Kitces: The Problem with FIREing at 4% and the Need for Flexible Spending Rules

Non-investing personal finance issues including insurance, credit, real estate, taxes, employment and legal issues such as trusts and wills.
User avatar
Topic Author
willthrill81
Posts: 32250
Joined: Thu Jan 26, 2017 2:17 pm
Location: USA
Contact:

Re: Kitces: The Problem with FIREing at 4% and the Need for Flexible Spending Rules

Post by willthrill81 »

countmein wrote: Thu Sep 05, 2019 4:44 pm
willthrill81 wrote: Thu Sep 05, 2019 4:39 pm
countmein wrote: Thu Sep 05, 2019 4:35 pm
willthrill81 wrote: Thu Sep 05, 2019 4:19 pm
https://www.etf.com/sections/index-inve ... nopaging=1

Well, you were right. My respect for Rick has gone down another notch.

I don't see how you were getting 2.5% real returns expected from a 60/40 portfolio. Rick is forecasting 1.9% real for 10 year Treasuries and 5% real for U.S. large-cap and higher returns for all other equity asset classes. That would work out to at least a 3.76% expected real return for a 60/40 AA, which is a lot better than 2.5%.
I was using their stated 4.5% from the podcast and subtracted 2% for inflation. Could use Rick's old 30 year forecast but you've got to set the bond side to 0 real. Stocks are 5% so that's 3% real for the portfolio. The PV model at 3% real expected return gives you an SWR of 2.0.
So you really believe that bonds will have a 0% real return for the next 50 years?
Yes, absolutely, that's the most reasonable forecast. Also I edited the above to include the 'worst years first' parameter, which really brings the SWR down even further (1.4% or less). Personally I'm willing to gamble a little bit on that one and will keep it near the 2.0 range but IMO SWRs in the 3+ range for very long horizons are just reckless.
Whether that's the most reasonable forecast is a highly debatable issue, but that's for another thread.

Out of curiosity, why would you hold bonds over such a long period of time if you believed their real return to be 0%? If I thought that bonds would have 0% real returns over the next 50 years, I wouldn't put a dime in them and would seek alternatives such as rental properties or even 100% stock portfolios.
The Sensible Steward
randomguy
Posts: 11285
Joined: Wed Sep 17, 2014 9:00 am

Re: Kitces: The Problem with FIREing at 4% and the Need for Flexible Spending Rules

Post by randomguy »

countmein wrote: Thu Sep 05, 2019 4:59 pm
Random Poster wrote: Thu Sep 05, 2019 4:49 pm

FIRECalc is showing me a 100% success rate (although I suppose that there are ways to play with the inputs to get a lower success rate). What is the calculator showing you?
Yes but they're probably using 5 year treasury returns in the neighborhood of 7%. How are you gonna get that?
No they aren't. They are including all the historical periods where interest rates were very similiar to our current rates. 1970-2000 is more of an anomaly than our current rates.
User avatar
Topic Author
willthrill81
Posts: 32250
Joined: Thu Jan 26, 2017 2:17 pm
Location: USA
Contact:

Re: Kitces: The Problem with FIREing at 4% and the Need for Flexible Spending Rules

Post by willthrill81 »

When I look at the table below from Kitces (also shown in the OP), I just can't help but think that starting one's withdrawals over a planned 50 year retirement, especially if Social Security benefits and Medicare are being discounted or ignored, at almost half of the rate (i.e. the 1.8% suggested above) that would never have failed in under 47 years with the AAs presented is far too pessimistic. Further, the 1.8% WR suggested above was 'only' at the 5th percentile and not the worse case! In the actual historic 5th percentile, the retiree would have more than 3.5 times their nominal starting portfolio balance after 50 years of withdrawals.

Image

That being said, I am not now nor have I ever suggested that any retiree blindly make withdrawals from their portfolio with complete disregard for its performance. But I find it incredulous for someone to believe that they need a portfolio almost double the size of what was necessary in essentially the worst 50 year period in history just in order to not be 'reckless'. Even for someone with a 50% savings rate, it has taken anywhere from 7-15 years to double their portfolio size in fairly recent history.
The Sensible Steward
visualguy
Posts: 2988
Joined: Thu Jan 30, 2014 12:32 am

Re: Kitces: The Problem with FIREing at 4% and the Need for Flexible Spending Rules

Post by visualguy »

jebmke wrote: Thu Sep 05, 2019 4:32 pm
dbr wrote: Thu Sep 05, 2019 11:14 am
Admiral wrote: Thu Sep 05, 2019 11:03 am Sounds suspiciously like longinvest's VPW scheme :D
There have been many approaches to flexible spending rules starting right from the beginning. Even FireCalc offers the Bernicke and Clyatt suggestions as scenarios.
Most people I know have variable spending. When things get dicey they cut back and spend a little more when things are OK. I think this goes back to the big bang.
That's like planning for a lower possible WR. It also assumes enough discretionary spending to cut back, and also assumes that you don't terribly mind cutting that discretionary spending. It's fair to wonder what the SWR (or other strategy) might be if you aren't able to cut back or you don't want to do that, and you don't want to have to go back to work potentially, etc.
User avatar
Topic Author
willthrill81
Posts: 32250
Joined: Thu Jan 26, 2017 2:17 pm
Location: USA
Contact:

Re: Kitces: The Problem with FIREing at 4% and the Need for Flexible Spending Rules

Post by willthrill81 »

visualguy wrote: Thu Sep 05, 2019 11:33 pm
jebmke wrote: Thu Sep 05, 2019 4:32 pm
dbr wrote: Thu Sep 05, 2019 11:14 am
Admiral wrote: Thu Sep 05, 2019 11:03 am Sounds suspiciously like longinvest's VPW scheme :D
There have been many approaches to flexible spending rules starting right from the beginning. Even FireCalc offers the Bernicke and Clyatt suggestions as scenarios.
Most people I know have variable spending. When things get dicey they cut back and spend a little more when things are OK. I think this goes back to the big bang.
That's like planning for a lower possible WR. It also assumes enough discretionary spending to cut back, and also assumes that you don't terribly mind cutting that discretionary spending. It's fair to wonder what the SWR (or other strategy) might be if you aren't able to cut back or you don't want to do that, and you don't want to have to go back to work potentially, etc.
When a SWR is calculated as such, it assumes no reductions in withdrawals, ever. But it also allows for portfolio depletion, and indeed in the worst case scenarios examined, the SWR purposefully depletes the portfolio entirely by the end of the period for which it is calculated (e.g. 30 years, 50 years, etc.). This means that in all other historic periods, there would have been an ending balance greater than zero.

Realistically, though, I think that jebmke is correct: virtually everyone naturally reduces spending, if even a little, when their portfolio's performance is poor. And I believe that this is generally prudent. We don't know for a fact that the future won't be worse than the worst of the past, nor do we know that we only need withdrawals for X years.
The Sensible Steward
visualguy
Posts: 2988
Joined: Thu Jan 30, 2014 12:32 am

Re: Kitces: The Problem with FIREing at 4% and the Need for Flexible Spending Rules

Post by visualguy »

willthrill81 wrote: Thu Sep 05, 2019 5:56 pm
countmein wrote: Thu Sep 05, 2019 4:44 pm
willthrill81 wrote: Thu Sep 05, 2019 4:39 pm
countmein wrote: Thu Sep 05, 2019 4:35 pm
willthrill81 wrote: Thu Sep 05, 2019 4:19 pm
https://www.etf.com/sections/index-inve ... nopaging=1

Well, you were right. My respect for Rick has gone down another notch.

I don't see how you were getting 2.5% real returns expected from a 60/40 portfolio. Rick is forecasting 1.9% real for 10 year Treasuries and 5% real for U.S. large-cap and higher returns for all other equity asset classes. That would work out to at least a 3.76% expected real return for a 60/40 AA, which is a lot better than 2.5%.
I was using their stated 4.5% from the podcast and subtracted 2% for inflation. Could use Rick's old 30 year forecast but you've got to set the bond side to 0 real. Stocks are 5% so that's 3% real for the portfolio. The PV model at 3% real expected return gives you an SWR of 2.0.
So you really believe that bonds will have a 0% real return for the next 50 years?
Yes, absolutely, that's the most reasonable forecast. Also I edited the above to include the 'worst years first' parameter, which really brings the SWR down even further (1.4% or less). Personally I'm willing to gamble a little bit on that one and will keep it near the 2.0 range but IMO SWRs in the 3+ range for very long horizons are just reckless.
Whether that's the most reasonable forecast is a highly debatable issue, but that's for another thread.

