Sequencing risk on the road to FIRE – returns aren’t smooth and why it matters

Sequencing risk for those of you who haven’t heard of this enthralling topic previously is the risk that the order and timing of your investment returns is unfavourable and results in you having less money than you would if your returns were nice and smooth, ie like the ones we all use in our projections.  We tend to hear about sequencing risk mostly in the context of already being retired and drawing down money from your portfolio, but it also applies when you’re adding more money to a lump sum as you would be when you’re building up your wealth.  Which to be honest is the stage most of us are currently in and certainly the stage that I’m currently in, so it’s pretty important!

Although it’s nice to pretend that our returns are constant over time, in the real world this isn’t the way it happens.  Markets are up a bit one year, down a bit the next, up again by a different amount again etc.  You may remember at some stage in your math classes in school your teacher telling you that when it comes to multiplication and division it doesn’t matter what order the sequence comes in, the result will be the same. 

Which is absolutely true, so long as you are only dealing with the one original number you are multiplying or dividing.  So if you have a lump sum investment and don’t add any more to it then it doesn’t matter what order the returns come in, the result at the end of it will be the same as shown in the below table.  The first column shows an actual return sequence, the second shows that sequence but in a low to high order, and the third shows the same sequence but high to low.  The next 3 columns then show the effect on the amount.  I’ve used a starting lump sum of $50,000 and as you can see the order of the returns doesn’t matter.

But for those of us who are adding more money to the investment over time (or subtracting it if you’ve already hit FIRE) it’s a different story.  In this case I’ve used a starting amount of $50,000, then added another $50,000 every year.  This is meant to be a somewhat realistic scenario for an individual or a couple who go from using every dollar that comes in to saving a decent amount of money each year, investing at the end of the first year and each subsequent year.  So for instance someone that has just paid off a mortgage and is now looking to build up wealth.  In this instance the sequence of returns becomes far more important. 

The order in which the returns occur now makes a huge difference.  As you can see that actual amount at the end of that time period when the person contributed $50,000 each year (so a total of $550,000) is $1,214,330.  If they’d had a low to high sequence of returns, ie worst returns when they had a small amount of money invested and best returns when they had the maximum amount, it grows to $1,486,275.  But if they had the worst possible sequence where the best returns where when they had a small amount invested and the worst were when they had the most invested, that money would have only grown to $704,012 or less than half the best case scenario and only 58% of the actual return.

As fascinating as all of this is, the big question of course is how it applies to the real world.  To play around with the numbers I used the returns from the All Ordinaries Accumulation index (ie capital growth and dividends)  and subtracted the inflation data from this going back to 1960 to give a net return.  Obviously this doesn’t take into account a whole bunch of things like taxes etc but it gives an idea of what actual returns might look like.  I’ve looked at a ten year time period with lump sums of a flat $50,000 being invested at the end of each calender year with a total of $500,000 being invested over the period although it’s only getting the benefit of compounding for 9 of those years.  So at the end of the time period, you’d certainly be hoping to have at least $500,000 in there and ideally a lot more.

What the results showed though was that out of the 48 time periods I looked at there were actually 7 periods where you ended up with less than what you’d put in.  Six of these were for consecutive periods starting back in 1964 so you ran into the bear markets and huge inflation in the early to mid 1970’s at the end or middle of the ten year period where you already had a lot invested. 

The worst example of this was starting to invest in 1965 where you put in $500,000 over that ten year period but only ended up with $295,021.80 because you ran into stock market losses of -23.3% and -26.9% as well as inflation of 9.1% and 15.8%.  That’s gotta hurt.  The other loss was of course if you started on the road to FIRE in 1999 and ended in 2008 where after building your portfolio up to $810,845.73 at the end of 2007 and could almost smell FIRE, then the GFC knocked you back down to $497,586.84.  Ouch. 

Hopefully investors would have stuck with the sharemarket in which case they would have made up their losses over time, but realistically there are probably a lot of people who would have sold all their holdings and gone to cash, thus locking in their losses.  It’s easy to say you just need to hold on, but it’s very hard to do.

If I look at the average net return (so accumulation index – inflation) over the 57 years covered in my calculations it was 9.5% and if I use a nice smooth 9.5% in my calculations then I get an end value of $778,014.53.  Looking at actual returns though only 21 out of the 48 results were above this level, so less than half, and as mentioned above 7 actually ended up with less money than was invested in the first place! 

This leaves 20 scenarios where there was more money at the end of the period than was put in, but less than there would have been had the returns been the average over the total time period.  Obviously the returns in any shorter time period can be substantially be different than the returns over a much longer time period so this is to be expected to some extent, but it further illustrates the point that using smooth long term returns in calculations isn’t likely to reflect what actually happens in the real world.

How does this affect FIRE

The two main methods of generating income once you have retired using shares seem to be the Bogle method which is selling down your portfolio over time and the Thornhill method which is just living off the income stream from dividends.  It’s pretty obvious that if you’ve got less money invested you’re going to be more likely to not have enough to declare FIRE if you’re relying on selling down shares so this definitely affects the Bogle method, and there is potentially a big impact there.

An argument can possibly be made that sequencing risk won’t affect the Thornhill method much, but it’s a pretty weak one in my opinion.  Yes dividends are less likely to get cut even if the sharemarket crashes, but they definitely can and have been cut in the past.  Generally nowhere near as dramatically as the fall in the sharemarket, but cut nonetheless.  And although again we tend to use a nice smooth figure of a 4% dividend yield on Aussie equities with the associated franking credits in our calculations, in reality there are times when the dividend yield is higher and times when it is lower so you’re not consistently buying the same amount of dividends with each dollar you invest.  If the market jumps 40% in one year chances are pretty good you’re not buying 40% more dividends than you were in the previous year.

Even if you’re buying LICs which are smoothing dividends over time it’s not as though they have some magic money generating machine so if the dividends of the underlying shares they hold are cut deep enough or for long enough they’re going to have to cut their own dividends as well, or start selling some of their underlying holdings to generate cash which will drop the future income they are going to receive.  Either way it hurts you.

So what’s the solution?

Unfortunately as sometimes happens there isn’t really one.  We don’t know what future returns are going to look like so we can’t plug these into our spreadsheets for calculating our FIRE date, let alone use them to invest.  Diversifying into other assets can certainly reduce the risk, but it may also mean delaying FIRE depending on the returns of the other asset classes you are investing in.  Depending on which method of generating income you’re using when you hit FIRE you might be able to build in some margin of safety, but there’s no guarantee it will be enough particularly if you’re selling down assets.   It is just one of those things that you have to learn to live with, and at the end of the day even if you do get hit with some big market falls at the wrong time you’re still going to be in a better position than those who have never saved anything.

As I’ve said previously when I get to FIRE I plan on having a year or two worth of living expenses (probably two because I like to play it safe) parked in HISAs or term deposits so I can ride out a downturn to some extent, but I don’t plan on implementing that until the last year or two before I pull the pin on work so it’s not going to smooth my returns while I am still getting to that FIRE number.  I’m also aiming for Fat FIRE so I will have plenty of room to cut costs if it should prove necessary, and I’m planning on relying on a more stable dividend income stream rather than selling down assets to generate enough income to support myself and my family.  All of these things should help, but as I said they’re no guarantee unfortunately.  Sometimes you just have to take a leap of faith, although if you can get it down to being a short hop so much the better!

Have you heard of sequencing risk previously, and what’s your plan to deal with it?  If you liked this post and would like to read more like it then please subscribe using the link on the right!