Home / 142

Safe withdrawal rates for Australian retirees

I love it when someone takes a complex question and answers it with something simple. The danger with elegant simplicity, though, is that people forget the details that sit behind it, and what question it was actually answering. This was one of the catalysts for the recent Morningstar research paper ‘Safe Withdrawal Rates for Australian Retirees’ that I co-authored with David Blanchett and Peter Gee. The ‘4% rule’ is often referenced in understanding what you can spend in retirement given a certain amount of savings, but where did it come from, and how relevant is it today?

1. What is this ‘4% rule’ we hear so much about, and where did it come from?

The ‘4% rule’ actually started in 1994 with an article published in the Journal of Financial Planning by William Bengen. He was a US-based financial planner who wanted to answer questions about how much his clients could spend in retirement. The way people interpret the 4% rule can vary, so let’s set out some important parameters that underpin the number:

  • 4% of the portfolio is used to calculate the first year’s payment only, and each subsequent year that amount is adjusted for inflation

    .

  • it assumed a minimum 30-year retirement period

    .

  • historical return data from 1926-1994 was used, based on a portfolio comprised of 50% US equities and 50% US bonds

    .

  • 4% was selected as ‘safe’, because at that level there was no past period where that rate would have exhausted all assets by the end of the 30-year period. So it was not a number based on an average return, but rather one that assumed returns at the very low end of the spectrum.

Before taking this framework forward, I’d like to tip my hat to Mr Bengen, who 22 years ago wrote a thoughtful and practical paper. The 51 simulations that he ran do not quite match up with the Monte Carlo simulators of today, but the paper still captured many important concepts.

An inflation-adjusted, constant income stream is pretty intuitive when you think about the way you want to plan retirement.

2. Does this 4% rule apply to Aussie retirees today?

The methodology can still apply in Australia today, but there are some important areas of improvement. First, we’ve included a fee assumption. Whether you’re paying for someone to manage the portfolio, an advisor, an accountant, an administration platform, or some combination of these, there are costs. For our calculations, we’ve assumed an annual fee of 1% per annum. If you repeated the same study as above with the 1% fee, using Australian share and bond returns, but increase the return history to 1900–2014, that 4% would have come out closer to 2.5%. Why lower? Apart from the impact of fees on the returns, the Australian equity market has been more volatile than the US, and our inflation higher in the 1970s and 1980s, so you need a lower withdrawal rate to weather the worst-case scenario.

The full paper also shows that Australia has experienced some of the highest historical returns from markets when compared to 19 other countries. While the US has led the world in retirement research, we need to be careful about localising those results. Australia has outperformed historically, but it’s arguable whether this will continue.

The next step was to replace past returns with our long-term expected returns, which take account of where equity markets and interest rates are today. In addition, if you diversify the portfolio further to include a mix of Australian and international assets, you get different answers again. If you want 99% certainty, the initial withdrawal rate is 2.8%, helped by the portfolio’s reduced volatility. If you’re prepared to lower that probability of success down to 80%, then that initial withdrawal rate can increase to 3.9%.

3. What is the probability of success or ‘success rate’?

This idea of a ’success rate’ is incredibly important. While it may be complex mathematically, the underlying principle isn’t. It speaks directly to the sort of trade-offs we all have to make. Quite often, people talk about ‘expected returns’, and use these to build their plans. Even if someone has made a good forecast, an expected return will only have a 50% probability of coming through, and the final result may be higher or lower. You might be happy around this level, or you may want to be more certain that the path you’re taking will meet your minimum goal. In our analysis, the goal is to make sure that whatever initial withdrawal rate you use, your account balance will run out exactly at the end of that period. Pick a success rate that you can be comfortable with, from the conservative 99% certainty, to the more optimistic 50% level, or somewhere in between.

4. What is the key message for Australians?

Equity returns over the next 20-30 years are likely to remain attractive relative to cash, but we’re projecting them to be 2% lower than history. We need to adjust our expectations and plan accordingly.

