Foolishness of the 4% rule

bostonerimus

OldScientist said:

Deleted_User said:

If (and that is a big if) you believe that the next 30 years of returns are going to be grim, then even today's relatively low annuity rates would be rather attractive.

Beliefs are for the church.  The future is an event tree rather than a certainty.   We don’t know the future so we try to  design our portfolio to withstand most scenarios.  In that respect annuities don’t really compete with equity returns; they are an option for fixed income.  If one has lots of DB income then there is no need.  Delaying state pension is always the number 1 option.  But annuities are better than todays bonds and this difference is more pronounced than in the past. 

With Pfau's paper, there are a number of questions

1) Are average US (and probably global) stock and bond returns going to be 2.5% lower in the next century than the previous one?
2)  Are US stock and bonds returns going to be 2.5% lower than the current average in the next 10-40 years (i.e. for recent or soon to be retirees)?

Considerations:
My use of the word 'believe' is founded on two further questions, is it possible and is it likely?

1) This is definitely possible and may even be likely (there are currently a number of problems to be solved, e.g. transitioning fuel sources, environment, ageing population,  etc.) - if so, it reduces the US MSWR from about 3.7 to 2.6% - so plan accordingly (and current US annuity prices look really attractive) .

2) This is definitely possibly and is looking highly likely (at least for the next decade or so given the high correlation of future interest rates with current ones presented Pfau's paper). Does this mean that the outcome will be worse than the current historical worst case? Looking at 1965-1995 (around the worst time for retirement), the S&P500 returned 4.8% real (i.e. about 3.8% below the average for US stocks), while US treasuries returned 2.6% real (i.e. average). Given this, conditions in the immediate future may not exceed the current worst case. Just for fun, I reduced the historical bond returns by 2.5% but left stock returns the same and got an MSWR~3.2%. Again, US annuity rates (with 3% COLA, 65 year old male the rate ~4.1%) look fairly attractive.

So, for the current (US) retiree is the SWR 2.6%, 3.2%, 3.7% or 4.0% (subtract 0.5-0.7% for UK given our higher inflation during the worst case retirements)? For me, illustrates why an unwavering index-linked withdrawal for essential spending is unwise.

I think we are agreeing here!

I think we agree that variable withdrawal methods are more natural that the 4% straight jacket and that when done correctly they can provide greater success with minimal changes in lifestyle. 

As always we must not fall into the trap of considering Pfau et al's paper as a prediction of the future. It just takes recent bond rates and implied stock returns and uses them to see what happens. For any sensible retirement plan this should have already been done by simply plugging in an range of portfolio returns. This stress testing will determine the range of possible futures that the plan will accommodate.

[Deleted User]

DT2001 said:

MK62 said:

Deleted_User said:

The UK's Office of National Statistics disagrees......they say that the chances of a 65yo British male reaching 100 are 2.9%.   (that is, only 2.9% of 65yo British males alive today, will make it to 100)

Its completely irrelevant.  Sick 65 year olds have different life expectancies and annuity rates.  They wouldn’t even need a longivity insurance.  

 I provided expectancies for healthy Americans from actuaries, the stats that matter for annuities:  https://www.longevityillustrator.org/  Its a 10% chance of getting passed 98. Average life expectancy is 3 years longer for Brits, so will be higher percentage although I don’t have numbers for healthy Brits.  And ones chances could be even higher if he is well off and understands money, which a lot of people on this board do.  

Your chart gives no detail about the % of 65yo who are considered to be in excellent health......for all we know it could be 5% excellent, 90% average and 5% poor.........not really representative if it only applies to a small percentage of 65 year olds.......even if it's 33%/33%/33%, it'd still mean it only applied to one third of 65yo Americans.

Also, while 2 years may not make much difference at 55, at 98 it does.......in the UK, a 65yo female has a 10% chance of making 98yo......it's less than half that for 100yo.

That said, whether it's 5% or 10% is irrelevant really, as neither could be described as an excellent chance in my book.........if I told you a withdrawal plan had 10% chance of making it to 100, would you describe that as an excellent chance?

If you go to the link you get a personalised quote which is more relevant to your own planning. There is the caveat that the more factors that are pertinent to you the better the ‘forecast’ will be.
It is a step up from a generalised life expectancy table.

