Joules360
30 posts
msg #33276
- Ignore Joules360 |
9/16/2004 6:29:08 PM
Ok..one more Tweak...thanks to all of you that helped me build this screen without reading the whole 149 page instruction manual...hehe... this is probably the best I can do with it for now...look at the charts...the daily stochastic is just turning up as the weekly stays nailed to the ceiling at least above 70...They can be changed to 60 or 50 get more picks..or tighted to 80 or 90 to be more selective. Ok here is the new and improved:
Joules360 SlingShot
|
Joules360
30 posts
msg #33277
- Ignore Joules360 |
9/16/2004 8:07:35 PM
By the way... for those of you that can short...reversing the slingshot parameters for shorts works just as well...maybe even better....You can go with long slingshots on up days...and short slingshots on down days...make the determination at the open...makes sense to me. As of this moment...I can only go long with this diversiation style broker.
Joules360
|
cegis
235 posts
msg #33291
- Ignore cegis |
9/17/2004 9:57:59 AM
Joules360,
Here's a way to do a short term backtest of the filter: Use, for example, "10 days ago" on your criteria (11 days ago in place of the 1 day ago...). This will give you "future data", in that you will have data available to you for 9 days ago (1 day into "future"), 8 days ago (2 days into "future"), etc. You can then use that to calculate potential returns using set{}, and show them in the results using add column. Here's your filter showing a 10 day "worst case" returns:
I say this is "worst case" in that it assumes I bought the stock at the high the day following a match on the filter, and sold at the low of the 10th day (first set command).
One thing I noticed about this, though, is that using offset gave a different list of stocks than using days ago (actually, days ago was a subset) when run for "10 days ago". I think this may be due to the fact that we're using a weekly value, which doesn't "slide" on a daily basis like daily values do, but I have not determined that for a fact yet.
Obviously, you can use just about anything you want for the buy and sell values, and do the calculations over any period. You can also use the count() function to determine what day you would have sold, and get a more accurate picture of your trading program. (I'll leave this as an exercise. I believe there are other posts by me that show this. This sometimes runs into the SF limit for nesting set{}s, so it may or may not work.)
Perhaps I'll make a post fully describing this method sometime soon.
HTH,
C
|
Joules360
30 posts
msg #33293
- Ignore Joules360 |
9/17/2004 11:56:31 AM
Cegis:
Thankyou....Its hard to do a backtest..I know...because of the variables involved...If I see a stock not performing back into its uptrend after the pullback...I usually dump it....
Live trading will always be different...but all I want is 1% net gain avg from all the scanned picks..so far I see that it is attainable...as long as we get the majority moving up....and drop the losers quickly....we will double our money every 2 years...
Its seems slow for a daytrader...hehe....but nice for a swingtrader....Once you develop a nice size account...this will far exceed the profits of a daytrader...for one...all of the money is invested and working at the same time. When I daytrade..I only use a small portion at a time to make my profits...the potential return is greater daytrading..but so is the risk...
I like this type of safe diversification.
Thanks!
Joules360
|
EWZuber
1,373 posts
msg #33302
- Ignore EWZuber |
9/18/2004 6:44:37 AM
I probably have a dozen variations on the filter you just posted with several different phaseing relationships.
Got a book by David Floyd on Daytrading. Over the past ~ 5 years he has annual returns of triple and quadruple digit percentage gains trading 4 hours a day.
|
Joules360
30 posts
msg #33309
- Ignore Joules360 |
9/18/2004 9:20:34 PM
EWZuber:
Thanks for the feedback....its encouraging to hear that...All I want is a safe and steady 1% week net income..I'll keep you updated on the weekly progress.
Joules360
|