Kevin_in_GA
4,599 posts
msg #115971
- Ignore Kevin_in_GA |
10/16/2013 2:13:33 PM
set{root,pow(close,0.25)}
++++
This is not the square root of the close, it is the 4th root (the square root of the square root). You need to say
set{root,pow(close,0.5)}
for it to be the square root.
|
dknoonan
27 posts
msg #116205
- Ignore dknoonan |
10/23/2013 11:11:09 PM
Today this filter flags YHOO and FFIV. Before that, there have been 8 days where nothing has been flagged. In order to generate more hits I've tried to relax the requirements, by making it select on ROC(7) < -1.5 instead of the original -2. Also played with ROC(80) > 15 instead of 20. Neither garnered many more hits. Any thoughts?
|
Kevin_in_GA
4,599 posts
msg #116207
- Ignore Kevin_in_GA |
10/23/2013 11:58:51 PM
Perhaps relaxing the longer term ROC might help. However, changing the filter parameters just to generate trades invalidates the results upon which the filter is based.
I systematically evaluated over 7000 combinations, and this was among the most profitable - but sure, go ahead and change it randomly. I'm sure you'll improve it.
|
mahkoh
1,065 posts
msg #116240
- Ignore mahkoh
modified |
10/25/2013 4:38:26 AM
I believe the basic idea behind this filter was to buy stocks in an uptrend that have pulled back recently.
Loosening the uptrend restrictions would require you to look at the chart and visually confirm that it is trending up.
And while you're there, I would make sure that the uptrend has not been broken..
|
tonyctl
7 posts
msg #116369
- Ignore tonyctl |
10/31/2013 3:25:08 AM
I notice the system sometimes generates a pick and then a few days later generates the same pick again, usually at a lower price point (eg. YHOO and MYL). Given maximum number of positions held hasn't been reached and no exit trigger generated, would "doubling down" when the stock is picked the second time be the right thing to do if one is to follow the system completely?
|