4% safe withdrawal rate?

[Colm Fagan]

For a lot of people, loss aversion is a bigger driver and minimising their losses is more important than maximising the gains

Steven. I agree that loss aversion is a major reason why savers of all ages put less in equities than they should and that it is more acute in older people for the reasons you mentioned.

It is a major hurdle to be overcome by anyone advising people where to put their money. It is especially intractable because loss aversion is hard-wired into us by evolution. (He who fights and runs away will live to fight another day - and to procreate). Unfortunately, some advisers are even more risk-averse than their clients and exacerbate the problem. From what I hear, you're not one of them. Fear of loss on the part of the adviser is perfectly understandable. If I were an adviser, I would be terrified that the client could lose money in the first month. There is close to a 50:50 chance that they will, irrespective of where markets are at the time I'm advising them. That could make me look stupid. Daniel Kahneman, who won a Nobel Prize for his work in behavioural economics, recognised this when he wrote that ".. decision makers who expect to have their decisions scrutinised with hindsight are driven .. to extreme reluctance to take risks.”

My proposals for smoothing returns over many years for Group ARF's and auto-enrolment were partly designed to address the problem of loss aversion.

 

[Colm Fagan]

While retirees do indeed take out less when the value decreases, there is a minimum income required too

The minimum income requirement isn't as much of a constraint on investment freedom as it is sometimes made out to be. The minimum income requirement in any month is (say) 1/12th of 4% of the value of the fund. That's only 0.3% of the fund value. If the pension account is on the ropes at that time, it's not the end of the world. Okay, it's a bigger problem if the downturn lasts for (say) two years, but even then the likelihood is that less than 10% of the fund has been permanently impaired, to borrow a phrase from the accountants.

 

[Sarenco]

Okay, it's a bigger problem if the downturn lasts for (say) two years...

What happens if stocks take 5 years to recover their value? Or 10 years? Or 25 years?

What would an all equity portfolio look like if you had retired in 1928 and drew down @4%pa? Or 1964? Or 1999?

I know I’ll need to eat - and sleep - in retirement whatever happens so my modest portfolio won’t be anything remotely like 100% in equities.

 

[Marc]

It would look like this ( in each of these graphs I am taking an income of 4% of the initial value of the pension) not adapting to market conditions

 

[Marc]

I think on reflection you meant to say 1974

 

[Marc]

 

[Marc]

I changed the last graph to Euro base currency, MSCI World Index and starting in December 2000.

 

[Marc]

Last one, this looks at 20:20 hindsight of getting an all equity portfolio spectacularly wrong compared to short term T bills and short-dated US Bonds. Funny how the really daft portfolio (bonds) does really well when the equity markets are going to hell in a hand-basket. Maybe there is something in this diversification lark after all

 

[Marc]

For the sake of completeness, if you had waited until May 1955 you would have got back to where you stated. So if you had been 65 in 1929 you would have been 94 when you returned to par but obviously still way behind a less risky portfolio.

 

[Duke of Marmalade]

Good stuff Marc!

For a lot of people, loss aversion is a bigger driver and minimising their losses is more important than maximising the gains.

Absolutely, but I want to challenge any suggestion (I am not saying you are making it, Steve) that this is irrational and that there is a role for "behavioural" advisors to protect folk from themselves.
Let's move along the wealth spectrum.
Say one's only income is State Basic Pension and Eddie Hobbs as Renua MoF decides to buy them all out at commercial annuity terms. It would be a brave advisor who would recommend 100% equities as you are risking dire poverty versus modest improvements in lifestyle.
Next think of an individual who can say just about afford a car and an annual holiday. They are risking no car or no holiday versus a better car or more exotic holiday.
Next an individual who is quite comfortable and has an expectation that they will leave some inheritance. They are risking having to downsize the car or the holiday versus more inheritance for the kids.
Next, someone who anticipates that their estate will be paying significant CAT. They will feel they can weather any reasonably foreseeable downturn in equities without affecting their lifestyle but on the other hand the upside is going not only to their estate but also to the taxman.
It is really only when you get to the Donald Trump end of the spectrum that you are indifferent as to the relative value of gains and losses, so for The Donald the ERP is truly a free lunch.

 

[elacsaplau]

Excellent post Duke. I, very literally, could not have said better myself.

