Author Topic: cFiresim SEVERELY overestimates success rates for Mustachians  (Read 75922 times)

gerardc

  • Pencil Stache
  • ****
  • Posts: 767
  • Age: 40
  • Location: SF bay area
cFiresim SEVERELY overestimates success rates for Mustachians
« on: September 19, 2016, 08:33:49 PM »
The probabilistic problem solved by cFiresim/FIRECalc has the following form: You start with $1m at a random year, you spend $40k yearly from then on, what's the probability of success over 50 years?

The actual plan people follow is more like: You start at 0, you save $50k yearly, when you reach $1m you stop working, then you start spending $40k yearly, what's your probability of success over 50 years?

The 2 formulations look very similar, but the probabilities are very different. Can you see why?

I calculated that Plan A has 98% success rate while Plan B only has 40%!! (Python yahoo-finance package)

The reason is that under Plan B, the year you stop working is not actually uniformly distributed -- you're much more likely to retire on a peak year, and unlikely to retire just after a crash. This makes a big difference, because cFiresim assumes you could have retired on $1m in 2009, but that wouldn't happen on Plan B because you'd have had $2m in 2007 and would have retired then. On Plan B, you're more likely to have only $500k in 2009.

A better approach to increase your probability of success is Plan C: same as Plan B, but you're only allow to retire when your target has been reached for 5 years in a row. Success rate is back at 90% but you just lost 5 years... So, please be careful.

Update with analysis and figure
« Last Edit: July 15, 2017, 03:17:56 AM by gerardc »

JLee

  • Walrus Stache
  • *******
  • Posts: 7621
Re: cFiresim SEVERELY overestimates success rates for Mustachians
« Reply #1 on: September 19, 2016, 08:41:37 PM »
The probabilistic problem solved by cFiresim/FIRECalc has the following form: You start with $1m at a random year, you spend $40k yearly from then on, what's the probability of success over 50 years?

The actual plan people follow is more like: You start at 0, you save $50k yearly, when you reach $1m you stop working, then you start spending $40k yearly, what's your probability of success over 50 years?

The 2 formulations look very similar, but the probabilities are very different. Can you see why?

I calculated that Plan A has 98% success rate while Plan B only has 40%!! (Python yahoo-finance package)

The reason is that under Plan B, the year you stop working is not actually uniformly distributed -- you're much more likely to retire on a peak year, and unlikely to retire just after a crash. This makes a big difference, because cFiresim assumes you could have retired on $1m in 2009, but that wouldn't happen on Plan B because you'd have had $2m in 2007 and would have retired then. On Plan B, you're more likely to have only $500k in 2009.

A better approach to increase your probability of success is Plan C: same as Plan B, but you're only allow to retire when your target has been reached for 5 years in a row. Success rate is back at 90% but you just lost 5 years... So, please be careful.

If you had $2mil in 2007, presumably you would've retired at $1mil well before 2007...?

Shane

  • Handlebar Stache
  • *****
  • Posts: 1665
  • Location: Midtown
Re: cFiresim SEVERELY overestimates success rates for Mustachians
« Reply #2 on: September 19, 2016, 09:10:54 PM »
In my opinion, cFIREsim underestimates actual success in FIRE.

cFIREsim assumes that you start withdrawing a certain percentage of your stash, and then every year, no matter what happens, you continue to increase the amount you withdraw based on inflation.

In reality, most people don't just robotically keep withdrawing the same amount and continuing to increase it for inflation no matter what happens in the markets. If a market crash happens, most rational people will adjust their spending accordingly. Some will get a job, others will earn money from a side business, rent out a room in their house or maybe just cut expenses.

gerardc

  • Pencil Stache
  • ****
  • Posts: 767
  • Age: 40
  • Location: SF bay area
Re: cFiresim SEVERELY overestimates success rates for Mustachians
« Reply #3 on: September 19, 2016, 09:49:11 PM »
I mean, I don't think anyone here will lay down and die on the street if the market crashes, so their "success" will always be 100%. Doesn't mean that their plan has worked though.
« Last Edit: September 19, 2016, 10:05:12 PM by gerardc »

aceyou

  • Handlebar Stache
  • *****
  • Posts: 1669
  • Age: 41
    • Life is Good - Aceyou's Journal
Re: cFiresim SEVERELY overestimates success rates for Mustachians
« Reply #4 on: September 19, 2016, 10:16:28 PM »
The probabilistic problem solved by cFiresim/FIRECalc has the following form: You start with $1m at a random year, you spend $40k yearly from then on, what's the probability of success over 50 years?

The actual plan people follow is more like: You start at 0, you save $50k yearly, when you reach $1m you stop working, then you start spending $40k yearly, what's your probability of success over 50 years?

The 2 formulations look very similar, but the probabilities are very different. Can you see why?

I calculated that Plan A has 98% success rate while Plan B only has 40%!! (Python yahoo-finance package)

The reason is that under Plan B, the year you stop working is not actually uniformly distributed -- you're much more likely to retire on a peak year, and unlikely to retire just after a crash. This makes a big difference, because cFiresim assumes you could have retired on $1m in 2009, but that wouldn't happen on Plan B because you'd have had $2m in 2007 and would have retired then. On Plan B, you're more likely to have only $500k in 2009.

A better approach to increase your probability of success is Plan C: same as Plan B, but you're only allow to retire when your target has been reached for 5 years in a row. Success rate is back at 90% but you just lost 5 years... So, please be careful.

This is a really solid post IMO.  I think you bring up some very valid and true points, and you wrote it out in a very easy to understand way, well done. 

I'm overall in the camp that says just go for it though.  You mentioned that cFiresim isn't reality because in reality we have a far higher than random probability of retiring at a peak.  That's a terrific observation, but I bet most of us have large safety margins built in already.  I've got a ton of them... 

- Inheritance.  Both my and my wife's parent's are likely to have a positive net worth at their death, and I will likely receive some of that money.  I don't count on it, but it's the most likely outcome. 
- SS.  Even as a 33 year old, I'm likely to get something in the line of social security.  Again, I don't count on it, but some amount of money will probably come my way from it. 
-  Pensions.  My wife and I will each have one. 
- Life Insurance.  My wife and I each have a 500,000 universal life policy on each other. 
- I'll have a paid off house that I could access income from in some way.  Heck, I could rent out rooms without even having to sell it if it really came to that. 
- Some things that I'll do will likely pay me at some point, even though earning money won't be why I do it. 

