The Ultimate Comprehensive Analysis of The 11-Year Returns of the STI ETF: Part 4 – DCA vs Lump Sum
This is the final part of a 4 part series. I started off trying to write a single article but the amount of analysis grew too large to fit into a single post, so I’ve decided to split it up into the following 4 parts:
- Part 1 : The 11-year performance of Dollar Cost Averaging into the STI ETF.
- Part 2 : Multi-Period (5-Year to 10-Year) Rolling IRR if you DCA into the STI ETF.
- Part 3 : Multi-Period (5-Year to 10-Year) Rolling IRR of Lump Sum Investing for the STI ETF.
- Part 4 : Comparing the performance of DCA vs Lump Sum.
It’s almost the end of the first quarter of 2019! Man, how time has flown.
That also means that it’s around that time of the year when we all start getting our annual bonuses! Some of you may have already received yours and some will soon be receiving it within the next month or so.
If you’re fortunate enough to be receiving an annual bonus, it may represent more than 1 month of your normal monthly income. Some people may even get as much as 6 months worth of pay in one go!
That’s a lot of money, and it’s important that we make the most of it by buying a 70-inch screen TV! (Please help convince my wife.)
Ahem, I mean, we should be investing it so that it can grow and build up our retirement nestegg. However, you might be nervous. It is a lot of money after all, you worked an entire year for this bonus. It won’t be fun to dump it into the market and have it just drop by 20% would it?
Given the market condition and a lot of uncertainty, you might want to play it safe and slowly spread out your investment over 3, 6, 9 or 12 months. This way if the market does tank tomorrow, a portion of your money is still safe. You have some funds left to invest while the market is down.
On the other hand, if you try to spread out the money and the market shoots up 50% a month from now, you would have lost all of that potential gain! Plus, now you’re forced to buy in at a much higher price. That won’t do either.
So what’s the approach with the best likelihood of better returns?
If you were to invest in international index like the S&P 500 or global ETF, there’s been extensive studies into Lump Sum vs DCA, and Lump Sum has been shown to produce better returns more often than DCA so for that, the case is closed. However, as Singapore investors, how would the STI ETF perform in the same scenario?
Well this is what the analysis is about!
If you’re ready for all the maths, as per my usual style, then let’s dive in!
I thought you already did the DCA and Lump Sum analysis? How is this analysis different?
The DCA analysis in part 1 of this series looked at how the STI ETF performed if you invested in fixed amount per month over the life-time of your investment – basically, how you’d normally invest your monthly income as it comes in. However, that may be as low as $500 or $1,000 per month, and extremely regular over years.
Then in part 3 we looked at how the STI ETF performs when you invest in one lump sum – only put money in once at the start of your investment period and let it grow.
In this post we will look at the performance of the investment if you received a large lump sum but instead decided to spread the money out over a 3, 6, 9, or 12 month period to reduce risk.
We will also be looking at the rolling returns over time to see how this strategy would have performed at different time and investment windows (I’ll explain this later.)
The investment setup
Before we dive into all the analysis, here is how we will be approaching it.
- I will assume that we have just received a SGD 10,800 annual bonus.
- For the control, will assume that we are investing it all immediately at the start of the investment period as a lump sum.
- For the DCA strategies, we will test out 4 strategies of spreading out the investment in equal amounts over 3, 6, 9 and 12 months.
Here’s what that looks like in a table:
|Investment Strategies||Investment Amount||Number of Investments|
|Lump Sum||SGD 10,800||1|
|3-Months DCA||SGD 3,600||3|
|6-Months DCA||SGD 1,800||6|
|9-Months DCA||SGD 1,200||9|
|12-Months DCA||SGD 900||12|
Charted out, this is how the cash will be injected for each investment strategy over the first 12 months of the investment period:
Pretty simple right? With lump sum, you’re concentrating the risk up front in month 1, with the 12 Month DCA you are spreading that risk out over 12 months so each month you’re only investing 1/12 of your bonus.
This will allow us to see whether lump sum or spreading out over 12 months – or anything in between – is the best option for eventual returns.
But… what periods of time should we be using?
