Author Topic: START worrying about the 4% rule  (Read 24275 times)

GenXbiker

  • Bristles
  • ***
  • Posts: 330
Re: START worrying about the 4% rule
« Reply #100 on: August 16, 2017, 10:25:33 AM »
I was just reviewing the 4% sticky thread which has been "cleaned up".  At one point, someone posted something similar to this, although edited here by me for clarification of what I'm curious about:

Quote
"pick data to look at risk of >21 CAPE only, use this calculator, under the investigate tab you can run simulations on individual years (ie you could pick the years only with CAPE >21).  Pick only the years of known CAPE >21.  Calculate the 4% WR success rate based on that sample and come back to tell us about your study."

I never saw that anyone actually ran that calculation to see the results vs. using the common larger sequential range historical sample.  There was another thread where someone had calculated the success rate to be 40% based on the fact that someone was more likely to hit their target and retire on a peak year than on a random year as used by the online calculators...

https://forum.mrmoneymustache.com/welcome-to-the-forum/cfiresim-severely-overestimates-success-rates-for-mustachians/

But I don't see any reference in that thread of him narrowing his historical sample to years with high CAPE, so it would be interesting to know how that changes the calculation, just out of curiosity as a mathematical exercise, not that it would change my confidence in using the 4% rule.

Runewell, are you still out there, and did you ever run that calculation?

While that 40% number is interesting and good analysis, it was assuming something like a 10% savings rate, which doesn't apply to many people on this forum.  The lower that rate, the bigger the market affects when you retire. 

In hindsight, I should not have bothered even mentioning that thread with the 40% figure because it really isn't approaching the calculation the way proposed in the 4% sticky thread edited quote I referenced.

Quote
The problem with comparing past CAPE's success rates is that accounting has changed relatively recently.  A CAPE of 21 today is not equivalent to a CAPE of 21 from 30 years ago.

Yes, I totally agree on CAPE 21 from 30 years ago not being equal to CAPE 21 today, and it has been mentioned quite a few times recently on this forum, but CAPE is around 30 today, so that calculation using 21 would give about a 10 point differential adjustment to compensate for those changes, which I think is why it was suggested to use 21 instead of 30 in the first place.  We don't know if that's the exact correct differential, but that's another reason why I mentioned that I'm just curious from a mathematical exercise standpoint, not putting any weight on what the results might be.   If it's an equally low or lower percentage, I still wouldn't worry about the 4% rule with the cushion I have planned for.
« Last Edit: August 16, 2017, 10:38:26 AM by GenXbiker »

Retire-Canada

  • Walrus Stache
  • *******
  • Posts: 6742
Re: START worrying about the 4% rule
« Reply #101 on: August 16, 2017, 10:26:27 AM »
Shiller himself says this:
https://www.nytimes.com/2017/03/31/upshot/trump-bull-market-stocks.html

I can't find where he says anything about accounting changes and CAPE at that link?

GenXbiker

  • Bristles
  • ***
  • Posts: 330
Re: START worrying about the 4% rule
« Reply #102 on: August 16, 2017, 10:45:20 AM »
The problem with comparing past CAPE's success rates is that accounting has changed relatively recently.  A CAPE of 21 today is not equivalent to a CAPE of 21 from 30 years ago.  It's also not obvious what it should compare to, other than something less than 21.  Shiller himself says this:
https://www.nytimes.com/2017/03/31/upshot/trump-bull-market-stocks.html

Shiller himself says this:
https://www.nytimes.com/2017/03/31/upshot/trump-bull-market-stocks.html

I can't find where he says anything about accounting changes and CAPE at that link?

Hmmm  Neither can I.  I see historical references to CAPE but nothing to invalidate or otherwise adjust them vs. today's CAPE due to accounting changes or otherwise.

Mr Mark

  • Handlebar Stache
  • *****
  • Posts: 1149
  • Location: Planet Earth
  • Achieved Financial Independence summer 2014. RE'18
Re: START worrying about the 4% rule
« Reply #103 on: August 16, 2017, 10:57:23 AM »
It's not just the issue of historic CAPE and accounting changes but also what about impact of QE and how dividends have been replaced by retained earnings and share buybacks. Or how many of our largest companies are now FAANG etc.

The market takes that into account too.

Using a subset of historical data to disprove results of other historical data when making predictions strikes me as self contradicted.

The sky is not falling. 

