glgene
616 posts
msg #72872
- Ignore glgene |
3/17/2009 8:45:58 PM
Tried it, and it produced a resulting number of 99.89. Can't be. Stock price of QQQQ is less than $30.
Look back at my note (2 msg. ago). According to Elder, you've got to use the 1-2 step process (for his example):
1) First, calculate an EMA(13) of price,
then,
2) Calculate an ROC(21) of the 13-day price EMA.
I think your SMROC(13,21) doesn't do the above. It doesn't address price at all. Just a smoothed ROC. I'm baffled that it came up with 99.89.
I went to Yahoo, and did a calculation of 13-day EMA price. It's $27.36. Price closed today at $29.33
I simply don't know how to SF script the 1-2 process above.
SOS.
Zack
|
chetron
2,817 posts
msg #72874
- Ignore chetron |
3/17/2009 9:19:17 PM
LOOKING UP THE FORMULA AT INVESTOPEDIA, IT LOOKS LIKE SF IS GOING IN THIS DIRECTIONISH
|
chetron
2,817 posts
msg #72875
- Ignore chetron
modified |
3/17/2009 9:27:46 PM
ALL TOGETHER NOW.......................
|
glgene
616 posts
msg #72876
- Ignore glgene |
3/17/2009 9:38:35 PM
OK, I'll put my "slow" magnifying glass to it.
Still need to put QQQQ in a Symlist
Symlist(QQQQ)
since that is the basis for the original project. Simply buy/sell QQQQ at buy/sell signals of TripleMomROC.
That would put QQQQ around 99 on your chart. Look at QQQQ's climb the last few days. It would have missed all the gain.
????
Zack
|
glgene
616 posts
msg #72877
- Ignore glgene |
3/17/2009 9:53:32 PM
I checked XLF with your scripting. It's at 99.81, short of a buy at 100. That means it would have just missed about a 40% run-up.
Put XLF in a symlist and you'll see what I mean.
Where did the number 100 come from?
Zack
|
chetron
2,817 posts
msg #72878
- Ignore chetron |
3/17/2009 10:01:09 PM
THE 100 WAS IN THE INVESTOPEDIA FORMULA. I DON'T KNOW WHY.
|
chetron
2,817 posts
msg #72879
- Ignore chetron
modified |
3/17/2009 10:03:14 PM
MAYBE XROC_OSC BOUNCING OFF THE ZERO LINE IS A BUY? THAT WOULD HAVE GOT XLF?????????????
100 A SELL????????????????
|
glgene
616 posts
msg #72880
- Ignore glgene |
3/17/2009 10:17:36 PM
Chetron,
Here is what I've done so far --- with MUCH your scripting help. I just added some fluff to it (displays, columns) for possible confirming signals (RSI, MACD-Histogram). Note that I added this to the script:
add column SMROC(13,21) ....... from your first msg. tonight. It's at the 6th line of scripting.
Look at the multitude of Symlist results: Practically all at 99+. Looks suspicious. XLF has had a superb run-up, and it would have been missed. ?????
Zack
|
Eman93
4,750 posts
msg #72884
- Ignore Eman93 |
3/18/2009 1:39:04 AM
This is pretty cool, it worked for BIG only one draw down day I started on the day it crossed 100 and just eyeballed the results.
Have you been using this? How is it working?
|
glgene
616 posts
msg #72887
- Ignore glgene |
3/18/2009 7:54:33 AM
This script is simply in the development stages. Haven't actually used it yet. Want to try and wrap up the smoothed ROC first (if that's doable). Chetron has been absolutely invaluable with this project!!
The book I got this from (written by Gerald Appel) says it (ROC crossover) has been backtested only with the NAZ index, but it says it should be suitable with other indices as well. It makes no comments about individual stocks, so that's just guesswork (I guess!).
I added RSI and MACD columns to be able to look at other indicator info at the same time as XROC. Also put Force Index into the graph (just learning about that) and ADX info.
I put BIG in the mix from a friend who bought that stock about a week ago. Thought he might be interested in seeing how it does with XROC.
Notice that I put 1-day ago info next to certain columns. Again, that's just to add more "look" at a stock's short-term performance. I figure, if you've got the columnar space, why not use it?
I'll keep working on it. Let me know if you (or anybody else reading this) have any suggestions for improvement.
Zack
|