50-60 year retirements - anyone have the numbers for withdrawal/success rates?

[Celda]

As we know, the Trinity Study was based on a 30-year timeframe. Does anyone have any links to articles or number crunching that show a 50 or 60-year timeframe, with varying withdrawal rates (2%, 3%, 4%) and their success rates (i.e. not running out of money)?

I am having a discussion with someone who is claiming that early retirement is too risky because you can never be sure if stocks will crash or if inflation goes up. They think that a 50 or 60-year timeframe has never been proven to be successful. I seem to remember that the numbers have been crunched before, can anyone point me in the right direction?

Thanks.

[FIPurpose]

There is an option to change the number of years to model on http://www.cfiresim.com/input.php.

Increase that number to 60 and model it yourself!

The default 75/25 portfolio over 60 year periods has an 82% success rate. Good odds if you ask me!

[Nords]

Raddr (a FIREd radiologist) did some long-term Monte Carlo studies a few years ago:
http://raddr-pages.com/research/

The issue is that the typical stock-market history (since 1871 or so) doesn't have enough 50-year periods to make a statistically significant historical simulation of retirements, so Monte Carlo is used instead.  This tends to make the analysis much more conservative than it may need to be, and it doesn't handle all the parameters for which there's less data-- like international mutual funds. 

Finally, due to the limitations on the simulation capabilities, it doesn't factor in annuity income or variable spending.  But here it is:
http://raddr-pages.com/research/enhanced_monte_simulation.htm

[kkbmustang]

Try www.cfiresim.com.

I'm not sure what percentage rate of success others recommend, but playing around on it, I've consistently gotten 96% success rates and higher, depending on different factors. And for a 50 year time frame. I'm not sure if this is one of those things that people just have different comfort levels with the success rate, but I'd like it to be darn near 100%.

[arebelspy]

EDIT: 3 answers while I was typing this.  You guys are helpful!

www.cfiresim.com lets you run those calculations yourself.

There are less 50 or 60 year periods than there are 30, of course, so a smaller sample size.

It looks like the standard scenario (4% SWR, inflation adjusted) is 93% successful over 30 years, 82% over 40 years, 78% over 50 years, 82% over 60 years*.

(*It's higher in the 60 case not because it's likely to last longer because you're FIRE'd longer, but due to the sample size I told you about.  That's probably one extra failure that happened in between 50 and 60 years ago that isn't caught by the 60-year samples.  Ditto on the below scenario with 40 years versus the higher ones.)

A 3.5% SWR gives you success rates of: 99% for 30 years, 95% for 40 years, 97% for 50 years, or 98% for 60 years.

If you want any nuances run, feel free to ask!

[trailrated]

25x expenses should last for life if invested decently.

[arebelspy]

The issue is that the typical stock-market history (since 1871 or so) doesn't have enough 50-year periods to make a statistically significant historical simulation of retirements, so Monte Carlo is used instead.

I used to believe that as well (based on the fact that I've heard it repeated so much, and indeed, have parroted it myself), but someone pointed out to me a few months ago that the 30-year simulations have a total of 115 runs, while the 50-year have a total of 95 cycles.  That's not THAT different, and if you think you can get statistically significant results from 115 runs, I don't see why you can't from 95.

In other words, I think FIRECalc/cFIREsim has limitations, but there aren't really any more limitations when running a 50-year retirement than when running a 30-year one - the limitations remain the same.

[Celda]

Thanks for the replies. I'll link him/her to cFireSim.

Here's the thread if anyone has advice on what I could say in addition:

http://www.reddit.com/r/casualiama/comments/28u7w0/i_save_over_half_my_income_and_plan_to_retire_in/ciele5a

[sol]

As Rebs pointed out, a 3.5% SWR has a higher success rate over 60 years than the traditional 4% does over 30 years.  And keep in mind the data say you can increase your SWR by about .5% by skewing more towards domestic small caps than the typical 60/40 split.

[arebelspy]

I suggest they read JLCollins' Stock Series:
http://jlcollinsnh.com/stock-series/

Then they won't be so scared of a market crash.

[arebelspy]

And keep in mind the data say you can increase your SWR by about .5% by skewing more towards domestic small caps than the typical 60/40 split.

Maybe historically.  I'm of the opinion that there's no such thing as a free lunch.

I have been known to be wrong though, and often.

[sol]

Maybe historically.  I'm of the opinion that there's no such thing as a free lunch.

Right, but this whole exercise is "historically". 

There are lots of reasons to think that the future will not look anything like the past, but unless you have something better to base your predictions on...