Out of curiosity, why would you hold bonds over such a long period of time if you believed their real return to be 0%? If I thought that bonds would have 0% real returns over the next 50 years, I wouldn't put a dime in them and would seek alternatives such as rental properties or even 100% stock portfolios.
I also think that it's quite possible that bonds will return 0% real over the long run. Yeah, owning rental properties is a lot more promising than bonds, but it's more work...
visualguy
Posts: 2988
Joined: Thu Jan 30, 2014 12:32 am

Re: Kitces: The Problem with FIREing at 4% and the Need for Flexible Spending Rules

Post by visualguy »

willthrill81 wrote: Thu Sep 05, 2019 11:37 pm When a SWR is calculated as such, it assumes no reductions in withdrawals, ever. But it also allows for portfolio depletion, and indeed in the worst case scenarios examined, the SWR purposefully depletes the portfolio entirely by the end of the period for which it is calculated (e.g. 30 years, 50 years, etc.). This means that in all other historic periods, there would have been an ending balance greater than zero.

Realistically, though, I think that jebmke is correct: virtually everyone naturally reduces spending, if even a little, when their portfolio's performance is poor. And I believe that this is generally prudent. We don't know for a fact that the future won't be worse than the worst of the past, nor do we know that we only need withdrawals for X years.
Yes, but it's perfectly reasonable to try to plan for no reductions in spending, not going back to work, etc. I know that I would hate to reduce spending during the first decade of retirement, for example, because those will be my fun years - likely the last good 10 years of my life in terms of health and the energy to fully enjoy various activities and experiences.
countmein
Posts: 653
Joined: Fri Dec 06, 2013 8:10 pm

Re: Kitces: The Problem with FIREing at 4% and the Need for Flexible Spending Rules

Post by countmein »

willthrill81 wrote: Thu Sep 05, 2019 5:56 pm
Out of curiosity, why would you hold bonds over such a long period of time if you believed their real return to be 0%? If I thought that bonds would have 0% real returns over the next 50 years, I wouldn't put a dime in them and would seek alternatives such as rental properties or even 100% stock portfolios.
What's wrong with zero? Beats the mattress. Clearly there's plenty of buyers at zero.

You buy bonds to decrease volatility, which, in a 50 year scenario, raises your SWR. Expected return also increases your SWR, so it's a trade off. Here's how it shakes out using PV's monte carlo (assume 5.2% real stock return with a us/intl mix):

Code: Select all

AA	        vol		expected real return	      	95% SWR
25/75       	4.9	 	1.3	                          1.8
50/50       	7.9	 	2.6	                          2.0
75/25       	11.4	 	3.9	                          2.0
User avatar
Topic Author
willthrill81
Posts: 32250
Joined: Thu Jan 26, 2017 2:17 pm
Location: USA
Contact:

Re: Kitces: The Problem with FIREing at 4% and the Need for Flexible Spending Rules

Post by willthrill81 »

countmein wrote: Thu Sep 05, 2019 11:53 pm
willthrill81 wrote: Thu Sep 05, 2019 5:56 pm
Out of curiosity, why would you hold bonds over such a long period of time if you believed their real return to be 0%? If I thought that bonds would have 0% real returns over the next 50 years, I wouldn't put a dime in them and would seek alternatives such as rental properties or even 100% stock portfolios.
What's wrong with zero? Beats the mattress. Clearly there's plenty of buyers at zero.

You buy bonds to decrease volatility, which, in a 50 year scenario, raises your SWR. Expected return also increases your SWR, so it's a trade off.
I get that. But there's still no way that I would put a large portion of my portfolio into an asset with a long-term 0% expected real return. If I had no other choice, I would definitely opt for a stock-heavy portfolio. Even if this didn't increase the SWR, it would result in a higher expected portfolio balance in most scenarios.

And I still don't see how Ferri is figuring on bonds' real return for the next 30 years being 0% (or any specific number).

Further, there's no way that I would form my withdrawal strategy on one person's forecast.
The Sensible Steward
visualguy
Posts: 2988
Joined: Thu Jan 30, 2014 12:32 am

Re: Kitces: The Problem with FIREing at 4% and the Need for Flexible Spending Rules

Post by visualguy »

willthrill81 wrote: Fri Sep 06, 2019 12:07 am I get that. But there's still no way that I would put a large portion of my portfolio into an asset with a long-term 0% expected real return. If I had no other choice, I would definitely opt for a stock-heavy portfolio. Even if this didn't increase the SWR, it would result in a higher expected portfolio balance in most scenarios.
What percentage would you be willing to put in bonds if you expected them to return 0% real in the long run?
international001
Posts: 2746
Joined: Thu Feb 15, 2018 6:31 pm

Re: Kitces: The Problem with FIREing at 4% and the Need for Flexible Spending Rules

Post by international001 »

willthrill81 wrote: Thu Sep 05, 2019 9:59 am
Kitces goes on to discuss several flexible withdrawal strategies. The first is his 'ratcheting rule', where retirees increase their spending by 10% any time the portfolio grows to 150% of its original value. I personally find this strategy to be very inefficient.

The second strategy he discusses is to withdraw every year the greater of 3.5% the portfolio's current balance or last year's withdrawal plus the inflation adjustment (i.e. 'resetting' the withdrawals). This seems much better than the ratcheting rule to me, although as Kitces notes, the decreasing life expectancy of the retiree means that 3.5% may become too low.

Kitces goes on to discuss the Guyton-Klinger 'guardrail' approach, which has been discussed extensively here in other threads.

Lastly, he discusses a strategy very similar to one that I've proposed before: segment one's spending such that essential spending is covered with something like his ratcheting rule from a portion of one's portfolio and that a more dynamic withdrawal method is used with the remainder of the portfolio.

I'm interested to hear others' thoughts on all of this.
I think this is an area on finance where all strategies have not been explored. It's more about of the algorithm, not the withdrawal rate parameter only.
To simplify, algorithm efficiency should be determined by:

1. Maximum spending overtime after retirement (and the $ you may leave to your heirs if you decide so)
2. Minimum variability of your spending

Guyton-Klinger is an approach, but by no means is justified to be the optimal

I would use a buffer approach, that smooths out withdraws and market gains. An example

- You start living of your retirement at 65
- 5 years of buffer
- Growth buffer 80/20
- Withdraw buffer with 20/80
- Growth buffer transfers 5% of its current value every year to withdraw buffer (after 60)
- You take 20% of the money in your withdraw buffer to spend every year. (after 65)

The you need to start simulating results with different parameters and optimize 1 and 2
trueblueky
Posts: 2365
Joined: Tue May 27, 2014 3:50 pm

Re: Kitces: The Problem with FIREing at 4% and the Need for Flexible Spending Rules

Post by trueblueky »

Consider this perspective. A pension fund (or annuity) can offer a higher payout because it doesn't have the same longevity risk.

https://earlyretirementnow.com/2019/08/ ... s-part-32/
dbr
Posts: 46137
Joined: Sun Mar 04, 2007 8:50 am

Re: Kitces: The Problem with FIREing at 4% and the Need for Flexible Spending Rules

Post by dbr »

trueblueky wrote: Fri Sep 06, 2019 9:22 am Consider this perspective. A pension fund (or annuity) can offer a higher payout because it doesn't have the same longevity risk.

https://earlyretirementnow.com/2019/08/ ... s-part-32/
Which is why the insurance principle of pooling risk makes sense and why any retirement analysis is incomplete without considering whether or not more annuitization might make sense. I say more because most people already have an annuity in Social Security.
User avatar
Topic Author
willthrill81
Posts: 32250
Joined: Thu Jan 26, 2017 2:17 pm
Location: USA
Contact:

Re: Kitces: The Problem with FIREing at 4% and the Need for Flexible Spending Rules

Post by willthrill81 »

visualguy wrote: Fri Sep 06, 2019 12:32 am
willthrill81 wrote: Fri Sep 06, 2019 12:07 am I get that. But there's still no way that I would put a large portion of my portfolio into an asset with a long-term 0% expected real return. If I had no other choice, I would definitely opt for a stock-heavy portfolio. Even if this didn't increase the SWR, it would result in a higher expected portfolio balance in most scenarios.
What percentage would you be willing to put in bonds if you expected them to return 0% real in the long run?
Probably no more than 25% and maybe less.
The Sensible Steward
4nursebee
Posts: 2642
Joined: Sun Apr 01, 2012 7:56 am
Location: US

Re: Kitces: The Problem with FIREing at 4% and the Need for Flexible Spending Rules

Post by 4nursebee »

I found the chart a little tough to interpret.
I find the topic headline kind of a common sense statement.
VPW solves such issues for me especially with a few years expenses readily available.
Pale Blue Dot
countmein
Posts: 653
Joined: Fri Dec 06, 2013 8:10 pm

Re: Kitces: The Problem with FIREing at 4% and the Need for Flexible Spending Rules

Post by countmein »

willthrill81 wrote: Fri Sep 06, 2019 12:07 am I get that. But there's still no way that I would put a large portion of my portfolio into an asset with a long-term 0% expected real return. If I had no other choice, I would definitely opt for a stock-heavy portfolio. Even if this didn't increase the SWR, it would result in a higher expected portfolio balance in most scenarios.
Right, SWR is still 2.0 at 100% stock but with a wider dispersion, so your good outcomes = more $ than otherwise.

