Bucketing strategy. When to de-risk?

Prism

Qyburn said:

Isnt the idea to allow choice, draw from equities if they're doing well, from non-equities if they aren't. Whereas I assume drawing from VLS60 you will always be taking the same proportion of each. 

Theoretically yes, but how would anyone make that choice with any level of accuracy? Just because one part of a portfolio falls more than the other, does not mean that it isn't about to fall more again. Taking a withdrawal from the bit that does better might be the worst thing to do. The easy option is to take from both in equal proportions, or rebalance for the same effect, or just use something like VLS60. 

Most safe withdrawal rate studies assume you take from both.

Linton

GazzaBloom said:

Qyburn said:

GazzaBloom said:

My initial plans for retirement are to have 15% (3 years of expenses) in cash/cash equivalent (MMF) and the remaining 85% in the same growth equity index funds I hold in accumulation, but not operate in 2 bucket mode, just draw proportionally and rebalance annually, rather than draw from cash solely in year 1.

If you you'll take 15% of each years income from cash and 85% from equities, what's your plan for years when equities fall substantially (or rise)?

Plough on with the same strategy but use Guyton's guardrails to reduce the drawdown amount taken in downturns or increase it in upswings, any excess gains could be used to increase the cash (risk-off) percentage over time as retirement rolls on.

How do you quickly reduce expenditure? Buy £8 bottles of wine rather than £10 ones? Cancel your planned meal out due that evening? Will that actually make any real difference? 

What about next years booked but not paid-for  £30K cruise to celebrate your 40th wedding anniversary? Sorry dear, Mr Guyton says we can’t afford it?

What about when the taps are turned on again? Would you know how to usefully spend the extra? Perhaps a wild binge?

In any case much expenditure is fixed, eg council tax, insurance, utilities

No, in my view it is far better to arrange your finances to avoid short/medium term fluctuations in income.  I do not believe you can significantly turn your standard of living up and down just like that.

Linton

Qyburn said:

Isnt the idea to allow choice, draw from equities if they're doing well, from non-equities if they aren't. Whereas I assume drawing from VLS60 you will always be taking the same proportion of each. 

Yes but the idea seems to be that you rebalance afterwards, which puts you back  in the same position as if you had drawn down the assets in the same proportions.

Linton

Prism said:

Qyburn said:

Isnt the idea to allow choice, draw from equities if they're doing well, from non-equities if they aren't. Whereas I assume drawing from VLS60 you will always be taking the same proportion of each. 

Theoretically yes, but how would anyone make that choice with any level of accuracy? Just because one part of a portfolio falls more than the other, does not mean that it isn't about to fall more again. Taking a withdrawal from the bit that does better might be the worst thing to do. The easy option is to take from both in equal proportions, or rebalance for the same effect, or just use something like VLS60. 

Most safe withdrawal rate studies assume you take from both.

Yes, with any switching between selling asset types you are into market timing.

Qyburn

Prism said:
Theoretically yes, but how would anyone make that choice with any level of accuracy? Just because one part of a portfolio falls more than the other, does not mean that it isn't about to fall more again. Taking a withdrawal from the bit that does better might be the worst thing to do. The easy option is to take from both in equal proportions, or rebalance for the same effect, or just use something like VLS60.

Drawing from whichever asset has done best (or least badly) at time of withdrawal means a given number of £ depletes that asset by the smallest %age.

Or maybe it would be better to draw a standard percentage each time from the highest risk component, topping up the low risk if that's more than you need, or drawing from the low risk if its not enough. 

I don't think I've seen any drawdown strategy without some element of guesswork or assumption about the future.

JohnWinder

Qyburn said:

Isnt the idea to allow choice, draw from equities if they're doing well, from non-equities if they aren't. Whereas I assume drawing from VLS60 you will always be taking the same proportion of each. 

Imagine a bad situation: equities crash just before you need to withdraw.

You have £60 in equities and £40 in bonds in one fund. Equities fall 25%. You now have £45 equities and £40 bonds. The fund manager rebalances daily or weekly, such that she sells £6 bonds to buy £6 of equities, giving you £51 equities and £34 bonds, for 60/40.

You now make your annual withdrawal of 4% of £100, ie £4. All the bonds you’ll be selling have not dropped in price; of all the equities you’ll be selling which is £2.40, about 12% were recently bought at bargain price after the crash. Check the maths, and do your own modelling, but a 60/40 fund withdrawal may not be as bad as appears at first blush.