Dicey

  • Walrus Stache
  • *******
  • Posts: 8458
  • Age: 60
  • Location: NorCal
Re: START worrying about the 4% rule
« Reply #104 on: August 16, 2017, 11:15:05 AM »
Dear Mods,
Please add Mr. Mark to the Forum Heroes list. ♡♡♡♡♡♡♡♡♡♡♡
Kind regards,
Dicey

Note to GenXbiker: Please leave
Spoiler: show
runewell
out of this! Have you not been paying any attention? Trolls are not welcome here!

Mr Mark

  • Handlebar Stache
  • *****
  • Posts: 1149
  • Location: Planet Earth
  • Achieved Financial Independence summer 2014. RE'18
Re: START worrying about the 4% rule
« Reply #105 on: August 16, 2017, 11:31:17 AM »
Dear Mods,
Please add Mr. Mark to the Forum Heroes list. ♡♡♡♡♡♡♡♡♡♡♡
Kind regards,
Dicey

Note to GenXbiker: Please leave
Spoiler: show
runewell
out of this! Have you not been paying any attention? Trolls are not welcome here!

Awwwww. Cheers.

GenXbiker

  • Bristles
  • ***
  • Posts: 330
Re: START worrying about the 4% rule
« Reply #106 on: August 16, 2017, 11:55:10 AM »
Using a subset of historical data to disprove results of other historical data when making predictions strikes me as self contradicted.

Perhaps I wasn't clear, which led to you misunderstanding me, but I was NOT trying to "disprove" any previous results, I was just curious about the results with the same calculator using historical years with high CAPE to be "potentially" more representative of someone retiring with high CAPE "today" vs. a random year after reading the comment in the other thread of the proposal.  The "existing" results of the wider historical range of sequential years are still as valid as they have ever been despite someone doing an additional calculation using a subset of "potentially" more relevant data points for someone retiring "today".  Again, I wasn't trying to "disprove" any existing data.  As I've mentioned various times, I plan to use a 4% WR myself when I retire at some unknown date in the future with an unknown CAPE.  I'm a believer.

Quote
The sky is not falling.
Did I seem worried?  I specifically stated that I wasn't worried about the 4% rule and was curious only from a mathematical exercise standpoint after reading the suggestion in the other thread.  Nothing more.  Nothing less.   Nothing changes for me.

GenXbiker

  • Bristles
  • ***
  • Posts: 330
Re: START worrying about the 4% rule
« Reply #107 on: August 16, 2017, 12:01:33 PM »
Dear Mods,
Please add Mr. Mark to the Forum Heroes list. ♡♡♡♡♡♡♡♡♡♡♡
Kind regards,
Dicey

Note to GenXbiker: Please leave
Spoiler: show
runewell
out of this! Have you not been paying any attention? Trolls are not welcome here!
Mr. Mark seems to have misunderstood that I was only making a inquiry out of curiosity, not trying to disprove anything.  The sky falling was completely irrelevant - I never suggested anything like that.

Sorry, I thought that was more of a concern in the sticky thread since this alternate thread was started for him.

GenXbiker

  • Bristles
  • ***
  • Posts: 330
Re: START worrying about the 4% rule
« Reply #108 on: August 16, 2017, 12:06:29 PM »
I was just reviewing the 4% sticky thread which has been "cleaned up".  At one point, someone posted something similar to this, although edited here by me for clarification of what I'm curious about:

Quote
"pick data to look at risk of >21 CAPE only, use this calculator, under the investigate tab you can run simulations on individual years (ie you could pick the years only with CAPE >21).  Pick only the years of known CAPE >21.  Calculate the 4% WR success rate based on that sample and come back to tell us about your study."

I never saw that anyone actually ran that calculation to see the results vs. using the common larger sequential range historical sample.  There was another thread where someone had calculated the success rate to be 40% based on the fact that someone was more likely to hit their target and retire on a peak year than on a random year as used by the online calculators...

https://forum.mrmoneymustache.com/welcome-to-the-forum/cfiresim-severely-overestimates-success-rates-for-mustachians/

But I don't see any reference in that thread of him narrowing his historical sample to years with high CAPE, so it would be interesting to know how that changes the calculation, just out of curiosity as a mathematical exercise, not that it would change my confidence in using the 4% rule.

Runewell, are you still out there, and did you ever run that calculation?

