Backtesting DCA methodology
Hi,
Thank you Shiny for your wonderful advice and response. You are a shiny stone in this forum

. I have learnt plenty from your posts.
I've been reading the thread and very interested in the Shiny's suggestion for ES3, A35 and IDWA portfolio. It seems to me the essence of this strategy is DCA. By adding money monthly, and then re-balance yearly.
I decided to back testing the DCA methodology for STI index using excel from 1990 to 2015 to gauge the return we expect why applying this methodology. I think maybe someone already did this, but I can't find so I do it myself.
I'm sharing the excel below. One is Microsoft excel and the other one is google speadsheet (some formula are not working on google spreadsheet since I create from excel)
Excel:
http://tinyurl.com/p6ean2y
G spreadsheet:
http://tinyurl.com/onur45x
Setup:
- 10000 Initial fund
- Monthly contribution $1500
- Commission $25
- Dividends are not re-invested
- No-rebalancing
- Database from Yahoo finance
I don't know how accurate my modelling is. If you find any faults, please let me know. I'm happy to re-visit it.
The result from the exercise suggests a
IRR of ~4% (excluding dividend). The reasone IRR is used instead of CAGR is because CAGR we are keep adding funds to the investment, so CAGR will not be accurate. Are there other ways to calculate return?
- If we include dividend, can I say that we could add about ~2.5-3% on top of 4%? Hence, 6.5% annual return (including dividend).
- This modeling does not have any bond component and re-balancing aspect. If these 2 are added, could the total return improve another 1% or would it actually get worse?
Another findings is that it does matter if we DCA monthly or quarterly. The result are quite similar. So, we can actually DCA quarterly to reduce maintenance effort. Does it make sense?
I also back-tested with IJR (iShares S&P SmallCap 600 Index). IJR has been consistently outperforming S&P500 or any other index by a
BIG margin in the past except this year. In the long run it will. Since 2000, IJR has gained a whopping 266% compared to a mere 52% on SPY. IWDA is quite new, but the result seems comparable with SPY.
See the performancce here
http://tinyurl.com/qzqzv6k
The back-testing shows that IJR yields an amazing
IRR of 11.1% excluding dividend. I have tried on SPY or IWDA but I think the performance may not be as good as IJR. someone can try.
I know Shiny suggests IWDA because it's listed on LSE which allows us to enjoy lower tax compared to US ETF. However, the amazing return in the long run could well compensate for the 30% dividend withholding tax on US ETF, doesn't it? or am I missing something here?
From this result, do you think it's make it's make sense to have IJR in our portfolio?
Regards,