SWRs when front loading drawdown

NedS

Hi everyone,

I guess I am not alone in being in a position where I have a mixture of both DB (plus state pension) and DC pensions to consider when planning my retirement. "The Plan" is to retire early, around 57-58, use the DC pot to "bridge the gap" by front loading drawdown in the early years until all DB/state pensions come on line. We are currently 52 and have 4 DB pensions between us which kick in at various points between 60-67 plus will both have full state pensions at 67. By the time we hit 67 we will have sufficient DB/SP income for the essentials but will still very much want some income from the remaining SIPP to cover larger/unexpected expenditure and those other things in life.

So I've done the basics - I have a spreadsheet of our current expenditure, I have a good idea of our fixed outgoings and what income we will need, and where that income will come from.

Since joining the forums here I've read up on SWRs and sequence of return risks etc. My question is around how do people plan/model more complicated scenarios where there are other fixed income streams to consider (DB / State Pension), and when a constant withdraw rate (e.g, 4% + CPI) is not appropriate as is the case here? Do people just use a bucket approach and say I need X amount of money in these years etc? What about investment strategies (for want of a better expression) pre the gap years, during the gap years, and post the gap years - do they change? How do they change?

I've read up on things like Guyton-Klinger rules. Being of mathematical mind, I like defined rules I can follow, but whilst Guyton-Klinger rules may work well for a DC only retirement, they do not seem applicable to my situation. Are there any other sets of rules or mathematical models I should be considering.

I'm guessing there are no easy answers to these questions otherwise I probably would have read about them online already, so I'm kind of hoping to start a discussion and find out how other people approach this situation as opposed to expecting someone to provide some hard and fast mathematical rule (although that would be nice!)

BritishInvestor

NedS said:

Hi everyone,

I guess I am not alone in being in a position where I have a mixture of both DB (plus state pension) and DC pensions to consider when planning my retirement. "The Plan" is to retire early, around 57-58, use the DC pot to "bridge the gap" by front loading drawdown in the early years until all DB/state pensions come on line. We are currently 52 and have 4 DB pensions between us which kick in at various points between 60-67 plus will both have full state pensions at 67. By the time we hit 67 we will have sufficient DB/SP income for the essentials but will still very much want some income from the remaining SIPP to cover larger/unexpected expenditure and those other things in life.

So I've done the basics - I have a spreadsheet of our current expenditure, I have a good idea of our fixed outgoings and what income we will need, and where that income will come from.

Since joining the forums here I've read up on SWRs and sequence of return risks etc. My question is around how do people plan/model more complicated scenarios where there are other fixed income streams to consider (DB / State Pension), and when a constant withdraw rate (e.g, 4% + CPI) is not appropriate as is the case here? Do people just use a bucket approach and say I need X amount of money in these years etc? What about investment strategies (for want of a better expression) pre the gap years, during the gap years, and post the gap years - do they change? How do they change?

I've read up on things like Guyton-Klinger rules. Being of mathematical mind, I like defined rules I can follow, but whilst Guyton-Klinger rules may work well for a DC only retirement, they do not seem applicable to my situation. Are there any other sets of rules or mathematical models I should be considering.

I'm guessing there are no easy answers to these questions otherwise I probably would have read about them online already, so I'm kind of hoping to start a discussion and find out how other people approach this situation as opposed to expecting someone to provide some hard and fast mathematical rule (although that would be nice!)

Surely there are free tools out there that offer the ability to alter desired income in retirement and give an SWR? This is (or should be) a trivial problem for a tool to solve.
I don't see why Guyton-Klinger rules wouldn't be applicable (might be misunderstanding what you are asking)? 

Thrugelmir

If you are planning to liquidise your investments over a 9-10 period then you need to minimise risk (volatility) as far as possible. As the time frame is to short to recover from poor periods of performance.  Perhaps split your portfolio into two parts and monitor each separately. 

NedS

Thrugelmir said:

If you are planning to liquidise your investments over a 9-10 period then you need to minimise risk (volatility) as far as possible. As the time frame is to short to recover from poor periods of performance.  Perhaps split your portfolio into two parts and monitor each separately. 

