Author Topic: SWR and Current Volatility  (Read 1883 times)

Montecarlo

  • Bristles
  • ***
  • Posts: 498
SWR and Current Volatility
« on: March 09, 2020, 08:00:24 PM »
Has current volatility changed anyone’s mindset about SWR?  We’re not in unprecedented territory yet, but if we get a 50% correction, I believe that’s the first for 3 in a 20 year time frame and I wonder how healthy a 4% 2000 early retiree  feels right now

maizefolk

  • Magnum Stache
  • ******
  • Posts: 4849
Re: SWR and Current Volatility
« Reply #1 on: March 09, 2020, 09:11:48 PM »
Has current volatility changed anyone’s mindset about SWR?

Not at all.

Quote
We’re not in unprecedented territory yet, but if we get a 50% correction, I believe that’s the first for 3 in a 20 year time frame and I wonder how healthy a 4% 2000 early retiree  feels right now

This is an easy trap to fall into: assuming a hypothetical outcome and then arguing that hypothetical outcome is so unexpected that something must be wrong about our assumptions.

We'll cross a 50% drop if it comes to that, but 20% drop happen fairly frequently and while they sometimes turn into 50% drops they more frequently do not.

If anything the long stretch from mostly upward stock market progress from '09-'20 is more historically unprecedented than anything else we've seen in the 21st century so far. 

MustacheAndaHalf

  • Magnum Stache
  • ******
  • Posts: 2998
Re: SWR and Current Volatility
« Reply #2 on: March 09, 2020, 09:21:24 PM »
I originally planned to adjust my SWR up or down slightly each year, based on stock market performance.  So maybe at a 4% withdrawal rate, it would be +0.50% when markets are strong, and -0.50% during a bad year.  But ultimately that was overly complicated and probably wasn't worth much, so I just used a simple withdrawal rate.

I've heard of another approach using buckets, where during downturns you spend from your cash (crash?) bucket.  Then you have a bonds bucket for years after that, and a stock bucket for the distant future.  Each bucket has a different time frame.

Montecarlo

  • Bristles
  • ***
  • Posts: 498
Re: SWR and Current Volatility
« Reply #3 on: March 10, 2020, 05:18:05 AM »
Take the hypothetical out of it.  S&P 500 is up 3.3% CAGR over past 20 years.  That sucks!  Doesn’t account for dividends, but call it 5-6%?  Add in the a SORR of a 50% haircut and a lost decade+ of basically no returns.  How much of the starting portfolio is left from someone who followed a strict 4% rule in early 2000?

AdrianC

  • Handlebar Stache
  • *****
  • Posts: 1131
  • Location: Cincinnati
Re: SWR and Current Volatility
« Reply #4 on: March 10, 2020, 06:46:29 AM »
How much of the starting portfolio is left from someone who followed a strict 4% rule in early 2000?
Not much. Try it:
https://www.portfoliovisualizer.com/backtest-portfolio

Using 90% SPY (for S&P500) and 10% Cash, $1M portfolio, they have an inflation adjusted $283k remaining. But no one would follow a strict 4% rule in such circumstances.

Starting in 2008, which is dear to my heart, they would have an inflation adjusted $1.16M remaining. Darn it!

ender

  • Walrus Stache
  • *******
  • Posts: 5577
Re: SWR and Current Volatility
« Reply #5 on: March 10, 2020, 06:52:20 AM »
Take the hypothetical out of it.  S&P 500 is up 3.3% CAGR over past 20 years.  That sucks!  Doesn’t account for dividends, but call it 5-6%?  Add in the a SORR of a 50% haircut and a lost decade+ of basically no returns.  How much of the starting portfolio is left from someone who followed a strict 4% rule in early 2000?

I don't know that it sucks at all.

If you told me that you could guarantee me 5% CAGR for the first 20 years after I retire, that'd be awesome. I plan on retiring around age 40, which means in this hypothetical I'd be 60 with about $300k in my portfolio as per @AdrianC's estimate.

Given that social security would be close to available for me at that point it seems like that would be pretty good still, even given that I was too dumb to realize SORR at any point in the 20 year early retirement.