[Grog]

why are there so few period (both 30y/50y)? Does cFireSim always just take period starting from January?
I'm asking this because if you take in account all period of 600 months (50 years x 12) you have way more period than 95.

I'm asking this because I was toying around with the MSCI USA index monthly data starting from 1.1.1969 and the difference in return starting from different months are not negligible.

I wanted to get a feeling of the accuracy of all this "average annual performance" and is crazy how sensible this is. If I use a compunding yearly average of 6.8%, starting with 100 from 1. Jan 1969, I reach 1760 points (today is 1810) with a good approximation.

If I have started just 6 months later, 1. July, with the index dropping 13% at 87, the average perforamnce would have been 7.5%, so an increase of overall performance of 10% (and over only 45 years). Not negligible to me, couldn't all this simulator use 600 months period, starting from different months, instead of 50 years period from Jan to Jan? I know that maybe then is a problem to retrieve past data, but probably you will end up with more simulation time frame by doing this. Like starting from 1969 and taking 360-months period you'll end up with ~180 different starting months and hence different 30-y period (from 1969 to 1984).

I was just wondering, maybe someone already dismissed this approach because the market curve are too similar? Although the may vary greatly form month to month?

[dragoncar]

Maybe historically.  I'm of the opinion that there's no such thing as a free lunch.

Right, but this whole exercise is "historically". 

There are lots of reasons to think that the future will not look anything like the past, but unless you have something better to base your predictions on...

Still waiting for cfiresim to add those return numbers from 2014-2050!  I like "Rebs" since ARS seems a bit formal.

Wrt number of periods, I've also heard people discuss "non overlapping periods" but not sure how 3 is that much better than 2

[Celda]

I think I am doing something wrong for the calculator.

I put in $1,000,000 as the starting point. $25,000 yearly spending, inflation adjusted for the spending plan. 60 year timeframe. 75% equities 25% bonds. 0.5% fees, and 0% cash growth. $0 Social Security.

Rebalance annually - yes; Gradually change allocation.

And it gives me a 100% success rate - and all the results end up with 6 million, 7 million, 10 million, etc.

Surely this can't be right?

[MDM]

25000/1000000 = 2.5%.  Not surprising that this works in all cases.  See comments above about 4%, 3.5%, 4.5%, etc.

[dude]

The Retire Early Homepage guys did this a while ago:

http://www.retireearlyhomepage.com/restud1.html

[arebelspy]

I think I am doing something wrong for the calculator.

I put in $1,000,000 as the starting point. $25,000 yearly spending, inflation adjusted for the spending plan. 60 year timeframe. 75% equities 25% bonds. 0.5% fees, and 0% cash growth. $0 Social Security.

Rebalance annually - yes; Gradually change allocation.

And it gives me a 100% success rate - and all the results end up with 6 million, 7 million, 10 million, etc.

Surely this can't be right?

Yes, if you have a 2.5% WR (25k on 1MM), you should never run out of money (or rather, you wouldn't have, historically, in the U.S. - you wouldn't have fared so well in other countries).  You should be left with a lot of money.

[Celda]

Ah yes, I see my mistake now, somehow I thought $25,000 was 4% of 1 mil.

However, when I do the same settings, except I do "% of portfolio" for spending at 4%, with the "Floor" as Never Less than 4% of original inflation adjusted portfolio, and the ceiling as never more than 5% of original inflation adjusted portfolio - I am still getting a 100% success rate.

Can that be right?

[brewer12345]

Allow me to whiz in someone's cornflakes:

For the vast majority of the population of would be ERs (not the heavily indebted general population) a 50 or 60 year survivable withdrawal rate is academic.  It is far more likely that you will have a fatal accident, contract a dread disease, commit suicide, live in a country that goes through a catastrophic event (losing a war, economic collapse, etc.) than it is that you will actually be around for the last 20 years of such a long draw period.  Who cares what the 60 year rate is?

[arebelspy]

Allow me to whiz in someone's cornflakes:

For the vast majority of the population of would be ERs (not the heavily indebted general population) a 50 or 60 year survivable withdrawal rate is academic.  It is far more likely that you will have a fatal accident, contract a dread disease, commit suicide, live in a country that goes through a catastrophic event (losing a war, economic collapse, etc.) than it is that you will actually be around for the last 20 years of such a long draw period.  Who cares what the 60 year rate is?

(Emphasis added.)

Hmm, what makes you say that?  The actuarial tables disagree.

