How much of an ARF should be in equities?

[LDFerguson]

Whilst I have you on the line, as it were, Liam! - anything else to add (solutions wise!) on de subject?!

I will admit that I haven't read all five pages of this thread. But let's go back to your example above of someone who has €600,000, takes the €150,000 tax-free lump sum, lives on it for as long as s/he can before touching the ARF, keeps the €150,000 in cash and invests the other €450,000 in equities etc., without worrying about sequence risk. (I know that this is not Sarenco. I guess this is Sarenco 2 - The Sequel. And like so many sequels it's not as good as the original. The original has €600,000 in the ARF. The sequel only has €450,000.)

Anyway, Sarenco has pointed out that he uses the tax-free cash to pay off the mortgage. So he can't rely on the tax-free lump sum for income. Perhaps it has been mentioned already but couldn't he invest 25% of the ARF assets (or any other percentage) into cash or other relatively safe, liquid assets and draw the mandatory income from that? Invest the other 75% in equities without worrying about sequence risk.

 

[Duke of Marmalade]

I attach the spreadsheet to test the sarenco conjecture. It is cruder than the ARF Projection and ignores complications such as tax and mandatory withdrawal. However it does have Macros and these may be rejected by your system.

If you enter the initial cash fund the initial equity fund will be the balance of €1M. Press Ctrl + s to run the 100 simulations. Two methods of withdrawal strategy are modelled. Method 1 withdraws cash first whilst the alternative would withdraw from cash and equities in equal proportions.

 

[Sarenco]

Perhaps it has been mentioned already but couldn't he invest 25% of the ARF assets (or any other percentage) into cash or other relatively safe, liquid assets and draw the mandatory income from that? Invest the other 75% in equities without worrying about sequence risk.

That's pretty close to what I'm suggesting Liam, except that I use a multiple of residual expenses, rather than a crude %, to determine the proportion of the ARF to be held in "safe" assets.

I'm proposing that the equity allocation be allowed to gradually drift higher (as safer assets are withdrawn), subject to a ceiling, but in some ways that's a detail. Retaining a static allocation would be another reasonable approach.

 

[Colm Fagan]

I haven't been following the more recent posts, but surely the allocation to equities is dynamic, depending on past performance at any point in time. In my own case, performance in the early years was good, so I was able to be braver, knowing that, even if the value were to fall (by x%), I should still be able to derive an adequate income. I'm now 100% in equities. Maybe I would have arrived there anyway, given my confidence in the Equity Risk Premium, but my decision was helped by past results (achieved when the allocation was less than 100% to equities).

 

[Duke of Marmalade]

I haven't been following the more recent posts, but surely the allocation to equities is dynamic, depending on past performance at any point in time.

Absolutely. The strategy should be to optimise expected drawdown subject to some measure of safety. At regular points the drawdown would be recalculated based on the portfolio's performance and an update of the safety net. What sarenco raises is the likelihood that the optimal strategy will involve drawing down cash first but obviously this cannot be the whole story as we surely don't want 90 year olds 100% in equities. In a sense Colm, you have been following the sarenco conjecture as the last port of call for your required living expenses seems to be actual decumulation.

 

[Colm Fagan]

Thanks Duke. Another observation is that most draw-down models assume that the amount required for "living expenses" each year is independent of the fund's performance. I think that's too simplistic. Assuming that the recipient is living above the breadline, there is some ability to flex each year's income, depending on performance, in much the same way as someone in employment or running their own business will have good years and bad years. In the good years, they can splurge a little, in the bad years they have to tighten their belt. It's the same post retirement.

 

[LDFerguson]

This is perhaps veering a little off the main road of this thread, but when I'm chatting with someone who is actually about to set up an ARF, there are a number of fundamental questions to ask. One of them will have quite an impact on this thread and indeed Sarenco's strategy, I think.

Is it important to you to leave behind a residual fund to your kids after your death?

Broadly I get two answers to this: -

(1) Yes. I want to try to draw only the investment growth from the ARF, thus making it a perpetual ARF which will be worth at least as much when I die as it is now, if not more. If possible, I don't want to eat into the original capital.

(2) Hell no. Those kids have got more than enough from me. I fed them, clothed them, schooled them, gave them a college education etc. They'll get the house, the holiday home and the priceless collection of porcelain wombats when I go, so the ARF is going to be spent by me before I die. (Another version of this answer is simply "I have no kids".)

If one is going for answer number (2) then there's an argument in favour of really living it up in the early years of the ARF, while you're still young and healthy enough to do all the things you promised yourself, even though this will involve eating into your original ARF capital. Perhaps in the Sarenco model, this answer would dictate a higher allocation to cash than answer number (1).

 

[Bob2018]

Hi Liam,

Interesting comments. The purpose of the thread, as I understand it, is to assume that the ARF is needed to provide income in retirement for the normal person and how best to achieve this. You might have to read all the back pages!

I know it's my own fault for asking the Duke to share his work with us but I must admit to be struggling with lognormal, etc. Up until this point in my life, this would have meant something that I'd lob into the fire which didn't have any remarkable features.

 

[Duke of Marmalade]

I know it's my own fault for asking the Duke to share his work with us but I must admit to be struggling with lognormal, etc. Up until this point in my life, this would have meant something that I'd lob into the fire which didn't have any remarkable features.

Correction, I provided the spreadsheet in response to a request from elac. I hope you are not struggling with the illustrations which were designed to be comprehensible by the most numerically challenged of folk. But perhaps you are the sort of guy that doesn't drive a car because you can't understand how the internal combustion engine works. Perhaps your brother can explain the lognormal to you

Bob I have to be honest, I find your contributions a tad confusing, even to the point that I suspect you of being a troll, a suspicion reinforced by your recent membership. I don't like trolls and so to be on the safe side I will be making no further contributions to this thread.

 

[Bob2018]

I know it's my own fault for asking the Duke to share his work with us but I must admit to be struggling with lognormal, etc. Up until this point in my life, this would have meant something that I'd lob into the fire which didn't have any remarkable features.

I think this post is very unfair. I did ask you to share your work (the Holy Grail) with us (we even joked about it). The other joke about lognormal was at my expense.

 

[Duke of Marmalade]

I think this post is very unfair. I did ask you to share your work (the Holy Grail) with us (we even joked about it). The other joke about lognormal was at my expense.

Okay Bob, I guess I just haven't got used to your style yet. Anyway the thread had run its course so far as I was concerned.