SWR Come What May

FIREDreamer

OldScientist said:

zagfles said:

You have a point but the example is ridiculously exaggerated. Dynamic asset allocation can help with SORR, personally I'll be using Prime Harvesting, where IIRC the worst case historical scenario you won't be selling equities for a decade from a 60/40 equities/bonds starting point. 

There's no point having a strategy if you don't trust it. Obviously if something totally unprecedented happened, something that has never happened before, you may need to rethink. But changing strategy in a panic because of eg a market crash on a similar scale to ones we've seen several times in recent decades and which the "SWR" has been measured to cope with is no different to those who sell in a panic or stop investing whenever the market goes down. 

The example can (and should for those wanting to see what happens 'if') be extended to a more sensible number of years than five (I wanted to keep the tables short, but illustrate the problem). A 50% reduction in real portfolio value is rare but not unprecedented, while persistent small or negative real returns are less dramatic, but, in the long run, equally problematic. According to McClung's own figures, Prime Harvesting either improved (by 40 basis points) or slightly decreased the SWR (10 bp) for US retirees (depending on what dataset was used), and improved it for UK (70 bp) and Japanese (50 bp) retirees. I also note it is possible for the retiree to end up with 100% equities if the bonds have been spent down and the equity returns have been insufficient to trigger a sale.

I'd agree that trust in the method is important. But I would contend that SWR is not trustworthy unless using very small withdrawal rates since while historical values are known (e.g., 3.0 to 3.5% for UK retirees) future values are not. For example, in the following graph the real value of the portfolio (upper panel) and the real withdrawal (lower panel) is shown for a UK retiree with a 60/20/20 (stocks/long bonds/cash - returns and inflation are from macrohistory.net) starting in 1937. I chose a conservative value of 3% for thew withdrawal rate.



The question you have to ask, is when would you decide to ditch SWR or otherwise panic? 10 years in when the portfolio has fallen, in real terms, to less than half of its initial value, after 15 years when it fell to just over 20% of the initial value, or after 20 years when it had fallen to about 15%? 

For anyone interested in real panic, then a search for 'Plan B' on the bogleheads forums during 2008-2009 is illuminating (in essence Plan B was getting out of equities and buying bonds during the prolonged and relentless market falls). This was directly contrary to the bogleheads stated aim of 'stay the course' and provoked heated discussion.

Plan B - buy an inflation linked annuity especially if the rate is nearer 4% as you get the same income but will never run out of money. Quite feasible now - something I did with 70% of my pot late last year which enabled me to have the courage to retire this summer.

Sarahspangles

FIREDreamer said:

Plan B - buy an inflation linked annuity especially if the rate is nearer 4% as you get the same income but will never run out of money. Quite feasible now - something I did with 70% of my pot late last year which enabled me to have the courage to retire this summer.

This also has the advantage that you don’t need to be hands on later in retirement with monitoring and decision making. From experience with two sets of parents, you don’t need to have dementia to see some decline in mental functioning.

zagfles

OldScientist said:

zagfles said:

You have a point but the example is ridiculously exaggerated. Dynamic asset allocation can help with SORR, personally I'll be using Prime Harvesting, where IIRC the worst case historical scenario you won't be selling equities for a decade from a 60/40 equities/bonds starting point. 

There's no point having a strategy if you don't trust it. Obviously if something totally unprecedented happened, something that has never happened before, you may need to rethink. But changing strategy in a panic because of eg a market crash on a similar scale to ones we've seen several times in recent decades and which the "SWR" has been measured to cope with is no different to those who sell in a panic or stop investing whenever the market goes down. 

The example can (and should for those wanting to see what happens 'if') be extended to a more sensible number of years than five (I wanted to keep the tables short, but illustrate the problem). A 50% reduction in real portfolio value is rare but not unprecedented, while persistent small or negative real returns are less dramatic, but, in the long run, equally problematic. According to McClung's own figures, Prime Harvesting either improved (by 40 basis points) or slightly decreased the SWR (10 bp) for US retirees (depending on what dataset was used), and improved it for UK (70 bp) and Japanese (50 bp) retirees. I also note it is possible for the retiree to end up with 100% equities if the bonds have been spent down and the equity returns have been insufficient to trigger a sale.

I'd agree that trust in the method is important. But I would contend that SWR is not trustworthy unless using very small withdrawal rates since while historical values are known (e.g., 3.0 to 3.5% for UK retirees) future values are not. For example, in the following graph the real value of the portfolio (upper panel) and the real withdrawal (lower panel) is shown for a UK retiree with a 60/20/20 (stocks/long bonds/cash - returns and inflation are from macrohistory.net) starting in 1937. I chose a conservative value of 3% for thew withdrawal rate.



The question you have to ask, is when would you decide to ditch SWR or otherwise panic? 10 years in when the portfolio has fallen, in real terms, to less than half of its initial value, after 15 years when it fell to just over 20% of the initial value, or after 20 years when it had fallen to about 15%? 

For anyone interested in real panic, then a search for 'Plan B' on the bogleheads forums during 2008-2009 is illuminating (in essence Plan B was getting out of equities and buying bonds during the prolonged and relentless market falls). This was directly contrary to the bogleheads stated aim of 'stay the course' and provoked heated discussion.

Yes PH could end you in 100% equities. If that would cause you psychological issues then it's obviously not the strategy for you. I presume the above is based on fixed asset allocation rather than PH. It would be interesting to see what would happen with PH in this case. 