If my wife and I retire at age 30, the odds that we're still around 50 years later is 57% for me, 66% for her, and 85% chance that at least one of us will be around.

https://personal.vanguard.com/us/insights/retirement/plan-for-a-long-retirement-tool

For 60 years it's 18/29/42.  So it looks slightly more likely that both of us will be dead than one of us alive after a 60-year time frame, and you can tweak that percentage based on what you think the odds of an economic collapse in the next 60 years are.

But I don't see where far more likely comes from.. just "slightly more likely" you'll be dead than survive a 60-year ER period, and for a 50 year ER period it's more likely that you will survive than die.

(And that's not taking into account if you have more potential longevity due to fitness, genes, etc., because we're talking average, obviously - though I'd bet the average ER is in better shape in those departments than the average person.)

[Nords]

The issue is that the typical stock-market history (since 1871 or so) doesn't have enough 50-year periods to make a statistically significant historical simulation of retirements, so Monte Carlo is used instead.

I used to believe that as well (based on the fact that I've heard it repeated so much, and indeed, have parroted it myself), but someone pointed out to me a few months ago that the 30-year simulations have a total of 115 runs, while the 50-year have a total of 95 cycles.  That's not THAT different, and if you think you can get statistically significant results from 115 runs, I don't see why you can't from 95.

In other words, I think FIRECalc/cFIREsim has limitations, but there aren't really any more limitations when running a 50-year retirement than when running a 30-year one - the limitations remain the same.

I don't think we get very good results from 115 runs, either, but... history is apparently perceived to be more credible than made-up synthetic market-performance parameters. 

My point is that somewhere between "30-year withdrawals" and "50-year withdrawals" we cross the line from "statistically marginal but useful results" to "really crappy and misleading results".  Neither one is particularly good for retirement forecasting, but the 50-year one is awful.  Ironically if a portfolio survives for 30 years of withdrawals then I bet the probability of it surviving for 50 years is very high.

I spend too much time talking people down from the ledge of "only 95% success" in their retirement calculations.  It's not a very good opportunity to point out that any success rate over 80% is meaningless.

Maybe if someone is hard over for a 100% success rate on a 50-year withdrawal then they should stick to living off their dividends.  As Brewer points out, they probably won't outlive their assets either way.

[arebelspy]

My point is that somewhere between "30-year withdrawals" and "50-year withdrawals" we cross the line from "statistically marginal but useful results" to "really crappy and misleading results".  Neither one is particularly good for retirement forecasting, but the 50-year one is awful.

I agree with the bolded part, but don't see the difference then between 30 and 50 year runs (an extra 20% data, from 95 to 115.. somewhat helpful, but they both have major limitations that one just has to accept for that model).

I spend too much time talking people down from the ledge of "only 95% success" in their retirement calculations.  It's not a very good opportunity to point out that any success rate over 80% is meaningless.

Maybe if someone is hard over for a 100% success rate on a 50-year withdrawal then they should stick to living off their dividends.  As Brewer points out, they probably won't outlive their assets either way.

Agreed on both.

[brewer12345]

Allow me to whiz in someone's cornflakes:

For the vast majority of the population of would be ERs (not the heavily indebted general population) a 50 or 60 year survivable withdrawal rate is academic.  It is far more likely that you will have a fatal accident, contract a dread disease, commit suicide, live in a country that goes through a catastrophic event (losing a war, economic collapse, etc.) than it is that you will actually be around for the last 20 years of such a long draw period.  Who cares what the 60 year rate is?

(Emphasis added.)

Hmm, what makes you say that?  The actuarial tables disagree.

If my wife and I retire at age 30, the odds that we're still around 50 years later is 57% for me, 66% for her, and 85% chance that at least one of us will be around.

https://personal.vanguard.com/us/insights/retirement/plan-for-a-long-retirement-tool

For 60 years it's 18/29/42.  So it looks slightly more likely that both of us will be dead than one of us alive after a 60-year time frame, and you can tweak that percentage based on what you think the odds of an economic collapse in the next 60 years are.

But I don't see where far more likely comes from.. just "slightly more likely" you'll be dead than survive a 60-year ER period, and for a 50 year ER period it's more likely that you will survive than die.

(And that's not taking into account if you have more potential longevity due to fitness, genes, etc., because we're talking average, obviously - though I'd bet the average ER is in better shape in those departments than the average person.)

There are very, very few ERs starting at 30.  The vast majority will be 45 or older, despite all of the people here who get worked up about bare bones retirements with skinny capital bases.

[arebelspy]

There are very, very few ERs starting at 30.  The vast majority will be 45 or older, despite all of the people here who get worked up about bare bones retirements with skinny capital bases.

Fair enough.

