Anyone in high equity allocation whilst retired?

Linton

michaels said:

Linton said:

DT2001 said:

As Linton says it depends on your objectives in retirement. If a secure income is paramount then a lower equity/bonds/cash mix or an annuity would be a better option.  I think having the ability to be flexible with your expenditure is key to being able to cope with a poor sequence of returns early in retirement. I have played with the idea of taking a fixed % of the portfolio and topping up from the cash buffer. I also added an IF function so that a poor sequence meant only part of the ‘shortfall’ was covered. Once we are both drawing SP less than 50% of our guesstimated expenditure will be from investments - a 30% total drop in the markets in the first 3 years would see a less than 10% reduction in total income.
A long slow recovery would mean more adjustments along the line but then in that situation most strategies would probably be struggling. 

The one final tweak I have considered is calculating my drawdown 1/4ly which will help smooth the ups and downs.

I dont see how being flexible with your expenditure in the face of short term investment volatility can be implemented:....
1) Much expenditure is fixed over the medium term - eg utilities and council tax.  You cant temporarily stop paying them.

2) Economising on expenditure takes time to have a useful effect.  Just turning down the central heating by 1 deg  will have minimal short term effect.  Reducing the quality of the wine you buy is not the answer to a market crash. You will have become used to a particular standard of living, making a substantial change up or down in a short time frame would be difficult.

3) Some expenditure may be fixed months in advance.  eg a cruise may need to be booked a year in advance with a final major payment a few months before departure.  Do you cancel if the market falls in the meantime?

4) As with any policy of responding to short term events, timing is very difficult.  At what point in a market wobble do you change your expenditure?  Take recent events - would you have cut expenditure when Trump came into office or would you still be wondering whether to cut it now? Similarly how do you decide when things have returned to normal?  

In my view a retiree largely dependent on investment income has to decouple expenditure from the release of capital in the short/medium term.  This implies some form of buffering, particularly for large one-off expenditures, and/or some other sources of ongoing income than selling capital to cover ongoing fixed expenditure.

I don't think annuities are a satisfactory complete answer because of their inflexibility.

But once you have spent that buffer then either you have no buffer (which you have just said is necessary) or you have to sell equities to replenish it.

I am not saying it doesn't make sense to use some sort of variable allocation strategy, but I just think how this strategy operates needs to be clearly stated so it can be back tested.

Yes, you need a strategy to replenish the buffer over time, but doing so is not time-critical unlike meeting your utility bills. I dont see having a tightly defined strategy is essential, ongoing monitoring with occasional payments is fine. Waiting until the buffer is empty and then refilling it would seem to run counter to its whole purpose.

 Most of my replenishing comes from planning for more income than required with the excess staying in the buffer.

You may need to devise an artificial replenishment strategy if you want to back test.  Personally I dont see much practical benefit in back testing other than as a one-off sanity check and teaching aid. 

Notfarfromtheborder

michaels said:

Linton said:

DT2001 said:

As Linton says it depends on your objectives in retirement. If a secure income is paramount then a lower equity/bonds/cash mix or an annuity would be a better option.  I think having the ability to be flexible with your expenditure is key to being able to cope with a poor sequence of returns early in retirement. I have played with the idea of taking a fixed % of the portfolio and topping up from the cash buffer. I also added an IF function so that a poor sequence meant only part of the ‘shortfall’ was covered. Once we are both drawing SP less than 50% of our guesstimated expenditure will be from investments - a 30% total drop in the markets in the first 3 years would see a less than 10% reduction in total income.
A long slow recovery would mean more adjustments along the line but then in that situation most strategies would probably be struggling. 

The one final tweak I have considered is calculating my drawdown 1/4ly which will help smooth the ups and downs.

I dont see how being flexible with your expenditure in the face of short term investment volatility can be implemented:....
1) Much expenditure is fixed over the medium term - eg utilities and council tax.  You cant temporarily stop paying them.

2) Economising on expenditure takes time to have a useful effect.  Just turning down the central heating by 1 deg  will have minimal short term effect.  Reducing the quality of the wine you buy is not the answer to a market crash. You will have become used to a particular standard of living, making a substantial change up or down in a short time frame would be difficult.