Just ran the numbers here:

http://www.firecalc.com/

Inputs different from default:
$1,000,000 portfolio
$40,000 expenses (4% WR)
30 Years
100% stocks (I don't think it would be accurate to use 75% stocks / 25% bonds because current bond yields are lower than the historical averages)
Clicked option to provide results in a spreadsheet format (this provided the end value for all 30 years given start dates)

This provided a data set for the years 1 - 30 end values for the start years 1871-1987

Went here to gather Cape Shiller PE ratios from 1881-1987:

http://www.multpl.com/shiller-pe/table

I then calculated the success ratio for the complete data set, then the success ratio for years where the CAPE Shiller PE ratio was > 21

I then repeated the exercise and changed the expenses to $30,000 (3% WR) and $35,000 (3.5% WR)

Results reported in success ratio (# of start years where end portfolio > 0 after 30 years / number of years in sample):
4%
All years: 93%
CAPE>21: 55%

3.5%
All years: 99%
CAPE > 21: 91%

3% All years: 100%
CAPE > 21: 100%


Start Years where CAPE > 21:

1899
1902
1929
1930
1937
1962
1964
1965
1966
1968
1969

Excellent.  a BIG THANKS to Moustaches!  You went above and beyond.  That's all I was curious about.  Thank you.

Dicey

  • Walrus Stache
  • *******
  • Posts: 8458
  • Age: 60
  • Location: NorCal
Re: START worrying about the 4% rule
« Reply #109 on: August 16, 2017, 12:23:07 PM »
Dear Mods,
Please add Mr. Mark to the Forum Heroes list. ♡♡♡♡♡♡♡♡♡♡♡
Kind regards,
Dicey

Note to GenXbiker: Please leave
Spoiler: show
runewell
out of this! Have you not been paying any attention? Trolls are not welcome here!
Mr. Mark seems to have misunderstood that I was only making a inquiry out of curiosity, not trying to disprove anything.  The sky falling was completely irrelevant - I never suggested anything like that.

Sorry, I thought that was more of a concern in the sticky thread since this alternate thread was started for him.
I just discovered today, by fat-fingering a thread I don't typically follow, that
Spoiler: show
runewell
has a history of this troll-ish behavior. My request was a reminder not to feed the trolls. The person you mentioned bowed out of this thread after saying they weren't going to participate because it didn't have a sticky. It would be a more helpful discussion if we could keep it that way.

Dicey

  • Walrus Stache
  • *******
  • Posts: 8458
  • Age: 60
  • Location: NorCal
Re: START worrying about the 4% rule
« Reply #110 on: August 16, 2017, 12:52:35 PM »
If I'm such a troll, how come you are still talking about the topics I brought up?
Excellent non-sequitur!

maizeman

  • Magnum Stache
  • ******
  • Posts: 2866
Re: START worrying about the 4% rule
« Reply #111 on: August 16, 2017, 12:58:54 PM »
Don't forget that 1994, 1996, 1997, 1998, 1999, 2000, 2001, 2002, 2003, 2004, 2005, 2006, 2007, 2008, 2011, 2012, 2013, 2014, 2015, 2016, 2017 are all years where the CAPE was above 21. So far none of them seem particularly likely to result in failures (although we don't have a full 30 years of data for all of them yet obviously). However at this point it would take something significantly worse than the Great Depression for 1994, for example, to end in a failure for someone mechanically withdrawing 4%/year.

More years with CAPE above 21 in the last 23 years than in all the history of the US student co market up to that point. That's why I find the argument at accounting convincing.

Retire-Canada

  • Walrus Stache
  • *******
  • Posts: 6742
Re: START worrying about the 4% rule
« Reply #112 on: August 16, 2017, 01:04:40 PM »
I'm curious as to whether anyone is convinced by my analysis that a 4% SWR with a CAPE of 30 is dangerous, considering that my analysis showed that it failed 45% of the time when the CAPE is over 21 considering we are at 30 (and I doubt accounting adjustments would reduce it below 21).

Your analysis has reduced the data set to 11 simulations based on the starting years you indicated CAPE has been above 21. That's not much data to work with. We haven't resolved the accounting issues at all in my mind so I am not moved by the argument to this point.

I expect to FIRE with a higher than 4%WR, but as has been discussed to death in the "Don't worry...." thread there are so many ways to mitigate the risk to a 4% Rule based retirement plan that even that higher withdrawal rate doesn't cause me a lot of concern.

I am far more worried about working too long and health/death risks than I am about running out of money.

GenXbiker

  • Bristles
  • ***
  • Posts: 330
Re: START worrying about the 4% rule
« Reply #113 on: August 16, 2017, 01:21:10 PM »
Don't forget that 1994, 1996, 1997, 1998, 1999, 2000, 2001, 2002, 2003, 2004, 2005, 2006, 2007, 2008, 2011, 2012, 2013, 2014, 2015, 2016, 2017 are all years where the CAPE was above 21.