I don't think of my retirement as "bare bones" (~60-70k net rental income for ~40k spending, with a NW over > 1MM), but to many it would be considered so.

You're right though, two 45 year olds have about a 1% chance of one of them making it 60 years, according to the Vanguard calculator above.

[brewer12345]

There are very, very few ERs starting at 30.  The vast majority will be 45 or older, despite all of the people here who get worked up about bare bones retirements with skinny capital bases.

Fair enough.

I don't think of my retirement as "bare bones" (~60-70k net rental income for ~40k spending, with a NW over > 1MM), but to many it would be considered so.

You're right though, two 45 year olds have about a 1% chance of one of them making it 60 years, according to the Vanguard calculator above.

I would suggest that your plan is over bare bones.  I am thinking more of the people who decide that their rock bottom expenses as healthy 25 year olds are $16k and decide they can bail forever with 400k or less.

Let's put it this way: I will ensure pension and SS income for life for DW who will likely outlive me, but at age 40 I am not worried about making sure my assets last 60 years.

[lauren_knows]

why are there so few period (both 30y/50y)? Does cFireSim always just take period starting from January?
I'm asking this because if you take in account all period of 600 months (50 years x 12) you have way more period than 95.

....

I was just wondering, maybe someone already dismissed this approach because the market curve are too similar? Although the may vary greatly form month to month?

Well, I do have all the monthly data, but the problem is basically efficiency of a web-application.  When cFIREsim does a simulation, it's calculating and storing into the web browsers memory, 30yrs * 115 runs  * 10 data fields= 34500 data points in @1-3seconds.  To do monthy simulations for a 30yr period (360 months), it'd be 143 years of data, or 1716 months.  So, 360months * 1356 runs * 10 data fields = 4.8 million data points.... which as far as I can tell, can't be performed in the time that a browser usually times out (30 seconds).   I've run proof-of-concepts on the idea, and the browser runs out of memory trying to do this calculation.

It's possible that I could do the calculation on the server, and have the user periodically check on it's status, but I'm guessing on cFIREsim's current "shared hosting" server, those simulation runs would take a lonnnnng time.

[warfreak2]

Neither one is particularly good for retirement forecasting, but the 50-year one is awful.

I agree with the bolded part, but don't see the difference then between 30 and 50 year runs (an extra 20% data, from 95 to 115.. somewhat helpful, but they both have major limitations that one just has to accept for that model).

A big difference is the 50-year runs overlap a lot more. You have about three independent 50-year runs, or about five independent 30-year runs, so it's more like 67% more data.

[EscapeVelocity2020]

Interesting thread.  I know of two 'real world' 30+ year retirements, The Kaderlis (http://www.retireearlylifestyle.com/profile.htm) and Terhorsts (https://sites.google.com/site/paulvicgroup/).  They aren't living like rockstars, all of the advertising on the homepage links notwithstanding, but it  certainly was possible (if you retired into a good market in the 80's and were flexible about living abroad).

I'd especially recommend reading the Therhorst interview: http://www.retireearlylifestyle.com/paul_vicki_interview.htm 

A couple key excerpts:

How did you handle the Bear Market of 2000-2002?
We cried a lot. All three years. We even extrapolated and figured if it got really bad we've have to move to Thailand full time. We love Thailand, so that actually cheered us up quite a bit. But the market turned around in late 2002, and by 2004 we'd recovered all our losses. We take comfort from having survived a terrible crash and come out whole.

From your treasury of experiences, what would your main piece of advice be to those considering Early Retirement these days?
Be courageous. Leap. As Argentines say, "you can't take away the dance that's already been danced." Especially if you want to travel--if you want to enjoy the whole world rather than just a tiny corner of it--early retirement makes so much sense.

[Grog]

Well, I do have all the monthly data, but the problem is basically efficiency of a web-application.  When cFIREsim does a simulation, it's calculating and storing into the web browsers memory, 30yrs * 115 runs  * 10 data fields= 34500 data points in @1-3seconds.  To do monthy simulations for a 30yr period (360 months), it'd be 143 years of data, or 1716 months.  So, 360months * 1356 runs * 10 data fields = 4.8 million data points.... which as far as I can tell, can't be performed in the time that a browser usually times out (30 seconds).   I've run proof-of-concepts on the idea, and the browser runs out of memory trying to do this calculation.

It's possible that I could do the calculation on the server, and have the user periodically check on it's status, but I'm guessing on cFIREsim's current "shared hosting" server, those simulation runs would take a lonnnnng time.

Cool thanks for the reply. I see the problem. Maybe you could test with every 6 months interval (double the period) and see if really makes a difference, if not is not worth it. Tks for the great job!

