Challenging the 4% "Rule"

[ysette9]

True, but on the flip side, other modeling by Michael Kitces and the Mad Fientist have shown that it is really the first ten years of FIRE where you need to be worried; if you can get out safely on the other side, your portfolio should essentially be too big to fail. The first years after retiring should be when you are paying most attention to spending, market performance, and when you are best able to return to work in some capacity or reduce spending in other ways.

[bacchi]

Mostly Big ERN isn't addressing unforeseen medical bills.  He's addressing situations where the 4% rule hasn't worked historically.  So it's not reducing your annual spending by 5% (which is what your scenario discusses).  It's more like 7.5% to 12% permanent reduction depending on what scenario you are looking at, and that assumes you have perfect foresight and know ahead of time to start with a lower safe withdrawal rate.  If you start off with 4% and have an awful sequence of returns, you can have to cut much deeper to compensate.

Eh, that's still pretty modest with enough slop in the budget. A 12% haircut sucks but it's not the end of the world unless you're living on the edge. Of course, the budget slop may be the result of OMY.

Still, it's playing the odds. Your ER might be a 1966 repeat. Or not. Retiring in 2000 with a 75/25 portfolio still has money left (and it's been struggling since the beginning -- surely this retiree would have adjusted withdrawals along the way). Even a 3.255% withdrawal might crater if we have the mother of all recessions, such as oil running low in 2040 and oil prices rising to $150/barrel without a tech replacement.

Smart people can't predict the future either.

[Jrr85]

Mostly Big ERN isn't addressing unforeseen medical bills.  He's addressing situations where the 4% rule hasn't worked historically.  So it's not reducing your annual spending by 5% (which is what your scenario discusses).  It's more like 7.5% to 12% permanent reduction depending on what scenario you are looking at, and that assumes you have perfect foresight and know ahead of time to start with a lower safe withdrawal rate.  If you start off with 4% and have an awful sequence of returns, you can have to cut much deeper to compensate.

Eh, that's still pretty modest with enough slop in the budget. A 12% haircut sucks but it's not the end of the world unless you're living on the edge. Of course, the budget slop may be the result of OMY.

Still, it's playing the odds. Your ER might be a 1966 repeat. Or not. Retiring in 2000 with a 75/25 portfolio still has money left (and it's been struggling since the beginning -- surely this retiree would have adjusted withdrawals along the way). Even a 3.255% withdrawal might crater if we have the mother of all recessions, such as oil running low in 2040 and oil prices rising to $150/barrel without a tech replacement.

Smart people can't predict the future either.

Yes.  I think having enough slop in the bucket is basically the equivalent of having a higher than 25x multiple of expenses.  You can look at it as retiring with a 4% withdrawal rate but having more than 10% of that going to luxury spending that you dont' mind cutting, or you can look at it as having a 3 and change % withdrawal rate but taking extra withdrawal each year to spend on luxuries as long as things are going decent.  Potato/Potato. 

If you're looking at 4% being your initial withdrawal rate and that gets you $100k, I wouldn't worry about working until that represents a 3.5 or 3.75% withdrawal rate.  If 4% was going to generate an initial withdrawal rate of $40k, I probably would (although plenty of people on here probably think $40k is already extravagant enough). 
 

[bacchi]

Yes.  I think having enough slop in the bucket is basically the equivalent of having a higher than 25x multiple of expenses.  You can look at it as retiring with a 4% withdrawal rate but having more than 10% of that going to luxury spending that you dont' mind cutting, or you can look at it as having a 3 and change % withdrawal rate but taking extra withdrawal each year to spend on luxuries as long as things are going decent.  Potato/Potato. 

Agreed. Having the luxury of having luxury in the ER budget is great if the SOR goes poorly. It also makes me think that I worked too long. OMY is psychologically tough to avoid for most of us, though, unless you finish with a windfall.

[steveo]

I can't help but think a lot of people haven't actually read Big ERN's posts.  The entire point of his posts is there is nothing "modest" about the flexibility (whether going back to work or cutting back on spending) required to avoid portfolio failures and portfolio failure (or the flexibility required to avoid it) is a much, much, much worse result for many people than simply working until they have a lower initial withdrawal rate.

