luc1grunt
622 posts
msg #70929
- Ignore luc1grunt |
1/24/2009 9:16:51 PM
Very insighful post gentlemen, thank you!
|
SFMc01
358 posts
msg #70946
- Ignore SFMc01 |
1/26/2009 8:32:21 AM
I have not found that one can profitably trade this scan ... as it now stands ... automatically. As a wise man says, ""it ain't what you trade, it's how you trade" This scan reports on what fills statistically but not necessarily profitably.
Other criteria need to be developed in order for this to be traded profitably. Wise traders have worked for years on developing this criteria which is often on a real-time basis.
|
Eman93
4,750 posts
msg #70969
- Ignore Eman93 |
1/26/2009 7:53:54 PM
I think if you back test on 1 min data that should be as close to live trading as one could get....
|
Walid
130 posts
msg #70970
- Ignore Walid |
1/26/2009 8:01:58 PM
Eman,
I got your email. I will print the manual and dig into the APIs. However, I may not be as fast as we both hope for. The work load these days is a little crazy .
Thanks
|
Eman93
4,750 posts
msg #70977
- Ignore Eman93
modified |
1/27/2009 12:06:52 AM
This is one filter, long only
/* 0.50 to 1.00 TRO MTC CHECKER */
Close below 100
set{ hiop, high - open }
set{ up50, count( hiop > .50, 100) }
set{ up100, count( hiop > 1.00, 100) }
set{ pct1, up100/ up50}
set{winpct, 100 * pct1 }
set{C1day, close 1 day ago}
set{GP, open-C1day}
set{Gfill, high-C1day}
set{FGL, count( Gfill > 0.05,100) }
set{GP$2.00, count( GP < 2, 100) }
and add column separator
add column up50
add column up100
add column winpct
and add column separator
add column open
add column C1day
add column GP
add column Gfill
and add column separator
add column low
add column high
add column Day Point Range(20)
add column FGL
add column GP$2.00
add column Day Range(20)
up50 above 70
average volume(90) above 1000000
sort column 7 descending
|
Eman93
4,750 posts
msg #70978
- Ignore Eman93 |
1/27/2009 12:11:09 AM
This is TROs original, I take out the nasdaq and the close is above 40......
/* TRO gap fill statistics */
set{Xgapped, open - close 1 day ago}
set{Xgapup, count(Xgapped > 0, 1)}
set{Xgapdn, count(Xgapped < 0, 1)}
set{absgap, abs(Xgapped)}
set{gappedup100, count(Xgapup > 0, 100)}
set{gappeddn100, count(Xgapdn > 0, 100)}
set{cl1lo, close 1 day ago - low}
set{hicl1, high - close 1 day ago }
set{fillup, count(cl1lo > 0, 1)}
set{filldn, count(hicl1 > 0, 1)}
set{gapupfilled, Xgapup * fillup }
set{gapdnfilled, Xgapdn * filldn }
set{gapup100f, count(gapupfilled > 0, 100)}
set{gapdn100f, count(gapdnfilled > 0, 100)}
set{gapfilled, gapup100f + gapdn100f}
/* DRAW PLOTS */
DRAW XGAPPED ON PLOT XGAPPED
/* DISPLAY COLUMNS */
and add column gappedup100
and add column gappeddn100
and add column gapfilled
and add column gapup100f
and add column gapdn100f
and add column cma(absgap, 100)
and add column absgap 100 day high
and add column absgap 100 day low
/* SELECTION CRITERIA */
CLOSE ABOVE 40
AVERAGE VOLUME(30) ABOVE 1000000
MARKET IS NASDAQ
/* SORT CRITERIA */
SORT COLUMN 7 DESCENDING
|
Eman93
4,750 posts
msg #70979
- Ignore Eman93 |
1/27/2009 12:19:25 AM
Walid
Lots of information.
Thanks,
|
Eman93
4,750 posts
msg #71015
- Ignore Eman93 |
1/28/2009 12:32:32 AM
SFMc01
- Ignore SFMc01 1/26/2009 8:32:21 AM
I have not found that one can profitably trade this scan ... as it now stands ... automatically. As a wise man says, ""it ain't what you trade, it's how you trade" This scan reports on what fills statistically but not necessarily profitably.
Other criteria need to be developed in order for this to be traded profitably. Wise traders have worked for years on developing this criteria which is often on a real-time basis.
-------------------------------------------------------------------------------------------------------------------------------------------
I think there are alot of things to be worked out,,, the back test is skewed it assums you get filled at your price every time.
we all know that is bunk, because most of the trades are in the first min of the day.... I was thinking of having it skip the first 5 min and see what happens.
Would you care to elaberate on the other criteria?
I can or I think I can see automated day trading in my charts, if you watch a 13/20 cross on a 3 min chart it pops every time on certin stocks,,, I can automate that entery pretty easy, its the exit that sucks.
Do you have any first hand knowlage on automation you would like to share?
|
SFMc01
358 posts
msg #71022
- Ignore SFMc01 |
1/28/2009 3:18:42 PM
Eman93: Thank you for your courteous response.
You said, "I can automate that entery pretty easy, its the exit that sucks" and "Do you have any first hand knowlage on automation you would like to share?"
Some really smart and experienced traders have been working on gap trading for years. Just 'Google" the words "fade the gap" and "gap trading" and you will get thousands of hits. Much can be learned by reading what others have experienced.
As you said, it is easy to code for an entry, but .... is it the right entry? The very knowledgeable TRO has written much about it here and I suggest his words of wisdom are right on target. One of the things he has said many times is ... (paraphrased)... "pick one stock" and learn to trade it. Implied in his suggestion is that automation might not apply across the board but may be possible on a stock by stock basis with specifically developed trading rules. I have not been able to prove anything different.
Steve
|
Eman93
4,750 posts
msg #71023
- Ignore Eman93 |
1/28/2009 3:51:28 PM
I was using the 10/30 cross "long" for realtime alerts "muddy style" on about 40 stocks, mostly RFR type or MTC.
I can only run the program on my home PC on Fridays...I am stuck at work the rest of the time.
I was having trouble with the program locking up while running in real time, they just came out with an upgrade I hope this will help, I can see the power of it, its like having your own trading floor all watching diffrent stocks.
I would look at the chart and make the call weather to enter the trade.
so no auto trading yet.
|