Correct, I couldn't put 1988-2017 as start years into my analysis because there hasn't been 30 years yet. 

I don't know the specific CAPE of those years, but since they are much more recent, I think a higher cut-off would be in order to correlate more closely with the earlier years of CAPE 21 and today's CAPE for the person retiring today.  That's if we actually had 30 years of historical data for those start years.

Retire-Canada

  • Walrus Stache
  • *******
  • Posts: 6742
Re: START worrying about the 4% rule
« Reply #114 on: August 16, 2017, 01:44:28 PM »
I'm afraid we can never resolve the accounting issues.  There is no way of having accurate numbers back to 1881 and normalizing for accounting standards.  The financial statements themselves are complex and hard to analyze.  There is accounting fraud.  There is just no freaking way we can do that.

Fair enough. I am not an expert in the area so I'll take your word for it. But the 11 simulations and that significant unresolved issue don't lead to a compelling argument in my mind regarding CAPE and 4%WR Rule failure.

NorthernBlitz

  • Bristles
  • ***
  • Posts: 297
Re: START worrying about the 4% rule
« Reply #115 on: August 16, 2017, 02:34:18 PM »
If you lower your WR to 2%, you guarantee that you will never run out of money in your stash, and guarantee that your retirement will never fail.

You also guarantee that your retirement will start many years later than if you use a higher WR, and you guarantee additional years working (and not retired).

At least you won't be able to worry about the 4% rule.

How much longer would you need to work to get from 4% (25x expenses) to 2% (50x expenses)?

Let's say that you're someone who makes $100k after taxes and you save 50% of that income and spend the other 50%.

You've got 25x expenses in your portfolio ($1.25M).

Every year, you save an additional $50k from your income. If you get 4% on your portfolio, you also save 1x your income in year 1.

Using this simple model, you'd reach 50x expenses in year 11.

If you plan on a 60 year retirement (closest thing to forever I could find with data) with 100% stocks, 4% has a success rate of 89%, 3.5% has a success rate of 98%, and 3% has a success rate of 100%.
https://earlyretirementnow.com/2016/12/07/the-ultimate-guide-to-safe-withdrawal-rates-part-1-intro/

Since 3% succeeds in 100% of the previous data, there is virtually no reason to go to 2% (and 11 extra years is a lot of time). The potential argument here is that there may not be enough data to draw conclusions with a 60 year time history (no one who retired after 1957 is in this data set). Asking for a 2% SWR is like building a strawman.

If you go from 4% to 3% (33.3x expenses), you hit your number in year 4.

The data in the link above suggests that 3% is the lowest SWR we should consider using (unless I guess you come into a bunch of money very early in life). The imaginary person above would need to work less than 4 years to go from failing in about 1 in 10 time periods (4% SWR) to failing in none of them (3% SWR). Not a bad trade for someone who likes their work, and wants to do all but guarentee they never have to work again.

If you go from 4% to 3.5% (28.6x expenses), you hit your number in year 2.

Getting from 4% to 3.5% requires less than 2 years for the case above. In this case you go from failing in about 1 in 10 time periods to failing 1 in 50 time periods. I think that this is probably what I'd do in this case. Doesn't totally eliminate risk, but doesn't make a significant impact if you think that advances in medicine will yield significant improvements in lifespan (try reading about CAR-T therapy that uses gene therapy to teach your immune system to attack tumors!).

From many people that I've read working after you've hit what you feel is your number can be pretty fun because you have so much more power to walk away.
- If you have a significant decline in those two years, you can still work until the market recovers. I'm pretty risk averse, so this is attractive to me.
- If the first year gives you 10%, you bank almost 4x expenses and you can walk away then.

Eric

  • Magnum Stache
  • ******
  • Posts: 3917
  • Location: On my bike
Re: START worrying about the 4% rule
« Reply #116 on: August 16, 2017, 04:31:32 PM »
I'm curious as to whether anyone is convinced by my analysis that a 4% SWR with a CAPE of 30 is dangerous, considering that my analysis showed that it failed 45% of the time when the CAPE is over 21 considering we are at 30 (and I doubt accounting adjustments would reduce it below 21).

Your analysis has reduced the data set to 11 simulations based on the starting years you indicated CAPE has been above 21. That's not much data to work with. We haven't resolved the accounting issues at all in my mind so I am not moved by the argument to this point.

I expect to FIRE with a higher than 4%WR, but as has been discussed to death in the "Don't worry...." thread there are so many ways to mitigate the risk to a 4% Rule based retirement plan that even that higher withdrawal rate doesn't cause me a lot of concern.