3) Some expenditure may be fixed months in advance.  eg a cruise may need to be booked a year in advance with a final major payment a few months before departure.  Do you cancel if the market falls in the meantime?

4) As with any policy of responding to short term events, timing is very difficult.  At what point in a market wobble do you change your expenditure?  Take recent events - would you have cut expenditure when Trump came into office or would you still be wondering whether to cut it now? Similarly how do you decide when things have returned to normal?  

In my view a retiree largely dependent on investment income has to decouple expenditure from the release of capital in the short/medium term.  This implies some form of buffering, particularly for large one-off expenditures, and/or some other sources of ongoing income than selling capital to cover ongoing fixed expenditure.

I don't think annuities are a satisfactory complete answer because of their inflexibility.

But once you have spent that buffer then either you have no buffer (which you have just said is necessary) or you have to sell equities to replenish it.

I am not saying it doesn't make sense to use some sort of variable allocation strategy, but I just think how this strategy operates needs to be clearly stated so it can be back tested.

And that for me will be the difficult one, what do I sell when I need income, Six months ago my pension was £585K, that's about where it is today (despite contributions over the past 6 months). During this 6 month period its peak was £615K and base £562K.

6 months ago if I had of needed funds I'd have probably sold equities as it was at its peak at this point and maintained the cash buffer, not sure what I would do now despite the pot being the same in value, as it is down from peak by around 5%, however, up from recent base by 4%. Maybe some sell some cash funds and some equities?

I suppose time will tell and I'll never be right all the time, just hope I'm not wrong all the time

Albermarle

Linton said:

DT2001 said:

As Linton says it depends on your objectives in retirement. If a secure income is paramount then a lower equity/bonds/cash mix or an annuity would be a better option.  I think having the ability to be flexible with your expenditure is key to being able to cope with a poor sequence of returns early in retirement. I have played with the idea of taking a fixed % of the portfolio and topping up from the cash buffer. I also added an IF function so that a poor sequence meant only part of the ‘shortfall’ was covered. Once we are both drawing SP less than 50% of our guesstimated expenditure will be from investments - a 30% total drop in the markets in the first 3 years would see a less than 10% reduction in total income.
A long slow recovery would mean more adjustments along the line but then in that situation most strategies would probably be struggling. 

The one final tweak I have considered is calculating my drawdown 1/4ly which will help smooth the ups and downs.

I dont see how being flexible with your expenditure in the face of short term investment volatility can be implemented:....
1) Much expenditure is fixed over the medium term - eg utilities and council tax.  You cant temporarily stop paying them.

2) Economising on expenditure takes time to have a useful effect.  Just turning down the central heating by 1 deg  will have minimal short term effect.  Reducing the quality of the wine you buy is not the answer to a market crash. You will have become used to a particular standard of living, making a substantial change up or down in a short time frame would be difficult.

3) Some expenditure may be fixed months in advance.  eg a cruise may need to be booked a year in advance with a final major payment a few months before departure.  Do you cancel if the market falls in the meantime?

4) As with any policy of responding to short term events, timing is very difficult.  At what point in a market wobble do you change your expenditure?  Take recent events - would you have cut expenditure when Trump came into office or would you still be wondering whether to cut it now? Similarly how do you decide when things have returned to normal?  

In my view a retiree largely dependent on investment income has to decouple expenditure from the release of capital in the short/medium term.  This implies some form of buffering, particularly for large one-off expenditures, and/or some other sources of ongoing income than selling capital to cover ongoing fixed expenditure.

I dont think annuities are a satisfactory complete answer because of their inflexibility.

Regarding your comment I have highlighted in bold ( which I agree with)

In addition you ( and your partner/family if you have one ) also may well have become used to not worrying about unplanned expenditure  in the short term too much. Car needs four new tyres- no problem; Fence needs replacing - no problem; Invited to wedding in another country - Great .
Suddenly you are stressing about these things. . That would be difficult and could well cause some family tension....

MK62

michaels said:

MK62 said:

michaels said:

I go with the maths/history (facts) approach that a 'cash buffer' is very much an artificial psychological prop rather than a serious de-risking strategy.

