Practicalities of living off the 4% SWR

[arebelspy]

I'm bad at jokes.

Acknowledged.

[lordmetroid]

Is the 4% withdrawal rate before taxes?
How much of the withdrawn money is lost to taxes?

[arebelspy]

Is the 4% withdrawal rate before taxes?
How much of the withdrawn money is lost to taxes?

4% should cover all your expenses. Including g taxes.

Some people will have zero taxes. Some a lot of taxes.

That will depend on a lot of things, including the amount you're withdrawing, the source of your funds (Roth versus 401k treated very differently), etc.

Calculate how much you need, total, including taxes (if any), then multiply by 25. This is the amount you need, so that withdrawing 4% covers all your expenses, including any necessary taxes.

Expense ratios should also be included in this. Any and all expenses.

[boarder42]

60/40 bonds split is way to conservative for a 4% SWR you need 80.20 minimum

Except that 60/40 was the benchmark used by both Bengen and the Trinity Studies that established the 4% rule in the first place, and continues to hold up in more recent studies.

Trinity study was based on 30 years we're not looking at 30 here we're looking at 60+

But if you succeed at 30, very likely your account is silly high and should have no problem succeeding at 60.

I was referring to the 60/40 stick to bond split. It is way worse at maintainingyour wealth then 80/20

Basically your asset allocation matters a lot more to the success than it does for people only looking at 30 years

You're setting yourself up to likely work more holding less than 80/20. 1MM portfolio on 4% see on cfiresim is a full 10% better success with 80/20 vs 60/40.

This goes back to the risk of opportunity cost. And people who are conservative gloss over the fact that they may actually be creating more risk trying to be less risky.

[arebelspy]

You're setting yourself up to likely work more holding less than 80/20.

I agree with this.

who are conservative gloss over the fact that they may actually be creating more risk trying to be less risky.

And this.

Trinity study was based on 30 years we're not looking at 30 here we're looking at 60+

I don't agree with this.  Because if you make it to 30 years, you should make it to 60, the vast, vast majority of the time.

I do think people will need to work longer with a lower equities allocation.  That may be worth it to them, for the less volatile portfolio.  They should realize and understand that tradeoff.  You're 100% correct.

The part I was disagreeing with is calling out 60 years as being that different from 30.

Run the numbers on the average portfolio (or, heck, even a portfolio that's at the bottom 20% of results) and see how much it ended up with as a final value at 30 years, and what the withdrawal amount was at that point.

Now take that amount and withdrawal and run it for another 30 years. 

I'm saying that the vast majority of the time, your portfolio will have grown (historically, obviously) so much that you're well under a 4% WR by the time you hit 30 years, and I'd bet your success rate for 30 more is close to 100%.  So the difference between 30 and 60 years isn't super significant.

I do think anyone looking for an AA less than 80/20 should run the numbers on how much longer they'd have to work to hit the same percent success rate with their preferred allocation, and weigh that extra amount of work versus the volatility.  It may be worth it for some, or not, but knowing the numbers is key.

[boarder42]

At a 60/40 split making it to 30 is not the same as making it to 60 as the Trinity study specifically looked at just a 30 year horizon. It's misleading to say if you make it to 30 you'll make it to 60 you can make it to 30 with 100 dollars left or 5MM dollars left. 60/40 does not grow like 80/20 that's the issue I have with your blanket statement. Of making it to 30.

[deborah]

OP is Australian. As such, they have access to a number of old aged safety nets from the age of 67, so a 60 year retirement would probably really be 30 years that they have to rely upon their stash. A mustachian probably won't get the old aged safety nets because their stash will be too big, but it is available if everything turns to mush.