I am far more worried about working too long and health/death risks than I am about running out of money.

I disagree that the data set is only 11 simulations.  I used 107 data points and got a 93% success rate and 11 data points had a 55% success rate so clearly there is an impact when you look at CAPE > 21.  I've also done a scatter plot of all 107 data points and there is a clear inverse relationship between Shiller PE and final portfolio value.

I'm not sure there's a ton of value in setting an arbitrary cutoff of 21 for CAPE and projecting a success rate from that.  I'm also not sure that we can figure out that CAPE of 30 today is equal to CAPE of XX from 20, 30, or 60 years ago.  (note, here's the great post explaining the differences)  However, I don't think the preciseness of either of those really matter, because what we do know is that the higher these current valuations climb, the higher the likelihood that we're looking at lower returns for the near future.  And this is what worries me, even if I can't put a hard number on it.  The first 10 years are the most important to any retirement.

Speaking as someone on the cusp of retirement, I'd advise anyone else who is thisclose to exercise caution.  I'll also add that the money has been flowing freely the last couple of years, so you're likely ahead of your original projections so it shouldn't be too much sacrifice to discount some of these outsized returns and simply continue to work until reaching your number based on "normal" projected returns.  This way you've oversaved, but you didn't actually have to work "longer", simply the same amount of time as you were expecting.  You don't want to end up a 2000 retiree.

But just to contradict myself a bit and add another variable to the CAPE adjustments conversation, consider that soon 2008 & 2009 will drop out of the calculation.  Those were two of the lowest earning years that we've had in a long time.  When those are replaced with current years, the denominator "E" will increase, lowering the ratio.  I did the math a few months ago, and from memory, it was about a 2 point drop just from eliminating those low earning years.  So add this to the pile of "CAPE may be overinflated, but it's not as bad as it looks" reasons for optimism. 

ChpBstrd

  • Handlebar Stache
  • *****
  • Posts: 1092
Re: START worrying about the 4% rule
« Reply #117 on: August 16, 2017, 07:40:47 PM »
I'm curious as to whether anyone is convinced by my analysis that a 4% SWR with a CAPE of 30 is dangerous, considering that my analysis showed that it failed 45% of the time when the CAPE is over 21 considering we are at 30 (and I doubt accounting adjustments would reduce it below 21).

Your analysis has reduced the data set to 11 simulations based on the starting years you indicated CAPE has been above 21. That's not much data to work with. We haven't resolved the accounting issues at all in my mind so I am not moved by the argument to this point.

I expect to FIRE with a higher than 4%WR, but as has been discussed to death in the "Don't worry...." thread there are so many ways to mitigate the risk to a 4% Rule based retirement plan that even that higher withdrawal rate doesn't cause me a lot of concern.

I am far more worried about working too long and health/death risks than I am about running out of money.

I disagree that the data set is only 11 simulations.  I used 107 data points and got a 93% success rate and 11 data points had a 55% success rate so clearly there is an impact when you look at CAPE > 21.  I've also done a scatter plot of all 107 data points and there is a clear inverse relationship between Shiller PE and final portfolio value.

I'm not sure there's a ton of value in setting an arbitrary cutoff of 21 for CAPE and projecting a success rate from that.  I'm also not sure that we can figure out that CAPE of 30 today is equal to CAPE of XX from 20, 30, or 60 years ago.  (note, here's the great post explaining the differences)  However, I don't think the preciseness of either of those really matter, because what we do know is that the higher these current valuations climb, the higher the likelihood that we're looking at lower returns for the near future.  And this is what worries me, even if I can't put a hard number on it.  The first 10 years are the most important to any retirement.

Speaking as someone on the cusp of retirement, I'd advise anyone else who is thisclose to exercise caution.  I'll also add that the money has been flowing freely the last couple of years, so you're likely ahead of your original projections so it shouldn't be too much sacrifice to discount some of these outsized returns and simply continue to work until reaching your number based on "normal" projected returns.  This way you've oversaved, but you didn't actually have to work "longer", simply the same amount of time as you were expecting.  You don't want to end up a 2000 retiree.

But just to contradict myself a bit and add another variable to the CAPE adjustments conversation, consider that soon 2008 & 2009 will drop out of the calculation.  Those were two of the lowest earning years that we've had in a long time.  When those are replaced with current years, the denominator "E" will increase, lowering the ratio.  I did the math a few months ago, and from memory, it was about a 2 point drop just from eliminating those low earning years.  So add this to the pile of "CAPE may be overinflated, but it's not as bad as it looks" reasons for optimism.