I can easily pick and choose any entry timing in the past 11 years to make either Lump Sum or DCA look better. If I wanted to make Lump Sum look better, then I will start the analysis at the lowest point in the market, if I wanted to make DCA to look better, I will start the analysis right before the market tanks.
So to keep this fair – and if you have read the previous posts in this series – we will be using our old friend “Rolling Returns Analysis.”
Adding in the lens of the rolling returns
For those who are uninitiated, a rolling returns analysis is when you calculate the rate of returns of a particular investment vehicle over a fixed length of time, but for all possible investment windows.
To illustrate what this means in simple terms, let’s take the STI ETF.
I want to find out how much the STI ETF returns over a 5 year period. So I can start taking the price of the STI ETF at the start of January 2001 and then look at the selling price again in January 2006. That gives us 1 rate of return. However, if it just happens that in January 2001 the price was unusually high and in January 2006, the price was unusually low. The return for that period will be low, but not representative of how the STI ETF really performs in general – it was just a coincidence.
So in order to remove the influence of luck and timing from the analysis, I’d do another analysis between February 2001 to February 2006, then March 2001 to March 2006 and so on and map out the returns over multiple periods.
How the data will look like in table form for Rolling Returns for 5 Year Investment:
|Start Date||End Date||Return|
|January 2001||January 2006||X%|
|February 2001||February 2006||Y%|
|January 2014||January 2019||Z%|
I can do this for investment periods of 5 years, 6 years, 7 years, to as many years as I wanted as long as I have the market data.
Then when I combine these numbers, it will give us a better idea of the likelihood of returns for a particular investment vehicle, knowing that the result wasn’t just due to picking the wrong time frame or a period which just happens to support my hypothesis.
So similar to all of my previous posts on the STI ETF returns, we will be making the analysis for 5, 6, 7, 8, 9 and 10-year rolling returns over the last 11 years.
Diving into the analysis
First let’s take a look at the 5-Year Rolling Return for all of our DCA periods against the Lump Sum.
During the last 11 years, there are a total of 73 different 5-Year investment periods. After charting out the returns for 3, 6, 9, and 12 months against the lump sum for all 73 periods, here’s what we get:
Each point on the chart represents the return of the investment if you started investing on that date and invest over 5 years.
As you can see, charting it out like that allows us to compare DCA directly with Lump Sum period-for-period.
Here are some conclusions we can make from this chart:
- DCA does really well when the market was entering the recession in 2008, as you’d expect. As prices decline over a year period before bottoming out in early 2009.
- Lump Sum did really well if you happen to put all your money in at the bottom of the market in February of 2009.
- However after that, there was really very little difference in performance between Lump Sum and DCA.
Out of 73 five-year rolling periods, here are how many times each of the DCA strategy beat Lump Sum investing:
|Periods Beat (%)||41 (56.2%)||37 (50.7%)||37 (50.7%)||31 (42.5%)|
What seems to be emerging is that if you keep your DCA short, 3 months, there seems to be some benefit over Lump Sum – at least that’s the case if your investment horizon is 5 years.
Well that was just one of the investment window – 5-years – so let’s do the same for all of the other windows: 6, 7, 8, 9 and 10 years. It will take a lot of numbers and spreadsheet wrangling, but lucky for you, I’ve already done it 😉
Here are the results! First, all in chart form:
6-Year Rolling Return
7-Year Rolling Return
8-Year Rolling Return
9-Year Rolling Return
10-Year Rolling Return
Note: The reason that the charts get shorter and shorter as you increase the investment window from 5-year to 10-year is due to lack of data. I only have market data from 2008 to 2019, so once you get to 10 year windows, I can only analyse start dates in 2008 as if you start in 2010, I need price data in 2020, which isn’t available yet. (If you have STI ETF price data going back further than 2008, please let me know.)
The 10-Year Rolling Return isn’t very informative since we only really have 13 periods and all the starting dates are within 2008 – while the market was tanking – a perfect time for DCA to work really well. That’s why you see that Lump Sum underperformed DCA most of the time here.
With that being said, let’s take a look at all the numbers to see how it plays out.
Once we compile the returns of each period, for every investment strategy, there are a total of 1,032 different investment periods. The performance of each investment period were then compared with the return of lump sum investing within the same period.
