Pension withdrawal tool

Albermarle

Toucan13 said:

This is in USD so will be using US data too (and obviously doesn't take into account SP) but does run Monte Carlo simulations and allows you to adjust the stock / bond / cash allocation (also gives you a probability that your savings will last for the number of years you choose) .  

Vanguard - Retirement Nest Egg calculator

I tried various scenarios of mix of equity + bonds + cash to see what the % would be that the money would last 30 years.
It gave almost exactly the same answer - 93% +/- 1% for a very wide range of asset allocations. You have to go below 30% equities or above 90% to even to move it down a couple of per cent , The only dramatic drop off is if you go very high in cash.
If it is anywhere near accurate it makes you wonder if all this mental energy used debating asset allocations , rebalancing etc is a waste of time . According to this as long as you do not go to the extremes then you get the same result anyway.

BritishInvestor

Toucan13 said:

This is in USD so will be using US data too (and obviously doesn't take into account SP) but does run Monte Carlo simulations and allows you to adjust the stock / bond / cash allocation (also gives you a probability that your savings will last for the number of years you choose) .  

Vanguard - Retirement Nest Egg calculator

So apart from using US data (returns and inflation?), using Monte Carlo (which has its own quirks) and not allowing you to adjust spending, it's a worthwhile exercise? 

Toucan13

BritishInvestor said:

Toucan13 said:

This is in USD so will be using US data too (and obviously doesn't take into account SP) but does run Monte Carlo simulations and allows you to adjust the stock / bond / cash allocation (also gives you a probability that your savings will last for the number of years you choose) .  

Vanguard - Retirement Nest Egg calculator

So apart from using US data (returns and inflation?), using Monte Carlo (which has its own quirks) and not allowing you to adjust spending, it's a worthwhile exercise? 

Certainly has it's flaws which I think I highlighted.  But surely all of this retirement planning malarkey is a degree of guesswork (barring access to a reliable crystal ball?). I'm personally leaning towards using a variety of different calculators to give myself a range of possible outcomes - all of which may prove to be wildly inaccurate in retrospect.  But therein is the problem with DC pensions.....

cfw1994

Brenster said:

Hi I have prepared my own on excel, it takes into account starting pot value (target), draw down age, assumed % return for each year (assumed at 4%), assumed inflation impact per year (assumed 2%), draw down amount per year, and i find it gives quite a useful insight into the length of time a pot will last and the massive impact a shift in drawdown amount, age you begin drawdown, % return, inflation impact can have cumuilitively.
It all has to be considered in todays money however, and excludes any tax which will come off your gross draw down numbers, to summarise,
If i reached a pot value of £550k at aged 60 (based on me being 60 now at todays values), i dont take the lump sum, and continue to earn 4% return, with 2% inflation impact, and i draw down £30k per year, until assumed aged 69 when my state pension retirement will kick in, at which point i reduce to £20k drawdown (remember my inflation 2% dealt with seperately), with the £10k state pension, it should still leave a pot of £150k + by the time i reach 80 (hopefully)
Note: There are lots of assumptions, and rounded numbers in these but i think it probably gives you the tools and answers you are looking for.  And can be tailored to suit, obviously the closer you are to retirement the more accurate / realistic it will be.  I would be happy to share if you wanted to message me.

Sounds similar to mine, also happy to share.
It gives the chance to model a poor sequence of returns (drop a -20% in year 2 and 8, for a random example!), but you can also perhaps model a reasonable 'average' annual return (3-4%) throughout, inflation at 2%...if you can figure on 1-2yrs in cash reserves, I believe that ought to work through a downturn.
It also allows you to look in 'real' terms, rather than today's numbers.   Useful, in some ways - no good accounting for todays pension value (around £9k for maximum) if you don't draw for another 10-15+ years...unless you want to be the richest in the graveyard!

BritishInvestor

Albermarle said:

Toucan13 said:

This is in USD so will be using US data too (and obviously doesn't take into account SP) but does run Monte Carlo simulations and allows you to adjust the stock / bond / cash allocation (also gives you a probability that your savings will last for the number of years you choose) .  

Vanguard - Retirement Nest Egg calculator

I tried various scenarios of mix of equity + bonds + cash to see what the % would be that the money would last 30 years.
It gave almost exactly the same answer - 93% +/- 1% for a very wide range of asset allocations. You have to go below 30% equities or above 90% to even to move it down a couple of per cent , The only dramatic drop off is if you go very high in cash.
If it is anywhere near accurate it makes you wonder if all this mental energy used debating asset allocations , rebalancing etc is a waste of time . According to this as long as you do not go to the extremes then you get the same result anyway.