Now if I was blindly withdrawing 4% every year without ever considering SORR? Yeah, that'd be unfortunate. But even in this "worse case" scenario it's still a pretty successful retirement for anyone who retired around age 40 or so (unless of course they were going more of a "fat fire" approach, but then they are complete idiots if they blindly spent the 4%).

Montecarlo

  • Bristles
  • ***
  • Posts: 498
Re: SWR and Current Volatility
« Reply #6 on: March 10, 2020, 10:46:43 AM »
But no one would follow a strict 4% rule in such circumstances.

This is the sort of tautological argument that always gets thrown out there.

"4% is always safe!"
"What about this real-life scenario?"
"Well, no one would stick with 4% then"

bacchi

  • Walrus Stache
  • *******
  • Posts: 5015
Re: SWR and Current Volatility
« Reply #7 on: March 10, 2020, 11:33:23 AM »
But no one would follow a strict 4% rule in such circumstances.

This is the sort of tautological argument that always gets thrown out there.

"4% is always safe!"
"What about this real-life scenario?"
"Well, no one would stick with 4% then"

The 4% SWR is not 100% infallible. First, it's based on historical data. Second, it's only 95% safe.

I don't know what "real-life" you're talking about but, to me, it means what most sensible people would do in that scenario. Given that, if one retired in 1966, it would make sense for someone to cut their expenses in the 1970s stagflation period.

To put it another way, this could be one of the 5% failures. It's not yet there.

bacchi

  • Walrus Stache
  • *******
  • Posts: 5015
Re: SWR and Current Volatility
« Reply #8 on: March 10, 2020, 11:58:00 AM »
How much of the starting portfolio is left from someone who followed a strict 4% rule in early 2000?
Not much. Try it:
https://www.portfoliovisualizer.com/backtest-portfolio

Using 90% SPY (for S&P500) and 10% Cash, $1M portfolio, they have an inflation adjusted $283k remaining. But no one would follow a strict 4% rule in such circumstances.

Starting in 2008, which is dear to my heart, they would have an inflation adjusted $1.16M remaining. Darn it!

About halfway down is the 2018 graph with a 60/40 portfolio.

https://www.bogleheads.org/forum/viewtopic.php?f=10&t=237334

It looks like the Y2k retiree will be able to make it to 2030.

MustacheAndaHalf

  • Magnum Stache
  • ******
  • Posts: 2998
Re: SWR and Current Volatility
« Reply #9 on: March 10, 2020, 01:59:06 PM »
Take the hypothetical out of it.  S&P 500 is up 3.3% CAGR over past 20 years.  That sucks!  Doesn’t account for dividends, but call it 5-6%?  Add in the a SORR of a 50% haircut and a lost decade+ of basically no returns.  How much of the starting portfolio is left from someone who followed a strict 4% rule in early 2000?
The past 20 years starts in 2000, which starts with the dot-com crash followed by 9/11.  And neatly misses the bull market 1990s that lead up to that point (with a +14.7% CAGR).  Or put another way:

30 year CAGR: +6.9%
20 year CAGR: +3.3%
10 year CAGR: +10.3%

https://www.portfoliovisualizer.com/backtest-asset-class-allocation

Telecaster

  • Handlebar Stache
  • *****
  • Posts: 2315
  • Location: Seattle, WA
Re: SWR and Current Volatility
« Reply #10 on: March 10, 2020, 03:13:01 PM »
This is the sort of tautological argument that always gets thrown out there.

"4% is always safe!"


I've never seen anyone knowledgeable on the topic say that 4% is always safe. 

FIRE 20/20

  • Bristles
  • ***
  • Posts: 462
Re: SWR and Current Volatility
« Reply #11 on: March 10, 2020, 03:55:13 PM »
But no one would follow a strict 4% rule in such circumstances.

This is the sort of tautological argument that always gets thrown out there.

"4% is always safe!"
"What about this real-life scenario?"
"Well, no one would stick with 4% then"

I've never seen this kind of argument here.  I suspect if you can find something like that, you'll also find a bunch of people correcting whoever makes it. 

maizefolk

  • Magnum Stache
  • ******
  • Posts: 4849
Re: SWR and Current Volatility
« Reply #12 on: March 10, 2020, 04:21:03 PM »
The only people I've seen arguing that the 4% rule says you will 100% safe all of the time if you follow a 4% WR strategy religiously are people trying to argue that the 4% rule is broken. We have a pinned thread for those discussions.

