Author Topic: Rich, Broke or Dead: Visualizing probabilities of outcomes in early retirement  (Read 16133 times)

CCCA

  • Bristles
  • ***
  • Posts: 499
  • Location: Bay Area, California
  • born before the 80's
    • FI programming
I really liked @maizeman 's graphs (see https://forum.mrmoneymustache.com/post-fire/for-those-who-follow-the-4-rule/msg1438246/#msg1438246 as one example) and I thought I'd make an interactive version of this graph so people could play with it and customize it to their particular parameters, rather than play with python and updating his script.  I DM'd maizeman and got his blessing to do so as I didn't want to step on anyone's toes. 

https://engaging-data.com/will-money-last-retire-early/
It should be relatively intuitive how to use it and the idea is to play with it and test out new theories/plans/scenarios and understand the options. You shouldn't be able to break it, but if you do, please let me know and I'll try to fix it.  It uses the same data as cFIREsim, historical periods from 1871 to 2016 to run through each scenario with inflation adjusted spending and then calculates the probability of success or failure, but adds, like maizeman's graphs did, the probability of death from social security life expectancy tables depending on your sex.

If you hide the "wedge of death", you can also use it as a simpler form of FIREcalc or cFIREsim without all the extras that complicate it.  I've gotten a few comments about things to add: social security, taxes, variable investment fees and I will probably do those, though I'm not sure about the timeline.  If you have other things you want to add, let me know and I'll see.  I'll be on vacation for awhile so I might not get to things for a few weeks at the earliest.

I'd appreciate any other comments or suggestions as well.
 
« Last Edit: June 26, 2018, 06:41:21 PM by CCCA »

hadabeardonce

  • Stubble
  • **
  • Posts: 240
  • It's never too early to learn the value of money.
    • My Journal
That dead % is scary...

Can you add something like a "life event" option that would add in Soc Sec or pension income? I'd like to combat that broke %.

Maybe a non-smoker checkbox to lower the dead % too...

rab-bit

  • Stubble
  • **
  • Posts: 105
Thanks for doing this, nice work!

Another suggestion would be to add a joint life option for a couple.

FiveSigmas

  • Bristles
  • ***
  • Posts: 464
Neat! Thanks for sharing, CCCA (and thanks Maizeman again for the static version).

I played around with a variety of inputs, and it almost always works great, but something weird seems to happen when I input an allocation of exactly 70/20/10 (Stock/Bond/Cash): it says the numbers don't add up to 100%. Any ideas?

OurTown

  • Handlebar Stache
  • *****
  • Posts: 1002
  • Age: 49
  • Location: Tennessee
Great tool, I am much more likely to die than to go broke.

Much Fishing to Do

  • Bristles
  • ***
  • Posts: 378
Wow, I loved the Maizeman graphs and so this is awesome.

Stupid question I'm sure... The Balance<Start and the Balance>2x wedges demarcations are all based on inflation adjusted dollars, right?, I'm not sure I can wrap my head around if that is definitely the case or not...

rab-bit

  • Stubble
  • **
  • Posts: 105
Neat! Thanks for sharing, CCCA (and thanks Maizeman again for the static version).

I played around with a variety of inputs, and it almost always works great, but something weird seems to happen when I input an allocation of exactly 70/20/10 (Stock/Bond/Cash): it says the numbers don't add up to 100%. Any ideas?

I saw this also for 60/30/10.

CCCA

  • Bristles
  • ***
  • Posts: 499
  • Location: Bay Area, California
  • born before the 80's
    • FI programming
First off, thanks everyone for the feedback.  Keep it coming!  And I'm glad that the tool is useful/fun for everyone here. 

That dead % is scary...

Can you add something like a "life event" option that would add in Soc Sec or pension income? I'd like to combat that broke %.

Maybe a non-smoker checkbox to lower the dead % too...

Hmm, I'll have to think about the smoker thing.  I will definitely add the SS annuity thing, though not sure when I'll get around to it.

Thanks for doing this, nice work!

Another suggestion would be to add a joint life option for a couple.

Great, glad that you like it.  I like the idea of a joint life option, will add to the list.


Great tool, I am much more likely to die than to go broke.

Yup, I think that's the stark reminder in the graph.  That wedge of death gets awfully big, awfully fast.  But still need to worry about longevity of your stache.

Neat! Thanks for sharing, CCCA (and thanks Maizeman again for the static version).