The interesting thing to me is how much of an effect volatility has on SWR. It basically cancels out any extra expected return in these models and I imagine that's under-appreciated by many.

I also don't understand why 5 year treasuries shouldn't return 0 real forever. It wouldn't occur to me to expect higher returns for such an instrument.
User avatar
Topic Author
willthrill81
Posts: 32250
Joined: Thu Jan 26, 2017 2:17 pm
Location: USA
Contact:

Re: Kitces: The Problem with FIREing at 4% and the Need for Flexible Spending Rules

Post by willthrill81 »

4nursebee wrote: Fri Sep 06, 2019 9:33 am I found the chart a little tough to interpret.
In what way? Kitces describes various instances (e.g. best/worst, 5th/95th percentiles, 25th/75th percentiles, etc.) of the '4% rule' for 30 year retirements, the '3.5% rule' for 50 year retirements, and the '3.5% rule' for 50 year retirements with an 80/20 AA.
Last edited by willthrill81 on Fri Sep 06, 2019 2:29 pm, edited 1 time in total.
The Sensible Steward
User avatar
Topic Author
willthrill81
Posts: 32250
Joined: Thu Jan 26, 2017 2:17 pm
Location: USA
Contact:

Re: Kitces: The Problem with FIREing at 4% and the Need for Flexible Spending Rules

Post by willthrill81 »

countmein wrote: Fri Sep 06, 2019 9:40 am
willthrill81 wrote: Fri Sep 06, 2019 12:07 am I get that. But there's still no way that I would put a large portion of my portfolio into an asset with a long-term 0% expected real return. If I had no other choice, I would definitely opt for a stock-heavy portfolio. Even if this didn't increase the SWR, it would result in a higher expected portfolio balance in most scenarios.
Right, SWR is still 2.0 at 100% stock but with a wider dispersion, so your good outcomes = more $ than otherwise.

The interesting thing to me is how much of an effect volatility has on SWR. It basically cancels out any extra expected return in these models and I imagine that's under-appreciated by many.
Yes, volatility and unexpected inflation are the real 'enemies' of a SWR. I hope to at least partially mitigate volatility with my trend following strategy, which I won't get into further here in order to avoid derailing the thread. Unexpected inflation can be addressed very effectively with TIPS.
countmein wrote: Fri Sep 06, 2019 9:40 amI also don't understand why 5 year treasuries shouldn't return 0 real forever. It wouldn't occur to me to expect higher returns for such an instrument.
Because it would be completely unprecedented over a long-term period, at least in U.S. history. Yes, there have been periods where 5 year or 10 year Treasuries had zero or negative real returns, but never over anything remotely approaching 50 years.

Since 1970, the longest holding period where intermediate-term Treasuries had a zero or negative real return was 14 years (i.e. 1970-1983). For the U.K., it was 16 years. For Germany, it was 4 years. For Canada, it was 13 years. For Australia, it was 19 years. (Data obtained via Portfolio Charts.)
The Sensible Steward
2pedals
Posts: 1988
Joined: Wed Dec 31, 2014 11:31 am

Re: Kitces: The Problem with FIREing at 4% and the Need for Flexible Spending Rules

Post by 2pedals »

dbr wrote: Fri Sep 06, 2019 9:25 am
trueblueky wrote: Fri Sep 06, 2019 9:22 am Consider this perspective. A pension fund (or annuity) can offer a higher payout because it doesn't have the same longevity risk.

https://earlyretirementnow.com/2019/08/ ... s-part-32/
Which is why the insurance principle of pooling risk makes sense and why any retirement analysis is incomplete without considering whether or not more annuitization might make sense. I say more because most people already have an annuity in Social Security.
+1
Excellent article, thanks for sharing. I love the graphics.

If you have retirement core spending needs not covered by pension, SS or an income stream, consider purchasing an annuity to help reduce issues with DIY pension fund planning, low SWR rate and/or flexible spending rules. It's not an all or nothing decision. A little bit more life long income stream might help you sleep well at night.
User avatar
nps
Posts: 1630
Joined: Thu Dec 04, 2014 9:18 am

Re: Kitces: The Problem with FIREing at 4% and the Need for Flexible Spending Rules

Post by nps »

willthrill81 wrote: Fri Sep 06, 2019 9:27 am
visualguy wrote: Fri Sep 06, 2019 12:32 am What percentage would you be willing to put in bonds if you expected them to return 0% real in the long run?
Probably no more than 25% and maybe less.
Huh?
willthrill81 wrote: Thu Sep 05, 2019 5:56 pm If I thought that bonds would have 0% real returns over the next 50 years, I wouldn't put a dime in them and would seek alternatives such as rental properties or even 100% stock portfolios.
countmein
Posts: 653
Joined: Fri Dec 06, 2013 8:10 pm

Re: Kitces: The Problem with FIREing at 4% and the Need for Flexible Spending Rules

Post by countmein »

willthrill81 wrote: Fri Sep 06, 2019 9:44 am
Because it would be completely unprecedented over a long-term period, at least in U.S. history. Yes, there have been periods where 5 year or 10 year Treasuries had zero or negative real returns, but never over anything remotely approaching 50 years.

Since 1970, the longest holding period where intermediate-term Treasuries had a zero or negative real return was 14 years (i.e. 1970-1983). For the U.K., it was 16 years. For Germany, it was 4 years. For Canada, it was 13 years. For Australia, it was 19 years. (Data obtained via Portfolio Charts.)
Today the 30 year treasury's expected return (for 30 years, based on yield and inflation) is between 0 and 0.5, and the 5YT is obviously less. But somehow 5 year Ts are going to get you 2-3% real over 30 years just because that's the way it used to be? I don't think that's how it works.
Seasonal
Posts: 3045
Joined: Sun May 21, 2017 1:49 pm

Re: Kitces: The Problem with FIREing at 4% and the Need for Flexible Spending Rules

Post by Seasonal »

countmein wrote: Fri Sep 06, 2019 11:19 am
willthrill81 wrote: Fri Sep 06, 2019 9:44 am
Because it would be completely unprecedented over a long-term period, at least in U.S. history. Yes, there have been periods where 5 year or 10 year Treasuries had zero or negative real returns, but never over anything remotely approaching 50 years.

Since 1970, the longest holding period where intermediate-term Treasuries had a zero or negative real return was 14 years (i.e. 1970-1983). For the U.K., it was 16 years. For Germany, it was 4 years. For Canada, it was 13 years. For Australia, it was 19 years. (Data obtained via Portfolio Charts.)
Today the 30 year treasury's expected return (for 30 years, based on yield and inflation) is between 0 and 0.5, and the 5YT is obviously less. But somehow 5 year Ts are going to get you 2-3% real over 30 years just because that's the way it used to be? I don't think that's how it works.
History is a reliable guide to the future and the unprecedented never happens, so it seems a perfectly reasonable prediction. :P
randomguy
Posts: 11285
Joined: Wed Sep 17, 2014 9:00 am

Re: Kitces: The Problem with FIREing at 4% and the Need for Flexible Spending Rules

Post by randomguy »

willthrill81 wrote: Fri Sep 06, 2019 9:44 am

Because it would be completely unprecedented over a long-term period, at least in U.S. history. Yes, there have been periods where 5 year or 10 year Treasuries had zero or negative real returns, but never over anything remotely approaching 50 years.

