Foolishness of the 4% rule

MK62

Deleted_User said:

  If this year you have 1000 potatoes and next year you may have 2000 potatoes or possibly 10 potatoes you can't "safely" withdraw 100 potatoes every year.

You do realise that withdrawing 100 potatoes would be a starting rate of 10%......

Stubod

..I don't see it as a "4% rule", more of a 4% "guide" as surely that is all it can ever be?

..and as such is good enough as you have to start somewhere when planning?

OldScientist

Terron said:

What real world retirees might do is irrelevant to whether the rule makes sense. The point of the research was to see is such a strategy could have coped with the worst the market has thrown at it in the past, and the answer was yes. Unless the future is worst that the past (including the great depression, WW2, and the 70s inflation) the strategy would cope. If people did adjust their spending that would just make it even safer.
The 4% (or 3.5% for the UK) rule has been shown to have worked for over 100 years. It is entirely reasonable and sensible to assume it is very likely to work for another 30, or that if things get so bad that it stops working you will have more serious things to worry about.

Certainly there are other options that are likely to be better, but that does not make the 4% rule nonsense or even wrong, just not optimal.

It is true that withdrawing 4% from a US portfolio only led to failure (i.e. running out of money) in about 5% of retirements (e.g. see https://cfiresim.com) with all of those failures occurring in retirements starting in the 1960s (assuming a 60/40 stock/bond portfolio with annual withdrawal and rebalancing and no fees) and the earliest failure happened just over 20 years after the start of retirement (i.e. around the life expectancy of someone retiring at 65). If you do not expect to live that long, do not rely on the income from the portfolio for essential spending (i.e. roof, food, etc.), or are wiling to take the 1:20 risk of failure, then this approach can make sense. As others have said it it also an incredibly useful tool for retirement planning.

I have no idea whether the future in this context will be worse than the past (it will be different!), but the current worst case is only the worst case we have seen to date (i.e. prior to the 1960s, if the modelling had been done earlier, the 'rule' would have been 4.5% instead). One problem with the 4% 'rule' is that as stated it makes no concession to the possibility of failure. That is easy to add (e.g. limiting the withdrawal to, say, 10% of the current portfolio value) thereby preventing failure in the sense of running out of money before death (although income could get quite small).

MK62

At the end of the day, it all depends on what "exactly" is meant by the 4% rule.

Taken as an absolute, for a UK retiree with no other form of income over the rest of his/her life, then the data suggests that it would not be "safe" for a 30yr retirement.....I say suggests, as the real 30yr data for retiring in the year 2000 is not in yet (and historically, it doesn't get much worse than retiring in 2000 with a stock market backed pension), but it doesn't look good....even 3.5% looks a bit touch and go......

Then again, this data is tracking UK RPI and the total returns of the FTSE All Share Index (minus 0.5% pa in costs)......the data will be different if this fictional retiree was invested differently (and in reality I suspect a large majority would/will have been).

However, that said, very few UK retirees will have to rely solely on a DC pension.......most will have a state pension if nothing else (though the level will be different for many). Throw that into the data in as well, and the picture changes.......4% now looks safe....even 4.5% does.

One major caveat is, of course, that relying on the year 2000 being the worst case could lead us into a black swan fallacy trap (ie just because worse hasn't been seen before, does not preclude it from being seen in the future).......what has been safe in the past, may not be so in the future.

However, a plan has to start somewhere, and in any plan, certain assumptions have to be made at the outset......we just have to accept that those assumptions may well turn out to be wrong, so the plan has to be adjusted along the way......I think few would find this surprising though......in the end, what would be foolish is just rigidly sticking to any withdrawal plan if looks like it won't meet it's target.

So, all things considered, as a starting point for a UK retiree with a 30 yr timeframe, 3.5-4% seems reasonable to me......just be prepared to alter course if, when the real data comes in, it starts looking like the plan is going awry.....

I accept Mordko's assertion that starting at 4% and then just blindly taking an index linked raise on that each year, no matter what, might well turn out to be foolish......but in reality, I don't think there are many doing that...even if that's how some start out.

Thrugelmir

MK62 said:

I accept Mordko's assertion that starting at 4% and then just blindly taking an index linked raise on that each year, no matter what, might well turn out to be foolish......but in reality, I don't think there are many doing that...even if that's how some start out.

That's a long established fact. Nothing new in that. 

OldScientist

Thrugelmir said:

MK62 said:

I accept Mordko's assertion that starting at 4% and then just blindly taking an index linked raise on that each year, no matter what, might well turn out to be foolish......but in reality, I don't think there are many doing that...even if that's how some start out.

That's a long established fact. Nothing new in that. 

