Historical Rolling Returns - DCA into STI

[nomoreliaola]

If you believe STI is low now, total returns can be slightly higher or more than 6%. Look at some of the months ending in 2014. They go up to ~4% pa excluding dividends.

You can choose not to reinvest your dividends. Your total returns is the same. Of course the final value is lower.

There are already 4 regular purchase programs in the market now (Phillip, OCBC, DBS, KE). I think it is more of a preference on which one to choose. Personally, I prefer Phillip because this was the one I participated before.

for me i no choice have to use OCBC cos thats the only one that can use SRS funds (other than poems, which charge higher)

 

[T H I N K]

Some food for thought

Your STI ETF DCA performance is very much dependent on when u exit (price of STI ETF, primary criterion) as well as when u start DCA and how u DCA (monthly start price, etc).

Look at the following examples (simulated case studies) of DCA:

Mthly $100 invested using start of mth price, valued at start of mth price after end of period, annual return using the calculator here ( http://www.calculator.net/interest-...xtrate=0&cinflationrate=0&printit=0&x=51&y=17)

1. May 2002 to April 2016 (14 years, since inception of STI ETF): capital 16,800 vs market value 19,528.75 = < 2.2%

2. May 2006 to Jun 2016 (10yrs): 12k vs 11,822.3 = loss

3. Jan 2009 to Jun 2016 (7.5yrs): 9k vs 8,867.21 = loss

4. Jan 2009 (7.5 yrs) with dividends reinvested = 9k vs 10,145.69 = < 3.2%

5. Jan 2011 to Jun 2016 (5.5yrs): 6.6k vs 6100.35 = loss

6. Jan 2011 (5.5yrs) with dividends reinvested = 6.6k vs 6742.22 = < 0.8%

7. July 1996 to Jun 2016 (20yrs): 24k vs 31368.10 = < 2.6%


Someone asked if STI continues to be range bound for the next 10-20 years, what would be the return? Take the results of 2009 (7.5yrs) add to results of 2006(10years) assuming STI rangebound in that pattern, it will still be a loss! Or rangebound with the pattern of 2011 to 2016, add 2009 and 2x 2011 results = 18.5 yrs, it is still a loss! Only consolation is your dividends? Only a simplified view of things!


If history is any indication of future performance?


The best is for people who have been doing DCA to share their portfolio results to confirm. Share your real cases to confirm the exercise being done here. Someone shared his portfolio from 2014 in another thread and read what he got in return?

(Feel free to verify my calculations, dun just depend/rely on what others calculate/post in hwz)

 

[Mecisteus]

Some food for thought

Your STI ETF DCA performance is very much dependent on when u exit (price of STI ETF, primary criterion) as well as when u start DCA and how u DCA (monthly start price, etc).
This is true. That is the reason why this exercise did a rolling study. To see all the possible returns from all possible windows of various ending months for various period.

Look at the following examples (simulated case studies) of DCA:

Mthly $100 invested using start of mth price, valued at start of mth price after end of period, annual return using the calculator here ( http://www.calculator.net/interest-...xtrate=0&cinflationrate=0&printit=0&x=51&y=17)

1. May 2002 to April 2016 (14 years, since inception of STI ETF): capital 16,800 vs market value 19,528.75 = < 2.2%

2. May 2006 to Jun 2016 (10yrs): 12k vs 11,822.3 = loss

3. Jan 2009 to Jun 2016 (7.5yrs): 9k vs 8,867.21 = loss

4. Jan 2009 (7.5 yrs) with dividends reinvested = 9k vs 10,145.69 = < 3.2%

5. Jan 2011 to Jun 2016 (5.5yrs): 6.6k vs 6100.35 = loss

6. Jan 2011 (5.5yrs) with dividends reinvested = 6.6k vs 6742.22 = < 0.8%

7. May 1996 to Jun 2016 (20yrs): 24k vs 31368.10 = < 2.6%

1) Your data above are biased. You specifically picked NON FAVORABLE windows just to prove that DCA does not work. I can also pick many more FAVORABLE windows to prove that DCA does work in the long term as what my general findings have mentioned. My FAVORABLE windows will outnumber your NON FAVORABLE windows for sure.

2) Month end or month begin? It doesn't matter. Because there is no study that can prove convincingly that the prices at month end are lower/higher and prices at month begin are lower/higher. They will be random. You can be buying at the month low prices at the beginning or you can be buying at the high prices at the beginning of the month. 1 thing for sure, you cannot be so UNLUCKY. actually it doesnt matter if you still believe that SG market is going to do well and trend is higher in the longer term. this is the basic assumption that you must have before you DCA.

3) Among all the 20-year windows, only about 5 windows ending late 2008 and early months of 2009 were generating <0% in annualized returns. That means if you had been DCA for 19.X years and it happened to be a down market, DONT try to redeem your shares at the bottom. You have 2 options. Defer your redemption, continue to DCA for the next few months OR at least defer your redemption. ie dont sell your STI ETF yet. Of course, you must believe that the market will recover after a big crash.

Someone asked if STI continues to be range bound for the next 10-20 years, what would be the return? Take the results of 2009 (7.5yrs) add to results of 2006(10years) assuming STI rangebound in that pattern, it will still be a loss! Or rangebound with the pattern of 2011 to 2016, add 2009 and 2x 2011 results = 18.5 yrs, it is still a loss! Only consolation is your dividends? Only a simplified view of things!
Again it depends on what you believe. If you are like some of the forummer like yyhwin who predicts doomsday, then don't DCA at all.

If history is any indication of future performance?
Yes. My rolling returns for various windows for a long term DCA have shown.

