Author Topic: Removing success bias from safe withdrawal rate  (Read 7447 times)

Vilgan

  • Bristles
  • ***
  • Posts: 451
  • Location: Seattle, WA
Removing success bias from safe withdrawal rate
« on: January 20, 2016, 05:24:14 PM »
Hi all,

One thing I've noticed with safe withdrawal rate discussions is that they always center on the US. Many here are US citizens so that makes sense, but it seems like we also have the tools necessary to look at safe withdrawal rates on a more global scale as well. The twentieth century was insanely positive for the US. World wars demolished the industrial capacity of our competitors and the US went from a power to a global power. In almost every way, the 20th century was fabulous for the US and "better than average".

One paper I was reading here: http://www3.grips.ac.jp/~pinc/data/10-12.pdf remarked on this and here are two paragraphs I found particularly striking:

Quote
It is widely acknowledged and understood that the applicability of these withdrawal rate
studies depend on the future behaving with the same patterns as the past. But a potential problem
with the findings of so many of the existing studies is that they are based on the same Ibbotson
Associates' Stocks, Bonds, Bills, and Inflation (SBBI) monthly data on total returns for US
financial markets since 1926. Either this data is used directly for historical simulations or
bootstrapping approaches, or it is used to calculate parameters for Monte Carlo simulations. The
problem is that the time period covered by this data may have been a particularly fortuitous one
for the United States. If one thinks of the world as a Monte Carlo simulation, then the single
simulated path observed in the twentieth-century United States may not represent the true
underlying distribution of returns, and future returns are more likely to be lower.

 This point is made forcefully in Dimson, Marsh, and Staunton (2004). They argue that
looking only to past US data for future predictions will lead to "success bias" and sampling error.
As the title of their article suggests, it is "irrational optimism." In the first case, they note that the
US stock market capitalization grew from about 22 percent of world's total in 1900 to 54 percent
in 2003, and such success cannot be extrapolated into the future. As for sampling error, the US
data does not reach over a long enough time to be confident about its characteristics, as there are
too few distinct periods. Examining asset returns for a larger selection of countries should provide
a better idea about the range of possibilities for the future. Their hope was to eliminate the
widespread belief that the stock market will always provide a positive real return over a 20 year
period, as while this was true in the United States, it was also true only in 3 of the remaining 15
countries they investigate.

I know that whenever we talk about markets going up, Japan is a great counter example. MMM would likely talk about optimism etc and that is definitely cherry picking a bad result. However, I wonder why so little of the discussion here and in other places around safe withdrawal rates looks at it from a more global/average viewpoint? Isn't focusing only on US results in our best century dramatically over optimistic? This doesn't mean we have to run around assuming that we will be Japan, but looking at more average results seems like it would help people make more informed decisions.

While its easy to recover quickly now because we are young/motivated/etc, it seems like that might change over time. I know it would be a lot easier for me to sock away a bit more money now than to try to get back into the tech world after 15 years away because my safe withdrawal rate was overly optimistic.

Note: If Arebelspy sees this thread, I'm sure he'll land on the opposite end since he feels like 4% is way too safe. I do, wonder, however if a 50+ year old female who would encounter severe age bias getting a job again would feel as gung ho about the potential risks of being wrong.

PhysicianOnFIRE

  • Bristles
  • ***
  • Posts: 458
  • Age: 44
  • Location: Up North
    • Physician On FIRE
Re: Removing success bias from safe withdrawal rate
« Reply #1 on: January 20, 2016, 05:32:22 PM »
I worry about this too.   My defenses are:

1. Diversification.  My holdings are 60% US stock, 20% International, 10% REIT, 10% cash & bond

2. A lower withdrawal rate.  If am technically FI with about 25x expenses (although the January correction isn't helping), but I plan on working another 5 years or more.  I have a physician's income and should be able to increase my nest egg aggressively to about 50x before I RE. 

3. Passive Income.  Well, I'd like to have some before I retire, anyway.

SwordGuy

  • Walrus Stache
  • *******
  • Posts: 6497
  • Location: Fayetteville, NC
Re: Removing success bias from safe withdrawal rate
« Reply #2 on: January 20, 2016, 06:37:30 PM »
I worry about the success bias, plus about the downward pressure on stock prices as the baby boomers retire and start raiding their 401Ks (at least until the millennials start buying for their 401Ks).

Well, actually worry is the wrong word.  I have concerns that I've acted on in order to mitigate those risks.   Now I don't worry. :)
Does English have a word that means that?

My solution is to diversify income streams plus over-engineer it.

Our normal living expenses will come from Social Security, Disability payments, Rental Properties and Farming income, supplemented by part or full time work, as needed.