Here's a tool to calculate CAPE on different timeframes. TLDR version: CAPE increases as you calculate it based on more years, because in general earnings have risen over time.

https://dqydj.com/shiller-pe-cape-ratio-calculator/

The point about earnings being based more on accounting regulations is well-taken. However, this does not necessarily mean companies have an incentive to minimize earnings. In the US, companies are allowed to report one earnings number to investors and another for tax purposes, a fact that blew the minds of the international students in my finance class years ago. "How is that not corrupt?!" If anything, companies try to boost earnings to improve the career prospects and bonuses of executives.

Another thing I learned in finance class: The cash flow don't lie - unless someone's going to jail. It sometimes makes more sense to analyze the stock market based on free cash flow than  earnings. Here's an article written this month talking about a side topic but pointing out that "today the free cash flow yield for the S&P 500 is currently 4.2%, near the average since 1990." That's certainly a mind-blowing observation about stock valuation, especially considering all the other metrics that seem historically high.

http://www.valuewalk.com/2017/08/free-cash-flow-yield-irrelevant/

Also, if I was going to worry about something, I'd worry about inflation surprising everybody and hitting a very historically normal 3.5% next year. The 10-year treasury might hit 5%. Banks and funds holding treasuries and corporate bonds would collapse, as billions of dollars worth of their long-dated assets lost double-digit percentages. A 2% increase in interest rates would also cause the payments on typical $250k 30y mortgages to increase by over 18%, potentially forcing prices down, potentially leading to another wave of defaults in the most frothy markets, leading to market panic. Meanwhile, investors using DCF analysis on either earnings or free cash flows would change their risk-free rate and get radically lower valuations for stocks. Pull out the assumption of low interest rates / low inflation, and the value of everything collapses.

Mr Mark

  • Handlebar Stache
  • *****
  • Posts: 1149
  • Location: Planet Earth
  • Achieved Financial Independence summer 2014. RE'18
Re: START worrying about the 4% rule
« Reply #118 on: August 17, 2017, 01:43:25 AM »
... I'm curious as to whether anyone is convinced by my analysis that a 4% SWR with a CAPE of 30 is dangerous, considering that my analysis showed that it failed 45% of the time when the CAPE is over 21 considering we are at 30 (and I doubt accounting adjustments would reduce it below 21). 
...

"dangerous" ? I'm not convinced at all. Why? Because it's a selective and arbitrary subset based on... 1 poorly estimated parameter [CAPE] combined with 100% equity [again, based on... a vague opinion that bond yields are too low even tho 25-50% bonds historically have increased success rates vs 100% equity] and rigid withdrawals. (But to clarify GenXBiker, it was an interesting experiment, and I do appreciate the effort!)

The Trinity Study et al - using all available data at the time - shows that increased withdrawal % lowers your historical portfolio survival chances, assuming rigid withdrawals and US equity/bond mix. It concludes (my summary) a <3.5% WR is really safe (100%), 4% safe-ish (95%), and it gets worse as you go higher. And because it looked over long periods in the past this takes account of all sorts of nasty events (wars, high inflation, multiple recessions and market crashes, techology, etc etc). OK. Hence "the 4% rule".

BUT...
The current CAPE 'markets are overvalued' argument is not a secret. Yet the market continues to climb. I just don't think you can use this one parameter in isolation from the myriad other key points that impact investment returns (and clearly neither does the market). Things that seem a bit different lately to historic norms:
- bond yields are historically very low (although still not as low as Germany or Japan)
- one off accounting changes that have increased CAPE ratios since 2001 (by about 20%)
- GFC effect increasing current CAPE (because of 10 yr lookback effect)
- higher earnings growth yet fewer dividend payments
- Quantitative easing
- increased share buybacks
- high valued big tech companies that don't currently make much in earnings
- higher international earnings % for US companies

Thus the impossibility of relying on any historic analysis to confidently predict the future with the sort of precision some people seems to be relying on. The solution IMHO is not to fear high valuations and target a 3% WR by working a lot longer, but to prepare to be a bit flexible* wrt expenditure and go with 4%. Heck, if I just plug my expected Social Security payments into FIREsim I can go to 5%. Some decent rental property might push it even higher.
 