Also, our equivalent of the 401k, SS, Roth... is superannuation. This is money that you cannot access until you are 60. It came into being by people not getting several pay rises, but instead making it somewhat mandatory (if you earn a really really small amount it doesn't apply) for every employer to put (9.5% currently) of your salary into the fund that you have chosen, in your name. Some employers put in more, but this is unusual. So most mustache Australians have a lump sum that they can do anything they like with once they reach 60. But, they must withdraw at least something between 4% a year (age 60) and 14% a year (age 95) - so the Trinity study, where people automatically take a certain percentage every year actually DOES apply to superannuation. This again means that most of us have an extra stash which means our initial stash only needs to last for 30 years - not 60.

We can top it up ourselves, and it is very tempting to do so because of the really good tax breaks it gives us. Most of us have too much in superannuation, and there have been a number of discussions about just how much you should put into super yourself. It is also worth while to top it up when you are about to reach 60 because the income you get from it is effectively tax free (again, this is a bit of a simplification, but not much).

[SpreadsheetMan]

.....

I do think anyone looking for an AA less than 80/20 should run the numbers on how much longer they'd have to work to hit the same percent success rate with their preferred allocation, and weigh that extra amount of work versus the volatility.  It may be worth it for some, or not, but knowing the numbers is key.

Actually I ran the numbers with Firecalc based on a 35 year period with the target of 100% to see min portfolio value with different equity/bond splits and my proposed draw. This was to see how much it mattered getting the allocation precise.

The lowest needed portfolio was at 40:60 and was pretty similar in the range 40:60 - 70:40, so not very sensitive at all. It was worse either side of that range with 30% and 100% about the same. Below 30% got pretty bad.

I concluded that moving to a higher proportion of bonds makes sense in drawdown when the important thing is the money not running out, but the exact split isn't that important as long as it is ballpark. More equities makes sense in accumulation to try and keep the time needed to build the stache as low as possible. Pretty unsurprising really, but good to check.

[arebelspy]

At a 60/40 split making it to 30 is not the same as making it to 60 as the Trinity study specifically looked at just a 30 year horizon. It's misleading to say if you make it to 30 you'll make it to 60 you can make it to 30 with 100 dollars left or 5MM dollars left. 60/40 does not grow like 80/20 that's the issue I have with your blanket statement. Of making it to 30.

I'm not talking about a study from 20 years ago, I'm talking about running the numbers yourself. cFIREsim makes this easy to do nowadays.

[Metric Mouse]

.....

I do think anyone looking for an AA less than 80/20 should run the numbers on how much longer they'd have to work to hit the same percent success rate with their preferred allocation, and weigh that extra amount of work versus the volatility.  It may be worth it for some, or not, but knowing the numbers is key.

Actually I ran the numbers with Firecalc based on a 35 year period with the target of 100% to see min portfolio value with different equity/bond splits and my proposed draw. This was to see how much it mattered getting the allocation precise.

The lowest needed portfolio was at 40:60 and was pretty similar in the range 40:60 - 70:40, so not very sensitive at all. It was worse either side of that range with 30% and 100% about the same. Below 30% got pretty bad.

I concluded that moving to a higher proportion of bonds makes sense in drawdown when the important thing is the money not running out, but the exact split isn't that important as long as it is ballpark. More equities makes sense in accumulation to try and keep the time needed to build the stache as low as possible. Pretty unsurprising really, but good to check.

Thanks for running the numbers! Focusing on not hitting zero, rather than worrying about making as much as possible, does probably lead to different assest allocations.  A more bond heavy-portfolio seems to limit the movement of the 'stache (both up and down), thereby reducing the sequence of return risks, assuming one has a large enough pile.

Would be interesting to run some numbers with a dynamic AA - retire with the 70:30 split, and move towards more equities as time goes on and the sequence of return risk is lower, allowing more 'upside' to be captured.

[arebelspy]

Would be interesting to run some numbers with a dynamic AA - retire with the 70:30 split, and move towards more equities as time goes on and the sequence of return risk is lower, allowing more 'upside' to be captured.

This has been researched (by Pfau & Kitces), and discussed.

Article here.

The phrase you want to search (both Google and here) for more info/discussion is "rising glidepath."