I'll likely be able to survive huge downturns with all those safety margins, even if I do FIRE right at the peak of a market before a downturn. 

Glenstache

  • Magnum Stache
  • ******
  • Posts: 3512
  • Age: 95
  • Location: Upper left corner
  • Plug pulled
Re: cFiresim SEVERELY overestimates success rates for Mustachians
« Reply #5 on: September 19, 2016, 10:56:58 PM »
OPs post is really solid. I wonder if the retirement year is speed to peaks or away from valleys. The peaks seem to be smooth slopes on the climb with cliffs on the backside as the markets panic. Based on that, I would expect a stronger aversion to RE in a year like 2008 than a bias towards 2007. Is there any data to test this?  While I ages that at some level, you have to go for it, it is a valid concern. It also dovetails with a post sol made a while back about parsing cfiresim percentages too closely.
 

msilenus

  • Pencil Stache
  • ****
  • Posts: 524
Re: cFiresim SEVERELY overestimates success rates for Mustachians
« Reply #6 on: September 19, 2016, 11:40:16 PM »
Right.  With PE this high, I'm looking at the lowest few curves when I run cFireSim.  80% doesn't mean 80%, and 100% doesn't even mean 100%... but 100% says your 'stache has withstood the worst history could have thrown at it in living memory, which is surely something.

At more typical valuations (if there is such a thing), I think the probabilities might be closer to real.  But, then, fewer people think they're ready to retire at more typical valuations. :D

beltim

  • Magnum Stache
  • ******
  • Posts: 2957
Re: cFiresim SEVERELY overestimates success rates for Mustachians
« Reply #7 on: September 20, 2016, 01:07:51 AM »
I calculated that Plan A has 98% success rate while Plan B only has 40%!! (Python yahoo-finance package)

The reason is that under Plan B, the year you stop working is not actually uniformly distributed -- you're much more likely to retire on a peak year, and unlikely to retire just after a crash.

This is a great post, but the difference between 98% and 40% seems a bit too high.  Can you share the distribution that you came up with?  98% suggests that only 2 years (3?) failed, and thus the 40% success rate suggests that 60% of all retirements occurred in those 2-3 years, which is obviously impossible.

dragoncar

  • Walrus Stache
  • *******
  • Posts: 9947
  • Registered member
Re: cFiresim SEVERELY overestimates success rates for Mustachians
« Reply #8 on: September 20, 2016, 01:23:34 AM »
I calculated that Plan A has 98% success rate while Plan B only has 40%!! (Python yahoo-finance package)

The reason is that under Plan B, the year you stop working is not actually uniformly distributed -- you're much more likely to retire on a peak year, and unlikely to retire just after a crash.

This is a great post, but the difference between 98% and 40% seems a bit too high.  Can you share the distribution that you came up with?  98% suggests that only 2 years (3?) failed, and thus the 40% success rate suggests that 60% of all retirements occurred in those 2-3 years, which is obviously impossible.

I'd guess the simulation has an assumption for savings rate which is much lower than typical here.  The higher your savings rate, the less market returns affect your retirement date (think ERE- at 90% SR, your money only grows for three years before you pull the plug).  Otoh at 15% savings rate, a bubble can add decades of savings value to your NW

But more to the point, at a low enough savings rate you will ONLY be able to retire at market peaks
« Last Edit: September 20, 2016, 01:25:13 AM by dragoncar »

tooqk4u22

  • Magnum Stache
  • ******
  • Posts: 2892
Re: cFiresim SEVERELY overestimates success rates for Mustachians
« Reply #9 on: September 20, 2016, 08:41:58 AM »
OP has a point in that psychology comes into play but this also another way of thinking about sequence of returns risk. 

The best time to retire is in a down market if you have your number after the downturn, but fear will come into play like OP says and most people will not FIRE then but should and anything further or more time will just add to the cushion. 

People who retire at the top when they hit their number is because it feels good, markets are performing well, economy is humming along, etc so you do it because you have your number and if it doesn't work out you simply go back because everything is awesome.  But then the rug gets pulled out.

People who want to FIRE now or near term should be thinking about this.....I don't know if it is a top or not but valuations are stretched, earnings are declining, and fed policies are manipulating markets for some time now.


Tjat

  • Pencil Stache
  • ****
  • Posts: 570
Re: cFiresim SEVERELY overestimates success rates for Mustachians
« Reply #10 on: September 20, 2016, 08:49:07 AM »
The other aspect that worries me about achieving "FIRE" is that the guidance here seems to be to just shoot for 25x of your expenses. What's more accurate is that you need 25x of your predicted expenses in retirement. Rosily subtracting out taxes and other job-related expenses really seems to ignore the fact that future health care costs (as one example) will substantially increase as you age. Maybe some people are okay with taking that chance and relying on the government to bail them out, but I really don't see how these people retiring with a 500K net worth will have what's considered successful retirements if we could look back in 50 years.


JLee

  • Walrus Stache
  • *******
  • Posts: 7621
Re: cFiresim SEVERELY overestimates success rates for Mustachians
« Reply #11 on: September 20, 2016, 09:01:04 AM »
The other aspect that worries me about achieving "FIRE" is that the guidance here seems to be to just shoot for 25x of your expenses. What's more accurate is that you need 25x of your predicted expenses in retirement. Rosily subtracting out taxes and other job-related expenses really seems to ignore the fact that future health care costs (as one example) will substantially increase as you age. Maybe some people are okay with taking that chance and relying on the government to bail them out, but I really don't see how these people retiring with a 500K net worth will have what's considered successful retirements if we could look back in 50 years.

From what I gather, the amount invested will often increase throughout retirement (something like a 50% chance of doubling your initial investment after 30 years, even with 4% WR?), and a lot of people will be doing side gigs / working part time, odd jobs, etc.