'The sky is not falling' comment is because so much commentary seems to be trying to make people afraid. I prefer to deploy the MMM Optimism Gun http://www.mrmoneymustache.com/2012/10/03/the-practical-benefits-of-outrageous-optimism/


*Now, if I was say, an actuary responsible for a company pension fund where a resulting need for 'flexibility' in payouts would get me tarred and feathered, OK, I can see why reducing forward total real return expectations down a bit is a wise recommendation given current high CAPE, lower bond yields and rising life expectancy. And maybe adding some international equity because their PE ratios are a lot lower.



steveo

  • Handlebar Stache
  • *****
  • Posts: 1646
Re: START worrying about the 4% rule
« Reply #119 on: August 17, 2017, 03:23:38 AM »
Mr Mark - good post.

Thus the impossibility of relying on any historic analysis to confidently predict the future with the sort of precision some people seem to be relying on.

This is where I honestly think people who are into data can really struggle. I work in databases and hacking data up has been a big part of my career. I've also traded and tried to trade via creating a system. I don't think data or rule based systems work very well at all when it comes to predicting the future. It is fun (as sad as this sounds) to analyse data and come up with smart inferences that you think will somehow let you control the future but it doesn't work out that way.

I agree that saving more money will increase your chances of not running out of money in retirement. I also think lots of other decisions can increase your chances of not running out of money in retirement. The point though is are you prepared to work longer to increase the chances of having a successful retirement from a money perspective when you could (more likely probably) be wrong. So the extra money comes with the trade off of a shorter retirement. The real question is are you prepared to take that trade. This of course assumes that you don't need to up your expenses. As previously stated though if it's an expense issue it doesn't have anything to do with the 4% rule.
« Last Edit: August 17, 2017, 03:25:12 AM by steveo »

Mr Mark

  • Handlebar Stache
  • *****
  • Posts: 1149
  • Location: Planet Earth
  • Achieved Financial Independence summer 2014. RE'18
Re: START worrying about the 4% rule
« Reply #120 on: August 17, 2017, 06:12:35 AM »
Mr Mark - good post.

Thus the impossibility of relying on any historic analysis to confidently predict the future with the sort of precision some people seem to be relying on.

This is where I honestly think people who are into data can really struggle. I work in databases and hacking data up has been a big part of my career. I've also traded and tried to trade via creating a system. I don't think data or rule based systems work very well at all when it comes to predicting the future. It is fun (as sad as this sounds) to analyse data and come up with smart inferences that you think will somehow let you control the future but it doesn't work out that way.

I agree that saving more money will increase your chances of not running out of money in retirement. I also think lots of other decisions can increase your chances of not running out of money in retirement. The point though is are you prepared to work longer to increase the chances of having a successful retirement from a money perspective when you could (more likely probably) be wrong. So the extra money comes with the trade off of a shorter retirement. The real question is are you prepared to take that trade. This of course assumes that you don't need to up your expenses. As previously stated though if it's an expense issue it doesn't have anything to do with the 4% rule.

cheers Steveo

I still like a quote from one of my (retired) colleagues: "you only get to control the length of your retirement by when you start"

OMY = guarenteed 1 less year of retirement

GenXbiker

  • Bristles
  • ***
  • Posts: 330
Re: START worrying about the 4% rule
« Reply #121 on: August 17, 2017, 04:51:25 PM »
(But to clarify GenXBiker, it was an interesting experiment, and I do appreciate the effort!)


Thanks.  And I definitely will have some flexibility in my WR.

Quote
OMY = guarenteed 1 less year of retirement

<2MY for me, I hope.

JohnSteed

  • 5 O'Clock Shadow
  • *
  • Posts: 15
Re: START worrying about the 4% rule
« Reply #122 on: August 20, 2017, 09:10:47 AM »
Mr Mark - good post.

Thus the impossibility of relying on any historic analysis to confidently predict the future with the sort of precision some people seem to be relying on.

This is where I honestly think people who are into data can really struggle. I work in databases and hacking data up has been a big part of my career. I've also traded and tried to trade via creating a system. I don't think data or rule based systems work very well at all when it comes to predicting the future. It is fun (as sad as this sounds) to analyse data and come up with smart inferences that you think will somehow let you control the future but it doesn't work out that way.

So you think CAPE is flawed.  Even though Vanguard says it is correlated with future returns, it's possible that the current CAPE isn't quite as bad as the number indicates.

There ar other studies out there that suggest that P/B ratios or average investor allocation to equities.  Someone put up a graph showing a 90% historical correlation between average investor allocation to equities and future 10-yr stock returns.  Nobody seemed to have a good answer as to why that should be dismissed.  You've got to admit a 90% correlation across a 60+ yr time frame runs circles around the CAPE indicator.