....but this isn't really supported by the facts, which actually show that in drawdown a cash buffer reduces the plan failure rate compared to a 100% equity portfolio........admittedly at the cost of a lower average final balance.

So it really depends what your main priority is.......however, a cash buffer is arguably of less use if a variable withdrawal is acceptable.

I agree that a portfolio with a fixed proportion of equities and bonds has historically had a higher (no failure) safe withdrawal rate than one only consisting of equities.

But that is very different to some sort of portfolio where the proportion of equities and cash is allocated dynamically according to some measure of equity performance. 

Can you show the proportions and strategy that you have in mind that reliably delivers a higher SWR?

Again, that's the issue......it's not about having a higher income using one method or the other......it's about having the same income with a lower plan failure rate.
If you compare Bill, with a £100k equity portfolio, and Ben with an £85k equity portfolio and a £15k cash buffer, for example, then it's not about Ben somehow being able to have a higher income than Bill because he is using a cash buffer......it's about whether Ben can have the same income as Bill, but with a lower chance of plan failure......on average.

If the income level desired is £2000pa (ie 2%), then it's highly likely that neither Bill nor Ben will suffer plan failure.......but if you push that to say £4500pa, what then?
As to the question of when Ben should replenish the cash buffer, it's not that much different to the question of when Bill should sell equities to provide the income withdrawal.....the main difference being that Bill may be forced to sell at a certain time, whereas Ben is not (though he could of course choose the same time as Bill if he wanted). Is this market timing?......perhaps, at least to some degree, but then one way or another both Bill and Ben must choose a time to sell portfolio assets......be that, in Bill's case, he needs money now, or on a fixed date each year, come what may, or throwing a dart at a calendar, or choosing a time yourself - it's unlikely you'll pick the best day of course, only hindsight will tell you that, but then it's equally unlikely that you'll pick the worst either.

As to outcomes, if markets are kind, then there is a good chance that neither Bill nor Ben will suffer plan failure........Bill may end up with a bigger residual pot......but that's the cost of using a cash buffer if it eventually turns out that Ben didn't need to. For myself, this is not a primary concern though.
However if the markets are unkind.......that is a primary concern.

GazzaBloom

I retired at the end of last year and switched from around a 80% equity portfolio to a rising equity glidepath strategy to start retirement. 8 years of inflated annual living expenses plus known one off spends (new car etc.) was put in cash/MMF and will be drawn first over the first 8 years. The percentage split between equities and cash is arbitrary, it was determined by the required cash allocation to last 8 years, but, came out at 45% cash and 55% equities.

Obviously the percentage allocation will change over time as the cash is consumed and the equity portion rises in percentage, ultimately putting us at 100% equities when our state pensions start paying out in 8 years time.

When modelled, this strategy shows a greater resilience to early sequence of returns risk than a fixed percentage allocation that is rebalanced annually, in the worst returns scenarios with only a modest sacrifice of end of life remaining balance. in the best returns scenarios.

We have around 30% fixed income from other sources in addition to the invested/cash portfolio.

OldScientist

Linton said:

DT2001 said:

As Linton says it depends on your objectives in retirement. If a secure income is paramount then a lower equity/bonds/cash mix or an annuity would be a better option.  I think having the ability to be flexible with your expenditure is key to being able to cope with a poor sequence of returns early in retirement. I have played with the idea of taking a fixed % of the portfolio and topping up from the cash buffer. I also added an IF function so that a poor sequence meant only part of the ‘shortfall’ was covered. Once we are both drawing SP less than 50% of our guesstimated expenditure will be from investments - a 30% total drop in the markets in the first 3 years would see a less than 10% reduction in total income.
A long slow recovery would mean more adjustments along the line but then in that situation most strategies would probably be struggling. 

The one final tweak I have considered is calculating my drawdown 1/4ly which will help smooth the ups and downs.

I dont see how being flexible with your expenditure in the face of short term investment volatility can be implemented:....
1) Much expenditure is fixed over the medium term - eg utilities and council tax.  You cant temporarily stop paying them.

2) Economising on expenditure takes time to have a useful effect.  Just turning down the central heating by 1 deg  will have minimal short term effect.  Reducing the quality of the wine you buy is not the answer to a market crash. You will have become used to a particular standard of living, making a substantial change up or down in a short time frame would be difficult.

