corsino
259 posts
msg #48939
- Ignore corsino |
1/5/2007 12:00:40 AM
UNITED
Although this is getting away from the main topic, I'll make a couple of additional comments on points you have raised.
I'm a little careful about buying on the "ten day low".The reason is that it may be due to a "falling knife". I think that trying to catch a falling knife is better suited for day traders,which I'm not.True, money can be made the day after the sharp drop because typically, the stock recoups some of the previous day's drastic loss. However,for the swing trader there's the danger of whipsaws. Too many times the price rises, or seems to stabilize for a couple of days,then there's another sharp drop. In other words, falling knives are typically caused by severe bad news, which may take considerable time to heal.
On the other hand,a slow steady decrease for two or three weeks may mean that the stock is temporarily out of favor because of a downgrade or a disappointing earnings report.After a couple of weeks,the price has dropped to a level where traders begin to think that it's a good buy again. Of course, all this is purely personal unproven theories !
As for your wish for additional backtesting, your subscription probably allows you a 4-month backtest. However, you can adjust the backtesting testing period backwards 4 months at a time to incrementally go farther into the past if you wish
|
itrademan
70 posts
msg #48955
- Ignore itrademan |
1/5/2007 10:38:23 AM
corsino,
Good explanation to identify the two types of fall. It might not be true always but do hold some good reasoning.
|
corsino
259 posts
msg #48960
- Ignore corsino |
1/5/2007 1:51:20 PM
Actually,the filter I submitted is not the complete filter I use. I just submitted a short form of it.There are two additional lines. Although they do not affect the filtering,they aid in the decision making:
and draw slow stochastics(5,3)fast %K
and draw williams %R(9)
They are my two favorite indicators. I should add that I monitor stocks starting with 7 down days and watch those two indicators, in addition to support regions,for signs of reversals.Whether I buy any or not,obviously depends on available funds.
|
sggc
4 posts
msg #48964
- Ignore sggc |
1/5/2007 2:53:00 PM
What is the actual per cent of all stocks that have gone up in the last 4 months. the last 4 months have been extremely bullish.
|
itrademan
70 posts
msg #48967
- Ignore itrademan |
1/5/2007 4:27:08 PM
corsino,
I too agree that filter will not give correct signal.
Since there are not many stocks returned it would be easy to track them every day after 7th day and look for:
* Resistance Break/ Trend Reversal during first hour of the trading day
* Oversold Indicators
on early 1st hour charts.
|
corsino
259 posts
msg #48975
- Ignore corsino |
1/5/2007 8:36:24 PM
Well, I don't know how realistic the SF backtest is. I agree that the stock market has been bullish for the last 4 months but, supposedly, it would affect all other filters too. I only submitted the filter for performance comparison with other filters. Unfortunately, no other filter was submitted to date, although I'm sure there are better ones out there.
|
lockwhiz
206 posts
msg #48979
- Ignore lockwhiz |
1/5/2007 10:28:45 PM
corsino,
How did you configure the back testing ... when I run your filter I don't get anywhere near the W/L ratio you stated.
Thanks
|
corsino
259 posts
msg #48981
- Ignore corsino |
1/5/2007 11:07:54 PM
lockwhiz
I ran the backtest for the latest 4 month duration, with a max. holding period of 5 days.
For comparison, since no other filter has been submitted, I'm submitting a basic highly respected RSI(2)<1 filter plus the 11-down-day filter modified to match the RSI filter in price range.
Actually, this comparison may not be apples-to-apples, since I consider the RSI(2) filter a daytrading filter,whereas the 11-down-day filter is more of a swing trade filter. Also, I agree that it's not only the filter, but how you trade it.
|
corsino
259 posts
msg #48985
- Ignore corsino |
1/6/2007 1:19:24 AM
RumpledOne
That's strange...adding RSI(2)<1 doesn't change the W/L ratio of 70/28. Or the number of stocks.But maybe it might for other 4-month periods.
|
sadelb
12 posts
msg #48986
- Ignore sadelb |
1/6/2007 1:20:36 AM
I like adding RSI < 1 to the filter but I changed the price parameters to close between 15 and 50. Anything under 15 wasnt doing to well.
|