Author Topic: CFireSim Success Rate  (Read 101719 times)

NaturallyHappier

  • Stubble
  • **
  • Posts: 130
  • Age: 58
  • Location: Near Philadelphia, PA
  • FIRED 3/10/2017
CFireSim Success Rate
« on: May 02, 2015, 12:25:58 PM »
What success rate do you consider safe when running CFireSim and using historical data -all years and constant withdrawal rate adjusted for inflation?

I know that Vanguard uses a Montecarlo simulator and considers anything above 80% safe.

I'm sure this a "based on personal opinion" type of thing, but I am wondering what other MMM folks use and how they rationalize it.

I am thinking that I should target between 90 and 100%  because it is running numbers against historical returns and I can afford to lower my withdraw by as much as $10K if I had too for several years.  Of course, that assumes that the future will be no worst than the past.

Chuck

  • Bristles
  • ***
  • Posts: 407
  • Age: 35
  • Location: Northern VA
Re: CFireSim Success Rate
« Reply #1 on: May 02, 2015, 12:34:51 PM »
I am comfortable with anything better than 80%

sol

  • Walrus Stache
  • *******
  • Posts: 8433
  • Age: 47
  • Location: Pacific Northwest
Re: CFireSim Success Rate
« Reply #2 on: May 02, 2015, 05:13:53 PM »
Of course, that assumes that the future will be no worst than the past.

If you assumed that the future would be no worse than the past, then you would target a 50% success rate.  Targeting 80% success means that you are expecting the future to be as bad as the worst 20% of historical periods.  Targeting 100% success, like many here do, means you expect the future to be worse than anything that has ever happened, including world wars, oil embargoes, presidential assassinations, and global catastrophic economic collapse.  All of that has happened.

NaturallyHappier

  • Stubble
  • **
  • Posts: 130
  • Age: 58
  • Location: Near Philadelphia, PA
  • FIRED 3/10/2017
Re: CFireSim Success Rate
« Reply #3 on: May 02, 2015, 05:23:15 PM »
Of course, that assumes that the future will be no worst than the past.

If you assumed that the future would be no worse than the past, then you would target a 50% success rate.  Targeting 80% success means that you are expecting the future to be as bad as the worst 20% of historical periods.  Targeting 100% success, like many here do, means you expect the future to be worse than anything that has ever happened, including world wars, oil embargoes, presidential assassinations, and global catastrophic economic collapse.  All of that has happened.

I'm not sure I understand.  Wouldn't 50% mean that in 50% of the cases in the past I would have run out of money?  If I want to make sure that I don't run out of money when my current situation is projected over all past time periods, wouldn't that require a 100% success rate using CFireSim?  Maybe I am not understanding what the program is doing.

forummm

  • Walrus Stache
  • *******
  • Posts: 7374
  • Senior Mustachian
Re: CFireSim Success Rate
« Reply #4 on: May 02, 2015, 06:10:08 PM »
Of course, that assumes that the future will be no worst than the past.

If you assumed that the future would be no worse than the past, then you would target a 50% success rate.  Targeting 80% success means that you are expecting the future to be as bad as the worst 20% of historical periods.  Targeting 100% success, like many here do, means you expect the future to be worse than anything that has ever happened, including world wars, oil embargoes, presidential assassinations, and global catastrophic economic collapse.  All of that has happened.

I'm not sure I understand.  Wouldn't 50% mean that in 50% of the cases in the past I would have run out of money?  If I want to make sure that I don't run out of money when my current situation is projected over all past time periods, wouldn't that require a 100% success rate using CFireSim?  Maybe I am not understanding what the program is doing.

You understand correctly. Sol has suggested in the past that a 50% success rate is sufficient for him because he is willing to cut back the fat from his budget (and/or get a job) if it looks like his retirement timing would put him in that 50% of scenarios that would be likely to fail. The idea is that you can generally tell within the first 5 years whether your portfolio is likely to succeed or fail based on what the market does during that time. If you have less flexibility to make those adjustments or are not as interested in taking that risk, then you can choose to work longer to save up enough extra money to make it much more likely that you will die rich and without many money worries.

sol

  • Walrus Stache
  • *******
  • Posts: 8433
  • Age: 47
  • Location: Pacific Northwest
Re: CFireSim Success Rate
« Reply #5 on: May 02, 2015, 06:39:39 PM »
I'm not sure I understand.  Wouldn't 50% mean that in 50% of the cases in the past I would have run out of money?

Yes, 50% of the time you would have run out of money with a 50% success rate.  In a large population of retirees, half of them would have saved too much and half would have not saved enough.  On average, they have saved the right amount.

When you reach the 50% success rate, that means that you have on average already saved enough.  Unless you expect the market future to be below the historical average, this might be sufficient for you.

You can never "make sure" you never run out of money.  You could get cancer or get sued or the world could get nuked.  Most people target an 80% success rate not because they're willing to risk not being in the 20%, but because they recognize that there are extrinsic risks not accounted for in their analysis that historical market analyses cannot represent.  By moving from an 80% success rate to a 90% success rate, you're just buying false security because your chances of some extrinsic risk probably exceeds the market performance risk you're analyzing.

NaturallyHappier

  • Stubble
  • **
  • Posts: 130
  • Age: 58
  • Location: Near Philadelphia, PA
  • FIRED 3/10/2017
Re: CFireSim Success Rate
« Reply #6 on: May 02, 2015, 06:59:17 PM »
Sol. This is a very good point.  We can always count on you for a deeper perspective.  Thanks.

Monkey Uncle

  • Handlebar Stache
  • *****
  • Posts: 1740
  • Location: West-by-god-Virginia
Re: CFireSim Success Rate
« Reply #7 on: May 03, 2015, 06:33:56 AM »
I'm a coward, so I'm shooting for over 90%.  Over 95% would be even better.  I've found that cFiresim is less conservative than the Monte Carlo simulator that I also use (Schwab).  When cFiresim is over 90%, the Monte Carlo is usually around 80%.  FireCalc, on the other hand, seems to be more liberal.  It gives a 100% success rate.  I'm sure the variability has to do with the fact that you can't set all the parameters exactly the same on all three.

forummm

  • Walrus Stache
  • *******
  • Posts: 7374
  • Senior Mustachian
Re: CFireSim Success Rate
« Reply #8 on: May 03, 2015, 07:45:51 AM »
I'm a coward, so I'm shooting for over 90%.  Over 95% would be even better.  I've found that cFiresim is less conservative than the Monte Carlo simulator that I also use (Schwab).  When cFiresim is over 90%, the Monte Carlo is usually around 80%.  FireCalc, on the other hand, seems to be more liberal.  It gives a 100% success rate.  I'm sure the variability has to do with the fact that you can't set all the parameters exactly the same on all three.

