Author Topic: 4% SWR (or 3.x) has never been modeled with current interest rates  (Read 22250 times)

TomTX

  • Walrus Stache
  • *******
  • Posts: 5345
  • Location: Texas
Re: 4% SWR (or 3.x) has never been modeled with current interest rates
« Reply #50 on: August 31, 2019, 03:10:08 PM »
Probably should be planning a 2 to 3 percent withdrawal rate the next few years. A recession is looming, best be prepared.

With a reasonable asset allocation, it's hard to find any past failures at around a 3.4-3.5% WR.

Why would you want to spend many additional years working to get below that?

chevy1956

  • Bristles
  • ***
  • Posts: 415
Re: 4% SWR (or 3.x) has never been modeled with current interest rates
« Reply #51 on: September 02, 2019, 01:31:37 AM »
Probably should be planning a 2 to 3 percent withdrawal rate the next few years. A recession is looming, best be prepared.

With a reasonable asset allocation, it's hard to find any past failures at around a 3.4-3.5% WR.

Why would you want to spend many additional years working to get below that?

Personally I think you have to at least think through the potential risks  when you retire. Recessions and crashes and wars and all sorts of stuff happen. There are lots of different approaches to mitigating your risk or not mitigating your risk and I suppose one of them is to reduce your WR to a really low figure. There is no chance I am getting to a 3% WR though.

Bloop Bloop

  • Handlebar Stache
  • *****
  • Posts: 2139
  • Location: Melbourne, Australia
Re: 4% SWR (or 3.x) has never been modeled with current interest rates
« Reply #52 on: September 02, 2019, 02:11:24 AM »
Probably should be planning a 2 to 3 percent withdrawal rate the next few years. A recession is looming, best be prepared.

With a reasonable asset allocation, it's hard to find any past failures at around a 3.4-3.5% WR.

Why would you want to spend many additional years working to get below that?

Personally I think you have to at least think through the potential risks  when you retire. Recessions and crashes and wars and all sorts of stuff happen. There are lots of different approaches to mitigating your risk or not mitigating your risk and I suppose one of them is to reduce your WR to a really low figure. There is no chance I am getting to a 3% WR though.

I have gone for the very conservative route of having a 3% WR, just for peace of mind, and if I'm looking really solid economically, I'll just increase the WR as time passes. I wouldn't mind splurging on something silly like business class round-the-world flights, if my path proves overly conservative, which it probably will.

Actually my partner intends to keep working so I have modelled the whole family's withdrawal based on only our passive income. I've told my partner she can do whatever she likes with her active income! I'm sure she'll enjoy splurging it on herself and me and whoever  else.

TomTX

  • Walrus Stache
  • *******
  • Posts: 5345
  • Location: Texas
Re: 4% SWR (or 3.x) has never been modeled with current interest rates
« Reply #53 on: September 02, 2019, 06:51:39 AM »
Probably should be planning a 2 to 3 percent withdrawal rate the next few years. A recession is looming, best be prepared.

With a reasonable asset allocation, it's hard to find any past failures at around a 3.4-3.5% WR.

Why would you want to spend many additional years working to get below that?

Personally I think you have to at least think through the potential risks  when you retire. Recessions and crashes and wars and all sorts of stuff happen. There are lots of different approaches to mitigating your risk or not mitigating your risk and I suppose one of them is to reduce your WR to a really low figure. There is no chance I am getting to a 3% WR though.

Far too many posters seem to model this only looking at "risk I run out of money" not all risk. For example, the risk of dying before retiring.  The risk of being decrepit enough that you can't really enjoy travel (travel is part of the plans for many retirees) et cetera.

If working longer to reduce portfolio failure risk from 5% to 1% increases the risk of dying before retiring from 5% to 10% and reduces the years you can enjoy travel from 10 to 5, is that a good tradeoff?

Mr Mark

  • Handlebar Stache
  • *****
  • Posts: 1229
  • Location: Planet Earth
  • Achieved Financial Independence summer 2014. RE'18
Re: 4% SWR (or 3.x) has never been modeled with current interest rates
« Reply #54 on: September 03, 2019, 02:24:10 PM »
It's like every week someone is shocked to discover that the 4% rule isn't infallible.

I'm absolutely fascinated by this.
No WR is safe no matter what mathematical model you use.
You can't Boglehead your way to total financial security.

The models are a best guess at how to make decisions today to anticipate the financial needs of tomorrow and to assess the possible risk mitigating effects of different options.

It's all just probabilities and not dissimilar to the weather forecast. The calculations behind it are solid and the best we've got for deciding whether or not to bring an umbrella.
Though, that decision will depend far more on the individual's willingness to risk getting wet.

Is it wise to retire on exactly 25X your base expenses and  to spend every single cent of your budget every single year, plus inflation, while blindly ignoring what the markets and global economy are doing, while keeping no doors open for future earnings if necessary???
Lol, probably not.

What these models do is allow us to decide today what mitigating strategies we think best fit our particular risk tolerances for what might happen tomorrow.

For one person, ageism might be a huge factor in their industry, so banking on going back to their career is a bad mitigating strategy. They may choose to be more conservative, or they may choose to build skills that aren't subject to ageism, or both.

For another, they might have citizenship and family in an extremely low cost geographic region and can easily geo-arbitrage their way to a much lower spend rate.

For yet another, they might not have a lot of flexibility and may have a lot of high fixed costs, say, for a medically complex child whose specialists are only in HCOL areas of a country with expensive healthcare. Their best bet may be an extremely low WR and a large cash reserve.

For someone whose stache is primarily in index funds, AA based strategies might be best for mitigating risks like SORR. For someone with a rock solid defined benefit pension that covers their entire bare bones spend, a bond tent or cash reserves might be overkill.

For someone whose stache is primarily in real estate...well, I have no idea what they might do since I don't read much about primarily RE strategies. But you get the point.

All plans can fail.
The best thing you can do is try to understand your possible and probable failures, know your particular risk factors, and modulate accordingly.

The 4% rule is a starting point, not an end goal.
If you just got a perm and wear a lot of silk, grab an umbrella. Me? I don't mind getting rained on and I don't worry about the 4% rule. YMMV

Nice post. As Eisenhower said, "Plans are useless, but planning is essential."

And the famous "No plan survives first contact with the enemy".

Having the understanding of the allocation you're running, why you have that allocation and the implication for risk, what SWR you are using, plus an awareness of tax optimisation enable a more flexible approach that is specific to you. My plan B/C/D/... factor in a whole host of 2nd order stuff, like social security, misc income, a few trees I'll cut down, bedrooms I could lease (in a pinch), etc.

Plus, to the OP's point, a lower than average nominal rate of return may be accompanied by lower than average inflation. But delaying retirement for the sake of moving from 4.0% -> 3.5% "calculated SWR with a 60/40 portfolio allocation" I think you're fooling yourself that this reduces risk and are a total scaredy-pants. You're exchanging a known 1 or more years of your remaining life for.. what exactly?

If you have your RE expenses properly estimated (ie not already assumed barebones minimum, like, I can live on $15k/yr, so I just need $375k!) and have 4%, go for it. Focus on flexibility, and living, not slaving at a cubical clutching a spreadsheet of assumptions and quaking in your boots about SORR. SORR should be addressed in your allocation and in your ability to skip that European vacation.




chevy1956

  • Bristles
  • ***
  • Posts: 415
Re: 4% SWR (or 3.x) has never been modeled with current interest rates
« Reply #55 on: September 05, 2019, 03:41:45 AM »
Probably should be planning a 2 to 3 percent withdrawal rate the next few years. A recession is looming, best be prepared.