Safe withdrawal rates for retirees now need to start at 2.5%, not 4%. Withdrawal rates could be even lower if life expectancy continues to increase. So we need to accept either spending less in retirement, OR saving more for retirement, OR running a greater risk of moving on to the aged pension sooner. It’s important to understand the trade-offs, and where you’re sitting.

The mandatory minimum withdrawal rates for account-based pensions in Australia are set higher than the safe minimums in our paper. The way these two rates operate is different after the first year, but the impact of the higher relative withdrawal rates still needs to be considered. Just because you’ve been paid an amount from an allocated pension doesn’t mean you have to spend it. Some retirees will need to invest some of their pension payments outside tax-concessional superannuation to ensure they still have savings in the future.

Once again, the benefits of a diversified, balanced portfolio shine through in the study. Adding equities can help a portfolio, but only if you accept a lower probability of success. Most of the incremental benefit to withdrawal rates of adding equities is achieved when 50 – 70% is allocated to growth assets.

Lastly, while the paper provides some useful pointers, the reality is that we’re all different, and reviewing your own personal circumstances will give you a much better answer to what you need in retirement than a rule of thumb.

 

Anthony Serhan, CFA, is Morningstar’s Managing Director Research Strategy, Asia-Pacific. For a full copy of the report and data, click here. This material has been prepared by Morningstar Australasia Pty Ltd for general use only, without reference to your objectives, financial situation or needs. You should seek your own advice and consider whether the advice is appropriate in light of your objectives, financial situation and needs.

RELATED ARTICLES

It's not a shock that retirement is different

Moving your SMSF into pension phase

Understand the retirement income challenge

20 Comments

Paul

June 16, 2017

Are we complicating what can be; and is currently for me, a simple strategy. Investing close to 100% in Australian shares with a foundation of bigger stable companies (Banks, Wesfarmers, Woolworths etc), pyramiding up with mid and small caps - all with a focus on growing dividends as well as capital growth (Nick Scali, Bapcor, Adelaide Brighton Cement, Think Childcare, ASX etc). A small portion is invested in small pure growth companies (Next DC, AMA group etc).

I have a cash contingency amount outside of super in case of GFC 2; purely to supplement any possible cut in dividends.

In a SMSF, the average 5% dividend yield produces $50,000 PLUS $21,000 franking credit refund (in Pension mode) pa on a $1,000 000 cost base for the share purchases. $2,000 000 produces $100,000 in dividends PLUS $42,000 in franking credit refunds - for a total of $142,000 pa;without ever selling one share!

As companies tend to increase their dividends over time, that 5% (on purchase price amount) can quickly grow. dividends were only cut for a year or two (in some cases) during the GFC and quickly recovered.

A simple strategy that requires some stock picking guidance / expertise (plenty of good subscription based Newsletters & info sites etc available. I pay virtually no fees.

Why aren't we hearing about this as an obvious strategy??? Also, those pure growth companies can be sold off without ever affecting dividends and in fact just add more cash into the equation ...or the investor's pocket!

Anthony Serhan

February 25, 2016

Alun, in addition to Andrew's comments and my earlier posts a few quick points. We did include an allowance for imputation credits in the Australian equity assumptions - Exhibit 10. Also, in Exhibit 13, if you look at the 70/30 portfolio and operate at the 50% Probability of Success (expected return), you get SWRs ranging from 4.3% to 6.0%, depending on the Retirement period. The 2.5% number you use comes from re-running the analysis using historical data and assuming a 99% probability of success. Take a look at Exhibit 13 for the SWRs generated using the forward looking assumptions. I look forward to getting on the road and chatting more about this face-to-face with people.