The best is for people who have been doing DCA to share their portfolio results to confirm. Share your real cases to confirm the exercise being done here. Someone shared his portfolio from 2014 in another thread and read what he got in return?
Again you are picking a selected window just to discredit this method. This exercise covers ALL the possible windows you can ever imagine.

See reply in red. 10 char.

 

[Mecisteus]

7. May 1996 to Jun 2016 (20yrs): 24k vs 31368.10 = < 2.6%

Thank you for confirming with my calculation.

My returns is also around the same number as you for that window. Mine is at 2.55% pa for ending in Jun 2016. The difference is because you picked the wrong starting month. May 1996 to Jun 2016 is more than 20 years. You need to verify your calculation.

1 year is 12 months.

If you start in May you must end in Apr the following year.

 

[wahkao3]

2) The 5th chart is a bit meaningless. Not supposed to combine them. It is like ranking all swimmers who are short, long distance, breast, free and butterfly together.

it answers the question: "DCA good or not? buy liao can huat? And I dunno my time horizon "

 

[Mecisteus]

it answers the question: "DCA good or not? buy liao can huat? And I dunno my time horizon "

so its like you dont know whether to swim 50m or 500m, fly, breast, free or butterfly.

 

[T H I N K]

Thank you for confirming with my calculation.

My returns is also around the same number as you for that window. Mine is at 2.55% pa for ending in Jun 2016. The difference is because you picked the wrong starting month. May 1996 to Jun 2016 is more than 20 years. You need to verify your calculation.

1 year is 12 months.

If you start in May you must end in Apr the following year.

the maths is correct, typo should be July 1996, will correct the post, mine only estimate <2.6, u confirmed with 2.55%

 

[nomoreliaola]

For me a reasonable is 20years which put me almost retiring

Sent from Samsung SM-N910G using GAGT

 

[T H I N K]

See reply in red. 10 char.

U are over suspicious and lack confidence of your analysis. I did not purposely choose those periods, some of which I have explained there.

Why I chose those periods:

1. STI ETF inception since May 2002, to show results if people had done it from 2002.

2. 2011 was a case discussed in another thread, consolidated here

3. 2006 was a case prepared by another forumer discussed in another thread, again I consolidate it here

4. 2009, as i explained in another thread, I want to show dividends reinvested. If u know, how much STI multiplied from the low, and what was the result when u continued to DCA?

5. 1996 to show a case of 20 years todate.

Sharing for people to know results of latest periods.

I believe many newbies who followed what was preached here are within these windows, so sharing what is relevant. That 2014 case was shared today, I happened to read it, not trying to discredit your DCA, am I that powerful?

(this will be my last post here)

 

[Mecisteus]

U are over suspicious and lack confidence of your analysis. I did not purposely choose those periods, some of which I have explained there.
You are just repeating the same adjectives used by the HDB man because you guys have something against me. Period.

I have done many such analysis in my financial modules. If I am not confident, I won't be posting here. Over suspicious just because I did a rolling study? Rolling studies are commonly done in financial analysis by financial analysts. It can cover all possible windows so why pick a particular window only?

Why I chose those periods:

blah blah blah

I believe many newbies who followed what was preached here are within these windows, so sharing what is relevant. That 2014 case was shared today, I happened to read it, not trying to discredit your DCA, am I that powerful?

You have numbers or facts to back up your claims? Or is it just your opinion?

see red. 10 char.

 

[Mecisteus]

(this will be my last post here)

You edited and ended your last post AFTER seeing that you can't answer to my question?

You have numbers or facts to back up your claims? Or is it just your opinion?

 

[Mecisteus]

I have already presented a very comprehensive and long term data in the 1st post. The data is for everyone to judge and make their own decision.

 

[Yachtmaster]

The essence of DCA is about leveraging on volatility of the price.

A standard deviation computation would be much more useful.

 

[Yachtmaster]

( http://www.calculator.net/interest-...xtrate=0&cinflationrate=0&printit=0&x=51&y=17)

DCA is never about straight line graphs. It's the intensity of the movements in between that does the magic.

 

[Mecisteus]

DCA is never about straight line graphs. It's the intensity of the movements in between that does the magic.

I think he is just making use of the formula from the website.

But if you understand correctly, the straight line graph just indicates how the reverse annuity like DCA money will grow when the annuity rate is found.

 

[winguy]

So long term total returns are about 6% after fees. Gone are the >=7% days.

A 80/20 portfolio (assuming 100% sti for stocks) means 4.8% contribution from stocks.

A 60/40 portfolio (assuming 100% sti for stocks) means 3.6% contribution from stocks.

With this kind of returns we just have to spend less (invest more) and work longer...

 

[Mecisteus]

So long term total returns are about 6% after fees. Gone are the >=7% days.

My numbers there don't include transaction costs yet. And also management fees. Dividend yield of 3% is also on the conservative side.

And the average annualized number is likely to go up because STI has inched up higher now compared to the closing price at the start of Jul/2016.

 

[Asphodeli]

don't forget, a whole-life can't cash out anytime you need...STI ETF you can.

for a 2.2% compounded annual interest, there aren't a lot of products that can offer you that much, allows compounding without adding in more capital, and is a low risk investment asset.

 

[Mecisteus]

Some of you might ask. What is the impact of transaction fees?

Assuming you are being charged a 1% fee to purchase monthly. ie $5 for every $500 of investment amount, your annualized returns will go down by 0.1% only. This is valid for the 20 years DCA only. This excludes sales charge.

 

[awful999]

just a question, if the any fund house or company that sell the ETF collapse before ? the chances very slim right ?