This thread was revised by Colm a few days ago. One of the problems that I have with Colm's post(s) generally is that he tends to segue from the specific (his own experience) to the general (the right general approach) without adequate consideration and/or caveating of the implications for different groups. My concern, in this instance, is that the cohorts that the Duke so eloquently described could get mushed in the all in equity crossfire!

By the way, Elacsaplau's first lau (law)....
As an on-line discussion thread lengthens, the probability of The Donald being mentioned approaches 1.

 

[Steven Barrett]

The minimum income requirement isn't as much of a constraint on investment freedom as it is sometimes made out to be. The minimum income requirement in any month is (say) 1/12th of 4% of the value of the fund. That's only 0.3% of the fund value. If the pension account is on the ropes at that time, it's not the end of the world. Okay, it's a bigger problem if the downturn lasts for (say) two years, but even then the likelihood is that less than 10% of the fund has been permanently impaired, to borrow a phrase from the accountants.

Or as I put it to clients, say you have €500,000 in an ARF and you get 4% a year, €20,000. If the ARF falls by 20% to €400,000 (always have to tell them the actual value, not the percentage ),your income will fall to €16,000. Can you still fund your lifestyle or do you have other savings to supplement your ARF income?

But yes, the length of the downturn will always be an issue, especially as none of us know what the next downturn will be or how bad it will be. But investing in quality companies should minimise the risk of permanent impairment. If Microsoft, Johnson & Johnson etc all go bust, we have a lot more problems on our hands than pension income!

Am currently reading Thinking Fast & Slow. A book that should be read like you'd eat an elephant...one bite at a time


Steven
http://www.bluewaterfp.ie (www.bluewaterfp.ie)

 

[Steven Barrett]

Good stuff Marc!


Absolutely, but I want to challenge any suggestion (I am not saying you are making it, Steve) that this is irrational and that there is a role for "behavioural" advisors to protect folk from themselves.
Let's move along the wealth spectrum.
Say one's only income is State Basic Pension and Eddie Hobbs as Renua MoF decides to buy them all out at commercial annuity terms. It would be a brave advisor who would recommend 100% equities as you are risking dire poverty versus modest improvements in lifestyle.
Next think of an individual who can say just about afford a car and an annual holiday. They are risking no car or no holiday versus a better car or more exotic holiday.
Next an individual who is quite comfortable and has an expectation that they will leave some inheritance. They are risking having to downsize the car or the holiday versus more inheritance for the kids.
Next, someone who anticipates that their estate will be paying significant CAT. They will feel they can weather any reasonably foreseeable downturn in equities without affecting their lifestyle but on the other hand the upside is going not only to their estate but also to the taxman.
It is really only when you get to the Donald Trump end of the spectrum that you are indifferent as to the relative value of gains and losses, so for The Donald the ERP is truly a free lunch.

It's not irrational at all!! That is why capacity for loss should be the biggest factor in advising on any investment strategy for a client, which is exactly what you have outlined in your post.

I see a lot of cases where a client is told to complete a risk profiling questionnaire and it spits out a fund strategy at the end. It's impossible to advise properly on just a risk profiling questionnaire. While I do use one, it has less and less importance with the advice I give as talking to clients about the circumstances and needs take more importance.


Steven
http://www.bluewaterfp.ie (www.bluewaterfp.ie)

 

[Colm Fagan]

It is really only when you get to the Donald Trump end of the spectrum that you are indifferent as to the relative value of gains and losses, so for The Donald the ERP is truly a free lunch.

Duke, the smoothing approach, by spreading losses and gains across generations, means everyone can avail of the free lunch.

What would an all equity portfolio look like if you had retired in 1928 and drew down @4%pa? Or 1964? Or 1999?

I don't have monthly figures for 1928 and 1964, but I do have them for 1999. The results are interesting.

First of all, 31 December 1999 marked the last day (literally) of the dot-com boom. Markets had more than doubled over the previous four years: +17% in 1996, +24% in 1997, +14% in 1998 and +24% in 1999. Investors paid the price over the following few years: -6% in 2000, -13% in 2001 and -23% in 2002. Then, just when people thought they were out of the woods after a few good years, the market fell 30% in 2008.

It really is a challenge to see how an ARF investor who was stupid enough to put all their money into the market at the very peak of the boom on 31 December 1999 and withdrew 4% each year (of their original investment, not of the value from time to time) would have done.

