM and got the following results with a 1,000,000 portfolio and 3% SWR with 60 years retirement period:

100% Stocks:

Success Rate Avg. Portfolio at Retirement
100% $1,000,000
Ending Portfolio Yearly Withdrawals Total Withdrawals
Average: $20,212,739 $30,000 $1,830,000
Median: $18,819,291 $30,000 $1,830,000
St. Dev.: $10,439,963 $0 $0
Highest: $46,539,389 $30,000 $1,830,000
Lowest: $2,548,686 $30,000 $1,830,000
Oh I see part of the problem. You're not adjusting your withdrawals for inflation at all.

$30,000 from $1 million is a pretty low withdrawal rate. Not adjusting for inflation makes it a lot worse. By year 30, you'd have to live on today's equivalent of $15,000 a year, and by year 50, you'd be living on the equivalent of $8000 a year in today's money.

And that's assuming low inflation.

No thanks.
Hi HomerJ, I did select adjust for inflation. I believe the numbers you see under withdrawal and total withdrawals are in inflation adjusted terms. If i select "not inflation-adjusted" the ending balances are higher.
Ah, I stand corrected then. I'm still surprised 3% worked during the Great Depression.

You definitely would have been down more than 70% three years into your retirement. That would be very scary. Start with $1 million, 3 years later, you only have $250,000 left with 57 years to go.
The J stands for Jay

User avatar
dodecahedron
Posts: 4245
Joined: Tue Nov 12, 2013 12:28 pm

Re: 100% stocks with 3% SWR with 60 years retirement period

Post by dodecahedron » Mon May 13, 2019 12:44 pm

badger42 wrote:
Mon May 13, 2019 10:43 am
The thing about blindly extending a simulation is that we have to remember 60 years is a REALLY long time.

60 years ago was 1959. The world was VERY different in 1959 (and not just because my parents were little kids). The global economy was still being rebuilt after World War II, computers were not yet really commercially viable, long distance telephone service was a fairly recent miracle, and stocks were traded by humans.
I was a little kid then. Indeed the world was very different. Also in 1959:
  • The US and major western economic powers were still attempting to adhere to some version of a gold standard.
  • OPEC was not yet born, let alone a prominent force on the world economic radar screen.
  • Inflation-protected securities were not even on a drawing board in the US or elsewhere.
  • Individual investors faced very high frictional trading costs compared to today. Minimum commissions were set by government regulation. Discount brokers did not exist. Spreads were necessarily higher, particularly in real terms, because stock prices were quoted in eighths of a dollar, rather than in cents.
  • Top income tax bracket rates were extremely high by modern standards (90% federal on ordinary income). This added further to frictional trading costs.
  • Tax-advantaged DC retirement plans for typical workers were almost non-existent. (Major exceptions were college employees, who had access to TIAA-CREF 403b plans.)
  • Taxes were not indexed for inflation, leading to bracket creep. Social Security was also not indexed for inflation.
  • Tax breaks for whole life insurance were far more generous than today. Estate taxes were a significant planning issue for far more American investors than today.
  • Major sectors of the economy (e.g., airlines, trucking, rail, telecommunication, banking) were subject to heavy price and entry government regulation, restricting competition and consumer choices.
  • International trade (exports) was a much smaller share of US GDP. In US macroeconomics models back in the day, we often neglected the international sector. Health care was also a much smaller share of GDP.

Amadis_of_Gaul
Posts: 93
Joined: Tue Dec 18, 2018 5:57 pm

Re: 100% stocks with 3% SWR with 60 years retirement period

Post by Amadis_of_Gaul » Mon May 13, 2019 1:18 pm

longinvest wrote:
Mon May 13, 2019 9:20 am

There was just another thread discussing the same subject. Maybe the answer is to question the idea of extracting a fixed inflation-indexed amount of money, every year, from a portfolio of fluctuating assets and consider using a variable withdrawal method instead: Firecalc is telling me to be [aggressive]: 85/15. Should I trust the advice?
I think it's possible to go one step further here. Recently, I've been doing a lot of reading on lifecycle investing, and it's got me convinced that articulating our goals and desigining/implementing a plan to achieve those goals is the most important way to think about investing. If your goal is, "Die with as much money as possible," a 100 percent stock allocation is a great way to get there.