Our stock holdings are for investing with, not for living on.   Our income should drop after we retire, but it should continue to climb afterwards for this reason.


Eric

  • Magnum Stache
  • ******
  • Posts: 4061
  • Location: On my bike
Re: Removing success bias from safe withdrawal rate
« Reply #3 on: January 20, 2016, 07:03:37 PM »
Wade updated this work in 2014 here.

And despite being one of the most pessimistic retirement researchers out there, he still concluded that:

Quote
With globally diversified 50/50 portfolios, the global historical success rate for the 4% rule increased to 78.3%. For US investors, the global portfolio supported a success rate of 80.9%.

So right around 80% with a 50/50 portfolio, and since most of us are heavier in equities than 50%, we should expect better success rates.  So while not quite the 4% slam dunk of past US-centric performance, it's still pretty encouraging.  Don't forget that this assumes that you'll always take an inflation adjusted spending increase and that you'll never earn any other money nor collect SS.

Mr. Green

  • Handlebar Stache
  • *****
  • Posts: 2126
  • Age: 36
  • Location: Wilmington, NC
Re: Removing success bias from safe withdrawal rate
« Reply #4 on: January 20, 2016, 08:07:48 PM »
If you look at any demographics that compare productivity gains of the industrial revolution to the gains from the "microchip and internet" era you'll find that statistics show it's highly likely that the next hundred years of advancement will dwarf the last hundred. I think that's a valid argument to expect a better performance over the next century than the last. A hundred years from now, they'll look back at people in 2015 like we look back at people in 1900. Hell maybe the advancement will be so great that 2015 to them will look like 1850 or 1800 to us.

dude

  • Handlebar Stache
  • *****
  • Posts: 2367
Re: Removing success bias from safe withdrawal rate
« Reply #5 on: January 21, 2016, 08:14:38 AM »
If you look at any demographics that compare productivity gains of the industrial revolution to the gains from the "microchip and internet" era you'll find that statistics show it's highly likely that the next hundred years of advancement will dwarf the last hundred. I think that's a valid argument to expect a better performance over the next century than the last. A hundred years from now, they'll look back at people in 2015 like we look back at people in 1900. Hell maybe the advancement will be so great that 2015 to them will look like 1850 or 1800 to us.

A hundred years from now we may have cooked, polluted and poisoned our planet into something unrecognizable to those of us alive today.  I'll be long dead, so I remain optimistic with a touch of melancholy with respect to my remaining time on earth, but beyond 30-50 years from now, not so much.

BBub

  • Pencil Stache
  • ****
  • Posts: 774
  • Age: 35
  • Location: Deep South
Re: Removing success bias from safe withdrawal rate
« Reply #6 on: January 21, 2016, 09:09:24 AM »
If you look at any demographics that compare productivity gains of the industrial revolution to the gains from the "microchip and internet" era you'll find that statistics show it's highly likely that the next hundred years of advancement will dwarf the last hundred. I think that's a valid argument to expect a better performance over the next century than the last. A hundred years from now, they'll look back at people in 2015 like we look back at people in 1900. Hell maybe the advancement will be so great that 2015 to them will look like 1850 or 1800 to us.

A hundred years from now we may have cooked, polluted and poisoned our planet into something unrecognizable to those of us alive today.  I'll be long dead, so I remain optimistic with a touch of melancholy with respect to my remaining time on earth, but beyond 30-50 years from now, not so much.

Agreed.  Looking out over the next, say, millenium the only sure fire way to preserve wealth will be via gold, agricultural land, fine art... tangible assets which have consistently maintained intrinsic value independent of the fate of specific nations.  But even these stores of value may be subject to seizure, destruction, devaluation, etc.

But I don't plan on living for thousands of years.  I'm willing to bet my financial independence on the view that the US will continue to advance and prosper over the course of my remaining expected lifetime of 50 more years.  Given the choice, I'll take my chance on profitable companies in a well-established, growing nation with major structural advantages, abundant natural resource, and a strong military.  The extreme alternative is to prepare for the worst, accept mediocre returns, lose (or barely maintain) purchasing power, and toil for decades to achieve the desired result of FI.  Continued advancement in the developed world just seems like the most probable outcome, barring a major cataclysmic event that disrupts life on earth as we know it.

fattest_foot

  • Pencil Stache
  • ****
  • Posts: 787
Re: Removing success bias from safe withdrawal rate
« Reply #7 on: January 21, 2016, 09:47:34 AM »
If you look at any demographics that compare productivity gains of the industrial revolution to the gains from the "microchip and internet" era you'll find that statistics show it's highly likely that the next hundred years of advancement will dwarf the last hundred. I think that's a valid argument to expect a better performance over the next century than the last. A hundred years from now, they'll look back at people in 2015 like we look back at people in 1900. Hell maybe the advancement will be so great that 2015 to them will look like 1850 or 1800 to us.

