Drawdown: safe withdrawal rates

edinburgher

2. Use Guyton's sequence of return risk taming (adds about 1% of pot size to withdrawal rate)

Having read through the materials provided and some more recent documents on similar themes by Pfau etc., I am none the wiser as to how sequence of returns risk taming can be applied practically to a model portfolio.

It seems easy enough to figure this out for a single asset class, but how can you/can you realistically calculate whether a fund such as one of the Life Strategy products is collectively overvalued against something like P/E 10? Something like calculate P/E 10 for each then weight by size of the market cap for each region?

jamesd

You'd have to look at how the particular mixed asset fund is invested then use the information for the various investments and markets it's holding. Then work out how to use other holdings to achieve the intended allocation. If you limit how much you hold in such funds you would be able to do the asset matching with the rest.

In the same way you can use something like a global equity tracker as a core fund and vary the allocations between countries with the rest.

What you shouldn't do is try to work out an average PE10 for the whole of a mixed asset fund. PE10's vary between different markets and asset classes at the same time and it's those underlying assets that need to have their allocations changed.

edinburgher

So it might be worth considering creating my own clone 'fund of funds' from low cost regional funds (with the side benefit of lower total costs) as an aid to reducing exposure to more overvalued regions? :think:

jamesd

Right. And as you say, you may be able to get that at lower total cost than the somewhat more expensive blended fund.

I've done something similar in the past myself going for regional and national investments to emulate a global equity investment at lower total cost with a close match to the investment mixture.

westv

With cfiresim is my understanding that the "individual dips" section of the results shows the lowest balance your portfolio could reduce to correct?

gallygirl

westv wrote: »

With cfiresim is my understanding that the "individual dips" section of the results shows the lowest balance your portfolio could reduce to correct?

That's my understanding as well.

michaels

Thanks JamesD, extremely useful.

2 questions.
How does cfiresim handle the uprating of state pension, it is currently say 8k but should I put in the likely cash value when it comes payable 5-10 years after my retirement date?
If deferring state pension makes sense should we include the deferral and uprating (how much?) in the model?

jamesd

for income streams you can choose to have the money inflation linked or not. Just make state pension inflation linked and it'll take care of most of the normal increases, though not the triple lock.

When deferring the state pension set the starting year to the year you start to take the pension after deferring. Set the amount to the level it will be at after deferring, including the increase due to deferral.

westv

I actually find deferring the state pension reduces our initial withdrawal rate but that must be due to when the various DB pensions come on stream.

AnotherJoe

westv wrote: »

With cfiresim is my understanding that the "individual dips" section of the results shows the lowest balance your portfolio could reduce to correct?

Nearly. But "could" implies definitely.

It shows the lowest balance it would reach tested against the past 100 years or so stock market performance. It's possible that results in the future might be worse.