With a reasonable asset allocation, it's hard to find any past failures at around a 3.4-3.5% WR.

Why would you want to spend many additional years working to get below that?

Personally I think you have to at least think through the potential risks  when you retire. Recessions and crashes and wars and all sorts of stuff happen. There are lots of different approaches to mitigating your risk or not mitigating your risk and I suppose one of them is to reduce your WR to a really low figure. There is no chance I am getting to a 3% WR though.

I have gone for the very conservative route of having a 3% WR, just for peace of mind, and if I'm looking really solid economically, I'll just increase the WR as time passes. I wouldn't mind splurging on something silly like business class round-the-world flights, if my path proves overly conservative, which it probably will.

Actually my partner intends to keep working so I have modelled the whole family's withdrawal based on only our passive income. I've told my partner she can do whatever she likes with her active income! I'm sure she'll enjoy splurging it on herself and me and whoever  else.

This sounds more like giving yourself buffer to spend a lot more. It might not happen (if you get a bad run) but if it does you can splurge.

chevy1956

  • Bristles
  • ***
  • Posts: 415
Re: 4% SWR (or 3.x) has never been modeled with current interest rates
« Reply #56 on: September 05, 2019, 03:44:22 AM »
Probably should be planning a 2 to 3 percent withdrawal rate the next few years. A recession is looming, best be prepared.

With a reasonable asset allocation, it's hard to find any past failures at around a 3.4-3.5% WR.

Why would you want to spend many additional years working to get below that?

Personally I think you have to at least think through the potential risks  when you retire. Recessions and crashes and wars and all sorts of stuff happen. There are lots of different approaches to mitigating your risk or not mitigating your risk and I suppose one of them is to reduce your WR to a really low figure. There is no chance I am getting to a 3% WR though.

Far too many posters seem to model this only looking at "risk I run out of money" not all risk. For example, the risk of dying before retiring.  The risk of being decrepit enough that you can't really enjoy travel (travel is part of the plans for many retirees) et cetera.

If working longer to reduce portfolio failure risk from 5% to 1% increases the risk of dying before retiring from 5% to 10% and reduces the years you can enjoy travel from 10 to 5, is that a good tradeoff?

I'm not going to argue against your points but they do seem a little bit unrealistic. I mean my parents are traveling in their late 70's. For me personally though it's not a trade off I am comfortable with even if all I do is stay at home. Some people maybe need to do this to manage their anxiety about portfolio failure.

thesis

  • Bristles
  • ***
  • Posts: 407
Re: 4% SWR (or 3.x) has never been modeled with current interest rates
« Reply #57 on: September 05, 2019, 05:34:14 PM »
Just wanted to say that the OP may be interested in the book "Living Off Your Money" by Michael McClung. I'm only 1/4 of the way through it, but it does loads of backtesting with several sets of data, including data from the UK and Japan, and analyzes 10 or so different withdrawal strategies in retirement. It's not loaded with too much jargon, but definitely requires an understanding of finance that most MMMers probably possess (it even references FireCalc!). That book is very relevant to this topic, and very intelligently written. I wish I could summon some nuggets from it to add to the conversation, but the book is at home and I'm not.

Sadly, it doesn't look too thoroughly at 3% rate, and shows most data for 4%, 4.5%, and 5%, with some interesting rationale.

TomTX

  • Walrus Stache
  • *******
  • Posts: 5345
  • Location: Texas
Re: 4% SWR (or 3.x) has never been modeled with current interest rates
« Reply #58 on: September 07, 2019, 06:10:36 AM »
I'm not going to argue against your points but they do seem a little bit unrealistic. I mean my parents are traveling in their late 70's. For me personally though it's not a trade off I am comfortable with even if all I do is stay at home. Some people maybe need to do this to manage their anxiety about portfolio failure.
My parents still "travel" in their mid-70's, but my Dad's health took a serious dive at age 70 (Parkinsons, etc) has to use a walker, has to have heavy meds to even be that mobile, plus multiple other issues - it takes him until 11AM or noon to get going in the morning. Needs pretty much full time care, which Mom does.

They're lucky if they manage 2 hours of activities beyond surviving in a given day. If it's purely a driving day, it's 4, maybe 5 hours of drive time in a day.

wenchsenior

  • Magnum Stache
  • ******
  • Posts: 3790
Re: 4% SWR (or 3.x) has never been modeled with current interest rates
« Reply #59 on: September 08, 2019, 09:45:46 AM »
I'm not going to argue against your points but they do seem a little bit unrealistic. I mean my parents are traveling in their late 70's. For me personally though it's not a trade off I am comfortable with even if all I do is stay at home. Some people maybe need to do this to manage their anxiety about portfolio failure.
My parents still "travel" in their mid-70's, but my Dad's health took a serious dive at age 70 (Parkinsons, etc) has to use a walker, has to have heavy meds to even be that mobile, plus multiple other issues - it takes him until 11AM or noon to get going in the morning. Needs pretty much full time care, which Mom does.

They're lucky if they manage 2 hours of activities beyond surviving in a given day. If it's purely a driving day, it's 4, maybe 5 hours of drive time in a day.

Yes, I think this is good point and very case dependent.  I've been thinking about it this week b/c my husband just unexpectedly got a big promotion that we weren't expecting him to have shot at for at least 4-6 more years (go, DH!)  This puts us on track to accelerate FI a few years over previous estimates.  OR, it could just give us a new cash flow to maybe do some of the big $ stuff that we keep putting off in favor of savings (for a retirement that honestly might not be what we envision when/if we finally make it there).  Looking at our relatives and our friends and the mobility/health issues that seem rampant once people hit their 50s (not to mention chronic health conditions I have), it seems prudent to start spending more now, rather than try to shave 2 or 3 years off our FI date.

It's kind of shocking to realize that my grandmothers were both relatively immobile by their mid to late 50s (I remember one of my grandmothers having to take frequent rest breaks while walking me around a relatively small zoo when I was around 10...the other grandmother was mostly immobile by then as well... used canes and such).  It's crazy to think what state they were in, physically, at the age my husband is NOW.  And my one living grandfather became essentially chairbound shortly after retirement in his mid 60s, as did my husband's single living grandmother and grandfather. 

Some of this seems purely generational...2 generations ago smoking and drinking were rampant, dietary knowledge was questionable, and exercise programs weren't 'a thing'.  The next generation did a little better...but all four of our parents took big downward health swings/reduction in functionality in their 70s.  Neither of my parents could safely/comfortably travel on their own now at 75, though they could do so if accompanied by us kids to help them.

Going by this generational metric, we should be hypothetically good until our mid-80s, and there are plenty of ways to increase our odds of a mobile retirement.  But I suspect there is a decent chance we won't be nearly as carefree and mobile as we think.  Even if no systemic illnesses strike, joints start to wear out and resiliency to setbacks really drops.

Food for thought, at least.  Currently, I'm leaning toward spending a good chunk of our new windfall.

vand

  • Handlebar Stache
  • *****
  • Posts: 2303
  • Location: UK
Re: 4% SWR (or 3.x) has never been modeled with current interest rates
« Reply #60 on: September 10, 2019, 03:00:15 AM »
Here is the connundrum;

- All the data shows that any portfolio will support higher withdrawal rates if those drawdowns coincide with the start of a bull market in whichever is the dominant asset in the portfolio. This is not rocket science..