Montecarlo

  • Bristles
  • ***
  • Posts: 498
Re: SWR and Current Volatility
« Reply #13 on: March 10, 2020, 07:53:44 PM »
This is the sort of tautological argument that always gets thrown out there.

"4% is always safe!"


I've never seen anyone knowledgeable on the topic say that 4% is always safe.

That’s also highly tautological

Telecaster

  • Handlebar Stache
  • *****
  • Posts: 2315
  • Location: Seattle, WA
Re: SWR and Current Volatility
« Reply #14 on: March 10, 2020, 08:42:59 PM »
This is the sort of tautological argument that always gets thrown out there.

"4% is always safe!"


I've never seen anyone knowledgeable on the topic say that 4% is always safe.

Let's get down to brass tacks then.  Who specifically said 4% is always safe with no qualifiers?

If the answer is "nobody" then you are just blowing smoke. 

That’s also highly tautological

AdrianC

  • Handlebar Stache
  • *****
  • Posts: 1131
  • Location: Cincinnati
Re: SWR and Current Volatility
« Reply #15 on: March 10, 2020, 09:15:31 PM »
But no one would follow a strict 4% rule in such circumstances.

This is the sort of tautological argument that always gets thrown out there.

"4% is always safe!"
"What about this real-life scenario?"
"Well, no one would stick with 4% then"
Would you continue to pull out your inflation adjusted 4% after your portfolio was cut in half less than a year into retirement? I wouldn’t. I’d be updating the resume.

You realize the 4% rule is not 100% successful historically, yes?

reeshau

  • Pencil Stache
  • ****
  • Posts: 887
  • Location: Houston, TX
Re: SWR and Current Volatility
« Reply #16 on: March 11, 2020, 09:44:36 AM »
But no one would follow a strict 4% rule in such circumstances.

This is the sort of tautological argument that always gets thrown out there.

"4% is always safe!"
"What about this real-life scenario?"
"Well, no one would stick with 4% then"
Would you continue to pull out your inflation adjusted 4% after your portfolio was cut in half less than a year into retirement? I wouldn’t. I’d be updating the resume.

You realize the 4% rule is not 100% successful historically, yes?

Actually, it is.  By definition, Bengen was looking for the rate that was successful through all historical 30-year periods he could test, which led him through periods beginning in the late 60's.  Even now, the only way to prove or disprove it, historically, is to wait 30 years:  so we are only up to retirement starts in 1990.  Pretty smooth sailing for another decade, I would say.  And Bengen's definiton of success is remaining above 0 after 30 years.

I think what you meant was *statistically* it isn't 100% successful, meaning tested against a Monte Carlo simulation.  If so, then agreed.

AdrianC

  • Handlebar Stache
  • *****
  • Posts: 1131
  • Location: Cincinnati
Re: SWR and Current Volatility
« Reply #17 on: March 11, 2020, 11:46:15 AM »
But no one would follow a strict 4% rule in such circumstances.

This is the sort of tautological argument that always gets thrown out there.

"4% is always safe!"
"What about this real-life scenario?"
"Well, no one would stick with 4% then"
Would you continue to pull out your inflation adjusted 4% after your portfolio was cut in half less than a year into retirement? I wouldn’t. I’d be updating the resume.

You realize the 4% rule is not 100% successful historically, yes?

Actually, it is.  By definition, Bengen was looking for the rate that was successful through all historical 30-year periods he could test, which led him through periods beginning in the late 60's.  Even now, the only way to prove or disprove it, historically, is to wait 30 years:  so we are only up to retirement starts in 1990.  Pretty smooth sailing for another decade, I would say.  And Bengen's definiton of success is remaining above 0 after 30 years.

I think what you meant was *statistically* it isn't 100% successful, meaning tested against a Monte Carlo simulation.  If so, then agreed.
No. I mean it did fail for some start years. Go to cFIREsim and try it for yourself.

bacchi

  • Walrus Stache
  • *******
  • Posts: 5015
Re: SWR and Current Volatility
« Reply #18 on: March 11, 2020, 12:22:29 PM »
But no one would follow a strict 4% rule in such circumstances.