"What's considered successful retirement" from who's perspective...those who are free from the 9-5 for the rest of their lives, or those who claim they didn't really retire because they had some side income along the way?

boarder42

  • Walrus Stache
  • *******
  • Posts: 9332
Re: cFiresim SEVERELY overestimates success rates for Mustachians
« Reply #12 on: September 20, 2016, 09:32:32 AM »
Assuming your real question deals more with being confident in being able to FIRE which is the best use of the cFIREsim tool. then i would just use that as a guage for incorporating external income and limited time expenses such as mortgages and houses ... as far as SWR though

one of the best indicators of a good SWR when FIREd is the shiller PE

See
http://www.madfientist.com/safe-withdrawal-rate/

there is a link in here to the original work supplied by Kitces.

its been a good historical metric to gauge what your SWR will be at FIRE.  how do i plan to use it ... if i hit 4% SWR based on savings and spending i'll evaluate where the inverse of the shiller PE is.  if its really low say 1999 low at 2% i'll wait if its really high i'll FIRE away.  if its right on or in the ballpark, then its just a game of can you pull the plug.

for information right now its siting a 3.7% SWR. 
« Last Edit: September 20, 2016, 09:34:16 AM by boarder42 »

bacchi

  • Walrus Stache
  • *******
  • Posts: 7423
Re: cFiresim SEVERELY overestimates success rates for Mustachians
« Reply #13 on: September 20, 2016, 09:34:54 AM »
The SWR has a 5% failure rate at 75/25. What is the likelihood that one will ER in one of those failure years? Is it more likely than any other year?

Looking at cFireSim, we're talking failures in the late 1960s. Even retiring in 1986, a year before the 1987 crash, isn't a failure over 30 years. Neither is 1928, a year before the Great Depression crash. (That analysis also doesn't include any non-investment income in the future: no old age pension, no inheritance, no part-time work, no hobby income, no house sale windfall, no mortgage payoff.)

Ok, take 1964-1969, the Failure Years.  Either before or during those years, there wasn't an especially strong run up. Sure, 1963 had a 22% return but 1962 was -8.81%. 1966 and 1969 were down -10% and -8%, respectively. That really leaves only the 3 years of positive returns from 1963-1965. Why are these so different from any other 3+ year period of positive returns, such as 1949-1952? The answer, of course, is sequence of returns. The 1970s stagflation killed any retirement portfolio.

The lesson, therefore, isn't to not retire in a "peak" year. The lesson is to not retire before a high inflation+shitty economy decade.


undercover

  • Pencil Stache
  • ****
  • Posts: 997
Re: cFiresim SEVERELY overestimates success rates for Mustachians
« Reply #14 on: September 20, 2016, 09:37:24 AM »
Yes, this is mostly another way of illustrating sequence of returns risk.

My hunch is this: the shorter your investing timeline, the more likely you are to be prone to sequence of returns risk, and vice-versa. But even then, you aren't necessarily more likely to be at the top of the market, as there's no way to predict that. I agree that you're less likely to retire in a bear market, but that doesn't mean the opposite is true.

It's easy to cherry pick peaks and assume that one retires on the peaks and then see how likely those 30 yr periods are likely to fail due to sequence risk. It's very possible that your overall savings increases faster than market returns, and you pull the plug well before the peak so that you're less exposed to sequence risk.

While you're much more likely to hit "your number" in a bull market, that doesn't necessarily mean that it's a peak.

The other aspect that worries me about achieving "FIRE" is that the guidance here seems to be to just shoot for 25x of your expenses. What's more accurate is that you need 25x of your predicted expenses in retirement. Rosily subtracting out taxes and other job-related expenses really seems to ignore the fact that future health care costs (as one example) will substantially increase as you age. Maybe some people are okay with taking that chance and relying on the government to bail them out, but I really don't see how these people retiring with a 500K net worth will have what's considered successful retirements if we could look back in 50 years.

Most people's numbers do include taxes as well as a buffer for potential increases in spending. If you're retiring on $500k, it's unlikely you'll be paying any taxes at all. And in all likelihood, you will spend less and have an even larger buffer than you thought.

brooklynguy

  • Handlebar Stache
  • *****
  • Posts: 2204
  • Age: 44
Re: cFiresim SEVERELY overestimates success rates for Mustachians
« Reply #15 on: September 20, 2016, 11:00:52 AM »
The 2 formulations look very similar, but the probabilities are very different. Can you see why?

There has been much discussion in the forum about this issue, most prominently in this thread: "Firecalc and cFIREsim both lie"

Classical_Liberal

  • Handlebar Stache
  • *****
  • Posts: 1171
  • Age: 48
Re: cFiresim SEVERELY overestimates success rates for Mustachians
« Reply #16 on: September 20, 2016, 11:10:46 AM »
I also think OP has a very valid point, as do many of the other posters.  To solve this problem with the calculators, why not just run two scenarios. 

Scenario 1) "The savings years";  get an idea or range of how long it takes you to get to your number (ex 25x expenses) with desired AA.

Scenario 2) "The FIRE years"; take your number as current asset balance and run retirement with the number of years you estimate you'll live.

This would provide you with some red flags when the situation plays out in reality.  For example, if you hit "your number" on the sooner, rather than later dates.  Was it because you saved more/spent less?  or was it because returns have been very good in the savings years? 

I agree that the savings rate (ie time to FIRE) is a huge factor in all of this.  Also agree that good metrics like PE10 can be very helpful in choosing a SWR.  I would also argue that making intelligent Asset Allocation choices at different times during the lifecycle of a Stache (ie quick savings, slow savings, early drawdown, late drawdown) can be very helpful in minimizing sequence risk.

There has been much discussion in the forum about this issue, most prominently in this thread: "Firecalc and cFIREsim both lie"

Popped up while I was typing, thanks for the link!


slowsynapse

  • Stubble
  • **
  • Posts: 103
  • Age: 52
Re: cFiresim SEVERELY overestimates success rates for Mustachians
« Reply #17 on: September 20, 2016, 11:57:55 AM »


The reason is that under Plan B, the year you stop working is not actually uniformly distributed -- you're much more likely to retire on a peak year, and unlikely to retire just after a crash. This makes a big difference, because cFiresim assumes you could have retired on $1m in 2009, but that wouldn't happen on Plan B because you'd have had $2m in 2007 and would have retired then. On Plan B, you're more likely to have only $500k in 2009.