- The start of a bull market normally follows on from the end of a bear market when asset prices have just fallen substantially and portfolios values will also have fallen accordingly

So how do you get to this happy state of having a RE-support portfolio at the start of a bull market without having it have to weather the prior bear market?  As most FIREes are strongly against market timing, it seems to me that... you can't.

IMO the safest option is to keep stashing throughout the entire bull market; continue working through the bear market and see what your stash looks like on the other side, or at least you can negotiate a nice redunduncy package that would put you in an considerably more assured position thant the person who simply quits with a x25 stash at the top of the bull market.

I would personally rather be managing a $800k portfolio one year after a 40% decline in stocks than I would be sitting on a $1m portfolio as of today (if that makes sense).
« Last Edit: September 10, 2019, 03:17:21 AM by vand »

ender

  • Walrus Stache
  • *******
  • Posts: 7402
Re: 4% SWR (or 3.x) has never been modeled with current interest rates
« Reply #61 on: September 10, 2019, 07:28:36 AM »
Any withdrawal rate has never been modeled in the current/future situation.


Wolfpack Mustachian

  • Handlebar Stache
  • *****
  • Posts: 1866
Re: 4% SWR (or 3.x) has never been modeled with current interest rates
« Reply #62 on: September 10, 2019, 10:14:29 AM »
Any withdrawal rate has never been modeled in the current/future situation.

That's a great point. First of all, take this with a grain of salt as I'm in the accumulation phase and will probably have my OMY stuff at the end like everyone else, but as I see it now, there's two possibilities - you'll have enough money or you won't. If you don't have enough money, you'll either reduce spending or generate income. If you decide/out of necessity generate income, you'll be in a great place to be able to do that - you'll have flexibility as you're already "retired." You'll could have a career you could fall back on. If you've got a side hustle, you'll have that to go on. Through everything, you'll only have to generate an amount above what you are willing to withdraw, not the full amount you'll need to live off of. You'll be in a great situation. So what you have to ask yourself is how many guaranteed years of your life are you willing to sacrifice working out of the risk that you can't reduce spending, will have to go back to work, you won't be able to go back with anything tied to your career (because if you are, then how is that worse than working another year in it now, it would be a similar job), and that whatever job you end up getting is so, so very bad that you regret taking the chance to retire years early because you lost your very solid gamble.

Boofinator

  • Handlebar Stache
  • *****
  • Posts: 1429
Re: 4% SWR (or 3.x) has never been modeled with current interest rates
« Reply #63 on: September 10, 2019, 01:08:57 PM »
Here is the connundrum;

- All the data shows that any portfolio will support higher withdrawal rates if those drawdowns coincide with the start of a bull market in whichever is the dominant asset in the portfolio. This is not rocket science..

- The start of a bull market normally follows on from the end of a bear market when asset prices have just fallen substantially and portfolios values will also have fallen accordingly

So how do you get to this happy state of having a RE-support portfolio at the start of a bull market without having it have to weather the prior bear market?  As most FIREes are strongly against market timing, it seems to me that... you can't.

IMO the safest option is to keep stashing throughout the entire bull market; continue working through the bear market and see what your stash looks like on the other side, or at least you can negotiate a nice redunduncy package that would put you in an considerably more assured position thant the person who simply quits with a x25 stash at the top of the bull market.

I would personally rather be managing a $800k portfolio one year after a 40% decline in stocks than I would be sitting on a $1m portfolio as of today (if that makes sense).

This makes perfect sense and is one of the deceptive qualities of the 4% rule. We usually look at the 4% rule (or enhanced versions of it like cFIREsim) to get an idea of portfolio success in retirement. The problem with this is that it looks at success from a time-based metric (years successful / total years), whereas success should be  measured as a portfolio-based metric (portfolios successful / total portfolios). And needless to say, way more people will have portfolios that meet the 4% criterion after a bull market than after a bear market. Or, in other words, if your stash has hit 25x expenses, you are much more likely to be at the tail of a bull market rather than at the start, and your corresponding probability of success (based on past performance) would drop from the oft-stated 95% to something a little less sure of itself.

RWD

  • Walrus Stache
  • *******
  • Posts: 6529
  • Location: Arizona
Re: 4% SWR (or 3.x) has never been modeled with current interest rates
« Reply #64 on: September 10, 2019, 01:55:54 PM »
Here is the connundrum;

- All the data shows that any portfolio will support higher withdrawal rates if those drawdowns coincide with the start of a bull market in whichever is the dominant asset in the portfolio. This is not rocket science..

- The start of a bull market normally follows on from the end of a bear market when asset prices have just fallen substantially and portfolios values will also have fallen accordingly

So how do you get to this happy state of having a RE-support portfolio at the start of a bull market without having it have to weather the prior bear market?  As most FIREes are strongly against market timing, it seems to me that... you can't.

IMO the safest option is to keep stashing throughout the entire bull market; continue working through the bear market and see what your stash looks like on the other side, or at least you can negotiate a nice redunduncy package that would put you in an considerably more assured position thant the person who simply quits with a x25 stash at the top of the bull market.

I would personally rather be managing a $800k portfolio one year after a 40% decline in stocks than I would be sitting on a $1m portfolio as of today (if that makes sense).

This makes perfect sense and is one of the deceptive qualities of the 4% rule. We usually look at the 4% rule (or enhanced versions of it like cFIREsim) to get an idea of portfolio success in retirement. The problem with this is that it looks at success from a time-based metric (years successful / total years), whereas success should be  measured as a portfolio-based metric (portfolios successful / total portfolios). And needless to say, way more people will have portfolios that meet the 4% criterion after a bull market than after a bear market. Or, in other words, if your stash has hit 25x expenses, you are much more likely to be at the tail of a bull market rather than at the start, and your corresponding probability of success (based on past performance) would drop from the oft-stated 95% to something a little less sure of itself.

There was a very interesting thread discussing this idea here:
https://forum.mrmoneymustache.com/welcome-to-the-forum/cfiresim-severely-overestimates-success-rates-for-mustachians/

Boofinator

  • Handlebar Stache
  • *****
  • Posts: 1429

vand

  • Handlebar Stache
  • *****
  • Posts: 2303
  • Location: UK
Re: 4% SWR (or 3.x) has never been modeled with current interest rates
« Reply #66 on: September 11, 2019, 02:52:33 AM »
There was a very interesting thread discussing this idea here:
https://forum.mrmoneymustache.com/welcome-to-the-forum/cfiresim-severely-overestimates-success-rates-for-mustachians/

Thanks for filling in the maths!!

Yes, an excellent deeper level of analysis.

This podcast episode from Financialmentor is also one of my go-tos in having a more nuanced discussion about withdrawal rates:
https://tunein.com/podcasts/Business--Economics-Podcasts/The-Financial-Mentor-Podcast-p938179/?topicId=109940735
(episode #9 if the link doesn't work properly)

Here is one of the articles referenced in course of the discussion:
https://www.onefpa.org/journal/Pages/Withdrawal%20Rates%20Savings%20Rates%20and%20Valuation-Based%20Asset%20Allocation.aspx
« Last Edit: September 11, 2019, 03:14:19 AM by vand »

Bloop Bloop

  • Handlebar Stache
  • *****
  • Posts: 2139
  • Location: Melbourne, Australia
Re: 4% SWR (or 3.x) has never been modeled with current interest rates
« Reply #67 on: September 11, 2019, 03:28:36 AM »
In that thread it's interesting to see proponents of the 25x rule who say "well if the market crashes you can cut spending or get a job again".