This is because the Monte Carlo simulators end up with some retirement cycles having multiple catastrophically bad years of stock returns one-after-the-other. Which is something that has never happened in modern developed financial markets. Monte Carlo simulators provide unrealistic sequence of returns risks. They also have the same problem with putting too many crazy boom years in a row as well. Both the highest and lowest returns on those simulators are incredibly unlikely.

brooklynguy

  • Handlebar Stache
  • *****
  • Posts: 2204
  • Age: 43
Re: CFireSim Success Rate
« Reply #9 on: May 03, 2015, 03:13:09 PM »
Sol's insight that a 50% success rate is optimal if your goal is minimization of total lifetime working hours (and assuming the future is no worse than the past) is in my view one of his more important contributions to the interdisciplinary field of FIREology (which is saying a lot).

I would just note, however, that in addition to assuming the future is no worse than the past, this idea also assumes that every retirement has an equal chance of success regardless of what market conditions exist when the retirement commences (which is also an implicit assumption inherent in the notion of using historical success rates as a proxy for chances of future retirement success in the first place).  When we use SWR-related concepts like "success rate," we all pretend this assumption is true, but in reality we all know it is false, and that is at least part of the reason most of us shoot for a WR that has historically succeeded in history's worst case (or near-worst case) scenarios.

Also, the idea that "any success rate over 80% constitutes false precision" (which is quickly becoming a cliche around here) is itself an example of false precision--it's not as if "80%" (or, in weaker versions of the expression, "around 80%") represents a bright line (or even a fuzzy bright line) under which every percentage point increase in historical success rate translates into a meaningful (and precise) corresponding increase in actual chances of your own retirement's success.

sol

  • Walrus Stache
  • *******
  • Posts: 8433
  • Age: 47
  • Location: Pacific Northwest
Re: CFireSim Success Rate
« Reply #10 on: May 03, 2015, 03:37:42 PM »
Sol's insight that a 50% success rate is optimal if your goal is minimization of total lifetime working hours

Every once on a while I have a moment of clarity, though most of the time I fail to communicate it very well and it is lost to the ages.  I'm glad you found something of value in what I've shared.

Quote
I would just note, however, that in addition to assuming the future is no worse than the past, this idea also assumes that every retirement has an equal chance of success regardless of what market conditions exist when the retirement commences

I agree that this is a very real problem, and one that we have discussed here before. An 80% success rate never means that 20% of this year's retirees will fail, it generally means either 100% or 0% of this year's retirees will fail and you have an 80% chance of this being one of the good 100% years.  At least it's usually been pretty obvious when you're headed towards one of the 20% failure scenarios, so that's something to be glad about.

I do wonder about the cohort effect, though.  If the class of 2015 retirees who use a 4% SWR all end up getting hosed, they are no less hosed even if the odds were in their favor.  Sometimes you just lose.

brooklynguy

  • Handlebar Stache
  • *****
  • Posts: 2204
  • Age: 43
Re: CFireSim Success Rate
« Reply #11 on: May 03, 2015, 04:42:30 PM »
Yeah. It makes me wonder what I would do if there truly were a way of knowing the precise odds of my own retirement plan's success, like if the government offered every worker the chance to win a guaranteed inflation-indexed pension in an amount equal to your projected spending under a "Wheel of Fortune" like scheme, where every year of additional work increases the share of winning slices put on the wheel when you decide to take your spin.  Of course, in the real world, we do almost have the equivalent of this option, given the availability for purchase of inflation-indexed single premium fixed annuities (except, unlike a government-backed annuity, the retiree bears some counterparty credit risk, and unlike the Wheel of Fortune gimmick, you know exactly what you're getting when and if you decide to purchase one), but the prohibitive expensiveness of this option for extremely early retirees effectively means (generally speaking, or maybe even in all cases?) that you would have to work many, many years beyond the otherwise doable early retirement age in order to obtain a significantly high confidence level.

Bateaux

  • Handlebar Stache
  • *****
  • Posts: 2318
  • Location: Port Vincent
Re: CFireSim Success Rate
« Reply #12 on: May 03, 2015, 05:49:39 PM »
I don't trust any calculator 100%.  However I'd never consider retirement at anything less than 100% based on past data.

sol

  • Walrus Stache
  • *******
  • Posts: 8433
  • Age: 47
  • Location: Pacific Northwest
Re: CFireSim Success Rate
« Reply #13 on: May 03, 2015, 06:08:02 PM »
I don't trust any calculator 100%.  However I'd never consider retirement at anything less than 100% based on past data.

So you're willing to work longer than required, that's fine.  If you like your work, and would prefer to leave a large legacy when you die, or you plan on increasing your spending in retirment.

It's that last one that I think causes most people to work longer than necessary.  Just because you can mathematically retire today with your current expenses doesn't mean there is no incentive to work longer so that you can spend more in the future than you do today.  Several of the regulars here have talked openly about our desire to spend significantly more in retirement than we do now, while working.

regulator

  • Bristles
  • ***
  • Posts: 469
Re: CFireSim Success Rate
« Reply #14 on: May 03, 2015, 06:25:33 PM »
Further to Sol's point, at some point you have to acknowledge two big risks that will tend to confound historical simulators and with which Monte Carlo struggles as well: exogenous risks and mortality/morbidity risks.  The exogenous risks could range from the mundane (US gets invaded and we all end up bowing to the head Imam or whatever) to the exotic (space aliens kill us all, environmental catastrophe, etc.).  No success rate in any calculator will account for any of these.  Exogenous risks as I describe them are low probability, high severity.  Mortality/morbidity risks OTOH are 100% certain at some point.  Look at a mortality table for your gender and age and you will realize that a big percentage of would-be retirees your age do not have to worry about outliving their savings. As for morbidity, my otherwise very healthy FIL is in advanced Parkinson's at age 70.  I'd blow my brains out in his shoes rather than live his life.

