Safe Withdrawal Rates for Aussies — Part 5: Retirement Length

Posted by on 23 December 2018 in Safe Withdrawal Rate Series

In part 3 of our Safe Withdrawal Rate Series we set out to determine the optimal mix of Australian equities and Australian bonds that would allow us to withdraw the most from our retirement portfolios without running out of money.

We found that we could safely withdraw 4.00% of our retirement balance per year (adjusted for inflation) as long as the portfolios had at least 75% Australian equities.

We also noticed that portfolios with 75% equities and 25% bonds actually performed better than a 100% equities portfolio. This was our first indication that diversification can be beneficial, even if we diversify into a lower return asset.

Then in part 4 of the series, we extended our analysis to determine whether international exposure would be better for Australian retirees. We found that we can safely withdraw more when Aussie retirement portfolios were made up of Australian equities, U.S. equities and U.S. bonds. We don’t even need to buy Australian bonds!

One conscious limitation of all that previous analysis is that we assumed a 30 year retirement length. However, in reality many early retirees may be retired for 40, 50 or even 60 years. So in this article we will examine the effect of longer retirement lengths on safe withdrawal rates.

The elegant asset mix for Aussies

Before we get into the analysis on safe withdrawal rates for longer retirement lengths, some comments on the asset mix that we uncovered in the last post.

It’s actually quite elegant.

As we explained, an Australian retiree’s portfolio should include some proportion of U.S. bonds. The amount of U.S. bonds will be determined by our risk tolerance, more U.S. bonds equals lower risk. Then the remainder of the portfolio should split between Australian and U.S. equities. This will give us the best return for your risk appetite.

Chart showing the best asset mix for an Australian retiree

Just note that the asset mixes in the above chart are simplified/rounded slightly. For example, in a perfectly optimised portfolio the equity split tends to be 1-4% more heavily weighted to Aussie equities and sometimes it includes 1-2% Australian bonds. You can check out the exact asset mix here.

But, practically speaking, we can accept that an optimised portfolio should look broadly like the asset mix in the chart above. So in the interest of simplicity, we will calculate using portfolios with these simplified asset mixes. This doesn’t change the outcome significantly — but makes the calculation much easier to follow.

How does retirement length affect SWRs?

Anyway, let’s get into the meat and potatoes. Calculating the effect of retirement length on safe withdrawal rates…

Background on the calculations

Before we start calculating we should run through a little house keeping:

  • We are using our data set of Australian equities, Australian bonds, U.S. equities and U.S. bonds for 1770 months from 1871 to 2018
  • For each monthly starting point we calculate the exact withdrawal rate that will result in a $0 balance at the end of the retirement period. We calculate using the approach that we outlined in Part 3 of this series
  • All of our calculations are adjusted for inflation (i.e. real dollars) and are net of foreign exchange movements

The relationship between retirement length and safe withdrawal rates

The chart below compares 30 year retirement length and 60 year retirement length safe withdrawal rates from 1871 to 2018 for a portfolio made up of 50% Australian equities and 50% U.S. equities.

As you can see, increasing the retirement length reduces safe withdrawal rates by about ~1%. So we would expect there to be higher failure rates for longer retirement lengths.

However, although we know what trend to expect, there are still a number of outstanding questions, such as:

  • How much do longer retirement lengths increase portfolio failure rates?
  • Does asset mix change the relationship between retirement length and safe withdrawal rates?
  • Does increasing retirement length reduce safe withdrawal rates in a linear manner?

The effect of retirement length on portfolio success rates

As always, to understand what safe withdrawal rates mean in practical terms, we convert them into success rates. It works like this: if a withdrawal rate of 4.00% has a success rate of 95%, this means that we would have run out of money during retirement in only 88 of 1770 monthly starting points.

The huuuuge table below outlines the success rates for various portfolio allocations, withdrawal rates and retirement lengths.

There’s quite a lot to digest there, so here are some interesting findings:

  • We replicated the findings from the previous post in this series that showed equity portfolios split 50/50 between Aussie and U.S. equities perform better than either 100% Aussie or 100% U.S. equities alone
  • We can maintain great success rates for high equity portfolios over a 60 year retirement length when we diversify across Aussie and U.S. equities. And this is the case for withdrawal rates as high as 4.25%!
  • High equity portfolios that are not diversified across countries will not always survive longer retirement. For example, if we choose to invest in 100% Aussie equities, then we need to reduce withdrawal rate to 3.75% or 3.50% to last 60 years

Visualising longer retirement lengths

The chart below plots portfolio success rates for a 50% Aussie equities, 50% U.S. equities portfolio of different retirement lengths. It’s essentially some of the same information in the table but presented in a visual format.

As we know, increasing withdrawal rate results in a drop in portfolio success rate and this happens faster for longer retirement lengths. And as we can see below, there relationship is non-linear. In other words, as safe withdrawal rates increase, portfolio success rates drop at an accelerating rate.

If we wanted to, we could build the same chart for different portfolio mixes — but it would be showing the same trend, so let’s leave it for now.

What did we learn and what are the caveats?

Before running this analysis, my hypothesis was that we’d see portfolio success rate fall quite significantly for all portfolio asset mixes when retirement length increases to 60 years. And we did see success rates fall for portfolios with 100% Aussie equities or 100% U.S. equities.

But portfolios with a high equity allocation that was split 50/50 between Aussie and U.S. equities maintained persistently high success rates. So it seems that high equity allocation and international diversification put together helps maintain acceptance portfolio success rates.

