taylorsk 106 posts msg #97725 - Ignore taylorsk |
12/2/2010 10:54:25 AM
Hey RumpledOne...
Found an error in my logic on the statistics filter looking for stocks that gap and looking for stocks that fill the gap but do not close in the direction of the gap.
I was using the code
set{gapuphold5, count(close > open, 5)}
set{gapdownhold5, count(close < open, 5)}
using the 5 period so I could manually check the numbers. This returns all days that the close>open or close< open and not just the gap days. .
I am trying to figure out a way to count only days that a stock closes in the direction of the gap
I have read the AND OR screens you wrote, but cannot get it to work or at least Have not as of yet figured it out to work.....
What I am trying to figure is
If a stock gaps up how often does it fill the gap down and not close higher
If a stock gaps down how often does it fill the gap up and not close lower
I would of coarse like to tie this into your overall Fade the Gap filter.....
Here is the 5 period scan I use to manually check things.
|
taylorsk 106 posts msg #97762 - Ignore taylorsk |
12/3/2010 9:53:37 AM
Further reaserch looks like I cant do the above due to the nesting issue.
Off to Exel I go.
|
TheRumpledOne 6,529 posts msg #98039 - Ignore TheRumpledOne modified |
12/17/2010 8:06:32 AM
|
TheRumpledOne 6,529 posts msg #98060 - Ignore TheRumpledOne |
12/17/2010 12:55:29 PM
INTERESTING...
http://www.sfomag.com/eSFO/eSFO2010_12.aspx?pg=49
|
TheRumpledOne 6,529 posts msg #98674 - Ignore TheRumpledOne modified |
1/20/2011 4:16:23 PM
Buy at market close, sell at the open the next session.
Statistics in your favor.
MAY ALL YOUR FILLS BE COMPLETE.
P.S. THEN YOU CAN FADE THE GAP... LOL!!
|
mahkoh 1,065 posts msg #98705 - Ignore mahkoh |
1/22/2011 6:51:31 PM
Funny that there are 37 pages of posting without anyone coming up with this approach to the filter!
A thing to consider though: When selling at the open the spread can be quite large unless you trade highly liquid stocks. I used settings that look for stocks between 5 and 50 and avg 30 day volume above 3 million. The highest gappedup returned is 72 times out of 100, which would hardly qualify as a candidate in the regular use of the filter.
As a matter of fact when using the filter this way wouldn't you prefer stocks that gapped up but tend to NOT fill the gap afterwards?
|
mahkoh 1,065 posts msg #98715 - Ignore mahkoh modified |
1/23/2011 3:03:14 PM
Here is a modified MTC filter that can be used for the "buy at the close-sell at the open filter"
After all when playing this strategy I would not at all be interested in whether the gap is to be filled or not, the edge would be that the stock has a statistical tendency to gap up more often than to gap down. Run the filter and sort the gapratio column descending. The higher the better, sort ascending for "short at the close-cover at the open" strategy.
Note that MAabsgap(30) and absgap-6weekhigh are in percentage in this filter. The columns filluppct and filldnpct should officially be multiplied by 100 but I never bothered.
|
TheRumpledOne 6,529 posts msg #98725 - Ignore TheRumpledOne modified |
1/24/2011 10:28:43 AM
|
TheRumpledOne 6,529 posts msg #98729 - Ignore TheRumpledOne modified |
1/24/2011 5:05:15 PM
Added last 5 days results columns.
|
TheRumpledOne 6,529 posts msg #98836 - Ignore TheRumpledOne |
2/1/2011 1:44:45 PM
http://www.tradingacademy.com/lessons/20110201/featured_article.htm
Meanwhile, we just MILK THE COWS everyday.
|