voidcomp
23 posts
msg #107431
- Ignore voidcomp |
8/7/2012 3:01:44 PM
How does this do with weekly periodicity, understanding that drawdown and return would probably be compromised in favor of fewer trades?
|
duke56468
683 posts
msg #107445
- Ignore duke56468 |
8/7/2012 9:27:49 PM
Shouldn't I be able to sort on column 5 descending, and have either SSO or SDS pop to the top? I thought this would help in backtesting. Didn't seem to work.
|
Kevin_in_GA
4,599 posts
msg #107448
- Ignore Kevin_in_GA |
8/7/2012 10:20:10 PM
By adding the line
ADD COLUMN SEPARATOR
before "add column go_long", the separator becomes column 5. Try sorting on column 6 descending.
|
olathegolf
119 posts
msg #107450
- Ignore olathegolf
modified |
8/7/2012 11:31:16 PM
Nice, Kevin. Thanks.
Anyone interested in backtesting manually can download (and modify) a stochastics spreadsheet from: http://stochastics.weebly.com/
**Update: Looks like this is not free but you may find the link useful.**
|
Kevin_in_GA
4,599 posts
msg #107463
- Ignore Kevin_in_GA |
8/8/2012 10:59:19 AM
Looking at today's values from stockcharts.com, freestockcharts.com and here on SF, the system will signal a move to go long on SSO at the end of the day (assuming no massive drop that would keep the current SDS trade going).
|
springhill
74 posts
msg #107473
- Ignore springhill |
8/8/2012 5:35:51 PM
Kevin,
Looks like SSO triggered long today.
But,,, it looks very oversold with Stochs above 75 and RSI(2) above 90
What do you think?
|
Kevin_in_GA
4,599 posts
msg #107474
- Ignore Kevin_in_GA |
8/8/2012 5:56:39 PM
I will be buying SSO tomorrow at the open, unless there is a gap down that would get me in at a better price.
I have been skeptical of this rally for a while, looking at historical trends for this time of year and also relying on my RS systems, which have me in bonds. But this approach has consistently delivered YTD with a success rate of 75%, and is now up over 50%. We could argue about the meaning of RSI(2) over 90 or any other indicator all day, but you really can't argue with actual performance.
These trades typically last 3-6 days. Any indicator can stay overbought for that long, or longer.
|
tconte
13 posts
msg #107475
- Ignore tconte |
8/8/2012 6:08:31 PM
Hello Kevin,
Thank you for sharing this filter.
I am pretty new at this, but I think that for SF Backtest to work, it needs to have in the Advance section "Maximum Entries Per Day" set to "2", so that you can see the two positions being switched on the same day at their respective open price, and "Maximum Open Positions" set to "1", so that you only end up long into one position. It is still not perfect, because sometimes both signals are long, or both signals are short, and SF misses the trade due to the fact that the exit crossover did not happen before or at the same time as the entry crossover. Nevertheless, very impressive results.
|
tombrown1
61 posts
msg #107476
- Ignore tombrown1 |
8/8/2012 9:38:20 PM
Sorry for the rookie question, but I haven't spent any time with stochastics. What are the three inputs represented by Stochastics %D(5,1,5)? Is that the K period, the D period, and the smoothing period?
|
drew9
171 posts
msg #107477
- Ignore drew9 |
8/8/2012 10:17:51 PM
Hi Kevin, you may want to try this as well with SPXL and SPXS? I entered SPXS near the close today even though this triggered this afternoon. When looking at the times when it fails, it tends to be when both lines merge near the top or bottom. They are right on top of each other now in the upper range. If you look back to the June 8th short, I think this is a repeat of what happened then. When the crosses occur near the middle, they are dead on with even a 1 or 2 day advance notice. With Bollinger, RSI and MACD all near the top, I just could not go long. MACD(2,3,1) diverged today and it has been very accurate lately. While this certainly could keep running, I sure hope I am right! Again, I really do like this as another tool. Thanks again and good trading!
|