That’s great news!

However, we must keep in mind that we have defined portfolio success as the portfolio having at least $1 at the end of retirement. As one reader mentioned, most retirees don’t think this way. It’s quite risky and there’s not much room for error. Would people be able to watch their portfolio fall close to zero and not react? It would be a scary ride!

In reality, most people will want to end retirement with money left in their portfolios. So in the next post we will look at how the target balance at the end of retirement effects safe withdrawal rates.

As always, if you have any questions, comments or feedback, please email me or comment below. I’d love to hear your thoughts.

The information in this website and the links provided are for general information only and should not be taken as constituting professional advice. Ordinary Dollar is not a financial adviser. You should consider seeking independent legal, financial, taxation or other advice to check how the website information relates to your unique circumstances. Ordinary Dollar does not guarantee the accuracy of any information on this website, and is not liable for any loss caused, whether due to negligence or otherwise arising from the use of, or reliance on, the information provided directly or indirectly, by use of this website.

There are 13 comments on this article

  1. Avatar for Dan Montgomery


    What about investing in LICS e.g. AFI, MLT or ETFs e.g. VAS? Their dividend yield is already 4% and thats before franking credit and you never have to sell your golden goose portfolio.

    • Avatar for Dan Montgomery

      Dan Montgomery

      It’s a good point! I haven’t really looked at LICs in much detail. I think it’s worth delving into, although we’ll need a high quality long-term data set of LIC returns. I’ll definitely do some investigating.

      • Avatar for Dan Montgomery

        Ben Murphy

        I agree with this. Would love to hear more on this subject.

  2. Avatar for Dan Montgomery

    The FI Explorer

    This is an excellent post and analysis, thank you! And your point on residual portfolio is an issue missed in so many discussions on safe withdrawals. Finishing with $1 in the bank is not a particularly comfortable thought!

    I was just curious, the focus is on US stock and bonds for diversification. Have you thought about doing some broader equity market diversification? EAFE, or some global index? This could potentially help the results be slightly less leveraged to past US asset performance. There is always a slight risk of ‘overfitting’ to past performance in these types of estimates.

    • Avatar for Dan Montgomery

      Dan Montgomery

      Thanks mate, appreciate the comment!

      The biggest challenge that I have with these calculations is finding a good data set (as they say, garbage in equals garbage out). I actually spent like 2-3 weeks solely on building and validating the data alone. Unfortunately, I haven’t come across a good free global index. But I do agree with you, if I can find something then it’s absolutely worth looking at 🙂

      You’re right, there’s always a risk when you base things on past performance. Despite what the data above says, I personally would only withdraw 3.75% or lower per year even with a high equity, highly diversified portfolio. You never know what’s going to happen!

  3. Avatar for Dan Montgomery

    Aussie HIFIRE

    Another fantastic post Dan, I’m thoroughly enjoying this series! It’s quite interesting that there really isn’t much of a difference between a 40/40/20 and a 50/50/0 split. I’m wondering how this plays out at higher withdrawal rates, presumably the 50/50/0 extends the lead if this is the case?

    It’s also interesting to see the times that the SWR has dropped significantly is where you would have run into sequencing risk early on, so the Great Depression, Oil/Inflation shock in the early 70s, tech wreck and then the GFC. Other than those times you probably could have had a much higher withdrawal rate.

    • Avatar for Dan Montgomery

      Dan Montgomery

      I haven’t looked at that specifically but I suspect you’re right, the 50/50/0 would presumably perform better than 40/40/20 at higher withdrawal rates.

      I haven’t spoken explicitly about sequence of returns risk but I think it’s worth it’s own post. Sequence of returns risk has a bigger effect on safe withdrawal rate than average returns themselves. I think most people understand the concept but we can spend some time teasing out the practical implications for retirees.

  4. Avatar for Dan Montgomery

    Retirement Investing Today

    A very interesting set of posts. Can you please clarify one point. Your US Equities Total Return (TR) is priced in USD, your AUS Equities TR is priced in AUD and your US Bonds TR is priced in USD. If I’m interpreting correctly then to replicate the portfolio’s being analysed in real life you’d need to currency hedge the US Equities and US Bonds portions of the portfolio to the AUD. Have I understood correctly and any idea how you might do that simply and for low cost?

    • Avatar for Dan Montgomery

      Dan Montgomery

      You’re right in saying that the Australian indices are priced in AUD and the U.S. indices are priced in USD. However, we take into account currency fluctuations when we look at U.S. returns (i.e. return equals index return plus currency gain/loss). Therefore, to replicate the findings you would not hedge the U.S. portions of the portfolios.

  5. Avatar for Dan Montgomery

    Keith Invests

    Hi Dan,

    Are you using US and Australian Market Indices as a proxy for the long term equities exposure returns?

    I am a firm believer in buying Direct Australian Equities and holding them for the long run, which is how I achieved Financially Independent Retirement Elective (FIRE) before my 40th Birthday.

    I recently bought my first direct US Stock exposure (Facebook) which has been a disappointment but this learning experience which should eventually teach me how to successfully diversify my equity exposure towards more direct international shares.

    • Avatar for Dan Montgomery

      Dan Montgomery

      My philosophy is that I can’t beat total market returns by picking individual stocks. I believe that buying into the market as a whole is the safest way to achieve comfortable returns, as the market trends upwards over a long enough time period.

      That’s not to say my approach is the only way. In fact, I’m sure there are people who can successfully pick individual equities and it sounds like you’re one of them! Nice work!

Leave a Reply