This is the sort of tautological argument that always gets thrown out there.

"4% is always safe!"
"What about this real-life scenario?"
"Well, no one would stick with 4% then"
Would you continue to pull out your inflation adjusted 4% after your portfolio was cut in half less than a year into retirement? I wouldn’t. I’d be updating the resume.

You realize the 4% rule is not 100% successful historically, yes?

Actually, it is.  By definition, Bengen was looking for the rate that was successful through all historical 30-year periods he could test, which led him through periods beginning in the late 60's.  Even now, the only way to prove or disprove it, historically, is to wait 30 years:  so we are only up to retirement starts in 1990.  Pretty smooth sailing for another decade, I would say.  And Bengen's definiton of success is remaining above 0 after 30 years.

I think what you meant was *statistically* it isn't 100% successful, meaning tested against a Monte Carlo simulation.  If so, then agreed.
No. I mean it did fail for some start years. Go to cFIREsim and try it for yourself.

The Trinity study does have some historical failures. It of course depends on your asset allocation.

ChpBstrd

  • Handlebar Stache
  • *****
  • Posts: 2330
Re: SWR and Current Volatility
« Reply #19 on: March 11, 2020, 01:07:44 PM »
Has current volatility changed anyone’s mindset about SWR?  We’re not in unprecedented territory yet, but if we get a 50% correction, I believe that’s the first for 3 in a 20 year time frame and I wonder how healthy a 4% 2000 early retiree  feels right now

It has me wondering if there are algorithmic moves one can make at 40-50% down that would seize certain assets for once-in-a-lifetime prices and retire on a 4-5% WR despite the US economy entering a Japan scenario. As someone who hedged before the correction, and has thus suffered less than half the paper losses, a pivot to a more aggressive portfolio near a probable bottom could seal the deal.

For example, nursing home REITs like OHI and SBRA are getting their arses kicked based on fears COVID-19 will turn them all into death camps. Dividend yields and p/ffo are over 10%. Could a portfolio of these cover a 5% WR, despite future market returns, assuming the panic is misplaced?

Similarly, we are about another 10-15% drop away from preferred stocks (PFF, PGF, PFXF) yielding in the 6.5% range. Could some of these cover a 4% WR even if financials languish for decades?

Then, near an actual bottom, there will be lots of companies with ROEs over 20% and PE’s under 10.

In sum, I’m thinking falling knives might not be so dangerous ;)

projekt

  • Bristles
  • ***
  • Posts: 337
Re: SWR and Current Volatility
« Reply #20 on: March 11, 2020, 03:56:14 PM »
If you are withdrawing 4%, that seems to be pretty safe for never running out of money, but it might not cover your expenses year-to-year. Luckily, as a retiree, you have several options:

1. Withdraw more this year, then when economics pick up, work for a year and refill your coffers.
2. Withdraw 4% and work a small job part time to make up the difference, assuming you're a mustachian and not a boglehead. A boglehead needs the income from a $3MM portfolio, a mustachian needs about $26k-30k/year or a 650k-750k stache.
3. Since some months, you're able to withdraw more than you need, you're reinvesting anyway, so that should get added in. Of course if you retire the day before the crash, you're going to have to decide what to do with options #1-2.

reeshau

  • Pencil Stache
  • ****
  • Posts: 887
  • Location: Houston, TX
Re: SWR and Current Volatility
« Reply #21 on: March 11, 2020, 05:09:17 PM »
But no one would follow a strict 4% rule in such circumstances.

This is the sort of tautological argument that always gets thrown out there.

"4% is always safe!"
"What about this real-life scenario?"
"Well, no one would stick with 4% then"
Would you continue to pull out your inflation adjusted 4% after your portfolio was cut in half less than a year into retirement? I wouldn’t. I’d be updating the resume.

You realize the 4% rule is not 100% successful historically, yes?

Actually, it is.  By definition, Bengen was looking for the rate that was successful through all historical 30-year periods he could test, which led him through periods beginning in the late 60's.  Even now, the only way to prove or disprove it, historically, is to wait 30 years:  so we are only up to retirement starts in 1990.  Pretty smooth sailing for another decade, I would say.  And Bengen's definiton of success is remaining above 0 after 30 years.