It was fairly novel a few years ago, as conventional wisdom is to get more conservative (i.e. more bonds) as you age, and this is the opposite.

It does seem, according to the research, to improve success rates.

[Metric Mouse]

Would be interesting to run some numbers with a dynamic AA - retire with the 70:30 split, and move towards more equities as time goes on and the sequence of return risk is lower, allowing more 'upside' to be captured.

This has been researched (by Pfau & Kitces), and discussed.

Article here.

The phrase you want to search (both Google and here) for more info/discussion is "rising glidepath."

It was fairly novel a few years ago, as conventional wisdom is to get more conservative (i.e. more bonds) as you age, and this is the opposite.

It does seem, according to the research, to improve success rates.

Glidepath - I was completely drawing a blank on the term. Thank you.

[boarder42]

At a 60/40 split making it to 30 is not the same as making it to 60 as the Trinity study specifically looked at just a 30 year horizon. It's misleading to say if you make it to 30 you'll make it to 60 you can make it to 30 with 100 dollars left or 5MM dollars left. 60/40 does not grow like 80/20 that's the issue I have with your blanket statement. Of making it to 30.

I'm not talking about a study from 20 years ago, I'm talking about running the numbers yourself. cFIREsim makes this easy to do nowadays.

i agree the comment i made you were countering was in response to the trinity study specifically that someone was trying to use to prove that a 60/40 split was good. its not its worse than not keeping a mortgage when it comes to safety in FIRE historically.

[arebelspy]

Regarding the discussion of 30 years v 60 years... Here we go, numbers!

cFIREsim.  60/40, leaving everything else default.

After 30 years, 60/40 had a 94.83% success rate.  80/20 had 96.55% success rate.

1MM portfolio on 4% see on cfiresim is a full 10% better success with 80/20 vs 60/40.

Not sure where you got a 10% difference, as I'm seeing under a 2% difference.  My inputs as above, what did you use?

Digging deeper into the numbers, I downloaded the year by year results for 60-40 and 80-20.  I took out the failures, since what I posited was that if you succeed at 30 years, you're likely to succeed at 60.  If you failed at 30, obviously you'd fail at 60.

Then I took the median.

60-40: 1,105,525
80-20: 1,724,751.5

More equities obviously gives you a lot more in the end, but at the cost of a much bumpier ride.

The inflation adjusted withdrawals at this point is still 40k, as it's gone up just to match inflation each year.  Running the numbers of 40k inflation-adjusted withdrawals on our new portfolio amounts, we get a 99.14% success rate (60-40) versus 100% (80-20).

Now, naturally, if times are worse in the future, having the extra cash from the 80/20 is good.  If you can stomach the volitility.

I think everyone should be at least 80/20, if not 90/10, and work on training themselves to not worry about market fluxuations.

But most of the time (most being defined as > 50%), if you make it with 60/40 to 30 years, you'll be able to make it to 60 years, as your WR will have declined as your portfolio increased in real terms.

About 3/4ths of the time, you'd have a 50/50 chance or better to make it.  The bottom 25% you might not, and that's where the extra money from the 80-20 comes in handy.

[arebelspy]

At a 60/40 split making it to 30 is not the same as making it to 60 as the Trinity study specifically looked at just a 30 year horizon. It's misleading to say if you make it to 30 you'll make it to 60 you can make it to 30 with 100 dollars left or 5MM dollars left. 60/40 does not grow like 80/20 that's the issue I have with your blanket statement. Of making it to 30.

I'm not talking about a study from 20 years ago, I'm talking about running the numbers yourself. cFIREsim makes this easy to do nowadays.

i agree the comment i made you were countering was in response to the trinity study specifically that someone was trying to use to prove that a 60/40 split was good. its not its worse than not keeping a mortgage when it comes to safety in FIRE historically.

Ah, we were cross posting (was typing the above post w/ data when you posted this).  I follow.

