4 Responses to “Stock Screens”

1. Mac Says:
   

   Thanks for the updates, Danny. I like the 11 days in a row screen - I should have thought of something like that for Telechart - it would be pretty easy to set up.

   You may want to look at CPSL - if you’re looking for a spec play that could run more on Monday, I think this looks a bit better than NED. I entered at the end of the day. Strictly short-term, however.

   Any insight as to how you set up the oversold scan? There is an overbought/oversold scan I use in Telechart that I found in one of the Worden Reports, and it works OK.

   Also curious if you use some of Pradeep Bonde’s scans in the “Time Machine” - it sounded similar and I use his scans in a “market monitor” type indicator (as he calls it). I am not a master of interpreting the signals yet like Pradeep is. Thanks.

   May 16th, 2008 at 4:32:16 pm
2. Danny Says:
   

   yeah cpsl was flying today, but I didn’t make any trades today. It really depends on if china complex keeps going. That’s why I don’t trust some of these breakouts.

   the oversold scan is a couple of different scans.

   First, simply, I do 5 down days in a row. That gives me a lot of hits, maybe 200. Then I weed out the bad earnings, bear stearns fiascos et al. I do some research on why the company is shitting a brick. I visually judge.

   Then, in blocks, I set up a comparison scan that is more complicated. I look at the RSI, and color code the chart to indicate when the price has moved outside of extreme boundaries (I believe 3 and 97 on RSI(2)). I sort this by sector, the rank of the stock within the sub industry, and versus the market’s performance those days.

   This usually leaves me with about 20 candidates to eyeball or investigate further.

   Then, like in the case of AMCN, you get an awesome bounce.

   GENC was a former IBD high flyer that fucked up earnings and lost a director (?) violating ndaq listing reqs. So that is a dicey play. I was hesitant to mention it, but like I said in the post I didn’t have much time to assess my scans at that point, but I wanted to make the post.

   GENC is shitty, but it is extremely oversold. Companies with problems like these tend to plunge further, in which case I am stopped out at the previous day’s low, or bounce nicely. I usually figure 1-2% downside, 5-9% upside, which to me is good risk reward.

   I can get deeper into how I construct the oversold scans and other buy scans, perhaps in another post.

   Re Pradeep, I think he has a lot of great info, but his methods and scans are different. His emphasis is on the stocks coming from his custom screens, and the resultant picks, which he outlines for his subscribers.

   I’ve ran his scans in TC, the ones that I have access to, and they do find a lot of similar stocks, as some of the parameters are similar. If you look in the archives of his free blog, I he gives some away in there. Generally all high volume, high percentage scans will find a lot of his stocks. Same thing with kirk report, who doesn’t share his scans, but I have kind of reverse programmed based on his results. he uses more fundamental screening in his approach, vs pradeep which is totally technical.

   Me, I prefer the fundamentally technical approach. Haha.

   TC is really awesome for scans. Really. You just have to have an idea of what to search for, which thankfully I do.

   Pradeep clearly designed his market monitor in the same vein as worden’s t2s. (they are all count based indicators).

   So in that sense they are similar, and they measure some similar things in terms of stocks making big moves, however, my scans are different, more critical, and the machine is structurally different.

   Using TC, you have to tabulate the data each day. So his “MM” goes backs far as he has data. Using blocks, I can run the scans for decades, and add a LOT more relevant criteria.

   I can judge the rarity of certain readings, the ratio and cumulative totals of others. It is really just phenomenal. Then you plot it against the sp500, and see if anything matches up.

   After many many hours, I found some extremely interesting relationships in some of the ideas I had, so I developed them further.

   While I don’t think I’ve reinvented the wheel, I DO think I am on to something. Proof in the pudding my committing capital to a dedicated account that is only trading off the machine.

   Or that it has correctly called market turns this year.

   May 16th, 2008 at 6:51:53 pm
3. Mac Says:
   

   Thanks for the info - I did not realize you also use Blocks. That explains why your scans would be a little different. I haven’t looked at it too much but from what I’ve seen on a few other sites, it looks like quite a powerful tool.

   I am still just using Telechart - agreed that it allows you to set up some very nice scans. I have just three main long scans in Telechart that I use, but I don’t really think I’ve missed a big-time stock this year in any of those scans. I like using some of their custom indicators on the charts as well. Unfortunately, I haven’t bought nearly enough of the ones my scans have produced. I also use his Market Monitor scans in Telechart - they are helpful as a secondary indicator - they aren’t perfect, but nothing really is in the stock market.

   I will follow your work on the “time machine” - it does seem interesting and very helpful. Thanks again,

   May 16th, 2008 at 8:27:20 pm
4. Danny Says:
   

   no problem, and thank you for the comments. I hope you make good use with it.

   I love TC for everything, but I use blocks because, in a word, it is infuckingsane. It is the most powerful stock program I’ve ever seen or heard of.

   May 16th, 2008 at 9:49:49 pm

Leave a Reply