FWIW a shift in asset allocation from 50% to 60% equities gave an increase in success rate from 30 to 44%, or looking at in another way, an increase in SWR from 2.48% to 2.67% for a case study I was looking at, so it does make a different 

BritishInvestor

cfw1994 said:

Brenster said:

Hi I have prepared my own on excel, it takes into account starting pot value (target), draw down age, assumed % return for each year (assumed at 4%), assumed inflation impact per year (assumed 2%), draw down amount per year, and i find it gives quite a useful insight into the length of time a pot will last and the massive impact a shift in drawdown amount, age you begin drawdown, % return, inflation impact can have cumuilitively.
It all has to be considered in todays money however, and excludes any tax which will come off your gross draw down numbers, to summarise,
If i reached a pot value of £550k at aged 60 (based on me being 60 now at todays values), i dont take the lump sum, and continue to earn 4% return, with 2% inflation impact, and i draw down £30k per year, until assumed aged 69 when my state pension retirement will kick in, at which point i reduce to £20k drawdown (remember my inflation 2% dealt with seperately), with the £10k state pension, it should still leave a pot of £150k + by the time i reach 80 (hopefully)
Note: There are lots of assumptions, and rounded numbers in these but i think it probably gives you the tools and answers you are looking for.  And can be tailored to suit, obviously the closer you are to retirement the more accurate / realistic it will be.  I would be happy to share if you wanted to message me.

Sounds similar to mine, also happy to share.
It gives the chance to model a poor sequence of returns (drop a -20% in year 2 and 8, for a random example!), but you can also perhaps model a reasonable 'average' annual return (3-4%) throughout, inflation at 2%...if you can figure on 1-2yrs in cash reserves, I believe that ought to work through a downturn.
It also allows you to look in 'real' terms, rather than today's numbers.   Useful, in some ways - no good accounting for todays pension value (around £9k for maximum) if you don't draw for another 10-15+ years...unless you want to be the richest in the graveyard!

"if you can figure on 1-2yrs in cash reserves, I believe that ought to work through a downturn."
The downturns that are going to hurt/impact the plan are the multi-year variety, potentially with spiky inflation, so the 1-2 year cash reserve isn't going to be of much use, unfortunately.

gm0

There is a fair amount of analysis out there on cash reserves and how long the market takes to regain prior highs (inflation adjusted) which can regularly be 5-7+ years in prior cycles.  So 1-2 years income cash will "help" but not solve - you sleep at night as you have a prop to short term spending and can avoid selling at what you think is the bottom or well down the down ramp.  You likely *should* rebalance into equities but this requires bravery (and bonds/cash/gold) beyond sitting on your hands alone.

My own view (and plan) is that it helps to have some cash as this gives you time - several years (with some spending pruning) to react to events and to wait and see more of the sequence without ravaging your equities holding. That feels a lot better even if it doesn't solve the "selling low" problem entirely.  So it does also help with the survivability of the full drawdown but this is non trivial to model out for the general case/all cohorts.

BritishInvestor

Toucan13 said:

BritishInvestor said:

Toucan13 said:

This is in USD so will be using US data too (and obviously doesn't take into account SP) but does run Monte Carlo simulations and allows you to adjust the stock / bond / cash allocation (also gives you a probability that your savings will last for the number of years you choose) .  

Vanguard - Retirement Nest Egg calculator

So apart from using US data (returns and inflation?), using Monte Carlo (which has its own quirks) and not allowing you to adjust spending, it's a worthwhile exercise? 

Certainly has it's flaws which I think I highlighted.  But surely all of this retirement planning malarkey is a degree of guesswork (barring access to a reliable crystal ball?). I'm personally leaning towards using a variety of different calculators to give myself a range of possible outcomes - all of which may prove to be wildly inaccurate in retrospect.  But therein is the problem with DC pensions.....

If you undertake a robust retirement planning exercise
1. Expenditure forecasting
2. Building a financial plan
3. Creating an investment engine
4. Designing a withdrawal strategy
and as importantly, revisiting it periodically to ensure you are on track, I think you will put yourself in a decent position. I just don't really see how that tool helps.

BritishInvestor

gm0 said:

There is a fair amount of analysis out there on cash reserves and how long the market takes to regain prior highs (inflation adjusted) which can regularly be 5-7+ years in prior cycles.  So 1-2 years income cash will "help" but not solve - you sleep at night as you have a prop to short term spending and can avoid selling at what you think is the bottom or well down the down ramp.  You likely *should* rebalance into equities but this requires bravery (and bonds/cash/gold) beyond sitting on your hands alone.

My own view (and plan) is that it helps to have some cash as this gives you time - several years (with some spending pruning) to react to events and to wait and see more of the sequence without ravaging your equities holding. That feels a lot better even if it doesn't solve the "selling low" problem entirely.  So it does also help with the survivability of the full drawdown but this is non trivial to model out for the general case/all cohorts.