3) Some expenditure may be fixed months in advance.  eg a cruise may need to be booked a year in advance with a final major payment a few months before departure.  Do you cancel if the market falls in the meantime?

4) As with any policy of responding to short term events, timing is very difficult.  At what point in a market wobble do you change your expenditure?  Take recent events - would you have cut expenditure when Trump came into office or would you still be wondering whether to cut it now? Similarly how do you decide when things have returned to normal?  

In my view a retiree largely dependent on investment income has to decouple expenditure from the release of capital in the short/medium term.  This implies some form of buffering, particularly for large one-off expenditures, and/or some other sources of ongoing income than selling capital to cover ongoing fixed expenditure.

I dont think annuities are a satisfactory complete answer because of their inflexibility.

Your first 4 points are difficult to argue with.

Your point I have taken the liberty of placing in bold is, IMV a critical one. Assuming other sources of income (e.g., state pension and DB pension) are insufficient to maintain an acceptable standard of living then there are several methods of reducing the coupling of portfolio income from market fluctuations
1) RPI annuities (which as you say are inflexible, but currently allow a strong baseline of income).
2) Inflation linked gilt ladders. As for annuities, with, potentially, more flexibility.
3) Increase the quantity of fixed income in the portfolio since, provided the duration is selected carefully*, this is likely to decrease volatility.
4) Use a 'smoothed' percentage of portfolio withdrawal method (Vanguard dynamic, Carlson's endowment, Guyton-Klinger, etc.)
5) Try to implement a dynamic asset allocation strategy (of which cash buffers and buckets are examples) -  McClung's free chapter (see https://livingoffyourmoney.com/wp-content/uploads/2016/05/LivingOffYourOwnMoney_eBook_FirstThreeChapters.pdf ) looks at a bucket strategy.

*Historical back testing has shown that 'intermediate' duration bond funds form a useful compromise since they have lower volatility than longer duration bonds and, generally, higher returns than short duration fixed income (inc 'cash'). 'Intermediate' duration falls around the 5 year mark (although it is slightly lower for US than the UK).

Pat38493

OldScientist said:

Linton said:

DT2001 said:

As Linton says it depends on your objectives in retirement. If a secure income is paramount then a lower equity/bonds/cash mix or an annuity would be a better option.  I think having the ability to be flexible with your expenditure is key to being able to cope with a poor sequence of returns early in retirement. I have played with the idea of taking a fixed % of the portfolio and topping up from the cash buffer. I also added an IF function so that a poor sequence meant only part of the ‘shortfall’ was covered. Once we are both drawing SP less than 50% of our guesstimated expenditure will be from investments - a 30% total drop in the markets in the first 3 years would see a less than 10% reduction in total income.
A long slow recovery would mean more adjustments along the line but then in that situation most strategies would probably be struggling. 

The one final tweak I have considered is calculating my drawdown 1/4ly which will help smooth the ups and downs.

I dont see how being flexible with your expenditure in the face of short term investment volatility can be implemented:....
1) Much expenditure is fixed over the medium term - eg utilities and council tax.  You cant temporarily stop paying them.

2) Economising on expenditure takes time to have a useful effect.  Just turning down the central heating by 1 deg  will have minimal short term effect.  Reducing the quality of the wine you buy is not the answer to a market crash. You will have become used to a particular standard of living, making a substantial change up or down in a short time frame would be difficult.

3) Some expenditure may be fixed months in advance.  eg a cruise may need to be booked a year in advance with a final major payment a few months before departure.  Do you cancel if the market falls in the meantime?

4) As with any policy of responding to short term events, timing is very difficult.  At what point in a market wobble do you change your expenditure?  Take recent events - would you have cut expenditure when Trump came into office or would you still be wondering whether to cut it now? Similarly how do you decide when things have returned to normal?  

In my view a retiree largely dependent on investment income has to decouple expenditure from the release of capital in the short/medium term.  This implies some form of buffering, particularly for large one-off expenditures, and/or some other sources of ongoing income than selling capital to cover ongoing fixed expenditure.

I dont think annuities are a satisfactory complete answer because of their inflexibility.