Well obviously, you can always do that. But that is a violation of the 25x rule. You are meant to be able to rely on your stash, not have to work to supplement it (or worry about doing so).

I model my retirement income on a 3.0% net / 4.0% gross annual return. I know the true return is closer to 7% gross / 5% net [don't forget you have to pay tax  - in my country, there are no tax concessions till age 60], but I prefer having a big safety net.

matchewed

  • Magnum Stache
  • ******
  • Posts: 4422
  • Location: CT
Re: 4% SWR (or 3.x) has never been modeled with current interest rates
« Reply #68 on: September 11, 2019, 06:39:31 AM »
In that thread it's interesting to see proponents of the 25x rule who say "well if the market crashes you can cut spending or get a job again".

Well obviously, you can always do that. But that is a violation of the 25x rule. You are meant to be able to rely on your stash, not have to work to supplement it (or worry about doing so).

I model my retirement income on a 3.0% net / 4.0% gross annual return. I know the true return is closer to 7% gross / 5% net [don't forget you have to pay tax  - in my country, there are no tax concessions till age 60], but I prefer having a big safety net.

I assure you that you in fact are allowed to be a proponent of a general strategy and also be a proponent of adjusting that strategy to suit your needs.

Metalcat

  • Senior Mustachian
  • ********
  • Posts: 17387
Re: 4% SWR (or 3.x) has never been modeled with current interest rates
« Reply #69 on: September 11, 2019, 07:02:05 AM »
In that thread it's interesting to see proponents of the 25x rule who say "well if the market crashes you can cut spending or get a job again".

Well obviously, you can always do that. But that is a violation of the 25x rule. You are meant to be able to rely on your stash, not have to work to supplement it (or worry about doing so).

I model my retirement income on a 3.0% net / 4.0% gross annual return. I know the true return is closer to 7% gross / 5% net [don't forget you have to pay tax  - in my country, there are no tax concessions till age 60], but I prefer having a big safety net.

I assure you that you in fact are allowed to be a proponent of a general strategy and also be a proponent of adjusting that strategy to suit your needs.

That and there's no "violation" of anything as the 4% rule isn't a promise or guarantee of anything. It's certainly not a guarantee of financial security under all possible future conditions. No WR is.

The 4% rule is a rough guideline made rougher by inputs that are guestimates at best and spending patterns that no human would follow. Its not a solemn promise of never having to worry about money ever again.

25X imagined future spending is a good enough amount for the vast majority of people to be very secure in retirement. Additional savings provide such diminishing degrees of security that sub-4% is not a rational starting point.

Also, let's not forget that 4% is extremely conservative for those who are in fact able and willing to modulate their spending significantly and/or generate income.

These numbers don't exist in a vacuum. The additional risk mitigation of a lower WR must be weighed against the guaranteed risk of working longer. For someone who isn't happy in their job and isn't thriving in their life as a result, that miniscule risk mitigation is simply not a rational trade off compared to the guaranteed risk.

Super low WRs tend to be seen as totally rational and reasonable by those who can easily achieve them with comfort and well being while working. It can seem irrational to do anything else.

Meanwhile, someone facing burnout, health problems as a result of their career, and damage to their marriage is probably better off not giving a fuck about the tiny added financial security benefit of working several more years. Especially if they have a projected budget where they could afford to cut back in down years.

Each person must assess their own risk, and honestly, that risk has so little to do with the math and SO MUCH to do with how happy they are in the PRESENT.

Bloop Bloop

  • Handlebar Stache
  • *****
  • Posts: 2139
  • Location: Melbourne, Australia
Re: 4% SWR (or 3.x) has never been modeled with current interest rates
« Reply #70 on: September 11, 2019, 07:22:28 AM »
"Also, let's not forget that 4% is extremely conservative for those who are in fact able and willing to modulate their spending significantly and/or generate income. "

Yep, but then, getting back to the first post of this thread, it's not a "safe" (as in SWR) FIRE any more - it's more of a conditional FIRE. As you say, that sort of condition may be completely to someone's tastes, if it means the difference between working away for the man versus having a pleasant-ish part-time blogging job at a pinch.

"Each person must assess their own risk, and honestly, that risk has so little to do with the math and SO MUCH to do with how happy they are in the PRESENT."

Another very good observation. Whether you want 90%, 99%, or 99.9% certainty with your SWR probably has a lot to do with how badly you want to quit your job in the present.

bacchi

  • Walrus Stache
  • *******
  • Posts: 7056
Re: 4% SWR (or 3.x) has never been modeled with current interest rates
« Reply #71 on: September 11, 2019, 07:53:17 AM »
"Also, let's not forget that 4% is extremely conservative for those who are in fact able and willing to modulate their spending significantly and/or generate income. "

Yep, but then, getting back to the first post of this thread, it's not a "safe" (as in SWR) FIRE any more - it's more of a conditional FIRE. As you say, that sort of condition may be completely to someone's tastes, if it means the difference between working away for the man versus having a pleasant-ish part-time blogging job at a pinch.

"Each person must assess their own risk, and honestly, that risk has so little to do with the math and SO MUCH to do with how happy they are in the PRESENT."

Another very good observation. Whether you want 90%, 99%, or 99.9% certainty with your SWR probably has a lot to do with how badly you want to quit your job in the present.

The SWR, as defined by the Trinity study, is not 100% historically "safe." It's 95% "safe."

It may well be that retiring today is the start of a 5% failure sequence, which means going back to work or paring back expenses.

nereo

  • Senior Mustachian
  • ********
  • Posts: 17498
  • Location: Just south of Canada
    • Here's how you can support science today:
Re: 4% SWR (or 3.x) has never been modeled with current interest rates
« Reply #72 on: September 11, 2019, 08:20:19 AM »

Another very good observation. Whether you want 90%, 99%, or 99.9% certainty with your SWR probably has a lot to do with how badly you want to quit your job in the present.
How much you hate/enjoy your current job is certainly one factor.  How much you earn (and by extension are able to save) is another.  In an earlier thread there was some deep discussion about how 'OMY-syndrome' hits high earners the hardest, because for them One More Year might be the difference between 25x and 28x excluding market gains (a WR of 4.0% to 3.6%, or a historical 'success' rate going from 95% to 99.6% (30y).  For those at/below median incomes it might be the difference between 25x and 25.5x (a WR of 4% vs 3.9%, or a success rate from 95% to 96.1%).
A third factor - as was mentioned - is 'adaptability'; being able to modulate spending or possibly earning money down the road.  For those who argue that this is some 'violation' in what the '4% rule' stands for, I'd counter that I have yet to meet a person who does not adjust their spending from time to time in response to life events.

Boofinator

  • Handlebar Stache
  • *****
  • Posts: 1429
Re: 4% SWR (or 3.x) has never been modeled with current interest rates
« Reply #73 on: September 11, 2019, 09:05:03 AM »
A third factor - as was mentioned - is 'adaptability'; being able to modulate spending or possibly earning money down the road.  For those who argue that this is some 'violation' in what the '4% rule' stands for, I'd counter that I have yet to meet a person who does not adjust their spending from time to time in response to life events.

I would only like to add that this can work both ways: Some people will inevitably have to increase their spending in response to life events.

DavidAnnArbor

  • Handlebar Stache
  • *****
  • Posts: 2266
  • Age: 58
  • Location: Ann Arbor, Michigan
Re: 4% SWR (or 3.x) has never been modeled with current interest rates
« Reply #74 on: September 11, 2019, 09:15:40 AM »
I recall there was a discussion once before about the false precision of going from a success rate of 95% to 99% to 99.9%. Though, I don't remember all the details of that discussion.