I also think that OP makes some really good points.  As a recent Re-Firee (is that the same as retiree), I went through many scenarios like this in my mind.  I highlighted the part about likelihood of retiring at a peak, not a crash because I think there is validity to this concern.  I worked for more than a year after reaching my FI number and I had quite a bit of cushion in my budget.  I can see where it might be a mistake to quit your job the day the market propels you to that magical FI number; particularly if you have either a lucrative career or one that you enjoy on some level.  I worked for about 13 months after, which was a little too long.  As others have mentioned here, market returns are just one element of the FIRE success.  It is also about cushion in the budget and what you think your ability to secure work again would be if you needed to.  Having some flexibility and back up plans is the biggest way you can control sequencing risk.

SnackDog

  • Handlebar Stache
  • *****
  • Posts: 1260
  • Location: Latin America
Re: cFiresim SEVERELY overestimates success rates for Mustachians
« Reply #18 on: September 20, 2016, 11:59:30 AM »
4% is just based on history. And remember it only applies for 30 years and includes inflation adjustments.  So, if we believe history, we can still retire on the highest of the high market peaks and still make it 30 years (not 50!).  In fact, we can even restart/reset our starting year each time our portfolio hits a new high and it will still last 30 forward years, according to the historic behavior.  If the market crashes the year AFTER we retire we can still continue our 4%+inflation withdrawals and we'll still make it 30 years, according to historic behavior.

IF we want additional insurance, we have many levers we can pull including those mentioned - trying to time the market for retirement year (easier to see market peaks than future stagflation), variable withdrawal rates according to market/economic behavior, variable withdrawal rates with more up front and less after age 75, a SWR less than 4%, a shorter retirement plan (less than 30 years), options to return to work, social security additions, deferred pension additions, property sales, reverse mortgages, and more.

What this group needs is some case studies from people who actually calculated all this, retired and ran out of money!

boarder42

  • Walrus Stache
  • *******
  • Posts: 9332
Re: cFiresim SEVERELY overestimates success rates for Mustachians
« Reply #19 on: September 20, 2016, 12:47:07 PM »
4% is just based on history. And remember it only applies for 30 years and includes inflation adjustments.  So, if we believe history, we can still retire on the highest of the high market peaks and still make it 30 years (not 50!).  In fact, we can even restart/reset our starting year each time our portfolio hits a new high and it will still last 30 forward years, according to the historic behavior.  If the market crashes the year AFTER we retire we can still continue our 4%+inflation withdrawals and we'll still make it 30 years, according to historic behavior.

IF we want additional insurance, we have many levers we can pull including those mentioned - trying to time the market for retirement year (easier to see market peaks than future stagflation), variable withdrawal rates according to market/economic behavior, variable withdrawal rates with more up front and less after age 75, a SWR less than 4%, a shorter retirement plan (less than 30 years), options to return to work, social security additions, deferred pension additions, property sales, reverse mortgages, and more.

What this group needs is some case studies from people who actually calculated all this, retired and ran out of money!

thats gonna be hard to come by

Classical_Liberal

  • Handlebar Stache
  • *****
  • Posts: 1171
  • Age: 48
Re: cFiresim SEVERELY overestimates success rates for Mustachians
« Reply #20 on: September 20, 2016, 01:15:54 PM »
4% is just based on history. And remember it only applies for 30 years and includes inflation adjustments.  So, if we believe history, we can still retire on the highest of the high market peaks and still make it 30 years (not 50!).  In fact, we can even restart/reset our starting year each time our portfolio hits a new high and it will still last 30 forward years, according to the historic behavior.  If the market crashes the year AFTER we retire we can still continue our 4%+inflation withdrawals and we'll still make it 30 years, according to historic behavior.

IF we want additional insurance, we have many levers we can pull including those mentioned - trying to time the market for retirement year (easier to see market peaks than future stagflation), variable withdrawal rates according to market/economic behavior, variable withdrawal rates with more up front and less after age 75, a SWR less than 4%, a shorter retirement plan (less than 30 years), options to return to work, social security additions, deferred pension additions, property sales, reverse mortgages, and more.

What this group needs is some case studies from people who actually calculated all this, retired and ran out of money!

thats gonna be hard to come by

Absolutely!  Anyone who has taken the time to consider the things being discussed on the threads in this forum is not going to allow themselves to run out of money. 

boarder42

  • Walrus Stache
  • *******
  • Posts: 9332
Re: cFiresim SEVERELY overestimates success rates for Mustachians
« Reply #21 on: September 20, 2016, 02:01:26 PM »
4% is just based on history. And remember it only applies for 30 years and includes inflation adjustments.  So, if we believe history, we can still retire on the highest of the high market peaks and still make it 30 years (not 50!).  In fact, we can even restart/reset our starting year each time our portfolio hits a new high and it will still last 30 forward years, according to the historic behavior.  If the market crashes the year AFTER we retire we can still continue our 4%+inflation withdrawals and we'll still make it 30 years, according to historic behavior.

IF we want additional insurance, we have many levers we can pull including those mentioned - trying to time the market for retirement year (easier to see market peaks than future stagflation), variable withdrawal rates according to market/economic behavior, variable withdrawal rates with more up front and less after age 75, a SWR less than 4%, a shorter retirement plan (less than 30 years), options to return to work, social security additions, deferred pension additions, property sales, reverse mortgages, and more.

What this group needs is some case studies from people who actually calculated all this, retired and ran out of money!

thats gonna be hard to come by

Absolutely!  Anyone who has taken the time to consider the things being discussed on the threads in this forum is not going to allow themselves to run out of money.

and there isnt a large enough sample ... i mean this question could be revisited in say 30-40 years to see how it effected people and 2 generations can benefit from our success and blunders.

Prairie Stash

  • Handlebar Stache
  • *****
  • Posts: 1795
Re: cFiresim SEVERELY overestimates success rates for Mustachians
« Reply #22 on: September 20, 2016, 02:10:14 PM »
CFiresim assumes you never work again. If I retire at 40 and 2 years later hit a crash, why can't I work again? 2 years of work and presto, retired again. So instead of working an extra 5 years there's a possibility of needing a job for a couple years, I'll take the possibility of never working over the guarantee of work.

