Sequence of Return Risk

Despite everything I've read over the years on retirement income I stick with a pretty simple method that others such as Thornhill, PC'er @truong and many Aussie SMSFs also use. Rather than bore you with more of my ramblings @truong's GFC experience is excellent reading and wonderful education in how to deal with major scary market events.

The following chart graphically illustrates the principle behind it. Note that the chart doesn't include franking credits and assumes all dividend income is consumed:


As a retiree here are my guiding principles:

1. Invest for enough dividends to comfortably meet living expenses, that is one is not dependent on every last cent of income. Having a little left over to reinvest is very beneficial.

2. The aim is to avoid drawing on one's capital for at least the first decade in retirement.

3. Invest in quality LICs where due to the company structure the Mgr can maintain profit reserves to help smooth dividends in bad times. That is if the investor is silly enough not to maintain their own reserves the LIC may potentially compensate for this.

4. Be sensible, in bad times adjust your spending accordingly.

5. Maintain a few years living expenses in a cash buffer. Importantly the cash buffer only needs to "TOP UP" any dividend shortfall that doesn't meet important living expenses. Note that the worst dividend cut in history was 50% during the Great Depression and it was temporary. HENCE A THREE OR SO YEAR CASH BUFFER MAY BE SUFFICIENT TO TOP UP DIVIDENDS FOR A DECADE OR MORE. I'm more conservative so maintain more than a three year cash buffer especially being earlier in retirement.

Finally when reading Safe Withdrawal research be mindful of the country it applies to. For example in Australia remember to consider Taxation (eg franking credits), Retirement investment structures (Super and tax free account based pensions), social security (Age pension) and numerous other factors.

I'll leave it to others to analyse and critique our retirement income approach but it lets us sleep well at night and I'm yet to see anything convincing enough to change what we, and others I know who have been retired a lot longer than us, do. And we're living it NOW for better or worse as opposed to many researchers who are still being paid to live in a world of acadamia!

Personal view only, not liscenced to give advice.

 

For the Kitces fans the Podcast Interview posted earlier by @dunno and linked again below seems to be trending heavily on Bogleheads:

Michael Kitces – The 4% Rule and Financial Planning for Early Retirement

Lots of discussion here:

Michael Kitces 4% rule podcast on Madfientist - Bogleheads.org

 

I more imagined it as a thread about managing sequence risk. The 4% rule was just the first approach to the issue that I’ve investigated. Seems to be lots of conflicting opinions from the ‘experts’ in the literature. My conclusion from applying the rule to Aus Historical Data is that’s it’s not right for me. Not dynamic enough resulting in both a high a probability of either running out of funds or conversely living a meagre life and leaving behind a fortune.

I’m interested in investigating sequence risk under the assumption of 100% equity exposure as I’m not enthusiastic about other asset classes which are the obvious way to dampen equity volatility.

But I think the thread could and should be used to examine all range of possibilities that anybody with an interest might like to explore.

 

This quote from Wade Pfau in your link completly sums up my thoughts on the 4% rule.

Although I also agree with the line in the FINSIA paper that was linked to earlier by Heinz57

making the point that this very simple rule is better than nothing.

I think I'll investigate next - the more dynamic approach of living off dividends. Seems there's lots of enthusiasm for that here. I’ll probably upset some as I like to investigate by trying to debunk ideas.

 

Based on this article, and others like it on the same website, I like to use the inverse Shiller CAPE (or CAEP) as a rough guide to what SWR to use if retiring soon. eg. The Shiller CAPE was 30.31 as of last Friday (see Shiller PE Ratio) of which the inverse puts the current conservative SWR at 3.3%, which I think would be quite prudent given how long the current bull market has run.

 

Starting valuations certainly matter. However I dunno that the inverse of CAPE to set a SWR is a conservative rule of thumb though.

Putting aside potential weaknesses in CAPE as a valuation measure AUS Cape is currently around 17-18 which by your measure sugests current retires setting a SWR of around ~5.5% for Aus Equities. 4% had too many failures for my liking on historical Aus data let alone 5.5%.

But along the lines of the 4% rule is dead - long live the 4% rule - I think as you are suggesting, some sort of valuation consideration of the starting point when assesing probable futre return paths is an improvement.

 

The immediate attraction to living only off dividends as a safe withdrawal level is that by definition capital must be preserved.

However, the exact same thing is true of living only off interest – the capital is preserved. Straight away I hope every self-respecting equity and real estate investor will reflexively see the dangers of capital destruction of the capital in purchasing power terms.

In theory Equity (and property) investment are an inflation hedge but there is some rather lengthy delays historically between inflation hitting and earnings/dividends/pricing catching up.

The viability of living off dividends to my mind must be addressed in “real” terms to see how volatile the real dividend stream is and how long in historical terms has it been supressed below past peaks.

I only have the split for Aus data between capital and dividend components back to 1974 which really irks me for investigating living off dividends alone. I would like a lot longer series to get a bigger picture so if anybody has a source I would be for ever grateful. The Shiller dividend data for USA goes back to 1871 which is much more usefull - however US centric.

 

This is a recreation of the nominal dividends in the chart Austing provided. The data is monthly and the nominal value is slightly different than Austings chart which I suspect is because I'm using whole market data and Austings chart is just for the industrials (ex resources) component. The difference is immaterial to the big picture of difference between nominal and real. We are still quite a way under our 2007 dividend levels in real terms. Somebody retiring in 2007 and relying on the whole of their dividend stream at that point to fund retirement would be hurting pretty badly by now I suspect in keeping their captial in tact. Given the pension has asset test applicable and they may be beyond producing wage income to supplement, there might not be many options but to start chewing into the capital or continue living a diminished lifestyle in real terms.