Get busy living, because before you know it you will get busy dying.

forummm

  • Walrus Stache
  • *******
  • Posts: 7374
  • Senior Mustachian
Re: CFireSim Success Rate
« Reply #15 on: May 03, 2015, 06:27:48 PM »
Yeah. It makes me wonder what I would do if there truly were a way of knowing the precise odds of my own retirement plan's success, like if the government offered every worker the chance to win a guaranteed inflation-indexed pension in an amount equal to your projected spending under a "Wheel of Fortune" like scheme, where every year of additional work increases the share of winning slices put on the wheel when you decide to take your spin.  Of course, in the real world, we do almost have the equivalent of this option, given the availability for purchase of inflation-indexed single premium fixed annuities (except, unlike a government-backed annuity, the retiree bears some counterparty credit risk, and unlike the Wheel of Fortune gimmick, you know exactly what you're getting when and if you decide to purchase one), but the prohibitive expensiveness of this option for extremely early retirees effectively means (generally speaking, or maybe even in all cases?) that you would have to work many, many years beyond the otherwise doable early retirement age in order to obtain a significantly high confidence level.

Social Security is pretty close to this. You have to assume it will still exist in its current form. And it's not exactly scaleable to your desired level of spending (unless you desire to spend within whatever level of benefit you have).

I think our current public policy on retirement is pretty dumb. By having pensions (other than SS) going by the wayside, we're forcing people to take on a lot of personal risk for the solvency of their retirement. They may be unsophisticated and trust financial advisors who put them into active management funds with lots of big load fees and ongoing fees. Or they may panic when the market crashes and sell. It would be much better to have a supplemental pension system to SS that you can take with you from job to job. Maybe the government sets some actuarial rate for your age and that's the amount of your salary you have to pay in to get a certain percent of your salary as a pension when you hit a certain age. Like SS does now, but as an additional benefit. That way the nation can target the 50% rate for everyone (or 55% if you want to make sure the government doesn't have to kick in public funds). And no one has to work until they get to 80% or 100% success rate just in case they retire just before a bad market period.

brooklynguy

  • Handlebar Stache
  • *****
  • Posts: 2204
  • Age: 43
Re: CFireSim Success Rate
« Reply #16 on: May 03, 2015, 06:54:09 PM »
Social Security is pretty close to this.

Except that annuity doesn't start until close to traditional retirement age (at the earliest), so it doesn't help the extremely early retiree in the decades before that point (other than to potentially reduce the amount needed to bridge the gap until then, but the gap is usually large enough that the amount needed to bridge it is equal or nearly equal to what would be needed indefinitely).

sol

  • Walrus Stache
  • *******
  • Posts: 8433
  • Age: 47
  • Location: Pacific Northwest
Re: CFireSim Success Rate
« Reply #17 on: May 03, 2015, 07:40:42 PM »
I think our current public policy on retirement is pretty dumb. By having pensions (other than SS) going by the wayside, we're forcing people to take on a lot of personal risk for the solvency of their retirement.

Someone recently suggested that this switch to defined contribution plans like 401ks instead of defined benefit plans like pensions was basically a reaction to the red scare.  The concern was that labor unions would end up owning all of the US stock market through their pension and retirement plans, making us a defacto socialist state where only union members are citizens.

In a time when average Joes couldn't easily invest and unions were ubiquitous and strong, that logic made more sense.  Today the markets are saturated by corporate players, hedge funds, day traders, and foreign investors so I think the fear is unfounded. In that light, maybe it's time to revisit the idea of defined benefit plans.

clifp

  • Pencil Stache
  • ****
  • Posts: 890
Re: CFireSim Success Rate
« Reply #18 on: May 03, 2015, 08:42:43 PM »


You understand correctly. Sol has suggested in the past that a 50% success rate is sufficient for him because he is willing to cut back the fat from his budget (and/or get a job) if it looks like his retirement timing would put him in that 50% of scenarios that would be likely to fail. The idea is that you can generally tell within the first 5 years whether your portfolio is likely to succeed or fail based on what the market does during that time. If you have less flexibility to make those adjustments or are not as interested in taking that risk, then you can choose to work longer to save up enough extra money to make it much more likely that you will die rich and without many money worries.

I am not sure if that is really true you can tell after 5 years if your retirement is in failure.  As a 99/2000 retiree arguable one of the worst years to retire (although I guess we won't know until 2030). my portfolio had recovered modestly by Dec 2005 from the dotcom bubble and bear markets.   Raddr's study http://www.raddr-pages.com/forums/viewtopic.php?f=2&t=1208&start=390puts it at $754,000 at the start of 2006. Would you really start to looking for work at the point? I suspect most all actual retirees did what I did during those years and gave up the annual inflation adjustment. (I switched in 2003/4 to focusing on dividends and only spending my dividends and interest.)   Now by the end of 2008 when your portfolio was cut in half, then you'd probably consider either looking or making significant cut backs. I know March of 2009 was when I seriously thought about going back to work.  Needless to say a 50 year old trying to look for a job in 2009, with almost 10 year gap in my resume in the midst of the great recession wasn't in fact a viable back up plan.

I am not really sure if there is huge difference on the employablilty of somebody with a 5 years off the job versus 10.  However I do think there is a big difference between 45, 50, 55, or 65 going back to work I think mothers who have returned to the work would be an interesting perspective to get. I would have had to take a massive (50-75%+) pay cut from my previous job was undeniable, while my pay cut would have been larger than most people I suspect it will be substantial for most everyone.   I think if you are truly trying to minimize the amount of your life spent working for the man, you also need to factor in the likelyhood that a market crash will correspond with a general economic downturn, and the lower wages of going back to work after quitting.


brooklynguy

  • Handlebar Stache
  • *****
  • Posts: 2204
  • Age: 43
Re: CFireSim Success Rate
« Reply #19 on: May 04, 2015, 08:20:45 AM »
I am not sure if that is really true you can tell after 5 years if your retirement is in failure.

I agree with this.  I think the general "outrageous optimism" ethos of mustachianism and the ingrained reliance on various levels of safety margin sometimes colors our perception a bit too rosy.

It's a common refrain around here (and elsewhere in the FIRE community) that the early years of retirement are the most critical, with the implicit (or explicit) implication that it's easy to recognize if your retirement is headed for failure in order to take appropriate corrective action.  Lest we get too cavalier about the potential difficulty of mitigating sequence of return risk, though, let's keep in mind that, although most of us tend to focus on the nightmare scenario of a sharp and sudden market drop shortly after commencing retirement, the more significant risk has historically been an extended period of low (but still positive, even in real terms) market returns during the first decade or so of retirement.  So it's not as if there will always be a 2008-like market crash to serve as a wake-up call that things aren't going according to plan; it's just as likely (or, based on history, more likely) that an extended period of subpar (but not drastically bad) real market returns will slowly doom a withdrawal plan to failure like a frog trapped inside a boiling pot.  Moreover, consistent with recent musings around the forum by skyrefuge and others, we shouldn't fall into the trap of believing that the market's next retirement-killer will necessarily resemble any of those of the past.