[Crushtheturtle]

Uh... So I have to ask: Am I correct to assume that the consensus view is not consistent with one world-famous Mr. Money Mustache and his blog post wherein he states

"This brings me to a critical point: this study defines “success” as not going broke during a 30-year test period. To people like you and me who will enjoy 60-year retirements, that would not be successful – we want our money to last much longer than 30 years. Luckily, the math in this case is pretty interesting: there is very little difference between a 30-year period, and an infinite year period, when determining how long your money will last. It’s much like a 30-year mortgage, where almost all of your payment is interest. Drop your payment by just $199 per month, and suddenly you’ve got a thousand-year mortgage that will literally take you 1000 years to pay off. Increase the payment by a few hundred, and you have a fifteen year payoff! In other words, above 30 years, the length of your retirement barely affects the safe withdrawal rate calculations."

http://www.mrmoneymustache.com/2012/05/29/how-much-do-i-need-for-retirement/

[matchewed]

Uh... So I have to ask: Am I correct to assume that the consensus view is not consistent with one world-famous Mr. Money Mustache and his blog post wherein he states

"This brings me to a critical point: this study defines “success” as not going broke during a 30-year test period. To people like you and me who will enjoy 60-year retirements, that would not be successful – we want our money to last much longer than 30 years. Luckily, the math in this case is pretty interesting: there is very little difference between a 30-year period, and an infinite year period, when determining how long your money will last. It’s much like a 30-year mortgage, where almost all of your payment is interest. Drop your payment by just $199 per month, and suddenly you’ve got a thousand-year mortgage that will literally take you 1000 years to pay off. Increase the payment by a few hundred, and you have a fifteen year payoff! In other words, above 30 years, the length of your retirement barely affects the safe withdrawal rate calculations."

http://www.mrmoneymustache.com/2012/05/29/how-much-do-i-need-for-retirement/

And that's why taking one paragraph out of a blog post and isolating it from the rest of the context doesn't help. He goes on to say that... well I'll just post his words.

The trinity study assumes a retiree will:

    never earn any more money through part-time work or self-employment projects

    never collect a single dollar from social security or any other pension plan

    never adjust spending to account for economic reality like a huge recession

    never substitute goods to compensate for inflation or price fluctuation (vacation in a closer place one year during  an oil price spike, or switch to almond milk in the event of a dairy milk embargo).

    never collect any inheritance from the passing of parents or other family members

    and never do what most old people tend to do according to studies – spend less as they age

In short, they are assuming a bunch of drooling Complete Antimustachians. You and I are Mustachians, meaning we have far more flexibility in our lifestyles. In short, we have designed a Safety Margin into our lives that is wider than the average person’s entire retirement plan.

So actually I think the consensus (if there is actually one) is quite consistent. We all pick apart the SWR rules of thumb to understand it better, test it to the limits so to speak. It doesn't mean that we'll blindly follow it as blindly following it is dumb. One must adjust for events in their lives and the external forces which impose upon our best laid plans.

[arebelspy]

Uh... So I have to ask: Am I correct to assume that the consensus view is not consistent with one world-famous Mr. Money Mustache and his blog post wherein he states

"This brings me to a critical point: this study defines “success” as not going broke during a 30-year test period. To people like you and me who will enjoy 60-year retirements, that would not be successful – we want our money to last much longer than 30 years. Luckily, the math in this case is pretty interesting: there is very little difference between a 30-year period, and an infinite year period, when determining how long your money will last. It’s much like a 30-year mortgage, where almost all of your payment is interest. Drop your payment by just $199 per month, and suddenly you’ve got a thousand-year mortgage that will literally take you 1000 years to pay off. Increase the payment by a few hundred, and you have a fifteen year payoff! In other words, above 30 years, the length of your retirement barely affects the safe withdrawal rate calculations."

http://www.mrmoneymustache.com/2012/05/29/how-much-do-i-need-for-retirement/

Ah, but here's the rub: it's a fine line.  Decrease your mortgage payment by a few hundred from a 30-year and it takes 1000 years.  Increase it by a few hundred and it takes half the time, or 15.

Same with SWR.  4% is approximately safe.  Decrease it by a single percent (to say 3%) and it's pretty rock solid.  Increase it by a single percent (to 5%) and you're adding tons of risk.

The 30 year versus 50 year is along that "fine line" of tipping the balance too far one way versus the other could leave you with tens of millions, or nothing.

[MDM]

Does anyone know the actual principal and interest used to calculate the following?  I think that was asked at some point but don't recall the answer.