Your first 4 points are difficult to argue with.

Your point I have taken the liberty of placing in bold is, IMV a critical one. Assuming other sources of income (e.g., state pension and DB pension) are insufficient to maintain an acceptable standard of living then there are several methods of reducing the coupling of portfolio income from market fluctuations
1) RPI annuities (which as you say are inflexible, but currently allow a strong baseline of income).
2) Inflation linked gilt ladders. As for annuities, with, potentially, more flexibility.
3) Increase the quantity of fixed income in the portfolio since, provided the duration is selected carefully*, this is likely to decrease volatility.
4) Use a 'smoothed' percentage of portfolio withdrawal method (Vanguard dynamic, Carlson's endowment, Guyton-Klinger, etc.)
5) Try to implement a dynamic asset allocation strategy (of which cash buffers and buckets are examples) -  McClung's free chapter (see https://livingoffyourmoney.com/wp-content/uploads/2016/05/LivingOffYourOwnMoney_eBook_FirstThreeChapters.pdf ) looks at a bucket strategy.

*Historical back testing has shown that 'intermediate' duration bond funds form a useful compromise since they have lower volatility than longer duration bonds and, generally, higher returns than short duration fixed income (inc 'cash'). 'Intermediate' duration falls around the 5 year mark (although it is slightly lower for US than the UK).

When you say intermedidate duration bonds, if it's in a fund do you mean "effective maturity" or "effective duration"?

I am holding some global bond funds "Vanguard Global Bond Index Fund GBP Hedged" and also the ETF version of that fund in a different account.  It says:
Effective Maturity 8.43
Effective duration 6.37

I am not sure what the difference is beetween them, but if duration as you discuss above is the same things, is 6.3 years too long then?

Pat38493

GazzaBloom said:

I retired at the end of last year and switched from around a 80% equity portfolio to a rising equity glidepath strategy to start retirement. 8 years of inflated annual living expenses plus known one off spends (new car etc.) was put in cash/MMF and will be drawn first over the first 8 years. The percentage split between equities and cash is arbitrary, it was determined by the required cash allocation to last 8 years, but, came out at 45% cash and 55% equities.

Obviously the percentage allocation will change over time as the cash is consumed and the equity portion rises in percentage, ultimately putting us at 100% equities when our state pensions start paying out in 8 years time.

When modelled, this strategy shows a greater resilience to early sequence of returns risk than a fixed percentage allocation that is rebalanced annually, in the worst returns scenarios with only a modest sacrifice of end of life remaining balance. in the best returns scenarios.

We have around 30% fixed income from other sources in addition to the invested/cash portfolio.

Interesting that you are finding that an 8 year cash bucket is working for you.  I think from memory you have a pretty front loaded bridging strategy like me?

If I play around with back testing, my probability of success is very similar no matter whether I pick anything bettween 20 to 80% equities or so.  If I tried to model an 8 to 10 years cash bucket my probability of success is goes from 95% to zero, but that may be because Timeline has a 0% return on cash in their model I think.

However I also look closely at when the first failure occurs, since my spending  in years 1 to 11 is much more heavily reliant on investments.  

Because most of my withdrawals are in the first 11 years, I get some pretty interesting results.

Overall mix of equities 15% bonds 85%, gives a success rate of 96%, but almost all the failures are before state pension age, with the earliest at age 65.  3% chance of running out of money before SP age.

On the other hand, using overall allocations of around 70% equities, I have a slightly lower overall chance of plan success at 93%, but only 1% chance of running out before SP age, with the first failurre near the end of when I am 66 years old.

Point being, carrying a large amount of cash (and to a lesser extent bonds) exposes you to inflation risk.  Unsurprisingly the first failure in my plan is 1915 - a year which was followed by quite a few years of double digit inflation.

michaels

MK62 said:

michaels said:

MK62 said:

michaels said:

I go with the maths/history (facts) approach that a 'cash buffer' is very much an artificial psychological prop rather than a serious de-risking strategy.

....but this isn't really supported by the facts, which actually show that in drawdown a cash buffer reduces the plan failure rate compared to a 100% equity portfolio........admittedly at the cost of a lower average final balance.

So it really depends what your main priority is.......however, a cash buffer is arguably of less use if a variable withdrawal is acceptable.