 

This chart shows historically how far and for how long Australian dividends have pulled back from their previous peak in real terms.



Worst in my relatively small sample for Aus data is 45% maximum pull back with 15 years to recover peak between 1990 & 2005.

GFC fallout hit nearly 40% contraction and we are still ~ 20% underwater 10 years on (in a low wage growth environment to boot).

 

One important point about the SWR in Australia that I have mentioned before is that if your dividends are over 4% then you need to reinvest the difference - as the SWR is calculated off total shareholder return.

The SWR is a dynamic problem and using a simple fixed solution is not efficient. I will outline a risk management approach with dynamic inputs:

Assumptions
1. Shiller Cape accurately measures long-term market returns.
2. The research shows that SWR failures occur as a result of early losses after retirement.
3. Avoid selling down equities.

Based on 1, Shiller Cape should define the asset allocation approaching and after retirement. A simple method might be increasing cash / bond allocation by Shilller Cape's percentage above it's long-term average.

Based on 2, the risk appetite approaching and immediately after retirement should be at their lowest. Something like an inverse normal curve where retirement is at 0.

Combining the Shiller Cape index and the phase of retirement should give an indication of the asset allocation required to manage the risk appropriately.

Here are a few scenarios:
A. Shiller Cape high (markets expensive after long bull market)
B. Shiller Cape around average
C. Shiller Cape low (markets cheap after long bear market or crash)

V. Far off retirement (> 8 years)
W. Approaching retirement (< 8 years)
X. Just retired (< 5 years)
Y. Retired for a while (> 5 years)

The matrix below defines a sample cash allocation (these are relative figures from the top of my head, do not use them as they need to be calculated by performing back testing):
A B C
V 5% 3% 1%
W 15% 10% 5%
X 30% 25% 15%
Y 10% 10% 5%

During years of large losses / expensive equities, withdrawing less money is advised and vice versa. That is, the SWR is variable based on the matrix above. Withdraw less when your cash allocation is higher as these are the higher risk scenarios.

Summary:
Risk appetite decreases as a person approaches and passes retirement while the inherent risk is higher when markets are expensive. A variable withdrawal rate and a defined allocation to cash can be calculated to reduce the chance of SWR failures.

 

Same depth and duration of real dividend declines for USA over a much longer data sample.

 

Where is the thumbs up emoji?

 

@dunno thanks for the work you're putting into this.

Most Aussie retirees who have made an effort to be self funded will likely be drawing their income from a "tax free" Super account based pension. That means full refund of franking credits and no tax on interest etc. This will make a meaningful difference in improving the real return of the income stream.

Your thoughts on this please?

 

How much cash? USA shows historical instance of 40 odd years below previous real dividend peak level - thats going to need a pretty big bucket.

"There’s a hole in my bucket, dear Liza, dear Liza! There’s a hole in my bucket, dear Liza, a hole"

 

Look out @ErYan's on fire.

Supposeably one of the safer options is to split retirement savings into a Risk Free Liability Matching Portfolio and a Risk Portfolio. Unfortunately given inflation the amount needed in risk free assets to fund all of ones retirement is substantial and impossible for many. So the next best thing is a risk free "Income Floor" to match liabilities during the earlier stages of retirement when sequencing risk is a real danger.

This reduces the chance of having to draw on the risk portfolio earlier in retirement.

This is why I maintain greater than a three year cash buffer and assume dividend income could be cut in half for a lengthy period. I'm trying to reduce the possibility of drawing on the Risk Portfolio during an extend crash / bear market. I need to be able to top up any dividend shortfall for potentially a long time.

 

Certainly, part of the equation. The return we are interested in is total 'real' return after tax and fees. The lower your fees and the lower your tax the higher your total return. The higher your total return the higher the sustainable withdrawal rate will be.

Imputation regime is adding somewhere between 1.5-2% extra return above corporate tax rates to exempt pension accounts. That’s not insignificant, the exact impact on sustainable withdrawal rates I don't know without modelling it. But volatility (sequence of bad returns) is more the devil than the ultimate return so impact may not be that substantial.

 

It would be very hard to think-build a overall theory-- only those that have experienced the down side would know how to mitigate the consequences . .One only has to look at the price of CBA it went from $85.00 july into the low $74 mid range a few minutes ago and that's the way free markets work and the ones that bought in at 85 bucks would by now understand how the trail and error process works..

 

From the Mad Fientist

Safe Withdrawal Rate for Early Retirees

The Morningstar Asset Class Gameboard for 2016/2017

best to worst was:

1. International Equity (unhedged): 14.7%
2. Australian Equity (ASX/S&P200 Total Return): 14.1%
3. Australian Small Caps: 7.0%
4. Cash: 1.8%
5. International Fixed Interest (hedged): 0.5%
6. Australian Fixed Interest: 0.2%
7. Australian Listed Property: -5.6%

 

@dunno in relation to All Ords historical real returns this Report might interest you. Refer to page 19 onward:

http://www.philocapital.com.au/imag...12.pdf?phpMyAdmin=0n6odT3,EGip9lBgzXdLgnIEe9f

For many investors it would make for scary reading.

 

Ahhhh the tranquillity of yester year, If only we can repeat it in the future.