Yes, I agree, in general if you're going for 60 years of FIRE, you're much safer, historically, with 80-20.  A few percent better success rate, and much higher balance for if TSHTF.

I think you can make it most times with 60-40, but you're right that 80-20 is less risky--even though it's more volatile.  Those who equate the two and go more "conservative" add more portfolio failure risk.

[Metric Mouse]

Regarding the discussion of 30 years v 60 years... Here we go, numbers!

Are those balance numbers inflation adjusted? I've always struggled with Cfiresim - it seems clunky to me.  If those balance numbers are not inflation adjusted after 30 years (obviously the withdrawl amount is) then that would change the numbers dramatically.

[boarder42]

Regarding the discussion of 30 years v 60 years... Here we go, numbers!

cFIREsim.  60/40, leaving everything else default.

After 30 years, 60/40 had a 94.83% success rate.  80/20 had 96.55% success rate.

1MM portfolio on 4% see on cfiresim is a full 10% better success with 80/20 vs 60/40.

Not sure where you got a 10% difference, as I'm seeing under a 2% difference.  My inputs as above, what did you use?

Digging deeper into the numbers, I downloaded the year by year results for 60-40 and 80-20.  I took out the failures, since what I posited was that if you succeed at 30 years, you're likely to succeed at 60.  If you failed at 30, obviously you'd fail at 60.

Then I took the median.

60-40: 1,105,525
80-20: 1,724,751.5

More equities obviously gives you a lot more in the end, but at the cost of a much bumpier ride.

The inflation adjusted withdrawals at this point is still 40k, as it's gone up just to match inflation each year.  Running the numbers of 40k inflation-adjusted withdrawals on our new portfolio amounts, we get a 99.14% success rate (60-40) versus 100% (80-20).

Now, naturally, if times are worse in the future, having the extra cash from the 80/20 is good.  If you can stomach the volitility.

I think everyone should be at least 80/20, if not 90/10, and work on training themselves to not worry about market fluxuations.

But most of the time (most being defined as > 50%), if you make it with 60/40 to 30 years, you'll be able to make it to 60 years, as your WR will have declined as your portfolio increased in real terms.

About 3/4ths of the time, you'd have a 50/50 chance or better to make it.  The bottom 25% you might not, and that's where the extra money from the 80-20 comes in handy.

yes i used a 40 year time horizon b/c kitces and Fientist both have posts that if it makes it 40 years it will make it FOREVER years.  so per that analysis. yes you're seeing 2% difference at 30 which is a small difference but put it on the 40 year horizon which should mean your money will make it forever and you see a large gap. the further you push the time horizon the more likely 60/40 fails ( i know we are going to start dropping some down years but it gets worse really fast as you go up) 

60/40
30 years - 94.83
40 years - 80
50 years - 72.63
60 years - 76.47

80/20
30 years - 96.55
40 years - 89.52
50 years - 85.26
60 years - 85.88 - higher b/c we drop some bad years b/c of the limitation of the data

[boarder42]

at the end of the day if you put a gun to my head and said you're working til 65 or you could retire today and cover all your expenses at exactly 4% SWR fixed rising with inflation and you have to have a 60/40 allocation and you cant earn another dollar.  i'd take my chances on quitting and the 72% success rate.

[arebelspy]

Regarding the discussion of 30 years v 60 years... Here we go, numbers!

Are those balance numbers inflation adjusted? I've always struggled with Cfiresim - it seems clunky to me.  If those balance numbers are not inflation adjusted after 30 years (obviously the withdrawl amount is) then that would change the numbers dramatically.

Yes, they are all in real dollars.  Talking in nominal dollars is almost always silly, but especially when looking at 30-60 year time frames.  :)

yes i used a 40 year time horizon

Ah, I see, 10% success rate difference at 40 years.  Thanks for clarifying.

at the end of the day if you put a gun to my head and said you're working til 65 or you could retire today and cover all your expenses at exactly 4% SWR fixed rising with inflation and you have to have a 60/40 allocation and you cant earn another dollar.  i'd take my chances on quitting and the 72% success rate.