Tyson

  • Magnum Stache
  • ******
  • Posts: 3216
  • Age: 52
  • Location: Denver, Colorado
Re: cFiresim SEVERELY overestimates success rates for Mustachians
« Reply #23 on: September 20, 2016, 02:51:09 PM »
Or just keep enough cash on hand to get you through the first year or 2 after a crash.

TheAnonOne

  • Handlebar Stache
  • *****
  • Posts: 1777
Re: cFiresim SEVERELY overestimates success rates for Mustachians
« Reply #24 on: September 20, 2016, 02:59:21 PM »
Or just keep enough cash on hand to get you through the first year or 2 after a crash.

This assumes you worked longer-than-needed to get this cash.

Classical_Liberal

  • Handlebar Stache
  • *****
  • Posts: 1171
  • Age: 48
Re: cFiresim SEVERELY overestimates success rates for Mustachians
« Reply #25 on: September 20, 2016, 03:41:33 PM »
4% is just based on history. And remember it only applies for 30 years and includes inflation adjustments.  So, if we believe history, we can still retire on the highest of the high market peaks and still make it 30 years (not 50!).  In fact, we can even restart/reset our starting year each time our portfolio hits a new high and it will still last 30 forward years, according to the historic behavior.  If the market crashes the year AFTER we retire we can still continue our 4%+inflation withdrawals and we'll still make it 30 years, according to historic behavior.

IF we want additional insurance, we have many levers we can pull including those mentioned - trying to time the market for retirement year (easier to see market peaks than future stagflation), variable withdrawal rates according to market/economic behavior, variable withdrawal rates with more up front and less after age 75, a SWR less than 4%, a shorter retirement plan (less than 30 years), options to return to work, social security additions, deferred pension additions, property sales, reverse mortgages, and more.

What this group needs is some case studies from people who actually calculated all this, retired and ran out of money!

thats gonna be hard to come by

Absolutely!  Anyone who has taken the time to consider the things being discussed on the threads in this forum is not going to allow themselves to run out of money.

and there isnt a large enough sample ... i mean this question could be revisited in say 30-40 years to see how it effected people and 2 generations can benefit from our success and blunders.

Even then, there is no guarantee the future will be like the past.  We would have more data, maybe even very different events to include in the data.  However, at the end of the day, there would only be more precision, not more accuracy.  The future is notoriously difficult to predict.  Until that changes, we'd do well to build our knowledge, tools, and willingness to adapt since these can help overcome even unpredictable events.

Or just keep enough cash on hand to get you through the first year or 2 after a crash.

This assumes you worked longer-than-needed to get this cash.

The idea of keeping a 2 year cash buffer for extraordinarily bad times was backtested for us here:

https://livingafi.com/2014/05/28/drawdown-part-4-examples/

Great blog, by the way, I love the job experience section.  It's reads like a great book!

Blueskies123

  • 5 O'Clock Shadow
  • *
  • Posts: 81
Re: cFiresim SEVERELY overestimates success rates for Mustachians
« Reply #26 on: September 20, 2016, 03:51:33 PM »
Or just keep enough cash on hand to get you through the first year or 2 after a crash.

This is why, at retirement, I put 8 years of base living expenses in treasures and short term bonds.  Eight years was the longest bear market I could find.  I still have 50% in stocks and 8 years in bonds.  I plan to live off my bonds which will slowly tilt my portfolio to 60% or 65% stocks.

MilesTeg

  • Handlebar Stache
  • *****
  • Posts: 1363
Re: cFiresim SEVERELY overestimates success rates for Mustachians
« Reply #27 on: September 20, 2016, 03:57:46 PM »
The idea of trying to plan for 30 or 40 years of retirement is, at best, a crap shoot. Life has a way of giving you a punch to the face every once in a while that you never see coming.

Classical_Liberal

  • Handlebar Stache
  • *****
  • Posts: 1171
  • Age: 48
Re: cFiresim SEVERELY overestimates success rates for Mustachians
« Reply #28 on: September 20, 2016, 04:46:24 PM »
Or just keep enough cash on hand to get you through the first year or 2 after a crash.

This is why, at retirement, I put 8 years of base living expenses in treasures and short term bonds.  Eight years was the longest bear market I could find.  I still have 50% in stocks and 8 years in bonds.  I plan to live off my bonds which will slowly tilt my portfolio to 60% or 65% stocks.

Have you played here?  https://portfoliocharts.com/calculators/

The types of stocks and bonds you choose matters, at least from a backtesting perspective.

Classical_Liberal

  • Handlebar Stache
  • *****
  • Posts: 1171
  • Age: 48
Re: cFiresim SEVERELY overestimates success rates for Mustachians
« Reply #29 on: September 20, 2016, 04:48:37 PM »
The idea of trying to plan for 30 or 40 years of retirement is, at best, a crap shoot. Life has a way of giving you a punch to the face every once in a while that you never see coming.

More like a game of blackjack... and you are the house.  Stick by house rules and the odds favor you over the long run.

SwordGuy

  • Walrus Stache
  • *******
  • Posts: 9044
  • Location: Fayetteville, NC
Re: cFiresim SEVERELY overestimates success rates for Mustachians
« Reply #30 on: September 20, 2016, 05:26:54 PM »
CFiresim assumes you never work again.

Not true.   There is an "Other Income" category.  You can put in income you might receive in the future, after you have FIREd.  I do that in all my planning sessions.  Just pay attention to the start/end dates.

For example, I use it for part time work, farm income and rental income.   

arebelspy

  • Administrator
  • Senior Mustachian
  • *****
  • Posts: 28467
  • Age: -997
  • Location: Seattle, WA
Re: cFiresim SEVERELY overestimates success rates for Mustachians
« Reply #31 on: September 20, 2016, 05:39:47 PM »
There has been much discussion in the forum about this issue, most prominently in this thread: "Firecalc and cFIREsim both lie"

Exactly what I thought of when I saw the thread title, and confirmed reading the OP.