Quote
I think if you are truly trying to minimize the amount of your life spent working for the man, you also need to factor in the likelyhood that a market crash will correspond with a general economic downturn, and the lower wages of going back to work after quitting.

On the other hand, we shouldn't veer too far down the path of pessimism and ignore the various levels of safety margin that we rightly take comfort in (the ability to return to work being, in most cases, only one of many).

sol

  • Walrus Stache
  • *******
  • Posts: 8433
  • Age: 47
  • Location: Pacific Northwest
Re: CFireSim Success Rate
« Reply #20 on: May 04, 2015, 09:10:20 AM »
I am not sure if that is really true you can tell after 5 years if your retirement is in failure.

While I still think that a sudden steep crash immediately after retirement should be a clear warning sign, I also think it's easier to recognize when you're comfortably into the "guaranteed success" side of the equation.  If you retired in 2009 and then saw your investments more than double over the next five years, chances are good that your retirement is not in jeopardy. 

And as long as both the really great outcomes and the really horrible outcomes are both recognizable, I don't worry as much about the great unknown middle ground.  I recognize that it exists, but it's not like no retiree can ever know how well his plan is unfolding.  In some cases, you know a lot.

just as likely (or, based on history, more likely) that an extended period of subpar (but not drastically bad) real market returns will slowly doom a withdrawal plan to failure like a frog trapped inside a boiling pot.

I think you're not adequately describing the early 1970s retiree, who clearly saw his low returns being coupled to crazy high inflation, making the relevant effective return pretty obviously terrible.  It's not like those folks were watching the stock market hold steady and celebrating because they didn't recognize the danger.

What other historical period of failure had a long period of moderate returns and still saw a 4% SWR portfolio failure?  cFIREsim suggests that 1906 is the only other year that 4% fails a 30 year retirement.

« Last Edit: May 04, 2015, 09:53:09 AM by sol »

Chuck

  • Bristles
  • ***
  • Posts: 407
  • Age: 35
  • Location: Northern VA
Re: CFireSim Success Rate
« Reply #21 on: May 04, 2015, 09:37:29 AM »
I think our current public policy on retirement is pretty dumb. By having pensions (other than SS) going by the wayside, we're forcing people to take on a lot of personal risk for the solvency of their retirement.

Someone recently suggested that this switch to defined contribution plans like 401ks instead of defined benefit plans like pensions was basically a reaction to the red scare. 
Doubtful.

If that were the case the movement away from defined benefit pensions would have started in the 50's, rather than the early 80's. Also it bears mentioning that no more than ~35% of Americans ever had access to a defined benefit pension. The myth that they were a widespread benefit is just that.

Quote
Start with private pension plans, which underpin nostalgic yarns about the golden age of the 1960s, when every man could raise a family on assembly-line wages and then retire in comfort. These pensions never were as widespread in fact as they are in popular legend—when the number of such plans peaked in the 1980s, they covered only about one-third of the workforce. And as it turned out, a lot of those plans failed catastrophically. Defined-benefit plans have a huge downside: they drastically discourage labor mobility. Not only do they make an economy less dynamic by tying workers to a given company, but they also leave the workers vulnerable if the company goes under, taking their retirement with it.


The reason pensions were dropped is because they are astoundingly expensive, and commit the company's future earnings in a very restrictive way. It is much better business to have the worker assume the risk and expense of their own retirement.
« Last Edit: May 04, 2015, 09:54:50 AM by Chuck »

brooklynguy

  • Handlebar Stache
  • *****
  • Posts: 2204
  • Age: 43
Re: CFireSim Success Rate
« Reply #22 on: May 04, 2015, 09:48:10 AM »
I think you're not adequately describing the early 1970s retiree, who clearly saw his low returns being coupled to crazy high inflation, making the relevant effective return pretty obviously terrible.  It's not like those folks were watching the stock market hold steady and celebrating because they didn't recognize the danger.

What other historical period of failure had a long period of moderate returns and still saw a 4% SWR portfolio failure?  cFIREsim suggests that 1906 is the only other year that 4% fails a 30 year retirement.

Well, not all the retiree cohorts undone by the period of stagflation in the 1970s had equally terrible returns in the early years of their retirements.  Take the 1962 starting year cohort, for example, which registers as a "success" in cFIREsim (using its default settings for all the variables) using a 4% SWR retirement plan, but these retirees only avoided failure by the skin of their teeth -- their remaining portfolio in year 30 was so drastically low that one or two more years of retirement would clearly deplete their portfolio to zero.  Yet the first decade of retirement for that cohort witnessed pretty decent real total annualized returns of 6.65% in the S&P 500 (according to IndexView).

But still, your point is well taken that an early retiree would have had to be pretty clueless to be living through that period without recognizing the writing on the wall.  My point was intended mostly to expand the anti-recency-bias campaign that's been undertaken in various threads around the forum, because my sense is that many people may be mentally steeling themselves against the types of risks history has presented in the last few years but not the more prevalent risks of the last few decades or the potential risks that history has yet to throw at us at all.

sol

  • Walrus Stache
  • *******
  • Posts: 8433
  • Age: 47
  • Location: Pacific Northwest
Re: CFireSim Success Rate
« Reply #23 on: May 04, 2015, 10:14:01 AM »
or the potential risks that history has yet to throw at us at all.

I'm fundamentally optimistic about the future. 

I think the odds of another major terror attack (and subsequent revenge invasion) are significantly greater than zero, and I have a short list of personal favorites that I can't believe no one has tried yet.  Flying airplanes into buildings was a good one, but there are a bunch more ways that bad people could do major harm with limited resources.  I think most people don't realize how fragile civilization really is.

I wouldn't be surprised by a devastating outbreak of some antibiotic-resistant pathogen taking 10% of the world population in my lifetime.  Such diseases already exist, and new ones are evolving faster than our drugs are.  I don't think anyone believes our ability to scale up the production of new antibiotics can possible compete with exponentially growing infection numbers, so this is one of those deals where we rapidly lose control of if we don't stop it up front.  The planet has never had this many people living in such close quarters before, we're ripe for infection.

Natural disasters happen with some regularity.  What would the US economy do if Mt. Rainier wiped out half of Seattle?  If the Hayward fault took down half of Northern California?  Katrina was a speedbump in a second-tier city.

War.  Always war.  It's what humans do.  There are lots of ways for this to unfold next time, but my money is on some combination of the Middle East and China.