If you compile the factors that appear relevant to life expectancy (e.g. smoker, location, wealth, family history, current health etc etc) you can get a better estimate of your life expectancy which you then use to help formulate your retirement plan.
It maybe then that annuities are an important part of your plan or nigh on irrelevant. Whether the purchase is at retirement, 10 years after or if you survive to a certain point.

We used an annuity for my mother when she was 70 following my father’s death for the exact reason that Mordko suggested - certainty of income - it was funded by about 20% of her available capital at that time. The rest of her capital was used to cover one offs until she downsized (not in my original plan but equity release was). It was a punt as she had indifferent health (still going 20 years later, just) but it served it’s purpose.

As Mordko says annuities CAN be part of your plan even in today’s scenario. Is your strategy to have enough income until you die or to maximise income with the inherent dangers that entails.

I would reiterate my believe that flexibility and diversity are key in income and expenditure. Having been self employed for 25+ years I think helps the mindset needed.

The “mindset” is an important point. 

In 2000 one financial advisor in Canada decided to track a year 2000 retiree’s fortunes using the 4% withdrawal rule and diversified investment vehicles available at the time. It is still being tracked.  The 100% stock option is doing the worst of all portfolio. 100% inflation linked bonds have done the best. Balanced portfolios are doing ok. 

The 100% stock retiree is in danger of running out of money before 2030, but he might manage. Touch and go.  Looked even worse at times. 

I wouldn’t want to be in the “mindset” of a real world retiree who has done it.  Can’t be nice.   What your mother did seems smart. 

[Deleted User]

2)  Are US stock and bonds returns going to be 2.5% lower than the current average in the next 10-40 years (i.e. for recent or soon to be retirees)?

Stock returns are impossible to predict but the current yield is an excellent predictor of what the bonds will return.  The answer is: nothing. 

bostonerimus

Deleted_User said:

Interesting article. The whole talk will have been interesting but it’s clear where he is going with this.  He does not like simplistic solutions for a complex problem.  

Milevsky noted at the FPAC webinar how “optimal decumulation” is a time-consuming process that requires quant skills. He equated decumulation expertise to tax and legal expertise — areas that may be out of the realm of a financial advisor’s knowledge base and that require special advising.

https://www.investmentexecutive.com/news/industry-news/4-rule-is-no-longer-an-ideal-retirement-strategy-milevsky-says/

I'm disappointed that the goal still seems to be "how much can you withdraw" and little attention is paid to budgeting and not trying to simply spend as much as you can. The mathematics of withdrawal is not hard, what makes the planning hard are the numerous variables and scenarios. So, as in many maths problems, it's useful to go to the extremes and simplify as this often makes the basic principles clear and shows you a general solution space.

bostonerimus

DT2001 said:

MK62 said:

Deleted_User said:

The UK's Office of National Statistics disagrees......they say that the chances of a 65yo British male reaching 100 are 2.9%.   (that is, only 2.9% of 65yo British males alive today, will make it to 100)

Its completely irrelevant.  Sick 65 year olds have different life expectancies and annuity rates.  They wouldn’t even need a longivity insurance.  

 I provided expectancies for healthy Americans from actuaries, the stats that matter for annuities:  https://www.longevityillustrator.org/  Its a 10% chance of getting passed 98. Average life expectancy is 3 years longer for Brits, so will be higher percentage although I don’t have numbers for healthy Brits.  And ones chances could be even higher if he is well off and understands money, which a lot of people on this board do.  

Your chart gives no detail about the % of 65yo who are considered to be in excellent health......for all we know it could be 5% excellent, 90% average and 5% poor.........not really representative if it only applies to a small percentage of 65 year olds.......even if it's 33%/33%/33%, it'd still mean it only applied to one third of 65yo Americans.

Also, while 2 years may not make much difference at 55, at 98 it does.......in the UK, a 65yo female has a 10% chance of making 98yo......it's less than half that for 100yo.

That said, whether it's 5% or 10% is irrelevant really, as neither could be described as an excellent chance in my book.........if I told you a withdrawal plan had 10% chance of making it to 100, would you describe that as an excellent chance?

If you go to the link you get a personalised quote which is more relevant to your own planning. There is the caveat that the more factors that are pertinent to you the better the ‘forecast’ will be.
It is a step up from a generalised life expectancy table.