However, that's not my goal. I want to provide a secure, stable retirement income for my wife and me, with some money for some nice extras from time to time. 100 percent stock with a 3 percent SWR is not the way to get there. In order to make sure that I don't run out of money, I am virtually guaranteeing that I will die with a vast unspent sum of money. I love my children, but I am not at all sure that dumping $10 million real on them would be doing them any favors!

Honestly, I am equally suspicious of VPR. It avoids the vast-unspent-sum problem, but it doesn't do a great job of smoothing retirement income. One year you might have $50K to spend, the next year $30K.

Instead, it makes the most sense to me to set up a floor of duration-matched inflation-adjusted bond funds. They can be relied on to generate the income that I want for 30 years within a few percentage points. Then, all that equity volatility is volatility in the whipped cream, not the sundae.

Of course, I don't think this helps the OP much, with his 60-year retirement horizon. I don't think duration-matching would work too well over 60 years, and without Social Security for most of those 60 years (and it won't be very impressive when it shows up anyway), it's tough to build a floor. I've read everything ERN has written, but I'm skeptical of the early-retirement project generally (except for the folks who by "retirement" mean "beginning a second career as a landlord/blogger/whatever"). That's a lot of faith to place in the U.S. stock market. Yes, it always has gone up, but there are no guarantees that it won't pull a Nikkei and fall never to rise again.

Valuethinker
Posts: 37840
Joined: Fri May 11, 2007 11:07 am

Re: 100% stocks with 3% SWR with 60 years retirement period

Post by Valuethinker » Tue May 14, 2019 3:55 am

Tyler9000 wrote:
Mon May 13, 2019 11:18 am
alibaba123 wrote:
Mon May 13, 2019 7:53 am
It seems to me that the safest allocation, safest meaning the lowest, average and median portfolio balance at the end of 60 years, is 100% stocks. However i often see that 100% stocks is strongly discouraged on this forum and considering the data i have retrieved I would like to know why? Is there something i'm missing or wrong about?
For me, the biggest issue is that not all stocks and bonds are created equal. Mix things up beyond the default US S&P500 and intermediate term bonds used in the calculators you're studying, and the results can change quite a bit where much lower percentages of stocks did significantly better. Basically, most old-school retirement calculations are ignorant to even the simplest ideas of portfolio theory such as international diversification. Just because there's lots of data does not mean the data applies to you!

Psychologically speaking, I imagine the reason that very high percentages of stocks are often discouraged here is that the board is generally a little older and experienced enough to remember just how painful a true market drawdown in the 50% range is to an investor. It's the difference between knowledge and wisdom -- even the theoretically best plan is useless if you can't stick with it through the worst times. Everyone has a different tolerance for that type of thing, but in my experience people are never as strong as they think they will be when it seems like the financial world is crumbling around them. It's important to know thyself.

And finally, I personally believe very long 60-year retirements require a different mindset. Rather than looking at safe withdrawal rates, the more practical measure is the perpetual withdrawal rate that preserved inflation-adjusted principal. And rather than thinking about maximizing end portfolio balances, it's really about building a comfortable, sustainable life. You're doing a great job studying the data, but make sure you're also thinking beyond the numbers right in front of you to the larger design challenge. There's so much more to life than simply maximizing returns.
You've nailed it.

Perpetual withdrawal rate not safe withdrawal rate.

That's the nub for very long time periods.