Climate Change.  In most places, changes will be slow and gradual and we'd have no problems adapting.  In some places, those slow changes will lead to very  rapid shifts.  Plate tectonics was slow, too, until South America suddenly disconnected from Antarctica and connected the Pacific and Atlantic oceans for the first time in a billion years and dramatically changed the planet's ocean circulation pattern.  That's the sort of problem with climate change I'm worried about. 

Generic population growth.  There are currently more people living in southeast Asia than in the rest of the world combined.  I don't claim to understand what impact that will have on the future of the planet, but I don't believe it will be irrelevant to the next century.

And despite all of that, I still think the future of the world economy looks better than the past.  Standards of living are rising. The rule of law is expanding.  Life expectancy and literacy rates are climbing.  We've already survived a bunch of that bad stuff listed above, and we'll get through some more of it, but I still think the general trend for humanity (and the stock market) is up up and away for the remainder of my own short life.
« Last Edit: May 04, 2015, 11:31:50 AM by sol »

brooklynguy

  • Handlebar Stache
  • *****
  • Posts: 2204
  • Age: 43
Re: CFireSim Success Rate
« Reply #24 on: May 04, 2015, 10:29:46 AM »
I love the irony of preceding and concluding a laundry-list of possible global calamities with serious potential for actual occurrence in the near-term with an expression of your optimism about the future :)

And I have to say, I agree 100%.  As a NYC resident (with no plans on ever moving) who lost a loved one on 9/11, catastrophic terrorist attack is at the forefront of my own personal list of exogenous risk factors.  But I remain optimistic about the future for the same reasons you described, plus the recognition that if the shit hits the fan in one of these ways then the triviality of the success of my own retirement will be thrown into razor-sharp relief.

Scandium

  • Magnum Stache
  • ******
  • Posts: 2827
  • Location: EastCoast
Re: CFireSim Success Rate
« Reply #25 on: May 04, 2015, 10:36:02 AM »
I'm not sure I understand.  Wouldn't 50% mean that in 50% of the cases in the past I would have run out of money?

Yes, 50% of the time you would have run out of money with a 50% success rate.  In a large population of retirees, half of them would have saved too much and half would have not saved enough.  On average, they have saved the right amount.

When you reach the 50% success rate, that means that you have on average already saved enough.  Unless you expect the market future to be below the historical average, this might be sufficient for you.

You can never "make sure" you never run out of money.  You could get cancer or get sued or the world could get nuked.  Most people target an 80% success rate not because they're willing to risk not being in the 20%, but because they recognize that there are extrinsic risks not accounted for in their analysis that historical market analyses cannot represent.  By moving from an 80% success rate to a 90% success rate, you're just buying false security because your chances of some extrinsic risk probably exceeds the market performance risk you're analyzing.

I still don't understand this. With a 100% success rate you would have survived every possible outcome that has occurred. I.e you expect the future to be exactly as bad as the past. If you expect the future to be twice as bad as the past you'd target a 200% success rate. With a 50% success rate you would have been ok in the 50% least-bad cycles (and failed in the other half), so you implicitly expect the future to be half as bad as the past. That is how I interpret it.

brooklynguy

  • Handlebar Stache
  • *****
  • Posts: 2204
  • Age: 43
Re: CFireSim Success Rate
« Reply #26 on: May 04, 2015, 10:47:35 AM »
I still don't understand this.

The idea is that if (i) you expect the future to be exactly as bad as the past and (ii) the chances that your own retirement start year will follow the path of any of history's previous retirement start years are exactly equal, then once you attain more than a 50% success rate your retirement is more likely than not to succeed, which means if you continue working you are more likely than not working longer than necessary.

sol

  • Walrus Stache
  • *******
  • Posts: 8433
  • Age: 47
  • Location: Pacific Northwest
Re: CFireSim Success Rate
« Reply #27 on: May 04, 2015, 10:58:17 AM »
With a 100% success rate you would have survived every possible outcome that has occurred. I.e you expect the future to be exactly as bad as the past.

No, with a 100% success rate you expect the future to exactly as bad as the absolute worst times of the past, not as bad as the past on average. 

See the difference?  I'm proposing that your expectations of the future should be based on the historical record, including lots of bad stuff but also lots of good stuff.  You seem to be suggesting that your expectations of the future should be based on the worst times in history you can find, and I don't think that's a reasonable expectation.

Unless you think the 2020s will be worse than the decades that included world wars and global depressions and presidential assassinations.  Not impossible, I just find it unlikely.

Scandium

  • Magnum Stache
  • ******
  • Posts: 2827
  • Location: EastCoast
Re: CFireSim Success Rate
« Reply #28 on: May 04, 2015, 11:03:44 AM »
I still don't understand this.

The idea is that if (i) you expect the future to be exactly as bad as the past and (ii) the chances that your own retirement start year will follow the path of any of history's previous retirement start years are exactly equal, then once you attain more than a 50% success rate your retirement is more likely than not to succeed, which means if you continue working you are more likely than not working longer than necessary.

OK, I see. But I think that agrees with what I wrote above too. Then if you happen to be in 49% of periods you are likely to fail too. "You're chance of not ending up in poverty is slightly better than a coin flip" is not really the level of assurance I'll aim for personally, but whatever floats your boat.

It's not like more money is a waste either. Sure, maybe I worked X years longer than I had to, but now I can buy myself a boat when I'm 65. I'd be ok with that outcome too.

Scandium

  • Magnum Stache
  • ******
  • Posts: 2827
  • Location: EastCoast
Re: CFireSim Success Rate
« Reply #29 on: May 04, 2015, 11:11:05 AM »
With a 100% success rate you would have survived every possible outcome that has occurred. I.e you expect the future to be exactly as bad as the past.

No, with a 100% success rate you expect the future to exactly as bad as the absolute worst times of the past, not as bad as the past on average. 

See the difference?  I'm proposing that your expectations of the future should be based on the historical record, including lots of bad stuff but also lots of good stuff.  You seem to be suggesting that your expectations of the future should be based on the worst times in history you can find, and I don't think that's a reasonable expectation.

Unless you think the 2020s will be worse than the decades that included world wars and global depressions and presidential assassinations.  Not impossible, I just find it unlikely.

aah ok. I understand what you mean now. I'm certainly not advocating for 100% success. But not sure I'm convinced I should aim for the average either. Which I think is generally the feeling on this board; most people aim for 80-95% numbers. This means you'd survive most of the bad events, but not the truly terrible ones (AKA: "The 60s"). Whichever approach is reasonable is obviously a personal decision. As an engineer most calculations end with x1.5 safety factor, but I'm trying to not go that crazy:)

NICE!