Yes, the bucket approach. That's kind of how I've been approaching it. Plan how much cash we will need for 10 year period aged 58-67, and put aside enough cash (3 years, bucket 1) and shorter term investments (7 years (3-10), bucket 2), with everything else in higher risk bucket 3 to which Guyton-Klinger type rules could certainly apply.

I guess my main issue is I don't know how best to invest the allocation in buckets 1 & 2 which will be needed to fund the period in 5-15 years time. I have a generous cash allocation already, I don't like the look of bonds and am unsure how much equity risk I should be taking with that 8-15 year tranche. I'm guessing the last 5 years worth could be in equities for now with a 10-15 year time frame.

BritishInvestor

Thrugelmir said:

If you are planning to liquidise your investments over a 9-10 period then you need to minimise risk (volatility) as far as possible. As the time frame is to short to recover from poor periods of performance.  Perhaps split your portfolio into two parts and monitor each separately. 

I don't see that in the arbitrary example that I cobbled together
For example, someone with a £100k pot taking £10k inflation-adjusted each year would've had a horrible time if they had retired in 1915 with a 20%/80% global equity bond split with the money running out in 6 years whereas an 80% equity/20% bond would've fared a tad better with the equity heavy portfolio having a greater overall success rate with the data set I have.
The limitations of bucketing have already been discussed.

Thrugelmir

With regards to equities. Only time will provide the answer. Challenging times lie ahead for many sectors. 

Audaxer

NedS said:

Thrugelmir said:

If you are planning to liquidise your investments over a 9-10 period then you need to minimise risk (volatility) as far as possible. As the time frame is to short to recover from poor periods of performance.  Perhaps split your portfolio into two parts and monitor each separately. 

Yes, the bucket approach. That's kind of how I've been approaching it. Plan how much cash we will need for 10 year period aged 58-67, and put aside enough cash (3 years, bucket 1) and shorter term investments (7 years (3-10), bucket 2), with everything else in higher risk bucket 3 to which Guyton-Klinger type rules could certainly apply.

I guess my main issue is I don't know how best to invest the allocation in buckets 1 & 2 which will be needed to fund the period in 5-15 years time. I have a generous cash allocation already, I don't like the look of bonds and am unsure how much equity risk I should be taking with that 8-15 year tranche. I'm guessing the last 5 years worth could be in equities for now with a 10-15 year time frame.

If it was me I would ensure I had enough in the cash bucket to cover from age 58 to 67, as you can't be certain than even low risk investments will not lose money over that sort of period. As you said that you have various DB pensions kicking in between 60 and 67, you won't even need 10 full years of cash income during that period. If the rest of your investments planned for year 11 onwards are 100% equities, you might want to de-risk the percentage of equities as you get nearer the time when you need to take income from them.

michaels

You could think about it as a pot that gives you the same amount as your state/DB pension until retirement age invested conservatively and the remainder topping up your income both now and then by the SWR held in equities.

Both cfiresim and the SWR toolbox let you model different income streams.

[Deleted User]

NedS said:

Hi everyone,

I guess I am not alone in being in a position where I have a mixture of both DB (plus state pension) and DC pensions to consider when planning my retirement. "The Plan" is to retire early, around 57-58, use the DC pot to "bridge the gap" by front loading drawdown in the early years until all DB/state pensions come on line. We are currently 52 and have 4 DB pensions between us which kick in at various points between 60-67 plus will both have full state pensions at 67. By the time we hit 67 we will have sufficient DB/SP income for the essentials but will still very much want some income from the remaining SIPP to cover larger/unexpected expenditure and those other things in life.

So I've done the basics - I have a spreadsheet of our current expenditure, I have a good idea of our fixed outgoings and what income we will need, and where that income will come from.

Since joining the forums here I've read up on SWRs and sequence of return risks etc. My question is around how do people plan/model more complicated scenarios where there are other fixed income streams to consider (DB / State Pension), and when a constant withdraw rate (e.g, 4% + CPI) is not appropriate as is the case here? Do people just use a bucket approach and say I need X amount of money in these years etc? What about investment strategies (for want of a better expression) pre the gap years, during the gap years, and post the gap years - do they change? How do they change?