JBeck
Posts: 42
Joined: Fri Apr 15, 2016 4:54 am

Re: 100% stocks with 3% SWR with 60 years retirement period

Post by JBeck » Tue May 14, 2019 7:00 am

Tyler9000 wrote:
Mon May 13, 2019 11:18 am
alibaba123 wrote:
Mon May 13, 2019 7:53 am
It seems to me that the safest allocation, safest meaning the lowest, average and median portfolio balance at the end of 60 years, is 100% stocks. However i often see that 100% stocks is strongly discouraged on this forum and considering the data i have retrieved I would like to know why? Is there something i'm missing or wrong about?
For me, the biggest issue is that not all stocks and bonds are created equal. Mix things up beyond the default US S&P500 and intermediate term bonds used in the calculators you're studying, and the results can change quite a bit where much lower percentages of stocks did significantly better. Basically, most old-school retirement calculations are ignorant to even the simplest ideas of portfolio theory such as international diversification. Just because there's lots of data does not mean the data applies to you!

Psychologically speaking, I imagine the reason that very high percentages of stocks are often discouraged here is that the board is generally a little older and experienced enough to remember just how painful a true market drawdown in the 50% range is to an investor. It's the difference between knowledge and wisdom -- even the theoretically best plan is useless if you can't stick with it through the worst times. Everyone has a different tolerance for that type of thing, but in my experience people are never as strong as they think they will be when it seems like the financial world is crumbling around them. It's important to know thyself.

And finally, I personally believe very long 60-year retirements require a different mindset. Rather than looking at safe withdrawal rates, the more practical measure is the perpetual withdrawal rate that preserved inflation-adjusted principal. And rather than thinking about maximizing end portfolio balances, it's really about building a comfortable, sustainable life. You're doing a great job studying the data, but make sure you're also thinking beyond the numbers right in front of you to the larger design challenge. There's so much more to life than simply maximizing returns.
Deleted
Last edited by JBeck on Tue May 14, 2019 2:35 pm, edited 1 time in total.

CraigTester
Posts: 113
Joined: Wed Aug 08, 2018 6:34 am

Re: 100% stocks with 3% SWR with 60 years retirement period

Post by CraigTester » Tue May 14, 2019 11:03 am

YRT70 wrote:
Mon May 13, 2019 11:06 am
Ern's numbers

Image

https://earlyretirementnow.com/2016/12/ ... t-1-intro/
Not sure I'd put a lot of faith in this chart.

Look at the box for 0% stock, 30 years, 3% Withdrawal, as an example:

If 100% of your portfolio is invested in something that simply keeps up with inflation, it is "guaranteed" to last 33 years.
(e.g. 3% of $1,000,0000 = $30,000....So using simple division: $1,000,000/$30,000/year = 33 years.)

Yet his chart says this combination only survives for 30 years, 89% of the time?

CraigTester

User avatar
Time2Quit
Posts: 198
Joined: Thu Nov 01, 2018 9:47 am
Location: Fridgid State

Re: 100% stocks with 3% SWR with 60 years retirement period

Post by Time2Quit » Tue May 14, 2019 11:26 am

CraigTester wrote:
Tue May 14, 2019 11:03 am
YRT70 wrote:
Mon May 13, 2019 11:06 am
Ern's numbers

Image

https://earlyretirementnow.com/2016/12/ ... t-1-intro/
Not sure I'd put a lot of faith in this chart.

Look at the box for 0% stock, 30 years, 3% Withdrawal, as an example:

If 100% of your portfolio is invested in something that simply keeps up with inflation, it is "guaranteed" to last 33 years.
(e.g. 3% of $1,000,0000 = $30,000....So using simple division: $1,000,000/$30,000/year = 33 years.)

Yet his chart says this combination only survives for 30 years, 89% of the time?