If you compile the factors that appear relevant to life expectancy (e.g. smoker, location, wealth, family history, current health etc etc) you can get a better estimate of your life expectancy which you then use to help formulate your retirement plan.
It maybe then that annuities are an important part of your plan or nigh on irrelevant. Whether the purchase is at retirement, 10 years after or if you survive to a certain point.

We used an annuity for my mother when she was 70 following my father’s death for the exact reason that Mordko suggested - certainty of income - it was funded by about 20% of her available capital at that time. The rest of her capital was used to cover one offs until she downsized (not in my original plan but equity release was). It was a punt as she had indifferent health (still going 20 years later, just) but it served it’s purpose.

As Mordko says annuities CAN be part of your plan even in today’s scenario. Is your strategy to have enough income until you die or to maximise income with the inherent dangers that entails.

I would reiterate my believe that flexibility and diversity are key in income and expenditure. Having been self employed for 25+ years I think helps the mindset needed.

Yes annuities can be a useful tool. Partial annuitization at the right time can provide an income floor and longevity insurance that reduces stress. I just can't see the utility right now of a 60 or 65 year old in the UK using one with payouts so low. The argument is better for them in the US where payouts are higher (6% for a 65 year old male) and 5 year fixed term saving accounts rates are just 1% so holding cash isn't quite as good as in the UK.

[Deleted User]

bostonerimus said:

Deleted_User said:

Interesting article. The whole talk will have been interesting but it’s clear where he is going with this.  He does not like simplistic solutions for a complex problem.  

Milevsky noted at the FPAC webinar how “optimal decumulation” is a time-consuming process that requires quant skills. He equated decumulation expertise to tax and legal expertise — areas that may be out of the realm of a financial advisor’s knowledge base and that require special advising.

https://www.investmentexecutive.com/news/industry-news/4-rule-is-no-longer-an-ideal-retirement-strategy-milevsky-says/

I'm disappointed that the goal still seems to be "how much can you withdraw" and little attention is paid to budgeting and not trying to simply spend as much as you can. The mathematics of withdrawal is not hard, what makes the planning hard are the numerous variables and scenarios. So, as in many maths problems, it's useful to go to the extremes and simplify as this often makes the basic principles clear and shows you a general solution space.

There are different problems and withdrawal is a big enough issue to warrant its own talk. 

Mathematics of withdrawal is too hard for many and, like Milewski says, for a typical advisor.  People routinely misuse stats because few understand statistics. For example a specialist can see Bergen’s problem of overlapping data but many appear to think his analysis has some validity.  Gets more complex still when applied to personal circumstances. 

And the question isn’t “how little am I able to withdraw”, no. 

MK62

Deleted_User said:

DT2001 said:

MK62 said:

Deleted_User said:

The UK's Office of National Statistics disagrees......they say that the chances of a 65yo British male reaching 100 are 2.9%.   (that is, only 2.9% of 65yo British males alive today, will make it to 100)

Its completely irrelevant.  Sick 65 year olds have different life expectancies and annuity rates.  They wouldn’t even need a longivity insurance.  

 I provided expectancies for healthy Americans from actuaries, the stats that matter for annuities:  https://www.longevityillustrator.org/  Its a 10% chance of getting passed 98. Average life expectancy is 3 years longer for Brits, so will be higher percentage although I don’t have numbers for healthy Brits.  And ones chances could be even higher if he is well off and understands money, which a lot of people on this board do.  

Your chart gives no detail about the % of 65yo who are considered to be in excellent health......for all we know it could be 5% excellent, 90% average and 5% poor.........not really representative if it only applies to a small percentage of 65 year olds.......even if it's 33%/33%/33%, it'd still mean it only applied to one third of 65yo Americans.

Also, while 2 years may not make much difference at 55, at 98 it does.......in the UK, a 65yo female has a 10% chance of making 98yo......it's less than half that for 100yo.

That said, whether it's 5% or 10% is irrelevant really, as neither could be described as an excellent chance in my book.........if I told you a withdrawal plan had 10% chance of making it to 100, would you describe that as an excellent chance?

If you go to the link you get a personalised quote which is more relevant to your own planning. There is the caveat that the more factors that are pertinent to you the better the ‘forecast’ will be.
It is a step up from a generalised life expectancy table.