I played around with a variety of inputs, and it almost always works great, but something weird seems to happen when I input an allocation of exactly 70/20/10 (Stock/Bond/Cash): it says the numbers don't add up to 100%. Any ideas?
I saw this also for 60/30/10.
Thanks!  I fixed the issue with the asset inputs.  My check for summation to 100% was a little too strict given how javascript converts numbers.  Anyway, I loosened the allowable sloppyness in the numbers and should be fine now. 


Wow, I loved the Maizeman graphs and so this is awesome.

Stupid question I'm sure... The Balance<Start and the Balance>2x wedges demarcations are all based on inflation adjusted dollars, right?, I'm not sure I can wrap my head around if that is definitely the case or not...
. Everything is in nominal dollars here.  I think  I will add a toggle (again, some point in the future) to switch between real and nominal dollars. 

wageslave23

  • Bristles
  • ***
  • Posts: 262
That dead % is scary...



I was looking at the red thinking "that's not that scary, but why is % death not increasing with age"... "oh shit, black is death %!"

2Birds1Stone

  • Magnum Stache
  • ******
  • Posts: 4681
  • Age: 31
  • Location: Earth
  • K Thnx Bye
Very cool/useful tool! Sell this to mass media for use on CNBC and the like.....

simmias

  • 5 O'Clock Shadow
  • *
  • Posts: 71
Loads fine for me.

I had a disheartening amount of red on the screen until I realized I had left a 0 off of my stash number!

OurTown

  • Handlebar Stache
  • *****
  • Posts: 1002
  • Age: 49
  • Location: Tennessee
If we all live long enough to get the Ray Kurzweil immortality treatments, that death wedge can drop back down to 0.

itchyfeet

  • Pencil Stache
  • ****
  • Posts: 638
Great visual tool. Still not sure what to make of a 70% probability Iíll be dead in 40 years, but if I do live that long thereís a 20% chance Iíll be broke. Hmmmm....

maizeman

  • Magnum Stache
  • ******
  • Posts: 2998
If we all live long enough to get the Ray Kurzweil immortality treatments, that death wedge can drop back down to 0.

Of course with paying for monthly injections of nanobots (or whatever form the treatment takes) added to the budget, I'd imagine that red wedge would rise to 100% pretty darn fast. ;-)

I'd told CCCA this earlier, but I really like the new calculator and display. Reading through some of the early comments, it would seem we're rapidly going to be adding a lack of good, detailed actuarial data for life expectancy at different ages with different combinations of health risks to my perennial complaint about the lack of good public sector datasets on stock and bond returns outside the USA.

albireo13

  • Bristles
  • ***
  • Posts: 438
  • Location: New England
very cool.  an interesting way of looking at the future landscape.
What is the death % calculation based on?

Rob

maizeman

  • Magnum Stache
  • ******
  • Posts: 2998
Social security provides this handy table with your risk of dying in the next year and average number of years you have left to live for any given age: https://www.ssa.gov/OACT/STATS/table4c6.html

DreamFIRE

  • Handlebar Stache
  • *****
  • Posts: 1345
First off, thanks everyone for the feedback.  Keep it coming!  And I'm glad that the tool is useful/fun for everyone here. 

That dead % is scary...

Can you add something like a "life event" option that would add in Soc Sec or pension income? I'd like to combat that broke %.

Maybe a non-smoker checkbox to lower the dead % too...

Hmm, I'll have to think about the smoker thing.  I will definitely add the SS annuity thing, though not sure when I'll get around to it.

Yeah, I mentioned this in another thread about this same graph.  With 15 years of FIRE until getting a decent SS benefit, it's quite significant:

https://forum.mrmoneymustache.com/investor-alley/maizeman's-trippy-script/msg2036794/#msg2036794

But even 92% without SS when it includes a large discretionary spending cushion in the WR is not bad in itself.  Still it would be nice to tack on some SS.  I expect that would show 100% here as with cFireSim.

shuffler

  • Bristles
  • ***
  • Posts: 305
That dead % is scary...
Yeah.  I think there's a bug.
No matter how much I jack up the size of the stash, the dead-region doesn't get any smaller.
That can't be how it's supposed to work.  Not in real life.  </s>

maizeman

  • Magnum Stache
  • ******
  • Posts: 2998
I needed that laugh tonight. Thank you @shuffler.

itchyfeet

  • Pencil Stache
  • ****
  • Posts: 638
That dead % is scary...
Yeah.  I think there's a bug.
No matter how much I jack up the size of the stash, the dead-region doesn't get any smaller.
That can't be how it's supposed to work.  Not in real life.  </s>

Lol

markbike528CBX

  • Pencil Stache
  • ****
  • Posts: 819
  • Location: the Everbrown part of the Evergreen State (WA)
That dead % is scary...
Yeah.  I think there's a bug.
No matter how much I jack up the size of the stash, the dead-region doesn't get any smaller.
That can't be how it's supposed to work.  Not in real life.  </s>

Sure you can, there's a toggle to turn the dead zone off .

