jpistell
123 posts
msg #33964
- Ignore jpistell |
11/16/2004 11:26:55 AM
yepher,
Am I missing something? ALl it returns is TZOO.
Joe
|
yepher
359 posts
msg #33965
- Ignore yepher |
11/16/2004 1:57:26 PM
;) I was just having some fun with the contest. My goal was a filter that would return the absolute best results for that period.. (I realize that is not the real intent of the contest but I did obey the rules, except the 20 matches per day one)
I guess the real point is, it is easy to make a filter that will work very well in a narrow chunk of time that will perform really lousy on a larger sample of or in the "Real World Markets". It is too easy to curve fit a filter for a given time period that has no real statistical significance and mis-lead the casual viewers.
|
rntrader7
10 posts
msg #33971
- Ignore rntrader7 |
11/16/2004 6:12:43 PM
Show stocks where MACD fast line(12,26) is approaching MACD slow line(12,26) from below
and Average Volume(90) is above 50000
and price is above 10
and MACD fast line(12,26)has been below 0 for 4 months
and MACD slow line(12,26) has been below 0 for 4 months
and RSI(10) is below 20
and date offset is 80
|
jpistell
123 posts
msg #33973
- Ignore jpistell |
11/16/2004 8:16:46 PM
rntrader7's filter at a glance:
(I grabbed a few dated randomly.)
Offset Date .. *Net% … vs. … Nasdaq% … Filter vs. Nasdaq
92 7/8/04 …… -2.96% … vs. … 7.40% ……...…. -10.4%
85 7/19/04 … 12.29% … vs. … 10.34% ………. 2.0%
80 7/26/04 … 17.77% … vs. … 13.03% ………. 4.7%
75 7/29/04 … 14.25% … vs. … 10.50% ………. 3.8%
SUM OF THE RESULT: FLAT TO SLIGHTLY AHEAD OF THE MARKET
|
jpistell
123 posts
msg #33975
- Ignore jpistell |
11/16/2004 8:41:43 PM
With this contest is rooted in the 1st screen of Dr Elder's Triple Screen system.
>>>...So-called 'modern' technical analysis uses approaches which include Dr Alexander Elder's 'Triple Screen Trading', which he outlines in his popular book Trading For a Living. The technique is very similar to the advice given to traders by W. D. Gann who suggested that traders should look at the big picture first. No doubt Gann and Elder thoroughly tested the technique and found it to be very effective - despite the fact that they undertook their research at the beginning and end of the 20th Century, respectively. <<<
http://www.marketmasters.com.au/80.0.html
There is no one way to build a filter, but one way to improve your filter profits is to Identify Long term Trend then wait for the pull back and let your "Trigger" catch the turn.
Elder's triple screen was
Weekly charts (for long term trends)
Daily Charts (for the pull back)
and Intraday (for the trigger).
Real Trading is managing our errors. If you build a well constructed long term 1st screen, your can tinker with trigger filters to enter a trade. Whats nice is that losses on a bad trigger should be less as you're trading with the strong uptrend, and, if your trigger missed the call, THE STRONG UPTREND WILL SAVE YOUR BUTT and give you an opportunity to bust out with small(er) losses.
May the Filter Be WIth You...
Joe
|
jpistell
123 posts
msg #33976
- Ignore jpistell |
11/16/2004 8:49:03 PM
jpistell's Filter Results:
Offset Date .. *Net% … vs. … Nasdaq% … Filter vs. Nasdaq
92 7/8/04 …… 13.76% … vs. … 7.40% ……...…. 6.4%
85 7/19/04 … 23.14% … vs. … 10.34% ………. 12.8%
80 7/26/04 … 32.30% … vs. … 13.03% ………. 19.3%
75 7/29/04 … 27.62% … vs. … 10.50% ………. 17.1%
SUMMARY: Well ahead of the market > 2:1
|
james549
53 posts
msg #33980
- Ignore james549 |
11/17/2004 12:08:06 AM
I'll give this a try.
|
defghca
150 posts
msg #33981
- Ignore defghca |
11/17/2004 7:02:59 AM
jpistell
why the 80 days offset?
|
jpistell
123 posts
msg #33985
- Ignore jpistell |
11/17/2004 12:10:52 PM
defgha,
Requirements:
Filter must produce at least 20 matching stocks
Stocks > $4
Avg Vol > 50k
Date SetBack: Anywhere from 75 to 90 days (July 8 to July 29, 2004)
We're looking for long term setups.
#1). This set back period allows for plenty of time for the filter to produce a profit.
#2). I've placed the set back time right in the middle of a recent market pull back so it'll place a premium on the filter writers scripting ability.
#3). To prevent a filter writer from optimizing on a single setback day, I offer a range of dates to prepare for.
As a side benefit, this contest will help everyone tune into the power of the SF's back testing ability. Build your script and then test it into a bear market. If it keeps your cash safe even in a bear market, you've got a winner.
Joe
|
jpistell
123 posts
msg #33988
- Ignore jpistell |
11/17/2004 12:26:15 PM
james549,
I just took a quick look, your filter looks very good indeed! I especially like the strong win/loss ratio AND the excellent 1st week performance. I also like how small your draw down is.
If you want a good laugh, scroll thru the weeks to either side of our set back. Some where around 120 days or so, you'll see NO hits. Then, as teh market rolls over, the hit count goes higher and higher. Even as the market falls apart, your filter hangs in there.
IMO, I'm saving you filter for use when I think the markets falling and I am hoping for a turn.
Great stuff!
Joe
james549:
|