OP, read that thread, you'll probably find it quite interesting. :)

IMO it's as valid point as one is inflexible.  The more inflexible one is, the more of a concern this is.  The more flexible, the less important it becomes.
I am a former teacher who accumulated a bunch of real estate, retired at 29, spent some time traveling the world full time and am now settled with three kids.
If you want to know more about me, this Business Insider profile tells the story pretty well.
I (rarely) blog at AdventuringAlong.com. Check out the Now page to see what I'm up to currently.

Tyson

  • Magnum Stache
  • ******
  • Posts: 3216
  • Age: 52
  • Location: Denver, Colorado
Re: cFiresim SEVERELY overestimates success rates for Mustachians
« Reply #32 on: September 20, 2016, 06:25:09 PM »
Or just keep enough cash on hand to get you through the first year or 2 after a crash.

This assumes you worked longer-than-needed to get this cash.

It's not working longer than needed if it is one of your hedges against the biggest cause of ER failure - poor returns during your first few years of retirement.

Blueskies123

  • 5 O'Clock Shadow
  • *
  • Posts: 81
Re: cFiresim SEVERELY overestimates success rates for Mustachians
« Reply #33 on: September 20, 2016, 06:32:28 PM »
The idea of trying to plan for 30 or 40 years of retirement is, at best, a crap shoot. Life has a way of giving you a punch to the face every once in a while that you never see coming.

This sounds a bit depressing.  A constructive approach would be look at history as a guide and build a conservative portfolio that should give you 30-50 years of income using history as a guide but also realizing that no can foresee the future so it would be wise to build plenty of cushion in your forecast.  One suggestion is to work a few extra years and bank as much as you can.  This way you can better off than 95% of the others that have not planned properly.

Shane

  • Handlebar Stache
  • *****
  • Posts: 1665
  • Location: Midtown
Re: cFiresim SEVERELY overestimates success rates for Mustachians
« Reply #34 on: September 20, 2016, 06:46:52 PM »
I mean, I don't think anyone here will lay down and die on the street if the market crashes, so their "success" will always be 100%. Doesn't mean that their plan has worked though.

If you haven't read it already, check out this post by Go Curry Cracker: "The Worst Retirement Ever." In the article, Jeremy details, step by step, simple changes that could've been made by a person who retired in 1965, the worst year ever, to easily salvage his retirement. Just because a retiree has to make a few changes to his plans doesn't mean that his plan has failed.


OurFirstFire

  • 5 O'Clock Shadow
  • *
  • Posts: 57
Re: cFiresim SEVERELY overestimates success rates for Mustachians
« Reply #35 on: September 20, 2016, 07:21:11 PM »
you're much more likely to retire on a peak year, and unlikely to retire just after a crash.

It's not obvious to me that this is true. 

Why would you be more likely to retire on a peak year than a positive year in the middle of a bull (or flatish) market?  I don't think you'd be any more likely to retire in 2007 than 1995 or 2013.

I also think there are plenty of people who would retire after a mild crash.  People who get laid off, hopefully with a severance, and decide it's the right time to retire?  Particularly if they're well diversified and not 100% stocks.

But overall yes, if you can only retire because the stock market went up 50% this year (Japan, China, or US 1929) then you should think twice.

sol

  • Walrus Stache
  • *******
  • Posts: 8435
  • Age: 47
  • Location: Pacific Northwest
Re: cFiresim SEVERELY overestimates success rates for Mustachians
« Reply #36 on: September 20, 2016, 07:45:15 PM »
There has been much discussion in the forum about this issue, most prominently in this thread: "Firecalc and cFIREsim both lie"

Exactly what I thought of when I saw the thread title, and confirmed reading the OP.

OP, read that thread, you'll probably find it quite interesting. :)

Everything old is new again. 

Wait around long enough, bell bottoms will be back in style and someone will re-ask every question this forum has already beaten to death.

Soon enough we should see new threads promoting dual momentum investment schemes, whether or not to invest a windfall immediately or to DCA into the market, whether or not a Mustachian can honestly invest in an exploitive capitalist system, and how to allocate your future charitable contributions.

Easy Does It FI

  • 5 O'Clock Shadow
  • *
  • Posts: 48
    • Easy Does It FI
Re: cFiresim SEVERELY overestimates success rates for Mustachians
« Reply #37 on: September 20, 2016, 07:51:21 PM »
The simulator could correct for this by doing backward-looking and pruning any runs (most likely success runs) where you hit FIRE multiple times in your savings phase.
You would need 1 additional assumption to let it know how far back to look/how high the old peak had to be:
1. You could provide a flat savings-per-month assumption. It could use historic return (before the trial's start date) to calculate how long you saved and how fast you approached FI.
2. You could tell it how long you saved to get to FI. And again- It could use the backward-looking return to calculate average amount contributed and speed of approach ("angle of attack").

This approach would eliminate trials where you somehow added principle despite a falling market.
It would also correctly consider the logarithmic effect that the closer to FI, the less important savings and the more important market return is.
*Apologies if this idea was already posted elsewhere.

dragoncar

  • Walrus Stache
  • *******
  • Posts: 9947
  • Registered member
Re: cFiresim SEVERELY overestimates success rates for Mustachians
« Reply #38 on: September 20, 2016, 08:26:40 PM »
There has been much discussion in the forum about this issue, most prominently in this thread: "Firecalc and cFIREsim both lie"

Exactly what I thought of when I saw the thread title, and confirmed reading the OP.

OP, read that thread, you'll probably find it quite interesting. :)

Everything old is new again. 

Wait around long enough, bell bottoms will be back in style and someone will re-ask every question this forum has already beaten to death.

Soon enough we should see new threads promoting dual momentum investment schemes, whether or not to invest a windfall immediately or to DCA into the market, whether or not a Mustachian can honestly invest in an exploitive capitalist system, and how to allocate your future charitable contributions.

Ok, but do we really need soap?

sol

  • Walrus Stache
  • *******
  • Posts: 8435
  • Age: 47
  • Location: Pacific Northwest
Re: cFiresim SEVERELY overestimates success rates for Mustachians
« Reply #39 on: September 20, 2016, 11:11:22 PM »
Networth dropped then came back and today is higher than when I first FIREd.

Sure, but that's entirely a product of the year you retired.  The market is up like 300% since 2009, one of the greatest bull markets in history.  OF COURSE the 4% rule worked out great for anyone who retired in the past decade (or three!), the market has been kind to them.  You're not the target audience.

