Doug, Andy, Derek
Re: Scans in QP. I ran some tests based on BBB and DNS to determine effect of various conditions and indicators. This may be “old hat” but it gave me a better idea of how they behave. I started by running a basic scan for only common stocks to get a count and then scanned that list as each condition was added. Of course, the final scan can be done in one pass. Data was from 11/8/96.
Number
selected Scan Conditions
7,847 Common stocks
256 BBB using 6 mos for vol & added close > 30
197 Added 5xMA > 13xMA
156 Added 13xMA > 40xMA
146 Added Dahl's, 50MA > 50MA 15 days ago
137 Added OBV > 40xMA
93 Added Avg Vol for last 6 mos > 20,000
92 Added MACD 8,17,9 above signal line
91 Added ROC > 0
56 Added QRS > 80 (Quotes Plus Relative Strength)
125 BBB as line 2 above plus only QRS > 80
Of the 92 stocks remaining after MACD scan, 20 were micro cap (<100MM). Another test showed that requiring the 13xMA to fall between 5 and 40xMA rather than just 5 > 40xMA reduced selected stocks by 20%.
Unfortunately we can not code parabolic SAR in QP and there is no earnings data yet. However, I think this is a viable alternative to starting with the IBD list, scanning with BBB then DNS.
One caution when looking at charts in QP that contain exponential moving averages, they are affected by the number of days displayed. QP is apparently using the displayed data not all of the data in the data base to calculate xMA's. If a short time frame such as 60 days is used to display a 40xMA, it will move relative to prices and short xMA's when the chart is expanded to more days. This does not happen in TC2000. I talked to Gary @ QP and he is going to review their calculation. Not a major issue but it is confusing when testing scans.
Bob
ps. The DOC COP is great!! |
