The purpose of this post is to keep track of my progress on optimizing stock screeners I’ve been using for many years.
Basically, for long candidates I search very strong stocks and try to buy them when a bullish reversal bar is formed on a way of a bigger pullback; For short candidates I search very strong downtrend stocks and try to short them when a bearish reversal bar is formed on a way of a bigger bounce. I have screener to search breakout on volume surge for long and breakdown on volume increase for short too. But I found generally, if without manually pick up the screen results, buy dip/sell bounce setup is marginally better. The manual pickup requires experiences and besides, psychologically most people aren’t comfortable in chasing high or selling low. So personally I feel perhaps it’s better to focus on screeners for buying dips and selling bounces. The chart below illustrates the back test result. As I don’t have super super powerful machine (Intel Core 2 Quad CPU shouldn’t be very bad, should it? So what more can I ask?), so testing on all stocks takes more than an hour so I have to use Nasdaq 100 components for testing each optimizing conditions to save time. Not perfect, but that’s the reality.
Below is the current back test for buying dip/selling bounce for all the stocks trading on the market. The results are not as good as that of Nasdaq 100, but not very bad I think.
我没看搞懂,这个backscanner是能帮忙挑出breakout的个股??但是截图上胜率什么的是说这个软件的有效性???但是软件推荐的个股在哪里?最好能下载一个玩玩看 azure4068 发表于 2010-1-17 20:08  不是。Backscanner只是用来back test的选股的有效性。具体还是要俺自己写的Scan Rule来选股。Backscanner怎么可能帮你选股呢?问题是你写的scan rule,能不能经得起考验,需要一个back test软件来验证,俺正好有Backscanner可以验证,所以就用他啦。 |