The 4% rule was written for people who retire at the start of a decades-long bear market brought on by global war or crashing commodities or failed monetary policy.  It is those unlucky souls for whom a 4% SWR might fail to pan out.  For everyone who retires in a business-as-usual-year, the SWR is more like 6 or even 6.5% (the statistical average SWR at which your money survives a 30 year retirement starting in an average or better-than-average start year).

This is the greatest existential risk I think this community faces.  It's not like the 95% success rate of the 4% rule means that 95 out of 100 of us who retire in any given year are going to make it on 4%.  We're either ALL going to make it, in that year, or we're ALL not going to make it.  Totally depends on what the future economy looks like.  If things go south on us bad enough, every singe retiree here is going to fail the 4% rule all at once, and then threads like this one are going to seem pretty silly.  All those early retirement rules and forecasts, though mathematically unchanged, will suddenly seem like a farce.

Shane

  • Handlebar Stache
  • *****
  • Posts: 1665
  • Location: Midtown
Re: cFiresim SEVERELY overestimates success rates for Mustachians
« Reply #40 on: September 20, 2016, 11:13:32 PM »
In my opinion, cFIREsim underestimates actual success in FIRE.

cFIREsim assumes that you start withdrawing a certain percentage of your stash, and then every year, no matter what happens, you continue to increase the amount you withdraw based on inflation.

In reality, most people don't just robotically keep withdrawing the same amount and continuing to increase it for inflation no matter what happens in the markets. If a market crash happens, most rational people will adjust their spending accordingly. Some will get a job, others will earn money from a side business, rent out a room in their house or maybe just cut expenses.
^This. I retired shortly before 2007 and easily cut expenses for a few years without any problems.  Networth dropped then came back and today is higher than when I first FIREd. Having wiggle room in your budget (something most ERees have) or being flexible in your lifestyle for a few years to ride out a market downturn, even a big one, shouldn't be a problem.

^+1

I quit working 15 months ago.

Before leaving my job I ran lots of scenarios on cFIREsim, mostly using a 4% WR. Our actual spending in retirement has been ~2.5%.

Even though we've been spending $2K+/month for the past 15 months, due to market gains our stash has actually grown by about $30K since I quit working.

If there were a big market downturn, we could easily just cut back on spending, all the way to zero if we needed to, and I wouldn't see that as a "failure" of our plan, because my only plan is to be flexible.

arebelspy

  • Administrator
  • Senior Mustachian
  • *****
  • Posts: 28467
  • Age: -997
  • Location: Seattle, WA
Re: cFiresim SEVERELY overestimates success rates for Mustachians
« Reply #41 on: September 20, 2016, 11:51:16 PM »
This is the greatest existential risk I think this community faces.  It's not like the 95% success rate of the 4% rule means that 95 out of 100 of us who retire in any given year are going to make it on 4%.  We're either ALL going to make it, in that year, or we're ALL not going to make it.  Totally depends on what the future economy looks like.  If things go south on us bad enough, every singe retiree here is going to fail the 4% rule all at once, and then threads like this one are going to seem pretty silly.  All those early retirement rules and forecasts, though mathematically unchanged, will suddenly seem like a farce.

Only if we ALL rigidly adhere to a rule without changing our behavior at all in spite of poor market returns.  The ones who adjust and earn a little money in ER, or who cut their spending, likely will be just fine (and this is ignoring those who planned for a lower WR to begin with).

I'd bet that if a 4% WR "failed" over a 30 year period, much less than half of us would actually run out of money and end up with only social security income.
I am a former teacher who accumulated a bunch of real estate, retired at 29, spent some time traveling the world full time and am now settled with three kids.
If you want to know more about me, this Business Insider profile tells the story pretty well.
I (rarely) blog at AdventuringAlong.com. Check out the Now page to see what I'm up to currently.

Glenstache

  • Magnum Stache
  • ******
  • Posts: 3512
  • Age: 95
  • Location: Upper left corner
  • Plug pulled
Re: cFiresim SEVERELY overestimates success rates for Mustachians
« Reply #42 on: September 21, 2016, 12:14:41 AM »
This is the greatest existential risk I think this community faces.  It's not like the 95% success rate of the 4% rule means that 95 out of 100 of us who retire in any given year are going to make it on 4%.  We're either ALL going to make it, in that year, or we're ALL not going to make it.  Totally depends on what the future economy looks like.  If things go south on us bad enough, every singe retiree here is going to fail the 4% rule all at once, and then threads like this one are going to seem pretty silly.  All those early retirement rules and forecasts, though mathematically unchanged, will suddenly seem like a farce.

Only if we ALL rigidly adhere to a rule without changing our behavior at all in spite of poor market returns.  The ones who adjust and earn a little money in ER, or who cut their spending, likely will be just fine (and this is ignoring those who planned for a lower WR to begin with).

I'd bet that if a 4% WR "failed" over a 30 year period, much less than half of us would actually run out of money and end up with only social security income.
... Assuming ss survived the hypothetical doomsday market condition. But yes, the principle of constant optimization applies

Villanelle

  • Walrus Stache
  • *******
  • Posts: 7054
Re: cFiresim SEVERELY overestimates success rates for Mustachians
« Reply #43 on: September 21, 2016, 02:11:11 AM »
Interesting take.  I don't think DH and I are more likely to declare FIRE in an up market for psychology reasons, but it makes sense that an up market might be enough to push people's stache up to the FIRE target, thus increasing the likelihood of FIRE during an up year. 

That said, I'm not too worried, and I tend to be a worrier!  Our FIRE plans are far, far from barebones, thus giving us lots of room for adjustment.  During a bear market, we can simply withdraw 3 or 2.5%, and make that the year we visit several national parks instead of the year we spend 2 weeks in Scotland, and I could read books from my list that are available at the library only, waiting on those I'd need to pay for until things loosen up again.  Because that flexibility is part of our plan, it can hardly be called a failure if/when it happens.  And if we have so put Scotland off for 5 or 10 years, okay.  Or if we don't even get to spend 2 weeks in the national parks and we spend that time exploring free museums in our area, well... that's doesn't sound like failure, either.  Heck, even a part time job, assuming one or both of us could find one, would't really be so awful, especially for a fairly short term (let's say 2 years or less), and again, is one of the fail-safes we've consciously included in our plan.   It would only be if we had to make adjustments that were truly uncomfortable that I might to consider it a failure or wish we'd worked OMY. 