I think that what is needed is multiple indicators of market conditions.  To dismiss other data-driven studies yet hold on to a 4% rule that is based on historical conditions which don't resemble today's situation seems like a giant contradicton. 

Mr Mark

  • Handlebar Stache
  • *****
  • Posts: 1149
  • Location: Planet Earth
  • Achieved Financial Independence summer 2014. RE'18
Re: START worrying about the 4% rule
« Reply #123 on: August 20, 2017, 10:50:24 PM »
correlation doesn't prove causation. The increase in investor allocation could be due to the increasing temptation of a bull market run and reduced attractiveness of alternatives (esp bond yields). It's just showing the typical mistake of retail investors who tend to buy too late into the bull run and sell after the drops/bear run. Plus, by definition, as the market goes up (if they don't rebalance) the equity allocation % rises and visa-versa after a big drop.

But again, what's the alternative? Sit on cash and try to time the market? LOL. We have a LOT of data that that is a losing strategy. If people want to work longer and have a 3% SWR, hey, their call.

I think a greater exposure for the FIREee is inflation and changes to tax law. The tax one especially is an exposure IMHO, as not many people are aware that under current US tax law one can pull $90k/yr in cap gains and dividends and pay no federal tax. Or the huge tax loopholes given to real estate landlords wrt depreciation and defered cap gains. Or using IRA-> roth conversion ladders to avoid tax. I think if 'ordinary' voters realised that you can be a millionaire living a cozy ER life and pay ZERO taxes all while getting subsidised health care, while they are slaving away paycheck to paycheck with no savings and paying significant taxes, they might support some changes to the tax code.

Some countries [Netherlands] even have a wealth tax that takes a % of your assets every year.

JohnSteed

  • 5 O'Clock Shadow
  • *
  • Posts: 15
Re: START worrying about the 4% rule
« Reply #124 on: August 27, 2017, 02:45:02 AM »
correlation doesn't prove causation.

I wasn't trying to prove causation.  If something else causes both, it's still worrying.

Quote
The increase in investor allocation could be due to the increasing temptation of a bull market run and reduced attractiveness of alternatives (esp bond yields).

I can believe this.  If so, equity prices can't remain stretched forever.

Quote
But again, what's the alternative? Sit on cash and try to time the market? LOL. We have a LOT of data that that is a losing strategy. If people want to work longer and have a 3% SWR, hey, their call.

Perhaps other nonUS markets are less expensive. 

Perhaps there is no alternative except to recognize that expected returns will be lower for the foreseeable future and assume that the next 10-15 years is projected to be lower returns.

Perhaps the viability of social security and pensions will be put into question and the government continues to run up debt.  Reductions, whenever they may come, will significantly affect consumer spending.



CanuckExpat

  • Magnum Stache
  • ******
  • Posts: 2727
  • Age: 35
  • Location: Travelling
    • Freedom35
Re: START worrying about the 4% rule
« Reply #125 on: October 31, 2017, 11:05:15 AM »
Someone who may agree about worrying about the 4% Rule: Hi, I'm Wade Pfau Ask me Anything

DrF

  • Bristles
  • ***
  • Posts: 464
Re: START worrying about the 4% rule
« Reply #126 on: October 31, 2017, 11:45:57 AM »
ERN at earlyretirementnow.com has delved into SWR quite extensively and his analyses suggest between 3-3.5% is a safe SWR going forward. With some trickery, he has shown how to get it up to ~3.75% or so.

Le Barbu

  • Handlebar Stache
  • *****
  • Posts: 1033
  • Age: 46
  • Location: Québec
  • I really didn’t say everything I said - Y. B.
Re: START worrying about the 4% rule
« Reply #127 on: August 05, 2018, 04:23:37 PM »
1 year since I started this thread...

Portfolio delivered almost 12% YOY and NW increased by +/-15%

Average PE of my index ETFs are the same than 2 years ago

We must be close to a drop anytime soon!

Retire-Canada

  • Walrus Stache
  • *******
  • Posts: 6742
Re: START worrying about the 4% rule
« Reply #128 on: August 06, 2018, 07:05:18 AM »
We must be close to a drop anytime soon!

In another thread on these forums I read that the Top Was In! ;)

Malkynn

  • Handlebar Stache
  • *****
  • Posts: 1218
Re: START worrying about the 4% rule
« Reply #129 on: August 06, 2018, 07:11:29 AM »
We must be close to a drop anytime soon!

In another thread on these forums I read that the Top Was In! ;)

LOL!