lctrcbddha 28 posts msg #68667 - Ignore lctrcbddha modified |
10/22/2008 10:26:02 AM
Many ways to skin a "gap"...
This is just a re-implementation of TRO's excellent "Fade the Gap Statistics" report, with some bells a whistles added.
I set out to add some modifications of my own to TRO's original. Eventually I started from scratch, re-engineering the internal logic as to how the statistics are derived. But the end result is not the filter I wanted it to be. My initial implementation turned out to have significant logic issues, and was scrapped. This version is inelegant and, frankly, not as good as TRO's original.
Along the way I added a couple of user-adjustable knobs in order to get statistics for more cases than just open-to-previous-close. The gap offset tuneable, such that a gap must be of a certain size to be considered. Also tuneable is how far within the previous close the stock must reach to be considered a successful fill. These can be combined, for example to show the success rate of gaps of at least 1.00 that filled to within .50 of the previous close. See the notes within the filter for details.
Hopefully some will find this report useful. Many thanks to TRO for the inspiration and the original filter.
|
lctrcbddha 28 posts msg #68710 - Ignore lctrcbddha modified |
10/23/2008 10:44:52 AM
Bugs fixed by starting from scratch again. This one works accurately, but the internal logic is awfully inelegant.
Another knob I'd like to add, but am not up to at the moment, is to report how many times the stock moved at least x.xx in the direction of the gap from the open. This is different than the "to within x.xx of the previous close" knob already in the report.
|
lctrcbddha 28 posts msg #68856 - Ignore lctrcbddha modified |
10/28/2008 12:42:53 PM
Version 2:
* Added a "minimum profit" knob.
* Combined knobs such that gaps down and up don't have to be defined separately.
* Reports display both fill successes and minimum profit successes.
* Renamed variables, hopefully they will make more sense to anyone modifying, reverse engineering or troubleshooting.
|
chetron 2,817 posts msg #68857 - Ignore chetron |
10/28/2008 12:54:52 PM
thanx for sharing.
|
TheRumpledOne 6,411 posts msg #68858 - Ignore TheRumpledOne modified |
10/28/2008 12:59:24 PM
Thanks for sharing.
What NEW information does this show?
I can't seem to figure it out.
I look at AAPL and don't see anything new.
Not busting your chops, just asking for clarification.
Thanks.
|
lctrcbddha 28 posts msg #68859 - Ignore lctrcbddha |
10/28/2008 1:17:40 PM
By default, there is no new information. It mimics your report, and will show the same information.
But if you edit any or all of the variables min_gap, min_fill, min_profit you'll get a whole new world of statistics.
|