Metalcat

  • Senior Mustachian
  • ********
  • Posts: 17387
Re: 4% SWR (or 3.x) has never been modeled with current interest rates
« Reply #75 on: September 11, 2019, 09:29:06 AM »
"Also, let's not forget that 4% is extremely conservative for those who are in fact able and willing to modulate their spending significantly and/or generate income. "

Yep, but then, getting back to the first post of this thread, it's not a "safe" (as in SWR) FIRE any more - it's more of a conditional FIRE. As you say, that sort of condition may be completely to someone's tastes, if it means the difference between working away for the man versus having a pleasant-ish part-time blogging job at a pinch.

"Each person must assess their own risk, and honestly, that risk has so little to do with the math and SO MUCH to do with how happy they are in the PRESENT."

Another very good observation. Whether you want 90%, 99%, or 99.9% certainty with your SWR probably has a lot to do with how badly you want to quit your job in the present.

There is no absolutely safe withdrawal rate though, and 4% is very safe for the vast, overwhelming majority of cases.

Safe is still and always will be a relative statement. A 4% WR is reasonable to consider safe unless an unlikely and very particular sequence of events occurs and the person involved is somehow unable to react reasonably to those circumstances.

Even Big ERN himself in a Choose FI podcasts said that 4% is more than conservative enough if you account for basic human behaviour, such as defining the annual spend amount with ample buffer/fat in it, as well as the natural instinct to cut costs if the person's nest egg takes a beating.

Say we have imaginary person A who plans to withdraw exactly 4%+inflation exactly annually and has no buffer in their budget at all. The unexpected realities of life are more likely to fuck them up than SORR ever will.

Adding buffer to a budget and lowering WR amount to basically the exact same thing conceptually in the end.

The set withdrawal rate of a plan is virtually irrelevant compared to what rationale makes up the defined "annual spend". The inputs matter far more than the outputs when it comes to these simulations, and those inputs are based on some pretty hand-wavy arithmetic.

If we account for actual normal human behaviour, such as padding in budgets and an instinct to cut spending when the 'stache drops significantly in value, then the 4% rule is about as safe as it gets, and working more to generate diminishing benefit is excessive for anyone who isn't delighted to stay at their job.

You simply CANNOT look at these models in a vacuum as if they actually model anything even remotely close to realistic spending patterns.

If someone looks at their savings amount, looks at the amount that 4% would provide and thinks "oh, that's plenty!" then they will probably be fine. If they cringe and think "Yikes! I think I can survive on that..." then it's more risky.

Nothing in life is 100% safe.


Wolfpack Mustachian

  • Handlebar Stache
  • *****
  • Posts: 1866
Re: 4% SWR (or 3.x) has never been modeled with current interest rates
« Reply #76 on: September 12, 2019, 10:11:47 AM »
"Also, let's not forget that 4% is extremely conservative for those who are in fact able and willing to modulate their spending significantly and/or generate income. "

Yep, but then, getting back to the first post of this thread, it's not a "safe" (as in SWR) FIRE any more - it's more of a conditional FIRE. As you say, that sort of condition may be completely to someone's tastes, if it means the difference between working away for the man versus having a pleasant-ish part-time blogging job at a pinch.

"Each person must assess their own risk, and honestly, that risk has so little to do with the math and SO MUCH to do with how happy they are in the PRESENT."

Another very good observation. Whether you want 90%, 99%, or 99.9% certainty with your SWR probably has a lot to do with how badly you want to quit your job in the present.

What would you propose as an alternative then, verbiage wise? Being data driven is generally best, and 4% has study backing for a relatively high success rate with historical precedents. Having a decent ball park figure out there with a study proven high track record for success is a good thing. It gives people some kind of metric and begins the conversation. Some people don't understand investing at all and who knows, might think that they have to save up almost everything they're going to spend like it's in a bank account with virtually no interest or recent historically low CD rates or whatnot. 4% is a conversation starter, and it's not a dishonest one at that. It's just the start, however, and I doubt anyone on here having a real conversation with someone who was interested would end there - it would proceed to budgeting, risk, alternatives given downturns in the economy, etc.. That doesn't mean that the rational beginning point is wrong.

Bloop Bloop

  • Handlebar Stache
  • *****
  • Posts: 2139
  • Location: Melbourne, Australia
Re: 4% SWR (or 3.x) has never been modeled with current interest rates
« Reply #77 on: September 12, 2019, 01:25:06 PM »
I think 4% is a rational starting point, for most people, and until any long-term data comes in to disprove it, it is the most balanced figure for most situations where you want reasonable security but are also willing to compromise if things don't pan out as expected. For my FIRE plans, I am less willing to compromise in terms of lifestyle (I don't want to ever contemplate having to go back to work, or cutting back on my desired spending) so I need to be more conservative.

TomTX

  • Walrus Stache
  • *******
  • Posts: 5345
  • Location: Texas
Re: 4% SWR (or 3.x) has never been modeled with current interest rates
« Reply #78 on: September 12, 2019, 07:08:52 PM »
I think 4% is a rational starting point, for most people, and until any long-term data comes in to disprove it, it is the most balanced figure for most situations where you want reasonable security but are also willing to compromise if things don't pan out as expected. For my FIRE plans, I am less willing to compromise in terms of lifestyle (I don't want to ever contemplate having to go back to work, or cutting back on my desired spending) so I need to be more conservative.

So, let say you are at a 4% SWR rate today. How much longer are you going to work to get your WR down? Lets call it X more years.

Keep in mind this approach is guaranteeing you will spend X more years working to avoid a small chance of having to go back to work* for a 1-3 years in the first 5 years of early retirement.

*Or otherwise make a moderate amount of income, around 25% of what you would normally withdraw.  Half time barista, selling stuff on Craigslist, cutting fat from the budget, whatever.

Radagast

  • Magnum Stache
  • ******
  • Posts: 2541
  • One Does Not Simply Work Into Mordor
Re: 4% SWR (or 3.x) has never been modeled with current interest rates
« Reply #79 on: September 12, 2019, 08:05:39 PM »
I recall there was a discussion once before about the false precision of going from a success rate of 95% to 99% to 99.9%. Though, I don't remember all the details of that discussion.
I don't recall the forum discussion either, but Taleb covers it in The Black Swan. His analogy is pulling colored balls out of a tub of unknown and possibly infinite size. Some of the balls are red, and some are black, but you can't see into the tub so you don't know what the color distribution is, you have to infer it by pulling them out one at a time. If you pull out 45 red and 55 black, you have a pretty good guess that the distribution, even if it is not 50/50, is pretty close to that. But what if you pull out 99 black balls and 1 red ball? Based on your single sample of a red ball's existence, you don't have the slightest freaking clue what the distribution of red balls actually is. Maybe it is 40:60 and you had an odd coincidence? 1 in 10? 1 in 100? 1 in 100,000? 1 in 10,000,000? Maybe they somehow migrate to the surface and appear in clusters? It is impossible to guess what reality is.

And what if you pull out 100 black balls in a row, and in fact have never even seen a red one? You may not even realize that red balls, as a concept, can exist. Which is where the name of the book and concept of "black swan" come from. In ancient times a black swan was a phrase used for an impossibility, similar to a pure black zebra. "Ha ha, silly person. By definition, zebras have stripes. If it is pure black it can't be a zebra." Taleb notes that AfroAmerEurAsians had perhaps billions of independent observations of white swans and knew with near statistical certainty that all swans were white (seriously, a billion observations with zero negatives can generate a high degree of statistical certainty). Then they discovered Australia, which had actual black swans, and the impossible happened.