CraigTester
You are forgetting to add CPI. The premise of the chart is % withdrawn the CPI every year in 30 year the amount withdrawn will be a lot more than the intital 3%($30k)
"It is not the man who has too little, but the man who craves more, that is poor." --Seneca

User avatar
Tyler9000
Posts: 410
Joined: Fri Aug 21, 2015 11:57 am

Re: 100% stocks with 3% SWR with 60 years retirement period

Post by Tyler9000 » Tue May 14, 2019 11:34 am

JBeck wrote:
Tue May 14, 2019 7:00 am
If a PWR was your objective, would you feel comfortable using the historical PWR for your desired asset allocation or would you dial it back further?
The PWR maintained the inflation-adjusted principal even in the very worst historical timeframe (stock crashes, skyrocketing interest rates, double-digit inflation, etc), and personally I believe it's plenty conservative for general planning purposes. Rather than planning for an even worse case and cutting your withdrawal rate to the bone, I would instead focus on building up your financial flexibility. Smart retirement planning is about more than just avoiding failure.

CraigTester
Posts: 113
Joined: Wed Aug 08, 2018 6:34 am

Re: 100% stocks with 3% SWR with 60 years retirement period

Post by CraigTester » Tue May 14, 2019 11:58 am

Time2Quit wrote:
Tue May 14, 2019 11:26 am
CraigTester wrote:
Tue May 14, 2019 11:03 am
YRT70 wrote:
Mon May 13, 2019 11:06 am
Ern's numbers

Image

https://earlyretirementnow.com/2016/12/ ... t-1-intro/
Not sure I'd put a lot of faith in this chart.

Look at the box for 0% stock, 30 years, 3% Withdrawal, as an example:

If 100% of your portfolio is invested in something that simply keeps up with inflation, it is "guaranteed" to last 33 years.
(e.g. 3% of $1,000,0000 = $30,000....So using simple division: $1,000,000/$30,000/year = 33 years.)

Yet his chart says this combination only survives for 30 years, 89% of the time?

CraigTester
You are forgetting to add CPI. The premise of the chart is % withdrawn the CPI every year in 30 year the amount withdrawn will be a lot more than the intital 3%($30k)
As explained in my example, I am adjusting for inflation.

E.g. Both the $1,000,000 and the $30,000 in my calculation are expressed in today's dollars.

So if anyone is relying on the posted chart, they should know that it has an error in it....Or at least an unexplained assumption.

CraigTester

User avatar
Tyler9000
Posts: 410
Joined: Fri Aug 21, 2015 11:57 am

Re: 100% stocks with 3% SWR with 60 years retirement period

Post by Tyler9000 » Tue May 14, 2019 12:09 pm

CraigTester wrote:
Tue May 14, 2019 11:03 am
If 100% of your portfolio is invested in something that simply keeps up with inflation, it is "guaranteed" to last 33 years. ... Yet his chart says this combination only survives for 30 years, 89% of the time?
Why do you assume that 10-year treasuries keep up with inflation?
Last edited by Tyler9000 on Tue May 14, 2019 1:27 pm, edited 2 times in total.

CraigTester
Posts: 113
Joined: Wed Aug 08, 2018 6:34 am

Re: 100% stocks with 3% SWR with 60 years retirement period

Post by CraigTester » Tue May 14, 2019 1:25 pm

Tyler9000 wrote:
Tue May 14, 2019 12:09 pm
CraigTester wrote:
Tue May 14, 2019 11:03 am
If 100% of your portfolio is invested in something that simply keeps up with inflation, it is "guaranteed" to last 33 years.
Why do you assume that 10-year treasuries keep up with inflation?
Because historically, 10-year treasuries have always had a "real" return whenever referring to a rolling 30 year time period..., which leads us to why they are so often used as a reasonable default for the "fixed income" choice when doing backtesting ....

So unless the author of the above table used something other than 10-year treasuries as his proxy for fixed-income, he made a calculation mistake....

And it is therefore possible/likely that he propagated this mistake through all of his results.....

But if he did use a fixed-income choice that didn't keep up with inflation, that alone would be reason enough to ignore the results of his table....because most people on this forum wouldn't store their entire fixed income allocation under their mattress for 30 years.....

Therefore, as a public service message, I'm letting everyone know to not draw any conclusions from the results displayed in the table....