I agree that a portfolio with a fixed proportion of equities and bonds has historically had a higher (no failure) safe withdrawal rate than one only consisting of equities.

But that is very different to some sort of portfolio where the proportion of equities and cash is allocated dynamically according to some measure of equity performance. 

Can you show the proportions and strategy that you have in mind that reliably delivers a higher SWR?

Again, that's the issue......it's not about having a higher income using one method or the other......it's about having the same income with a lower plan failure rate.
If you compare Bill, with a £100k equity portfolio, and Ben with an £85k equity portfolio and a £15k cash buffer, for example, then it's not about Ben somehow being able to have a higher income than Bill because he is using a cash buffer......it's about whether Ben can have the same income as Bill, but with a lower chance of plan failure......on average.

If the income level desired is £2000pa (ie 2%), then it's highly likely that neither Bill nor Ben will suffer plan failure.......but if you push that to say £4500pa, what then?
As to the question of when Ben should replenish the cash buffer, it's not that much different to the question of when Bill should sell equities to provide the income withdrawal.....the main difference being that Bill may be forced to sell at a certain time, whereas Ben is not (though he could of course choose the same time as Bill if he wanted). Is this market timing?......perhaps, at least to some degree, but then one way or another both Bill and Ben must choose a time to sell portfolio assets......be that, in Bill's case, he needs money now, or on a fixed date each year, come what may, or throwing a dart at a calendar, or choosing a time yourself - it's unlikely you'll pick the best day of course, only hindsight will tell you that, but then it's equally unlikely that you'll pick the worst either.

As to outcomes, if markets are kind, then there is a good chance that neither Bill nor Ben will suffer plan failure........Bill may end up with a bigger residual pot......but that's the cost of using a cash buffer if it eventually turns out that Ben didn't need to. For myself, this is not a primary concern though.
However if the markets are unkind.......that is a primary concern.

By SWR I am talking about the highest 'no historic failure' real terms constant withdrawal amount expressed as a percentage of the initial pot size.  This definition builds in risk of failure as it is set to zero.

It is obviously easier to model the situation where the portfolio is set to a fixed asset mix proportion and rebalanced regularly and this is what underpins most of the SWR analysis so we can say over the 'worst' possible period what fixed proportions of equities vs bonds vs cash/MM would have given the highest no failure withdrawal rate.

Once you start going to 'replenishable cast pots' then you need to add another set of rules on how those pots are managed.  Rather that each year you rebalance to 70% equities, 25% bonds, 5% cash or whatever, you now need a rule that says 'if equity markets are more than 10% below market high / AP is above 25 or whatever) then rebalance to 80% / 20% / 0% (=markets are down, spend from cash and bonds) otherwise go back to default 70% / 25% / 5%.

Even this simple set of rules makes the modelling of that initial SWR/annual real terms amount much more difficult but in order to justify this or any other set of rules it would seem to make sense to either back test against history or use a montecarlo simulation in order to work out what your SWR/initial safe withdrawal amount is to determine if it is larger than the traditional fixed proportion SWR and indeed simply to know how much you should be contemplating withdrawing.

Are you suggesting that with a cash pot and some sort of dynamic allocation strategy you can withdraw more than the traditional SWR no failure amount from year one in real terms and still have zero failures?

Linton

OldScientist said:

Linton said:

DT2001 said:

As Linton says it depends on your objectives in retirement. If a secure income is paramount then a lower equity/bonds/cash mix or an annuity would be a better option.  I think having the ability to be flexible with your expenditure is key to being able to cope with a poor sequence of returns early in retirement. I have played with the idea of taking a fixed % of the portfolio and topping up from the cash buffer. I also added an IF function so that a poor sequence meant only part of the ‘shortfall’ was covered. Once we are both drawing SP less than 50% of our guesstimated expenditure will be from investments - a 30% total drop in the markets in the first 3 years would see a less than 10% reduction in total income.
A long slow recovery would mean more adjustments along the line but then in that situation most strategies would probably be struggling. 

The one final tweak I have considered is calculating my drawdown 1/4ly which will help smooth the ups and downs.