Since 1970, the longest holding period where intermediate-term Treasuries had a zero or negative real return was 14 years (i.e. 1970-1983). For the U.K., it was 16 years. For Germany, it was 4 years. For Canada, it was 13 years. For Australia, it was 19 years. (Data obtained via Portfolio Charts.)
From 1940-1981, the 10 year had a real return of -.6%. 42 years is not quite 50 but is is a lot longer than 14:). I am not sure if there is a longer one. I just picked 1940 cause I know the rates were low and inflation sort of high for the 40s and parts of the 50s. https://dqydj.com/treasury-return-calculator/ for the data.
User avatar
Topic Author
willthrill81
Posts: 32250
Joined: Thu Jan 26, 2017 2:17 pm
Location: USA
Contact:

Re: Kitces: The Problem with FIREing at 4% and the Need for Flexible Spending Rules

Post by willthrill81 »

nps wrote: Fri Sep 06, 2019 10:32 am
willthrill81 wrote: Fri Sep 06, 2019 9:27 am
visualguy wrote: Fri Sep 06, 2019 12:32 am What percentage would you be willing to put in bonds if you expected them to return 0% real in the long run?
Probably no more than 25% and maybe less.
Huh?
willthrill81 wrote: Thu Sep 05, 2019 5:56 pm If I thought that bonds would have 0% real returns over the next 50 years, I wouldn't put a dime in them and would seek alternatives such as rental properties or even 100% stock portfolios.
You missed where I said this.
If I had no other choice, I would definitely opt for a stock-heavy portfolio.
The Sensible Steward
User avatar
Topic Author
willthrill81
Posts: 32250
Joined: Thu Jan 26, 2017 2:17 pm
Location: USA
Contact:

Re: Kitces: The Problem with FIREing at 4% and the Need for Flexible Spending Rules

Post by willthrill81 »

randomguy wrote: Fri Sep 06, 2019 12:11 pm
willthrill81 wrote: Fri Sep 06, 2019 9:44 am

Because it would be completely unprecedented over a long-term period, at least in U.S. history. Yes, there have been periods where 5 year or 10 year Treasuries had zero or negative real returns, but never over anything remotely approaching 50 years.

Since 1970, the longest holding period where intermediate-term Treasuries had a zero or negative real return was 14 years (i.e. 1970-1983). For the U.K., it was 16 years. For Germany, it was 4 years. For Canada, it was 13 years. For Australia, it was 19 years. (Data obtained via Portfolio Charts.)
From 1940-1981, the 10 year had a real return of -.6%. 42 years is not quite 50 but is is a lot longer than 14:). I am not sure if there is a longer one. I just picked 1940 cause I know the rates were low and inflation sort of high for the 40s and parts of the 50s. https://dqydj.com/treasury-return-calculator/ for the data.
The main reason for that was unexpected inflation in the 1970s. Do you think that we can forecast unexpected inflation decades into the future? I certainly don't.
The Sensible Steward
User avatar
Topic Author
willthrill81
Posts: 32250
Joined: Thu Jan 26, 2017 2:17 pm
Location: USA
Contact:

Re: Kitces: The Problem with FIREing at 4% and the Need for Flexible Spending Rules

Post by willthrill81 »

Regarding bonds paying so little, Harry Markowitz said this in an interview back in April.
Did you win your bet?

The equities had a big spike and then came down, but I remained long. I got out of almost all my bonds, even tax-deferred or nontaxable ones because they were paying little. And I’ve been waiting.

For what?

When long-term tax exempt bonds get to 4%, I’m going to rebalance. I’m just letting it ride and ignoring the bond market until interest rates come up to a level that interests me.
https://www.thinkadvisor.com/2019/04/15 ... ested-now/
The Sensible Steward
User avatar
nps
Posts: 1630
Joined: Thu Dec 04, 2014 9:18 am

Re: Kitces: The Problem with FIREing at 4% and the Need for Flexible Spending Rules

Post by nps »

willthrill81 wrote: Fri Sep 06, 2019 2:30 pm
nps wrote: Fri Sep 06, 2019 10:32 am
willthrill81 wrote: Fri Sep 06, 2019 9:27 am
visualguy wrote: Fri Sep 06, 2019 12:32 am What percentage would you be willing to put in bonds if you expected them to return 0% real in the long run?
Probably no more than 25% and maybe less.
Huh?
willthrill81 wrote: Thu Sep 05, 2019 5:56 pm If I thought that bonds would have 0% real returns over the next 50 years, I wouldn't put a dime in them and would seek alternatives such as rental properties or even 100% stock portfolios.
You missed where I said this.
If I had no other choice, I would definitely opt for a stock-heavy portfolio.
You originally said you wouldn't put a dime in bonds if you expected 0% real returns over 50 years. That's very unequivocal.

Then you said you wouldn't put a "large portion" of your portfolio in an asset expected to return 0% real over the long term. Then further stated that you might put up to 25 percent.

I don't know what you meant by having "no other choice," so I assumed you had free will to invest or not invest in bonds in both cases.

Could you clarify the conditions under which you wouldn't invest a dime and those in which you might put up to 25 percent in expected long term 0% return bonds?
User avatar
Topic Author
willthrill81
Posts: 32250
Joined: Thu Jan 26, 2017 2:17 pm
Location: USA
Contact:

Re: Kitces: The Problem with FIREing at 4% and the Need for Flexible Spending Rules

Post by willthrill81 »

nps wrote: Fri Sep 06, 2019 3:41 pm
willthrill81 wrote: Fri Sep 06, 2019 2:30 pm
nps wrote: Fri Sep 06, 2019 10:32 am
willthrill81 wrote: Fri Sep 06, 2019 9:27 am
visualguy wrote: Fri Sep 06, 2019 12:32 am What percentage would you be willing to put in bonds if you expected them to return 0% real in the long run?
Probably no more than 25% and maybe less.
Huh?
willthrill81 wrote: Thu Sep 05, 2019 5:56 pm If I thought that bonds would have 0% real returns over the next 50 years, I wouldn't put a dime in them and would seek alternatives such as rental properties or even 100% stock portfolios.
You missed where I said this.
If I had no other choice, I would definitely opt for a stock-heavy portfolio.
You originally said you wouldn't put a dime in bonds if you expected 0% real returns over 50 years. That's very unequivocal.

Then you said you wouldn't put a "large portion" of your portfolio in an asset expected to return 0% real over the long term. Then further stated that you might put up to 25 percent.

I don't know what you meant by having "no other choice," so I assumed you had free will to invest or not invest in bonds in both cases.

Could you clarify the conditions under which you wouldn't invest a dime and those in which you might put up to 25 percent in expected long term 0% return bonds?
I don't think that bonds will have 0% real returns for the next 50 years, so it's a moot point as far as I'm concerned.
The Sensible Steward
randomguy
Posts: 11285
Joined: Wed Sep 17, 2014 9:00 am

Re: Kitces: The Problem with FIREing at 4% and the Need for Flexible Spending Rules

Post by randomguy »

willthrill81 wrote: Fri Sep 06, 2019 2:31 pm
randomguy wrote: Fri Sep 06, 2019 12:11 pm
willthrill81 wrote: Fri Sep 06, 2019 9:44 am

Because it would be completely unprecedented over a long-term period, at least in U.S. history. Yes, there have been periods where 5 year or 10 year Treasuries had zero or negative real returns, but never over anything remotely approaching 50 years.

Since 1970, the longest holding period where intermediate-term Treasuries had a zero or negative real return was 14 years (i.e. 1970-1983). For the U.K., it was 16 years. For Germany, it was 4 years. For Canada, it was 13 years. For Australia, it was 19 years. (Data obtained via Portfolio Charts.)
From 1940-1981, the 10 year had a real return of -.6%. 42 years is not quite 50 but is is a lot longer than 14:). I am not sure if there is a longer one. I just picked 1940 cause I know the rates were low and inflation sort of high for the 40s and parts of the 50s. https://dqydj.com/treasury-return-calculator/ for the data.
The main reason for that was unexpected inflation in the 1970s. Do you think that we can forecast unexpected inflation decades into the future? I certainly don't.
You think the events of the 70s caused the period from 1940-1970 to have negative returns? Is there some type of time travel involved in bond investing?:) The 70s didn't help but they weren't the problem. They were set up by low rates and normal inflation. Don't like the 40s? How about 1895-1920 for another 25 year period of 0%.:)

I have no clue if we are going to have unexpected inflation or not. But I can easily imagine 10 years paying 2%+-2% and 2% +-2% inflation for the next 50 years and ending up with about 0% real. But I know looking at 40years of a bull market set up by 10 bad years (and ignoring the 30 bad years before it) is not a great way to come to decent conclusions. We have a whole generation with unrealistic expectations of what bond rates should be.
trueblueky
Posts: 2365
Joined: Tue May 27, 2014 3:50 pm

Re: Kitces: The Problem with FIREing at 4% and the Need for Flexible Spending Rules

Post by trueblueky »

randomguy wrote: Fri Sep 06, 2019 4:08 pm
willthrill81 wrote: Fri Sep 06, 2019 2:31 pm
randomguy wrote: Fri Sep 06, 2019 12:11 pm
willthrill81 wrote: Fri Sep 06, 2019 9:44 am

Because it would be completely unprecedented over a long-term period, at least in U.S. history. Yes, there have been periods where 5 year or 10 year Treasuries had zero or negative real returns, but never over anything remotely approaching 50 years.