CraigTester

User avatar
Tyler9000
Posts: 410
Joined: Fri Aug 21, 2015 11:57 am

Re: 100% stocks with 3% SWR with 60 years retirement period

Post by Tyler9000 » Tue May 14, 2019 1:37 pm

It's not that simple. If nominal bonds always kept up with inflation there would be no such thing as TIPS. And because of interest rate risk, all bonds have sequence of returns issues just like stocks.

I'll vouch for ERN's calculations here, and also point out that the results are reasonably consistent with the numbers from the Trinity Study which found that a 3% WR on a 100% bond portfolio succeeded only 80% of the time over 30 years (see table 3). What you're missing is the fact that nobody has the luxury in retirement to wait 30 years for their long term average returns to manifest before paying the bills along the way. So sequence of returns matters, and it always reduces the SWR from what you'd expect from the average alone. It's just how the math works.

redmaw
Posts: 3
Joined: Mon Apr 22, 2019 7:20 am

Re: 100% stocks with 3% SWR with 60 years retirement period

Post by redmaw » Tue May 14, 2019 4:16 pm

CraigTester wrote:
Tue May 14, 2019 1:25 pm
Tyler9000 wrote:
Tue May 14, 2019 12:09 pm
CraigTester wrote:
Tue May 14, 2019 11:03 am
If 100% of your portfolio is invested in something that simply keeps up with inflation, it is "guaranteed" to last 33 years.
Why do you assume that 10-year treasuries keep up with inflation?
Because historically, 10-year treasuries have always had a "real" return whenever referring to a rolling 30 year time period..., which leads us to why they are so often used as a reasonable default for the "fixed income" choice when doing backtesting ....

So unless the author of the above table used something other than 10-year treasuries as his proxy for fixed-income, he made a calculation mistake....

And it is therefore possible/likely that he propagated this mistake through all of his results.....

But if he did use a fixed-income choice that didn't keep up with inflation, that alone would be reason enough to ignore the results of his table....because most people on this forum wouldn't store their entire fixed income allocation under their mattress for 30 years.....

Therefore, as a public service message, I'm letting everyone know to not draw any conclusions from the results displayed in the table....

CraigTester
FYI in the article he says he uses 10Y treasury yields.

In this link https://earlyretirementnow.com/2016/05/ ... tock-risk/ he has a table that includes several long time spans where the 10Y bond has negative real returns, including 1971-1981 at -5.6% per year. If you started drawing 3% during that decade you will run out of money. 1915-1920 it was -13%, on top of the the preceding 15 years average -.5%. So lets say your real return is flat for 15 years pulling 3% out each year, you are now down 45% and then hit a 5 year run where your return is down 13% on top of 3% (of the original balance) withdraws. These sound like certain failures to me. It may not be intuitive but you can't say every 30 year period eventually averages to at least keep up with inflation therefore bonds can always support at least a 3.3% withdraw rate for 30 years.

columbia
Posts: 1484
Joined: Tue Aug 27, 2013 5:30 am

Re: 100% stocks with 3% SWR with 60 years retirement period

Post by columbia » Tue May 14, 2019 5:30 pm

CraigTester wrote:
Tue May 14, 2019 1:25 pm
Tyler9000 wrote:
Tue May 14, 2019 12:09 pm
CraigTester wrote:
Tue May 14, 2019 11:03 am
If 100% of your portfolio is invested in something that simply keeps up with inflation, it is "guaranteed" to last 33 years.
Why do you assume that 10-year treasuries keep up with inflation?
Because historically, 10-year treasuries have always had a "real" return whenever referring to a rolling 30 year time period..., which leads us to why they are so often used as a reasonable default for the "fixed income" choice when doing backtesting ....

So unless the author of the above table used something other than 10-year treasuries as his proxy for fixed-income, he made a calculation mistake....

And it is therefore possible/likely that he propagated this mistake through all of his results.....

But if he did use a fixed-income choice that didn't keep up with inflation, that alone would be reason enough to ignore the results of his table....because most people on this forum wouldn't store their entire fixed income allocation under their mattress for 30 years.....