Remarkably, the answer (to end 2017 - I didn't input 2018 figures) isn't nearly as bad as one might think. In the 18 years to end 2017, someone who invested €100,000 on 31 December 1999 would have withdrawn €72,000 in "income" (18 years at 4% a year) and their account would be worth €85,000 at the end of 2017. This equates to a compound annual return of 3.4% a year, despite all the adversity this poor sod would have had to endure. Allowing for charges of 0.75% a year, the balance at end 2017 would have been €67,000, equivalent to 2.5% a year.

If they had put their money in a year earlier, at 31 December 1999, they would still be going in on a high - markets having gone up 14% that year, 24% the previous year and 17% the year before that again. Despite that, at end 2017 their original €100,000 would be worth €133,000 (no charges) or €106,000 (charging 0.75% a year) despite having withdrawn €76,000 in the 19 intervening years.

 

[Colm Fagan]

When I revisited this thread last evening (Note to @elacsaplau : Not "revised" . Me revise something? Never!) I wrote the following:

I discovered that the money-weighted return on my retirement savings over the last five years (i.e. the “real” return, based on actual amounts withdrawn each month) is a full 1% a year more than the time-weighted return, i.e. the return assuming that there were no withdrawals and no deposits in the period. I believe that this 1% difference between the real return and the notional return is due in large part to my practice of withdrawing more in good times and less in bad times.

My own personal financial adviser (who gets rewarded with an odd coffee) came back to me privately with the following:
"I did a few doodles on Excel. I am not sure that the discrepancy is in the main down to your withdrawal strategy. The period is characterised by a long bull run followed by a correction. Even with a regular withdrawal rate that would mean the money weighted return would be higher than the time weighted return, as you have less exposure to the correction having made withdrawals."

I confirmed (going from the specific to the general, @elacsaplau : I was always better at arithmetic than algebra) that he's right. I looked at the figures for a 4% withdrawal strategy from end 1998 (results quoted in the previous post) and varied the strategy slightly by hypothesising that the ARF pensioner would only make withdrawals on months when the price had increased, none when the price fell the previous month. Every time he/she made a withdrawal, they would take whatever was due since the previous withdrawal. The results were almost identical to those derived on the assumption of a constant withdrawal each month.

My argument isn't completely dead, though. I really believe that the market is manifestly undervalued or overvalued on the odd occasion. Needless to say, it's hard to incorporate a "manifestly overvalued/undervalued" rule into an algorithm. It played some role in the excess of the money-weighted return over the time-weighted return over the last five years, maybe not the full 1% but possibly 0.5%.

 

[Duke of Marmalade]

My own personal financial adviser (who gets rewarded with an odd coffee)...

Don't you think he is worth a proper coffee?

 

[Sarenco]

Remarkably, the answer (to end 2017 - I didn't input 2018 figures) isn't nearly as bad as one might think. In the 18 years to end 2017, someone who invested €100,000 on 31 December 1999 would have withdrawn €72,000 in "income" (18 years at 4% a year) and their account would be worth €85,000 at the end of 2017. This equates to a compound annual return of 3.4% a year, despite all the adversity this poor sod would have had to endure. Allowing for charges of 0.75% a year, the balance at end 2017 would have been €67,000, equivalent to 2.5% a year.

Interesting.

Mind you, if our 2000 retiree had increased his 4% drawdown each year to account for inflation, he would be pretty much broke at this stage.

Adding a reasonable allocation to Government bonds would have improved the picture considerably.

 

[Steven Barrett]

Interesting.

Mind you, if our 2000 retiree had increased his 4% drawdown each year to account for inflation, he would be pretty much broke at this stage.

But in reality that's not what we see. Retirees tend to be very cautious with their ARF, taking out the required amount and not much more. Usually, if they are prudent enough to save for their retirement, they have other savings, which is used to supplement the ARF income (use up the taxed income first before drawing down an income that is liable to income tax). And after the retiree has got past the loads of holidays, giving money away part of retirement, they tend to spend less so there is less need to inflation proof their retirement income.


Steven
http://www.bluewaterfp.ie (www.bluewaterfp.ie)

 

[Sarenco]

Fair enough but the CPI is up nearly 50% since 2000 - that's pretty significant.

Also, imputted distributions rise to 5% at 71....

My main point really is that adding a reasonable allocation to Government bonds can increase the risk-adjusted return of an equity heavy portfolio.

 

[elacsaplau]

Duke, the smoothing approach, by spreading losses and gains across generations, means everyone can avail of the free lunch.

Colm - are we now not segueing from the specific to the general to the theoretical - the thread is about real-world safe withdrawals at an individual level?!