If you take a simple case of 
£100k starting pot, split into 60% equities, 40% bonds, with annual rebalancing
Every year £3k is taken out at the same time as rebalancing
If an equity crash happens in year 1 (30% fall), the withdrawal money is likely to come from the bonds, and you will most likely be buying equities as part of the rebalancing. If equities carry on falling the same may be true in year 2.
If both equities and bonds fall that may well imply we are in serious trouble (e.g. WW1), and this might well take longer than 1-2 years to play out.
So I wonder in what scenario the 1-2 year cash bucket really works, other than acting as a comfort blanket (which I can understand)

gm0

Yes - this is why the FIRE crowd (ERN) etc. and others are sceptical in their analysis of buffers. Viewed across the 40-50+ year term and all cohorts it's impacts are mild to marginal (albeit this is based on simplifying assumptions about behaviour and long term returns).  I recall that particular blog concluded mildly helpful - but questioned the complexity/lost return potential (I haven't gone back to check the reference and quote directly from the epic safe WR series).  McClung looks at formal buckets and buffering ideas in his initial drawdown strategy survey and discovers that the details matter - a lot as to whether it slightly over performs or more commonly underperforms annual rebalancing.  Most simple rules or "rational" combinations of nice sounding rules for buckets do in fact fail the general case test.  Whether backtesting mild outperformance (by his ultimately preferred) options is actually a meaningful gain in WR/safety is a conversation for another thread.

For the individual retiree - on a single start date -  who has a choice to suspend (or reduce) drawdown (or not) upon a hard equity crash in the first decade or two of four - then the difference between having the two years income available (which can then be stretched by varying consumption down towards essential if you choose) is the difference between having money to pay the bills and not or a forced "at a perceived bottom" growth asset firesale which may persist for several years unknown at that moment.  There is no rule that bonds must be up when equities are down in all possible corrections.  Often yes.  Always?  Brave decision minister.  

Sure you may have an established credit line or be able to get credit - or alternatively you may not be able to get credit depending upon economic and credit conditions at the time.  Your asset/income potential position outside property has just soured by 60-80%

Once the equities are oversold repeatedly to maintain income you lock in your path regardless of how sunny the returns become in later years 20-40.  All those happy retirees with plump pots at the end and the higher average from the 2nd half returns - you are not part of it. Had it - spent it.  Your money is now in the hands of the cohort who were buying through that dip and recovery. 

While the buffered retiree doesn't sell equities in the "worst of it" whatever it is - for - maybe 3 years.  For many shapes of recession they move *their* curve back up from "retirement at risk of failing - terrible start date" cohort to something better - closer to the average.  The price being the lost inflationary and real equity returns on the cash buffer during the "not a crash" years prior net of whatever deposit/other asset return can be found for it.  Which is high if the good years are the first 18. And not if the first pre-crash years are 2.

Perhaps they then also use that time provided by the buffer to initiate a downsizing property in retirement move (normal move not a crisis) and then release non-correlated capital and prolong the pause in equity sales - if that is needed due to a longer shape of crash/recovery. 

Nothing can both generate a decent income and cope with a major correction followed by a decade long slump along the bottom.  There is no planning for that other than largely non-growth assets and linear depletion or the annuity route where you pool the cohort problem and pay the fee to be rid of the longevity issue.   Most history shows bonds moving against equities to a degree and also that 5-7 years is "enough" - except when it isn't (1920s (diversification doesn't help) or Japan (global holdings helps here).

This is an asymmetric risk.  Failed retirement consequence bad.  Slightly smaller inheritance - consequence less bad.  

I could setup my plan based on maximum investments and no buffer, set a slightly lower WR believed backtesting safe and just run with it.  And probably that would be equivalent or even very slightly better than many buffering models - on average. 

But I am not going to do that.  It will be buffered.   And yet from some perspectives - even with the buffer - I will still embracing a lot of risk and more than I maybe should.  My thread the other day on the Bernstein Liability Matched Portoflio ideas (from Rational Expectations) explores this a bit although it wanders onto diverse subjects

I really do think comfort blanket underplays this.  But I agree it undoubtedly is helpful in that dimension of having confidence in a plan and sticking to it. Which is important to avoid panic and unhelpful actions.  But it is far from a free lunch. 

If CAPE10 was lower at my retirement I would likely buffer less.  With it high - I will be building a buffer into my plans.  Just a function of my risk appetite and family factors to keep things "calm".  Other people will make different judgements.

One final supporting example - there was a thread the other day where someone was confused and dismayed by their IFA on 2020 drawdown reduction / suspension due to an absence of buffering (somewhere - either in their finances more generally or within the drawdown plan).  And there was an amount of tutting around portfolio setup that their IFA had left them in a position to need to do that during the (relatively mild) events of Feb 2020 or at least had not explained the variable nature of it better.