CCCA

  • Bristles
  • ***
  • Posts: 499
  • Location: Bay Area, California
  • born before the 80's
    • FI programming
If we all live long enough to get the Ray Kurzweil immortality treatments, that death wedge can drop back down to 0.

Of course with paying for monthly injections of nanobots (or whatever form the treatment takes) added to the budget, I'd imagine that red wedge would rise to 100% pretty darn fast. ;-)

I'd told CCCA this earlier, but I really like the new calculator and display. Reading through some of the early comments, it would seem we're rapidly going to be adding a lack of good, detailed actuarial data for life expectancy at different ages with different combinations of health risks to my perennial complaint about the lack of good public sector datasets on stock and bond returns outside the USA.

Since maizeman is in the thread I’d like to thank him again for his idea to put the death projection and portfolio projections together.

I think one key value of this is that it reframes the considerations from purely focused on the few percentage points we normally agonize over in cfiresim (trying to go from 10% failure to 5% or 0%). But when viewed with the giant “wedge of death”, the differences between 5, 10 or even 20% seem relatively small compared to the chance of death.

FIRE 20/20

  • Stubble
  • **
  • Posts: 141
If we all live long enough to get the Ray Kurzweil immortality treatments, that death wedge can drop back down to 0.

Or if the Nick Bostrom / Eliezer Yudkowsky wing of the AI/singularity conversation is correct then the death wedge will spike to 100% for all of us at some unknown point.

I want to thank both maizeman and CCCA for this.  I've shared maizeman's graphs with a lot of people, and now I will need to send them CCCA's link.  You guys are awesome!

markbike528CBX

  • Pencil Stache
  • ****
  • Posts: 819
  • Location: the Everbrown part of the Evergreen State (WA)
thanks maizeman and CCCA!   

I downloaded the Ken Fisher, "when can I retire" guide  (ad link at side), just for fun. Hope you get a kickback and enable your FIRE earlier.

No, I did NOT use my real name and email address.   Did that twice before and was bugged for years (Fisher and some gold thingy).

CCCA

  • Bristles
  • ***
  • Posts: 499
  • Location: Bay Area, California
  • born before the 80's
    • FI programming
thanks maizeman and CCCA!   

I downloaded the Ken Fisher, "when can I retire" guide  (ad link at side), just for fun. Hope you get a kickback and enable your FIRE earlier.

No, I did NOT use my real name and email address.   Did that twice before and was bugged for years (Fisher and some gold thingy).


Thanks for the thought, though you don't need to click on ads.
If we all live long enough to get the Ray Kurzweil immortality treatments, that death wedge can drop back down to 0.

Or if the Nick Bostrom / Eliezer Yudkowsky wing of the AI/singularity conversation is correct then the death wedge will spike to 100% for all of us at some unknown point.

I want to thank both maizeman and CCCA for this.  I've shared maizeman's graphs with a lot of people, and now I will need to send them CCCA's link.  You guys are awesome!
.


Yeah, I'm sort of on the pessimistic side of the AI/singularity thing but who knows what will happen. 
Please share the link.  I'm happy that people find it useful.

dude

  • Handlebar Stache
  • *****
  • Posts: 2239
Boy, I'm not sure how I feel about the fact that I'm far more likely to be dead than broke!

dude

  • Handlebar Stache
  • *****
  • Posts: 2239
Though I suspect the high educational attainment of Mustachians -- a protective factor -- probably skews towards a significantly lower death rate curve.

maizeman

  • Magnum Stache
  • ******
  • Posts: 2998
Though I suspect the high educational attainment of Mustachians -- a protective factor -- probably skews towards a significantly lower death rate curve.

Huh, it makes a bigger difference than I would have guessed. So based on this paper* from back in 2010, a 25 year old man with a graduate degree can expect to live to 85, while a 25 year old man who only graduated from high school can expect to live to 76.

The difference is smaller for women: 87 (graduate degree) and 81 (high school) respectively.

*https://www.ncbi.nlm.nih.gov/pubmed/25093685

letsdoit

  • Bristles
  • ***
  • Posts: 270
it looks amazing.  it Is not running. 
there is no button to run it

CCCA

  • Bristles
  • ***
  • Posts: 499
  • Location: Bay Area, California
  • born before the 80's
    • FI programming
it looks amazing.  it Is not running. 
there is no button to run it


Hmm, if you change the numbers in the input boxes, and it doesn't update, you can press "Enter" on your keyboard to make sure that the change is registered, or just click on the "Update Asset Allocation" button, which will redraw the graph based on the updated inputs.

letsdoit

  • Bristles
  • ***
  • Posts: 270
i was using a work cpu

at home it works like a charm
Thank you!  so, we're all gonna die, huh?