Qyburn

JohnWinder said:

You have £60 in equities and £40 in bonds in one fund. Equities fall 25%. You now have £45 equities and £40 bonds. The fund manager rebalances daily or weekly, such that she sells £6 bonds to buy £6 of equities, giving you £51 equities and £34 bonds, for 60/40.

You now make your annual withdrawal of 4% of £100, ie £4. All the bonds you’ll be selling have not dropped in price; of all the equities you’ll be selling which is £2.40, about 12% were recently bought at bargain price after the crash. Check the maths, and do your own modelling, but a 60/40 fund withdrawal may not be as bad as appears at first blush.

You're correct that I need to do the maths. My impulse was that you'd better off taking all £4 from the bonds. But I think if you then rebalance it might end up the same as your example.

Prism

Qyburn said:

Prism said:
Theoretically yes, but how would anyone make that choice with any level of accuracy? Just because one part of a portfolio falls more than the other, does not mean that it isn't about to fall more again. Taking a withdrawal from the bit that does better might be the worst thing to do. The easy option is to take from both in equal proportions, or rebalance for the same effect, or just use something like VLS60.

Drawing from whichever asset has done best (or least badly) at time of withdrawal means a given number of £ depletes that asset by the smallest %age.

Or maybe it would be better to draw a standard percentage each time from the highest risk component, topping up the low risk if that's more than you need, or drawing from the low risk if its not enough. 

I don't think I've seen any drawdown strategy without some element of guesswork or assumption about the future.

I would actually say that most drawdown strategies make no assumptions of the future. Take the original 4% Trinity or Bengen models. They are based on yearly withdrawals of an inflation adjusted 4% along with a rebalance back to the 50/50 or 60/40 split. Forget about for another year and repeat.

Guyton Klinger takes the same idea but then reduces or increases withdrawals based on a formula. Uses a set allocation.

The bucket models move value from one pot to another based upon yearly expenditure. Sure, a bit of guesswork based on future inflation and costs, but not on market performance which is irrelevant to the plan.

GazzaBloom

Linton said:

GazzaBloom said:

Qyburn said:

GazzaBloom said:

My initial plans for retirement are to have 15% (3 years of expenses) in cash/cash equivalent (MMF) and the remaining 85% in the same growth equity index funds I hold in accumulation, but not operate in 2 bucket mode, just draw proportionally and rebalance annually, rather than draw from cash solely in year 1.

If you you'll take 15% of each years income from cash and 85% from equities, what's your plan for years when equities fall substantially (or rise)?

Plough on with the same strategy but use Guyton's guardrails to reduce the drawdown amount taken in downturns or increase it in upswings, any excess gains could be used to increase the cash (risk-off) percentage over time as retirement rolls on.

How do you quickly reduce expenditure? Buy £8 bottles of wine rather than £10 ones? Cancel your planned meal out due that evening? Will that actually make any real difference? 

What about next years booked but not paid-for  £30K cruise to celebrate your 40th wedding anniversary? Sorry dear, Mr Guyton says we can’t afford it?

What about when the taps are turned on again? Would you know how to usefully spend the extra? Perhaps a wild binge?

In any case much expenditure is fixed, eg council tax, insurance, utilities

No, in my view it is far better to arrange your finances to avoid short/medium term fluctuations in income.  I do not believe you can significantly turn your standard of living up and down just like that.

Well you certainly seem quite set on your plan.

30% of our annual planned day to day spend is going to be discretionary and a 5% reduction if SWR goes up 20% is what is modelled and it makes a significant difference over time.

So, go with me here, and help me understand in a little more detail how your bucketing plan works. Let's say I need £32K a year in retirement (after tax - I calculate a drawdown amount of £29,700 from DC pension UFPLS) with £6K coming from a DB pension, which after tax should give the £32K or thereabouts. State Pensions are 8 years away. Ignore one off spend items for now.

So for arguments sake let¡s say I need to drawdown £30K in year 1, how much would you put in each of your 3 buckets and how does your drawdown work year to year? And how do your bucket top ups work?

Qyburn

Prism said:

I would actually say that most drawdown strategies make no assumptions of the future. 

Broadly they all do, if not in detail. I don't think any would succeed if the investments dropped 50% followed by zero growth for the next ten years.