Foolishness of the 4% rule

[Deleted User]

MoJoeGo said:

jamesd said:

kuratowski said:

As described above this is a pure strawman.  How many people would just blindly pick a number the day they retire and then draw that much down forever, refusing to take any other precautions such as 3 years' cash buffer, and stubbornly refuse to adjust despite their portfolio shrinking.

...

I'm pretty keen on retiring early, so I have been working on the basis of 3%, i.e. to cover a longer time period than 4% was designed for.  This is despite being on track for full new state pension, and having a deferred DB pension amount to about half SP, together these two would cover my basic needs, so I like to believe my plans are pretty cautious.

People are expected to use the constant inflation-adjusted income rule though it is recommended by assorted people including me to recalculate income since it doesn't incorporate market gains and assumes you're living through the historic worst case. If you do live through something worse you'd be expected to notice and react. But US average for this rule is 7% and it starts at only 4.1 or 4.2%, both for 30 years, so it's extremely cautious. Start at the UK 4% rule level and blindly follow it and your chance of failure is very low.

At 3% you're planning on either worse than historic UK performance or living more than 45 years or using less than optimal investments, since 45 years US calculated 4% rule starts at 4.1% with 65% equities. Deducting the usual 0.3 for the UK that's 3.8%. Deducting a third* of say 0.6% in total costs cuts it to 3.6% for 45 years.

*you don't deduct 100% of the costs on the initial balance because the balance and hence the costs decrease during the almost but not quite failing worst case. The right number turns out to be around a third of costs but there is variation. I pretty uniformly use a third or 30% because it's close enough, correct for the common 30 year US case and the error margin is lost in the market unpredictability noise.

Or of course 3% is enough for your needs - for the missus and me, that should produce about 30k each, and then there's the state pension to come later. My guess is that will be more than enough for a very comfy retirement, and also a good amount to pass on to our children...

Lets say you have 1 million pounds at the start of your retirement in 2022. Your portfolio drops to 300K on January 2nd 2023.  How much are you going to withdraw? 

bostonerimus

Ibrahim5 said:

You would have to have a different figure for people who use IFAs because the IFA takes out a large proportion of the investment returns for their fees.

No you wouldn't. SWR is only about what level of withdrawal a pot can sustain, not what you spend it on. Two retirees with identical pots that are identically managed will have identical SWR, but the one without an IFA will have a bit more money to spend on other things. So if you have investment fees, taxes, debt payments, holidays, food etc they are all just expenses you need to pay for with your SWR.

[Deleted User]

bostonerimus said:

Ibrahim5 said:

You would have to have a different figure for people who use IFAs because the IFA takes out a large proportion of the investment returns for their fees.

No you wouldn't. SWR is only about what level of withdrawal a pot can sustain, not what you spend it on. Two retirees with identical pots that are identically managed will have identical SWR, but the one without an IFA will have a bit more money to spend on other things. So if you have investment fees, taxes, debt payments, holidays, food etc they are all just expenses you need to pay for with your SWR.

The comment is valid because people mean different things when they say “4% SWR”. Many use the number to apply after taxes and investment fees.  Regardless, fundamental problem is letter “S” in “SWR”. Mathematically there can be no “safe” and constant rate of withdrawal from a highly variable portfolio.

michaels

My current working on a UK post fees, 45 year retirement horizon, 70% global equities, 30% UK cash/short bonds is a baseline SWR of between 3 and 3.25% (lowered because of the current CAPE).  I then uplift this by about 1% to take into account two full state pensions.

One thing I find interesting is if you look at the historic SWR capital pot paths you see many where valuations fall to pretty scarily low multiples of the annual withdrawal amount with not many year passed before the markets soar away.  In reality no one would hold there nerve if they were down to 6 years cover 15 year in so SWR is probably a rule that will never get tested to destruction.

Audaxer

The SWR of 4% may be a useful as a yardstick and have worked historically, but if I really needed to withdraw at least 4% (rising with inflation) each year from my investments, with no cash buffer, I wouldn't consider it that safe. I think if possible you should aim for a big enough retirement pot to have enough income for a comfortable retirement by drawing only 3% rising with inflation each year. That seems a much safer figure to aim for, and you would still have the option to increase withdrawals if you get a good sequence of returns.

MoJoeGo

Deleted_User said:

MoJoeGo said:

jamesd said:

kuratowski said:

As described above this is a pure strawman.  How many people would just blindly pick a number the day they retire and then draw that much down forever, refusing to take any other precautions such as 3 years' cash buffer, and stubbornly refuse to adjust despite their portfolio shrinking.