I've read up on things like Guyton-Klinger rules. Being of mathematical mind, I like defined rules I can follow, but whilst Guyton-Klinger rules may work well for a DC only retirement, they do not seem applicable to my situation. Are there any other sets of rules or mathematical models I should be considering.

I'm guessing there are no easy answers to these questions otherwise I probably would have read about them online already, so I'm kind of hoping to start a discussion and find out how other people approach this situation as opposed to expecting someone to provide some hard and fast mathematical rule (although that would be nice!)

Variable Percentage Withdrawal does this for you. The spreadsheet accounts for future DB pensions kicking in. 

https://www.finiki.org/wiki/Variable_percentage_withdrawal

BritishInvestor

Audaxer said:

NedS said:

Thrugelmir said:

If you are planning to liquidise your investments over a 9-10 period then you need to minimise risk (volatility) as far as possible. As the time frame is to short to recover from poor periods of performance.  Perhaps split your portfolio into two parts and monitor each separately. 

Yes, the bucket approach. That's kind of how I've been approaching it. Plan how much cash we will need for 10 year period aged 58-67, and put aside enough cash (3 years, bucket 1) and shorter term investments (7 years (3-10), bucket 2), with everything else in higher risk bucket 3 to which Guyton-Klinger type rules could certainly apply.

I guess my main issue is I don't know how best to invest the allocation in buckets 1 & 2 which will be needed to fund the period in 5-15 years time. I have a generous cash allocation already, I don't like the look of bonds and am unsure how much equity risk I should be taking with that 8-15 year tranche. I'm guessing the last 5 years worth could be in equities for now with a 10-15 year time frame.

If it was me I would ensure I had enough in the cash bucket to cover from age 58 to 67, as you can't be certain than even low risk investments will not lose money over that sort of period. As you said that you have various DB pensions kicking in between 60 and 67, you won't even need 10 full years of cash income during that period. If the rest of your investments planned for year 11 onwards are 100% equities, you might want to de-risk the percentage of equities as you get nearer the time when you need to take income from them.

Without knowing the numbers involved it's tricky to say whether ensuring sufficient cash was available to cover inflation-linked (I assume) withdrawals that worked over historical periods would jeopardise the remainder of the retirement plan.
To reiterate, this is a very common retiree objective.
At retirement target a certain withdrawal amount.
At SP age (or when DB pensions(s) start) reduce DC withdrawals to cater for this.
In later life assume reduced spending. 
Why not start with the easiest solution? Given:

Spending plans (from the DC pot)
Longevity assumptions
Asset allocation
Fees/slippage

What success rate does this give based on historical outcomes?
No need to start thinking about bucketing, Guyton rules etc at this stage IMO

kangoora

I think you're over-complicating things. SWR, whether using Guyton-Klinger or not, is used as a means of withdrawing a certain amount of cash per year whilst (hopefully) maintaining the capital. You are going to be drawing down 'x' amount of money to maintain a lifestyle until your pensions kick in, so SWR is immaterial in your drawdown phase UNLESS you are close to the wire on DC pot lasting until you get your pensions. If this is the case I'd question the wisdom of retiring at your planned point and maybe think 'just one more year'. Once you have reached full pensions then SWR can be used to maintain what remains of your capital/extra spending money after sustaining you for 10 years.

I'm doing what you are planning, am at the age you plan to start and am broadly similar in that my DB and SP pensions will cover general spending at age 67. I just keep 3 years expenses in cash/ISA's/some 2 year bonds. The rest is invested in VLS60 and equvalent for (hopefully) some capital growth and to draw down a top-up to existing DBs before age 67. I've also got 2 reasonable TFLS coming in before 67, each of which is around 1/2 my annual spends so some years I'll draw even less from the DC pot.

I'd recommend cfiresim or Firecalc to run your numbers (just remember to up the charges to reflect what your pension charges are as they assume very low rates), they are american based but should let you know if you have a good/reasonable or barking mad chance of achieving your aims.