Then consider being in a natural, social, and economic system which adapts to past events and in which every single new year may have no relationship to the past, and you realize that predicting extremely unusual events is basically just a funny movie. The actual odds of success of any withdrawal rate with a low historical probability of failure is unknown, and 95% is statistically nearly identical to 99 or even 100%.

Bloop Bloop

  • Handlebar Stache
  • *****
  • Posts: 2139
  • Location: Melbourne, Australia
Re: 4% SWR (or 3.x) has never been modeled with current interest rates
« Reply #80 on: September 12, 2019, 08:38:25 PM »
I think 4% is a rational starting point, for most people, and until any long-term data comes in to disprove it, it is the most balanced figure for most situations where you want reasonable security but are also willing to compromise if things don't pan out as expected. For my FIRE plans, I am less willing to compromise in terms of lifestyle (I don't want to ever contemplate having to go back to work, or cutting back on my desired spending) so I need to be more conservative.

So, let say you are at a 4% SWR rate today. How much longer are you going to work to get your WR down? Lets call it X more years.

Keep in mind this approach is guaranteeing you will spend X more years working to avoid a small chance of having to go back to work* for a 1-3 years in the first 5 years of early retirement.

*Or otherwise make a moderate amount of income, around 25% of what you would normally withdraw.  Half time barista, selling stuff on Craigslist, cutting fat from the budget, whatever.

For me, the difference is about 5-6 years. I'd rather have the extra reassurance. I also don't terribly mind my job in the meantime and there are certain work goals/easter eggs I'd like to unlock and that will take a few years. The reassurance means that if (as predicted) I have been way too conservative, I can splurge on things as I get older, and I like having that option, even though I may or may not use it.

I am someone who is very conservative and defensive when it comes to evaluating spending contingencies (strangely I am high-risk when it comes to evaluating income opportunities, and I don't mind speculating in that sense).

DavidAnnArbor

  • Handlebar Stache
  • *****
  • Posts: 2266
  • Age: 58
  • Location: Ann Arbor, Michigan
Re: 4% SWR (or 3.x) has never been modeled with current interest rates
« Reply #81 on: September 12, 2019, 08:48:58 PM »
Thank you Radagast, that was a really clear and thoughtful explanation of that idea.

Telecaster

  • Magnum Stache
  • ******
  • Posts: 3551
  • Location: Seattle, WA
Re: 4% SWR (or 3.x) has never been modeled with current interest rates
« Reply #82 on: September 12, 2019, 09:29:57 PM »

Then consider being in a natural, social, and economic system which adapts to past events and in which every single new year may have no relationship to the past, and you realize that predicting extremely unusual events is basically just a funny movie. The actual odds of success of any withdrawal rate with a low historical probability of failure is unknown, and 95% is statistically nearly identical to 99 or even 100%.

Great post.  We'll never go through a 1929, 1966, or 2007 again.  We'll go through something different in the future, but not those things.  At some point you have to decide "close enough" and pull the pin. 

Metalcat

  • Senior Mustachian
  • ********
  • Posts: 17387
Re: 4% SWR (or 3.x) has never been modeled with current interest rates
« Reply #83 on: September 13, 2019, 04:31:29 AM »
I think 4% is a rational starting point, for most people, and until any long-term data comes in to disprove it, it is the most balanced figure for most situations where you want reasonable security but are also willing to compromise if things don't pan out as expected. For my FIRE plans, I am less willing to compromise in terms of lifestyle (I don't want to ever contemplate having to go back to work, or cutting back on my desired spending) so I need to be more conservative.

So, let say you are at a 4% SWR rate today. How much longer are you going to work to get your WR down? Lets call it X more years.

Keep in mind this approach is guaranteeing you will spend X more years working to avoid a small chance of having to go back to work* for a 1-3 years in the first 5 years of early retirement.

*Or otherwise make a moderate amount of income, around 25% of what you would normally withdraw.  Half time barista, selling stuff on Craigslist, cutting fat from the budget, whatever.

For me, the difference is about 5-6 years. I'd rather have the extra reassurance. I also don't terribly mind my job in the meantime and there are certain work goals/easter eggs I'd like to unlock and that will take a few years. The reassurance means that if (as predicted) I have been way too conservative, I can splurge on things as I get older, and I like having that option, even though I may or may not use it.

I am someone who is very conservative and defensive when it comes to evaluating spending contingencies (strangely I am high-risk when it comes to evaluating income opportunities, and I don't mind speculating in that sense).

And that's great that you know yourself.

For someone else, an extra 5-6 years is an absolutely horrible trade off of very likely dying rich.

Now, to be clear, I'm not at all against your choice, what I'm against is people generalizing that their personal preferences are somehow fundamentally superior plans, such as advocating that aiming for a 3ish% WR is fundamentally a better plan than 4%.

There's nothing wrong with a sub 4% WR, for the record, mine is looking like it will be somewhere between 1.5-2.5%, which is insane overkill, but it's not because I'm conservative.

Bloop Bloop

  • Handlebar Stache
  • *****
  • Posts: 2139
  • Location: Melbourne, Australia
Re: 4% SWR (or 3.x) has never been modeled with current interest rates
« Reply #84 on: September 13, 2019, 06:39:32 AM »
I don't think I've ever said 3% is "superior" in a general sense. It's more fitted if you're conservative like me and you hate the idea of ever having to lift a finger again, and you enjoy knowing that recession or other negative events will likely never impact you financially. But I've never claimed that this is the appropriate choice for the 'median' FIRE route.

Metalcat

  • Senior Mustachian
  • ********
  • Posts: 17387
Re: 4% SWR (or 3.x) has never been modeled with current interest rates
« Reply #85 on: September 13, 2019, 08:36:14 AM »
I don't think I've ever said 3% is "superior" in a general sense. It's more fitted if you're conservative like me and you hate the idea of ever having to lift a finger again, and you enjoy knowing that recession or other negative events will likely never impact you financially. But I've never claimed that this is the appropriate choice for the 'median' FIRE route.

Forgive me, but your posts very much came across to me as though you saw the 4% rule as fundamentally flawed and that a lower WR to be a superior benchmark to be considered "safe".

Radagast

  • Magnum Stache
  • ******
  • Posts: 2541
  • One Does Not Simply Work Into Mordor
Re: 4% SWR (or 3.x) has never been modeled with current interest rates
« Reply #86 on: September 15, 2019, 11:22:42 PM »
I recall there was a discussion once before about the false precision of going from a success rate of 95% to 99% to 99.9%. Though, I don't remember all the details of that discussion.
I don't recall the forum discussion either, but Taleb covers it in The Black Swan. His analogy is pulling colored balls out of a tub of unknown and possibly infinite size. Some of the balls are red, and some are black, but you can't see into the tub so you don't know what the color distribution is, you have to infer it by pulling them out one at a time. If you pull out 45 red and 55 black, you have a pretty good guess that the distribution, even if it is not 50/50, is pretty close to that. But what if you pull out 99 black balls and 1 red ball? Based on your single sample of a red ball's existence, you don't have the slightest freaking clue what the distribution of red balls actually is. Maybe it is 40:60 and you had an odd coincidence? 1 in 10? 1 in 100? 1 in 100,000? 1 in 10,000,000? Maybe they somehow migrate to the surface and appear in clusters? It is impossible to guess what reality is.

