Donate to Remove ads

Got a credit card? use our Credit Card & Finance Calculators

Thanks to barchid,Tortoise1000,NotSure,csearle,brightncheerful, for Donating to support the site

A thought experiment of sorts - retirement drawdown

Including Financial Independence and Retiring Early (FIRE)
Newroad
Lemon Slice
Posts: 262
Joined: November 23rd, 2019, 4:59 pm
Been thanked: 85 times

A thought experiment of sorts - retirement drawdown

#401723

Postby Newroad » April 4th, 2021, 9:16 pm

Hi All.

Just doing a bit of a thought experiment.

People often cite 4% drawdown during retirement as "safe", historically at least. I wondered instead whether assuming a number of years or retirement, e.g. 31 in this example, then recalibrate each year (i.e. 1/31st in the first year, 1/30th of the remaining in the second year etc) might work OK. Then I thought about combining the two, e.g. 2% plus 1/62nd in the first year, 2% plus 1/60th in the second year etc) - this also has the dual merits of ensuring you wouldn't completely run out of money if you live a very long time and your initial income is higher (when you might want to do more)!

For some test data, I took the FTSE 100 index since 1990 (i.e. I assumed whatever is left remains 100% in equities). I have plugged in constant yield of 4.7% (can't remember where I saw that number) as I don't have access to either FTSE 100 total return data (or standalone yield data) for that long a period. The initial capital is meaningful to me, but irrelevant for others.

I then calibrated the draw down percentage (to 4.98%) so that the total income received plus capital remaining was about the same for the two standalone options, i.e. I have assumed all income taken was spent.

The resultant data appears below. Interestingly, the combined option appears a little greater than the two individual ones. It also overtakes the annual income for "Fixed Percentage" around 1/3 of the way through (for the Year 2000) - £28,571 vs £28,174.



What I would really like, if anyone could point me to it or provide it, is the total return data and/or FTSE 100 yields for the same 30 year period - to check - it may or may not be material as to the conclusions which might be drawn?

Regards, Newroad

JohnB
Lemon Quarter
Posts: 1572
Joined: January 15th, 2017, 9:20 am
Has thanked: 187 times
Been thanked: 438 times

Re: A thought experiment of sorts - retirement drawdown

#401729

Postby JohnB » April 4th, 2021, 9:35 pm

https://www.swanlowpark.co.uk/ftseannual FTSE 100 and FTSE All-Share since 1986 includes TR

Newroad
Lemon Slice
Posts: 262
Joined: November 23rd, 2019, 4:59 pm
Been thanked: 85 times

Re: A thought experiment of sorts - retirement drawdown

#401738

Postby Newroad » April 4th, 2021, 10:07 pm

Thanks, JohnB.

Now I can model more accurate and extended figures :)

Also now able to pretend retirement at 65 and assuming (for the "year fraction" portion) a 35 year period, i.e. leaving the mortal coil shortly after the telegram from QE II.



Still decent performance, it seems, for the combined option?

Of course, it depends on start year etc, but this time, it would have taken only 6 years for the combined option to overtake the income of the percentage option.

Regards, Newroad

dealtn
Lemon Quarter
Posts: 3488
Joined: November 21st, 2016, 4:26 pm
Has thanked: 119 times
Been thanked: 1160 times

Re: A thought experiment of sorts - retirement drawdown

#401748

Postby dealtn » April 4th, 2021, 10:47 pm

How confident are you that a single experiment of the past will resemble the future in your model?

tjh290633
Lemon Half
Posts: 5809
Joined: November 4th, 2016, 11:20 am
Has thanked: 533 times
Been thanked: 2351 times

Re: A thought experiment of sorts - retirement drawdown

#401756

Postby tjh290633 » April 4th, 2021, 11:11 pm

Where is inflation in your calculations?

TJH

Newroad
Lemon Slice
Posts: 262
Joined: November 23rd, 2019, 4:59 pm
Been thanked: 85 times

Re: A thought experiment of sorts - retirement drawdown

#401765

Postby Newroad » April 5th, 2021, 12:14 am

Hi Dealtn & TJH.

To answer your questions ...

{Dealtn} Not at all.

{TJH} Only in that the starting capital assumed in the figures is based on a real world starting number divided by 2.5102 (to reflect modern pounds with 1986 pounds). However, that real world starting number is only meaningful to me - hence giving some useful context in the putative incomes per year.