Glenstache

  • Handlebar Stache
  • *****
  • Posts: 2070
  • Age: 186
  • Location: Seattle!
  • Target FI date 2027 (maybe?)
i was using a work cpu

at home it works like a charm
Thank you!  so, we're all gonna die, huh?
given enough time, yes. Hence, the point is to do things that add meaning to your life, not just scrimp and save.

CCCA

  • Bristles
  • ***
  • Posts: 499
  • Location: Bay Area, California
  • born before the 80's
    • FI programming
i was using a work cpu

at home it works like a charm
Thank you!  so, we're all gonna die, huh?


if there was no picture at all, maybe your work computer disabled javascript?

centwise

  • 5 O'Clock Shadow
  • *
  • Posts: 90
I love this! Thank you (and thanks also to Maizeman). I'm posting mainly so I'll be able to easily find in the future. I think it will make a great teaching tool to illustrate the general idea of the 4% rule to people, and then (of course) it goes way beyond that because of all the scenarios you can play with.

less4success

  • Bristles
  • ***
  • Posts: 255
  • Location: Just moved
    • Seldom updated blog
Are the different degrees of success based on comparing inflation-adjusted portfolio values or nominal?

CCCA

  • Bristles
  • ***
  • Posts: 499
  • Location: Bay Area, California
  • born before the 80's
    • FI programming
Are the different degrees of success based on comparing inflation-adjusted portfolio values or nominal?


Nominal values.  I have plans to update it to allow for an inflation adjusted view as well.

SugarMountain

  • Pencil Stache
  • ****
  • Posts: 578
It would complicate this beautifully simple tool, but it would be nice if it modeled a couple.  When I do simulations, it seems that the odds that both DW and I will survive 45 years is sadly very small (although my GPs on one side both made it to 95).  If we imagine that spending for 1 of us would be 75% of the two of us, it really lowers the chances of ever going broke, since the chances of both of us making it to our 80s is less than 30%.

dude

  • Handlebar Stache
  • *****
  • Posts: 2239
Though I suspect the high educational attainment of Mustachians -- a protective factor -- probably skews towards a significantly lower death rate curve.

Huh, it makes a bigger difference than I would have guessed. So based on this paper* from back in 2010, a 25 year old man with a graduate degree can expect to live to 85, while a 25 year old man who only graduated from high school can expect to live to 76.

The difference is smaller for women: 87 (graduate degree) and 81 (high school) respectively.

*https://www.ncbi.nlm.nih.gov/pubmed/25093685

Yeah, and from what I've read, the gap has been widening. With all the talk about how "college isn't worth it" going on now (primarily because of the rising expenses), the fact remains that not only do those with college degree significantly out-earn their non-college peers over a lifetime, but they are also now living significantly longer.  That's the uncomfortable inequality -- longevity inequality -- most people aren't talking about.
« Last Edit: June 19, 2018, 07:30:20 AM by dude »

Chaplin

  • Pencil Stache
  • ****
  • Posts: 981
  • Location: Le Canada
Though I suspect the high educational attainment of Mustachians -- a protective factor -- probably skews towards a significantly lower death rate curve.

Huh, it makes a bigger difference than I would have guessed. So based on this paper* from back in 2010, a 25 year old man with a graduate degree can expect to live to 85, while a 25 year old man who only graduated from high school can expect to live to 76.

The difference is smaller for women: 87 (graduate degree) and 81 (high school) respectively.

*https://www.ncbi.nlm.nih.gov/pubmed/25093685

Yeah, and from what I've read, the gap has been widening. With all the talk about how "college isn't worth it" going on now (primarily because of the rising expenses), the fact remains that not only do those with college degree significantly out-earn their non-college peers over a lifetime, but they are also now living significantly longer.  That's the uncomfortable inequality -- longevity inequality -- most people aren't talking about.