And what if you pull out 100 black balls in a row, and in fact have never even seen a red one? You may not even realize that red balls, as a concept, can exist. Which is where the name of the book and concept of "black swan" come from. In ancient times a black swan was a phrase used for an impossibility, similar to a pure black zebra. "Ha ha, silly person. By definition, zebras have stripes. If it is pure black it can't be a zebra." Taleb notes that AfroAmerEurAsians had perhaps billions of independent observations of white swans and knew with near statistical certainty that all swans were white (seriously, a billion observations with zero negatives can generate a high degree of statistical certainty). Then they discovered Australia, which had actual black swans, and the impossible happened.

Then consider being in a natural, social, and economic system which adapts to past events and in which every single new year may have no relationship to the past, and you realize that predicting extremely unusual events is basically just a funny movie. The actual odds of success of any withdrawal rate with a low historical probability of failure is unknown, and 95% is statistically nearly identical to 99 or even 100%.
I just saw a great example of this! Remember a couple years ago when they saw a space rock that was on a trajectory such that it must have come from beyond the solar system? It was the only one ever seen, and the astronomic community was very fun to watch as it fluttered around guessing how often these things were likely to pass through. I recall estimates ranged from annually to once per several million years.

Well, they just found another one! Now that we have seen two of these ever, an expert said the frequency of them passing within observable distance had been narrowed down to somewhere from 1/year to 1/10,000 years. 

And astronomy is one of the most precise and mathematically predictable fields.

ChpBstrd

  • Walrus Stache
  • *******
  • Posts: 6659
  • Location: A poor and backward Southern state known as minimum wage country
Re: 4% SWR (or 3.x) has never been modeled with current interest rates
« Reply #87 on: September 16, 2019, 09:53:43 AM »
Going from a 4% WR to a 3% WR means going from a 25x portfolio to a 33x portfolio. To determine how much time it would take one to attain the lower WR we have to solve the question “how long does it take to save 33-25=8x of your expenses when you have a job plus a 25x portfolio?”

This obviously depends on 2 variables: one’s savings rate and one’s ROI. To keep things simple, let’s assume a 50% savings rate and a 4% ROI.

The 50% savings rate is equal to 1x spending.  savings/spending = savings ratio

The 4% ROI on the 25x portfolio is also equal to 1x spending. (ROI x 100 x (portfolio/spending))/spending = portfolio return ratio

Thus, in our example, the worker saves 2x spending per year because they save 1x from their pay and earn 1x from their portfolio. Savings ratio + portfolio return ratio = multiple saved per year

To save 8x at a rate of 2x saved per year will cost this worker about 4 years of labor, not accounting for the minor effect of compounding. Similarly, going from 4% WR to 3.5% would require 2 years.

Your math may vary based on expected returns and savings rates, but just plug your #s into the equations above. The results become a bit tragic as the savings rate goes down. A person with a 25% SR, for example, only accumulates 0.33X spending per year from their pay. Add in the 1x portfolio return and you are looking at 0.33x + 1x = 1.33x per year. At that pace we’re looking at 8/1.33=6 years of work to attain the lower WR. The good news is that if you dialed back earnings to the point of saving nothing (e.g. part time or passion work), your portfolio would reach 33x in 8 years at a 4% ROI all by itself. On the other hand, if you are making crazy money and can save, say 75% of your income, you would be saving 3x spending per year from work, plus 1x from the portfolio, which means you’ve saved the 8x in just 2 years.

So when we ask whether it is “worth it” to save the extra 8x, understand we might be talking 2 years for some people and 8 years for others.

Metalcat

  • Senior Mustachian
  • ********
  • Posts: 17387
Re: 4% SWR (or 3.x) has never been modeled with current interest rates
« Reply #88 on: September 16, 2019, 09:58:10 AM »
Going from a 4% WR to a 3% WR means going from a 25x portfolio to a 33x portfolio. To determine how much time it would take one to attain the lower WR we have to solve the question “how long does it take to save 33-25=8x of your expenses when you have a job plus a 25x portfolio?”

This obviously depends on 2 variables: one’s savings rate and one’s ROI. To keep things simple, let’s assume a 50% savings rate and a 4% ROI.

The 50% savings rate is equal to 1x spending.  savings/spending = savings ratio

The 4% ROI on the 25x portfolio is also equal to 1x spending. (ROI x 100 x (portfolio/spending))/spending = portfolio return ratio

Thus, in our example, the worker saves 2x spending per year because they save 1x from their pay and earn 1x from their portfolio. Savings ratio + portfolio return ratio = multiple saved per year

To save 8x at a rate of 2x saved per year will cost this worker about 4 years of labor, not accounting for the minor effect of compounding. Similarly, going from 4% WR to 3.5% would require 2 years.

Your math may vary based on expected returns and savings rates, but just plug your #s into the equations above. The results become a bit tragic as the savings rate goes down. A person with a 25% SR, for example, only accumulates 0.33X spending per year from their pay. Add in the 1x portfolio return and you are looking at 0.33x + 1x = 1.33x per year. At that pace we’re looking at 8/1.33=6 years of work to attain the lower WR. The good news is that if you dialed back earnings to the point of saving nothing (e.g. part time or passion work), your portfolio would reach 33x in 8 years at a 4% ROI all by itself. On the other hand, if you are making crazy money and can save, say 75% of your income, you would be saving 3x spending per year from work, plus 1x from the portfolio, which means you’ve saved the 8x in just 2 years.

So when we ask whether it is “worth it” to save the extra 8x, understand we might be talking 2 years for some people and 8 years for others.

And for some people who love their work and can do it on their own terms, 8 years may be easy and for a burnt-out person whose marriage on the rocks, 2 years more might be 7 years too many.

"Work" is a relative term.

NorCal

  • Handlebar Stache
  • *****
  • Posts: 1464
Re: 4% SWR (or 3.x) has never been modeled with current interest rates
« Reply #89 on: September 16, 2019, 11:39:59 AM »
This thread brings up some excellent points that I've been thinking about.

After reflection, my biggest problem with the 4% rule is that it doesn't distinguish between returns from interest/dividends and principal drawdown.

The breakout of these two actually matters for portfolio survival even if they don't matter much during the acquisition phase.

If you recall, the 1970's were one of the worst decade's for portfolio survival out there.  So we build that into our backtesting.  What the back-testing doesn't take into account is that the dividend yield on the S&P 500 was around 4% in 1970.  If you retired on the 4% rule in 1970, you would be drawing $0 in principal in year 1, and only withdrawing principal as inflation increased.

If/when another decade like the 1970's starts, you're withdrawing a lot more principal if your starting point is a yield of ~1.8%.

I'd love to see something similar to the Trinity study, but with conclusions stated as "what is the safe amount of principal I can sell every year". 

vand

  • Handlebar Stache
  • *****
  • Posts: 2303
  • Location: UK
Re: 4% SWR (or 3.x) has never been modeled with current interest rates
« Reply #90 on: September 16, 2019, 12:29:44 PM »
At the risk of going over old ground, a static 4% model is a convenient starting point for the discussion, but it doesn't very well reflect the ups and downs we see in real life.

During the accumulation phase we have an idea of what savings rate we can acheive each year, but as life unfolds the actual rate actually end up achieving will be slightly different. We may assume 7% portfolio growth, but in real life we almost certainly won't get that exact number. Why should the distribution side of FI be any different? Markets and life are dynamic, using a static model for something as important as your financial bluprint is sub-optimal at best.