Hah.  Me too.

Good thing we don't have to make that choice.  90/10 it is*!

;)


*Though I obviously have only a small, but growing, part of my investments in paper assets, due to real estate.

[Metric Mouse]

at the end of the day if you put a gun to my head and said you're working til 65 or you could retire today and cover all your expenses at exactly 4% SWR fixed rising with inflation and you have to have a 60/40 allocation and you cant earn another dollar.  i'd take my chances on quitting and the 72% success rate.

If that were the choice, I would probably inflate my expenses as much as I could... :D

Or decide to work as a romance novelist and claim writers block for the next 40 years...

[boarder42]

yep 90/10 is even better yet.  that is actually where i will reside but there are really mostly diminishing returns once you hit 80/20

[dude]

You're setting yourself up to likely work more holding less than 80/20.

I agree with this.

who are conservative gloss over the fact that they may actually be creating more risk trying to be less risky.

And this.

Trinity study was based on 30 years we're not looking at 30 here we're looking at 60+

I don't agree with this.  Because if you make it to 30 years, you should make it to 60, the vast, vast majority of the time.

I do think people will need to work longer with a lower equities allocation.  That may be worth it to them, for the less volatile portfolio.  They should realize and understand that tradeoff.  You're 100% correct.

The part I was disagreeing with is calling out 60 years as being that different from 30.

Run the numbers on the average portfolio (or, heck, even a portfolio that's at the bottom 20% of results) and see how much it ended up with as a final value at 30 years, and what the withdrawal amount was at that point.

Now take that amount and withdrawal and run it for another 30 years. 

I'm saying that the vast majority of the time, your portfolio will have grown (historically, obviously) so much that you're well under a 4% WR by the time you hit 30 years, and I'd bet your success rate for 30 more is close to 100%.  So the difference between 30 and 60 years isn't super significant.

I do think anyone looking for an AA less than 80/20 should run the numbers on how much longer they'd have to work to hit the same percent success rate with their preferred allocation, and weigh that extra amount of work versus the volatility.  It may be worth it for some, or not, but knowing the numbers is key.

Agree with you (arebelspy) on all  -- in the accumulation phase, 60/40 will almost certainly take you longer.  But the 4% SWR applies to the withdrawal phase -- which is of course what the seminal studies looked at, and for that phase, 60/40 was the benchmark. Not saying others won't work, but saying that 60/40 won't cut it (as the person I originally responded to stated), simply isn't what that 4% rule says at all. In fact, it says the opposite.

[boarder42]

You're setting yourself up to likely work more holding less than 80/20.

I agree with this.

who are conservative gloss over the fact that they may actually be creating more risk trying to be less risky.

And this.

Trinity study was based on 30 years we're not looking at 30 here we're looking at 60+

I don't agree with this.  Because if you make it to 30 years, you should make it to 60, the vast, vast majority of the time.

I do think people will need to work longer with a lower equities allocation.  That may be worth it to them, for the less volatile portfolio.  They should realize and understand that tradeoff.  You're 100% correct.

The part I was disagreeing with is calling out 60 years as being that different from 30.

Run the numbers on the average portfolio (or, heck, even a portfolio that's at the bottom 20% of results) and see how much it ended up with as a final value at 30 years, and what the withdrawal amount was at that point.

Now take that amount and withdrawal and run it for another 30 years. 

I'm saying that the vast majority of the time, your portfolio will have grown (historically, obviously) so much that you're well under a 4% WR by the time you hit 30 years, and I'd bet your success rate for 30 more is close to 100%.  So the difference between 30 and 60 years isn't super significant.

I do think anyone looking for an AA less than 80/20 should run the numbers on how much longer they'd have to work to hit the same percent success rate with their preferred allocation, and weigh that extra amount of work versus the volatility.  It may be worth it for some, or not, but knowing the numbers is key.