I have no data to support this, but I would guess that it has a lot to do with which group is more likely to smoke. If you took out the smokers, I bet the longevity would be considerably closer.

maizeman

  • Magnum Stache
  • ******
  • Posts: 2998
That's a good point. Lower educational attainment is also associated with a higher probability of drunk driving.* I'm guessing you'd see a lot of other high risk behaviors that show similar patterns. And also suicides: "Men with a high school education were twice as likely to die by suicide compared with those with a college degree in 2014."**

* https://www.ncbi.nlm.nih.gov/pubmed/12711655

** https://www.ncbi.nlm.nih.gov/pubmed/28756896

dragoncar

  • Walrus Stache
  • *******
  • Posts: 8114
  • Registered member
Subscribing to learn when we get the inflation-adjusted option.  2x starting stache doesn't mean much if it ain't inflation adjusted

Slee_stack

  • Pencil Stache
  • ****
  • Posts: 853
As with a pension...I presume this doesn't factor any SS income either?

Loren Ver

  • Bristles
  • ***
  • Posts: 285
  • Location: Indianapolis IN
Neato!  Thanks for making and sharing.

DreamFIRE

  • Handlebar Stache
  • *****
  • Posts: 1345
As with a pension...I presume this doesn't factor any SS income either?

See the original post.

Roland of Gilead

  • Handlebar Stache
  • *****
  • Posts: 1663
Why is everyone dead at 105?  100%

There is always the chance that we discover ways to prevent cell death/decay.

Or download your mind into whatever quantum computer thingy they have in 40 years (but then not sure how large of a stash one needs...maybe a lot more, maybe not as much)

maizeman

  • Magnum Stache
  • ******
  • Posts: 2998
Given current health and medical science, out of every 100,000 male babies born today, we would expect 76 (so 0.076%) to live to 105, so I suspect CCCA's website is just rounding that value to zero. Women do a little better, 0.3% of them can expect to make it to 105.

But yes, there is indeed a non-zero (but difficult if not impossible to estimate) longevity "risk" from discovering a transformational cure for old age.

EscapeVelocity2020

  • Handlebar Stache
  • *****
  • Posts: 1816
  • Age: 45
  • Location: Houston
    • EscapeVelocity2020
One thing I always struggled with on the FIRE discussions was the enthusiastic discussion of the fact that people die or get incapacitated earlier than 'traditional retirement age', so that's a great reason to ER.  As a corollary, there is also enthusiastic talks about how, if we manage to live a really long time, we will spend less and leave lots of money to others.  Both of these discussions made me a little sad as to why, if you really are 'betting' to die in 10 - 20 years, then do ER and don't complain if you don't die, but stop talking about it like you are 60 or 70 years old.  Also, if you live below your means for much longer than expected, then enjoy figuring out to do with all that excess then.  But in the meantime, just figure out a good way to optimize for the foreseeable future.  The long term for you will inevitably be something different from whatever an average projects, and only you can be the best judge if you, say, eat better than average, exercise more than average, have better genetics than average, build net worth better than average, etc.

In my engineering job, Operations complains when we try to make things inflexible and reduce their job to the point where they don't have to think.  That is apparently the most dangerous situation, when inexperienced people rely on safeguards as a matter of course.

Replacing 'expected outcomes' with 'normalized outcomes' subjects the individual to inevitable disappointment.
« Last Edit: June 21, 2018, 10:31:40 PM by EscapeVelocity2020 »

Glenstache

  • Handlebar Stache
  • *****
  • Posts: 2070
  • Age: 186
  • Location: Seattle!
  • Target FI date 2027 (maybe?)
Given current health and medical science, out of every 100,000 male babies born today, we would expect 76 (so 0.076%) to live to 105, so I suspect CCCA's website is just rounding that value to zero. Women do a little better, 0.3% of them can expect to make it to 105.

But yes, there is indeed a non-zero (but difficult if not impossible to estimate) longevity "risk" from discovering a transformational cure for old age.
If this were found, and it were somehow available to the masses, I wonder how that would affect the hidden assumptions of the Trinity study? Would working habits and productivity change? That type of demographic shift has the potential to really disturb economies (not to mention population growth if the balance between adding and subtracting shifts dramatically).

maizeman

  • Magnum Stache
  • ******
  • Posts: 2998
Yup, radical life extension (that works and is available to many people) definitely falls into the that category of black swan/long tail events where historical data probably would lose most of its ability to forecast the future.

If the reality of a person's body or mind being too old to work went away, would the concept of regular retirement vanish entirely? Or would people continue to retire and the balance of workers to retirees would rapidly spiral out of control? (If timed perfectly this could work well with the accelerating pace we're seeing blue and white collar jobs being replaced by robotics and AI respectively.)

In the past increases in life expectancy in a given country tend to be correlated with declines in birth rates (usually with a lag time of a generation or so, so you still see a dramatic uptick in population). Would that relationship hold true if lifespans doubled or tripled?

boarder42

  • Walrus Stache
  • *******
  • Posts: 7854
this is amazing