Judging by my own journey along these lines there are several levels of sophistication in understanding that fact

1) I'll just take money out when the market is up - oh dear, that means some years you don't get anything
2) I'll take out money at the mean rate of stock market returns - read about sequence of returns risk and hence
3) Trinity study and '4% rule' - hurrah! Oh dear, it can still fail and I want to leave a legacy (in fact, personally, the legacy component is more important than the portfolio income component since my essential spend is currently covered by DB income - in my view, MK62's point about state pension and other forms of income is spot on*)
4) Arrive at variable withdrawal methods (a read of McClung and Bogleheads)

* However, the exact amounts are critical. For example, assuming a state pension (approx 10k), a DC pot of 50k (probably close enough to current UK median - it depends on which source you believe as a google search will testify, but 40k-60k seems typical), and essential expenses of 13k (single person, https://www.which.co.uk/money/pensions-and-retirement/starting-to-plan-your-retirement/how-much-will-you-need-to-retire-atu0z9k0lw3p), then the withdrawal amount from the portfolio is about 6% (which, for US data, is likely to exhaust the portfolio as soon as 15 years after retirement with a failure rate of 50%). Under these circumstances, being frugal (and who of us moneysavers is not?) would be essential since knocking 1k off expenditure (or saving harder, retiring later, working part-time, etc. to increase income) would at least reduce the historical failure rate by a factor of 10.

Malthusian

MK62 said:

So, all things considered, as a starting point for a UK retiree with a 30 yr timeframe, 3.5-4% seems reasonable to me......just be prepared to alter course if, when the real data comes in, it starts looking like the plan is going awry.....

I accept Mordko's assertion that starting at 4% and then just blindly taking an index linked raise on that each year, no matter what, might well turn out to be foolish......but in reality, I don't think there are many doing that...even if that's how some start out.

Nobody in the history of UK drawdown pensions has ever withdrawn 4% (or 3.5% or a similar amount) of the starting amount in the first year and then increased their withdrawals by the rate of inflation, each year every year, without paying any attention to what the market is doing, until death. Or for more than, say, 10 years (drawdown is only 30 years old so "till death" is a bit of an unfair challenge).

If you disagree then name two people who did and I'll happily update my knowledge.

People have started with 4% of the starting value and then left it there, or increased it based on "gut feel", or even applied elaborate rules like Guyton-Klinger, but that isn't the same thing.

MK62

Malthusian said:

MK62 said:

So, all things considered, as a starting point for a UK retiree with a 30 yr timeframe, 3.5-4% seems reasonable to me......just be prepared to alter course if, when the real data comes in, it starts looking like the plan is going awry.....

I accept Mordko's assertion that starting at 4% and then just blindly taking an index linked raise on that each year, no matter what, might well turn out to be foolish......but in reality, I don't think there are many doing that...even if that's how some start out.

Nobody in the history of UK drawdown pensions has ever withdrawn 4% (or 3.5% or a similar amount) of the starting amount in the first year and then increased their withdrawals by the rate of inflation, each year every year, without paying any attention to what the market is doing, until death. Or for more than, say, 10 years (drawdown is only 30 years old so "till death" is a bit of an unfair challenge).

If you disagree then name two people who did and I'll happily update my knowledge.

People have started with 4% of the starting value and then left it there, or increased it based on "gut feel", or even applied elaborate rules like Guyton-Klinger, but that isn't the same thing.

I'm a bit confused by this tbh....as that's pretty much what I said......"I don't think there are many doing that"

CloesUnc

It seems to me that all of this talk of 4% foolishness is squarely aimed at those who wish to draw from their pensions long-term, say 30 to 40 years. There are very many variants of these rules and even the variable withdrawal rules; someone mentioned the book "How to Live off Your Money" earlier, and much of that is about the various rules. I think 4% is a good starting point for those wishing to get an idea about how much pension that someone can expect to get from a DC pot. Otherwise they are at the mercy of their work, or other pension provider, and they could be provided a lot less income per month than otherwise "possible".

These rules also apply to ISAs as well as pensions of course. I need a strategy to draw from my ISA to cover part of my income from age 55 to when SP kicks in at 67. So for 12 years I will use Guyten-Klinger with an initial 5.5%, giving me more than the 4% rule.

ukdw

Malthusian said:

Nobody in the history of UK drawdown pensions has ever withdrawn 4% (or 3.5% or a similar amount) of the starting amount in the first year and then increased their withdrawals by the rate of inflation, each year every year, without paying any attention to what the market is doing, until death. Or for more than, say, 10 years (drawdown is only 30 years old so "till death" is a bit of an unfair challenge).

If you disagree then name two people who did and I'll happily update my knowledge.

People have started with 4% of the starting value and then left it there, or increased it based on "gut feel", or even applied elaborate rules like Guyton-Klinger, but that isn't the same thing.

Probably also the case in the US too.