Also consider that all the data shows that spending tends to fall as we age by about 10% per decade. The stats say that if you make it to 90yo then you will spend about 65% of what you did at 50, in real terms. As well, most countries have a state welfare programme which will kick in hopefully somewhere along the line. Given all this and because I consider myself very well diversified with good risk management and the ability to cut down to live on very low expenses if required, I will personally consider myself FI with quite a bit less than a x25 stash.

TomTX

  • Walrus Stache
  • *******
  • Posts: 5345
  • Location: Texas
Re: 4% SWR (or 3.x) has never been modeled with current interest rates
« Reply #91 on: September 16, 2019, 07:45:41 PM »
After reflection, my biggest problem with the 4% rule is that it doesn't distinguish between returns from interest/dividends and principal drawdown.

I think your biggest problem is that you're espousing an economic philosophy appropriate for Victorian England - not the modern era.

NorCal

  • Handlebar Stache
  • *****
  • Posts: 1464
Re: 4% SWR (or 3.x) has never been modeled with current interest rates
« Reply #92 on: September 16, 2019, 10:10:04 PM »
After reflection, my biggest problem with the 4% rule is that it doesn't distinguish between returns from interest/dividends and principal drawdown.

I think your biggest problem is that you're espousing an economic philosophy appropriate for Victorian England - not the modern era.

Nonsense.  Principal drawdowns do matter.  I dare you to do the math yourself.  Assume two portfolios of risky assets with identical return and risk profiles.  The only difference between the portfolios is whether returns are distributed through dividends or through capital appreciation.

Run that through a Monte Carlo simulation and let me know what you find.

The capital appreciation portfolio will outperform during the accumulation phase (it's more tax efficient), but you'll find the dividend portfolio does better during the drawdown phase, as you're not forced to sell principal during down markets.

bacchi

  • Walrus Stache
  • *******
  • Posts: 7056
Re: 4% SWR (or 3.x) has never been modeled with current interest rates
« Reply #93 on: September 16, 2019, 10:50:06 PM »
The capital appreciation portfolio will outperform during the accumulation phase (it's more tax efficient), but you'll find the dividend portfolio does better during the drawdown phase, as you're not forced to sell principal during down markets.

How were the dividend stocks chosen?

RWD

  • Walrus Stache
  • *******
  • Posts: 6529
  • Location: Arizona
Re: 4% SWR (or 3.x) has never been modeled with current interest rates
« Reply #94 on: September 17, 2019, 06:06:49 AM »
After reflection, my biggest problem with the 4% rule is that it doesn't distinguish between returns from interest/dividends and principal drawdown.

I think your biggest problem is that you're espousing an economic philosophy appropriate for Victorian England - not the modern era.

Nonsense.  Principal drawdowns do matter.  I dare you to do the math yourself.  Assume two portfolios of risky assets with identical return and risk profiles.  The only difference between the portfolios is whether returns are distributed through dividends or through capital appreciation.

Run that through a Monte Carlo simulation and let me know what you find.

The capital appreciation portfolio will outperform during the accumulation phase (it's more tax efficient), but you'll find the dividend portfolio does better during the drawdown phase, as you're not forced to sell principal during down markets.

https://earlyretirementnow.com/2019/02/13/yield-illusion-swr-series-part-29/

Car Jack

  • Handlebar Stache
  • *****
  • Posts: 2141
Re: 4% SWR (or 3.x) has never been modeled with current interest rates
« Reply #95 on: September 17, 2019, 07:30:32 AM »
All of the modeling for withdrawal rates has been done by taking into account higher interest rates. If this is wrong, can you point me to any research or articles?

Now we're at all-time lows in yields after a 38-year bond bull market. Global negative rates 16T+

For the 10 year treasury, let's call normal/average around 4-5%. Right now it's nearly 1.5% and could be headed below 50 basis points.

All returns price from this risk free rate.

This is not a thread on valuations or asset allocation. However, with this in mind, what does it mean for withdrawal rates? Especially if we have a Japan-like scenario where rates stay here 10+ years.

Side node: I'm planning to FIRE next year. I've got enough capital for a 3.5% SWR with an 80/20 Asset Allocation.

Any thoughts?

Have you not studied the Trinity Study?  It studied 15 and 30 year periods from 1925 to 1995, using various AAs of stocks and bonds.  The conclusion was that one would be unlikely to exhaust funds, withdrawing 3 and 4% in the first year, increasing with the CPI rate.  There have been periods with interest rates as low as today. 

From a 10,000 foot view, if your investments are returning more than inflation, your total dwindles slower, regardless of interest rates.

Anyone can add their own padding to the percentages.  I do, planning on 2% and padding on top of that, even.  But I'm a huge safety net guy and if something fails, that net catches me.

NorCal

  • Handlebar Stache
  • *****
  • Posts: 1464
Re: 4% SWR (or 3.x) has never been modeled with current interest rates
« Reply #96 on: September 17, 2019, 07:41:43 AM »
After reflection, my biggest problem with the 4% rule is that it doesn't distinguish between returns from interest/dividends and principal drawdown.

I think your biggest problem is that you're espousing an economic philosophy appropriate for Victorian England - not the modern era.

Nonsense.  Principal drawdowns do matter.  I dare you to do the math yourself.  Assume two portfolios of risky assets with identical return and risk profiles.  The only difference between the portfolios is whether returns are distributed through dividends or through capital appreciation.

Run that through a Monte Carlo simulation and let me know what you find.

The capital appreciation portfolio will outperform during the accumulation phase (it's more tax efficient), but you'll find the dividend portfolio does better during the drawdown phase, as you're not forced to sell principal during down markets.

https://earlyretirementnow.com/2019/02/13/yield-illusion-swr-series-part-29/

We are clearly talking about two entirely different topics. 

matchewed

  • Magnum Stache
  • ******
  • Posts: 4422
  • Location: CT
Re: 4% SWR (or 3.x) has never been modeled with current interest rates
« Reply #97 on: September 17, 2019, 10:08:29 AM »
This thread brings up some excellent points that I've been thinking about.

If you recall, the 1970's were one of the worst decade's for portfolio survival out there.  So we build that into our backtesting.  What the back-testing doesn't take into account is that the dividend yield on the S&P 500 was around 4% in 1970.  If you retired on the 4% rule in 1970, you would be drawing $0 in principal in year 1, and only withdrawing principal as inflation increased.

I don't think that's entirely correct. The backtesting does include dividends.

https://forum.mrmoneymustache.com/welcome-to-the-forum/cfiresim-dividends/

h82goslw

  • Stubble
  • **
  • Posts: 182
Re: 4% SWR (or 3.x) has never been modeled with current interest rates
« Reply #98 on: September 18, 2019, 09:15:16 AM »
Has anyone here not heard of VPW?   Variable percentage withdrawal.....from my perspective it alleviates many of the concerns being brought up here.

RWD

  • Walrus Stache
  • *******
  • Posts: 6529
  • Location: Arizona
Re: 4% SWR (or 3.x) has never been modeled with current interest rates
« Reply #99 on: September 18, 2019, 09:29:28 AM »
Has anyone here not heard of VPW?   Variable percentage withdrawal.....from my perspective it alleviates many of the concerns being brought up here.

Of course we've heard of it. Here is a thread discussing it from a few years back:
https://forum.mrmoneymustache.com/investor-alley/variable-percentage-withdrawal-(vpw)/

cFIREsim also supports simulations with VPW. There's a pretty glaring downside of some very low income years.

 

Wow, a phone plan for fifteen bucks!