MARY4MONEY
806 posts
msg #78151
- Ignore MARY4MONEY |
8/28/2009 10:31:25 AM
weekly macd 2,3,1----could some one desigen a filter for when this becomes pos and neg------thanks
|
dwiggains
441 posts
msg #78155
- Ignore dwiggains |
8/28/2009 10:55:57 AM
Hi
Please remember using weekly data --- ONLY --- uses Fridays data.
Every time I use a weekly indicator I get into trouble.
See ya
David
|
chetron
2,817 posts
msg #78164
- Ignore chetron
modified |
8/28/2009 2:11:59 PM
maybe..........................................
|
chetron
2,817 posts
msg #78165
- Ignore chetron |
8/28/2009 2:27:18 PM
or even....
|
Kevin_in_GA
4,599 posts
msg #78193
- Ignore Kevin_in_GA |
8/28/2009 6:49:31 PM
Or this - sorted on largest divergence.
Kevin
|
trendscanner
265 posts
msg #78199
- Ignore trendscanner |
8/28/2009 7:45:23 PM
Building on Kevin's filter, adding an additional line, and playing with the time frame a bit.
This one dredged up fewer hits than Kevin's but includes some real stinkers. Lots of ProShares short funds meeting these criteria. Many of these also show a weekly divergence even with conventional MACD parameters. Maybe a correction really is not far off.
|
Kevin_in_GA
4,599 posts
msg #78202
- Ignore Kevin_in_GA |
8/28/2009 8:09:13 PM
Check that last line. Should be "weekly" MACD Fast line ...
also, wouldn't you want this week's fast line to be above last week's, if the slope is above 0?
|
Kevin_in_GA
4,599 posts
msg #78203
- Ignore Kevin_in_GA
modified |
8/28/2009 8:10:27 PM
I went with a shorter timeframe because the sensitivty of the slope was poor when looking at recent moves.
It would be simpler if there were a ROC function for indicators.
|
trendscanner
265 posts
msg #78205
- Ignore trendscanner |
8/28/2009 8:38:34 PM
Kevin, you're right, it should have been the weekly MACD. Here it is revised. For whatever reason, it still captures most of the same suspects but got 54 hits instead of 37, so something a little different is going on.
My rationale for weekly fast line being below last week's fast line was to reduce the number of hits and to limit it to those where MACD was not yet increasing (i.e., not yet past the point of maximum rebound). The 6 week slope can still be positive with this condition since it started at a much more negative place 6 weeks ago than it is today. By the time the weekly MACD fast line is above last week's, price has already increased from its nadir.
Here's the 4 week version for comparison
This version has more hits and a very different group. Not sure what the cosmic implications of that are. Probably just reflects the character of the last 6 week's overall market movement.
|
Kevin_in_GA
4,599 posts
msg #78233
- Ignore Kevin_in_GA |
8/29/2009 4:25:35 PM
I keep thinking that we are looking at divergence wrong here - the way I usually see it presented is that, during two recent maxima/minima in MACD a line is drawn from the max of the previous signal to the current max. What we are doing is just defining a recent slope, rather than the slope between the two maxima.
Not sure how to code this.
|