Double ticker in PascalA_list

Pascal,

Since a few days there is a ticker double in the repository file: CLNE.
Is it easy for you to delete one of them?
This double row causes my import-program some (minor) problems.

PdP

OK. I just removed this.
Feel free to use private mails for minor issues.

By the way, I am now working on the XLX 20DMF correlation to try to find out the best XLX to invest in long or short when a 20DMF signal is issued.
The table bellow is interesting.

The top part is for long trades. The bottom show short trades.
In yellow, we have the 20DMF trades and in green the trades using "the best" XLX ETF.
We can see that the difference is almost always positive (XLX show better returns), but mostly in 2008/2009.
Lets' not forget that the XLX correlation is higher than 90%, while the SPY/GDX correlation is often lower than 50%.

I believe that this is mainly a volatility issue: many ETFs are more volatile (because less diversified than SPY). Hence, when the market moves much in one or another direction, a specific ETF will also move more.

Note that the dates refer to the trades that started during the specific year. Hence, the last 20DMF long is still continuing now, even if its started in 2012.

I will continue that work and see if the ETF selection can be improved


Pascal.

[ATTACH=CONFIG]16990[/ATTACH]

Pascal,

I am still struggling with my strategy.
I believe one of the answers is an improved ETF selection as you have mentioned in your post on January 19.
Have you made any progress on this research?

Regards

PdP-Brugge

[QUOTE=pdp-brugge;24808]Pascal,

I am still struggling with my strategy.
I believe one of the answers is an improved ETF selection as you have mentioned in your post on January 19.
Have you made any progress on this research?

Regards

PdP-Brugge[/QUOTE]

I am sorry, I put this project on the back-burner simply because we are putting our efforts on a stock based alert system to detect in real-time the stocks that attract good money (based on the Thrust analysis.)

Since we have an EV_RT system under test, we want to add alerts regarding a list of interesting stocks.

However, regarding the ETF selection, I compiled the table below, which is simply adding long/short trades per year.
We can see that since 2010, trading ETFs has only slightly been better than trading the 20DMF. Since I use SDS/SSO to trade the 20DMF signals, this means that I can easily have a better return than trading separate ETFs. Of course, people can also go to triple leverage on the ETFs, but I do not do that.

On the positive side (for ETFs), we can see that in 2012, the 20DMF underperformed the S&P500, but trading the ETFs outperformed.

[ATTACH=CONFIG]17194[/ATTACH]

I will come back to this project with the following idea: since the best ETFs signals are always Buy Oversold and Short Overbought signals, I will combine these to the 20DMF and see if we get better returns.



Pascal

I did a quick work on these Overbought/Oversold signal trades when the 20DMF also issues a Buy/Short signal.
I only did the comparison from 2010, but it is clear that this is a viable strategy.
If produces better returns than the 20DMF, but not overwhelmingly better.
This means that if you use leveraged ETFs, then you'll do pretty well.



Pascal

[ATTACH=CONFIG]17195[/ATTACH]