I think what you meant was *statistically* it isn't 100% successful, meaning tested against a Monte Carlo simulation.  If so, then agreed.
No. I mean it did fail for some start years. Go to cFIREsim and try it for yourself.

The Trinity study does have some historical failures. It of course depends on your asset allocation.

Yes, the Trinity study is probabilistic because it tested a wider range of factors, including asset mix and length of retirement.  Bengen's original study determined "SAFEMAX,"  which was the rate with zero failures.  It was a simplified study, because it was done with a '90s spreadsheet.  But no, no failures:  probability wasn't how he phrased his question.

http://www.retailinvestor.org/pdf/Bengen1.pdf
« Last Edit: March 11, 2020, 05:14:40 PM by reeshau »

MMMdude

  • Bristles
  • ***
  • Posts: 322
Re: SWR and Current Volatility
« Reply #22 on: March 11, 2020, 06:55:28 PM »
No change in my plans as thought it was bunk for an early retiree.  I always intended to live on interest and dividends which before the fall was 2.7% yield on my portfolio.  With the drop it's abit over 3% but no doubt some dividends will be cut and I can expect my bond funds to pay less with the Fed rate cut.  So for me it was always around that 2.75% mark

MMMdude

  • Bristles
  • ***
  • Posts: 322
Re: SWR and Current Volatility
« Reply #23 on: March 11, 2020, 06:58:21 PM »
And may I remind everyone that success is defined as having more than 0 cents to your name at the end of the SWR period.  No thanks - that alone makes this SWR concept worthless

projekt

  • Bristles
  • ***
  • Posts: 337
Re: SWR and Current Volatility
« Reply #24 on: March 12, 2020, 10:04:57 AM »
And may I remind everyone that success is defined as having more than 0 cents to your name at the end of the SWR period.  No thanks - that alone makes this SWR concept worthless
If that endpoint is death, well, you can't take it with you!

Montecarlo

  • Bristles
  • ***
  • Posts: 498
Re: SWR and Current Volatility
« Reply #25 on: March 12, 2020, 12:09:24 PM »
And may I remind everyone that success is defined as having more than 0 cents to your name at the end of the SWR period.  No thanks - that alone makes this SWR concept worthless
If that endpoint is death, well, you can't take it with you!

I think you're being a little tongue in cheek, but living the last few years of your life worried about running out of cash, that sucks.  I'm personally planning for capital preservation through my retirement.

nereo

  • Senior Mustachian
  • ********
  • Posts: 13549
  • Location: Just south of Canada
    • Here's how you can support science today:
Re: SWR and Current Volatility
« Reply #26 on: March 12, 2020, 01:03:40 PM »
And may I remind everyone that success is defined as having more than 0 cents to your name at the end of the SWR period.  No thanks - that alone makes this SWR concept worthless
No, that does not make the concept "worthless".  It's merely the starting point, and how simulations have been modeled.  After all, if the output is binary you need to define the terms.

We've been through this dozens of times, but the wealth of data and analyses is far greather than simply "it succeeded [has >$0] at the end of X years".  We can look at the spread, how many runs finished with greater than the starting amount, and SORR, among others.  What's more, others have looked at this extensively (!)

It's like people only read the headline numbers regarding a 4% WR and conclude that's all we know. 

vand

  • Handlebar Stache
  • *****
  • Posts: 1064
  • Location: UK
Re: SWR and Current Volatility
« Reply #27 on: March 12, 2020, 01:19:30 PM »
We're only 1 month into the slide.

That's, at most, 1 month of living expenses that FIREEs will have had to liquidate to cash to meet living expenses.  I suspect that most MMMers will not even have liquidated anything and will instead be living on any cash buffer they may have.

This questionis worth revisiting if we get an extended period of time where stocks remain downbeat and folks will have had to liquidate to meet living expenses.. the psychology of being a forced seller at depressed prices will be interesting, I wager.

AdrianC

  • Handlebar Stache
  • *****
  • Posts: 1131
  • Location: Cincinnati
Re: SWR and Current Volatility
« Reply #28 on: March 12, 2020, 03:33:48 PM »

Would you continue to pull out your inflation adjusted 4% after your portfolio was cut in half less than a year into retirement? I wouldn’t. I’d be updating the resume.