Since 1970, the longest holding period where intermediate-term Treasuries had a zero or negative real return was 14 years (i.e. 1970-1983). For the U.K., it was 16 years. For Germany, it was 4 years. For Canada, it was 13 years. For Australia, it was 19 years. (Data obtained via Portfolio Charts.)
From 1940-1981, the 10 year had a real return of -.6%. 42 years is not quite 50 but is is a lot longer than 14:). I am not sure if there is a longer one. I just picked 1940 cause I know the rates were low and inflation sort of high for the 40s and parts of the 50s. https://dqydj.com/treasury-return-calculator/ for the data.
The main reason for that was unexpected inflation in the 1970s. Do you think that we can forecast unexpected inflation decades into the future? I certainly don't.
You think the events of the 70s caused the period from 1940-1970 to have negative returns? Is there some type of time travel involved in bond investing?:) The 70s didn't help but they weren't the problem. They were set up by low rates and normal inflation. Don't like the 40s? How about 1895-1920 for another 25 year period of 0%.:)

I have no clue if we are going to have unexpected inflation or not. But I can easily imagine 10 years paying 2%+-2% and 2% +-2% inflation for the next 50 years and ending up with about 0% real. But I know looking at 40years of a bull market set up by 10 bad years (and ignoring the 30 bad years before it) is not a great way to come to decent conclusions. We have a whole generation with unrealistic expectations of what bond rates should be.
In the 1900s, bouts of annual inflation >10% in the U.S. were associated with world events (WWI, WWII, oil shocks), which leads me to believe it's not if, but when. I have no idea when.
countmein
Posts: 653
Joined: Fri Dec 06, 2013 8:10 pm

Re: Kitces: The Problem with FIREing at 4% and the Need for Flexible Spending Rules

Post by countmein »

willthrill81 wrote: Fri Sep 06, 2019 2:31 pm
The main reason for that was unexpected inflation in the 1970s. Do you think that we can forecast unexpected inflation decades into the future? I certainly don't.
Will,

The 10yT also returned -0.375% from 1934 - 1969
fujiters
Posts: 548
Joined: Tue Mar 06, 2018 1:17 pm

Re: Kitces: The Problem with FIREing at 4% and the Need for Flexible Spending Rules

Post by fujiters »

For small time investors, like me, EE and I bonds can hold a significant portion of one's bond holdings ($40k/year in additions for a couple MFJ) and get better rates than are currently offered on marketable treasuries.

I bonds are currently offering a 0.5% real return (for purchases before November 1), and EE bonds double in value in 20 years, giving an effective 3.54% nominal yield on a 20 year treasury. As a bonus, no taxes on interest until they're redeemed.
“The purpose of the margin of safety is to render the forecast unnecessary.” -Benjamin Graham
MikeG62
Posts: 5054
Joined: Tue Nov 15, 2016 2:20 pm
Location: New Jersey

Re: Kitces: The Problem with FIREing at 4% and the Need for Flexible Spending Rules

Post by MikeG62 »

There is a lot of interesting back and forth in this thread and I am not going to jump directly into the fray. However, I will say two things regarding the belief in the need for a 2.0% withdrawal rate.

First, if future returns are such that a 2.0% WD rate is required to sustain a long retirement period, a very significant % of the US population is likely to be living in or close to poverty. This includes those who derive comfort because they have pensions (after all, just where is it that those pension plans are going to invest to get the returns needed to satisfy the plan obligations) and probably a considerable percentage of those who bought annuities thinking that those were somehow going to protect them from bleak future financial markets (those are invested in the same markets too).

Second, those who work long enough to generate enough assets to live well off a 2.0% WD rate will end up working a whole lot longer (probably a decade or more longer) than they otherwise had to. They may not think about this while working, but they will once they are finally retired (and most certainly when on their death bed).

I am not by the way suggesting one retire early with a plan to withdraw 4.5% or more of their assets for 40 years. However, something in the 3.5% range +/- is very, very, very likely to work just fine (especially if a sizable % of one's annual spend is in discretionary items).
Real Knowledge Comes Only From Experience
EddyB
Posts: 2431
Joined: Fri May 24, 2013 3:43 pm

Re: Kitces: The Problem with FIREing at 4% and the Need for Flexible Spending Rules

Post by EddyB »

Random Poster wrote: Thu Sep 05, 2019 1:48 pm
EddyB wrote: Thu Sep 05, 2019 12:56 pmIf you think this is "the answer," I'd like to know what you thought the question was.
What are some reasonable methodologies to ensure that a FIRE'd portfolio adapts to whatever markets may bring.
EddyB wrote: Thu Sep 05, 2019 12:56 pmThe initial 2% withdrawal rate seems to needlessly delay retirement vs. the scenarios described in willthrill81's précis, and the relatively conservative allocation has in the past given up considerable upside while also negatively affecting the 5th percentile outcome.
Upside potential may be "nice to have," but I suspect that everyone would agree that it is more important not to run out of money.
Terms like “reasonable” and “ensure” are open to interpretation, but isn’t Kitces claiming that a 75% higher withdrawal rate (and different allocation) has historically also been an answer to that question? Am I right that you’re not disputing that, but merely suggesting a much higher starting point and annually-determined withdrawals?
EnjoyIt
Posts: 8244
Joined: Sun Dec 29, 2013 7:06 pm

Re: Kitces: The Problem with FIREing at 4% and the Need for Flexible Spending Rules

Post by EnjoyIt »

countmein wrote: Thu Sep 05, 2019 4:44 pm
willthrill81 wrote: Thu Sep 05, 2019 4:39 pm
countmein wrote: Thu Sep 05, 2019 4:35 pm
willthrill81 wrote: Thu Sep 05, 2019 4:19 pm
https://www.etf.com/sections/index-inve ... nopaging=1

Well, you were right. My respect for Rick has gone down another notch.

I don't see how you were getting 2.5% real returns expected from a 60/40 portfolio. Rick is forecasting 1.9% real for 10 year Treasuries and 5% real for U.S. large-cap and higher returns for all other equity asset classes. That would work out to at least a 3.76% expected real return for a 60/40 AA, which is a lot better than 2.5%.
I was using their stated 4.5% from the podcast and subtracted 2% for inflation. Could use Rick's old 30 year forecast but you've got to set the bond side to 0 real. Stocks are 5% so that's 3% real for the portfolio. The PV model at 3% real expected return gives you an SWR of 2.0.
So you really believe that bonds will have a 0% real return for the next 50 years?
Yes, absolutely, that's the most reasonable forecast. Also I edited the above to include the 'worst years first' parameter, which really brings the SWR down even further (1.4% or less). Personally I'm willing to gamble a little bit on that one and will keep it near the 2.0 range but IMO SWRs in the 3+ range for very long horizons are just reckless.
Countmein, You have the right to do whatever you want with your money, your portfolio and how long you want to stay employed. But, I think recommending sub 2% withdrawal rates is reckless. Most investors will be dead or close to it before they ever get there.

EDIT: I wanted to add. If you expect real returns to be 1.8% and you are able to save 1/2 a years expenses every year, it will take you 21.5 years to go from 25x to 50x expenses. For someone who is looking to retire at 67, they may very well be long dead before they get there.
A time to EVALUATE your jitters: | viewtopic.php?p=1139732#p1139732
User avatar
Topic Author
willthrill81
Posts: 32250
Joined: Thu Jan 26, 2017 2:17 pm
Location: USA
Contact:

Re: Kitces: The Problem with FIREing at 4% and the Need for Flexible Spending Rules

Post by willthrill81 »

EnjoyIt wrote: Sat Sep 07, 2019 9:08 am
countmein wrote: Thu Sep 05, 2019 4:44 pm
willthrill81 wrote: Thu Sep 05, 2019 4:39 pm
countmein wrote: Thu Sep 05, 2019 4:35 pm
willthrill81 wrote: Thu Sep 05, 2019 4:19 pm
https://www.etf.com/sections/index-inve ... nopaging=1

Well, you were right. My respect for Rick has gone down another notch.