  • Pencil Stache
  • ****
  • Posts: 682
  • Location: Africa
Re: CFireSim Success Rate
« Reply #30 on: May 04, 2015, 11:12:46 AM »
Katrina was a speedbump in a second-tier city.

I agree with literally everything you've said except for classifying New Orleans as a second-tier city. The acquisition of New Orleans and the corresponding ability to ship our goods cheaply to the world via the Mississippi is one of the essential reasons that the United States is a superpower. Stratfor's George Freidman had a great article about this a few years back. The port of New Orleans is absolutely critical to America.

End of tangent!
« Last Edit: May 04, 2015, 01:01:59 PM by NICE! »

Zinsch

  • 5 O'Clock Shadow
  • *
  • Posts: 43
Re: CFireSim Success Rate
« Reply #31 on: May 04, 2015, 12:22:21 PM »
@sol: If you go for 50% success rate, then that translates to roughly a 6% withdrawal rate, right?
Is that what you use(d) for your retirement decision?

sol

  • Walrus Stache
  • *******
  • Posts: 8433
  • Age: 47
  • Location: Pacific Northwest
Re: CFireSim Success Rate
« Reply #32 on: May 04, 2015, 01:02:20 PM »
@sol: If you go for 50% success rate, then that translates to roughly a 6% withdrawal rate, right?
Is that what you use(d) for your retirement decision?

The short answer is that we're shooting for something above 5%, but there are many caveats to that number.

The longer answer is that percent SWR recommendations are just rules of thumb anyway, and I wouldn't advocate for anyone to make a retirement decision based solely on one number.

My personal situation, like everyone else's, is complicated by details.  My wife and I both have pensions and social security to collect one day.  We have real estate we intend to sell, providing cashflow for now and large lump sum influxes of cash in the future.  We have investments tied up in a variety of tax sheltered accounts that constrain how those funds can be accessed or spent. We will have the option to annuitize a portion of our investments upon retirement.  We have some investments, like solar panels, that produce real income each year but are not counted as part of our net worth.  We may or may not receive an inheritance one day, and we don't know how much money our children might need for college.  We also don't need our investments to support a traditional 30 year retirement, and SWRs rise for shorter time periods.

My retirement is planned out using a spreadsheet that lists all these numbers by month, along with our projected expenses, and then tries to figure out which sources can be accessed at what times to meet our anticipated living expenses.  It's a bit more complicated than "4% SWR means save 25x your expenses." 


brooklynguy

  • Handlebar Stache
  • *****
  • Posts: 2204
  • Age: 43
Re: CFireSim Success Rate
« Reply #33 on: May 04, 2015, 03:43:20 PM »
But I remain optimistic about the future for the same reasons you described, plus the recognition that if the shit hits the fan in one of these ways then the triviality of the success of my own retirement will be thrown into razor-sharp relief.

Also, it occurs to me that we may be improperly ignoring the other side of the coin when we focus on the downside exogenous risk factors (like sol's parade of horribles above) to the exclusion of the possibility that upside transformative dislocations will come to pass (which the recent discussions in the forum related to artificial intelligence have opened my eyes to).  Very smart and knowledgable people seriously believe that, due to the exponential nature of the rate of technological progress, we could be on the cusp of achieving previously unimaginable breakthroughs, like conquering our own mortality and manipulating the physical world on an atom-by-atom level.  If we're going to wade into the waters of "false precision," shouldn't we also consider the possibility that utopia will be realized?  Or am I now stretching the wild optimism of this thread beyond its breaking point?

Eric

  • Magnum Stache
  • ******
  • Posts: 4057
  • Location: On my bike
Re: CFireSim Success Rate
« Reply #34 on: May 04, 2015, 03:51:16 PM »
Or am I now stretching the wild optimism of this thread beyond its breaking point?

Your optimism gun is now an optimism raygun!  Pew pew!

Cpa Cat

  • Handlebar Stache
  • *****
  • Posts: 1692
Re: CFireSim Success Rate
« Reply #35 on: May 04, 2015, 04:12:38 PM »
Very smart and knowledgable people seriously believe that, due to the exponential nature of the rate of technological progress, we could be on the cusp of achieving previously unimaginable breakthroughs, like conquering our own mortality and manipulating the physical world on an atom-by-atom level.  If we're going to wade into the waters of "false precision," shouldn't we also consider the possibility that utopia will be realized?  Or am I now stretching the wild optimism of this thread beyond its breaking point?

Is it a best-case scenario or a worst-case scenario if I turn into a lich who feeds its immortality by sucking away the lifeforces of lesser humans?

I'm leaning toward best-case... but I wonder if 4% is appropriate?

neil

  • Stubble
  • **
  • Posts: 215
Re: CFireSim Success Rate
« Reply #36 on: May 04, 2015, 04:38:47 PM »
I agree some pretty bad stuff has happened that gives a lot of leeway with the 4% number.  But we also haven't lost a significant war and haven't fought much on actual US soil.

Germany's worst year is 1914 which I assume is a consequence of reparations after WWI.  1.14% withdrawal rate
Japan's worst is 1940. 0.47%.

Considering UK was shelled pretty heavily, their SWR (3.77) is not actually that bad.  So that would suggest being at war is fine for retirees as long as your country doesn't lose (badly).

Is our investment in the military a form of retirement planning?

I think the weird part about the idea of SWR is if you were to extend market data out 1000 years, eventually there would be some period where a generation of people could not "retire" on any amount of money.  But in any such scenarios, there are no winners and everyone else is in the same situation.  But I personally feel like that is also motivation to pull the trigger.  If I retired at 40 with the assumption I could retire under most conditions, and the world went to hell when I was 60, at least I got twenty years of quality retirement.

NICE!

  • Pencil Stache
  • ****
  • Posts: 682
  • Location: Africa
Re: CFireSim Success Rate
« Reply #37 on: May 04, 2015, 04:43:12 PM »
But in any such scenarios, there are no winners and everyone else is in the same situation.  But I personally feel like that is also motivation to pull the trigger.  If I retired at 40 with the assumption I could retire under most conditions, and the world went to hell when I was 60, at least I got twenty years of quality retirement.

