KSK8
561 posts
msg #159115
- Ignore KSK8 |
5/24/2022 1:43:48 PM
Another gift from THE KING...This one sports a high win rate...
Have fun with it!
Entry:
Long next day's open
Exit:
At market close
|
mursemoney
13 posts
msg #159150
- Ignore mursemoney |
6/2/2022 12:29:10 AM
Thank you my grace
|
bert123
53 posts
msg #159473
- Ignore bert123 |
8/18/2022 7:21:15 PM
King - This is a great filter. Very high percentage of 1-day pops. Is it possible for you to loosen the parameters to get more plays although the win percentage would come down? If not, thanks again for this little gem!
|
KSK8
561 posts
msg #159476
- Ignore KSK8 |
8/20/2022 8:57:25 PM
Good question Bert,
Whilst loosening the criteria does enable more plays to pop up, it unfortunately reduces the accuracy. On the other hand, increased criteria, increases the accuracy, but further decreases the # of plays.
Because of SF's limits with backtesting and because programs like Stratasearch aren't working as they used to, it's hard to experiment around with the filter's potential.
However, the remedy to increase the # of plays, whilst not decreasing the accuracy would be to decrease part of the criteria (perhaps removing the b-band condition) in conjunction with conducting technical analysis on the intraday timeframe to inspect the likelihood of a 1 bar reversal for the corresponding stock.
But if you want something that is in the wall's of SF, then deploying a S/R criteria into the filter would be the solution. However, I'm no expert with utilizing S/R in stockfetcher filters, but I'd render it the best case solution.
|
bert123
53 posts
msg #159478
- Ignore bert123 |
8/21/2022 2:44:44 AM
Thanks King. I'll take a look at eliminating criteria one variable at a time. As a day trading vehicle,
one option would be to create a watchlist of the candidates for the next day and then monitor the open with the idea of taking one or more that go green. While the BB requirement is great, it limits the plays, and since the RSI indicates an oversold condition, it may be that eliminating one or the other will still give you the oversold condition while increasing the plays.
|
KSK8
561 posts
msg #159490
- Ignore KSK8 |
8/25/2022 2:19:05 AM
If you end up furnishing a modified version, do feel free to post it here, I'd be intrigued to check it out.
|
snappyfrog
731 posts
msg #159491
- Ignore snappyfrog |
8/25/2022 6:32:11 AM
I opened this up by removing the BB, ATR and changed RSI < 5 to RSI < 10.
This give more hits and I must say, some make good day scalps, but also longer runs.
|
bert123
53 posts
msg #159492
- Ignore bert123 |
8/25/2022 12:36:03 PM
Hi King,
I haven't had the time to really dive into this as much as I'd like but I came up with some modifications that create more plays and still capture the tenets of the filter (I think). Based purely on observation, I think the most important elements of the filter are the candles leading up to the trigger day, the COG and the lower bollinger band requirement. The filter below incorporates these parameters and give plenty of opportunities almost daily.
If I were to trade the filter I would create a watchlist the night before and Wait For Green as
Muddy used to say after the Open. Furthermore, I would wait for the market noise to calm down after the Open and not jump on the trade until then. This is personal preference of course but my guess is that there is plenty of time to capture a good run on one or more of these picks as many tend to close much higher than their Open. Then it becomes how you trade it, as Rumpled One would say.
Unfortunately for me, I am shuttling grandkids to school during the first 90 minutes of the trading day (West Coast) so I can't do the kind of live study I would like to do. If I were it would involve figuring out the earliest I could enter a trade and limiting the false positives. Is it 5, 10, 15, 30 minutes after the Open? Whatever the case, the losers and winners will be much easier to identify after the market noise of the Open. Perhaps some in the Stockfetcher will experiment with this and share their findings.
Here is the modified filter and thanks again for the initial filter.
market is not OTCBB
set{v1,min(open,close)}
set{v2,max(open,close)}
set{v3,v2 - v1}
set{v5, v1 - low}
set{v4, v3 * 2}
set{v6, high - v2}
set{a1, count(v5 is greater then .2,1)}
set{a2, count(v5 is greater then v4,1)}
set{a3, count(v3 is greater then 0,1)}
set{a4, a1+a2}
set{a5, a4+a3}
/*enable the line below for increased accuracy*/
/*a5 > 2*/
set{c1, close 1 day ago}
set{o1, open 1 day ago}
set{c2, close 2 days ago}
set{o2, open 2 days ago}
set{rc1, c1-o1}
set{rc2, c1-o2}
set{x1, c1*0.02}
set{x2, c2*0.02}
set{b0, count(rc2 is below 0,1)}
set{b1, count(rc1 is below 0,1)}
set{b2, count(abs(rc1) is above abs(x1),1)}
set{b3, count(abs(rc2) is above abs(x2),1)}
set{b4, b1+b2}
set{b5, b4+b3}
set{b6, b5+b0}
b6 > 3
COG(5) > COG(5) 1 day ago
volume > 100000
close > open
open is below Lower Bollinger Band(20,2)
do not draw ATR(30)
do not draw Bollinger Band(20)
do not draw rsi(2)
do not draw b6
add column rsi(2)
sort column 5 descending
|