...

I'm pretty keen on retiring early, so I have been working on the basis of 3%, i.e. to cover a longer time period than 4% was designed for.  This is despite being on track for full new state pension, and having a deferred DB pension amount to about half SP, together these two would cover my basic needs, so I like to believe my plans are pretty cautious.

People are expected to use the constant inflation-adjusted income rule though it is recommended by assorted people including me to recalculate income since it doesn't incorporate market gains and assumes you're living through the historic worst case. If you do live through something worse you'd be expected to notice and react. But US average for this rule is 7% and it starts at only 4.1 or 4.2%, both for 30 years, so it's extremely cautious. Start at the UK 4% rule level and blindly follow it and your chance of failure is very low.

At 3% you're planning on either worse than historic UK performance or living more than 45 years or using less than optimal investments, since 45 years US calculated 4% rule starts at 4.1% with 65% equities. Deducting the usual 0.3 for the UK that's 3.8%. Deducting a third* of say 0.6% in total costs cuts it to 3.6% for 45 years.

*you don't deduct 100% of the costs on the initial balance because the balance and hence the costs decrease during the almost but not quite failing worst case. The right number turns out to be around a third of costs but there is variation. I pretty uniformly use a third or 30% because it's close enough, correct for the common 30 year US case and the error margin is lost in the market unpredictability noise.

Or of course 3% is enough for your needs - for the missus and me, that should produce about 30k each, and then there's the state pension to come later. My guess is that will be more than enough for a very comfy retirement, and also a good amount to pass on to our children...

Lets say you have 1 million pounds at the start of your retirement in 2022. Your portfolio drops to 300K on January 2nd 2023.  How much are you going to withdraw? 

I think if you are seriously suggesting a 70% drop that is sustained for a period then I'll probably be seeing you in the soup kitchen! Except of course it would be even more than 70% drop in the wider equity markets because we'd have a few years set aside in cash and other lower risk holdings. So g'night John Boy!

Seriously though, that would still give 18k between us at 3%, and whilst it would be frugal, we'd not exactly be on the breadline. Perhaps downsize, assuming the house was still worth more than when it was built in 1730...

Oh and in case I didn't mention before, Mrs MJG is fairly conservative and we have a lot of cash knocking around that I'd be wanting to invest if there really was a slump of that magnitude. That could be an interesting discussion!

My VaR models can only go so far...

bostonerimus

Deleted_User said:

bostonerimus said:

Ibrahim5 said:

You would have to have a different figure for people who use IFAs because the IFA takes out a large proportion of the investment returns for their fees.

No you wouldn't. SWR is only about what level of withdrawal a pot can sustain, not what you spend it on. Two retirees with identical pots that are identically managed will have identical SWR, but the one without an IFA will have a bit more money to spend on other things. So if you have investment fees, taxes, debt payments, holidays, food etc they are all just expenses you need to pay for with your SWR.

The comment is valid because people mean different things when they say “4% SWR”. Many use the number to apply after taxes and investment fees.  Regardless, fundamental problem is letter “S” in “SWR”. Mathematically there can be no “safe” and constant rate of withdrawal from a highly variable portfolio.

Yes it's a bit of a mine field because people make assumptions and don't take the time to understand the parameters of the modeling. "SWR" depends greatly on a whole set of assumptions, but it does not include anything about what you spend the money on be it taxes, investments fees or rent. If it was to include things like investment fees it would be even more of a moving target. And, being cynical, I can see how financial advisors like Bergen would want to exclude them as they might take up to half of your initial spending when your retire.

Thrugelmir

bostonerimus said:

Deleted_User said:

bostonerimus said:

Ibrahim5 said:

You would have to have a different figure for people who use IFAs because the IFA takes out a large proportion of the investment returns for their fees.

No you wouldn't. SWR is only about what level of withdrawal a pot can sustain, not what you spend it on. Two retirees with identical pots that are identically managed will have identical SWR, but the one without an IFA will have a bit more money to spend on other things. So if you have investment fees, taxes, debt payments, holidays, food etc they are all just expenses you need to pay for with your SWR.

The comment is valid because people mean different things when they say “4% SWR”. Many use the number to apply after taxes and investment fees.  Regardless, fundamental problem is letter “S” in “SWR”. Mathematically there can be no “safe” and constant rate of withdrawal from a highly variable portfolio.