Granted, if you don't mind working part time and/or can do so at a high pay, then flexibility is fine.  But unless you can jump back in the job market at a high pay, it's not working part time for a couple of years.  It's working part time for potentially a decade or more, compared to sticking with a job (that in his case he likes) for another year or two.  Definitely not applicable to everybody, but I think he's right that many people just saying "I'll be flexible" are looking at it as "I'll work part time for a few years", not "I'll work part time for more than a decade of my 'retirement'" or alternatively they are thinking "I'll reduce my withdrawal rate by half a percentage point for a few years, no big deal", and not "I'll reduce my withdrawal rate by more than 12% for basically forever". 

 

The audience is different. If you're ERE and FIRE on $300k, having to pay $50k out of pocket for medical expenses is an "oh shit" moment.

If, however, you retire on $1 million, spending $50k is a reduction of $2000/year in withdrawals. With some flexibility in the budget, that's traveling internationally 1/3 years instead of 1/2 years. That's pretty modest.

Mostly Big ERN isn't addressing unforeseen medical bills.  He's addressing situations where the 4% rule hasn't worked historically.  So it's not reducing your annual spending by 5% (which is what your scenario discusses).  It's more like 7.5% to 12% permanent reduction depending on what scenario you are looking at, and that assumes you have perfect foresight and know ahead of time to start with a lower safe withdrawal rate.  If you start off with 4% and have an awful sequence of returns, you can have to cut much deeper to compensate.

I think people should understand this point. Flexibility is definitely important but it needs to be put into perspective and flexibility works both ways. I may have to increase my spending at some point. For instance we typically spend 40k per year for a family of 5 but my kids are turning 15 & 17 this year and we are spending more on them. So flexibility may require increase spending.

If the market crashes and I have to get a crappy part time job plus my expenses increase it could be a disaster.

I'm still okay with spending 5% but I have other back-ups. I can sell my house. I could potentially decrease my spending. I will receive social security a lot earlier than a 30 year window.

My take is understanding the limitations of flexibility is important. It's not just a hand waving exercise.

[steveo]

Yes.  I think having enough slop in the bucket is basically the equivalent of having a higher than 25x multiple of expenses.  You can look at it as retiring with a 4% withdrawal rate but having more than 10% of that going to luxury spending that you dont' mind cutting, or you can look at it as having a 3 and change % withdrawal rate but taking extra withdrawal each year to spend on luxuries as long as things are going decent.  Potato/Potato. 

Agreed. Having the luxury of having luxury in the ER budget is great if the SOR goes poorly. It also makes me think that I worked too long. OMY is psychologically tough to avoid for most of us, though, unless you finish with a windfall.

I view my spending this way but there are limitations to this as well. My wife has a terrible job so once we get to 5% at bare minimum spending she can quit. I don't want to push her to work longer. I intend to work a little past this point to allow some flexibility in spending and make my situation a little safer. I'll still though withdraw 5% when I RE based on market conditions.

[DreamFIRE]

I'm still okay with spending 5% but I have other back-ups. I can sell my house. I could potentially decrease my spending. I will receive social security a lot earlier than a 30 year window.

16 year window here (SS will cover my barebones), plus my barebones in FIRE will be less than 2% of my current stash, so more than half of a 4% SWR would be discretionary spending.  I'm a heavy saver, so it might be more difficult for me to spend it than to cut back.  lol

[steveo]

I'm still okay with spending 5% but I have other back-ups. I can sell my house. I could potentially decrease my spending. I will receive social security a lot earlier than a 30 year window.

16 year window here (SS will cover my barebones), plus my barebones in FIRE will be less than 2% of my current stash, so more than half of a 4% SWR would be discretionary spending.  I'm a heavy saver, so it might be more difficult for me to spend it than to cut back.  lol

I'm interested how people model this. I use cfiresim and model variable expenses (highish to bare bones but realistic) within the timeframe that I am looking to retire at. I aim to get to SS which as per your situation should cover my barebones spending (which I would be cool with).

The whole 30 or 30+ year retirement with fixed expenses is completely unrealistic for me.

We are also heavy savers and really don't have a huge urge to spend a bunch of money. Our highest predicted expenses are something we have never done so we could end up with the ability to spend more in retirement that what we spend now.