If you compile the factors that appear relevant to life expectancy (e.g. smoker, location, wealth, family history, current health etc etc) you can get a better estimate of your life expectancy which you then use to help formulate your retirement plan.
It maybe then that annuities are an important part of your plan or nigh on irrelevant. Whether the purchase is at retirement, 10 years after or if you survive to a certain point.

We used an annuity for my mother when she was 70 following my father’s death for the exact reason that Mordko suggested - certainty of income - it was funded by about 20% of her available capital at that time. The rest of her capital was used to cover one offs until she downsized (not in my original plan but equity release was). It was a punt as she had indifferent health (still going 20 years later, just) but it served it’s purpose.

As Mordko says annuities CAN be part of your plan even in today’s scenario. Is your strategy to have enough income until you die or to maximise income with the inherent dangers that entails.

I would reiterate my believe that flexibility and diversity are key in income and expenditure. Having been self employed for 25+ years I think helps the mindset needed.

In 2000 one financial advisor in Canada decided to track a year 2000 retiree’s fortunes using the 4% withdrawal rule and diversified investment vehicles available at the time. It is still being tracked.  The 100% stock option is doing the worst of all portfolio. 100% inflation linked bonds have done the best. Balanced portfolios are doing ok. 

The 100% stock retiree is in danger of running out of money before 2030, but he might manage. Touch and go.  Looked even worse at times.

For most UK retirees though, you need to factor in the state pension - the effect of doing that will depend on when the SP becomes payable and the size ratio between the SP and the 4% rule's withdrawal, assuming you reduce that withdrawal by the equivalent of the SP when it does become payable.

While strictly speaking, it would then not be the vanilla 4% rule, it's probably more representative for most UK retirees. Once you do that even a 100% equity portfolio becomes a lot less touch and go.....

lisyloo

Deleted_User said:

Is it?...  .......out of interest, how much does one of those cost, today?

A 55 year old female in the US would get an 8.05% payout rate at 65 based on 2012 mortality data. A male would get a better payout rate. You would need to get a quote for something accurate. 

Don’t we have gender neutral Annuity rates in the uk? Which is a benefit for women?

Terron

OldScientist said:

[snip]

So, for the current (US) retiree is the SWR 2.6%, 3.2%, 3.7% or 4.0% (subtract 0.5-0.7% for UK given our higher inflation during the worst case retirements)? For me, illustrates why an unwavering index-linked withdrawal for essential spending is unwise.

I think we are agreeing here!

When talking about essential spending it can't be reduced, because it is essential, and an increase beyond inflation is likely not essential.  So the only option that makes sense is for unwavering index-linked withdrawal. 

Ideally that would come from an index-linked DB pension, but we can't all work in jobs that give them, and the SP is not enough for the majority.
An index-linked annuity would be safe, but (currently at least) is more expensive (for most people) than using drawdown per the SWR for your country. So the 4% rule is wise for someone in the US who only just has enough to meet their essentials in retirement (~3.5% for the UK).

Of course it is wise to save more than the minimum, but knowing what the minimum is is useful.

BTW I believe the UK's SWR is lower because WW2 was even worse for the British economy than the 1970s, whereas the war was good for the US economy. The worst case for Britain is in the mid-30s not the late 60s as it is in the US.

MK62

DT2001 said:

We used an annuity for my mother when she was 70 following my father’s death for the exact reason that Mordko suggested - certainty of income - it was funded by about 20% of her available capital at that time. The rest of her capital was used to cover one offs until she downsized (not in my original plan but equity release was). It was a punt as she had indifferent health (still going 20 years later, just) but it served it’s purpose.

As Mordko says annuities CAN be part of your plan even in today’s scenario. Is your strategy to have enough income until you die or to maximise income with the inherent dangers that entails.

I would reiterate my believe that flexibility and diversity are key in income and expenditure. Having been self employed for 25+ years I think helps the mindset needed.

Fair enough, but what if that same payout had cost 40-50% of her available capital.......annuity rates were a lot higher 20 years ago than they are today.

Not arguing with the idea of an annuity, just the current cost.

Is your strategy to have enough income until you die or to maximise income with the inherent dangers that entails.

Like most people I suspect, it's a bit of both (plus other considerations like partner provision, legacy etc). The problem is that imho, currently, annuities fail to accomplish the former.  Just to replicate the state pension would cost around £460k.

I would reiterate my believe that flexibility and diversity are key in income and expenditure

Me too......but there's not really much flexibility with an annuity.