Yes it's a bit of a mine field because people make assumptions and don't take the time to understand the parameters of the modeling. "SWR" depends greatly on a whole set of assumptions, but it does not include anything about what you spend the money on be it taxes, investments fees or rent. If it was to include things like investment fees it would be even more of a moving target. And, being cynical, I can see how financial advisors like Bergen would want to exclude them as they might take up to half of your initial spending when your retire.

Investment returns are influenced by rates of corporate tax.  That's a minefield that's already being laid ahead. Though hasn't appeared to have registered yet with many investors given their expected growth rates in the years ahead. 

[Deleted User]

MoJoeGo said:

Deleted_User said:

MoJoeGo said:

jamesd said:

kuratowski said:

As described above this is a pure strawman.  How many people would just blindly pick a number the day they retire and then draw that much down forever, refusing to take any other precautions such as 3 years' cash buffer, and stubbornly refuse to adjust despite their portfolio shrinking.

...

I'm pretty keen on retiring early, so I have been working on the basis of 3%, i.e. to cover a longer time period than 4% was designed for.  This is despite being on track for full new state pension, and having a deferred DB pension amount to about half SP, together these two would cover my basic needs, so I like to believe my plans are pretty cautious.

People are expected to use the constant inflation-adjusted income rule though it is recommended by assorted people including me to recalculate income since it doesn't incorporate market gains and assumes you're living through the historic worst case. If you do live through something worse you'd be expected to notice and react. But US average for this rule is 7% and it starts at only 4.1 or 4.2%, both for 30 years, so it's extremely cautious. Start at the UK 4% rule level and blindly follow it and your chance of failure is very low.

At 3% you're planning on either worse than historic UK performance or living more than 45 years or using less than optimal investments, since 45 years US calculated 4% rule starts at 4.1% with 65% equities. Deducting the usual 0.3 for the UK that's 3.8%. Deducting a third* of say 0.6% in total costs cuts it to 3.6% for 45 years.

*you don't deduct 100% of the costs on the initial balance because the balance and hence the costs decrease during the almost but not quite failing worst case. The right number turns out to be around a third of costs but there is variation. I pretty uniformly use a third or 30% because it's close enough, correct for the common 30 year US case and the error margin is lost in the market unpredictability noise.

Or of course 3% is enough for your needs - for the missus and me, that should produce about 30k each, and then there's the state pension to come later. My guess is that will be more than enough for a very comfy retirement, and also a good amount to pass on to our children...

Lets say you have 1 million pounds at the start of your retirement in 2022. Your portfolio drops to 300K on January 2nd 2023.  How much are you going to withdraw? 

Seriously though, that would still give 18k between us at 3%, and whilst it would be frugal, we'd not exactly be on the breadline. Perhaps downsize, assuming the house was still worth more than when it was built in 1730...

Ok, so you are using variable percentage withdrawal rather than Safe Withdrawal Rate.  In the latter case you’d be withdrawing 30k plus inflation. 

bostonerimus

Thrugelmir said:

bostonerimus said:

Deleted_User said:

bostonerimus said:

Ibrahim5 said:

You would have to have a different figure for people who use IFAs because the IFA takes out a large proportion of the investment returns for their fees.

No you wouldn't. SWR is only about what level of withdrawal a pot can sustain, not what you spend it on. Two retirees with identical pots that are identically managed will have identical SWR, but the one without an IFA will have a bit more money to spend on other things. So if you have investment fees, taxes, debt payments, holidays, food etc they are all just expenses you need to pay for with your SWR.

The comment is valid because people mean different things when they say “4% SWR”. Many use the number to apply after taxes and investment fees.  Regardless, fundamental problem is letter “S” in “SWR”. Mathematically there can be no “safe” and constant rate of withdrawal from a highly variable portfolio.

Yes it's a bit of a mine field because people make assumptions and don't take the time to understand the parameters of the modeling. "SWR" depends greatly on a whole set of assumptions, but it does not include anything about what you spend the money on be it taxes, investments fees or rent. If it was to include things like investment fees it would be even more of a moving target. And, being cynical, I can see how financial advisors like Bergen would want to exclude them as they might take up to half of your initial spending when your retire.

Investment returns are influenced by rates of corporate tax.  That's a minefield that's already being laid ahead. Though hasn't appeared to have registered yet with many investors given their expected growth rates in the years ahead. 

The might be the case, and its history will be included in the model as part of the investment return data sets. SWR models don't tell you anything about the future. Maybe the misunderstanding of SWR models is quite well expressed in the title of this thread. 4% SWR isn't foolish, but it can be applied foolishly, misunderstood and over interpreted.