It’s much like a 30-year mortgage, where almost all of your payment is interest. Drop your payment by just $199 per month, and suddenly you’ve got a thousand-year mortgage that will literally take you 1000 years to pay off.

Table below shows interest and principal pairs for which the difference between 30-year and 1000-year monthly payments is $199. 

Code: [Select]

Rate/yr Principal
0% $73,900
1% $83,500
2% $98,100
3% $116,000
4% $138,100
5% $165,600
10% $449,800


[Tyler]

Still waiting for cfiresim to add those return numbers from 2014-2050!

Me too. I'd also like a check box option to adjust my projected spending according to all future tax law changes.   ;)

The key to retirement success is not a perfect cfiresim score before you start, but the skill and confidence to safely adjust course as you go.

[bwall]

Yes, if you have a 2.5% WR (25k on 1MM), you should never run out of money (or rather, you wouldn't have, historically, in the U.S. - you wouldn't have fared so well in other countries).

I'm interested in the quote: "or rather, you wouldn't have, historically, in the U.S. - you wouldn't have fared so well in other countries".

Where can I find more information about the comparison of investments in the USA vs. investments in other countries? Obviously, the first half of the 20th C. was a bad time for lots of western European countries and the second really half sucked for eastern European savers and investors. Can you point me in the right direction to find out more about this statement? For example, would this also hold true for Canada and Australia? Thanks!

[EscapeVelocity2020]

This is right up my alley.  Between re-reading 'A Tale of Two Cities' and re-watching Les Miserables (2012), I feel like this forum and MMM in general are on to something with feeling imbued with optimism in these heady times (we do live in America, and there are many examples of countries that haven't matched our GDP growth), but you are right that ridiculously long retirement and wealth has little modern precedent.  Being a wealthy landowner nowadays does not seem to be a secure 60 year FIRE strategy (ARS may disagree, but Trump declared bankruptcy and Buffet didn't).  And it's certainly not like we are at the point where Buffet likes to jump in, when 'there is blood is in the streets' (http://www.investopedia.com/articles/financial-theory/08/contrarian-investing.asp)

If you want the dim view, go to:
http://www.early-retirement.org/forums/f29/er-on-hold-47618.html
http://www.bogleheads.org/forum/viewtopic.php?t=11742

But making it up as you go along isn't as ridiculous as it was before the internet, where knowledge and access to improvement is quite accessible and powerful.  Free time is certainly much more valuable. 

Who knows what existence will look like in 50-60 years, I appreciate the idea that we worry far too much about what will happen, but I am a skeptic and an academic by nature (which served me well in 2008, and is serving me well now - ERE, where have you gone?).  I question MMM's motives in leading America to his Badass Utopia.  Maybe he has good intentions, but he has also profited handsomely off of his relatively thin internet persona, so be sure you are not contributing to his FI at the expense of your own.  Not everyone is going to reach the MMM / JD Roth nirvana of not having to worry about money for the rest of their life.  ERE had it in his posts, MMM seems to have more in the comments sections, as I noted in many of my own rejected comments.  It would be an interesting day to see what would happen if all the comments were automatically accepted... 

Many leaders have sold similar promises (Branch Dividians, Amish (measles outbreak in OH is really sad), Madhoff, Kozlowski, Lehman Bros, Enron, Long Term Capital, ...).  At the end of the day, you are given their content and shown what they accept as 'questioning' and answers.  If they have not answered all of your concerns, then by all means, keep searching I am blogging to try to bridge my perceived gap (http://escapevelocity2020.com/).  If you are drinking the kool-aid, then be happy.  If you are commenting, you are probably going to be OK. 

Going back to work in your 40's or 50's isn't the end of the world, but MMM is pretty well set.  Just don't think you're following his original footsteps, his posts no longer apply to himself, and he already said so (http://www.mrmoneymustache.com/2014/05/29/give-yourself-the-gift-of-not-worrying-about-money/comment-page-2/#comments)  He is going to make seven figures if he sells his blog, or more just by sitting on it. 

[arebelspy]

Yes, if you have a 2.5% WR (25k on 1MM), you should never run out of money (or rather, you wouldn't have, historically, in the U.S. - you wouldn't have fared so well in other countries).

I'm interested in the quote: "or rather, you wouldn't have, historically, in the U.S. - you wouldn't have fared so well in other countries".

Where can I find more information about the comparison of investments in the USA vs. investments in other countries? Obviously, the first half of the 20th C. was a bad time for lots of western European countries and the second really half sucked for eastern European savers and investors. Can you point me in the right direction to find out more about this statement? For example, would this also hold true for Canada and Australia? Thanks!