I don't see how you were getting 2.5% real returns expected from a 60/40 portfolio. Rick is forecasting 1.9% real for 10 year Treasuries and 5% real for U.S. large-cap and higher returns for all other equity asset classes. That would work out to at least a 3.76% expected real return for a 60/40 AA, which is a lot better than 2.5%.
I was using their stated 4.5% from the podcast and subtracted 2% for inflation. Could use Rick's old 30 year forecast but you've got to set the bond side to 0 real. Stocks are 5% so that's 3% real for the portfolio. The PV model at 3% real expected return gives you an SWR of 2.0.
So you really believe that bonds will have a 0% real return for the next 50 years?
Yes, absolutely, that's the most reasonable forecast. Also I edited the above to include the 'worst years first' parameter, which really brings the SWR down even further (1.4% or less). Personally I'm willing to gamble a little bit on that one and will keep it near the 2.0 range but IMO SWRs in the 3+ range for very long horizons are just reckless.
Countmein, You have the right to do whatever you want with your money, your portfolio and how long you want to stay employed. But, I think recommending sub 2% withdrawal rates is reckless. Most investors will be dead or close to it before they ever get there.

EDIT: I wanted to add. If you expect real returns to be 1.8% and you are able to save 1/2 a years expenses every year, it will take you 21.5 years to go from 25x to 50x expenses. For someone who is looking to retire at 67, they may very well be long dead before they get there.
So much for early retirement. :(
The Sensible Steward
stlutz
Posts: 5585
Joined: Fri Jan 02, 2009 12:08 am

Re: Kitces: The Problem with FIREing at 4% and the Need for Flexible Spending Rules

Post by stlutz »

Just reading the thread, there is some mixing up going on between 30 year and 50 year retirements. I haven't read anyone as suggesting a 2% WR for a 30 year retirement (those posters can correct me if I'm wrong).

A couple other notes:

--If one is going to make a 30 year bond forecast, I'd suggest using the 30 year TIPS rate, which is .4% currently. Doesn't make a big difference, but 0 may well be closer to the truth than 2.5 unless you can get a bout of high inflation thrown in for a couple of decades which creates other issues.

--A 45 year old couple can buy an inflation adjusted annuity that pays out 2.03% or an annuity with a 3% COLA that pays out 2.33%. I don't think anyone would recommend that, but that seems like the absolute floor withdrawal rate range that would be worth considering.

--Even 45 year old retirees will get some Social Security--if they have enough to retire at 45 they'll get quite a bit. Your withdrawal calculations need to include that, which will bump up your withdrawal rate numbers a not-insignificant amount (depending on expected consumption levels).

--As has been brought in multiple articles on the super-ER topic, it's extremely rare for people to actually fully retire at such a young age. It's perhaps more problematic that someone who is very well off by any objective standard is not willing to considering taking a couple of years off unless they can fund taking 50 years off. At the end of the day whole 50 year retirement scenario seems to be a theoretical exercise trying to solve the wrong problem.
konic
Posts: 58
Joined: Fri Jan 05, 2018 4:26 pm

Re: Kitces: The Problem with FIREing at 4% and the Need for Flexible Spending Rules

Post by konic »

willthrill81 wrote: Sat Sep 07, 2019 9:45 am
EnjoyIt wrote: Sat Sep 07, 2019 9:08 am
countmein wrote: Thu Sep 05, 2019 4:44 pm
willthrill81 wrote: Thu Sep 05, 2019 4:39 pm
countmein wrote: Thu Sep 05, 2019 4:35 pm

I was using their stated 4.5% from the podcast and subtracted 2% for inflation. Could use Rick's old 30 year forecast but you've got to set the bond side to 0 real. Stocks are 5% so that's 3% real for the portfolio. The PV model at 3% real expected return gives you an SWR of 2.0.
So you really believe that bonds will have a 0% real return for the next 50 years?
Yes, absolutely, that's the most reasonable forecast. Also I edited the above to include the 'worst years first' parameter, which really brings the SWR down even further (1.4% or less). Personally I'm willing to gamble a little bit on that one and will keep it near the 2.0 range but IMO SWRs in the 3+ range for very long horizons are just reckless.
Countmein, You have the right to do whatever you want with your money, your portfolio and how long you want to stay employed. But, I think recommending sub 2% withdrawal rates is reckless. Most investors will be dead or close to it before they ever get there.

EDIT: I wanted to add. If you expect real returns to be 1.8% and you are able to save 1/2 a years expenses every year, it will take you 21.5 years to go from 25x to 50x expenses. For someone who is looking to retire at 67, they may very well be long dead before they get there.
So much for early retirement. :(
Or Retirement at all :oops:

With such conservative planning, we might as well just prepare to never retire, just keep piling it up for the beneficiary. This might be fine with the individual suggesting it, but it is terrible advice to set the goal so far away.
mariezzz
Posts: 989
Joined: Mon Oct 02, 2017 11:02 pm

Re: Kitces: The Problem with FIREing at 4% and the Need for Flexible Spending Rules

Post by mariezzz »

willthrill81 wrote: Fri Sep 06, 2019 3:43 pm
nps wrote: Fri Sep 06, 2019 3:41 pm
willthrill81 wrote: Fri Sep 06, 2019 2:30 pm
nps wrote: Fri Sep 06, 2019 10:32 am
willthrill81 wrote: Fri Sep 06, 2019 9:27 am

Probably no more than 25% and maybe less.


Huh?

willthrill81 wrote: Thu Sep 05, 2019 5:56 pm If I thought that bonds would have 0% real returns over the next 50 years, I wouldn't put a dime in them and would seek alternatives such as rental properties or even 100% stock portfolios.


You missed where I said this.

If I had no other choice, I would definitely opt for a stock-heavy portfolio.


You originally said you wouldn't put a dime in bonds if you expected 0% real returns over 50 years. That's very unequivocal.

Then you said you wouldn't put a "large portion" of your portfolio in an asset expected to return 0% real over the long term. Then further stated that you might put up to 25 percent.

I don't know what you meant by having "no other choice," so I assumed you had free will to invest or not invest in bonds in both cases.

Could you clarify the conditions under which you wouldn't invest a dime and those in which you might put up to 25 percent in expected long term 0% return bonds?


I don't think that bonds will have 0% real returns for the next 50 years, so it's a moot point as far as I'm concerned.


As I read the 2 statements, what differed was context. "Long run" does not imply the entire 50 years - it just implies a long period.
There's a difference between
wouldn't put a dime in bonds if you expected 0% real returns over 50 years

and
visualguy wrote: ↑05 Sep 2019 22:32
What percentage would you be willing to put in bonds if you expected them to return 0% real in the long run?
Willthrill replied Probably no more than 25% and maybe less."
As has already been pointed out, it seems quite extremist to me to talk about expecting 0% real returns over 50 years - or anything close to that. I'll add ... if that scenario ever occurs, you can expect that the US will be involved in a major war during that period that takes place (in part) on US territory, so all your financial planning will just go out the window. Extreme? Yes. But no more extreme than what was postulated. In fact, I'd suggest less extreme and less ridiculous.
randomguy
Posts: 11285
Joined: Wed Sep 17, 2014 9:00 am

Re: Kitces: The Problem with FIREing at 4% and the Need for Flexible Spending Rules

Post by randomguy »

mariezzz wrote: Sat Sep 07, 2019 10:45 am
As has already been pointed out, it seems quite extremist to me to talk about expecting 0% real returns over 50 years - or anything close to that. I'll add ... if that scenario ever occurs, you can expect that the US will be involved in a major war during that period that takes place (in part) on US territory, so all your financial planning will just go out the window. Extreme? Yes. But no more extreme than what was postulated. In fact, I'd suggest less extreme and less ridiculous.

0% for 50 years might be extreme. But 0% for 20-30 years is something that has happened several times in US History including times where there weren't major wars on US territories. Given SOR risks that might be enough to give you a really bad SWR. Historically bond returns just haven't been that good. The post 1981 period is much more of an exception than the rule where bonds had decent returns.

Obviously we are talking about extremes. We are talking about stock returns that are much lower than the worst that we have had and bond returns that are slightly worse (47 years versus 50) than what we have ever had. Then you combine them together and get epically bad results.

As far as society falling apart, Japan has done OK for the last 30 years after a stock market collapse and pretty low bond yields. Now if something that happens to a somewhat small country would apply if the whole world collapsed is a different issue.
countmein
Posts: 653
Joined: Fri Dec 06, 2013 8:10 pm

Re: Kitces: The Problem with FIREing at 4% and the Need for Flexible Spending Rules

Post by countmein »

There's a lot of pushback along the lines of "but that means I'd have to work another 25 years" or "pensions will go broke" or "website X said 4% and I like the sound of that", or, my favorite, "0% bonds means WAR" (yikes).