A hundred years from now we may have cooked, polluted and poisoned our planet into something unrecognizable to those of us alive today.  I'll be long dead, so I remain optimistic with a touch of melancholy with respect to my remaining time on earth, but beyond 30-50 years from now, not so much.

Don't be so sure about that. This is off topic, but you may potentially never die.

If you've got some time to kill, I found this incredibly engrossing: The AI Revolution: The Road to Superintelligence

soccerluvof4

  • Walrus Stache
  • *******
  • Posts: 5795
  • Location: Artic Midwest
  • Retired at 50
    • My Journal
Re: Removing success bias from safe withdrawal rate
« Reply #8 on: January 21, 2016, 12:49:20 PM »
This fear/ question has always been around just in different disguises. Stay the course.

faramund

  • Bristles
  • ***
  • Posts: 331
Re: Removing success bias from safe withdrawal rate
« Reply #9 on: January 23, 2016, 12:46:35 AM »
If you look at any demographics that compare productivity gains of the industrial revolution to the gains from the "microchip and internet" era you'll find that statistics show it's highly likely that the next hundred years of advancement will dwarf the last hundred. I think that's a valid argument to expect a better performance over the next century than the last. A hundred years from now, they'll look back at people in 2015 like we look back at people in 1900. Hell maybe the advancement will be so great that 2015 to them will look like 1850 or 1800 to us.

A hundred years from now we may have cooked, polluted and poisoned our planet into something unrecognizable to those of us alive today.  I'll be long dead, so I remain optimistic with a touch of melancholy with respect to my remaining time on earth, but beyond 30-50 years from now, not so much.

Agreed.  Looking out over the next, say, millenium the only sure fire way to preserve wealth will be via gold, agricultural land, fine art... tangible assets which have consistently maintained intrinsic value independent of the fate of specific nations.  But even these stores of value may be subject to seizure, destruction, devaluation, etc.

But I don't plan on living for thousands of years.  I'm willing to bet my financial independence on the view that the US will continue to advance and prosper over the course of my remaining expected lifetime of 50 more years.  Given the choice, I'll take my chance on profitable companies in a well-established, growing nation with major structural advantages, abundant natural resource, and a strong military.  The extreme alternative is to prepare for the worst, accept mediocre returns, lose (or barely maintain) purchasing power, and toil for decades to achieve the desired result of FI.  Continued advancement in the developed world just seems like the most probable outcome, barring a major cataclysmic event that disrupts life on earth as we know it.
In a thousand years, I expect we won't be limited to one planet, or even one star. So, the relative value of gold and agricultural land by then is really uncertain. If agricultural productivity/land increases greater than population, then the value of land will go down.. And as for gold, there's gold in them thar asteroids, and if there's lots - its value will go down as well. Fine art, well whatever that is, maybe. Art from the last 100 years, seems to have a greater auction value then art from around 1000 AD. Who knows what people 1000 years in the future will think of what counts as fine art today.

Vilgan

  • Bristles
  • ***
  • Posts: 451
  • Location: Seattle, WA
Re: Removing success bias from safe withdrawal rate
« Reply #10 on: January 23, 2016, 09:17:21 AM »
This fear/ question has always been around just in different disguises. Stay the course.

Meh, the answer to "is there a more accurate number we can use?" is to clap hands over your ears and "stay the course" ? Really? :P

Article in OP would suggest a slightly lower SWR might be more accurate and more representative of average experience if you don't plan on having other income after retirement. It seems worthwhile to at least understand factors that may be over inflating the 4% SWR figure that many people use as an input into a lot of other decisions and calculations.

ender

  • Walrus Stache
  • *******
  • Posts: 5290
Re: Removing success bias from safe withdrawal rate
« Reply #11 on: January 23, 2016, 09:29:57 AM »
Retiring will always have risk.

Even if your plan is a government pension, there is still risk.

We can debate 80% vs 95% but the fact is:
  • Someone blindly following a plan without adjusting ever will do worse than others
  • Being FIRE allows you the opportunity to reduce spending significantly given you can now trade time for DIY/savings/etc
  • Even a 4/5 chance of success is still quite good (which I suspect if you added lifestyle adjustments into would be higher)
  • Most ER plan failures could be very easily mitigated by part-time jobs, even making $10k/year has a huge effect
  • Many people who FIRE end up doing hobbies/etc which could be monetized if necessary
  • People who ER really early may have some inheritance later in their lives
  • Many people will run into pensions/Social Security supplemental income later in their lives

There's definitely no guarantee.