Here you go:
http://wpfau.blogspot.com/2012/05/may-i-add-part-vi-to-retirement.html
http://wpfau.blogspot.com/2011/10/cliff-notes-for-international.html

4% holds for US and Canada.

US SAFEMAX for 1900-1981 was 3.96%
Canada SAFEMAX for 1900-1981 was 3.96%
Australia SAFEMAX for 1900-1981 was 3.5%

Historically withdraw that retiring any year from 1900-1981 (multiple World Wars, etc.) in those countries and you wouldn't have run out of money after 30 years.

Hope that helps.  :)

[arebelspy]

[A mess of stuff]

[EscapeVelocity2020]

Put your last two responses together and what do you get?

 

Very, very deep.  This is the kind of high level intelligence that keeps humans one step ahead of AI :)

[Another Reader]

Donald Trump is not as much an investor as he is a speculator and he's had his share of "margin calls."  He lives off his reputation, not his long-term investing skill, IMO.

I retired in early 2007, just before the SHTF in real estate and a year and a half before it did in the stock market.  Did I have to go back to work?  Heck, no.  For the most part, my tenants continued to pay the rent and the dividend checks still arrived and were reinvested (except those darn banks, but that's what diversification is for).  Yeah, I took huge haircuts in portfolio value, and some properties are still years away from recovering peak value.  However, my income is significantly HIGHER than it was six years ago, because over time rents have gone up, I bought a number of properties on sale and I trimmed a couple of losers.  And the dividends and the paper portfolio value have recovered plus I have added to the portfolio.

I agree we have no idea what existence will be like in 60 years and the 4 percent safe withdrawal rate IMO is bogus.  Past performance is not indicative of future performance, yet we model future performance on the past to determine how safe a particular withdrawal rate is.  That's a little schizo in my book.  The US is likely no longer going to produce double digit economic growth and the risk in investing in the economies that are capable of that is a lot higher.  That alone could torpedo the 4 percent SWR.  Forecasting this stuff to figure out what I can spend is an ulcer inducing activity I prefer to avoid anyway.

So I refuse to be a hand wringing worrier, like a lot of folks over at early-retirement.org are.  That mindset is a good indication you really don't have "enough."  I think MMM, who is probably is in the $2.5 to $3M net worth range without the blog value per my guesstimate in another thread, has now reached the point where he really doesn't ever have to worry about money again.  Hence the recent blog post on that topic.  And I think he's correct.  Once you have reached the point where you do not have to decumulate to fund your life with a decent margin on top, why would you worry about money?

You folks can keep projecting the future and playing with withdrawal rates if you want to.  For me, I will stick to my diversified rental portfolio and dividend-tilted stock portfolio that has served me well so far, and I will remain both vigilant and flexible so I can deal with the future challenges, whatever they may be.

[arebelspy]

Put your last two responses together and what do you get?

...

Very, very deep.  This is the kind of high level intelligence that keeps humans one step ahead of AI :)

(Scrolling, scrolling) I don't see how those are are my last two responses.  If you mean "here are two posts, arebelspy has made, ever" then... sure?  Out of my 9,500 posts, there's probably a few dozen animated gifs, which I feel pretty clearly express what I wanted them to. 

/eyeroll

[arebelspy]

So I refuse to be a hand wringing worrier, like a lot of folks over at early-retirement.org are.  That mindset is a good indication you really don't have "enough."  I think MMM, who is probably is in the $2.5 to $3M net worth range without the blog value per my guesstimate in another thread, has now reached the point where he really doesn't ever have to worry about money again.  Hence the recent blog post on that topic.  And I think he's correct.  Once you have reached the point where you do not have to decumulate to fund your life with a decent margin on top, why would you worry about money?

You folks can keep projecting the future and playing with withdrawal rates if you want to.  For me, I will stick to my diversified rental portfolio and dividend-tilted stock portfolio that has served me well so far, and I will remain both vigilant and flexible so I can deal with the future challenges, whatever they may be.

Well said.

[EscapeVelocity2020]

So that it's not a [mess of stuff]

1.  It's amazing how far humanity has come from the French (and American) Revolution to worrying about if a 60 year retirement is sustainable.
2.  If your funds get low, you scale back on spending to water and rice, or get a job.
3.  The wealth of real life examples are a whole lot more informative than academic papers, cFIRE simulations, and a MMM blog post.
4.  Who knows what life will be like in 50-60 years, AI is already passing the Turing test. (http://www.independent.co.uk/life-style/gadgets-and-tech/computer-becomes-first-to-pass-turing-test-in-artificial-intelligence-milestone-but-academics-warn-of-dangerous-future-9508370.html)

Thanks Another Reader for a helpful comment.  Sorry ARS, I just thought those two gifs next to each other (your last two replies TO ME) were funny.

