xxcheck
53 posts
msg #79439
- Ignore xxcheck |
9/16/2009 8:39:42 PM
Can anybody help setting up a basic filter for this search. Took from Trading Markets looked interesting.
Thanks Mark
The rules for the example were:
1. The ETF is above the 200 moving average.
2. Buy the ETF when the 2-period RSI closes under 2.
3. Buy a second unit it prices close lower than your first entry.
4. Exit when the 2-period RSI closes above 70.
And today we’ll look at the test results on this very easy to follow model.
We tested this model on a universe of over 100 of the most liquid 1x ETFs going back to the first day of trading of each of these ETFs. Using the above rules created the following test results:
442 trades
81.22% correct
1.44% per trade (that’s all trades, not just the winning trades).
Average hold is 5 trading days.
So there you have it. 4 simple rules looking back over 15 years and triggering over 400 signals with the more than 80% of those signals being profitable in the testing.
Strong, quantified strategies exist, especially in ETFs. The model above proves this out. And the nice thing about the above rules is that the entries and exit levels are not magic numbers. You will see healthy edges even with higher RSI entry levels and different exit levels.
The key is to keep things simple: buy on pullbacks, sell into strength, and quantify everything you do. It’s a good recipe for long term success in your trades.
|
trendscanner
265 posts
msg #79443
- Ignore trendscanner |
9/16/2009 9:55:18 PM
From what I can tell, there's no available feature on SF to differentiate ETFs from stocks.
The easiest way to do this might be to get a list of ETF symbols and put them in a watchlist, then run the RSI filter against the watch list to look for candidates.
|
durgin
60 posts
msg #79445
- Ignore durgin |
9/16/2009 10:08:49 PM
You can just say -- Show ETFs -- and you can filter 1257 of them with SF.
|
trendscanner
265 posts
msg #79449
- Ignore trendscanner |
9/17/2009 12:44:08 AM
Would it be simply this then?
|
marine2
963 posts
msg #79450
- Ignore marine2 |
9/17/2009 1:42:36 AM
Nice filter. It did backtest during my positive stock environment time frame like you backtested at. Mine was 90% though. Thanks for sharing it with us.
|
jnafach
74 posts
msg #79453
- Ignore jnafach |
9/17/2009 6:28:10 AM
I think it is nice concept but need differene exit criteria for example if you apply on sep 1 and exit when RSI 70 you would have missed a lot of profit
|
Kevin_in_GA
4,599 posts
msg #79482
- Ignore Kevin_in_GA |
9/17/2009 11:36:22 AM
xxcheck:
I would look at modifying your exit criteria. To sell when the close crosses above 70 on the RSI(2) is too soon. I would think that you could do better if you reversed this - "close when the RSI(2) crosses below 70 from above". Then it can run for a while on strength.
|
durgin
60 posts
msg #79523
- Ignore durgin |
9/17/2009 6:24:33 PM
An alternative approach is to look at the upper end (high relative strength).
RSI(2) above 99.
Performance is similar or slightly better than "below 2".
|
durgin
60 posts
msg #79527
- Ignore durgin |
9/17/2009 6:42:34 PM
|
xxcheck
53 posts
msg #79528
- Ignore xxcheck
modified |
9/17/2009 6:55:41 PM
Thanks all for the help and ideas. Kev are u out of ETFC? Durgin --- Would'nt this filter out overbought ETF'S with the RSI so high?
|