[FIRE 20/20]

Granted, if you don't mind working part time and/or can do so at a high pay, then flexibility is fine.  But unless you can jump back in the job market at a high pay, it's not working part time for a couple of years.  It's working part time for potentially a decade or more, compared to sticking with a job (that in his case he likes) for another year or two.  Definitely not applicable to everybody, but I think he's right that many people just saying "I'll be flexible" are looking at it as "I'll work part time for a few years", not "I'll work part time for more than a decade of my 'retirement'" or alternatively they are thinking "I'll reduce my withdrawal rate by half a percentage point for a few years, no big deal", and not "I'll reduce my withdrawal rate by more than 12% for basically forever". 

You acknowledge this in your post, but Big ERN doesn't (I don't think).  His assumptions are pretty far off from what I think anyone would reasonably do.  If a mustachian who retired at 25x found herself retiring in 1932 or 1967, she wouldn't just earn $10k / year.  She'd cut expenses to the bone, travel to a low cost country, and/or earn substantially more than $10k.  And ERN's concern about oversaving in these scenarios is amusing because the plans he advocates for are a nearly 100% guarantee of overworking and oversaving to avoid a very small chance of preventing an unlikely failure scenario.  Again, I am quite convinced (and Big ERN's posts only really confirm it for me) that someone who is willing to be flexible - really flexible, not the minor edge tweaks he talks about - should be more than comfortable FIREing at 25x expenses.  I think he is correct that someone who wants absolute security and is totally unwilling to really be flexible probably needs a lot more.  Whether that's worth guaranteeing that you're giving up the best remaining year of your life for a small chance and an unwillingness to be flexible - that's for the FIREee to decide. 

[Threshkin]

Here are a couple of ERN posts on flexibility

https://earlyretirementnow.com/2018/02/07/the-ultimate-guide-to-safe-withdrawal-rates-part-23-flexibility/
https://earlyretirementnow.com/2018/05/09/the-ultimate-guide-to-safe-withdrawal-rates-part-24-flexibility-myths-vs-reality/

Edited to fix links.

[Acastus]

Whatever SWR you choose, I think of it as a course heading, not a destination. Everyone will encounter trouble, some you can predict and some you can't. There will be many course corrections. Most of the time they will be leisurely. A single year's spending decision will not make or break the plan. Imagining you can map out the next 40 years with a simple equation you only use once is pretty simplistic.

All models are wrong. Some are useful.

[Threshkin]

Whatever SWR you choose, I think of it as a course heading, not a destination. Everyone will encounter trouble, some you can predict and some you can't. There will be many course corrections. Most of the time they will be leisurely. A single year's spending decision will not make or break the plan. Imagining you can map out the next 40 years with a simple equation you only use once is pretty simplistic.

All models are wrong. Some are useful.

Yes, Yes, YES!

[boarder42]

Whatever SWR you choose, I think of it as a course heading, not a destination. Everyone will encounter trouble, some you can predict and some you can't. There will be many course corrections. Most of the time they will be leisurely. A single year's spending decision will not make or break the plan. Imagining you can map out the next 40 years with a simple equation you only use once is pretty simplistic.

All models are wrong. Some are useful.

Yes, Yes, YES!

It's highly unlikely you'll face any issues. The higher probability is you can just spend at your goal spending forever.  And actually increase spending over time. Historically. It's no concern at all.

So saying everyone is a massive overstatement.

[exit2019]

Its no different than using data from the past to come up with a 3.5 or 4% SWR.  This analysis is much more valuable at predicting if you're going to be ok than just writing 29 blog posts to come up with what worked in the past an historically didnt fail - that takes 4 minutes on cFIREsim to determine a historical withdrawal rate that never failed

Most of ERNs series is about understanding both what failed actually means and what success and failure look like.

cFiresim doesn't really work well for really extended retirements using the default parameters because the data set is relatively small for retirements spanning 50 ot 60 years.

I already did all the cfiresim/firecalc/VPW/personal spreadsheet stuff long before ERNs series and yet his series was incredibly valuable.  In particular, it was value to remove confidence in a number of naive assumptions that cfiresim encourages by default.  In particular, there is no way to know at any given time whether you're OK or not, despite historical sequences, and so in the moment one will need to have a good grip on the fundamentals.

