brank20 41 posts msg #92758 - Ignore brank20 modified |
5/14/2010 10:56:36 PM
I just think of all kinds of random ways to measure the performance of a stock, heres one of them. Not sure if it will help anyone or not, nor do I know how well this filter will perform, but here it is anyway. Feel free to adjust the scoring/ranking as you see fit. I plan to incorporate actual performance values into it like how much it outperformed/underperformed versus the index. Right now its strickly did the stock/index close up or down.
|
Kevin_in_GA 4,599 posts msg #92809 - Ignore Kevin_in_GA |
5/16/2010 9:31:37 AM
No, you're not crazy. This is a clever way to look at stocks - in essence you are trying to quantify what most investors look at qualitatively.
I like your coding structure and the fact that you are careful and complete in your annotation of what you are doing at each step. That is the sign of a good programmer and a good mentor.
After running this filter against some others I look at, my only comment on its approach is that it is semi-quantitative (it gives a score based on the occurance of an event, but ignores the magnitude of the event). A stock might close one penny up against a small drop in the ^INX, and it gets three points, but if it drops 10% against a 0.01% drop in the same index, it gets one point rather than it raising a big red flag. Unlike other statistical filters here (MTC, 5% A WEEK/CROCK POT, RFR) it is not looking at a minimum performance threshold that one would bet on.
This is a momentum indicator, and when I compare it against other momentum indicators such as relative strength(^INX,45) I think it does not give as solid a signal. I might also add in a minimum volume requirement to weed out stocks with poor liquidity.
Thanks for sharing this.
Kevin
|
four 5,087 posts msg #92810 - Ignore four |
5/16/2010 10:30:08 AM
Thank you (Both of you).
|
brank20 41 posts msg #92818 - Ignore brank20 modified |
5/16/2010 1:57:05 PM
Thanks for the comments. Ill take what you said and improve on it when i have a few minutes to spare. I might have to upgrade to the premium as im beginning to run into the performance error when doing certain things with this filter, like adding the line score > 100 as a filter.
Ill see what i can do to measure magnitude properly.
|
reels 29 posts msg #92828 - Ignore reels |
5/16/2010 8:09:31 PM
hi Brank20,
------
/*Scoring System - If a stock closes down, when a major index is closed up, it scores 0 points for negatively closing against an up day on the index*/
set{var7, iusd * 0}
set{var8, var4 + var5}
set{var9, var6 + var8}
set{totals, var9 - var7} .....
------
iusd*0 is always 0 so, how do var7 counts on totals?
is iusd*1 weight correct?
thanks
Reels
|
brank20 41 posts msg #92856 - Ignore brank20 |
5/17/2010 1:36:47 PM
Good catch. I use different scoring sometimes. Replacing the zero with a different number will give you a negative drawback on the score if you want to penalize stocks that close down against an up index. It was intentionally coded that way to give the end user the ability to customize the penalty for performing poorly.
|
ChizzleMeTimbers 36 posts msg #92919 - Ignore ChizzleMeTimbers |
5/20/2010 8:52:03 AM
ur crazy
|