sulldt
31 posts
msg #43081
- Ignore sulldt |
4/20/2006 9:53:54 AM
Since this is for general discussion I thought I'd generally discuss. I've been subscribed to S.F. for a little over a year and am interested in opinions on something I did back before the meltdown in '01. It was really simple, if I had had S.F. the filter would have been "Show stocks where close gained 100% in the last month and average volume(30) is above 100000 and close between 1 and 5". I got tons of stocks back then that did that and all I had to do was follow them until they quit going up and started to retreat, I'd wait until they retreated to 66% of whatever that high might have been and then buy. Sometimes the rebound was the same day but generally I kept for 1 or 2 weeks and sold. If I was in it for a month I would sell, figuring I had made a mistake. I had 5,000 when I started doing that, 10 months latter I had 65,000. Biggest mistake was buyiny 2,000 shares of something for 1.375 and selling 2 weeks latter for 2.75, 10 weeks latter, right before the crash it hit 48.00 a share. I had a 10 week run where I made money 30 out of 31 trades. I sure wish it had lasted for another year or two but it didn't and it hasn't worked since but now I'm starting to re-look at that approach. The filter now reads "Show stocks where close is 75% above close 1 month ago and average volume(30) is above 100000 and ema(13) is above 0". I use ema(13) now because it generaly appears to be about the same as 66% of the high. With S.F. you can run that filter and then back it up week by week or day by day for as far back as you want to go until you find one that has dropped below the ema(13).The only problem I have with it now is it only works about half the time. I've looked at all kinds of other indicators to add to that basic filter to increase the winning percentage but as yet haven't found any.If anyone has any ideas as to what indicator or indicators that could be added I would appreciate their input. That is if there is anything that could be added to that filter that would improve it. Don't really know if that is possible.
|
heyen
124 posts
msg #43082
- Ignore heyen |
4/20/2006 11:18:04 AM
One possible problem cause sudden jumps in price. Rather increasing interest in a stock causes its run to continue.
|
guru_trader
485 posts
msg #43084
- Ignore guru_trader
modified |
4/20/2006 1:53:52 PM
Here is your filter tweaked in the backtester for the following dates: 12/30/2005 - 4/19/2006
Exit on day: 9
Win% = 100
Lose% = 0
R/R = 0
ROI% = 367.87
Unfortunately, the backtest doesn't hold up going 2 years back.
|
rtucker
318 posts
msg #43093
- Ignore rtucker
modified |
4/20/2006 4:17:59 PM
This a flexible method to define the retracement from the high rather than using a pullback to ma(xx). .318 is one number that is looked at ,and, close to your 33% retracement requirement. Above allows 30% to 40% retracement from high.
|
sulldt
31 posts
msg #43095
- Ignore sulldt |
4/20/2006 4:22:55 PM
Thanks guys!
|
marmac
2 posts
msg #43158
- Ignore marmac |
4/23/2006 4:13:39 PM
SULLDT> I do basically what you do but I use my buy point when slow stoch(13,3) fast turns up to meet slow below 25 and rsi(15) also turned up that day. See NECX 20 days ago for what I look for.
|
EWZuber
1,373 posts
msg #43172
- Ignore EWZuber |
4/24/2006 3:53:47 AM
You may find that this type of filter only works adequately when the index involved is under accumulation. Once the Monthly Chart moves into a distribution phase typically all you will get is short violent pops to test resistance then a resumption of the downtrend.
To play long positions under these circumstances is to trade against the longer term trend. Not much slop is tolerated in entries or exits or the market will eat your lunch, IMO.
|