Comparing Index ETFs to their Indexes for trading signals

I like to test strategies using 20 years of data utilizing ETFs based on Indexes.

Seems like a number of ETFs based on non US based assets, don't have data going back to the year 2000.

I tested on these Indexes and their related ETFs. However, I am getting signals different for both, some more differences than others. When viewing charts for both the Indexes and related ETFs I do see minor differences. Not sure how to account for this:



So I am not confident to base my ETF trading that is different for this. For my testing I don't have enough ETF data to give me confidence. My only thought is to skip these ETFs. Any suggestions. I don't want to use less years data.

Thank you,
Larry

 

better to compare the ETF to the iNAV of the etf.

 

If your ETF prices are adjusted for splits and dividends, that could change signals.

Another possible reason is your signal generator is too sensitive in some way. One way to test for this is compare signals on ETFs that are highly correlated. For example, one would assume timing signals for ETFs that track the same stock index (e.g., IVV VOO SPY) or exclusively hold the same precious metal (e.g., GLD GLDM IAU SGOL OUNZ) would be close to identical. If the signals are for when to buy or sell are too different for a set of almost identical ETFs, your signal generator might have issues.

 

Ph,
What an insightful question on the signal generator. So if data is different, of course different results. (Duh... lol).

For my testing, I don't want to trust an index going back 20 years data with its base of an ETF going back 8 years. What I also did was test the index and the ETF going back the shorter period to see the results. Looks like some exact trades, but a number of them are off by a bit.

Your solution from the other thread:
https://www.elitetrader.com/et/thre...history-on-yahoo-finance.371134/#post-5717604

Finding a Web screener to find ETFs by their inception date, saved me a lot of headaches. My initial testing did not give me the results I was looking for.

Thank you,
Larry

 

Two possible reasons I can think of:

The index closes at a different time than the ETF, especially when the index and the ETF trade in different time zones.

I'm assuming the index is priced in each country's local currency whereas the US trade ETF is priced in US Dollars. Could currency effects account for the discrepancies?