But acting as if once you ER with a 4% SWR you have to withdraw and spend exactly 4% every year blindly without ever reevaluating your plan or adjusting is a bit of a pointless thought exercise.

Capsu78

  • Pencil Stache
  • ****
  • Posts: 727
  • Location: Chicagoland
Re: Removing success bias from safe withdrawal rate
« Reply #12 on: January 23, 2016, 10:29:25 AM »
I wonder how much thought Lemmy, Bowie, Rickman and Frey invested in confidence calculations of their SWR's!... Date of Demise calculators would have been most helpful to each of them.

Hey, that would make for a pretty prestigious law firm name, wouldn't it?

Bertram

  • Bristles
  • ***
  • Posts: 258
  • I'm not a chef
Re: Removing success bias from safe withdrawal rate
« Reply #13 on: January 23, 2016, 12:01:54 PM »
Retiring will always have risk.

Even if your plan is a government pension, there is still risk.

We can debate 80% vs 95%

Exactly! IMHO there is a spurious precision in these probabilities anyway. It's not like your gut/feeling can discern the difference between an 85% chance and a 95% chance, and there's as much impression in the model as there is in your subjective feeling of security. Sure you define thresholds, but in the end what matters is whether or not you are successful or not. And you could fail even if the success probability is 99,9%.

I think, just like Ender said, the important thing is not the exactness of the probability, but that you keep evaluating where you are, what the market is doing and that you adjust earnings/expenses where sensible/possible, especially in the early years of FI which AFAIK most people agree are the most important that determine whether you are successful or not.

RetiredAt63

  • Senior Mustachian
  • ********
  • Posts: 12140
  • Location: Eastern Ontario, Canada
Re: Removing success bias from safe withdrawal rate
« Reply #14 on: January 23, 2016, 12:58:49 PM »
This may be a bit simplistic, but if the return on investment is enough to finance retirement income (i.e. the 4% withdrawal rate), the original investments are not needing to be sold.  Some retirement vehicles may force this (i.e. RRIFs in Canada) and if that turns out to cause major economic problems the likely thing is for governments to relax the rules.  If people don't need all the pension money (i.e. when I am forced to take 20%) they may simply use it to buy investments outside the retirement vehicle.  Otherwise the original investment may never be cashed out if the retiree is fine for income, and the retirement portfolio will pass to the estate.  This handing down of wealth has already been discussed in the popular press for boomers, but really it is a more general situation.

For the record, in my own case, my pension income (Caisse de dépôt et placement du Québec - investment returns, not capital sell-offs) and private income (same) is fine, and so far (2 years in) my investments are still making more money than I can spend.

shotgunwilly

  • Pencil Stache
  • ****
  • Posts: 548
Re: Removing success bias from safe withdrawal rate
« Reply #15 on: January 25, 2016, 02:08:41 PM »
If you look at any demographics that compare productivity gains of the industrial revolution to the gains from the "microchip and internet" era you'll find that statistics show it's highly likely that the next hundred years of advancement will dwarf the last hundred. I think that's a valid argument to expect a better performance over the next century than the last. A hundred years from now, they'll look back at people in 2015 like we look back at people in 1900. Hell maybe the advancement will be so great that 2015 to them will look like 1850 or 1800 to us.

A hundred years from now we may have cooked, polluted and poisoned our planet into something unrecognizable to those of us alive today.  I'll be long dead, so I remain optimistic with a touch of melancholy with respect to my remaining time on earth, but beyond 30-50 years from now, not so much.

Don't be so sure about that. This is off topic, but you may potentially never die.

If you've got some time to kill, I found this incredibly engrossing: The AI Revolution: The Road to Superintelligence

I've read the article before, which is incredibly interesting and about something we should absolutely be worried about.  But has nothing to do with an individual living forever. 

If you look at any demographics that compare productivity gains of the industrial revolution to the gains from the "microchip and internet" era you'll find that statistics show it's highly likely that the next hundred years of advancement will dwarf the last hundred. I think that's a valid argument to expect a better performance over the next century than the last. A hundred years from now, they'll look back at people in 2015 like we look back at people in 1900. Hell maybe the advancement will be so great that 2015 to them will look like 1850 or 1800 to us.

A hundred years from now we may have cooked, polluted and poisoned our planet into something unrecognizable to those of us alive today.  I'll be long dead, so I remain optimistic with a touch of melancholy with respect to my remaining time on earth, but beyond 30-50 years from now, not so much.

Agreed.  Looking out over the next, say, millenium the only sure fire way to preserve wealth will be via gold, agricultural land, fine art... tangible assets which have consistently maintained intrinsic value independent of the fate of specific nations.  But even these stores of value may be subject to seizure, destruction, devaluation, etc.