{Both} The purpose of the "thought experiment" was to be indicative - to see how the two approaches (or a hybrid of them) would have borne out in a practical example. I'm very unlikely to hold 100% FTSE 100 when I retire - though that said, it wouldn't be the worst thing to do IMO.

Regards, Newroad

1nvest
Lemon Slice
Posts: 919
Joined: May 31st, 2019, 7:55 pm
Has thanked: 89 times
Been thanked: 224 times

Re: A thought experiment of sorts - retirement drawdown

#401767

Postby 1nvest » April 5th, 2021, 2:11 am

Here's some data for the UK from 1969 to 1999 inclusive, overall a pretty good total return/accumulation period that outpaced inflation by over 6% annualised over those years. BUT that endured earlier year sequence of returns risk that would have been harsh when also drawing a income.

In this case I've applied a 4% SWR i.e. 4% of the start data portfolio value drawn as income for the first year, and then in subsequent years that amount is uplifted by inflation (so maintains the same purchase power) as that amount drawn in those subsequent years.


By the 6th year the portfolio was in difficulties. Inflation had pushed consumer prices up by approaching 100%, whilst stock total returns weren't good after a couple of really bad years in 1973/74. 1975 somewhat came to the rescue with a near 150% nominal stock total return, but even with that the portfolio value was still down in inflation adjusted terms, around a third of its former value and where inflation had pushed consumer prices up to 2.5 times the 1969 start date level. Despite that the portfolio battled on and did very well to survive until 1989

It's important to factor in inflation. If consumer prices more than double then taking the same/similar nominal income is paramount to cutting your spending by half.

Yet another factor to consider is that under times of stress, when interest rates/inflation etc. tend to spike, so also do taxation/confiscation policies. 1968 for instance Roy Jenkins (then Labour) opted to tax retrospectively that for some meant a 130% tax rate. Yep! For each £1 end up owing £1.30 in tax. Around those years the Beatles were singing 'Taxman' ... 19 for you, 1 for me ... in reflection of the 95% tax rates they were paying. Many artists and others opted to flight the country due to such Labour policies. During such times, non interest/dividend paying assets are more highly valued as gains compound up internally rather than being paid out/taxed/reinvested (or spent) i.e. tax only falls due when the asset is sold and if policies are harsh at the time selling might be deferred until more 'normal service' resumes.

Note also how in nominal terms (nominal portfolio value) 1980 had 89% of the start date portfolio value available, which looks decent, however look across to the real (after inflation) portfolio value for that year and it was only 21.6% of the start date portfolio value. Whilst to maintain spending power the SWR value had risen for 0.04 at the start to 0.165. Looked pretty much doomed however a sequence of pretty great stock gains through the 1980's enabled it to hang in there for longer than perhaps expected.

A 3% SWR would have been much better, got through the bad times and subsequently recovered 'losses'. But again that assumes you got to keep all of gross gains/benefits and paid low fees/costs and that you'd selected the right index/stocks, and not something perhaps like the then popular FT30 index/stocks. Always remember that the financial sector is the worlds richest sectors and much of media is funded in order to direct towards indications of how 'easy' it is to win. Selective time periods, ignoring inflation, ignoring cost/taxes, selective 'Index' of 'average' returns etc. are all biased towards marketing their products/services that enables them to pay high wages and own expensive real estate/buildings.

Over some decades you might reasonably anticipate 0% or worse stock total (dividends reinvested) return rewards. Funds often hide costs such as a average 20% foreign dividend withholding tax being applied. Add on market makers spreads, brokers fees/platform costs, tax mans cut, fund managers fees etc. and historically that could collectively be north of 4%/year. Summed with drawing a income and its those bad decades that present great risk, typically more so if they are endured during earlier years of retirement.

Newroad
Lemon Slice
Posts: 262
Joined: November 23rd, 2019, 4:59 pm
Been thanked: 85 times

Re: A thought experiment of sorts - retirement drawdown

#401861

Postby Newroad » April 5th, 2021, 1:34 pm

Hi 1nvest.

As per my reply to TJH, inflation hasn't been (completely at least) ignored - I converted a relevant (to me) pot or two of money back into 1986 pounds then worked forward from there. I know how much I was being paid during each year of the period, so have a frame of reference for the income implied.