Bingo! And after those 20 glorious years you'd go back to living with family and help raise whatever kids were around. You'd tell them about the glorious days of your retirement and teach them about how to FIRE in the new environment.

forummm

  • Walrus Stache
  • *******
  • Posts: 7374
  • Senior Mustachian
Re: CFireSim Success Rate
« Reply #38 on: May 04, 2015, 05:22:46 PM »
I agree some pretty bad stuff has happened that gives a lot of leeway with the 4% number.  But we also haven't lost a significant war and haven't fought much on actual US soil.

Germany's worst year is 1914 which I assume is a consequence of reparations after WWI.  1.14% withdrawal rate
Japan's worst is 1940. 0.47%.

Considering UK was shelled pretty heavily, their SWR (3.77) is not actually that bad.  So that would suggest being at war is fine for retirees as long as your country doesn't lose (badly).

Is our investment in the military a form of retirement planning?

I think the weird part about the idea of SWR is if you were to extend market data out 1000 years, eventually there would be some period where a generation of people could not "retire" on any amount of money.  But in any such scenarios, there are no winners and everyone else is in the same situation.  But I personally feel like that is also motivation to pull the trigger.  If I retired at 40 with the assumption I could retire under most conditions, and the world went to hell when I was 60, at least I got twenty years of quality retirement.

tl;dr
You have to diversify internationally.

sol

  • Walrus Stache
  • *******
  • Posts: 8433
  • Age: 47
  • Location: Pacific Northwest
Re: CFireSim Success Rate
« Reply #39 on: May 05, 2015, 03:07:30 PM »
You have to diversify internationally.

I'd definitely agree with the first four words. 

I'm not sure international exposure is the only viable option to get that diversification, though.  If you live in the US there are tons of other options for diversifying your investments.  Annuities and real estate, for example, can both provide a more protected steady cashflow that can offset most of the downside risk at the expense of the upside growth.

forummm

  • Walrus Stache
  • *******
  • Posts: 7374
  • Senior Mustachian
Re: CFireSim Success Rate
« Reply #40 on: May 05, 2015, 07:20:49 PM »
You have to diversify internationally.

I'd definitely agree with the first four words. 

I'm not sure international exposure is the only viable option to get that diversification, though.  If you live in the US there are tons of other options for diversifying your investments.  Annuities and real estate, for example, can both provide a more protected steady cashflow that can offset most of the downside risk at the expense of the upside growth.

People like to invoke Japan, so I'll be trendy and do that here. In the 80's Japan looked like it was ruling the business world and going to take over everything, had a rip-roaring economy, international business, the Nikkei was 40% of global market cap, yadda yadda. Well, the US has a lot of similarities there, and we're something like 55% of global market cap. Why can't it happen here? In 10 years or 20? For whatever reason. Some international diversification has historically improved returns and decreased volatility.

beltim

  • Magnum Stache
  • ******
  • Posts: 2957
Re: CFireSim Success Rate
« Reply #41 on: May 05, 2015, 07:38:58 PM »
You have to diversify internationally.

I'd definitely agree with the first four words. 

I'm not sure international exposure is the only viable option to get that diversification, though.  If you live in the US there are tons of other options for diversifying your investments.  Annuities and real estate, for example, can both provide a more protected steady cashflow that can offset most of the downside risk at the expense of the upside growth.

People like to invoke Japan, so I'll be trendy and do that here. In the 80's Japan looked like it was ruling the business world and going to take over everything, had a rip-roaring economy, international business, the Nikkei was 40% of global market cap, yadda yadda. Well, the US has a lot of similarities there, and we're something like 55% of global market cap. Why can't it happen here? In 10 years or 20? For whatever reason. Some international diversification has historically improved returns and decreased volatility.

It could.  But it would also probably happen after an unbelievable stock run-up.  For context, for the S&P 500 to reach the same p/e valuation as the Nikkei in 1989, it would have to go up more than fivefold from current levels.

It doesn't take much thought to realize that a 4% withdrawal rate after a fivefold increase in the stock market is nowhere near as safe as a 4% withdrawal rate calculated before that runup.

http://www.economist.com/node/9370662

OldPro

  • Stubble
  • **
  • Posts: 167
Re: CFireSim Success Rate
« Reply #42 on: May 05, 2015, 09:07:06 PM »
Who was it who said, 'life is what happens while you're busy making plans'?  I have to chuckle at people who think they can predict their future whether just financially or in any other way.

I also like the saying, 'man plans and the Gods laugh.'

brooklynguy

  • Handlebar Stache
  • *****
  • Posts: 2204
  • Age: 43
Re: CFireSim Success Rate
« Reply #43 on: May 05, 2015, 09:22:08 PM »
Who was it who said, 'life is what happens while you're busy making plans'?  I have to chuckle at people who think they can predict their future whether just financially or in any other way.

I also like the saying, 'man plans and the Gods laugh.'

And who was it that told a story about an ant and a grasshopper?  To paraphrase one of sol's refrains on these boards, imperfect planning is still better than no planning.

morning owl

  • Stubble
  • **
  • Posts: 230
  • Location: Canada
Re: CFireSim Success Rate
« Reply #44 on: May 06, 2015, 03:26:06 AM »
I have a question about this 80% success rate idea. I've been working towards a 100% success rate, simply because there are so many unknowns in the numbers I'm entering into cfiresim. For example, my income varies greatly, so I enter a conservative number for that (pretty much my lowest year's income as a freelancer.) We will be getting a significant inheritance at some point, but of course I have to guess when, so I'm conservative with that, as well as the amount. We will eventually sell our house and downsize, but with the canadian real estate market right now, that is a guess too, as we could be in a huge RE bubble. My DH is looking to retire soon, so I'd rather shoot for a higher success rate than a lower one. Am I being too conservative overall? It just doesn't seem like we have enough saved yet. With the 4% withdrawal rate, the actual money we have in investments gives us only about 1/3 what we would need to spend, not counting all these other variables.

When I enter all my "safe" guesses we get a success rate of about 95% with DH retiring next year, but there are still so many unknowns, and the actual money-in-the-bank figure seems low to me. It's scary to be thinking of withdrawing from this already.
« Last Edit: May 06, 2015, 03:27:48 AM by morning owl »

NICE!

  • Pencil Stache
  • ****
  • Posts: 682
  • Location: Africa
Re: CFireSim Success Rate
« Reply #45 on: May 06, 2015, 04:22:10 AM »
When I enter all my "safe" guesses we get a success rate of about 95% with DH retiring next year, but there are still so many unknowns, and the actual money-in-the-bank figure seems low to me. It's scary to be thinking of withdrawing from this already.