But I don't plan on living for thousands of years.  I'm willing to bet my financial independence on the view that the US will continue to advance and prosper over the course of my remaining expected lifetime of 50 more years.  Given the choice, I'll take my chance on profitable companies in a well-established, growing nation with major structural advantages, abundant natural resource, and a strong military.  The extreme alternative is to prepare for the worst, accept mediocre returns, lose (or barely maintain) purchasing power, and toil for decades to achieve the desired result of FI.  Continued advancement in the developed world just seems like the most probable outcome, barring a major cataclysmic event that disrupts life on earth as we know it.
In a thousand years, I expect we won't be limited to one planet, or even one star. So, the relative value of gold and agricultural land by then is really uncertain. If agricultural productivity/land increases greater than population, then the value of land will go down.. And as for gold, there's gold in them thar asteroids, and if there's lots - its value will go down as well. Fine art, well whatever that is, maybe. Art from the last 100 years, seems to have a greater auction value then art from around 1000 AD. Who knows what people 1000 years in the future will think of what counts as fine art today.

Elon Musk is already on his mission to colonize mars.  He plans to personally make a trip there by the time he's dead.  Waitbutwhy.com has incredibly interesting articles on AI and SpaceX.  Some people may call Musk's ambitions crazy but you can't deny what the man has done.

He started a company that is outrunning NASA.  They just achieved the first rocket return-to-earth landing (first stage landing) by any entity in history. And were told it wouldn't be done.  (And this landing was just a side project experiment as the launch was to deliver a satellite for NASA.)  Every move him and his company makes is a stepping stone towards eventually colonizing mars and the progress has been amazing.

Personally, I believe we (humans) will destroy ourselves and the world within 150-200 years. 

Reynold

  • Stubble
  • **
  • Posts: 245
Re: Removing success bias from safe withdrawal rate
« Reply #16 on: January 25, 2016, 03:51:58 PM »
I've thought about the point you (the OP) mention as well, particularly in connection with Kitces international analysis. 
About a third of developed countries have less than a 50% success rate at 4% withdrawal;

http://retirementresearcher.com/the-shocking-international-experience-of-the-4-rule/

One concern I have is whether the US has been an outlier, and will revert to the mean, which would look more like the other countries mentioned.  Mitigating somewhat against that is that the U.S. has, for a long period of history, had a more dynamic economy than most other developed countries, so maybe it will continue to "lead the pack". 

The other concern I have is that with the aging of the U.S. population, it will start looking more like Europe, which is starting to look more like Japan, demographically.  Historically that leads to a pretty stagnant economy and stock market.  Mitigating somewhat against that is the fact that the U.S. is more welcoming of immigrants than other countries, so shouldn't have quite as bad a retiree-to-worker ratio as they would.  Capital tends to flow to the least bad economy, which could be us, even if our economy in 2030 is nothing like it was in 1980. 

So my own personal planning involves making sure we can keep our spending to more like a 3% withdrawal rate.  That seems like a reasonable precaution, but I could understand that people who really don't like their jobs, unlike me, might not want to be as conservative. 

Vilgan

  • Bristles
  • ***
  • Posts: 451
  • Location: Seattle, WA
Re: Removing success bias from safe withdrawal rate
« Reply #17 on: January 25, 2016, 04:44:07 PM »
I've thought about the point you (the OP) mention as well, particularly in connection with Kitces international analysis. 
About a third of developed countries have less than a 50% success rate at 4% withdrawal;

http://retirementresearcher.com/the-shocking-international-experience-of-the-4-rule/

One concern I have is whether the US has been an outlier, and will revert to the mean, which would look more like the other countries mentioned.  Mitigating somewhat against that is that the U.S. has, for a long period of history, had a more dynamic economy than most other developed countries, so maybe it will continue to "lead the pack". 

The other concern I have is that with the aging of the U.S. population, it will start looking more like Europe, which is starting to look more like Japan, demographically.  Historically that leads to a pretty stagnant economy and stock market.  Mitigating somewhat against that is the fact that the U.S. is more welcoming of immigrants than other countries, so shouldn't have quite as bad a retiree-to-worker ratio as they would.  Capital tends to flow to the least bad economy, which could be us, even if our economy in 2030 is nothing like it was in 1980. 

So my own personal planning involves making sure we can keep our spending to more like a 3% withdrawal rate.  That seems like a reasonable precaution, but I could understand that people who really don't like their jobs, unlike me, might not want to be as conservative.