What can I tell you? The model is what it is. You can model 50 years using historical data or synthetic. Historical data gives you a somewhat rosier output. Since this is an exercise in finding the "safest" withdrawal rate based on worst case scenarios, I think it's important to not disregard the less-rosy version of the exercise. This is the paradox of the SWR endeavor: the point is to find the floor but there's an overwhelming emotional bias pulling us towards finding the ceiling. I feel the bias too-- I don't want to have to live with a tiny SWR any more than anyone else. I am suspicious that attempts to discredit the forward models are driven by rosy-outcome bias.

But just because the model spits out a 2% SWR for a 50 year retirement doesn't mean that's the end of the story. From there, you can get quite the boost from considering VPW. Tweak the AA a little to find the sweet spot (but make sure you can handle the AA). Or, as stlutz points out, consider your Social Security expectation (but don't forget taxes as well. Personally, I let those two cancel each other out). And of course, in reality, you're not going to never earn another dime as long as you live.

Also, keep in mind that the SWR exercise, as opposed to the PWR exercise, seeks to deplete the entire portfolio on the day you die. Is that really how you want it to go down? More likely, there's a preference to have a little more of a cushion there.

The real debatable points are the inputs to the model, i.e. the return expectations. I think stlutz has it right by inferring 0.4 for bonds for 30 years. Ok but what about 50 years-- you have to extrapolate. This is your rosy-outcome bias's opportunity to show up and say "it's fine, Kitces said it was all good, we can call it 2%". To me this is problematic and counter to the objective.

As for stocks, the 2% SWR comes from a pretty generous (IMO) 5+% real return estimate. Read Triumph of the Optimists and tell me that's pessimistic. I don't think so.
EddyB
Posts: 2431
Joined: Fri May 24, 2013 3:43 pm

Re: Kitces: The Problem with FIREing at 4% and the Need for Flexible Spending Rules

Post by EddyB »

countmein wrote: Sat Sep 07, 2019 12:51 pm There's a lot of pushback along the lines of "but that means I'd have to work another 25 years" or "pensions will go broke" or "website X said 4% and I like the sound of that", or, my favorite, "0% bonds means WAR" (yikes).

What can I tell you? The model is what it is.
Not knowing what “the model” is here, I wonder what SWRs it “forecasts” when applied at various points in the past 90 years, and how they compare to what turned out to be the SWRs from those starting points (whether for 30, 40, or 50 years).
EnjoyIt
Posts: 8244
Joined: Sun Dec 29, 2013 7:06 pm

Re: Kitces: The Problem with FIREing at 4% and the Need for Flexible Spending Rules

Post by EnjoyIt »

countmein wrote: Sat Sep 07, 2019 12:51 pm There's a lot of pushback along the lines of "but that means I'd have to work another 25 years" or "pensions will go broke" or "website X said 4% and I like the sound of that", or, my favorite, "0% bonds means WAR" (yikes).

What can I tell you? The model is what it is. You can model 50 years using historical data or synthetic. Historical data gives you a somewhat rosier output. Since this is an exercise in finding the "safest" withdrawal rate based on worst case scenarios, I think it's important to not disregard the less-rosy version of the exercise. This is the paradox of the SWR endeavor: the point is to find the floor but there's an overwhelming emotional bias pulling us towards finding the ceiling. I feel the bias too-- I don't want to have to live with a tiny SWR any more than anyone else. I am suspicious that attempts to discredit the forward models are driven by rosy-outcome bias.

But just because the model spits out a 2% SWR for a 50 year retirement doesn't mean that's the end of the story. From there, you can get quite the boost from considering VPW. Tweak the AA a little to find the sweet spot (but make sure you can handle the AA). Or, as stlutz points out, consider your Social Security expectation (but don't forget taxes as well. Personally, I let those two cancel each other out). And of course, in reality, you're not going to never earn another dime as long as you live.

Also, keep in mind that the SWR exercise, as opposed to the PWR exercise, seeks to deplete the entire portfolio on the day you die. Is that really how you want it to go down? More likely, there's a preference to have a little more of a cushion there.

The real debatable points are the inputs to the model, i.e. the return expectations. I think stlutz has it right by inferring 0.4 for bonds for 30 years. Ok but what about 50 years-- you have to extrapolate. This is your rosy-outcome bias's opportunity to show up and say "it's fine, Kitces said it was all good, we can call it 2%". To me this is problematic and counter to the objective.

As for stocks, the 2% SWR comes from a pretty generous (IMO) 5+% real return estimate. Read Triumph of the Optimists and tell me that's pessimistic. I don't think so.
It's all about risk tolerance. You have very little risk tolerance in conjunction with the view of an extreme pessimist who reads into the worst of the studies and makes them appear even worse than the writer expresses. It is your prerogative to do what is right for you and your family. I will admit that there is a small chance that you are correct. I will also state that your beneficiaries are likely going to be wealthy at some point. If that is one of your goals, I think you are right on track.

Me, I do not read into the studies as you have. I am comfortable being slightly wrong and needing to spend a little less. I am ok with depleting the majority of my portfolio upon death but as you stated would like a small cushion. I will be using a variable withdrawal strategy using 4% as my baseline. I have a very high likelihood of success especially with the boost that SS will one day give me. I also have a very high likelihood of not dying at work or shortly after retiring. I like my odds.
A time to EVALUATE your jitters: | viewtopic.php?p=1139732#p1139732
countmein
Posts: 653
Joined: Fri Dec 06, 2013 8:10 pm

Re: Kitces: The Problem with FIREing at 4% and the Need for Flexible Spending Rules

Post by countmein »

EnjoyIt wrote: Sat Sep 07, 2019 1:36 pm
It's all about risk tolerance. You have very little risk tolerance in conjunction with the view of an extreme pessimist who reads into the worst of the studies and makes them appear even worse than the writer expresses. It is your prerogative to do what is right for you and your family. I will admit that there is a small chance that you are correct. I will also state that your beneficiaries are likely going to be wealthy at some point. If that is one of your goals, I think you are right on track.

Me, I do not read into the studies as you have. I am comfortable being slightly wrong and needing to spend a little less. I am ok with depleting the majority of my portfolio upon death but as you stated would like a small cushion. I will be using a variable withdrawal strategy using 4% as my baseline. I have a very high likelihood of success especially with the boost that SS will one day give me. I also have a very high likelihood of not dying at work or shortly after retiring. I like my odds.
I do not have "little" risk tolerance, and I haven't "read into" any studies. I just ran the numbers on PV's monte carlo for a 50 year retirement and reported back.

"Pessimist" is also a mischaracterization. In fact I allowed my own rosy-outcome bias to use 95% survival instead of 100%, and I totally disregarded the sequence of returns portion of the tool, both of which would make the output worse. Try the damn tool for yourself, don't come after me.
MikeG62
Posts: 5054
Joined: Tue Nov 15, 2016 2:20 pm
Location: New Jersey

Re: Kitces: The Problem with FIREing at 4% and the Need for Flexible Spending Rules

Post by MikeG62 »

countmein wrote: Sat Sep 07, 2019 1:50 pm
EnjoyIt wrote: Sat Sep 07, 2019 1:36 pm
It's all about risk tolerance. You have very little risk tolerance in conjunction with the view of an extreme pessimist who reads into the worst of the studies and makes them appear even worse than the writer expresses. It is your prerogative to do what is right for you and your family. I will admit that there is a small chance that you are correct. I will also state that your beneficiaries are likely going to be wealthy at some point. If that is one of your goals, I think you are right on track.

Me, I do not read into the studies as you have. I am comfortable being slightly wrong and needing to spend a little less. I am ok with depleting the majority of my portfolio upon death but as you stated would like a small cushion. I will be using a variable withdrawal strategy using 4% as my baseline. I have a very high likelihood of success especially with the boost that SS will one day give me. I also have a very high likelihood of not dying at work or shortly after retiring. I like my odds.
I do not have "little" risk tolerance, and I haven't "read into" any studies. I just ran the numbers on PV's monte carlo for a 50 year retirement and reported back.

"Pessimist" is also a mischaracterization. In fact I allowed my own rosy-outcome bias to use 95% survival instead of 100%, and I totally disregarded the sequence of returns portion of the tool, both of which would make the output worse. Try the damn tool for yourself, don't come after me.
You may want to read this before putting too much stock in worst case outcomes your Monte Carlo analysis...

https://www.kitces.com/blog/monte-carlo ... l-returns/

Here is the conclusion, but please do read the entire article.