You realize the 4% rule is not 100% successful historically, yes?

Actually, it is.  By definition, Bengen was looking for the rate that was successful through all historical 30-year periods he could test, which led him through periods beginning in the late 60's.  Even now, the only way to prove or disprove it, historically, is to wait 30 years:  so we are only up to retirement starts in 1990.  Pretty smooth sailing for another decade, I would say.  And Bengen's definiton of success is remaining above 0 after 30 years.

I think what you meant was *statistically* it isn't 100% successful, meaning tested against a Monte Carlo simulation.  If so, then agreed.
No. I mean it did fail for some start years. Go to cFIREsim and try it for yourself.

The Trinity study does have some historical failures. It of course depends on your asset allocation.

Yes, the Trinity study is probabilistic because it tested a wider range of factors, including asset mix and length of retirement.  Bengen's original study determined "SAFEMAX,"  which was the rate with zero failures.  It was a simplified study, because it was done with a '90s spreadsheet.  But no, no failures:  probability wasn't how he phrased his question.
That's interesting. I'm used to seeing the Trinity study findings and the results I get myself from CFIREsim and Firecalc, not the original Bengen one.

So, did Bengen get something wrong? Bad data? In CFIREsim or Firecalc there's always one or more starting years that fail, even with zero investment fees.

Bloop Bloop

  • Handlebar Stache
  • *****
  • Posts: 2140
  • Location: Melbourne, Australia
Re: SWR and Current Volatility
« Reply #29 on: March 12, 2020, 05:31:54 PM »
This is why I am aiming for a 2% SWR instead of 4% SWR. Yes, it means that I have to work till 46 instead of 37, but I don't dislike working that much and I don't ever want to have to worry about downturns affecting the safety of my FIRE plans.

reeshau

  • Pencil Stache
  • ****
  • Posts: 887
  • Location: Houston, TX
Re: SWR and Current Volatility
« Reply #30 on: March 13, 2020, 07:07:20 AM »

Would you continue to pull out your inflation adjusted 4% after your portfolio was cut in half less than a year into retirement? I wouldn’t. I’d be updating the resume.

You realize the 4% rule is not 100% successful historically, yes?

Actually, it is.  By definition, Bengen was looking for the rate that was successful through all historical 30-year periods he could test, which led him through periods beginning in the late 60's.  Even now, the only way to prove or disprove it, historically, is to wait 30 years:  so we are only up to retirement starts in 1990.  Pretty smooth sailing for another decade, I would say.  And Bengen's definiton of success is remaining above 0 after 30 years.

I think what you meant was *statistically* it isn't 100% successful, meaning tested against a Monte Carlo simulation.  If so, then agreed.
No. I mean it did fail for some start years. Go to cFIREsim and try it for yourself.

The Trinity study does have some historical failures. It of course depends on your asset allocation.

Yes, the Trinity study is probabilistic because it tested a wider range of factors, including asset mix and length of retirement.  Bengen's original study determined "SAFEMAX,"  which was the rate with zero failures.  It was a simplified study, because it was done with a '90s spreadsheet.  But no, no failures:  probability wasn't how he phrased his question.
That's interesting. I'm used to seeing the Trinity study findings and the results I get myself from CFIREsim and Firecalc, not the original Bengen one.

So, did Bengen get something wrong? Bad data? In CFIREsim or Firecalc there's always one or more starting years that fail, even with zero investment fees.

Remember, Bengen was a practicing financial planner, with a 90's-era spreadsheet who dreamt up this question in the first place.  He also assembled the data himself.  (it was available, obviously, but the dataset was not freely available at the time)  So, he kept the analysis necessarily simple.

The below chart shows the Trinity study results (as interpreted by Wade Pfau's extension into 2008, at the time)  Bengen's original scenario is highlighted in red, by me.  So, 100%--again, speaking historically, not statistically (i.e. with a monte carlo analysis).

« Last Edit: March 13, 2020, 07:12:23 AM by reeshau »

AdrianC

  • Handlebar Stache
  • *****
  • Posts: 1131
  • Location: Cincinnati
Re: SWR and Current Volatility
« Reply #31 on: March 13, 2020, 08:34:30 AM »
Remember, Bengen was a practicing financial planner, with a 90's-era spreadsheet who dreamt up this question in the first place.  He also assembled the data himself.  (it was available, obviously, but the dataset was not freely available at the time)  So, he kept the analysis necessarily simple.