Interesting article. I thought the 3.26% SAFEMAX number for ex-US (vs 3.96 for US) was particularly interesting as it draws the stark difference between the US and developed nations as a whole.

Telecaster

  • Handlebar Stache
  • *****
  • Posts: 2195
  • Location: Seattle, WA
Re: Removing success bias from safe withdrawal rate
« Reply #18 on: January 25, 2016, 04:59:16 PM »

He started a company that is outrunning NASA.  They just achieved the first rocket return-to-earth landing (first stage landing) by any entity in history. And were told it wouldn't be done.  (And this landing was just a side project experiment as the launch was to deliver a satellite for NASA.)  Every move him and his company makes is a stepping stone towards eventually colonizing mars and the progress has been amazing.

Slight nit:   The space shuttle main engines returned to Earth and were used.   The shuttle solid rocket boosters were also recovered and re-used (parachute recovery).   

Fastfwd

  • Stubble
  • **
  • Posts: 198
Re: Removing success bias from safe withdrawal rate
« Reply #19 on: January 26, 2016, 08:05:21 AM »
... Looking out over the next, say, millenium the only sure fire way to preserve wealth will be via gold, agricultural land, fine art... tangible assets which have consistently maintained intrinsic value independent of the fate of specific nations.  But even these stores of value may be subject to seizure, destruction, devaluation, etc.

Land should keep its value but it is also the first thing that falls to new political regimes. Also it can change in value quite a lot depending on importance that is given to location. Imagine a world with free air transport or with 90% of people working from home. Suddenly living close to downtown is not so valuable. New techniques in agriculture could make land worth less or population increase could make it worth more. Lands with lakes and rivers could be worth a LOT more.

I never understood gold. It's just a make believe value. It has no more worth than many things and far less than other metals or gases. Helium is going away and we are not making any new one. Gold has very limited real world uses.

I have little use for art but it seems some people will always be ready to pay a premium for old things at least as long as there is a class with large disposable income.

zephyr911

  • Magnum Stache
  • ******
  • Posts: 3624
  • Age: 41
  • Location: Northern Alabama
  • I'm just happy to be here. \m/ ^_^ \m/
    • Pinhook Development LLC
Re: Removing success bias from safe withdrawal rate
« Reply #20 on: January 26, 2016, 08:36:55 AM »
I worry about this too.   My defenses are:

1. Diversification.  My holdings are 60% US stock, 20% International, 10% REIT, 10% cash & bond

2. A lower withdrawal rate.  If am technically FI with about 25x expenses (although the January correction isn't helping), but I plan on working another 5 years or more.  I have a physician's income and should be able to increase my nest egg aggressively to about 50x before I RE. 

3. Passive Income.  Well, I'd like to have some before I retire, anyway.
Pretty much my strategy, along with some (easy/casual) part-time work that I'll keep doing after I hit 25x so that I don't need any of the investment income until it's well above that.
I'll actually quit FT work before I hit 25x with this strategy, but still avoid total reliance on a 4% WR. If there's one thing I never wanna do again in my life, it's leave my field and have to go back just to pay the bills. Once was enough, and god did I hate it.

BBub

  • Pencil Stache
  • ****
  • Posts: 774
  • Age: 35
  • Location: Deep South
Re: Removing success bias from safe withdrawal rate
« Reply #21 on: January 26, 2016, 12:55:27 PM »
... Looking out over the next, say, millenium the only sure fire way to preserve wealth will be via gold, agricultural land, fine art... tangible assets which have consistently maintained intrinsic value independent of the fate of specific nations.  But even these stores of value may be subject to seizure, destruction, devaluation, etc.

Land should keep its value but it is also the first thing that falls to new political regimes. Also it can change in value quite a lot depending on importance that is given to location. Imagine a world with free air transport or with 90% of people working from home. Suddenly living close to downtown is not so valuable. New techniques in agriculture could make land worth less or population increase could make it worth more. Lands with lakes and rivers could be worth a LOT more.

I never understood gold. It's just a make believe value. It has no more worth than many things and far less than other metals or gases. Helium is going away and we are not making any new one. Gold has very limited real world uses.

I have little use for art but it seems some people will always be ready to pay a premium for old things at least as long as there is a class with large disposable income.

Agreed.  Nothing is guaranteed & I'm certainly not advocating ownership of these assets for the purposes of helping people around here achieve FIRE.  It was more of an observation about which assets have held value over the centuries.  If one's chief concern is wealth preservation and the time horizon is several centuries, those 3 are probably the best bet.