Andrew Barnett

February 25, 2016

But Alun, you're assuming 100% investment in Australian equities? And, the 2.5% is indexed to inflation. If you look at the historical record from 1900, there have been periods where a diversified fund would have product zero real returns for a decade. If you draw down 4% pa, plus another 1.5% for fees, you've withdrawn 55% from the account, and the possibility of pension exhaustion is high.

Alun Stevens

February 25, 2016

Not sure what Anthony was thinking of when setting up the paper. Dividend yields are above 4% in Australia and have been very stable over a long period even with capital value volatility as described by Anthony. If dividend imputation is taken into account the yield is higher.

Locking into 2.5% is just a method for entrenching a lower standard of living and ensuring a bigger inheritance for the kids.

If I had drawn down my super at 4% over the last 10 years (i.e. including the GFC) my asset base would have increased. It would have increased even if I had drawn it down at 4% of the high watermark value of assets (i.e. maintained the drawdown even when asset values were smashed during the GFC).

The real question is how much above 4% one can go and still maintain the money for life. With proper asset liability matching and cash flow management, a rate of 6% is more realistic.

Lower minima would simply enhance the tax sheltered estate planning aspect of superannuation.

Aaron

February 25, 2016

Alun
I also get the result for Australian equities that the nominal value would have been maintained after 10 years- about a 22% reduction in real terms but I don't see how capital has increased if you started at the peak in 2007. I find a reduction in capital of around 30% (44% in real terms), (even after including franking credits).
Also, with the recent cuts announced, dividend levels are lower in real terms than what they were 10 years ago!

There are flaws in the safe withdrawal approach in that it assumes no adjustment after the point of retirement so the conservative approach always generates a large balance for the estate.

I assume your reference to proper asset liability matching is advocating for some form of pooled longevity insurance to cope with uncertainty of death. Retirees certainly don't know when their cash flow needs will end in advance.

David

February 16, 2016

In relation to your opening paragraph Anthony there is a wonderful quote that goes... "For every complex problem there is always a simple solution. And it is almost always wrong".

Personally I am quite confused and bemused about how minimum withdrawal rates for superannuation pensions has been interpreted as some sort of edict about the "right" amount of income in retirement.

The minimum withdrawal amount does not have to be spent, there is no maximum withdrawal amount (other than in TTR mode), and most people will have other non super resources that can potentially be used to fund their living expenses. There is hardly any relationship between the minimum withdrawal amount and actual retirement living costs for most people.

The minimum withdrawal amount just forces people to take some money out of a tax free environment to either spend, or accumulate in a potentially taxable environment. It is another example of checks and balances on the alleged "tax rorts" of the superannuation system. Like contribution caps in reverse.

Anthony Serhan

February 15, 2016

Mark, let me expand my comment about equities (I agree, when it comes to investing in equities it is about the underlying businesses). Look at the portfolios in Exhibit 13 of the paper. At the 50% success rate (expected return) level you see the classic relationship between higher returning equities and the SWR – more equities produces a higher SWR. As soon as you start increasing the desired success rate, the payoff for adding more equities starts to diminish, and at the 90% success rate and above, the relationship starts to reverse. While the math behind this is complex the principles are quite intuitive. The important point is that we are just not working in the risk and return dimension, but have added a liability we are trying to match to – think objectives based investing. If you accept equities are riskier than cash and bonds, than the more certain you want to be about the outcome (higher success rate), the less equities will help. I am not advocating a particular point on the curve but rather, to understand the relationship.

Paul, Peter, Greg thanks for the comments and observations. You are right, this sort of modelling starts becoming more useful when it operates at the individual level and can incorporate more real life circumstances. This is something we have invested in heavily to help our clients and it is exciting to see the range of tools now becoming available in the market.

Greg Einfeld

February 15, 2016

It is a shame most retirees don't understand this. Most of them say "I am earning returns of 7% p.a. therefore I can afford to spend 7% of my capital each year." They are overlooking inflation, investment risk, and lower expected future returns.