”The bottom line, though, is simply to recognize that despite the common criticism that Monte Carlo analysis and normal distributions understate “fat tails”, when it comes to long-term retirement projections, Monte Carlo analysis actually overstates the risk of extreme drawdowns relative to the actual historical record – yielding a material number of projections that are worse (or better) than any sequence that has actually occurred in history. On the one hand, this suggests that Monte Carlo analysis is actually a more conservative way of projecting the safety of a retirement plan than “just” relying on rolling historical returns. Yet on the other hand, it may actually lead prospective retirees to wait too long to retire (and/or spend less than they really can), by overstating the actual risk of long-term volatility and sequence of return risk!”
Real Knowledge Comes Only From Experience
randomguy
Posts: 11285
Joined: Wed Sep 17, 2014 9:00 am

Re: Kitces: The Problem with FIREing at 4% and the Need for Flexible Spending Rules

Post by randomguy »

EddyB wrote: Sat Sep 07, 2019 1:24 pm
countmein wrote: Sat Sep 07, 2019 12:51 pm There's a lot of pushback along the lines of "but that means I'd have to work another 25 years" or "pensions will go broke" or "website X said 4% and I like the sound of that", or, my favorite, "0% bonds means WAR" (yikes).

What can I tell you? The model is what it is.
Not knowing what “the model” is here, I wonder what SWRs it “forecasts” when applied at various points in the past 90 years, and how they compare to what turned out to be the SWRs from those starting points (whether for 30, 40, or 50 years).
The problem is you have no good way to validate models. It is incredibly easy for computers to spit out numbers. To have any basis in reality is a lot harder. How are things like valuations (mean reversion), inflation and its links to future returns, the effects of bond rates on future returns, and so on all being accounted for. A lot of models just tread stocks, bonds, and inflation as 3 separate sources of data and you get all sorts of results that border on imaginary. It is easy to have computers spit out numbers. If those numbers have any basis in reality is another.

But lets think about what 5% real means. It means that the AVERAGE over the next 50 years will be worse than the worst 50 year period.

Nobody will ever convince people about the other point. There have been chicken littles for 30+ years talking about the next great depression. So far the worse we have had is 2000-9 which was bad but far from end of the world. Maybe next time will be different. Maybe the US is about to burst its stock bubble and we are getting japan 1989-2019 type performance.
EnjoyIt
Posts: 8244
Joined: Sun Dec 29, 2013 7:06 pm

Re: Kitces: The Problem with FIREing at 4% and the Need for Flexible Spending Rules

Post by EnjoyIt »

countmein wrote: Sat Sep 07, 2019 1:50 pm
EnjoyIt wrote: Sat Sep 07, 2019 1:36 pm
It's all about risk tolerance. You have very little risk tolerance in conjunction with the view of an extreme pessimist who reads into the worst of the studies and makes them appear even worse than the writer expresses. It is your prerogative to do what is right for you and your family. I will admit that there is a small chance that you are correct. I will also state that your beneficiaries are likely going to be wealthy at some point. If that is one of your goals, I think you are right on track.

Me, I do not read into the studies as you have. I am comfortable being slightly wrong and needing to spend a little less. I am ok with depleting the majority of my portfolio upon death but as you stated would like a small cushion. I will be using a variable withdrawal strategy using 4% as my baseline. I have a very high likelihood of success especially with the boost that SS will one day give me. I also have a very high likelihood of not dying at work or shortly after retiring. I like my odds.
I do not have "little" risk tolerance, and I haven't "read into" any studies. I just ran the numbers on PV's monte carlo for a 50 year retirement and reported back.

"Pessimist" is also a mischaracterization. In fact I allowed my own rosy-outcome bias to use 95% survival instead of 100%, and I totally disregarded the sequence of returns portion of the tool, both of which would make the output worse. Try the damn tool for yourself, don't come after me.
The problem with tools like these has a lot to do with the user. Garbage in, garbage out.

Just to re-emphasize:
MikeG62 wrote: Sat Sep 07, 2019 2:08 pm You may want to read this before putting too much stock in worst case outcomes your Monte Carlo analysis...

https://www.kitces.com/blog/monte-carlo ... l-returns/

Here is the conclusion, but please do read the entire article.

”The bottom line, though, is simply to recognize that despite the common criticism that Monte Carlo analysis and normal distributions understate “fat tails”, when it comes to long-term retirement projections, Monte Carlo analysis actually overstates the risk of extreme drawdowns relative to the actual historical record – yielding a material number of projections that are worse (or better) than any sequence that has actually occurred in history. On the one hand, this suggests that Monte Carlo analysis is actually a more conservative way of projecting the safety of a retirement plan than “just” relying on rolling historical returns. Yet on the other hand, it may actually lead prospective retirees to wait too long to retire (and/or spend less than they really can), by overstating the actual risk of long-term volatility and sequence of return risk!”
A time to EVALUATE your jitters: | viewtopic.php?p=1139732#p1139732
stlutz
Posts: 5585
Joined: Fri Jan 02, 2009 12:08 am

Re: Kitces: The Problem with FIREing at 4% and the Need for Flexible Spending Rules

Post by stlutz »

Aswath Damodaran who is arguably the world's foremost authority on valuation and who does not hang around in the "my forecast is more pessimistic than your forecast" circles currently estimates the Equity Risk Premium as being 5.3%. So, if long term bonds yield .4% real, then a reasonable expectation of US equity returns to plug into the model would be ~5.7%.

People need to remember that SWR is not based on the average outcome; it's based on unfavorable outcomes--not necessarily the worst but on the bottom 5 or 10%. Also keep in mind that unlike historical returns, Monte Carlo considers the case of bad circumstances combined with poor decision-making that makes those circumstances worse. I'm always a fan of the Death of Equities article because when you read it you realize the problems it outlines were fixed by good policy decisions. That didn't have to happen. If it didn't, then 4% (or slightly less) likely wouldn't have worked for the 1967 retirement cohort.

Someone who did an SWR analysis in 1950 based on global returns would have concluded that there was a high probability of everything going to zero. After all, that happened in quite a few countries in the first half of the 20th century. Basing the "best" or "worst" case on a limited selection of history definitely has its limitations. Monte Carlo is certainly not a perfect tool--but it does at least address this problem.
randomguy
Posts: 11285
Joined: Wed Sep 17, 2014 9:00 am

Re: Kitces: The Problem with FIREing at 4% and the Need for Flexible Spending Rules

Post by randomguy »

stlutz wrote: Sat Sep 07, 2019 2:31 pm Aswath Damodaran who is arguably the world's foremost authority on valuation and who does not hang around in the "my forecast is more pessimistic than your forecast" circles currently estimates the Equity Risk Premium as being 5.3%. So, if long term bonds yield .4% real, then a reasonable expectation of US equity returns to plug into the model would be ~5.7%.

People need to remember that SWR is not based on the average outcome; it's based on unfavorable outcomes--not necessarily the worst but on the bottom 5 or 10%. Also keep in mind that unlike historical returns, Monte Carlo considers the case of bad circumstances combined with poor decision-making that makes those circumstances worse. I'm always a fan of the Death of Equities article because when you read it you realize the problems it outlines were fixed by good policy decisions. That didn't have to happen. If it didn't, then 4% (or slightly less) likely wouldn't have worked for the 1967 retirement cohort.
Of course. On the same token if Nixon didn't favor winning an election instead of the economy, maybe the 70s wouldn't have been remotely as bad. If we didn't sign up for a trade war, would the 30s been better? What about if we didn't repeal. Glass Stegal? Would we have had financial crisis? There will constantly be choices being made that effect outcomes. How much of the US versus Europe returns for the past 10 years a result of political choices? Who knows. But it seems likely that the conditions that cause 0% real from stocks for 10-15 years are likely to cause political action to booster the economies. How effective will those actions be? Who knows.

It is also fun to read the articles form the 90s talking about poor bond yields and how inflation is about to make a big comeback. Back then the 70s was still somewhat recent memory and people worried about a rebound Sort of like how everyone is looking for a real estate bubble these days.
milktoast
Posts: 625
Joined: Wed Jul 10, 2019 8:17 pm

Re: Kitces: The Problem with FIREing at 4% and the Need for Flexible Spending Rules

Post by milktoast »

2% SWR over 50yrs assumes 0% real return on portfolio.

That seems extreme to me. But if it seems reasonable to you, I highly recommend going for 0.5% real returns via a TIPS only portfolio.

If you think 1.8% is the number of 50 years, run don’t walk to your TIPs auction.
Post Reply