No Name Guy

  • Bristles
  • ***
  • Posts: 449
  • Location: Western Washington
Re: Removing success bias from safe withdrawal rate
« Reply #22 on: January 26, 2016, 01:03:05 PM »
http://youarenotsosmart.com/2016/01/02/yanss-065-survivorship-bias-rebroadcast/

Gist in re survivorship bias:  Those that fail are forgotten, or aren't around to be study subjects to see what REALLY went wrong so that corrections could be made to prevent failure in the future.

Everyone knows Warren Buffet - who knows the person who ran that shop down the way that recently closed / failed?  Why did it close / fail?  Studying the failure is more important than the success.  Success, in one measure, is simply not failing.  Learning how to not fail is probably more important that how to "succeed".

The YANSS episode talks about bomber losses during WW2.  Somebody wanted to know what changes could be made to reduce losses.  The trouble is they only airplanes they could study in any detail were those that actually made it back to base - only the survivors could be studied, not the bombers that were actually shot down (and the damage that was actually the fatal damage).  At best, studying the survivors only gave limited information as to what could be done to actually avoid or mitigate the damage that actually brought down the bombers.

More relevantly - aviation has made huge strides in safety by actually studying failures.  Unlike war losses, typically the wreckage (and data recorders, cockpit recorders, radar tapes, ATC tapes, maintenance records, etc) is fully available for study post crash in a civil arena.  With each crash - the Comet and metal fatigue, multiple controlled flight into terrain, Value Jet fire / crash and the hazardous cargo, the Dallas Delta microburst crash, and the biggest body count of them all, the Tenerife KLM / Pan Am collision and crew resource management, valuable lessons were learned about HOW failures occur and actions were taken (changing design standards, improved crew training, better maintenance practices, better operational procedures, etc) to reduce or eliminate these causes of failure.

Studying successful flights on the other hand tells you little of what it takes to avoid failure (crashing).

The whole "do you regret paying off your mortgage" thread is one of success versus not failure.  My paid off mortgage means I will never fail to have shelter (at the super low annual cost of property taxes plus ~1% for maintenance).  Leveraging / keeping a mortgage and investing for income however adds all kinds of failure mechanisms into the provision of shelter.  Thanks, but I'll take the avoid failure approach - the same one that makes modern commercial aviation so boringly safe.

BBub

  • Pencil Stache
  • ****
  • Posts: 774
  • Age: 35
  • Location: Deep South
Re: Removing success bias from safe withdrawal rate
« Reply #23 on: January 26, 2016, 01:12:54 PM »
I worry about this too.   My defenses are:

1. Diversification.  My holdings are 60% US stock, 20% International, 10% REIT, 10% cash & bond

2. A lower withdrawal rate.  If am technically FI with about 25x expenses (although the January correction isn't helping), but I plan on working another 5 years or more.  I have a physician's income and should be able to increase my nest egg aggressively to about 50x before I RE. 

3. Passive Income.  Well, I'd like to have some before I retire, anyway.
Pretty much my strategy, along with some (easy/casual) part-time work that I'll keep doing after I hit 25x so that I don't need any of the investment income until it's well above that.
I'll actually quit FT work before I hit 25x with this strategy, but still avoid total reliance on a 4% WR. If there's one thing I never wanna do again in my life, it's leave my field and have to go back just to pay the bills. Once was enough, and god did I hate it.

I worry about it too.  My plan has been rightfully criticized as too conservative, but since you two spilled your guts I'll join.  My goals, in order:

1. 33x Stash.  70/30 mix
2. 2 yrs cash
3. 3-5 Rentals (partially leveraged)

I figure it should only take about 2 extra yrs after reaching 33x to accumulate the cash & seed a decent real estate portfolio.  The combination of savings, dividends, interest should help me get the rental portfolio built up & stabilized up pretty quickly.  This structure should provide:

A.) prolonged downside protection
B.) continued exponential growth, i.e. a runaway stash via the equities and principal paydown of leveraged rentals.
C.) diversified income streams

I realize all of these benefits are already built into the 4% SWR portfolio, but what can I say, I'm a belt & suspenders guy.

Eric

  • Magnum Stache
  • ******
  • Posts: 4061
  • Location: On my bike
Re: Removing success bias from safe withdrawal rate
« Reply #24 on: January 26, 2016, 01:20:14 PM »
The whole "do you regret paying off your mortgage" thread is one of success versus not failure. 

Yes, totally agree.

My paid off mortgage means I will never fail to have shelter (at the super low annual cost of property taxes plus ~1% for maintenance).  Leveraging / keeping a mortgage and investing for income however adds all kinds of failure mechanisms into the provision of shelter.  Thanks, but I'll take the avoid failure approach - the same one that makes modern commercial aviation so boringly safe.