Peter Vann

February 14, 2016

Anthony, it is great to see Morningstar’s addition to the retirement outcome discussion in Australia using analysis that accounts for uncertainties such as investment volatility.

Taking this one or two steps further, all members should be able to obtain proper stochastic forecasts of their safe withdrawal rates whether in accumulation or retirement phase. Such forecasts would obviate the need to debate simplistic rules of thumb, such as the 4% rule. They are necessary to enable members to see the potential impact of choices they make today on their safe withdrawal rates optionally including the age pension in their total retirement outcome.

One of the impediments to superannuation funds providing individualised stochastic forecasts as part of each member’s annual report is that Monte Carlo simulation technology is generally too slow. But it is possible to do this using closed form mathematical techniques that allow each member report to be produced in a fraction of a second. Moreover, these techniques can drive very responsive “what if” functionality of a superannuation fund’s website.

I look forward to further additions to the Australian retirement landscape from Morningstar supporting use of stochastic analysis.

samir

September 21, 2016

Dear Peter (Vann),
I came across an old post of yours in morningstar (http://discuss.morningstar.com/NewSocialize/forums/t/104612.aspx). I am looking for ways to compare my money weighted portfolio return against some benchmark however I have no idea as to how to calculate money weighted return for an index. I have found your idea in that post immensely useful. Is there anyway you can provide more details so that I can go through it and try to understand how to do it myself (this is for my personal portfolio comparison). I appreciate your help very much in this regard. Thanks samir

Paul Goodwin

February 12, 2016

Thanks for the research Anthony. Your report gives a less optimistic view than comparable Australian studies such as Drew & Walk (2014). I assume using forecast returns rather than historical returns, and including the impact of fees would account for these differences.
Using an asset/ liability matching strategy (or bucket strategy) to protect against sequence risk should increase the safe withdrawal rate. Reducing spending in the later years (80yo +) would further increase safe withdrawal rates. A dynamic spending policy which adapts to market movements would also assist. In most cases a 2.5% withdrawal rate would result in very considerable wealth at death- and the associated missed opportunity of an improved lifestyle.
The biggest factor potentially increasing safe withdrawal rates would be consideration of the Age Pension. However, the impact of the Age Pension requires fairly sophisticated retirement planning software and is subject to legislative change.
Good luck to the average punter planning for retirement without some professional assistance!
Alex- this isn’t necessarily an article about superannuation, it is about retirement planning in general and your dream of being financially independent even without work. A retiree can hold significant assets outside superannuation/allocated pensions and still fall under the tax free threshold. I encourage this to guard against legislative risk.

Mark Hayden

February 12, 2016

I respectfully disagree with a number of points raised. Whilst appropriately referring to the inflationary increases and the need to include fees, this paper has a number of short-comings.

As an example the comment “adding equities can help a portfolio, but only if you accept a lower probability of success” is short-sighted for two reasons. Firstly some of the invested capital is long-term in nature; eg some is not accessed for 20 years. Secondly, equities appear to be treated by Anthony and others as a commodity (eg as a bit of paper to be traded) rather than as an intangible asset in the form of a business. The Hayden Investment Model addresses these matters in considerable detail. The Model is currently being back-tested and it will then include a robust critique of the 4% rule.

Anthony Serhan

February 12, 2016

Ashley, You are right 4.0% = 25 times desired income, 2.5% = 40 times. A big difference in what capital you require. If you take a look at Exhibit 13 in the original paper, you will the SWR approaches 6% if the time period is shorter and you lower the required success rate. 6.0% = 16.7 times, which is getting closer to the multiples you mentioned. One of the main points of the paper though is that rules of thumb built off past return expectations aren't that useful looking forward. More importantly, lets not pretend we know what is going to happen over the next 30 years and talk about the range of possibilities. If you are "optimistic" and believe returns will be at least as good as what you expect and probably better, spend away, if not be a little cautious. Donald, the 2.5% figure was not an expected real return, but a first year withdrawal rate that would still provide an inflation indexed income stream over 30 years IF you experienced one of the worst return patterns possible in the 10,000 simulations run for the portfolio. As a side point, we know from the 70s and the first 10 years of this century that there have been periods where real returns ARE LESS than 2.5%. To quote Bengen's paper "You have prepared them to survive the worst that has ever occurred, and should circumstances be better than that, they will prosper. After all, isn't that what they hired you for? And isn't that what you wish for them?" I am more optimistic than that, but at least I know and accept the trade-offs I am making.