I dont see how being flexible with your expenditure in the face of short term investment volatility can be implemented:....
1) Much expenditure is fixed over the medium term - eg utilities and council tax.  You cant temporarily stop paying them.

2) Economising on expenditure takes time to have a useful effect.  Just turning down the central heating by 1 deg  will have minimal short term effect.  Reducing the quality of the wine you buy is not the answer to a market crash. You will have become used to a particular standard of living, making a substantial change up or down in a short time frame would be difficult.

3) Some expenditure may be fixed months in advance.  eg a cruise may need to be booked a year in advance with a final major payment a few months before departure.  Do you cancel if the market falls in the meantime?

4) As with any policy of responding to short term events, timing is very difficult.  At what point in a market wobble do you change your expenditure?  Take recent events - would you have cut expenditure when Trump came into office or would you still be wondering whether to cut it now? Similarly how do you decide when things have returned to normal?  

In my view a retiree largely dependent on investment income has to decouple expenditure from the release of capital in the short/medium term.  This implies some form of buffering, particularly for large one-off expenditures, and/or some other sources of ongoing income than selling capital to cover ongoing fixed expenditure.

I dont think annuities are a satisfactory complete answer because of their inflexibility.

Your first 4 points are difficult to argue with.

Your point I have taken the liberty of placing in bold is, IMV a critical one. Assuming other sources of income (e.g., state pension and DB pension) are insufficient to maintain an acceptable standard of living then there are several methods of reducing the coupling of portfolio income from market fluctuations
1) RPI annuities (which as you say are inflexible, but currently allow a strong baseline of income).
2) Inflation linked gilt ladders. As for annuities, with, potentially, more flexibility.
3) Increase the quantity of fixed income in the portfolio since, provided the duration is selected carefully*, this is likely to decrease volatility.
4) Use a 'smoothed' percentage of portfolio withdrawal method (Vanguard dynamic, Carlson's endowment, Guyton-Klinger, etc.)
5) Try to implement a dynamic asset allocation strategy (of which cash buffers and buckets are examples) -  McClung's free chapter (see https://livingoffyourmoney.com/wp-content/uploads/2016/05/LivingOffYourOwnMoney_eBook_FirstThreeChapters.pdf ) looks at a bucket strategy.

*Historical back testing has shown that 'intermediate' duration bond funds form a useful compromise since they have lower volatility than longer duration bonds and, generally, higher returns than short duration fixed income (inc 'cash'). 'Intermediate' duration falls around the 5 year mark (although it is slightly lower for US than the UK).

1) Yes IL annuities may be useful for safeguarding essential ongoing income.  However they on their own are not a complete solution. Apart from their inflexibility when circumstances change you also have the problem of major one-off expenditure. 

2) with gilt ladders you gain some flexibility over annuities but on the other hand lose the advantages of longevity insurance.  They are therefore a more expensive solution to retitrement income than an annuity since you need to deliberately plan for a longer than average lifetime. Furthermore they also dont readily support major one-offs. 

3) I would agree that generating ongoing income from a portfolio rather than just selling capital is important..  It is essential if you are going to maintain a steady standard of living.  This income can also come from equity dividends.  The problem with only using fixed interest is inflation. That is why I run a growth portfolio alongside the income one.

4) I do not understand how Guyton Klinger and similar schemes work in the real world.  Yes you can cut income with say a month's notice.  How do you cut expediture quickly without severe disruption to your day-to-day life?

5) Yes I agree you need to split your total portfolio.  Both dealing with inflation and enabling a mixture of steady expenditure and planned major one-offs independent of market prices are incompatible.  You need separate portfolios structured to meet the different objectives.  When you split the portfolios volatility ceases to be a major problem. The ongoing income from one portfolio is unaffected by capital volatility and long term growth from the second portfolio is not compromised by the short term volatility.  Both can be largely ignored.

In principle I believe that back testing is of limited use anyway since we have no idea of how representative the past 100 years are of the infinite universe of possible economic outcomes.  However the back testing I have seen simply compares say a 100/0 portfolio with a 60/40 one, presumably with an assumption of ongoing rebalancing. What I am talking about is entirely separate portfolios, one for income and the other for growth alongside a cash fund primarily intended to support major one-offs.