But anyway, if you retired in 1937 then I think you'd have had far bigger things to panic about than your pension running out in 1965! Let's face it, in retirement various horrible things could happen. You might get cancer, heart disease, your partner might, you might become seriously disabled or have health problems that prevent you doing all the things you wanted to do. Or your drawdown pot may run dry and you have to live on the state pension and means tested benefits.

Of those things, the latter is by far the least bad. Particularly if you have some other provision like DB pension. I know I'll be OK, even if I didn't have the DB. It's not something that would ever cause me "panic" even if it did happen. Anyone who can't cope with the risk of running out of money really shouldn't be using drawdown. What's the point of drawing 3% or less just to cater for the worst case when you can buy an index linked annuity paying 4.5% at 65.

The usual stuff people come out with are "what if I die young, the insurance company just keeps my money!" (err, no they don't, they use it to cross subsidise those who live to 100) or "I want to leave the kids an inheritance" (what, so they need a big inheritance if you die at 70 but not if you live to 100?)

NedS

Sarahspangles said:

FIREDreamer said:

Plan B - buy an inflation linked annuity especially if the rate is nearer 4% as you get the same income but will never run out of money. Quite feasible now - something I did with 70% of my pot late last year which enabled me to have the courage to retire this summer.

This also has the advantage that you don’t need to be hands on later in retirement with monitoring and decision making. From experience with two sets of parents, you don’t need to have dementia to see some decline in mental functioning.

This for me is a massive concern for my children (the next generation), who are unlikely to have substantial index-linked DB pensions in retirement. How reasonable is relying on drawdown once one inevitably reaches an age where it becomes unrealistic for them to manage (even if they were capable when younger, which is not a given).

SouthCoastBoy

NedS said:

Sarahspangles said:

FIREDreamer said:

Plan B - buy an inflation linked annuity especially if the rate is nearer 4% as you get the same income but will never run out of money. Quite feasible now - something I did with 70% of my pot late last year which enabled me to have the courage to retire this summer.

This also has the advantage that you don’t need to be hands on later in retirement with monitoring and decision making. From experience with two sets of parents, you don’t need to have dementia to see some decline in mental functioning.

This for me is a massive concern for my children (the next generation), who are unlikely to have substantial index-linked DB pensions in retirement. How reasonable is relying on drawdown once one inevitably reaches an age where it becomes unrealistic for them to manage (even if they were capable when younger, which is not a given).

To counter that issue, if I'm still around I expect I will buy an annuity in my early 70s to cover basic living e.g maybe 15k per year in today's money. 

Linton

SWR is fine for sanity checking a strategy but to actually rely on it during retirement is in my view madness...,

1) The idea is that if your strategy successfuly survives 70 30-year runs in the past 100 years it really must be safe.  But the problem is that these runs are highly correlated.  For example the GFC occurred some 8 years after the Tech crash.  No runs explore what would happen had it been 6 years.   We know what happened with 2 world wars.  What would the results have been for 3 or 1 world war?  The test data covers an infinitesimally small % of the situations that could occur.

2) When you start retirement you have 70 examples to start form.  WIthin a few years the economic situation is unlikely to match any of them.  To rely on SWR would be like flying blind with only a map to guide you.  If you dont know where you are and do not react accordingly your flight wont last long.

barnstar2077

I don't intend to use a SWR, but will be using drawdown.  What are people's thoughts on the number they will use as inflation?

I am not going to just take how much I withdrew last year (in retirement) plus what google tells me inflation is.  I will probably use the percentage my bills go up every year as that years guide to what inflation means to me.  I think Increasing my withdrawals by 11% because that is the number I saw on the news etc would end up with me having too much or too little each year.

I assume this is how everyone else is, or will be, doing it too?

Hoenir

Sea_Shell said:

We now have £100k+ more than we started with 😇

We've experienced an era of free/cheap money. Going to be a price to pay. 

Linton

barnstar2077 said:

I don't intend to use a SWR, but will be using drawdown.  What are people's thoughts on the number they will use as inflation?

I am not going to just take how much I withdrew last year (in retirement) plus what google tells me inflation is.  I will probably use the percentage my bills go up every year as that years guide to what inflation means to me.  I think Increasing my withdrawals by 11% because that is the number I saw on the news etc would end up with me having too much or too little each year.

I assume this is how everyone else is, or will be, doing it too?

I simply take what I need/want independent of inflation rates or returns or any other factor. The plan was based on my standard of living remaining constant at the level it was when I retired 19 years ago and there has never been any need to change it.  When the assets have built up then that may give the opportunity for an expensive holiday or other one-off expenditure.

Never being forced to cut expenditure was a major requirement in designing the investment strategy.  

Linton

Grumpy_chap said:

[Deleted User] said:

I'm not going to use my thumb to measure anything important.

You would have to if you wanted to know what a thumb's width was

On a serious note, why wouldn't someone do this?

[Deleted User] said:

 living off their Safe Withdrawal Rate like it's set in stone. 

There is nothing brave in living off a SWR.
It is not a SAFE withdrawal rate otherwise.
Now, whether the SWR is correctly set is another matter entirely.

FWIW, I always think that the SWR percentages generally mentioned are too conservative as the typical values referenced (4% or thereabouts) seem to be to preserve the original capital value and only spend growth.  I will be perfectly happy to spend the original capital value over the period of retirement.  At 4% withdrawal rate, growth could be zero and I still last 25 years...

Almost all of the time I believe you are correct in saying that normally quoted SWR rates are too low.  But the problem is that SWR simulations tend to end up eventually in you either going bust or ending up with far more money than you started, with few in between.  To what extent are you willing to take the risk?

My approach would be to start off with drawdown around the SWR level and allow it to steadily increase as funds build up.