Michael

February 12, 2016

Does anyone remember Alex? Sincerely, Michael :))))))))))) PLANET EARTH

Graham Hand

February 12, 2016

Hi Alex,

Thanks for the feedback. We want to reach a broad church, not only old codgers like me. Super is important, but hopefully, this week's edition on China, January results, RBA decisions and Asia have more to do with investing than retiring. But point taken, young fella.

Cheers, Graham

Alex

February 12, 2016

Hello, Am I one of your very rare readers under 40 ? I eagerly await your newsletters, and I am sorry to say but the tone this year seems to be shifting to more and more focus on superannuation. As an Australian expat I must be a rare species of Australian investor who isn't obsessed with my super fund, but still very interested in getting financially secure and daring to dream, even independent (of work). Despite the results of your reader survey, don't forget us! We are your next generation of readers.... ;) Kind regards Alex, Rotterdam, NL.

Donald Hellyer

February 11, 2016

If returns are only going to be 2.5% (real) perhaps we should stop superannuation contributions. The private sector wouldn't accept such a low return. We would be better off to pay down debt, invest privately, take more risk (something superfunds find very difficult to do).

Paul

February 11, 2016

Graham, re your comments on super changes in the newsletter ... Spot on predictions.

Comany

February 11, 2016

Introduce death taxes and this will sharpen retireees focus on spending their money whilst alive.

Ashley

February 11, 2016

2.5% withdrawal rate (or even 3.9% at 80% confidence) will scare many people. That’s a multiple of 25 times the required living expense budget. Most retires are told multiples of 12 to 15 times, not 25.

· Problem is that NOBODY has nothing but their super to rely on. EVERYBODY has other assets to rely on – combinations of house, investment properties, businesses, inheritances, gov pension, etc.

· So, since something like 90% of Australians die with about the same level of total assets they retired with, they can afford to draw down more of their assets each year and still not worry about ‘longevity’ since it is not an issue.


 

Leave a Comment:

     

Most viewed in recent weeks

Retirees facing steep increases for basic items

ASFA has updated its tables on how much money is needed for a 'comfortable' or 'modest' lifestyle in retirement, but there are some prices rising well ahead of inflation.

Adele Ferguson on ‘Banking Bad’ and weaving magic

The journalist most responsible for the calling of the Royal Commission takes care not to be roped in by everyone with a complaint to push. It takes experienced judgement to gather the right information.

Let’s stop calling them ‘bond proxies’

With cash and term deposit rates at all-time lows, and fixed interest bonds not much better, investors are looking for ‘bond proxies’ to deliver more income. But is ‘proxy’ a misnomer, and what are they anyway?

Six warning bells against property spruikers

Property spruikers use common techniques, and con men will increasingly target older people who feel they do not have enough financial independence for their retirement years.

Helping your children build their super

It has become more difficult to build large superannuation balances with contribution caps and more people paying off home loans for longer. How can wealthy parents help their adult children?

Should retirees spend more and worry less?

When more than half of retired Australians restrict their spending to less than the age pension and fear running out of money more than death itself, they may be denying a better lifestyle for themselves.

banner

banner

Sponsors

Alliances

Special eBooks

Specially-selected collections of the best articles 

Read more

Earn CPD Hours

Accredited CPD hours reading Firstlinks

Read more

Pandora Archive

Firstlinks articles are collected in Pandora, Australia's national archive.

Read more