If you are an SWR purist, when does your retirement start?

michaels

If my retirement had started when I handed in my notice 3 months ago then my SWR would be about 5k pa higher than it is now I am about to hit my retirement date.

To me this points up a pretty big flaw with a pure SWR approach.

dunstonh

To me this points up a pretty big flaw with a pure SWR approach.

Not really.  Everyone has a starting point and anything is possible from that starting point.   The SVR is meant to get you close to the worst outcome.

If you started at 3.5% SVR 3 months ago, then large drops (larger than what we have had) would be taken into account on that higher value.    If you start it now, then those same drops would still be taken into account on the lower value and lower draw  

I just had a little play to get close to running out to see the range of outcomes on a 50% equities portfolio:



The red line is the worst case at 3.61% withdrawal rate, and it had a 99% success rate using 1915-2024. 

So, it's less likely that you would suffer the worst-case scenario after a recent drop, but the whole point of the SVR is that it's the worst-case scenario.     Its not a guide to what is likely.  Its a guide to what could be the worst.  You would have to be statistically unlucky to get that worst case but look at just how wide the range of outcomes has been.

Triumph13

If you want to look at it another way, SWR's are defined based on the worst period in history.  Pretty much by definition, the worst period will always have started from a peak, given the disproportionate impact of bad returns early on.  It's therefore probably not unreasonable to regard past history as giving support for the figures when you gave notice rather than the figures now.

OTOH of course, the SORR alarm bells are ringing pretty damn loud at the moment, so I know I wouldn't have the cojones myself to use the higher figure.

michaels

But you have to believe in efficient markets 100% and that market history provides no information about possible future paths.  Analysis I have seen says if you only start take periods where the stock market is at least 10% of its peak then you get a materially higher SWR - of course this is not a pure SWR

However, personally I can't see that sliding doors me who retired 3 months ago would really have a safe lifetime spend 5% higher than real me who retires at the end of next week.

Hoenir

When people are speculating rather than investing. Markets do become inefficient.  Viewing (often manufactured) historical data through the prism of today is totally misleading on so many counts.  

Bostonerimus1

The retirement simulations that produce index linked withdrawal numbers like 3% or 4% generally assume a 30 year retirement and a 5% failure rate. Those failures will often come from scenarios where there are big losses early in retirement. It's always good to remember that we will only ever experience one particular sequence of market returns not the thousands of manufacture sets of returns inherent in Monte Carlo simulations. I think the interest in annuities will increase in the near term and the SWR purist will become a little more rare. I know a guy who bought an annuity 3 months ago...he's pretty happy about that right now.

MK62

michaels said:

If my retirement had started when I handed in my notice 3 months ago then my SWR would be about 5k pa higher than it is now I am about to hit my retirement date.

To me this points up a pretty big flaw with a pure SWR approach.

It's not really a flaw as such......pretty much any portfolio with a sizable chunk invested in equities will show a similar pattern, even if the numbers don't match exactly. The global equity market is down about 10% over the last 3 months, and some equity funds a lot more than that (obv it depends on exactly what your equity investments are), so crunching the numbers now, compared to 3 months ago is going to show a lower starting withdrawal with pretty much any drawdown plan.

PS - personally I'm not really a big believer in the "pure" SWR drawdown plan......in my view it's basically gambling that the future will be no worse than the worst period in the past......it might well not be, but nobody knows!

westv

MK62 said:

michaels said:

If my retirement had started when I handed in my notice 3 months ago then my SWR would be about 5k pa higher than it is now I am about to hit my retirement date.

To me this points up a pretty big flaw with a pure SWR approach.

It's not really a flaw as such......pretty much any portfolio with a sizable chunk invested in equities will show a similar pattern, even if the numbers don't match exactly. The global equity market is down about 10% over the last 3 months, and some equity funds a lot more than that (obv it depends on exactly what your equity investments are), so crunching the numbers now, compared to 3 months ago is going to show a lower starting withdrawal with pretty much any drawdown plan.

PS - personally I'm not really a big believer in the "pure" SWR drawdown plan......in my view it's basically gambling that the future will be no worse than the worst period in the past......it might well not be, but nobody knows!

If the future is worse than the worst period in the past then we are all in big trouble!

QrizB

westv said:

MK62 said:

......in my view it's basically gambling that the future will be no worse than the worst period in the past......it might well not be, but nobody knows!

If the future is worse than the worst period in the past then we are all in big trouble!

Immediately before the current "worst period in the past", it hadn't happened.

There is (hopefully) a lot of future for the current world financial system, and there is relatively little documented past.

It's therefore almost certain that there will be a period in the future that is worse than any previous period.

michaels

MK62 said:

michaels said:

If my retirement had started when I handed in my notice 3 months ago then my SWR would be about 5k pa higher than it is now I am about to hit my retirement date.

To me this points up a pretty big flaw with a pure SWR approach.

It's not really a flaw as such......pretty much any portfolio with a sizable chunk invested in equities will show a similar pattern, even if the numbers don't match exactly. The global equity market is down about 10% over the last 3 months, and some equity funds a lot more than that (obv it depends on exactly what your equity investments are), so crunching the numbers now, compared to 3 months ago is going to show a lower starting withdrawal with pretty much any drawdown plan.

PS - personally I'm not really a big believer in the "pure" SWR drawdown plan......in my view it's basically gambling that the future will be no worse than the worst period in the past......it might well not be, but nobody knows!

But had I retired 3 months ago and was wedded to SWR then Alt-me would be happy pulling my 65k pa for ever after despite what has happened to the markets whereas me-me who retires next Friday will instead have to be happy drawing 60k.  I would say that realistically my risk of failure is lower than Alt-me's.  It is only SWR purists who would say otherwise.

OldScientist

westv said:

MK62 said:

michaels said:

If my retirement had started when I handed in my notice 3 months ago then my SWR would be about 5k pa higher than it is now I am about to hit my retirement date.

To me this points up a pretty big flaw with a pure SWR approach.

It's not really a flaw as such......pretty much any portfolio with a sizable chunk invested in equities will show a similar pattern, even if the numbers don't match exactly. The global equity market is down about 10% over the last 3 months, and some equity funds a lot more than that (obv it depends on exactly what your equity investments are), so crunching the numbers now, compared to 3 months ago is going to show a lower starting withdrawal with pretty much any drawdown plan.

PS - personally I'm not really a big believer in the "pure" SWR drawdown plan......in my view it's basically gambling that the future will be no worse than the worst period in the past......it might well not be, but nobody knows!

If the future is worse than the worst period in the past then we are all in big trouble!

The worst case in the past depends on a) whose past (e.g., for 30 year retirements there is a difference between the well-known '4% rule' for the US and the 3.0 %to 3.5% 'rule' for the UK and b) the data used (e.g., even for the well-researched US, there is a difference in SWR between different data sets, for different durations of fixed income and with the resolution of the data - monthly resolution gives a lower SWR than annual resolution).

Notwithstanding SORR, even small changes in the annualised real returns make a difference to how long the portfolio lasts. For example, with a annualised returns of 0% real, a withdrawal of 3.33% would last 30 years, +1% it would last 35 years, -1% it would last 26 years (using the nper function). In other words, just a +/- 1% change in annualised real returns results in a change in portfolio longevity of about +/- 5 years.

While SWR might be good for planning, it is not a terribly robust practical approach.