Agree with you (arebelspy) on all  -- in the accumulation phase, 60/40 will almost certainly take you longer.  But the 4% SWR applies to the withdrawal phase -- which is of course what the seminal studies looked at, and for that phase, 60/40 was the benchmark. Not saying others won't work, but saying that 60/40 won't cut it (as the person I originally responded to stated), simply isn't what that 4% rule says at all. In fact, it says the opposite.

see above its great for 30 years - drops off dramatically as time goes on.  just moving the yard sticks to 40 years makes it a 10% worse decision than 80/20. 

and as ARS stated who the F cares about the damn trinity study or what it said we have market data we can back test multiple withdrawal startegies against with cfiresim and others. 

And since we have 50 page threads on why paying down your mortgage is bad to gain 2-3% ... 10% is huge.   AA should be more extensively fought topic around here than paying down a mortgage or keeping it in FIRE>

[dude]

Would be interesting to run some numbers with a dynamic AA - retire with the 70:30 split, and move towards more equities as time goes on and the sequence of return risk is lower, allowing more 'upside' to be captured.

This has been researched (by Pfau & Kitces), and discussed.

Article here.

The phrase you want to search (both Google and here) for more info/discussion is "rising glidepath."

It was fairly novel a few years ago, as conventional wisdom is to get more conservative (i.e. more bonds) as you age, and this is the opposite.

It does seem, according to the research, to improve success rates.

But not by much, and not in all cases.  I shared another of their studies not too long ago, and the surprising takeaway was that the 60/40 combo was as good or better in many cases, though contrary to intuition (mine anyway), it worked best when the 40% was in treasuries as opposed to bonds.

[dude]

You're setting yourself up to likely work more holding less than 80/20.

I agree with this.

who are conservative gloss over the fact that they may actually be creating more risk trying to be less risky.

And this.

Trinity study was based on 30 years we're not looking at 30 here we're looking at 60+

I don't agree with this.  Because if you make it to 30 years, you should make it to 60, the vast, vast majority of the time.

I do think people will need to work longer with a lower equities allocation.  That may be worth it to them, for the less volatile portfolio.  They should realize and understand that tradeoff.  You're 100% correct.

The part I was disagreeing with is calling out 60 years as being that different from 30.

Run the numbers on the average portfolio (or, heck, even a portfolio that's at the bottom 20% of results) and see how much it ended up with as a final value at 30 years, and what the withdrawal amount was at that point.

Now take that amount and withdrawal and run it for another 30 years. 

I'm saying that the vast majority of the time, your portfolio will have grown (historically, obviously) so much that you're well under a 4% WR by the time you hit 30 years, and I'd bet your success rate for 30 more is close to 100%.  So the difference between 30 and 60 years isn't super significant.

I do think anyone looking for an AA less than 80/20 should run the numbers on how much longer they'd have to work to hit the same percent success rate with their preferred allocation, and weigh that extra amount of work versus the volatility.  It may be worth it for some, or not, but knowing the numbers is key.

Agree with you (arebelspy) on all  -- in the accumulation phase, 60/40 will almost certainly take you longer.  But the 4% SWR applies to the withdrawal phase -- which is of course what the seminal studies looked at, and for that phase, 60/40 was the benchmark. Not saying others won't work, but saying that 60/40 won't cut it (as the person I originally responded to stated), simply isn't what that 4% rule says at all. In fact, it says the opposite.

see above its great for 30 years - drops off dramatically as time goes on.  just moving the yard sticks to 40 years makes it a 10% worse decision than 80/20. 

and as ARS stated who the F cares about the damn trinity study or what it said we have market data we can back test multiple withdrawal startegies against with cfiresim and others. 

And since we have 50 page threads on why paying down your mortgage is bad to gain 2-3% ... 10% is huge.   AA should be more extensively fought topic around here than paying down a mortgage or keeping it in FIRE>

Okay, on that we agree.