No, wait, that's the exact opposite conclusion you should be making.  You're looking at a data set of all of the people who have successfully paid off their house and concluding that it's the optimal thing to do.  You're ignoring the failures, the exact same thing the whole rest of your post warns about. You're ignoring all of the people who paid extra to their mortgage and then lost their jobs and had nothing to show for it after foreclosure.  Once your home is paid off, it's like landing that WWII plane back at the base.  Of course it's the right decision if you make it!

Retire-Canada

  • Walrus Stache
  • *******
  • Posts: 7332
Re: Removing success bias from safe withdrawal rate
« Reply #25 on: January 26, 2016, 01:37:57 PM »
Being older solves a lot of the fear and what if's about the 4% SWR. I'm too old to realistically save and invest more than 25x my FIRE COL.

Heck I'm likely not going to bother even getting to 25x. Unless I get super lucky with markets/income I'll downshift to part-time and use a variable withdrawal rate to make a smaller starting portfolio work.

Ultimately the certainty of spending years of my prime-time working a full-time gig vs. being flexible and dealing with what actually happens comes out in favour of time off sooner.

When you have 40yrs+ of retirement ahead of you there is time to adjust and figure stuff out. Having more money or stuff may not help in the future we are actually headed towards.

BigoteGato

  • 5 O'Clock Shadow
  • *
  • Posts: 15
Re: Removing success bias from safe withdrawal rate
« Reply #26 on: January 26, 2016, 04:46:33 PM »
If you look at any demographics that compare productivity gains of the industrial revolution to the gains from the "microchip and internet" era you'll find that statistics show it's highly likely that the next hundred years of advancement will dwarf the last hundred. I think that's a valid argument to expect a better performance over the next century than the last. A hundred years from now, they'll look back at people in 2015 like we look back at people in 1900. Hell maybe the advancement will be so great that 2015 to them will look like 1850 or 1800 to us.
I agree with your point of view, but for an alternative point of view, check out  ‘The Rise and Fall of American Growth’ by Robert J. Gordon. NYT review at http://www.nytimes.com/2016/01/31/books/review/the-powers-that-were.html.

No Name Guy

  • Bristles
  • ***
  • Posts: 449
  • Location: Western Washington
Re: Removing success bias from safe withdrawal rate
« Reply #27 on: January 27, 2016, 01:39:47 PM »
No, wait, that's the exact opposite conclusion you should be making.  You're looking at a data set of all of the people who have successfully paid off their house and concluding that it's the optimal thing to do.  You're ignoring the failures, the exact same thing the whole rest of your post warns about. You're ignoring all of the people who paid extra to their mortgage and then lost their jobs and had nothing to show for it after foreclosure.  Once your home is paid off, it's like landing that WWII plane back at the base.  Of course it's the right decision if you make it!

Actually Eric, you're talking about HOW one goes about paying off a mortgage, not the risk factors of actually having a paid off mortgage.

You correctly point out that there are failure modes in making early payments - a person can do that and still, if the sh*t hits the fan bad enough, lose their house AND the extra they've put into the mortgage.  Fair enough.  I realized those failure mechanism's by studying those I knew who actually did suffer them (lost a job, lost a house after paying extra)....so I thought "let's eliminate or substantially mitigate that failure mechanism".

One way to mitigate this failure mechanism is to save, in low risk (capital preservation / low volatility / low risk of capital loss) vehicles, the excess funds used to pay off a mortgage.  In the meantime, keep making the required minimum mortgage payments.  (So, instead of throwing 500 or 1k / month extra at the mortgage, put that money into a high yield savings account, CD's, or short term AAA tax free muni bonds, or what ever fit's a person's definition of low risk savings vehicle).  When savings in low risk vehicles exceeds remaining balance, lump sum pay off the mortgage using the low risk savings vehicle funds.

If, as in your scenario, a person has the sh*t hit the fan during this process, they have a fat lump of cash to draw upon to continue making the minimum monthly payments - house saved, cash not wasted (still providing housing), time bought to mitigate the situation (finding a new job, selling the house without an impending foreclosure, etc).

I was ridiculed here on MMM some time ago for this very strategy, but as I said, it's a "not fail" (or more correctly, a "minimize failure mechanism") versus a "succeed" way of going about things.  Enough is enough after all, and my preferred method of "minimize failure" gets me to "enough" fast enough.

It worked great for me - I did that very strategy some years ago.  Here I sit, with a paid off home.  I could pay the taxes and maintenance from a part - part time, minimum wage job if it came to that.....and never mind the income potential for renting out rooms, etc.

So yes, I'm a "success", but I got to success by studying those who failed, how & why they failed and took measures in the "how" of paying off my mortgage to eliminate, mitigate, hedge or minimize those failure mechanisms to the maximum extent practical.