Therefore, as a public service message, I'm letting everyone know to not draw any conclusions from the results displayed in the table....

CraigTester

I’ve seen some posts about ST treasuries beating inflation a Roth. Then again, few here would hold them in such an account.

User avatar
One Ping
Posts: 651
Joined: Thu Sep 24, 2015 4:53 pm

Re: 100% stocks with 3% SWR with 60 years retirement period

Post by One Ping » Tue May 14, 2019 5:56 pm

redmaw wrote:
Tue May 14, 2019 4:16 pm
CraigTester wrote:
Tue May 14, 2019 1:25 pm
Tyler9000 wrote:
Tue May 14, 2019 12:09 pm
CraigTester wrote:
Tue May 14, 2019 11:03 am
If 100% of your portfolio is invested in something that simply keeps up with inflation, it is "guaranteed" to last 33 years.
Why do you assume that 10-year treasuries keep up with inflation?
Because historically, 10-year treasuries have always had a "real" return whenever referring to a rolling 30 year time period..., which leads us to why they are so often used as a reasonable default for the "fixed income" choice when doing backtesting ....

So unless the author of the above table used something other than 10-year treasuries as his proxy for fixed-income, he made a calculation mistake....

And it is therefore possible/likely that he propagated this mistake through all of his results.....

But if he did use a fixed-income choice that didn't keep up with inflation, that alone would be reason enough to ignore the results of his table....because most people on this forum wouldn't store their entire fixed income allocation under their mattress for 30 years.....

Therefore, as a public service message, I'm letting everyone know to not draw any conclusions from the results displayed in the table....

CraigTester
FYI in the article he says he uses 10Y treasury yields.

In this link https://earlyretirementnow.com/2016/05/ ... tock-risk/ he has a table that includes several long time spans where the 10Y bond has negative real returns, including 1971-1981 at -5.6% per year. If you started drawing 3% during that decade you will run out of money. 1915-1920 it was -13%, on top of the the preceding 15 years average -.5%. So lets say your real return is flat for 15 years pulling 3% out each year, you are now down 45% and then hit a 5 year run where your return is down 13% on top of 3% (of the original balance) withdraws. These sound like certain failures to me. It may not be intuitive but you can't say every 30 year period eventually averages to at least keep up with inflation therefore bonds can always support at least a 3.3% withdraw rate for 30 years.
I believe what redmaw is referring to is known as Sequence of Returns risk. It's a b _ _ _ h. :wink:
"Re-verify our range to target ... one ping only."

dbr
Posts: 28881
Joined: Sun Mar 04, 2007 9:50 am

Re: 100% stocks with 3% SWR with 60 years retirement period

Post by dbr » Tue May 14, 2019 6:13 pm

One Ping wrote:
Tue May 14, 2019 5:56 pm

I believe what redmaw is referring to is known as Sequence of Returns risk. It's a b _ _ _ h. :wink:
There are different things that can happen. One can compare two periods and a sequence of returns problem would mean that the average return for both periods was the same but good returns happened early in one period and bad returns later while the opposite sequence happened in the other period. Good returns early is good for supporting withdrawals and bad returns early is bad for supporting withdrawals. A different thing that can happen is that the average return for one period is less than the average return for the other period. A low average return is bad for withdrawals. These two issues can be convoluted depending on the time spans involved.

User avatar
Cycle
Posts: 1213
Joined: Sun May 28, 2017 7:57 pm
Location: Minneapolis

Re: 100% stocks with 3% SWR with 60 years retirement period

Post by Cycle » Tue May 14, 2019 6:38 pm

Am I the only sucker who looks at this as an opportunity to market time. "Well I'm 85% stocks now, but if we ever get a 30% decline I'll change my AA to 100% stocks since I'm only 35yo and in the long run bonds are a drag".

Also having been investing thru the great recession, my foolish memory always tells me there should be a 40% drop every decade

Praise be to Bogle I'm 100% in target funds in 401k and total world in taxable.

Post Reply