The below chart shows the Trinity study results (as interpreted by Wade Pfau's extension into 2008, at the time)  Bengen's original scenario is highlighted in red, by me.  So, 100%--again, speaking historically, not statistically (i.e. with a monte carlo analysis).
Thanks. So why the discrepancy between Bengen/Pfau and cFIREsim/FIREcalc? I expect it's the bonds.

Portfoliocharts gives an SWR of 4.2% (no failures) for 50% LCB/50% Int treasuries.

Tyler

  • Handlebar Stache
  • *****
  • Posts: 1172
Re: SWR and Current Volatility
« Reply #32 on: March 13, 2020, 09:19:47 AM »
So why the discrepancy between Bengen/Pfau and cFIREsim/FIREcalc? I expect it's the bonds.

Yep, that's it.  The biggest issue between the various sources is the bonds they studied.  Bengen used 5-year treasuries, while Trinity used 10-year corporates.  Pfau commonly uses T-bills and Kitces uses "intermediate treasuries" which I believe are closer to the 10-year variety.  I'm not 100% sure what cFIREsim and Firecalc use, but I suspect it's the 10-year treasury data from Robert Shiller.  While they all refer to them simply as "bonds", those are all different assets and will result in different withdrawal rates.  Not all bonds are created equal!

FWIW, you can read a bit about the history of different SWR studies on my Withdrawal Rates page.  Just scroll down below the calculator.  I also have some comparisons of SWR numbers from Bengen, Pfau, and Kitces in the FAQ for anyone who is curious about how different data sources for bonds can affect the numbers.  And of course you can also play with my charts to explore for yourself how asset allocation affects withdrawal rates. :)  For reference, my bond numbers use government treasuries/gilts/bunds/etc. and model ladders of multiple maturities that match the most common definitions and methodologies of bond index funds.
« Last Edit: March 13, 2020, 12:56:37 PM by Tyler »

Tyler

  • Handlebar Stache
  • *****
  • Posts: 1172
Re: SWR and Current Volatility
« Reply #33 on: March 13, 2020, 09:39:52 AM »
Regarding the discussion of the Bengen vs. Trinity study (safemax vs. 95% rule), I think it's important to understand that neither study was necessarily more sophisticated or "right". They just used different methodologies to solve the same problem. 

- Bengen studied every start date and calculated the withdrawal rate that resulted in exactly $0 left after 30 years.  The endpoint was fixed, and the WR was variable.  Then he looked at all of the different SWRs and found the minimum number, which he called the "safemax".

- Trinity studied the account balances for a set withdrawal rate and counted how many failed before 30 years.  The WR was fixed, and the endpoint was variable.  Then they determined the "safe withdrawal rate" by looking at their tables for the withdrawal rate that succeeded 95% of the time. 

Note that the 95% criteria is sorta arbitrary.  You can set it to 100% to follow the more conservative Bengen safemax criteria with Trinity's methodology. Or you could exclude the worst 5% of Bengen's calculated WRs to follow the more forgiving Trinity 95% criteria using Bengen's methodology.

TL;DR - Both methods are perfectly valid and will get you to the same endpoint.  It's just a difference in approach. 
« Last Edit: March 13, 2020, 11:01:20 PM by Tyler »

NWOutlier

  • Stubble
  • **
  • Posts: 115
  • Age: 52
  • Location: Seattle Area
Re: SWR and Current Volatility
« Reply #34 on: March 13, 2020, 09:54:18 AM »

[/quote]

The 4% SWR is not 100% infallible. First, it's based on historical data. Second, it's only 95% safe.

[/quote]

I'm a guy that follows the "rule" that the 4% rule is a guideline... personally I am focusing on a 3% worst case withdrawal or dividend only (approx 1.7%)... BigErn has lots of good articles about this if you're into the details... https://earlyretirementnow.com/safe-withdrawal-rate-series/

for me, I over invest (to support a significant drop up to 30-50%) and I reduce the rule to ensure I will have money when I pass... nothing is perfect.. I may fail... but this is what I'm doing