Going forwards, I don't think the management overheads are an issue - for example, it would be straightforward to get an ETF to track the FTSE 100 selling units annually (or whatever frequency suited). Yes taxation/confiscation (more generally, legislative) risk is real, but that's also true of SIPP and ISA havens etc.

Also, I would not be proposing to "uplift" based on inflation each year - that is the point of the 1/"remaining year" part of the methodology - it recalibrates. As part of going for this approach, you accept variability in income. This is less of an issue when you are likely to own your own home, have a decent car etc.

The above said, thank you for the data - I may take the inflation information and factor that in*. This would be for others moreso than me - I'm comfortable I understand adequately what the figures tell me.

Like I said at the start, it's more of a thought experiment for me - and based on the results, I may well take something like the "combined approach" when the time comes - though with different underlying investments quite likely.

Regards, Newroad

* just realised I can't - it only goes to 1999

Hariseldon58
Lemon Slice
Posts: 502
Joined: November 4th, 2016, 9:42 pm
Has thanked: 30 times
Been thanked: 280 times

Re: A thought experiment of sorts - retirement drawdown

#402085

Postby Hariseldon58 » April 6th, 2021, 4:52 pm

My experience since retiring early in 2007 is that I saw a substantial drop in the portfolio, with an inflation adjusted recovery by 2012, as of April 2021 the inflation adjusted capital is up by a factor just under 2.

The income withdrawal percentage has averaged around 3% ( within the bounds of 2% to 5%) the real issue is that whatever happened previously, we really don't know what will happen next.

I had an equity heavy portfolio in 2007 and it worked out despite major market falls, if my inflation related spending was constant, then my withdrawal rate as a % would have halved, but in practice some years one spends a lot more than others and if you can afford to travel more etc then one does.

I started using the calculations for McClungs Prime Harvesting for the tax year beginning April 2019 and this suggested spending 3.9% but this has risen to 5.06 % for the coming year, in practice this is probably £40k + more than I will spend.

The issue with any formula is how much do you trust the output ? In my case the suggested figures have been higher than I need and this results in reinvestment and hopefully the portfolio grows and the margin of safety increases.

The danger with a formula led answer to how much I can spend is if the suggested figure is very close or lower than what you need to spend, what then ?

Newroad
Lemon Slice
Posts: 262
Joined: November 23rd, 2019, 4:59 pm
Been thanked: 85 times

Re: A thought experiment of sorts - retirement drawdown

#402115

Postby Newroad » April 6th, 2021, 6:55 pm

Hi HariSeldon.

My thoughts on

"The danger with a formula led answer to how much I can spend is if the suggested figure is very close or lower than what you need to spend, what then?"

are as follows: You use your drawn income for sustenance and what's left for discretionary spend. If you have a one-off expense you can't accommodate in year from income*, you take it from capital. This is where my, say, "2% + 1/(remaining years*2)" of capital approach differs from the 4% raised by inflation approach. The former recalibrates annually on the fly, the later is relatively more likely to run out of money (as per 1nvest's example).

Regards, Newroad

* e.g. you need a new car or a new boiler or whatever

Hariseldon58
Lemon Slice
Posts: 502
Joined: November 4th, 2016, 9:42 pm
Has thanked: 30 times
Been thanked: 280 times

Re: A thought experiment of sorts - retirement drawdown

#402126

Postby Hariseldon58 » April 6th, 2021, 7:28 pm

@newroad

The McClung formula is not a simple % but adjusts to the prior years returns and takes account of inflation. The issue is in the down years....

I tended to think along the lines that you suggest, what is essential, then a comfortable lifestyle , then very comfortable. The formula results have been ahead of my needs and that makes life easy but it’s an extended down period that can cause problems and being able to cut spending if required plus some cash always helps!

Newroad
Lemon Slice
Posts: 262
Joined: November 23rd, 2019, 4:59 pm
Been thanked: 85 times

Re: A thought experiment of sorts - retirement drawdown

#402138

Postby Newroad » April 6th, 2021, 8:45 pm

Indeed, Hariseldon.

One day, I'll back-test my then ideas against (say) Prime Harvesting using UK data - with a few more years added by then, most likely - including a likely extensive period of low interest rates. It will be interesting to see the prognosis.

Like McClung, I (think I am at least) fine with variable income in retirement.

Regards, Newroad


Return to “Retirement Investing (inc FIRE)”

Who is online

Users browsing this forum: No registered users and 3 guests