[arebelspy]

So that it's not a [mess of stuff]

1.  It's amazing how far humanity has come from the French (and American) Revolution to worrying about if a 60 year retirement is sustainable.
2.  If your funds get low, you scale back on spending to water and rice, or get a job.
3.  The wealth of real life examples are a whole lot more informative than academic papers, cFIRE simulations, and a MMM blog post.
4.  Who knows what life will be like in 50-60 years, AI is already passing the Turing test. (http://www.independent.co.uk/life-style/gadgets-and-tech/computer-becomes-first-to-pass-turing-test-in-artificial-intelligence-milestone-but-academics-warn-of-dangerous-future-9508370.html)

Thanks Another Reader for a helpful comment.  Sorry ARS, I just thought those two gifs next to each other (your last two replies TO ME) were funny.

That's funny, I disagreed with most of the mess of stuff, but I agree with most of that.  Once you cut out the negativity and doom and gloom, it becomes quite reasonable.

(Although - side note - the Turing test wasn't really passed, it's just click bait that it was.  No one was able to tell a man versus AI as a woman though the computer.)

Interesting that your posts inspire that reaction (expressive animated gifs, rather than substantive posts) from me.

[dragoncar]

Put your last two responses together and what do you get?

...

Very, very deep.  This is the kind of high level intelligence that keeps humans one step ahead of AI :)

(Scrolling, scrolling) I don't see how those are are my last two responses.  If you mean "here are two posts, arebelspy has made, ever" then... sure?  Out of my 9,500 posts, there's probably a few dozen animated gifs, which I feel pretty clearly express what I wanted them to. 

/eyeroll

Haha, I actually thought recently "Hmm... Rebs must have recently found a repository of reaction GIFs!"

P.S.  "Rebs"... approved?

[arebelspy]

Haha, I actually thought recently "Hmm... Rebs must have recently found a repository of reaction GIFs!"

Yeah, it's definitely a recent thing, but I was just confused because it wasn't my last two posts - I see what he means now.

With the sheer volume I post, it's still a small percent, I think, even of just recent posts, but lmk if it's getting annoying.  :)

/REBS OUT

[dragoncar]

Haha, I actually thought recently "Hmm... Rebs must have recently found a repository of reaction GIFs!"

Yeah, it's definitely a recent thing, but I was just confused because it wasn't my last two posts - I see what he means now.

With the sheer volume I post, it's still a small percent, I think, even of just recent posts, but lmk if it's getting annoying.  :)

/REBS OUT

They are probably only annoying to those on dial-up, but I'd suggest that those people use a plugin to optimize their browsing experience (for example, not loading large images unless requested).

[brooklynguy]

I question MMM's motives in leading America to his Badass Utopia.  Maybe he has good intentions, but he has also profited handsomely off of his relatively thin internet persona, so be sure you are not contributing to his FI at the expense of your own.  Not everyone is going to reach the MMM / JD Roth nirvana of not having to worry about money for the rest of their life.  ERE had it in his posts, MMM seems to have more in the comments sections, as I noted in many of my own rejected comments.  It would be an interesting day to see what would happen if all the comments were automatically accepted... 

EscapeVelocity, I'm curious what you meant by this.  Like ARS, I agreed with most of the restated and condensed version of your post.  But I'm not sure what you were getting at with the quoted portion of your initial post.  I think you may have meant something along the lines of the following: "MMM (perhaps only subconsciously) doesn't really have enough confidence in the 4% rule plus his built-in safety margins to enjoy a worry-free lifetime of retirement, so part of the reason he built the MMM enterprise is to increase the size of his stash and thereby reduce his withdrawal rate to a level that can be considered safe beyond the wildest stretches of imagination, which is a luxury that most of MMM's disciples cannot afford."  Is that correct?

[tooqk4u22]

Given where the US and global economy stands from a low growth and high debt perspective, low interest rates, and relatively high equity values I am more comfortable with a 3% SWR if I were retiring today (or at a time with similar characteristics).

Then as time goes by and the above normalizes in a methodical way (not to mention the simple fact of getting older and not needing the funds as long) then increasing to 4% makes more sense to me. 

And if the above doesn't normalize methodically and markets sink and my portfolio declines by 25% then I am back to a 4% SWR and all good.

Just my opinion.