SnackDog

  • Handlebar Stache
  • *****
  • Posts: 1260
  • Location: Latin America
Re: cFiresim SEVERELY overestimates success rates for Mustachians
« Reply #44 on: September 21, 2016, 04:21:43 AM »
My grandfather retired from public school teaching in the worst retirement year on record, 1965, and lived 35 more years.   He lived well his first ten years, buying a new house in the 70s and new cars periodically, then cut back a bit as he slowed down.   He had a penchant for periodic new lawnmowers and Zenith televisions (color!), and dined out for lunch 7 days a week for at least 20 years. He didn't even come close to running out of money and in fact gave away quite a bit to his brother and sister and supported five grandkids' college tuition.  He had no Excel spreadsheets, no cFiresim, no internet, no computer, no chat board, no accountant, no advice, no 4% rules, no stock market investments. Nothing.  Just a pension, savings account and savings bonds.

Everyone on this board will be just fine.

boarder42

  • Walrus Stache
  • *******
  • Posts: 9332
Re: cFiresim SEVERELY overestimates success rates for Mustachians
« Reply #45 on: September 21, 2016, 05:33:42 AM »
This is the greatest existential risk I think this community faces.  It's not like the 95% success rate of the 4% rule means that 95 out of 100 of us who retire in any given year are going to make it on 4%.  We're either ALL going to make it, in that year, or we're ALL not going to make it.  Totally depends on what the future economy looks like.  If things go south on us bad enough, every singe retiree here is going to fail the 4% rule all at once, and then threads like this one are going to seem pretty silly.  All those early retirement rules and forecasts, though mathematically unchanged, will suddenly seem like a farce.

Only if we ALL rigidly adhere to a rule without changing our behavior at all in spite of poor market returns.  The ones who adjust and earn a little money in ER, or who cut their spending, likely will be just fine (and this is ignoring those who planned for a lower WR to begin with).

I'd bet that if a 4% WR "failed" over a 30 year period, much less than half of us would actually run out of money and end up with only social security income.
... Assuming ss survived the hypothetical doomsday market condition. But yes, the principle of constant optimization applies

really you think the US govt will take away a giant pension plan they setup ... this would be an upheaval. and this hypothetical doomsday turns the entire western world on its head and likely in to a state of anarchy.  so rather than hoard money you should be hoarding canned goods bullets and guns if you really think this doomsday is plausible

brooklynguy

  • Handlebar Stache
  • *****
  • Posts: 2204
  • Age: 44
Re: cFiresim SEVERELY overestimates success rates for Mustachians
« Reply #46 on: September 21, 2016, 07:37:51 AM »
This is the greatest existential risk I think this community faces.

I think the list of existential risk factors for the early retirement community is nearly as short as the one for Western civilization itself (which, admittedly, is alarmingly long).  If the world were to throw us an ostensibly 4%-rule-busting scenario, our ranks would undoubtedly thin as softer-core members drop out and Yahoo Finance stops publishing feel-good stories of retired 30-something millionaires to send waves of new recruits our way, but (short of true catastrophe) a critical mass of the community would remain and, to it, idea-exchange-places like this forum would become even more important.

Ok, but do we really need soap?

Foam?

ender

  • Walrus Stache
  • *******
  • Posts: 7402
Re: cFiresim SEVERELY overestimates success rates for Mustachians
« Reply #47 on: September 21, 2016, 07:53:55 AM »
The single greatest risk to an early retiree is inflexibility. Not what % you withdraw.

It's not obvious to me that this is true. 

Why would you be more likely to retire on a peak year than a positive year in the middle of a bull (or flatish) market?  I don't think you'd be any more likely to retire in 2007 than 1995 or 2013.

It has to do with the ratio of where portfolio growth comes from. The closer you get to FIRE, the more that (hopefully) comes from growth and not contributions. If your number is $1M and you start with $900k at the beginning of the year, 5% growth in the market is $45k. That's almost 2x what you can contribute as an individual to a 401k/IRA combined.

The net effect is that if you are waiting to hit a "number" it's much more likely you will hit it due to market growth than contributions. The most consistent market growth is when you are near a peak. And the market is often near a peak when it is growing.

Mmm_Donuts

  • Bristles
  • ***
  • Posts: 409
Re: cFiresim SEVERELY overestimates success rates for Mustachians
« Reply #48 on: September 21, 2016, 08:46:35 AM »
Would using the Variable CAPE method of withdrawing resolve this issue? When I look at the Shiller PE ratio chart linked to above via the MadFientist site, the peak point was in December 1999, at 44.14, which would imply a 2.2% withdrawal rate for that year for those following this method. I like the idea of lowering/raising SWR rate due to measurable indicators like this, but want to make sure I'm understanding the OP correctly. 

The Happy Philosopher

  • Bristles
  • ***
  • Posts: 342
    • thehappyphilosopher
Re: cFiresim SEVERELY overestimates success rates for Mustachians
« Reply #49 on: September 21, 2016, 09:11:46 AM »
Predictions are hard to makes, especially about the future.

When the 4% rule fails (and it probably will) it will likely be for reasons no one saw coming and no one prepared for. Doesn't mean it's not a great starting point though and it should hold up in most scenarios.

The biggest threat I see to MMM style early retirement is there is so much inflexibility to cutting further spending. If you are retiring on 25k/yr it is much harder/impossible to cut spending by 10k than if you are spending 70k. All the fat has already been cut out of the budget. There is more danger of sequence of return risk. Also everyone is healthy in their 30's (more or less). As I see insurance become thinner and more restrictive over time I worry about a medical event causing failure. It is hard to be frugal or go back to work when someone in the family is going through multiple rounds of chemotherapy. In this case even though the 4% rule may work, it may fail if you are unable to keep spending at 4%. We all talk about being flexible and spending less in times of market stress, but few talk about the potential need to spend MORE during these times.

At the end of the day it's about being comfortable with risk. The more redundant systems in place the better.