In addition, his review of Kitces work (which cherry picks bond classes after the fact) and of rising equity glidepaths is very interesting.  Also very interesting is that his actual retirement approach, despite all of that, is to become a long term options trader rather than truly retiring it.

[exit2019]

I can't help but think a lot of people haven't actually read Big ERN's posts.  The entire point of his posts is there is nothing "modest" about the flexibility (whether going back to work or cutting back on spending) required to avoid portfolio failures and portfolio failure (or the flexibility required to avoid it) is a much, much, much worse result for many people than simply working until they have a lower initial withdrawal rate. 

another way to say it is this: people who talk flexibility have mostly never quantitatively analyzed what that actually means, nor have they considered that you lack information at the time flexibility decisions need to be made - going back to work, cutting costs, depleting a cash cushion, etc. are all decisions one has to make without knowing what the future holds.

ERN makes a very strong quantitative case that an extra two or three years of work, especially since each of those years will typically be at your earning peak as you generally make more annually per year until your 50s - to get to a lower and safer retirement rate is the right move.  The main argument against it is that people don't want to work another few years.

It's especially weird IMHO given how much uncertainty there is about economic growth and healthcare expenses going forward; nothing even remotely like the current situation is really to be found in the historical data so short of simulations (as opposed to backtesting) with some reasonable parameters confidence should be low.

[PizzaSteve]

I can't help but think a lot of people haven't actually read Big ERN's posts.  The entire point of his posts is there is nothing "modest" about the flexibility (whether going back to work or cutting back on spending) required to avoid portfolio failures and portfolio failure (or the flexibility required to avoid it) is a much, much, much worse result for many people than simply working until they have a lower initial withdrawal rate. 

another way to say it is this: people who talk flexibility have mostly never quantitatively analyzed what that actually means, nor have they considered that you lack information at the time flexibility decisions need to be made - going back to work, cutting costs, depleting a cash cushion, etc. are all decisions one has to make without knowing what the future holds.

ERN makes a very strong quantitative case that an extra two or three years of work, especially since each of those years will typically be at your earning peak as you generally make more annually per year until your 50s - to get to a lower and safer retirement rate is the right move.  The main argument against it is that people don't want to work another few years.

It's especially weird IMHO given how much uncertainty there is about economic growth and healthcare expenses going forward; nothing even remotely like the current situation is really to be found in the historical data so short of simulations (as opposed to backtesting) with some reasonable parameters confidence should be low.

While not necessarily advocating a higher WR, I thought it worth stating the obvious that one might also consider that having more available disposable income is an additional 'upside' of working a few extra years.

This is a factor that each individual might consider in their plans. 

I think he is assuming that people reading his work are first world types, who have work that is reasonably enjoyable, and where the money is relatively good.  Hence the conclusion to advocate a plan where people might choose to work a few extra years. 

With a little extra cushion one is both prepared for an unexpected medical issue and has more flexibility to splurge on a once and a lifetime travel or hobby opportunities, that might be expensive. If ones priority is to see the world and spending extra money brings no joy, then go for 6%+, as that fulfills your life dream.

Its all individual choices, as has been often discussed.

[exit2019]

I think he is assuming that people reading his work are first world types, who have work that is reasonably enjoyable, and where the money is relatively good.  Hence the conclusion to advocate a plan where people might choose to work a few extra years. 

I agree strongly with this.  specifically, ERNs audience is well-paid white collar professionals for whom the age-compensation curve increases steadily and sharply with experience.  these are people who basically make enough in the age range, say, 40-45 than they are able to save more in a few years than they were able to in their entire career up to that point.  Karsten himself notes that most of his net worth was after he moved to the capital management company in terms of compensation, delayed compensation and the ability to buy housing relatively low thanks to the high-salary the job afforded.

That's a very tech/finance point of view, and there's a lot to learn from it regardless of your situation, but a lot of his caution is driven by it: once you step out of tech or finance, you are not going to be stepping back in 3-4 years later at anything even slightly close to your comp at the moment you left.  if your "go back to work" scenario involves working 10Y for 20%-30% of your pre-FIRE annual compensation (this is easily possible in the bay area, for example) it's far better just to work another two-three years.