I think you need to figure out why you feel scared. Do you have no confidence in the other "safety margins" that MMM and other forum members have discussed, such as additional work, Social Security (or its equivalent elsewhere), downsizing, cutbacks, etc? Shouldn't that be more than enough, coupled with the 95% success rate based upon conservative inputs?

morning owl

  • Stubble
  • **
  • Posts: 230
  • Location: Canada
Re: CFireSim Success Rate
« Reply #46 on: May 06, 2015, 05:27:46 AM »
When I enter all my "safe" guesses we get a success rate of about 95% with DH retiring next year, but there are still so many unknowns, and the actual money-in-the-bank figure seems low to me. It's scary to be thinking of withdrawing from this already.

I think you need to figure out why you feel scared. Do you have no confidence in the other "safety margins" that MMM and other forum members have discussed, such as additional work, Social Security (or its equivalent elsewhere), downsizing, cutbacks, etc? Shouldn't that be more than enough, coupled with the 95% success rate based upon conservative inputs?

I'd say I'm uneasy because I live in a very HCOL area where for a normal retirement people think they need upwards of 2 million. By next year we'll be at roughly MMM-approved retirement savings amounts but we spend way more than MMM (about 3x). I don't see people mentioning inheritances too much on this site - it seems taboo to be using that as part of a retirement plan. And from what I inderstand canadian old age security is much less than US social security and should not be relied upon in retirement, either. (Though I used a conservative estimate for both in my cfiresim inputs.)

The good news is that I plan to work another 15-20 years because I enjoy what I do. So I'm hoping I can increase my income if I need to, or DH can potentially work part time or something as well. I am probably being over cautious. We would be pretty resourceful people if the shit ever hit the fan.


forummm

  • Walrus Stache
  • *******
  • Posts: 7374
  • Senior Mustachian
Re: CFireSim Success Rate
« Reply #47 on: May 06, 2015, 07:04:04 AM »
You have to diversify internationally.

I'd definitely agree with the first four words. 

I'm not sure international exposure is the only viable option to get that diversification, though.  If you live in the US there are tons of other options for diversifying your investments.  Annuities and real estate, for example, can both provide a more protected steady cashflow that can offset most of the downside risk at the expense of the upside growth.

People like to invoke Japan, so I'll be trendy and do that here. In the 80's Japan looked like it was ruling the business world and going to take over everything, had a rip-roaring economy, international business, the Nikkei was 40% of global market cap, yadda yadda. Well, the US has a lot of similarities there, and we're something like 55% of global market cap. Why can't it happen here? In 10 years or 20? For whatever reason. Some international diversification has historically improved returns and decreased volatility.

It could.  But it would also probably happen after an unbelievable stock run-up.  For context, for the S&P 500 to reach the same p/e valuation as the Nikkei in 1989, it would have to go up more than fivefold from current levels.

It doesn't take much thought to realize that a 4% withdrawal rate after a fivefold increase in the stock market is nowhere near as safe as a 4% withdrawal rate calculated before that runup.

http://www.economist.com/node/9370662

The market is at a CAPE of 27--right where it was a few months before Black Tuesday and the following Great Depression. And quite a few people on this forum are getting ready to retire now. Who knows what will happen. Maybe CAPE doesn't mean now what it used to. If people are willing to sacrifice returns to buy some portfolio stability by owning bonds they should be willing to maintain or increase returns and increase portfolio stability by owning cheap international funds too.

beltim

  • Magnum Stache
  • ******
  • Posts: 2957
Re: CFireSim Success Rate
« Reply #48 on: May 06, 2015, 07:15:01 AM »
You have to diversify internationally.

I'd definitely agree with the first four words. 

I'm not sure international exposure is the only viable option to get that diversification, though.  If you live in the US there are tons of other options for diversifying your investments.  Annuities and real estate, for example, can both provide a more protected steady cashflow that can offset most of the downside risk at the expense of the upside growth.

People like to invoke Japan, so I'll be trendy and do that here. In the 80's Japan looked like it was ruling the business world and going to take over everything, had a rip-roaring economy, international business, the Nikkei was 40% of global market cap, yadda yadda. Well, the US has a lot of similarities there, and we're something like 55% of global market cap. Why can't it happen here? In 10 years or 20? For whatever reason. Some international diversification has historically improved returns and decreased volatility.

It could.  But it would also probably happen after an unbelievable stock run-up.  For context, for the S&P 500 to reach the same p/e valuation as the Nikkei in 1989, it would have to go up more than fivefold from current levels.

It doesn't take much thought to realize that a 4% withdrawal rate after a fivefold increase in the stock market is nowhere near as safe as a 4% withdrawal rate calculated before that runup.

http://www.economist.com/node/9370662

The market is at a CAPE of 27--right where it was a few months before Black Tuesday and the following Great Depression. And quite a few people on this forum are getting ready to retire now. Who knows what will happen. Maybe CAPE doesn't mean now what it used to. If people are willing to sacrifice returns to buy some portfolio stability by owning bonds they should be willing to maintain or increase returns and increase portfolio stability by owning cheap international funds too.

Oh I don't disagree with that.  I'm just saying the example of the Nikkei is an extreme one - more like Nasdaq at the dot com bubble peak, rather than the S&P ever.  Although international equities are a strange beast - they have had lower returns than US markets, with higher risk.  They can increase portfolio stability, sure, but I wouldn't count on them to increase returns long term.

brooklynguy

  • Handlebar Stache
  • *****
  • Posts: 2204
  • Age: 43
Re: CFireSim Success Rate
« Reply #49 on: May 06, 2015, 08:12:48 AM »
It doesn't take much thought to realize that a 4% withdrawal rate after a fivefold increase in the stock market is nowhere near as safe as a 4% withdrawal rate calculated before that runup.

As recently as two months ago I would've agreed with this statement as being tautologically true, but, as you (beltim) recently pointed out here, as counterintuitive as it seems, retiring in the wake of a large market run-up has historically had zero impact on portfolio success rate.

Now, as we recognized in the linked thread, there are flaws in the statistical analysis you conducted that prevent us from drawing statistically significant conclusions (most prominently, the lack of sufficient data in the historical record), and perhaps if we run the same analysis again a few years from now (when the more recent market crashes are included in the data set) that answer will change.  But it just isn't true that retiring on a stash equal to a given withdrawal-multiple after a big market run-up is necessarily less safe than retiring on a stash equal to the same withdrawal-multiple before a big market run-up, because both options may turn out to be 100% safe -- the problem is that it's impossible to ever know whether that's the case in advance, before we have benefit of hindsight to confirm it for us.

 

Wow, a phone plan for fifteen bucks!