This produces the below table – the percentage of investment periods where each DCA strategy beats lump sum investing:
All in all, the best strategy seems to be 3-Month DCA, which beats Lump Sum more than 56% of the time across all rolling periods. The worst looks like the 12-Month DCA, which has a beat rate of only 33% overall.
However, the performance of the 10-Year rolling window really jumps out as odd. Every DCA strategy seems to beat lump sum more often than not which matches what we saw earlier in the chart. This is because the starting period of 10-Year rolling window all lies between January 2008 to January 2009, a period when the market was on a massive downward trend, which really skews the results in the DCA strategy’s favor.
Since we don’t have data going back far enough to give a better representation of what a “normal” 10-Year return looks like, I decided to remove that set of data set:
After the removal of the higher-than-usual 10-Year rolling window, on average, all DCA strategies except the 3-Month strategy underperforms lump sum more than 50% of the time. So it looks like if you’re going to DCA, you should put your funds in over 3 months and you’re likely to overperform lump sum.
Wait, even if 3-Month DCA outperforms lump sum, but how MUCH does it outperform? Is it worth the risk?
Fair enough. We should not just look at how many times DCA beats Lump Sum strategy, we should also be looking at:
- When DCA overperforms Lump Sum, how much does it outperform?
- When DCA underperforms to Lump Sum, how much does it underperform?
We need to be calculating the average difference in returns.
If DCA overperforms Lump Sum 60% of the time and underperforms 40% of the time and when DCA beats Lump Sum, the returns are 7.5% on average (difference of 0.5%). However, if when DCA underperform it generates a return of just 1% (difference of -6%) then the average difference in returns is:
(0.6*0.005) + (0.4*-0.06) = 0.003 – 0.01 = -0.021 = -2.1%
This means that – when taken in aggregate – DCA will underperforming Lump Sum by 2.1% over all. So even if DCA beats Lump Sum 60% of the time, you’re more likely to underperform Lump Sum by 2.1% on average just due to bad timing. If that was the case, it would be better to stick to Lump Sum investing.
Great, so what are the average difference in returns for each DCA strategy?
After crunching the numbers again for all the DCA strategy and rolling windows, here’s what we got, the average difference in returns by DCA strategy and rolling window:
And the Median:
What does this tell us?
- Although 3-Month DCA seems to beat Lump Sum more than 50% of the time, but on average, the difference in percentage return is only around 0.10%. Nothing to write home about. Also since the average difference is lower than the median, when 3-Month DCA underperforms, it underperforms a bit more significantly than when it overperforms. So even though it beats Lump Sum more often, the difference isn’t significant for me to really bother with.
- Oddly enough, it looks like 6-Month DCA beat at a higher amount on average, around 0.20% even though the chances of a beat is a hair lower than 50% of the time. This results in a low median difference, all of them are negative which means more that 50% of the returns underperforms lump sum.
- 9-Month DCA has a high average difference in returns but pretty low median, which means that when the 9-Month DCA beat lump sum, it beats by a pretty high amount. However, the likelihood of beating Lump Sum with 9-Month DCA is less than 50% the longer the investment horizon, so it would be akin to gambling.
- Like always, stay away from the 12-Month DCA, the average returns are all negative.
So what’s the conclusion?
Well after all this analysis, I know what I will be doing with my bonus when I receive it: invest all of it immediately into the STI ETF as a Lump Sum and don’t bother with DCA.
- Most of the time, a DCA strategy underperforms lump sum in the last 11 years.
- On average, the average difference in returns between DCA and Lump Sum isn’t significant enough to make a big enough difference.
- DCA strategies only does really well if you happen to start right in the beginning of a massive recession like in 2008.
- I’m lazy. Lump Sum does not require me to remember to come back on the same day each month to invest again, I just do it once and be done with it.
However, in the end, if Dollar Cost Averaging provides you with peace of mind and gets you to start investing then I think DCA is definitely better than not starting at all. So if DCA gives you that, then please use DCA and sleep better at night!
I hope that this analysis was interesting for you and also helped inform your decision on how to invest your hard-earned bonus. So after all that, do you agree with my conclusion? What would you be doing with your bonus. I’d like to know!
Please let me know what you’ll be doing down in the comments below or tweet at me @firepathlion, I’d love to hear from you.
Until next time.